Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
STATE OF THE MARKET 2018
published
May 15, 2019
Southwest Power Pool, Inc. Market Monitoring Unit
State of the Market 2018
Acknowledgement
The following members of the Market Monitoring Unit
contributed to this report:
Keith Collins Jodi Woods John Luallen
Greg Sorenson Kevin Bates
Jason Bulloch Jared Greenwalt
Esat Serhat Guney David Hurtado James Lemley
Mark Rouse Cristie Turner
Will Vestal Cuiping Xu
Report lead: Greg Warren
Southwest Power Pool, Inc. Market Monitoring Unit
State of the Market 2018
Southwest Power Pool, Inc. Table of contents Market Monitoring Unit
State of the Market 2018 i
TABLE OF CONTENTS
Table of contents .................................................................................................................................. i
List of figures ........................................................................................................................................ v
1 Executive summary .................................................................................................................1 1.1 Market highlights .................................................................................................................. 1 1.2 Overview ................................................................................................................................ 3 1.3 Day-ahead and real-time market performance ................................................................. 4 1.4 Transmission congestion and hedging .............................................................................. 6 1.5 Uplift costs ............................................................................................................................. 6 1.6 Competitiveness assessment .............................................................................................. 7 1.7 Structural issues .................................................................................................................... 9 1.8 Recommendations ............................................................................................................ 10
1.8.1 New recommendations for 2018 ............................................................................. 10 1.8.2 Clarification of previous recommendations ............................................................ 13 1.8.3 Outstanding recommendations ............................................................................... 14
2 Load and resources .............................................................................................................. 17 2.1 The Integrated Marketplace ............................................................................................. 18
2.1.1 SPP market footprint .................................................................................................. 18 2.1.2 SPP market participants............................................................................................. 19
2.2 Electricity demand ............................................................................................................. 21 2.2.1 System peak demand ................................................................................................ 21 2.2.2 Market participant load ............................................................................................. 22 2.2.3 SPP system energy consumption ............................................................................. 24 2.2.4 Heating and cooling degree days ........................................................................... 25
2.3 Installed generation capacity ........................................................................................... 28 2.3.1 Capacity additions and retirements ......................................................................... 31 2.3.2 Generation interconnection...................................................................................... 32
2.4 Generation.......................................................................................................................... 34 2.4.1 Generation by technology ........................................................................................ 34 2.4.2 Generation on the margin ........................................................................................ 38
2.5 Growing impact of wind generation capacity ................................................................ 40 2.5.1 Wind capacity and generation ................................................................................. 40 2.5.2 Wind impact on the system ...................................................................................... 44 2.5.3 Wind integration ........................................................................................................ 47 2.5.4 Challenges with non-dispatchable variable energy resources ............................ 50
2.6 External transactions ......................................................................................................... 51 2.6.1 Exports and imports .................................................................................................. 51 2.6.2 Market-to-market coordination ................................................................................ 56
Southwest Power Pool, Inc. Table of contents Market Monitoring Unit
State of the Market 2018 ii
3 Unit commitment and dispatch processes ....................................................................... 61 3.1 Commitment process ........................................................................................................ 62
3.1.1 Resource starts ........................................................................................................... 63 3.1.2 Demand for reliability ................................................................................................ 66 3.1.3 Quick-start resources commitment .......................................................................... 69 3.1.4 Generation scheduling .............................................................................................. 74
3.2 Dispatch .............................................................................................................................. 77 3.2.1 Scarcity pricing ........................................................................................................... 77 3.2.2 Ramping ...................................................................................................................... 83 3.2.3 Ramp capability product ........................................................................................... 88
3.3 Must-offer provision .......................................................................................................... 94 3.4 Virtual trading .................................................................................................................... 95
4 Prices.................................................................................................................................... 105 4.1 Market prices and costs .................................................................................................. 106
4.1.1 Energy market prices and fuel prices .................................................................... 106 4.1.2 Hub prices ................................................................................................................. 111 4.1.3 Energy price volatility .............................................................................................. 113 4.1.4 Day-ahead and real-time price convergence ....................................................... 115 4.1.5 Negative prices ........................................................................................................ 122 4.1.6 Operating reserve market prices ........................................................................... 126 4.1.7 Market settlement results ........................................................................................ 132
4.2 Make-whole payments .................................................................................................... 136 4.2.1 Make-whole payment allocation ............................................................................ 141 4.2.2 Regulation mileage make-whole payments ......................................................... 142 4.2.3 Potential for manipulation of make-whole payments .......................................... 144 4.2.4 Jointly-owned unit make-whole payments ........................................................... 145
4.3 Total wholesale market costs and production costs ................................................... 147 4.4 Long-run price signals for investment ........................................................................... 149
5 Congestion and transmission congestion rights market .............................................. 153 5.1 Transmission congestion ................................................................................................ 154
5.1.1 Pricing patterns and congestion ............................................................................ 155 5.1.2 Congestion by geographic location ...................................................................... 156 5.1.3 Transmission expansion .......................................................................................... 158 5.1.4 Transmission constraints ......................................................................................... 160 5.1.5 Planned transmission projects ................................................................................ 164 5.1.6 Geography and marginal losses ............................................................................ 164 5.1.7 Frequently constrained areas and local market power ....................................... 166 5.1.8 Market congestion management ........................................................................... 167 5.1.9 Congestion payments and uplifts .......................................................................... 168 5.1.10 Distribution of marginal loss revenues (over-collected losses) .......................... 171
Southwest Power Pool, Inc. Table of contents Market Monitoring Unit
State of the Market 2018 iii
5.2 Congestion hedging market .......................................................................................... 173 5.2.1 Market design ........................................................................................................... 173 5.2.2 Market transparency ................................................................................................ 175 5.2.3 Funding ..................................................................................................................... 181 5.2.4 Intra-auction sales and bulletin board transactions ............................................. 186 5.2.5 Issues, progress, and next steps ............................................................................ 188
6 Planning process ................................................................................................................ 191 6.1 Resource adequacy ......................................................................................................... 193
6.1.1 Capacity additions and retirements ....................................................................... 193 6.1.2 Generation capacity compared to peak load ....................................................... 194
6.2 Integrated transmission plan .......................................................................................... 196 6.2.1 2020 ITP generator retirement assumptions ........................................................ 197 6.2.2 2020 ITP renewable capacity accreditation assumptions ................................... 199 6.2.3 Other topics .............................................................................................................. 202
6.3 Assessment ....................................................................................................................... 203 6.4 Recommendations going forward ................................................................................. 204
7 Competitive assessment ................................................................................................... 207 7.1 Structural aspects of the market .................................................................................... 209 7.2 Behavioral aspects of the market ................................................................................... 216
7.2.1 Offer price markup .................................................................................................. 216 7.2.2 Mitigation performance and frequency ................................................................ 219 7.2.3 Output gap (measure for potential economic withholding) ............................... 223 7.2.4 Unoffered generation capacity (measure for potential physical withholding) . 227
7.3 Offer behavior due to mitigation threshold ................................................................. 231 7.4 Start-up and no-load behavior ....................................................................................... 231 7.5 Generator retirements .................................................................................................... 232 7.6 Competitive assessment summary ................................................................................ 234
8 Recommendations ............................................................................................................. 237 8.1 New recommendations .................................................................................................. 237
8.1.1 Limit the exercise of market power by creating a backstop for parameter changes .................................................................................................................................... 237 8.1.2 Enhance credit rules to account for known information in assessments ........... 238 8.1.3 Develop compensation mechanism to pay for capacity to cover uncertainties ...... .................................................................................................................................... 239 8.1.4 Enhance ability to assess a range of potential outcomes in transmission planning .................................................................................................................................... 240 8.1.5 Improve regulation mileage price formation ....................................................... 241
8.2 Clarification of previous recommendations ................................................................. 243 8.2.1 Address inefficiency caused by self-committed resources ................................. 243 8.2.2 Enhance commitment of resources to increase ramping flexibility ................... 245 8.2.3 Further enhance alignment of planning processes with operational conditions .... .................................................................................................................................... 246
Southwest Power Pool, Inc. Table of contents Market Monitoring Unit
State of the Market 2018 iv
8.3 Address previous recommendations ............................................................................ 248 8.3.1 Develop a ramping product ................................................................................... 248 8.3.2 Enhance market rules for energy storage resources ........................................... 249 8.3.3 Address inefficiency when forecasted resources under-schedule day-ahead . 250 8.3.4 Convert non-dispatchable variable energy resources to dispatchable ............ 251 8.3.5 Address gaming opportunity for multi-day minimum run time resources ....... 252 8.3.6 Replace day-ahead must offer, add physical withholding provision ................. 253
8.4 Recommendations update ............................................................................................. 253
Common acronyms ........................................................................................................................ 255
The data and analysis provided in this report are for informational purposes only and shall not be considered or relied upon as market advice or market settlement data. All analysis and opinions contained in this report are solely those of the SPP Market Monitoring Unit (MMU), the independent market monitor for Southwest Power Pool, Inc. (SPP). The MMU and SPP make no representations or warranties of any kind, express or implied, with respect to the accuracy or adequacy of the information contained herein. The MMU and SPP shall have no liability to recipients of this information or third parties for the consequences that may arise from errors or discrepancies in this information, for recipients’ or third parties’ reliance upon such information, or for any claim, loss, or damage of any kind or nature whatsoever arising out of or in connection with:
i. the deficiency or inadequacy of this information for any purpose, whether or not known or disclosed to the authors;
ii. any error or discrepancy in this information;
iii. the use of this information, and;
iv. any loss of business or other consequential loss or damage whether or not resulting from any of the foregoing.
Southwest Power Pool, Inc. Market Monitoring Unit
State of the Market 2018 v
LIST OF FIGURES
Figure 2—1 SPP market footprint and RTO/ISO operating regions ....................................... 19 Figure 2—2 Market participants by type .................................................................................... 20 Figure 2—3 Capacity by market participant type ...................................................................... 20 Figure 2—4 Monthly peak system demand ............................................................................... 21 Figure 2—5 Cleared demand bids in day-ahead market ......................................................... 22 Figure 2—6 System energy usage .............................................................................................. 23 Figure 2—7 System energy consumption, monthly .................................................................. 24 Figure 2—8 Load duration curve................................................................................................. 25 Figure 2—9 Heating and cooling degree days ......................................................................... 26 Figure 2—10 Degree days and loads compared with a normal year ....................................... 27 Figure 2—11 Generation nameplate capacity by technology type .......................................... 29 Figure 2—12 Aggregate supply curve, typical summer day ...................................................... 30 Figure 2—13 Capacity additions ................................................................................................... 31 Figure 2—14 Capacity retirements ............................................................................................... 32 Figure 2—15 Active generation interconnection requests, megawatts ................................... 33 Figure 2—16 Executed generation interconnection requests, on-schedule ........................... 34 Figure 2—17 Generation by technology type, real-time, annual .............................................. 35 Figure 2—18 Generation by technology type, real-time, monthly ............................................ 36 Figure 2—19 Implied heat rate ...................................................................................................... 37 Figure 2—20 Generation on the margin, real-time, annual ....................................................... 38 Figure 2—21 Generation on the margin, real-time ..................................................................... 39 Figure 2—22 Generation on the margin, day-ahead .................................................................. 40 Figure 2—23 Wind speed map...................................................................................................... 41 Figure 2—24 Wind capacity and generation, annual ................................................................. 42 Figure 2—25 Wind capacity factor ................................................................................................ 43 Figure 2—26 Wind capacity factor by month, real-time, three year ......................................... 44 Figure 2—27 Wind generation as a percent of load ................................................................... 45 Figure 2—28 Wind production curve ........................................................................................... 46 Figure 2—29 Net demand by hour of day.................................................................................... 47 Figure 2—30 Dispatchable and non-dispatchable wind generation ........................................ 48 Figure 2—31 Manual dispatches for wind resources .................................................................. 49 Figure 2—32 Exports and imports, SPP system ........................................................................... 52 Figure 2—33 Exports and imports, ERCOT interface .................................................................. 53 Figure 2—34 Exports and imports, Southwestern Power Administration interface ................ 53 Figure 2—35 Exports and imports, MISO interface..................................................................... 54 Figure 2—36 Exports and imports, Associated Electric Cooperative interface ....................... 54 Figure 2—37 Imports and export transactions by type, day-ahead .......................................... 56 Figure 2—38 Market-to-market settlements ................................................................................ 57 Figure 2—39 Market-to-market settlements by flowgate ........................................................... 58 Figure 3—1 Commitment process timeline ............................................................................... 62 Figure 3—2 Start-up instructions by resource count ................................................................ 63 Figure 3—3 Start-up instructions by resource capacity ............................................................ 64
Southwest Power Pool, Inc. List of figures Market Monitoring Unit
State of the Market 2018 vi
Figure 3—4 Origin of start-up instructions for gas resources .................................................. 65 Figure 3—5 Average hourly capacity increase from day-ahead to real-time ........................ 67 Figure 3—6 Instantaneous load capacity required ................................................................... 68 Figure 3—7 Instantaneous load capacity upper bound default levels ................................... 69 Figure 3—8 Deployment of quick-start resources .................................................................... 70 Figure 3—9 Operation of quick-start resources ........................................................................ 72 Figure 3—10 Day-ahead market offers by commitment status and participant type ............. 74 Figure 3—11 Day-ahead market offers by commitment status ................................................. 74 Figure 3—12 On-line capacity as a percent of load .................................................................... 75 Figure 3—13 Generation outages ................................................................................................. 76 Figure 3—14 Short-term generation outages, fall months ......................................................... 77 Figure 3—15 Scarcity intervals and marginal energy cost, after scarcity implementation ..... 80 Figure 3—16 Scarcity events by interval of the hour ................................................................... 81 Figure 3—17 Spinning reserve shortages .................................................................................... 83 Figure 3—18 Frequency of net load change, real-time .............................................................. 84 Figure 3—19 Volatility of interval-to-interval net load change .................................................. 85 Figure 3—20 Average up-rampable capacity by month ............................................................ 86 Figure 3—21 Average down-rampable capacity by month ....................................................... 87 Figure 3—22 Average up-rampable capacity by hour ............................................................... 87 Figure 3—23 Average down-rampable capacity by hour .......................................................... 88 Figure 3—24 Interval length of short-term price spikes ............................................................. 89 Figure 3—25 Interval length of short-term price spikes, percentage ....................................... 90 Figure 3—26 Cleared virtual transactions as percent of real-time load ................................... 96 Figure 3—27 Cleared virtual offers by settlement location type ............................................... 97 Figure 3—28 Cleared virtual bids by settlement location type ................................................. 98 Figure 3—29 Virtual offers and bids, day-ahead market ............................................................ 99 Figure 3—30 Virtual profits with distribution charges .............................................................. 100 Figure 3—31 Profit and loss per cleared virtual, average ........................................................ 101 Figure 3—32 Virtual portfolios by market participant ............................................................... 102 Figure 4—1 Energy price versus natural gas price, annual .................................................... 106 Figure 4—2 Day-ahead energy price versus natural gas cost ............................................... 107 Figure 4—3 Fuel price indices and wholesale power prices ................................................. 108 Figure 4—4 Fuel-adjusted marginal energy cost .................................................................... 110 Figure 4—5 Hub prices, annual average .................................................................................. 111 Figure 4—6 Hub prices, day-ahead, monthly .......................................................................... 112 Figure 4—7 Hub prices, real-time, monthly ............................................................................. 112 Figure 4—8 Comparison of RTO/ISO average on-peak, day-ahead prices ......................... 113 Figure 4—9 System price volatility ............................................................................................ 114 Figure 4—10 Price volatility by market participant.................................................................... 115 Figure 4—11 Day-ahead and real-time prices ........................................................................... 116 Figure 4—12 Difference between day-ahead and real-time prices ........................................ 116 Figure 4—13 Price divergence .................................................................................................... 118 Figure 4—14 On-peak marginal energy prices, excluding scarcity hours ............................. 119 Figure 4—15 Off-peak marginal energy prices, excluding scarcity hours ............................. 120
Southwest Power Pool, Inc. List of figures Market Monitoring Unit
State of the Market 2018 vii
Figure 4—16 Average hourly real-time generation incremental to day-ahead market ........ 121 Figure 4—17 Negative price intervals, day-ahead, monthly .................................................... 122 Figure 4—18 Negative price intervals, real-time, monthly ....................................................... 123 Figure 4—19 Negative price intervals, day-ahead, by hour ..................................................... 124 Figure 4—20 Negative price intervals, real time, by hour ........................................................ 124 Figure 4—21 Negative price intervals, real-time, by asset owner ........................................... 125 Figure 4—22 Operating reserve product prices, real-time ...................................................... 126 Figure 4—23 Regulation-up service prices ................................................................................ 127 Figure 4—24 Regulation-down service prices ........................................................................... 128 Figure 4—25 Spinning reserve prices ......................................................................................... 128 Figure 4—26 Supplemental reserve prices ................................................................................ 129 Figure 4—27 Regulation mileage prices, real-time ................................................................... 130 Figure 4—28 Regulation mileage payments and charges ....................................................... 132 Figure 4—29 Energy settlements, load ...................................................................................... 133 Figure 4—30 Energy settlements, generation ........................................................................... 134 Figure 4—31 Operating reserve product settlements .............................................................. 135 Figure 4—32 Make-whole payments by fuel type, day-ahead ................................................ 137 Figure 4—33 Make-whole payments by fuel type, reliability unit commitment .................... 137 Figure 4—34 Make-whole payments, commitment reasons .................................................... 138 Figure 4—35 Make-whole payments for voltage support ........................................................ 139 Figure 4—36 Concentration of make-whole payments by resource ...................................... 140 Figure 4—37 Market participants receiving make-whole payments ....................................... 141 Figure 4—38 Make-whole payments by market uplift allocation, real-time ........................... 142 Figure 4—39 Regulation mileage make-whole payments ....................................................... 143 Figure 4—40 All-in price of electricity and natural gas cost ..................................................... 147 Figure 4—41 Production cost, daily average, day-ahead......................................................... 148 Figure 4—42 Production cost, daily average, real-time ........................................................... 149 Figure 4—43 Net revenue analysis assumptions ....................................................................... 150 Figure 4—44 Net revenue analysis results ................................................................................. 151 Figure 4—45 Net revenue analysis by zone ............................................................................... 151 Figure 5—1 Price map, day-ahead and real-time market ...................................................... 155 Figure 5—2 Marginal congestion cost map, day-ahead market ........................................... 157 Figure 5—3 SPP transmission expansion in service, 2018 ..................................................... 159 Figure 5—4 SPP transmission expansion plan ......................................................................... 160 Figure 5—5 Congestion by shadow price, top ten flowgates ............................................... 161 Figure 5—6 Central Kansas congestion ................................................................................... 162 Figure 5—7 Southwest Missouri congestion ........................................................................... 163 Figure 5—8 Top ten congested flowgates with projects ....................................................... 164 Figure 5—9 Marginal loss component map, day-ahead ........................................................ 165 Figure 5—10 Breached and binding intervals, day-ahead market .......................................... 167 Figure 5—11 Breached and binding intervals, real-time .......................................................... 168 Figure 5—12 Annual congestion payment by load-serving entity .......................................... 169 Figure 5—13 Total congestion payments .................................................................................. 170 Figure 5—14 Over-collected losses, real-time ........................................................................... 172
Southwest Power Pool, Inc. List of figures Market Monitoring Unit
State of the Market 2018 viii
Figure 5—15 Total congestion and hedges ............................................................................... 176 Figure 5—16 Net congestion revenue by market participant .................................................. 177 Figure 5—17 Transmission outages by reporting lead time .................................................... 179 Figure 5—18 Transmission outages by duration ....................................................................... 180 Figure 5—19 Transmission outages by market ......................................................................... 180 Figure 5—20 Transmission congestion right funding levels, annual ...................................... 182 Figure 5—21 Transmission congestion right funding levels, monthly .................................... 183 Figure 5—22 Transmission congestion right funding, daily .................................................... 184 Figure 5—23 Auction revenue right funding levels, annual ..................................................... 185 Figure 5—24 Auction revenue right funding levels, monthly .................................................. 186 Figure 5—25 Intra-auction sales and bulletin board transactions ........................................... 187 Figure 6—1 Capacity by age of resource ................................................................................. 193 Figure 6—2 Capacity additions and retirements by year ....................................................... 194 Figure 6—3 Peak available capacity percent ........................................................................... 195 Figure 6—4 Average age of retirements for coal and gas resources ................................... 198 Figure 6—5 Average age of generators in the SPP market ................................................... 199 Figure 6—6 MMU calculation for wind and solar capacity factors ........................................ 201 Figure 7—1 Market share of largest supplier .......................................................................... 210 Figure 7—2 Market concentration level, real-time .................................................................. 211 Figure 7—3 Herfindahl-Hirschman Index ................................................................................. 212 Figure 7—4 Herfindahl-Hirschman Index, annual ................................................................... 213 Figure 7—5 Market participants with a marginal resource, real-time ................................... 214 Figure 7—6 Hours with at least one pivotal supplier .............................................................. 216 Figure 7—7 Average offer price markup of marginal resource, monthly ............................ 217 Figure 7—8 Average on-peak offer price markup by fuel type of resource, monthly ........ 218 Figure 7—9 Average offer price markup, annual .................................................................... 219 Figure 7—10 Mitigation frequency, day-ahead market ............................................................ 221 Figure 7—11 Mitigation frequency, real-time market ............................................................... 222 Figure 7—12 Mitigation frequency, start-up offers ................................................................... 222 Figure 7—13 Output gap, monthly ............................................................................................. 225 Figure 7—14 Output gap, SPP footprint .................................................................................... 226 Figure 7—15 Output gap, Lubbock frequently constrained area ........................................... 226 Figure 7—16 Unoffered economic capacity .............................................................................. 228 Figure 7—17 Unoffered economic capacity at various load levels, SPP footprint ................ 229 Figure 7—18 Unoffered economic capacity at various load levels, Lubbock area ............... 230 Figure 7—19 Mitigation start-up offer mark-up by fuel category ............................................ 232 Figure 8—1 Annual State of the Market recommendations update ..................................... 253
Southwest Power Pool, Inc. Executive summary Market Monitoring Unit
State of the Market 2018 1
1 EXECUTIVE SUMMARY
The Southwest Power Pool (SPP) Market Monitoring Unit’s (MMU) Annual State of the Market
report for 2018 presents an overview of market design and market outcomes, assesses
market performance, and provides recommendations for improvement. The purpose of this
report is to provide SPP market stakeholders with reliable and useful analysis and information
to use in making market-related decisions. The MMU emphasizes that economics and
reliability are inseparable and that an efficient wholesale electricity market provides the
greatest benefit to the end user both presently and in the years to come.
1.1 MARKET HIGHLIGHTS
The following list identifies key observations in the SPP marketplace over the past year.
• SPP market results were workably competitive, with infrequent mitigation of offers and
high resource participation levels.
• Total wholesale market costs—including energy, operating reserve, and uplift
payments—averaged around $28/MWh in 2018, which was about 15 percent higher
than in 2017.
• Day-ahead and real-time prices both averaged around $25/MWh for the year, up from
$23/MWh in 2017.
• While the annual peak load of 49,926 MW was two percent lower this year compared
to last year, total electricity consumption was up about six percent.
• The incidence of negative prices in 2018, 3.5 percent of intervals, was about half of
the 2017 level of seven percent.
• While natural gas-fired resources frequently set prices in the SPP market, other factors
such as load increases, had a large impact on market prices. Natural gas prices
decreased about two percent in 2018 compared to 2017.
• When system prices are controlled for changes in fuel prices, they averaged about 13
percent higher in 2018 compared to 2017.
Southwest Power Pool, Inc. Executive summary Market Monitoring Unit
State of the Market 2018 2
• In 2018, there were 167 intervals with operating-reserve scarcity. This was roughly
five times as many intervals of operating reserve scarcity as was seen in 2017.
• Day-ahead and real-time congestion costs totaled over $450 million in 2018, a four
percent decrease from 2017.
• Wind generation peaked at 16.3 GW and peak wind penetration was almost 64
percent of load in December. Wind capacity increased to almost 20.6 GW in 2017, up
about 17 percent from 2017.
• Wind generation totaled 24 percent of all generation in 2018, up slightly from 23
percent in 2017. Coal generation fell from 46 percent in 2017 to 42 percent in 2018.
• New capacity additions were just over 2,300 MW at nameplate capacity, with wind
representing 97 percent of the new capacity. Capacity retirements increased to just
under 2,000 MW in 2018, split about evenly between coal and gas resources.
• The generator interconnection process includes just over 84 GW of additional
resources, of which 99 percent are renewable or storage.
• SPP continues to have significant excess capacity at peak loads. The MMU estimates
that capacity at peak is 35 percent higher than the peak demand level in 2018.
• Market prices themselves do not signal new investment in generation. Furthermore,
MMU analysis shows that market revenues do not support going forward costs for coal
resources.
• Market uplifts remained low at about $73 million in 2018, which was up slightly from
2017 levels.
• Combined ancillary reserve costs totaled $76 million last year, a decrease of three
percent from 2017.
• Auction revenue rights were funded at 145 percent in 2018, down from just over 160
percent in 2017.
• Transmission congestion rights funding was virtually unchanged from 2017 to 2018, at
94 percent.
Southwest Power Pool, Inc. Executive summary Market Monitoring Unit
State of the Market 2018 3
1.2 OVERVIEW
Overall, the SPP market produced highly competitive market results with total market costs
around $28/MWh. As with previous years, the largest component of total wholesale costs
remains energy costs, which represented almost 98 percent of total costs in 2018. While total
costs increased by 15 percent in 2018 compared to 2017, a main driver for the increase in
energy costs was an increase in load, as well as several other factors. When adjusted for fuel
prices, average SPP marginal energy prices increased by 13 percent.
While the annual peak load of 49,926 MW was two percent lower this year compared to last
year, total electricity consumption was up about six percent. Over 97 percent of the 2,300
MW increase in nameplate generation capacity last year was from wind resources. This
continues a pattern that has occurred over the past several years. Wind generation as a
percent of total generation continued to increase as it represented 24 percent of system
generation, up from 23 percent in 2017 and 18 percent in 2016. Conversely, coal generation
continued to decline, representing around 42 percent of total generation last year, down
from 46 percent in 2017 and 49 percent in 2016. Prior to 2016, coal represented 55 percent
or more of generation in SPP.
Significant market events in 2018 included:
• A second circuit from Woodward to Mathewson was added.1 Our analysis shows that
this transmission addition helped shift power further east in the SPP footprint and
reduce the frequency of negative prices.
• Market participants continued to express concerns with the effectiveness of the
auction revenue rights process to allow participants to receive sufficient hedges for
congestion.2 We observe that while many participants were able to manage
congestion, a handful of participants did not have sufficient hedges. There were
multiple reasons for this, including the nature of congestion patterns, outages, and
market participant strategies.
1 See Section 5.1.3 for further discussion. 2 Auction revenue rights and transmission congestion rights are covered in Section 5.2.
Southwest Power Pool, Inc. Executive summary Market Monitoring Unit
State of the Market 2018 4
• SPP launched the Holistic Integrated Tariff Team effort. This group comprised
members, Board of Directors, Regional State Committee members, and supported by
SPP staff, focused on developing problem statements, reviewing analysis, and
proposing solutions to related areas of concern including transmission cost allocation,
auction revenue right allocation, market enhancements, and load integration. The
findings of the group are anticipated to be presented to the April 2019 Board
meeting.
1.3 DAY-AHEAD AND REAL-TIME MARKET PERFORMANCE
Overall, energy prices were about $2/MWh higher in 2018 compared to 2017. This can
primarily be attributed to several factors, including increased loads, increased exports,
transmission expansion, lower wind capacity factors, and increased generator outages.
These offset other factors that would typically lower prices, including lower gas prices and a
large and increasing reserve margin. Historically, day-ahead prices have been higher than
real-time prices. However, in 2017 real-time prices were higher than day-ahead prices in nine
months primarily because of higher real-time price volatility. This trend reversed in 2018 as
eight out of twelve months had higher day-ahead average prices.
While load participation in the day-ahead market continued to be strong in 2018, generation
participation, particularly from wind resources was substantially less in the day-market market
compared to the real-time market. For instance, the average level of participation for the
load assets was between 99 percent and 101 percent of the actual real-time load. However,
we found that on average for the year, wind generation was over 1,000 MW higher in the
real-time market compared to the amount scheduled in the day-ahead market on an hourly
basis. This represents an increasing challenge to the market as wind generation has
continued to increase substantially over the past few years.
Virtual bids and offers may theoretically offset the under-scheduling of renewable supply in
the day-ahead market, however, in net they did not as they averaged around 500 MW of net
virtual supply. Furthermore, it is important to recognize that even if virtual transactions were
to match the quantity of under-scheduled renewables, the prices associated with the virtual
offers are not likely to fully represent the offer prices of the renewable resources in order to
preserve a profit margin.
Southwest Power Pool, Inc. Executive summary Market Monitoring Unit
State of the Market 2018 5
In general, virtual transactions were profitable in the SPP market. However, total profits
decreased in 2018 to about $44 million, down from $54 million in 2017. When transaction
fees are included, net profit for virtual transactions was $18 million in 2018. Net virtual profits
were highest in January ($4.9 million) and October ($4.0 million) when wind generation is
typically high and loads are low.
Self-commitment of generation continues to be a concern because it does not allow the
market software to determine the most economic market solution. Furthermore, it can
contribute to market uplifts and low prices. Some of the reasons for self-committing may
include contract terms for coal plants, low gas prices that reduce the opportunity for coal
units to be economically cleared in the day-ahead market, long startup times, and a risk-
averse business practice approach. Generation offers in the day-ahead market averaged
almost 53 percent as “market” commitment status followed by “self-commit” status at 30
percent of the total capacity commitments for 2018.3 These levels almost exactly match
those in 2017, however the overall trend is still downward, as 2016 had 48 percent as
“market” commitment status, and 35 percent as “self-commit” status. While the overall
increase in market commitments and decrease in self-commitments highlights an
improvement, self-commitments still represent over 30 percent of generation, a trend that
has existed since the Integrated Marketplace began in 2014. In order to improve market
commitment in the SPP market, we recommend that SPP and stakeholders look to find ways
to address this issue.
Scarcity events in 2018 were generally in line with 2017 levels, with the exception of October.
October 2018 experienced significant wind volatility, which among other factors, led to an
uptick in operating reserve scarcity events. Additionally, looking at the intervals where
scarcity occurs each hour shows that 30 percent of all scarcity events, and two-thirds of
regulation-down scarcity events, in 2018 occurred in the first interval of the hour. One
potential reason for this trend is that SPP does not preposition regulating resources to be at
their regulating effective maximum and minimum limits prior to the period that the resource
is cleared for regulation. Unlike some other RTO/ISO markets, the current SPP model does
not account for forecasted ramping needs. Our analysis shows that accounting for ramping
needs would greatly assist in preparing and compensating generation for both anticipated
3 Other resource commitment statuses are “reliability”, “not participating”, and “outage” at two percent, three percent, and 11 percent, respectively.
Southwest Power Pool, Inc. Executive summary Market Monitoring Unit
State of the Market 2018 6
and unanticipated ramping needs. As such, we continue to recommend, as discussed below,
that SPP and stakeholders complete design for a ramping product in 2019.
1.4 TRANSMISSION CONGESTION AND HEDGING
Locational marginal prices reflect the sum of the marginal cost of energy, the marginal cost of
congestion, and the marginal cost of losses for each pricing interval at any given pricing
location in the market. Although the SPP market currently maintains a high reserve margin,
certain locations of the footprint experience significant price movements resulting from
congestion caused by high wind generation and transmission limitations.
Areas that experienced the highest congestion costs in 2018 were in central Kansas,
southeastern Oklahoma, and southwest Missouri. Western Nebraska experienced the lowest
congestion costs for the year. The frequently constrained area study for 2018 saw the
addition of the central Kansas and southwest Missouri frequently constrained areas, while the
Texas panhandle frequently constrained area was removed.
In total, congestion costs were just over $450 million in 2018. This was down slightly from
$465 million in 2017. While most load-serving entities were able to successfully hedge their
congestion exposure with auction revenue rights and transmission congestion rights, a
handful of participants were under-hedged. In 2018, the total of all transmission congestion
right and auction revenue right net payments to load-serving entities of $458 million was
more than the total day-ahead and real-time markets congestion costs of $381 million.
However, on an individual basis, some participants were over-hedged, whereas others were
under-hedged. The largest amount over-hedged was by just over $16 million, while the
largest amount under-hedged was just under $10 million.
1.5 UPLIFT COSTS
Generators receive make-whole payments to ensure that they receive sufficient revenue to
cover energy, start-up, no-load, and operating reserve costs for both market and local
reliability commitments. Make-whole payments are additional market payments in cases
where prices result in revenue that is below a resource’s cleared offers. These payments are
intended to make resources whole to energy, commitment, and operating reserve costs.
In 2018, total make-whole payments were approximately $73 million, up slightly from $68
million in 2017. Make-whole payments averaged about $0.26/MWh in 2018, which was
Southwest Power Pool, Inc. Executive summary Market Monitoring Unit
State of the Market 2018 7
nearly the same as in 2017. In comparison to other RTO/ISO markets, SPP’s make-whole
payments were in the range of uplift costs, which varied from $0.08/MWh to $0.29/MWh in
2017 and 2018.4
Day-ahead make-whole payments constituted about 38 percent of the total make-whole
payments in 2018. SPP pays about 80 percent of all make-whole payments to gas-fired
resources.
In December 2017, FERC found SPP’s quick-start pricing practice may be unjust and
unreasonable, and argued that pricing enhancements would reduce uplift payments.5 Our
analysis shows that uplift payments remain fairly low. Furthermore, our assessment of FERC’s
proposal suggests that, if anything, FERC’s proposal would transfer, if not increase, uplifts.6
FERC has yet to rule on SPP’s quick-start pricing practices.
1.6 COMPETITIVENESS ASSESSMENT
The SPP market provides effective incentives and mitigation measures to produce
competitive market outcomes even during periods when the potential for the exercise of
local market power could be a concern. The MMU’s competitive assessment using structural
and behavioral metrics indicate that market results in 2018 were workably competitive and
that the market required mitigation of local market power infrequently to achieve competitive
outcomes.
Structural competitiveness metrics—which review the structural potential for the exercise of
market power—indicate minimal to moderate potential structural market power in SPP
markets outside of areas that are frequently congested. With the merger of Great Plains
Energy and Westar, creating Evergy, Inc., the percent market share of the largest supplier
saw an increase, with 35 percent of hours for the year exceeding the 20 percent threshold
4 2017 ISO NE State of Market Report https://www.potomaceconomics.com/wp-content/uploads/2018/06/ISO-NE-2017-EMM-SOM-Report_6-17-2018_Final.pdf, page 39; 2018 State of the Market Report for PJM http://www.monitoringanalytics.com/reports/Presentations/2019/IMM_MC_SOM_20190321.pdf, page 9. 5 Order Instituting Section 206 Proceeding and Commencing Hearing Procedures and Establishing Refund Effective Date, 161 FERC ¶ 61,296 (2017). 6 Southwest Power Pool Market Monitoring Unit Reply Brief, Docket No. EL18-35-000, March 13, 2018.
Southwest Power Pool, Inc. Executive summary Market Monitoring Unit
State of the Market 2018 8
that indicates potential market power, with a high value of 29 percent. This is up from 2017,
when no hours were above the 20 percent threshold, and the highest value was 17 percent.
An additional measure of structural market power is the Herfindahl-Hirschman Index (HHI).
This analysis, based on actual generation, indicates that 12 percent of hours in 2018 had
values between 1,000 and 1,800, which indicates a moderate level of concentration. The
market had been considered unconcentrated since the addition of the Integrated System in
October 15, up until the creation of Evergy in June 2018. Prior to the addition of the
Integrated System, approximately 45 percent of all hours were considered moderately
concentrated.
While moderately concentrated hours increased following the creation of Evergy, an increase
in market share and HHI in themselves does not pose a threat to the structural
competitiveness of the SPP market. Other relevant market data including pivotal supplier
hours and local market power mitigation must also be evaluated for competitive assessment.
For the two frequently constrained areas, where potential for concerns of local market power
is the highest, existing mitigation measures serve well to prevent pivotal suppliers from
unilaterally raising prices.
Behavioral indicators—which assess the actual exercise of market power—show low levels of
mitigation frequency. Mitigation of day-ahead energy, operating reserve, and no-load offers
each occurred less than 0.2 percent of the time and real-time mitigation occurred about
0.015 percent of the time. The overall mitigation frequency of start-up offers was the lowest
since the market began in 2014, as it decreased in 2018 relative to 2016 levels to just under
three percent.
The decline in mitigation may be the result of declining offer price mark-ups. Both off-peak
and on-peak average offer markups were at the lowest levels since implementation of the
Integrated Marketplace at around −$2.05/MWh and −$2.41/MWh, respectively. Although a
lower offer price markup level in itself would indicate a competitive pressure on suppliers in
the SPP market, the observed continuous downward trend may raise questions about the
commercial viability of generating units and the possibility of generation retirements.
The monthly average output gap—which measures economic withholding—shows very low
levels of economic withholding in all months in 2018. Specifically, there was miniscule
Southwest Power Pool, Inc. Executive summary Market Monitoring Unit
State of the Market 2018 9
amounts of measurable output withheld in the frequently constrained areas. These low levels
of economic output withholding reflect highly competitive participation in the market.
Another method of competitive assessment is unoffered generation capacity for potential
physical withholding. Specifically, any economic generation capacity that is not made
available to the market through derates, outages, or otherwise not offered to the market is
considered for this analysis.
Annually for the SPP footprint, the total unoffered capacity (as a percent of total resource
reference levels) equaled 2.0 percent in 2016, 1.9 percent in 2017, and 3.1 percent in 2018.
When short and long-term outages are removed, the remaining unoffered capacity was 0.22
percent, 0.23 percent, and 0.39 percent, respectively. The majority of the outages were long-
term outages due to maintenance during the shoulder fall and spring months. From a
competitive market perspective, the results indicate reasonable levels of total unoffered
economic capacity and are consistent with the results in other RTO/ISO markets.
1.7 STRUCTURAL ISSUES
Installed generation capacity in the SPP market has grown rapidly over the past several years.
This has contributed to high levels of capacity at peak loads. Specifically, the MMU estimates
that capacity was 35 percent higher than the peak load in 2018. SPP’s current annual
planning capacity requirement is 12 percent.
Wind capacity has more than doubled from 8.6 GW in 2014 to 20.6 GW in 2018. At the same
time, wind generation has constituted a growing and significant part of the total annual
generation, from around 12 percent in 2014 to 24 percent in 2018. Furthermore, the
interconnection process includes just over 84 GW of additional resources, of which 99
percent are renewable resources.
The shift in generation mix towards renewable resources is a significant development and
carries both market and operational challenges. Furthermore, these challenges are further
exacerbated by the fact that currently 29 percent of the total wind capacity is non-
dispatchable.
A recent wave of generator retirements, particularly of coal-fired generation, has been widely
observed throughout the country. The SPP market should be expected to follow this trend
Southwest Power Pool, Inc. Executive summary Market Monitoring Unit
State of the Market 2018 10
because of excess capacity, aging fleet, and cost disadvantages of certain types of
generation technologies vis-à-vis the prevailing market prices.
The MMU believes that SPP stakeholders should prepare for the challenges these changes,
and potential changes, to the market present. However, additional changes from planning to
operations needs to be developed to improve market outcomes. As such, we make several
recommendations to address these growing market concerns.
1.8 RECOMMENDATIONS
One of the primary responsibilities of a market monitoring unit is to evaluate market rules and
market design features for market efficiency and effectiveness. When we identify issues with
the market, one of the ways to correct them is to make recommendations on market
enhancements. These recommendations are highlighted in detail in Chapter 8. Below is a
summary of our 2018 recommendations.
1.8.1 NEW RECOMMENDATIONS FOR 2018
• Limit the exercise of market power by creating a backstop for parameter changes –
The market is currently vulnerable to the exercise of market power by the
manipulation of resources’ non-dollar based parameters. Non-dollar-based
parameters should not be manipulated to affect market clearing. Although SPP’s tariff
and protocols have well-defined expectations and precise limitations for the basis of
dollar-based offer components in the presence of local market power, the
expectations for the basis of non-dollar-based offer components are much less clearly
defined and the limitations are much less precise.
The MMU recommends that SPP strengthen the language regarding non-dollar-based
parameters so that the expectation for the basis of these values is clear and the
exercise of market power is limited.
• Enhance credit rules to account for known information in assessments – In 2018,
GreenHat Energy, a financial-only market participant in PJM, defaulted on its portfolio
of congestion hedging products in the PJM markets. Current estimates of the default
Southwest Power Pool, Inc. Executive summary Market Monitoring Unit
State of the Market 2018 11
exposure exceed $400 million.7 While the SPP market is different from the PJM
market, SPP’s credit policy is similar to PJM’s in some respects. For instance, SPP’s
financial security requirements for transmission congestion rights are based only on
historic congestion patterns even when significant transmission upgrades are planned
or have occurred.
The MMU recommends that changes to the SPP credit policy be developed to protect
stakeholders from risks of default, particularly with respect to information that is
known such as transmission expansion.
• Develop compensation mechanism or product to pay for capacity to cover
uncertainties – Because of unexpected variations in wind output, SPP operations often
needs to manually commit resources (often in excess of 50 units) in order to meet
instantaneous load capacity requirements. Often, however, these resources receive
real-time wake whole payments and are not compensated specifically for the need.
While the MMU recognizes that SPP operators may need to commit units to account
for unforeseen circumstances, manual capacity commitments occur often enough that
systematic solutions should be developed. The large number of manual capacity
commitments may indicate that there is a need for a new uncertainty product that
might better address problems in the 30-minute to 3-hour time frame.8
The MMU recommends and supports developing products to reduce the need for the
large number of manual capacity commitments to ensure appropriate compensation
is being provided for the reliability services provided.
• Enhance ability to assess a range of potential outcomes in transmission planning –
SPP’s transmission planning process develops an annual look-ahead plan. This plan
evaluates transmission needs over a 5- and 10-year time horizon. The plan typically
7 https://pjm.com/-/media/committees-groups/committees/mic/20190206/20190206-item-01a-informational-update-ferc-order-denying-waiver-request.ashx. 8 The exact time periods of potential uncertainty products should be determined through the stakeholder process.
Southwest Power Pool, Inc. Executive summary Market Monitoring Unit
State of the Market 2018 12
evaluates two scenarios.9 The first case is a base case, and the second case is an
emerging technology case. The process could consider a third scenario. In 2018,
stakeholders requested that SPP staff consider an additional case. The third case
would have considered a shift in environmental regulations—such as a carbon tax or
adder—and a change in technologies/market trends including accelerated
deployment of storage devices or electric vehicles, higher levels of generation
retirement, and higher penetration of renewables.10
We recommend that SPP enhance their study process to allow the ability to study a
range of potential outcomes. If such range of potential outcomes are not captured in
a third case study, the MMU recommends that they be factored into either an existing
case study, as part of the 20-year assessment, or other assessment.
• Improve regulation mileage price formation – In addition to regulation capacity
payments, resources that are deployed for regulation also receive payments for costs
incurred when moving from one set point instruction to another. These payments,
known as mileage, are paid directly through regulation-up and regulation down
payments in the day-ahead market. The MMU is concerned that the mileage clearing
price does not correctly reflect price formation of mileage given that this price does
not represent the expected or realized mileage deployment. Furthermore, the MMU
is concerned that participants with resources frequently deployed for regulation will
have an incentive to inflate the mileage prices by offering in $0 regulation offers and
high mileage offers. The MMU also identified systematic overpayment of regulation
mileage in the day-ahead market, which appears to be the result of the mileage factor
being set consistently too high relative to actual mileage deployed.
We recommend that SPP staff review the performance of regulation mileage, and
develop potential approaches to improve regulation mileage price formation.
9 These scenarios are also known as futures cases. 10 This scenario was similar to a scenario used in the MISO transmission plan, see https://cdn.misoenergy.org/MTEP19%20Futures%20Summary291183.pdf.
Southwest Power Pool, Inc. Executive summary Market Monitoring Unit
State of the Market 2018 13
1.8.2 CLARIFICATION OF PREVIOUS RECOMMENDATIONS
The MMU has provided recommendations to improve market design in our previous annual
reports. The following are clarifications made to those recommendations from the prior
report.
• Address inefficiency caused by self-committed resources – The MMU recommends
that SPP and its stakeholders continue to explore and develop ways to reduce the
incidence of self-commitment of resources outside of the market solution. With
regards to the development of multi-day forecasting of prices or schedules or a multi-
day unit commitment process, the MMU supports these as attempts to reduce self-
commitment of generation, but we will carefully weigh the benefits and costs before
we support any specific proposal especially in light of the accuracy of forecasted
information and the potential changing generation mix. We continue, however, to
view reducing self-commitment of generation as a high priority for SPP and its
stakeholders as this will enhance market efficiency and improve price signals.
• Enhance commitment of resources to increase ramping flexibility – As noted in last
year’s annual report, ramping flexibility is becoming increasingly important to
integrate higher levels of renewable generation.11 Over-commitment of resources in
real-time reduces flexibility, suppresses prices, and leads to increased make-whole
payments. This can be caused by changing conditions between the time a resource is
locked into a commitment by the market software and the time the resource actually
comes on-line. The MMU recommends that SPP and its stakeholders address this
issue by enhancing its markets rules to enhance the commitment of resources to
increase resource flexibility. Last year, the MMU described this as enhancing
decommitment of resources. However, having explored this issue further, the issue is
not just about decommitment of resources, but is also about improving how and when
resources are committed.
• Further enhance alignment of planning processes with operation conditions –
Enhancing the accuracy of planning processes with operational realities enables SPP
and its members to more effectively plan for future system needs and
11 2017 Market Monitoring Unit Annual State of the Market Report, https://www.spp.org/documents/57928/spp_mmu_asom_2017.pdf , page 194.
Southwest Power Pool, Inc. Executive summary Market Monitoring Unit
State of the Market 2018 14
conditions. Many of the challenges outlined in this report—including congestion,
negative prices, and low generator net revenues—as well as improvements—such as
transmission additions—are, in part, a reflection of planning decisions. The more the
planning process can learn from and incorporate operational information, the more
planning can identify and address concerns in advance of market operations. SPP and
stakeholders over the past few years have worked to improve and align the planning
and operational processes.
Based on our experience in this process over the last year, we recommend that SPP
staff continue to collaborate with the MMU to build upon the work in 2018, by
reviewing data and best practices in other markets, factoring in market conditions and
trends in decision-making process, and performing analysis to support and inform
assumptions. Doing so will help to improve assumptions and ultimately study results.
1.8.3 OUTSTANDING RECOMMENDATIONS
The MMU has provided recommendations to improve market design in our previous annual
reports. Overall, SPP and its stakeholders have found ways to effectively address many of our
concerns. However, there are a number of recommendations that remain outstanding. A
description of each of these outstanding recommendations are outlined below.
• Develop a ramping product – A ramping product that incents actual, deliverable
flexibility can send appropriate price signals to value resource flexibility. This
resource flexibility can help prepare the system for fluctuations in both demand and
supply that result in transient short-term positive and negative price spikes. This is a
high priority initiative in the SPP stakeholder process. We concur and continue to
recommend that a ramping product design be completed in 2019.
• Enhance market rules for energy storage resources – With the increase in wind
penetration in the SPP market, there is not only a need for resource flexibility, but also
for storage. Stored energy resources have the potential to address both the need for
flexibility and reduce the incidence of negative prices. The MMU views integration of
storage resources in the SPP markets as an ongoing high priority as several
outstanding items beyond compliance with FERC Order No. 841 need to be
addressed in order to fully integrate electric storage resources in the SPP markets.
Southwest Power Pool, Inc. Executive summary Market Monitoring Unit
State of the Market 2018 15
• Address inefficiency when forecasted resources under-schedule day-ahead –
Systematic under-scheduling of wind resources in the day-ahead market can
contribute to distorting market price signals, suppressing real-time prices, and
affecting revenue adequacy for all resources. Therefore, the MMU continues to
recommend that SPP and its stakeholders address this issue through market rules
changes that address the consequences of under-scheduling of forecasted supply of
resources in the day-ahead market. We consider this a high priority as it helps to
enhance market efficiency and improve price signals.
• Convert non-dispatchable variable energy resources to dispatchable – Non-
dispatchable variable energy resources exacerbate congestion, reduce prices for
other resources, increase the magnitude of negative prices, cause the need for
market-to-market payments, and force manual commitments of resources that can
increase uplifts. Going forward, resource flexibility is essential to integrate an
increasing volume of wind generation in the SPP market. While the MMU’s preferred
solution did not pass the SPP stakeholder process, an acceptable amended solution
passed the board and was submitted to FERC in late 2018. FERC approved the filing
in April 2019.
• Address gaming opportunity for multi-day minimum run time resources – Resources
with minimum run times greater than two days have the opportunity to game the
market. The current implementation of the market rules limit make-whole payments
to the as-committed market offers for the first two days of a resource’s minimum run
time. However, after the second day, no rule exists to limit make-whole payments for
a resource that increases its offers from the third day onward until the resource’s
minimum run time is satisfied. For resources with minimum run times greater than two
days, the market participant knows that the resource is required to run and can
increase their market offers after the second day to increase make-whole payments.
The SPP board passed a proposal12 at the July 2018 meeting. However, subsequent
to board approval of the proposal, SPP legal staff identified internally inconsistent
tariff language that was revealed by but not addressed in the revisions. The MMU
strongly recommends that SPP and its stakeholders clean up the tariff language
12 Revision request 306 (2014 ASOM MWP MMU Recommendation [3-Day Minimum Run Time])
Southwest Power Pool, Inc. Executive summary Market Monitoring Unit
State of the Market 2018 16
necessary to implement the changes to address the gaming issue and provide an
anticipated implementation date for the system changes. Addressing this matter in a
timely manner is a high priority for the MMU.
• Replace day-ahead must-offer, add physical withholding provision – FERC rejected
SPP’s proposal to remove the day-ahead must offer requirement and indicated that it
would consider removal of the requirement if it were paired with additional physical
withholding provisions.
While the MMU remains concerned with the current day-ahead must offer
requirement, we recommend that further consideration of this issue be a low priority.
The MMU will continue to track market performance concerns related to this provision
and will consider raising the priority on this matter if further issues are identified or
current issues are exacerbated. Otherwise, we regard other matters as having higher
impacts to the market and priority for development at this time.
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 17
2 LOAD AND RESOURCES
This chapter reviews load and resources in the SPP market for 2018. Key points from this
chapter include:
• Total system energy consumption was up six percent from 2017 to 2018.
• Variations in demand continue to trend with seasonal temperature changes and
departures from normal temperatures
• Nearly 2,300 MW of wind generation was added to the market in 2018. Wind
generation now accounts for 23 percent of installed nameplate capacity in the SPP
market.
• The generation interconnection queue continues to grow, with over 83,000 MW of
projects in the queue. Only 312 MW is from fossil fuel generation, with the remainder
renewable or storage resources.
• Of total energy produced in 2018, 42 percent was from coal resources, while wind
resources produced nearly 24 percent of energy produced. For comparison, in 2014
coal generation represented 60 percent of the total, and wind generation accounted
for 12 percent of the total.
• SPP remained a net exporter for 2018 increasing from an hourly average of just over
350 MW in 2017 to just over 500 MW in 2018.
• Exports to MISO and AECI decreased from 2017, but tight supply in ERCOT over the
summer drove exports higher than previous years.
• September experienced the highest amount of exports when SPP’s real-time price of
$21/MWh was lower than both ERCOT and MISO prices.
• Dispatchable transactions in the day-ahead increased, but still only contributed to 10
percent of total volume for the year.
• Market-to-market payments from MISO to SPP totaled almost $25 million for 2018
with almost half occurring in the first two months of the year.
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 18
2.1 THE INTEGRATED MARKETPLACE
SPP is a Regional Transmission Organization (RTO) authorized by the Federal Energy
Regulatory Commission (FERC) to ensure reliable power supplies, adequate transmission
infrastructure, and competitive wholesale electricity prices. FERC granted RTO status to SPP
in 2004. SPP provides many services to its members, including reliability coordination, tariff
administration, regional scheduling, reserve sharing, transmission expansion planning,
wholesale electricity market operations, and training. This report focuses on the 2018
calendar year of the SPP wholesale electricity market referred to as the Integrated
Marketplace, which started on March 1, 2014.
The Integrated Marketplace is a full day-ahead market with transmission congestion rights,
virtual trading, a reliability unit commitment process, a real-time balancing market, and a
price-based operating reserves market. SPP simultaneously put into operation a single
balancing authority as part of the implementation of the Integrated Marketplace. The
primary benefit of the introduction of a day-ahead market was to improve the efficiency of
daily resource commitments. Another benefit of the new market includes the joint
optimization of the available capacity for energy and operating reserves.
2.1.1 SPP MARKET FOOTPRINT
The SPP market footprint is located in the westernmost portion of the Eastern
Interconnection, with Midcontinent ISO (MISO) to the east, Electric Reliability Council of Texas
(ERCOT) to the south, and Western Electricity Coordinating Council (WECC) to the west.
Figure 2—1 shows the current operating regions of the nine RTO/ISO markets in the United
States and Canada, as well as a more detailed view of the SPP footprint. The SPP market also
has connections with other non-RTO/ISO areas such as Saskatchewan Power Corporation,
Associated Electric Cooperative, and Southwestern Power Administration.13
13 Southwestern Power Administration belongs to the SPP RTO, Reliability Coordinator (RC), and Reserve Sharing Group (RSG) footprints. Associated Electric Cooperative belongs to the SPP RSG.
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 19
Figure 2—1 SPP market footprint and RTO/ISO operating regions
2.1.2 SPP MARKET PARTICIPANTS
At the end of 2018, 226 entities were participating in the SPP Integrated Marketplace. SPP
market participants can be divided into several categories: regulated investor-owned
utilities, electric cooperatives, municipal utilities, federal and state agencies, independent
power producers, and financial only market participants that do not own physical assets.
Figure 2—2 shows the distribution of the number of the 77 resource owners registered to
participate in the Integrated Marketplace.
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 20
Figure 2—2 Market participants by type
The number of independent power producers is high because most wind producers are
included in this category. Market participants referred to as an “agent” represent several
individual resource owners that would individually be classified as different types, such as
municipal utilities, electric cooperatives, and state agencies.
Figure 2—3 shows generation nameplate capacity owned by the type of market participant.
Investor-owned utilities and cooperatives own two-thirds of the nameplate generation
capacity in the SPP market.
Figure 2—3 Capacity by market participant type
Cooperatives7
Investor owned utilities7
Independent power producers
45
Marketer3
Municipal9
Industrial1
Agent1
State agency3
Federal agency1
77 resource owners totalas of December 31, 2018
Cooperatives11,503
Investor owned utilities47,630
Independent power producers
10,477 Marketer2,197
Municipal3,690
Industrial20
Agent4,957
State agency6,094
Federal agency2,598
89,166 total nameplate capacity (MW) as of December 31, 2018
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 21
Although investor-owned utilities represent only a small percent of the total market
participants at nine percent, they own the majority of the SPP generation capacity at 53
percent. This is in contrast to the “independent power producer” category, which has a large
number of participants (58 percent) representing only a small portion (12 percent) of total
nameplate capacity.
2.2 ELECTRICITY DEMAND
2.2.1 SYSTEM PEAK DEMAND
One way to evaluate load is to review peak system demand statistics over an extended
period of time. The market footprint has changed over time as participants have been added
to or withdrawn from the market. The peak demand values reviewed in this section are
coincident peaks, calculated out of total generation dispatch across the entire market
footprint that occurred during a specific market interval. The peak experienced during a
particular year or season is affected by events such as unusually hot or cold weather, daily
and seasonal load patterns, and economic growth and change.
Figure 2—4 shows a month-by-month comparison of peak-day demand for the last three
years. The monthly peak demand in eight months in 2018 was higher than the monthly peak
demand in both 2016 and 2017. Heating and cooling demand increases contributed to this
change (see Section 2.2.4).
Figure 2—4 Monthly peak system demand
20,000
30,000
40,000
50,000
60,000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Peak
load
(MW
)
2016 2017 2018
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 22
The SPP system coincident instantaneous peak demand in 2018 was 49,926 MW, which
occurred on July 12 at 3:35 PM. This is 2.5 percent lower than the 2017 system peak of
51,181 MW.
2.2.2 MARKET PARTICIPANT LOAD
Load continued to participate in the day-ahead market at high levels in 2018 as shown in
Figure 2—5.
Figure 2—5 Cleared demand bids in day-ahead market
The average monthly participation rates in the day-ahead market for load assets on an
aggregate level were between 99 and 101 percent of the actual real-time load. Accurate
reflection of demand in the day-ahead market economically incents generation to participate
in the day-ahead market. Additionally, accurate reflection of the load helps to converge
prices. Load participation in the day-ahead market increased by about one percent in 2018
from 2017.
Figure 2—6 depicts 2018 total energy consumption and the percentage of energy
consumption attributable to each entity in the market.
94%
96%
98%
100%
102%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Cle
ared
dem
and
as
a p
erce
nt o
f lo
ad
Off-peak On-peak
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 23
Figure 2—6 System energy usage 2016 2017 2018
Energy
consumed (GWh)
Percent of system
Energy consumed
(GWh)
Percent of system
Energy consumed
(GWh)
Percent of system
American Electric Power 42,746 17.2% 41,887 17.0% 43,109 16.6%
Oklahoma Gas and Electric 28,078 11.3% 27,747 11.3% 29,411 11.3%
Southwestern Public Service Company 25,658 10.3% 25,826 10.5% 27,359 10.5%
Westar Energy * 23,885 9.6% 23,845 9.7% 25,101 9.7%
Basin Electric Power Cooperative 17,859 7.2% 18,665 7.6% 20,165 7.8%
Kansas City Power and Light, Co * 15,528 6.3% 15,194 6.2% 16,049 6.2%
The Energy Authority, NPPD 13,248 5.3% 13,206 5.4% 12,932 5.0%
Omaha Public Power District 11,168 4.5% 11,066 4.5% 11,431 4.4%
Kansas City Power and Light, GMOC * 8,423 3.4% 8,275 3.4% 8,815 3.4%
Western Farmers Electric Cooperative 8,448 3.4% 8,046 3.3% 8,312 3.2%
Grand River Dam Authority 5,957 2.4% 5,581 2.3% 5,805 2.2%
Golden Spread Electric Cooperative Inc. 5,132 2.1% 4,817 2.0% 5,684 2.2%
Empire District Electric Co. 5,144 2.1% 4,984 2.0% 5,413 2.1%
Sunflower Electric Power Corporation 4,732 1.9% 4,693 1.9% 4,906 1.9%
Western Area Power Administration, Upper Great Plains 4,477 1.8% 4,534 1.8% 4,471 1.7%
Arkansas Electric Cooperative Corporation 3,708 1.5% 3,675 1.5% 4,258 1.6%
Lincoln Electric System Marketing 3,515 1.4% 3,441 1.4% 3,570 1.4%
The Energy Authority, CU 3,332 1.3% 3,227 1.3% 3,424 1.3%
Oklahoma Municipal Power Authority 2,857 1.1% 2,766 1.1% 2,846 1.1%
Kansas City Board of Public Utilities 2,427 1.0% 2,347 1.0% 2,530 1.0%
Midwest Energy Inc. 1,710 0.7% 1,715 0.7% 1,762 0.7%
Northwestern Energy 1,651 0.7% 1,632 0.7% 1,748 0.7%
Kansas Municipal Energy Agency 1,480 0.6% 1,473 0.6% 1,557 0.6%
Tenaska Power Service Company 1,363 0.5% 1,365 0.6% 1,418 0.5%
Missouri River Energy Services 1,260 0.5% 1,226 0.5% 1,309 0.5%
City of Independence 1,065 0.4% 1,030 0.4% 1,094 0.4%
Municipal Energy Agency of Nebraska 1,015 0.4% 1,022 0.4% 1,057 0.4%
East Texas Electric Cooperative 983 0.4%
Kansas Power Pool 860 0.3% 843 0.3% 866 0.3%
City of Chanute 482 0.2% 487 0.2% 497 0.2%
City of Fremont 441 0.2% 433 0.2% 450 0.2%
Missouri Joint Municipal Electrical Utility Commission 450 0.2% 430 0.2% 445 0.2%
Big Rivers Electric Corporation 310 0.1%
MidAmerican Energy Company 284 0.1% 280 0.1% 285 0.1%
South Sioux City, Nebraska 225 0.1% 261 0.1%
Harlan Municipal Utilities 19 0.0% 17 0.0% 18 0.0%
NSP Energy 4 0.0% 5 0.0% 5 0.0%
Otter Tail Power Company 41 0.0% 3 0.0% 1 0.0%
System Total 248,446 246,009 259,653
* Evergy, Inc. 49,965 19.2%
*Evergy was formed in June 2018 and is the corporate parent of Westar Energy, Kansas City Power and Light, and Kansas City Power and Light GMOC.
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 24
The largest five entities account for 56 percent of the total system energy usage. This
concentration is understandable as SPP’s market is primarily composed of vertically
integrated investor-owned utilities, which tend to be large. When Evergy is considered one
entity in the market, it represents the largest user of energy in the SPP market footprint at 19
percent of the total, and the four largest entities would then comprise 58 percent of energy
consumed in the market. Overall, the total system energy usage in 2018 was almost six
percent above the 2017 level. This was driven by increased loads in 2018 related to
increased heating and cooling demand.
2.2.3 SPP SYSTEM ENERGY CONSUMPTION
Figure 2—7 shows the monthly system energy consumption in thousands of gigawatt-hours.
Figure 2—7 System energy consumption, monthly
Energy consumed was higher in 2018 during the first six months of the year, and again in
November. August 2017 saw lower than normal temperatures across the footprint, as
evidenced by the lower level of energy consumption in that period. For the year, total system
energy consumption was up by over 6 percent in 2018 compared to 2017.
Figure 2—8 depicts load duration curves from 2016 to 2018. These load duration curves
display hourly loads from the highest to the lowest for each year.
10
14
18
22
26
30
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
GW
h (t
hous
and
s)
2016 '17 '18
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 25
Figure 2—8 Load duration curve
In 2018, the maximum hourly average load was 47,522 MW, down from both prior years. The
minimum hourly load for 2018 was 20,203 MW, which was slightly above both previous years.
Comparing annual load duration curves shows differentiation between cases of extreme
loading events and more general increases in system demand. If only the extremes are
higher or lower than the previous year, then short-term loading events are likely the reason.
However, if the load curve is higher than the previous year, as it was in 2018, it indicates that
total system demand has increased.
Reference percentage lines indicate higher load levels in 2018 at all but the highest
percentages, above five percent. As mentioned above, the 2018 load duration curve reflects
an increase in overall load compared to the previous two years.
2.2.4 HEATING AND COOLING DEGREE DAYS
Changes in weather patterns from year-to-year have a significant impact on electricity
demand. One way to evaluate this impact is to calculate heating degree days (HDD) and
cooling degree days (CDD). These values can then be used to estimate the impact of actual
weather conditions on energy consumption, compared to normal weather patterns.
15000
25000
35000
45000
55000
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000
Load
(MW
)
Hour2016 2017 2018
25%50%
75%
5%
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 26
To determine heating degree days and cooling degree days for the SPP footprint, several
representative locations14 were used to calculate system daily average temperatures.15 In this
report, the base temperature separating heating and cooling periods is 65 degrees
Fahrenheit. If the average temperature of a day at a location is 75 degrees Fahrenheit, there
would be 10 (=75–65) cooling degree days at that location. If a day’s average temperature is
50 degrees Fahrenheit, there would be 15 (=65–50) heating degree days at that location.
Using statistical tools, the daily estimated load impact of a single cooling degree day is much
higher than the impact of a single heating degree day. This is in part because of more
electric cooling than electric heating.
Figure 2—9 shows monthly heating and cooling degrees over the last three years compared
to the total monthly load.
Figure 2—9 Heating and cooling degree days
14 Amarillo TX, Topeka KS, Oklahoma City OK, Tulsa OK, and Lincoln NE. After October 1, 2015, Bismarck ND was added to represent SPP’s expanded market footprint. 15 Daily average temperature is calculated as the average of the daily lowest and highest temperatures. The source of temperature data is the National Oceanic and Atmospheric Administration (NOAA).
0
10,000
20,000
30,000
0
1,000
2,000
3,000
Jan
Feb
Mar
Ap
rM
ayJu
nJu
lA
ug Sep
Oct
No
vD
ec Jan
Feb
Mar
Ap
rM
ayJu
nJu
lA
ug Sep
Oct
No
vD
ec Jan
Feb
Mar
Ap
rM
ayJu
nJu
lA
ug Sep
Oct
No
vD
ec
2016 2017 2018
Load
(GW
h)
Deg
ree
day
s
Heating degree days Cooling degree days Monthly load
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 27
As shown in the chart, cooling degree days are more prevalent in the higher load months of
May through September, whereas heating degree days are more prevalent in the other
months. The SPP market footprint experienced much warmer temperatures in May, June,
and August 2018 compared to the previous year, which resulted in higher demand during
those months. July temperatures were milder than last year, explaining the slight drop in
market demand in July.
Figure 2—10 shows heating degree days, cooling degree days, and load levels from 2016
through 2018 compared to a normal year.16 Normal 2018 load was derived from a
regression analysis of actual footprint heating degree days, cooling degree days, weekends,
and holidays, substituting footprint normal temperatures.
The figures indicate loads are influenced by cooling demand in the late spring and summer
months, whereas late fall and winter loads are, to a lesser degree, influenced by heating
demand. Moreover, the figures show that cooling degrees in 2018 were higher than the
average but similar to 2016 in most months, whereas heating degree days were more similar
to average but higher than both 2016 and 2017 in most months.
Figure 2—10 Degree days and loads compared with a normal year
16 30 year normal temperatures are from the 1981-2010 U.S. Climate normals product from NOAA.
14,000
18,000
22,000
26,000
30,000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
GW
h
SPP system load
2016 2017 2018 normal
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 28
2.3 INSTALLED GENERATION CAPACITY
Figure 2—11 depicts the Integrated Marketplace installed generation for the SPP market
footprint at the end of the year. Total installed nameplate generation in the SPP Integrated
Marketplace was 89,167 MW by the end of 2018, representing an increase of over two
0
1,000
2,000
3,000
4,000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
deg
ree
day
scooling degree days
2016 2017 2018 normal
0
1,000
2,000
3,000
4,000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
deg
ree
day
s
heating degree days
2016 2017 2018 normal
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 29
percent from 2017.17 This increase was driven by a 17 percent increase in nameplate wind
capacity in 2018.
Figure 2—11 Generation nameplate capacity by technology type
Fuel type 2016 2017 2018 Percent as of year-end 2018
Coal 26,939 25,717 25,064 28%
Gas, simple-cycle 24,024 23,737 22,846 26%
Wind 16,114 17,596 20,589 23%
Gas, combined-cycle 12,870 12,618 13,248 15%
Hydro 3,428 3,422 3,431 4%
Nuclear 2,107 2,061 2,061 2%
Oil 1,684 1,639 1,639 2%
Solar 215 215 215 0%
Other 74 74 74 0%
Total 87,453 87,079 89,167 Note: Capacity is nameplate rating at year-end.
Natural gas-fired installed generation capacity still represents the largest share of generation
capacity in the SPP market at 42 percent (gas simple-cycle 26 percent, gas combined-cycle
15 percent), with coal being the second largest type at 28 percent. Wind continues to
increase due largely to new additions, with a 2018 market share of 23 percent of total
capacity in the SPP market. In terms of nameplate capacity, coal resources decreased from
17 The change in total generation capacity from year to year includes additions, retirements, and nameplate rating changes that occur during the year.
Coal28%
Gas, simple cycle26%
Wind23%
Gas, combined cycle15%
Hydro4%
Nuclear2%
Oil2%
Solar0% Other
0%
Year end 2018
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 30
30 percent of total nameplate capacity in 2017 to 28 percent in 2018. Wind nameplate
capacity more than offset the decline in nameplate coal capacity, increasing from 20 percent
of nameplate capacity in 2017 to 23 percent in 2018.
Figure 2—12 shows the total SPP aggregate real-time generation supply curves by offer price,
peak demand, and average demand for the summers of 2016 to 2018. Resources in
“outage” status were excluded from the supply curve. To calculate the summer supply
curves, a typical summer day was used for each analysis year. We calculated the aggregate
generation supply curves using the real-time offers of non-wind resources and wind forecast
data for wind resources.
Figure 2—12 Aggregate supply curve, typical summer day
Total aggregate real-time generation supply for summer 2018 was 65,295 MW, compared to
62,252 MW for summer 2017, an increase of nearly five percent. Just over 2,300 MW of
capacity was added to the SPP market in 2018, and most of the 1,900 MW of retirements
occurred after the summer season. The system peak-demand of 2018 was the lowest in the
last three years at about 2.5 percent lower than 2017. On the other hand, there was a six
percent increase of average demand of 2018 compared to 2017. Based on the heating and
cooling degree days analysis in Section 2.2.4, the SPP market footprint experienced much
warmer temperatures in May, June, and August 2018, which resulted in higher demand
during those months as compared to the previous year. July temperatures were milder than
last year, explaining the slight drop in market peak demand in 2018.
-$100
$0
$100
$200
$300
0 10 20 30 40 50 60 70
$/M
Wh
GW
2016 supply 2016 avg demand 2016 peak demand2017 supply 2017 avg demand 2017 peak demand2018 supply 2018 avg demand 2018 peak demand
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 31
Also evident is the approximately 24 GW gap between this maximum supply and the total
installed nameplate generation capacity on a typical summer day. This is primarily a result of
excluding the resources in “outage” status from the available supply (approximately 4 GW),
reduced summer capacity due to high ambient temperatures (approximately 6 GW), as well
as the difference between the wind forecast and installed capacity of wind resources
(approximately 13 GW).
The section of the offer curve below $0/MWh is mostly due to wind and solar energy and can
vary between 1,000 and 16,000 megawatts, based on wind and solar availability. Negative
offers typically reflect opportunity costs associated with state and federal tax incentives. The
sharp uptick in price at the top of the supply curves represents the transition from natural gas
units to oil units.
2.3.1 CAPACITY ADDITIONS AND RETIREMENTS
Figure 2—13 shows the capacity by the technology and the number of resources added in
2018.
Figure 2—13 Capacity additions
Just over 2,300 MW of generation capacity was added to the SPP market during 2018. Of the
new capacity added in 2018, 15 were wind resources accounting for 2,266 MW of capacity,
three were gas resources representing 66 MW of capacity, and one was a hydro resource
representing nine megawatts of capacity. All of the wind and gas capacity was new
construction.
Gas
Wind
Hydro
0 400 800 1,200 1,600 2,000 2,4000
4
8
12
16
20
Capacity (MW)
Res
our
ces
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 32
In 2018, the SPP market had generation retirements amounting to just over 1,900 MW of
installed capacity, shown in Figure 2—14.
Figure 2—14 Capacity retirements
Six medium-sized coal resources representing 954 MW of capacity, five gas resources
representing 924 MW of capacity, and one oil resource representing 54 MW of capacity were
retired in 2018.
A look at annual trends in additions and retirements can be found in Section 6.1.1.
2.3.2 GENERATION INTERCONNECTION
SPP is responsible for performing engineering studies to determine if the interconnection of
new generation within the SPP footprint is feasible, and to identify any transmission
development that would be necessary to facilitate the proposed generation. The generation
interconnection process involves a cluster study methodology allowing participants several
windows to submit requests for evaluation.18
Figure 2—15 shows the megawatts of capacity by generation technology type in all stages of
development. Included in this figure are interconnection agreements in the process of being
18 See Guidelines for Generator Interconnection Requests to SPP’s Transmission System http://sppoasis.spp.org/documents/swpp/transmission/studies/GuidelinesAndBusinessPracticesForGIP.pdf.
Coal
Gas
Oil
0 200 400 600 800 1000 12000
1
2
3
4
5
6
7
8
Capacity (MW)
Res
our
ces
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 33
created; those under construction; those already completed, but not yet in commercial
operation; and those in which work has been suspended as of year-end 2018.
Figure 2—15 Active generation interconnection requests, megawatts Prime mover 2016 2017 2018
Wind 32,690 28,147 55,118 Gas 1,080 3,122 312
Solar 3,770 15,306 24,583 Battery 40 1,135 4,305 Total 37,580 47,710 84,318
As can be seen in the table, generation capacity from renewable resources accounts for the
vast majority of proposed generation interconnection, with only 312 MW of non-renewable
generation interconnection requests in 2018. After a slight decrease of wind generation in
the queue from 2016 to 2017, the amount of wind generation of the queue nearly doubled
from 2017 to 2018. Interconnection requests for solar generation increased from 2017 to
2018, but not nearly as much as from 2016 to 2017. Battery interconnection requests also
increased from 2017 to 2018 with just over 4 GW in the queue at the end of 2018.
Development of renewable generation in the SPP region is expected to continue and the
proper integration of wind and solar generation is fundamental to maintaining the reliability
of the SPP system.
0
20
40
60
80
100
2016 2017 2018
Nam
epla
te c
apac
ity (G
W)
Wind Gas Solar Battery
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 34
Figure 2—16 Executed generation interconnection requests, on-schedule
As the chart above shows, at the end of 2018, just under 3 GW of generation is currently
slated to be added to the market in 2019. This is consistent with the levels of generation
additions in SPP in recent years. In the next three years, just over 100 MW of solar generation
is currently on-schedule to be developed. This would be the first solar generation added to
the market since 2016. It is important to note that generation can still be added or removed
from the list, even in the current year. However, there is more surety to the levels of
generation to go into production closer to the current year, and additions and deletions to
on-schedule projects are more typical in future years.
Development of renewable generation in the SPP region is expected to continue and the
proper integration of wind and solar generation is fundamental to maintaining the reliability
of the SPP system. Additional wind impact analysis follows in the Section 2.5.
2.4 GENERATION
2.4.1 GENERATION BY TECHNOLOGY
An analysis of generation by technology type used in the SPP Integrated Marketplace is
useful in understanding pricing, as well as the potential impact of environmental and
additional regulatory requirements on resources in the SPP system. Information on fuel types
and fleet characteristics is also useful in understanding market dynamics regarding
congestion management, price volatility, and overall market efficiency.
2019
2020
2021
Wind
Solar
0
2
4
6
Nam
epla
te c
apac
ity (G
W)
by year by fuel type
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 35
Figure 2—17 depicts annual generation percentages in the SPP real-time market by
technology type for the years 2007 through 2018.
Figure 2—17 Generation by technology type, real-time, annual
The long-term trend for coal-fired generation had been relatively flat through 2014 making
up around 60 to 65 percent of total generation, but has declined to under 50 percent
beginning in 2016. In 2018, coal generation accounted for 42 percent of total generation,
down from 46 percent in 2017. This decrease can primarily be attributed to increasing wind
generation and low natural gas prices.
The wind generation share continues to increase from three percent in 2007 to nearly 24
percent in 2018. Generation from simple-cycle gas units such as gas turbines and gas steam
turbines has seen a decline over the past few years, decreasing share from 13 percent in
2007 to just under five percent in 2017. However, with low gas prices during much of 2018,
the share increased to nearly eight percent. Gas combined-cycle generation has remained
relatively stable at about sixteen percent for the past three years, which can mostly be
attributed to low gas prices.
Some of the annual fluctuations in generation by technology type shares are driven by the
relative difference in primary fuel prices, namely natural gas versus coal. Gas prices in 2012,
and 2015 to 2018 were extremely low, resulting in some displacement of coal by efficient gas
generation, as can be seen in the higher generation from combined-cycle gas plants.
Another trend appears to be the increase in wind generation pushing simple-cycle gas
generation up the supply curve, making it less competitive.
0%
20%
40%
60%
80%
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Perc
ent
of t
ota
l gen
erat
ion
Coal Gas, simple cycle Gas, combined cycle Wind Hydro Nuclear
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 36
Retirement of older coal generation, environmental limits, along with competition from wind
and natural gas technologies are some of the factors that will continue to put pressure on
coal generation levels. Wind generation is expected to continue to increase in the years
ahead.
Figure 2—18 depicts the 2018 monthly fluctuation in generation by technology type.
Figure 2—18 Generation by technology type, real-time, monthly
Wind generation as a percentage of total generation is generally lowest in the summer
months, with peaks increasing above 30 percent in the highest wind generation months of
March and April. October is typically a high wind month, but wind generation was down in
October 2018, and was higher in November and December. The increase in wind
generation accompanied with low natural gas prices resulted in coal-fired generation market
share falling to below 40 percent from March through June 2018.
The monthly generation percentages show a general inverse relationship between monthly
levels of wind and coal generation in 2018. As wind generation increases, coal generation
decreases and vice versa. However, in November and December both coal and wind as a
percentage of total generation increased. This was primarily attributed to an increase in
natural gas prices and a subsequent decrease in natural gas-fired generation during the last
two months of 2018.
One method commonly used to assess price trends and relative efficiency in electricity
markets originating from non-fuel costs is the implied heat rate. The implied heat rate is
calculated by dividing the electricity price, net of a representative value for variable
0%
20%
40%
60%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Perc
ent
of t
ota
l gen
erat
ion
Coal Gas, simple cycle Gas, combined cycle Wind Hydro Nuclear
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 37
operations and maintenance (VOM) costs, by the fuel (gas) price.19 For a gas generator, the
implied heat rate serves as a “break-even” point for profitability such that a unit producing
output with an operating (actual) heat rate below the implied heat rate would be earning
profits, given market prices for electricity and gas. If the price of natural gas was $3/MMBtu,
and the electricity price was $24/MWh, the implied heat rate would be (24/3) = 8
MMBtu/MWh (8,000 Btu/kWh). This implied heat rate shows the relative efficiency required
of a generator to convert gas to electricity and cover the variable costs of production, given
market prices.
Figure 2—19 shows the monthly implied heat rate using real-time electricity prices for 2016 to
2018, along with an annual average for those years.
Figure 2—19 Implied heat rate
The chart shows a decline in implied heat rates in 2017 compared to 2016 and 2018. Of
note, wind generation and load affected electricity prices in 2017. For instance, high levels of
wind generation affected electricity prices in May and October, lowering average heat rates,
while high temperatures in July increased load and electricity prices. Implied heat rates in
December 2018 were low as a result of coal resources setting electricity prices more
19 For the implied heat rate calculation, natural gas units are assumed to be on the margin and accordingly, gas prices are taken as the relevant fuel cost. Emission costs are ignored in fuel cost as they rarely apply in the SPP market.
6,000
8,000
10,000
12,000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Imp
lied
hea
t rat
e (B
tu/K
Wh)
2016 2017 2018
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 38
frequently (see Section 2.4.2) as higher natural gas prices caused more gas resources to be
uneconomic.
2.4.2 GENERATION ON THE MARGIN
The system marginal price represents the price of the next increment of generation available
to meet the next increment of total system demand. The locational marginal price at a
particular pricing node is the system marginal energy price plus any marginal congestion
charges and marginal loss charges associated with that pricing node.
Figure 2—20 illustrates the frequency with which different technology types were marginal
and price setting. For a generator to set the marginal price, the resource must be: (a)
dispatchable by the market; (b) not at the resource economic minimum or maximum; and (c)
not ramp limited. In other words, it must be able to move to provide the next increment of
generation.
Figure 2—20 Generation on the margin, real-time, annual
This chart illustrates the dramatic shift in technology on the margin with natural gas
representing about 75 percent in the first year of an SPP market in 2007 to only about 52
percent in 2018. There is a corresponding shift in coal generation on the margin from about
25 percent in 2007, peaking at around 50 percent in 2015, then declining to 34 percent by
2018. With low natural gas prices, coal-fired plant owners are experiencing more daily
swings in dispatch level, which are reflected in this generation on the margin analysis.
0%
20%
40%
60%
80%
100%
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
% in
terv
als
on
the
mar
gin
Wind Coal Gas Gas, combined cycle Gas, simple cycle Other
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 39
It is worth noting the increase in wind generation being on the margin—from five percent in
2014 and 2015 to over 12 percent in 2018. With the growing amount of dispatchable wind
generation and an overall quantity of 23 percent of total nameplate capacity, wind
generation is increasingly becoming the marginal technology a higher percentage of the
time. At the end of 2017, 64 percent of wind capacity was dispatchable, compared to 60
percent at the end of 2016, 46 percent at the end of 2015, and 27 percent at the beginning
of the Integrated Marketplace in March 2014. The recommendation to convert non-
dispatchable variable energy resources to dispatchable variable energy resources is
discussed in Section 8.3.4.
Figure 2—21 shows monthly values for real-time generation on the margin for 2018.
Figure 2—21 Generation on the margin, real-time
Intervals with coal generation on the margin are typically lower in the spring and fall months,
resulting in more coal-fired units running as base load units with less cycling. The increased
wind generation is also affecting prices to some extent in every month of the year. The
higher wind generation on the margin values in the spring and fall are as expected given that
these periods are the windiest time of the year, as well as the lowest demand periods in the
SPP footprint.
Day-ahead generation on the margin, shown in Figure 2—22, is different from real-time in that
the day-ahead market includes virtual transactions. The real-time market does not include
0%
20%
40%
60%
80%
100%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec '16 '17 '18
% in
terv
als
on
the
mar
gin
Wind Coal Gas, combined cycle Gas, simple cycle Other
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 40
virtual transactions and is required to adjust to unforeseeable market conditions such as
unexpected plant and transmission outages.
Figure 2—22 Generation on the margin, day-ahead
Wind generation on the margin is comparable in the day-ahead and real-time markets with a
similar annual cyclical pattern. Both coal and gas generation on the margin in the day-ahead
market is noticeably lower than in the real-time market. The most significant difference is the
displacement of natural gas-fired generation by virtual offers in the day-ahead market. Virtual
energy offers on the margin have been declining over the past three years. However, virtual
energy offers represent 28 percent of the marginal offers in the day-ahead market in 2018,
compared to 25 percent in 2017 and 27 percent in 2016. While marginal virtual offers occur
at all types of settlement locations, 67 percent of marginal virtual offers are at resource
settlement locations, with a significant amount of activity at non-dispatchable wind
generation resource locations.
2.5 GROWING IMPACT OF WIND GENERATION CAPACITY
2.5.1 WIND CAPACITY AND GENERATION
The SPP region has a high potential for wind generation given wind patterns in many areas of
the footprint. Federal incentives and state renewable portfolio standards are additional
factors that have resulted in significant wind investment in the SPP footprint during the last
several years.
0%
20%
40%
60%
80%
100%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec '16 '17 '18
% in
terv
als
on
the
mar
gin
Wind Coal Gas, combined cycle Gas, simple cycle Other Virtual
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 41
Figure 2—23 is a wind speed map of the United States of America with the SPP footprint
outlined in black.
Figure 2—23 Wind speed map
Outside of coastal areas, much of the SPP footprint highlighted on the map is covered with
some of the highest wind speeds in the country. As has been discussed, there continues to
be a high potential for additional wind resource development in the SPP footprint going
forward.
Figure 2—24 depicts annual nameplate capacity and total generation of SPP wind facilities
since 2007.
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 42
Figure 2—24 Wind capacity and generation, annual
Total registered wind nameplate capacity at the end of 2018 was 20,589 MW, an increase of
17 percent from 2017. Wind generation output increased 10 percent in 2018 to nearly
65,000 GWh produced.
Wind resources comprise about 23 percent of the installed capacity in the SPP market,
behind only natural gas with 42 percent and coal with 28 percent. Consistent with previous
years, wind generation fluctuated seasonally as summer as usual was the low wind season,
while spring and fall were the high wind seasons. Also typical of wind patterns is lower
production during on-peak hours than off-peak. Furthermore, higher levels of wind
generation tend to coincide with the morning ramp periods.
Figure 2—25 shows the wind capacity factor. Note that the wind capacity factor is reported for
the entire month.20
20 Wind resources may be considered in-service, but not yet in commercial operation. In this situation, the capacity will be counted but the resource may not be providing any generation to the market.
0
20,000
40,000
60,000
80,000
0
6,000
12,000
18,000
24,000
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Gen
erat
ion
(GW
h)
Cap
acity
(MW
)
Capacity Generation
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 43
Figure 2—25 Wind capacity factor
The wind capacity factor in the real-time market dropped from 40.3 percent in 2017 to 39.5
percent in 2018, while the day-ahead wind capacity factor rose from 31.5 percent to 32.4
percent during the same period. Lower wind generation totals, coupled with increased wind
capacity, drove this drop in real-time capacity factor. The spread between the real-time and
the day-ahead wind capacity indicates a disconnect in the amount of wind in the real-time
market, compared to the forecasted wind in the day-ahead market.
Figure 2—26 shows the monthly real-time wind capacity factor for the past three years.
0%
10%
20%
30%
40%
50%
60%
Jan18
Feb18
Mar18
Apr18
May18
Jun18
Jul18
Aug18
Sep18
Oct18
Nov18
Dec18
2016 '17 '18
Win
d c
apac
ity fa
cto
r
Day-ahead capacity factor Real-time capacity factor
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 44
Figure 2—26 Wind capacity factor by month, real-time, three year
As shown above, the wind capacity factor for October and November 2018 was well below
the levels of the prior two years. These months, along with the spring, typically have the
highest wind generation of the year.
2.5.2 WIND IMPACT ON THE SYSTEM
Average annual wind generation as a percent of load continues to increase as shown in
Figure 2—27. The chart shows the trend for average and maximum wind generation as a
percent of load since 2007, illustrating the dramatic increase since the start of the SPP
markets.
0%
10%
20%
30%
40%
50%
60%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Win
d c
apac
ity fa
cto
r
2016 2017 2018
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 45
Figure 2—27 Wind generation as a percent of load
Average wind generation as a percent of load in the real-time market increased one percent
to nearly 25 percent in 2018. The growth of average wind generation as a percent of load
has leveled off after climbing much more sharply from 2015 to 2017. Wind generation
peaked at 16,329 MW in 2018 on a five-minute interval basis, an increase of over four percent
from 15,675 MW in 2017. Wind generation as a percent of load for any five-minute interval
reached a maximum value of 64 percent, which was up from 50 percent in 2016 and 57
percent in 2017.
Figure 2—28 shows wind production duration curves that represent wind generation as a
percent of load by real-time (five-minute) interval for 2016 through 2018.
0%
20%
40%
60%
80%
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Win
d g
ener
atio
n as
a p
erce
nt o
f lo
ad
average maximum
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 46
Figure 2—28 Wind production curve
The shift upward for the curve from year-to-year reflects an increase in total wind generation
on an annual basis. The wind production curves for 2017 and 2018 are nearly identical for
the lower two-thirds of the curve, while in the upper third of the curve, 2018 values climb
above 2017. Wind generation in both 2017 and 2018 served at least 23 percent of the total
load during half of the year, compared to 17 percent in 2016. It is also important to note that
the low end curve continues to gradually approach zero while the high end of the curve
increases very steeply.
Figure 2—29 below shows average hourly demand, along with wind generation, and net
demand (demand minus wind generation) for 2018.
0%
10%
20%
30%
40%
50%
60%
70%
0 10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000 90,000 100,000
Win
d g
ener
atio
n as
per
cent
of l
oad
Wind production intervals
2016 2017 2018
10% 50%
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 47
Figure 2—29 Net demand by hour of day
With wind generation at the highest levels in the overnight hours, net demand climbs more
steeply than total demand, as wind generation begins to taper off when approaching the
peak hours of the day. This can have an impact on the market as generation needed from
traditional resources climbs more quickly than demand. When this occurs, ramp scarcity is
more likely. This is discussed in Section 3.2.1
2.5.3 WIND INTEGRATION
Wind integration brings low cost generation to the SPP region but does not count for much
accredited capacity. There are a number of operational challenges in dealing with
substantial wind capacity. For instance, wind energy output varies by season and time of day.
This variability is estimated to be about three times more than load when measured on an
hour-to-hour basis. Moreover, wind is counter-cyclical to load. As load increases (both
seasonally and daily), wind production typically declines. The increasing magnitude of wind
capacity additions since 2007, along with the concentration, volatility, and timeliness of wind,
can create challenges for grid operators with regard to managing transmission congestion
and resolution of ramping constraints (which began being reflected in scarcity pricing in May
2017) as well as challenges for short- and long-run reliability. Several price spikes occurred
because of wind forecast errors. Wind forecast errors are also the leading cause of day-
ahead and real-time price divergence.
In the SPP market, wind and other qualifying resources were allowed to register as non-
dispatchable variable energy resources, provided the resource had an interconnection
0
10
20
30
40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
GW
h
Hour ending
Demand Net demand Wind generation
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 48
agreement executed by May 21, 2011 and was commercially operated prior to October 15,
2012. Because 29 percent (5,933 MW) of the existing installed wind capacity is composed of
non-dispatchable variable energy resources, and these generally produce without regard to
price, SPP operators must still issue manual dispatch instructions to reduce or limit their
output at certain times.
Figure 2—30 illustrates dispatchable variable energy resources (DVERs) and non-dispatchable
variable energy resources (NDVERs) wind output since 2016, with dispatchable variable
energy resource output mirroring the increasing percentage of installed wind capacity.
March 2018 saw over 6,300 GWh of monthly wind production, which was the highest since
the start of the Integrated Marketplace and 30 percent of this output originated from non-
dispatchable variable energy resource capacity.
Figure 2—30 Dispatchable and non-dispatchable wind generation
Figure 2—29 also shows the amount of reduced output of dispatchable variable energy
resources below their forecast (grey line). This depicts the increase in reductions of
dispatchable variable energy resource dispatch output from 2016 to 2017 but fell back to
2016 levels in 2018. This change was likely the result of increased transmission that
alleviated congestion bottlenecks and higher loads. Reductions in dispatchable resource
generation also follow the seasonal pattern of lower wind output during the summer months,
resulting in the decrease in need to reduce dispatchable variable energy resource output
during these times. This increase in dispatchable wind capacity has helped in the
management of congestion caused by high levels of wind generation in some of the western
parts of the SPP footprint.
0
100
200
300
400
0
2,000
4,000
6,000
8,000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Red
uced
dis
pat
ch (G
Wh)
Gen
erat
ion
(GW
h)
Non-dispatchable Dispatchable Dispatchable resource reduced dispatch
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 49
Substantial transmission upgrades in the SPP footprint over the past few years have provided
an increase in transmission capability for wind-producing regions, helping to address
concerns related to high wind production, and resulting congestion. The increased
transmission capability directly reduces localized congestion, creating a more integrated
system with higher diversity and greater flexibility in managing high levels of wind
production. However, given the historical growth of wind capacity and indicators of future
additions in the generation interconnection queue, additional transmission upgrades may
entice further development of wind capacity.
Figure 2—31 shows the number of out-of-merit energy directives (manual dispatches) initiated
for dispatchable and non-dispatchable variable energy wind resources for the past three
years.
Figure 2—31 Manual dispatches for wind resources
As expected, manual dispatches are fewer during the lower wind output and higher demand
months of summer. In 2018, 20 percent of the 444 manual dispatches were for dispatchable
variable energy wind resources, whereas almost 49 percent were for non-dispatchable
variable energy wind resources. Line loading in excess of 104 percent, operating guides,
0
20
40
60
80
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Mau
nal d
isp
atch
es
2016 2017 2018
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 50
and outages caused 75 percent of dispatchable manual dispatches. These same factors plus
transmission switching21 caused 80 percent of non-dispatchable manual dispatches.
SPP is at the forefront among RTOs in managing wind energy integration. The Integrated
Marketplace has reliably managed wind generation when it represented more than 60
percent of load. Even though the use of manual dispatch is limited and SPP continues to see
an expanding dispatchable wind generation fleet, ramping capability is needed because of
the variability of wind. Since May 2017, ramp shortages are reflected in prices. Section 3.2.1
discusses the pricing of ramp shortages.
2.5.4 CHALLENGES WITH NON-DISPATCHABLE VARIABLE ENERGY RESOURCES
A non-dispatchable variable energy resource is defined as “a variable energy resource not
capable of being incrementally dispatched down by the transmission provider.”22 This
definition does not delve into the requirements of a non-dispatchable variable energy
resource. However, the market design requires that these resources, barring absence of fuel
or mechanical limitations, follow close to their current output or forecast. This concept also
applies to dispatchable variable energy resources not receiving a signal to follow dispatch
and all resources in manual control status that are not in start-up or shutdown. Significant
deviation from the most recent actual or forecasted output causes market inefficiencies that
will be evaluated by the MMU.
Large swings in generation from non-dispatchable variable energy resources responding to
the ex-ante real-time price is known as “price chasing”. This behavior introduces oscillations
on constraints, adversely impacting prices and dispatch instructions for other resources as
well as an impact on regulation products. Price chasing occurs when non-dispatchable
variable energy resources or resources on manual control respond to prices by curtailing
output in response to lower prices and increasing production when prices rise. Such
behavior can cause operational and market issues. For instance, it can create breaches on
flowgates when these resources raise output in response to a price increase. This in turn
21 Transmission switching out-of-merit instructions are issued to accommodate switching of 345kV transmission lines, because of stability concerns during the switching process. Typically, these instructions last from two hours prior to switching to two hours after switching is completed, whereas the 345kV line may be out of service for a longer time frame. 22 SPP Tariff, Attachment AE, Section 1.1.
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 51
causes more relief than necessary and security constrained economic dispatch effectiveness
declines. Other impacts include additional volatility in the real-time market, more regulation
needs, and more output loss due to increased regulation. Operators have at times resorted
to reducing line ratings to ensure system reliability. As a result, out-of-merit energy directives
are issued to other resources, which means extra cost (uplift) to the system, which translates
into reduced market efficiency.
In addition to the inefficiencies introduced to the market because of price chasing behavior,
there are also inefficiencies introduced by non-dispatchable variable energy resources when
they are physically incapable of responding to dispatch signals or when they are acting as
“price takers”. These inefficiencies exist at times when a non-dispatchable variable energy
resource is operating uneconomically, even when considering state and/or federal subsidies.
This results in transmission congestion at times not relieved by the most price effective unit
possible, creating greater price differences and volatility.
The Market Working Group passed a proposal by SPP at the February 2018 meeting that
required full conversion of non-dispatchable variable energy resources. The proposal failed
to pass the Market Operations and Policy Committee meeting in April and was appealed to
the SPP board. The SPP board tabled discussion in April. In July, the Market Worked Group
passed a modified version of the proposal that allowed exemptions for resources exercising
rights under PURPA23 and for run-of-river hydro that could not be dispatchable. This
subsequently passed the July Market Operations and Policy Committee meeting and SPP
board.
SPP filed with FERC to require conversion of non-dispatchable resources to dispatchable
resources in late 2018 and received approval from FERC in April 2019.
2.6 EXTERNAL TRANSACTIONS
2.6.1 EXPORTS AND IMPORTS
The SPP Integrated Marketplace has greater than 6,000 megawatts of AC interties with MISO
to the east, 810 megawatts of DC ties to ERCOT to the south, and over 1,000 megawatts of
DC ties to WECC to the west. Additionally, SPP has over 1,500 megawatts of interties with
23 Public Utilities Regulatory Policies Act of 1978
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 52
the Southwestern Power Administration (SPA) in Arkansas, Missouri, and Oklahoma, and over
5,000 megawatts of AC interties the Associated Electric Cooperative (AECI) in Oklahoma and
Missouri.
As shown in Figure 2—32, SPP has been a net exporter in real-time since 2016. It is important
to note that because of the addition of the Integrated System to SPP in October 2015,
transactions that were external imports from Western Area Power Administration (WAPA)
became internal transactions within the SPP footprint.
Figure 2—32 Exports and imports, SPP system
Generally, SPP exports follow the wind production curve for the day. As wind generation
increases, exports typically increase. However, in 2018, exports were highest in September
and August, reaching almost 2,000 MW and 1,700 MW of gross exports, respectively. This
was driven by lower real-time prices in SPP during these summer months compared to higher
prices in ERCOT and MISO.
Exports to ERCOT were driven by tight supply conditions and high prices during the summer
months. Southwestern Power Administration hydro power is imported to serve municipals
tied to SPP transmission and is highest during on-peak hours, but is scheduled day-ahead.
MISO interchange generally follows wind production, while AECI interchange is coordinated
on an ad hoc basis. DC tie imports and exports are scheduled hourly, and the DC ties are not
responsive to real-time prices. Nonetheless, many exports and imports with ERCOT and
MISO are adjusted based on day-ahead price differences in the organized markets and
-1,500
-1,000
-500
0
500
1,000
1,500
2,000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Ave
rag
e M
W
Exports Imports Net
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 53
expectations of renewable generation. Interchange with SPA, AECI, and Western
Interconnection parties is less responsive to prices. Figure 2—33 through Figure 2—36 show
the data for the four most heavily used interfaces in real-time, namely ERCOT, SPA, MISO,
and AECI.
Figure 2—33 Exports and imports, ERCOT interface
Figure 2—34 Exports and imports, Southwestern Power Administration interface
-200
0
200
400
600
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Ave
rag
e M
W
Exports Imports Net
-800
-600
-400
-200
0
200
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Ave
rag
e M
W
Exports Imports Net
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 54
Figure 2—35 Exports and imports, MISO interface
Figure 2—36 Exports and imports, Associated Electric Cooperative interface
Interchange transactions in the SPP market can be scheduled in the real-time market, as well
as in the day-ahead market. The day-ahead market has three types of interchange
transactions:
• Fixed interchange transactions are physical transactions that bring energy into or out
of the SPP balancing authority. Energy prices are settled at the price at the applicable
external interface settlement location. Submitters of this type of transaction in the
Integrated Marketplace are price takers for that energy.
-500
-250
0
250
500
750
1000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Ave
rag
e M
W
Exports Imports Net
-200
-100
0
100
200
300
400
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Ave
rag
e M
W
Exports Imports Net
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 55
• Dispatchable interchange schedules are physical transactions that bring energy into
or out of the SPP balancing authority and specify a bid or offer for an amount of
megawatts. These schedules are supported in the day-ahead market only and also
must meet all market requirements. Prices are determined in the day-ahead market at
the appropriate external interface settlement location representing the interface
between the SPP balancing authority and the applicable external balancing authority.
• An up-to-transmission usage charge (or up-to-TUC) offer on an interchange
transaction specifies both a megawatt amount and the maximum amount of
congestion cost and marginal loss cost the customer is willing to pay if the transaction
is cleared in the day-ahead market.
All interchange transactions cleared in the day-ahead market, regardless of type, become
fixed interchange transactions in the reliability unit commitment and real-time market.24
As shown in Figure 2—37, of the 1,657 MW of day-ahead import and export transactions in
2018, 90 percent were fixed in the day-ahead market, ten percent were dispatchable, and
none were up-to-TUC. Dispatchable transactions increased threefold from an average of 56
MW in 2017 to an average of 173 MW in 2018. Dispatchable transactions were highest in the
summer months peaking at almost 400 MW in August.
24 Per Market Protocols section 4.2.2.7 Import Interchange Transaction Offers.
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 56
Figure 2—37 Imports and export transactions by type, day-ahead
Some reasons for the fixed transactions that make up the vast majority of interchange
transactions include bilateral contracts with external entities, Southwestern Power
Administration hydro contracts, and generally lower prices of the SPP market compared to
other RTOs. To enhance market efficiency, market participants should consider further use of
the dispatchable and up-to-TUC imports and exports, which allow for a specific strike price to
be set, allowing for more economic imports and exports.
2.6.2 MARKET-TO-MARKET COORDINATION
SPP began the market-to-market (M2M) process with MISO in March 2015 as part of a FERC
requirement that also included regulation compensation and long-term congestion rights.
These were required to be implemented one year after go-live of the SPP Integrated
Marketplace. The market-to-market process under the joint operating agreement allows the
monitoring RTO and non-monitoring RTO to efficiently manage market-to-market constraints
by exchanging information (shadow prices, relief request, control indicators, etc.) and using
the RTO with the more economic redispatch to relieve congestion.25
25 Essentially, the RTO which manages the limiting element of the constraint is the monitoring RTO. In most cases, the monitoring RTO has most of the impact and resources that provide the most effective relief of a congested constraint.
0
500
1,000
1,500
2,000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
MW
h
Fixed Dispatchable Up to TUC
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 57
Each RTO is allocated property rights on market-to-market constraints. These are known as
firm flow entitlements (FFE), and each RTO calculates its real-time usage, known as market
flow. RTOs exchange money (market-to-market settlements) for redispatch based on the
non-monitoring RTO’s market flow in relation to its firm flow entitlement. The non-monitoring
RTO receives money from the monitoring RTO if its market flow is below its firm flow
entitlement. It pays if above its firm flow entitlement. Figure 2—38 shows payments by month
between SPP and MISO (positive is payment from MISO to SPP and negative is payment from
SPP to MISO.)
Figure 2—38 Market-to-market settlements
For 2018, total market-to-market payments from MISO to SPP totaled almost $25 million,
while market-to-market payments from SPP to MISO totaled nearly $9 million, resulting in a
net payment of approximately $16 million from MISO to SPP for the year. Almost half of the
payments from MISO to SPP occurred in the first two months of 2018. Of the $12 million of
payments in January and February, $9 million is attributed to the Neosho – Riverton 161kV
and Nashua transformer constraints. These constraints are discussed more below.
Figure 2—39 shows market-to-market payments (over $100,000 either from SPP to MISO, or
MISO to SPP) by flowgate for 2018.
-$8
$0
$8
$16
$24
$32
-$2
$0
$2
$4
$6
$8
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Mill
ions
Mill
ions
Payments from SPP to MISO Paymnents from MISO to SPP Net payments
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 58
Figure 2—39 Market-to-market settlements by flowgate
Seven flowgates had payments from MISO to SPP over $500,000. Three flowgates had
payments from SPP to MISO over $500,000 while the highest was just over $1 million. The
Neosho-Riverton 161kV flowgate was highly congested in the latter half of 2017 and first few
months of 2018. The line was uprated later in 2018, which alleviated some of the congestion
on this constraint. This constraint is impacted by wind and external flows and is discussed in
more detail in Section 5.1.4.2. Market-to-market payments from MISO to SPP in 2018 on this
flowgate totaled over $9 million which was half of the $18 million in payments in 2017. The
Nashua transformer market-to-market payments totaled over $4 million compared to over
$900,000 in 2017.
Market-to-market allows for a coordinated approach between markets to provide a more
economical dispatch of generation to solve congestion. In most cases, MISO is paying SPP to
help resolve congestion at a lower cost than what was available to MISO and in a few cases,
SPP pays MISO to help resolve congestion. The following are points of interest on the SPP
and MISO seam. The MMU feels continued improvement in these areas would lead to
improved benefits for both markets.
2.6.2.1 Transmission planning
SPP and MISO continue to discuss options to overcome hurdles in coordinated projects
along the seam. Differing cost allocation methodologies, assumptions for separate
transmission planning studies, and separate approval processes can lead to varying results
-$2
$0
$2
$4
$6
Mill
ions
Payments from MISO to SPP Payments from SPP to MISO
$9
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 59
within each RTO/ISO. Benefits are also viewed differently within each region due to the cost
allocation methodology differences.
2.6.2.2 Price/power swings
SPP and MISO implemented the ability to transfer monitoring and non-monitoring RTO roles
in December 2017. MISO and PJM have been utilizing this function to address constraint
volatility or power swings when the non-monitoring RTO may have more “effective control”
on certain constraints. This added feature helps alleviate power swings in cases where an
RTO has more effective control than the other. However, instances where both RTOs have
almost equal amount of flow on a constraint still results in oscillations. SPP and MISO also
implemented the ability to utilize shadow price overrides which will limit the non-monitoring
RTO’s upward swing when limited capability exists on a constraint.
2.6.2.3 Third party impacts
The transmission loading relief (TLR) process is used when tagged impacts or other external
market/non-market flows (other than MISO) impacts are present on an SPP constraint. The
market monitor believes that the transmission loading relief process is not needed when the
SPP and MISO markets have the majority of impacts, but is still needed when external impacts
from non-market (third party) entities are significant. Assuming interface price definitions
correctly reflect congestion, tagged transactions should respond to the market conditions
and either withdraw or delay submitting tags during congestion. Thus, this should alleviate
the need for transmission loading relief when impacts on the constraint are mostly between
SPP and MISO.
When third-party impacts exist, the MMU believes transmission loading relief is warranted to
subject the third party to redispatch.26 A scenario observed between SPP and MISO entailed
third-party firm network and native load impacts that are not subjected to redispatch by
either market. The third party does not have a market signal in the form of a price and by the
absence of a transmission loading relief will not have an obligation to provide relief on the
constraint. Transmission loading relief is not as efficient as a market using price and dispatch
to manage congestion on a constraint. Market-to-market is the preferred method in
addressing congestion along the seams, but until further development is made in areas
26 Third parties impacting the SPP-MISO seam typically include Tennessee Valley Authority, Associated Electric Cooperative Incorporated, and Southwestern Power Administration.
Southwest Power Pool, Inc. Load and resources Market Monitoring Unit
State of the Market 2018 60
outside RTO markets, transmission loading relief is the current mechanism to manage
impacts between markets and non-markets.
2.6.2.4 Market flow methodology
SPP, MISO, and PJM calculate market flows differently. MISO and PJM use a marginal zone
methodology (although the margins are derived in different manners), and SPP uses a
tagging impact approach. This topic is not discussed by SPP and MISO, but because market
flow is a component used in market-to-market settlements, the MMU suggests this topic
should be revisited to ensure consistency and equitable measurements across RTOs.
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 61
3 UNIT COMMITMENT AND DISPATCH PROCESSES
This chapter covers unit commitment and dispatch, scarcity pricing, ramp, and virtual
transactions. Key points from this chapter include:
• In 2018, day-ahead commitments increased by six percent and self-commitments
decreased by four percent. This is a positive trend.
• During the fall months, short-term outages increased by 13 percent from 2017 to
2018.
• In 2018, there were 167 intervals with operating-reserve scarcity. The average scarcity
price for these events was $362/MW. This was roughly five times as many intervals of
scarcity as was seen in 2017. Sixty-four of the 167 intervals of operating reserve
scarcity happened in October.
• Over one-third of the regulation-down scarcity events happened in the first interval of
the hour. This trend has held since the inception of the SPP marketplace.
• SPP is in the process of designing a ramp product which should help to reduce the
frequency of scarcity events.
• Cleared virtual energy offers as a percentage of load were flat year-over-year;
however, cleared virtual energy bids as a percentage of load increased 12 percent
year-over-year.
• Virtual energy offers continue to transact most frequently at wind resources, whereas
virtual energy bids continue to occur most frequently at non-variable energy
resources.
• Virtual profits before fees decreased 19 percent year-over-year and virtual profits after
fees decreased 42 percent year-over-year.
• Average profit per cleared virtual megawatt decreased almost 50 percent in 2018.
• Virtual profits before fees exceeded total virtual losses by more than forty fold,
whereas virtual profits after fees exceeded total virtual losses by almost seven fold.
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 62
3.1 COMMITMENT PROCESS
The Integrated Marketplace uses centralized unit commitment to determine an efficient
scheduling and dispatch of generation resources to meet energy demand and operating
reserve requirements. The principal component of the commitment process is the day-ahead
market, which determines a least cost commitment schedule that meets day-ahead energy
demand and operating reserve requirements simultaneously. Most of the time it becomes
necessary to commit additional capacity outside the day-ahead market to ensure all reliability
needs are addressed and to adjust the day-ahead commitment for real-time conditions. This
is done through the reliability unit commitment (RUC) processes and manual commitments.
SPP employs five reliability commitment processes:
• multi-day reliability assessment;
• day-ahead reliability unit commitment (DA RUC) process;
• intra-day reliability unit commitment (ID RUC) process;
• short-term intra-day reliability unit commitment (ST RUC) process; and
• manual commitment instructions issued by the RTO.
Figure 3—1 shows a timeline describing when the various commitment processes are
executed.
Figure 3—1 Commitment process timeline
Multi-day reliability assessments are made for at least three days prior to an operating day.
This assessment determines if any long lead-time generators are needed for capacity for the
operating day. Any generator committed from this process is treated as a “must commit” in
the day-ahead market. The day-ahead closes at 0930 Central time and is executed on the
day before the operating day, with the results posted at 1400. The day-ahead reliability unit
commitment process is executed approximately 45 minutes after the posting of the day-
ahead market results. This allows market participants time to re-offer their uncommitted
resources.
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 63
The intra-day reliability unit commitment process is run throughout the operating day, with at
least one execution occurring every four hours. The short-term intra-day reliability unit
commitment may be executed as needed to assess resource adequacy over the next two
hour period as part of the intra-day process. SPP operators may also issue manual
commitment instructions for capacity, transmission, or local reliability during the operating
day to address reliability needs not fully reflected in the security constrained unit
commitment algorithm used in the day-ahead and reliability unit commitment processes.
Transmission operators occasionally also issue local reliability commitments.
3.1.1 RESOURCE STARTS
The SPP resource fleet, excluding variable energy resources, had 35,568 starts during 2018.
This is up 26 percent from 28,166 starts last year. This increase can primarily be attributed to
an increase in starts by gas combustion turbine resources. The following two tables and
graphs provide a breakdown of the origins of the commitment decisions for resources. For
all generation participation offers in the day-ahead market by commitment status see Figure
3—10.
Figure 3—2 Start-up instructions by resource count 2016 2017 2018 Day-ahead market 59% 68% 74% Self-commitment 19% 15% 11% Intra-day RUC 10% 8% 7% Manual, regional reliability 5% 4% 5% Short-term RUC 3% 2% 3% Day-ahead RUC 2% <1% <1% Manual, local reliability 2% 1% 1%
Day-ahead market74%
Self-commitment11%
Intra-day RUC7%
Manual, regional reliability
5%Short-term RUC
3%Day-ahead RUC
0%Manual, local reliability
1%2018 total starts: 35,568
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 64
Seventy-four percent of start-up instructions in 2018 were a result of the day-ahead market.
As shown in Figure 3—2, each year there has been an increase in day-ahead market starts and
a decrease in self-commitment starts. This is an encouraging trend as it leads to greater
market efficiency and suggests that resource owners may be gaining confidence in allowing
the market to start-up resources. However, a limiting factor on the number of day-ahead
commitments is that the optimization algorithm is restricted to a 48-hour window; hence,
large base-load resources with long lead-times and long run times may not appear economic
to the day-ahead market commitment algorithm. Some market participants choose to self-
commit these resources, which contributes to many self-commitments. Nonetheless, many
market participants have improved their operating practices to decrease the start-up times
on the units to take advantage of the market commitment process. The day-ahead, intra-day,
short-term, and manual reliability unit commitments represent 13 percent of the resource
start-ups.
Figure 3—3 is based on capacity committed and provides a slightly different look at the data
with the percentages based on capacity committed to start-up.
Figure 3—3 Start-up instructions by resource capacity 2016 2017 2018 Day-ahead market 59% 72% 75% Self-commitment 21% 15% 13% Intra-day RUC 9% 7% 5% Manual, regional reliability 4% 3% 4% Short-term RUC 3% 2% 2% Day-ahead RUC 3% 1% <1% Manual, local reliability 2% 1% 1%
Day-ahead market75%
Self-commitment13%
Intra-day RUC5%
Manual, regional reliability
4%Short-term RUC
2%Day-ahead RUC
0%
Manual, local reliability1%2018 total starts: 35,568
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 65
One important observation is the percentage differences between Figure 3—2 and Figure 3—
3. This is the result of larger resources either self-committed or committed by the day-ahead
market, and smaller resources with shorter lead times committed in the day-ahead reliability
unit commitment, intra-day reliability unit commitment, and manual commitment processes.
Within the operating day, commitment flexibility is limited by resource start-up times. As the
operating hour approaches, fewer resources are eligible to be started. Quick-start units can
be directly dispatched by the real-time balancing market and the reliability unit commitment
processes is not designed to start other units if the quick-start units are able to resolve the
problem. SPP issued 75 percent of all start-up instructions to gas-fired generators in 2018,
up from 72 percent of all start-ups in 2017, and 59 percent in 2016.
Figure 3—4 shows that almost all start-up instructions issued to combined-cycle generators
are the result of the day-ahead market. This result is expected given the lower variable costs
and different operating parameters for these resources relative to other gas units.
Figure 3—4 Origin of start-up instructions for gas resources
Commitment Process
Combined-cycle Simple-cycle,
combustion turbine Simple-cycle, steam turbine
2016 2017 2018 2016 2017 2018 2016 2017 2018 Day-ahead market
92% 94% 93% 59% 69% 76% 46% 64% 73%
Day-ahead RUC
1% 0% 0% 2% 0% 0% 16% 3% 2%
Intra-day RUC
1% 1% 2% 15% 12% 8% 18% 15% 14%
Short-term RUC
0% 0% 0% 6% 4% 4% 4% 4% 2%
Manual instruction
0% 0% 0% 11% 7% 7% 9% 2% 3%
Self 6% 5% 5% 7% 7% 5% 8% 11% 6%
SPP issued day-ahead starts for gas-fired generators with simple-cycle combustion turbine
technology accounted for 76 percent of their total starts. This is an increase from 59 percent
in 2016 and 69 percent in 2017. Steam turbine starts saw a marked increase in the day-
ahead market starts, with 73 percent in 2018, compared to 64 percent the year before. The
MMU is encouraged by the increase in units being started by the day-ahead market, and the
accompanying decrease in manual commitments, as well as self-commitments. Starts in the
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 66
day-ahead market are preferred since the market clearing engine will optimize starts and
dispatch to the most benefit to the market.
Some reliability unit commitments are made to meet instantaneous load capacity
requirements; however, this is not a product generators are directly compensated for by the
market. Therefore, reliability commitment processes, more often than the day-ahead market,
make commitments that may not be supported by real-time price levels. These situations
often lead to make-whole payments. The next section discusses the drivers behind reliability
commitments.
3.1.2 DEMAND FOR RELIABILITY
The previous section noted that eight percent of SPP start-up instructions by capacity
originated from SPP reliability unit commitment processes. To understand the need for the
reliability commitments it is useful to discuss the different assumptions, requirements, and
rules that are used in the reliability unit commitment processes versus the day-ahead market.
A fundamental difference is the definition of energy demand between the two studies. The
energy demand in the day-ahead market is determined by bids submitted by the market
participants, and averages between 99 to 101 percent of the real-time values, as shown in
Figure 2—5.
Another important difference between the two studies is virtual transactions. Market
participants submit virtual bids to buy and virtual offers to sell energy in the day-ahead
market. A virtual transaction is not tied to an obligation to generate or consume energy;
rather, it is a financial instrument that is cleared by taking the opposite position in the real-
time market. Because the reliability unit commitment processes must ensure sufficient
generation is on-line to meet energy demand, virtual transactions are not included in the day-
ahead, intra-day, or short-term reliability unit commitment algorithms.
The assumptions regarding wind generation differ as well. Eighty-five percent of the real-
time wind production cleared in the day-ahead market on an average hourly basis in 2018.
While the market participants determine the participation levels for their wind generators in
the day-ahead market through the use of supply offers, a wind forecast is used by the
reliability unit commitment processes. Import and export transaction data are updated to
include the latest information available for the reliability unit commitment processes.
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 67
These types of differences are referred to as resource gaps (i.e., a gap in meeting demand)
between the day-ahead and real-time markets. The resource gap is the difference between
the (1) excess supply between day-ahead and real-time markets, and the (2) excess demand
between the day-ahead and real-time markets.
The primary drivers for the negative resource gaps are differences in virtual supply net of
virtual demand, differences in real-time wind generation compared to wind cleared in the
day-ahead market, and real-time net exports exceeding day-ahead net exports. It is generally
true that real-time wind generation exceeds the clearing of wind in the day-ahead market.
The mismatch between real-time and day-ahead wind is partly because some market
participants with wind generation assets do not participate or offer the full amount of
forecasted capacity in the day-ahead market. Instead, they take real-time positions given the
uncertainty of the wind generation.
The resource gaps can help explain why additional commitments occur after the day-ahead
market has cleared. Figure 3—5 compares on-line capacity between the day-ahead and real-
time markets.
Figure 3—5 Average hourly capacity increase from day-ahead to real-time
The chart indicates that in 2018 there was, on average, around 1,800 MW of additional
capacity on-line during the real-time market relative to the capacity cleared in the day-ahead
market, a reduction of six percent compared to 2017. However, this is a 33 percent reduction
from 2016. The lower capacity increases in July through November can be primarily
$0
$10
$20
$30
$40
0
1,000
2,000
3,000
4,000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 18
Pric
e ($
/MW
h)
MW
Average capacity increase Real-time price Day-ahead price
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 68
attributed to resources producing closer to their day-ahead totals. The month of May also
had wind producing closer to the day-ahead schedule, but the load came in higher than what
was forecasted in the day-ahead market. This increased the total capacity in real-time.
One of the reasons for reliability commitments is the need for ramp capability. The
instantaneous load capacity constraint may commit additional resources to ensure there is
adequate ramping capacity to meet the instantaneous peak demand for any given hour. The
instantaneous load capacity constraint is defined as the greater of the forecasted
instantaneous peak load, or an SPP defined default value. A value is calculated for upper
bound (upward ramp) and a lower bound (downward ramp). Because the default value is
used the majority of the time, the MMU believes that instantaneous load capacity constraint
can contribute to reliability commitments in excess of the resource gaps. Figure 3—6 shows
the percentage of hours for which the default value is used for the upper bound and lower
bound of instantaneous load capacity.
Figure 3—6 Instantaneous load capacity required
The percent of hours at various upper bound default levels is shown in Figure 3—7. The most
frequent observations were from 500 MW to 599 MW at around 23 percent. Overall, the
2018 levels were less than the values observed after the change in August 2017. The lower
bound default values remained at 500 for all hours of the day.
0%
20%
40%
60%
80%
100%
Upper bound Lower bound
Perc
ent
of t
ota
l ho
urs
Minimum requirement Greater than minimum requirement
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 69
Figure 3—7 Instantaneous load capacity upper bound default levels
Resources committed to provide ramp capability, whether as a result of applying the
instantaneous load capacity constraint in a reliability commitment algorithm or a manual
process, can affect real-time prices. Without the appropriate scarcity pricing rules that reflect
the market value of capacity shortages due to ramp capability, the cost of bringing the
resource on-line may not be fully reflected in the real-time prices.
Reliability commitments, along with wind exceeding the day-ahead forecast, can dampen
real-time price signals, as is evidenced by 42 percent of make-whole payments made for
reliability unit commitments as shown in Figure 4-35.
3.1.3 QUICK-START RESOURCES COMMITMENT
A quick-start resource can be started, synchronized, and begin injecting energy within 10
minutes of SPP notification. This section will detail the ways quick-start resources are used,
which includes quick-start resources dispatched by the real-time market27. This section has
changed from prior years because of the inclusion of real-time market dispatches, which are
not considered commitments since they get dispatched by the real-time market rather than
committed. Quick-start resources do not show up in the current operating plan.
27 Add Protocol Section Reference for quick start logic
0%
5%
10%
15%
20%
25%
Perc
ent
of h
our
s
Head-room requirement (MW)
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 70
For 2018, the definition of a quick-start resource has been updated to include only resources
that were dispatched by the real-time market and were cleared by the market clearing
engine. This differs from the 2017 definition which was any resource that could be on-line in
less than 10 minutes. The new method results in 97 resources counting as quick-start
resources all of which have been offered and used by the real-time market at least once
during 2018 in this capacity.
The more encompassing term “deployment” is being used to account for both commitment
and dispatch of resources. Figure 3—8 summarizes the deployment method available for
quick-start resources along with the number of times each method was selected, lead time,
original commitment hours, and actual hours on-line.
Figure 3—8 Deployment of quick-start resources
Commitment process
Number of starts
Committed available
capacity (MW)
Lead time
(hours)
Hours in original
commitment
Actual hours
on-line
Day-ahead RUC 0 N/A N/A N/A N/A
Intra-day RUC 748 26,331 1.8 3.8 6.0
Short-term RUC 453 18,761 0.3 2.5 4.2
Manual 508 26,642 0.6 4.0 6.4
Day-ahead market 16,991 603,551 21.4 6.1 12.8
Real-time market 12,225 416,994
The real-time market is the only time when these resources are being deployed as quick-start
resources. There is some overlap in the numbers from the real-time market and day-ahead
market commitments because there are a number of the day-ahead market committed
resources having intervals where their commitment period is moved up or extended by the
real-time market.
The level of make-whole payments associated with the commitment of quick-start resources
in the reliability processes is noteworthy. In 2018, 76 percent of the reliability commitments
for quick-start units resulted in real-time make-whole payments, which is similar to 2017 and
2018 at 78 percent. In 2018, quick-start resources received $7 million in real-time make-
whole payments and $350,000 in day-ahead make-whole payments. This is a decrease from
their respective 2017 make-whole payments of $9.5 million and $485,000. The MMU
believes that the commitment costs of quick-start resources needs to be included in the
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 71
commitment decisions of these resources, which is consistent with changes developed
through the stakeholder process.28
With the inception of the short-term reliability unit commitment process in February 2016,
there has been a significant reduction in the number of day-ahead reliability unit
commitments for these units. The short-term reliability unit commitment can commit units in
as little as 15 minutes ahead, increasing certainty of the need for the unit. Before the short-
term reliability unit commitment process was implemented, units had often been committed
hours ahead of the actual start time—sometimes more than a day—ignoring the value of their
flexible capability. The 15 minute lead-time leaves time to commit these quick-start
resources when needed, but allows the commitment to be held off longer providing, more
certainty of the need of the resource. This also minimizes the time these units are at
minimum load levels with market prices below their marginal costs.
The Integrated Marketplace protocols29 describe the real-time market dispatch and
registration options of resources with quick-start capability. Resources use these quick-start
capabilities in their real-time offer, and registered as quick-start capable, can choose not to
receive start-up and/or shutdown orders out of the short-term, intra-day RUC.
Figure 3—9 shows the percent of time quick-start resources generated power and the
relationship of prices to their offer.
28 See MMU comments to FERC regarding pricing of quick-start resources, https://elibrary.ferc.gov/idmws/common/opennat.asp?fileID=14820032. 29 Integrated Marketplace protocols, Section 4.4.2.3.1 and Section 6.1.1.
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 72
Figure 3—9 Operation of quick-start resources
Over the last three years, about 25 percent of the megawatt-hours produced by quick-start
resources30 had energy prices below real-time energy offers. In 2018, this is about the same
as the percentage of offers to energy price for other resources in the SPP footprint, which is
represented by the green line in Figure 3—9. Quick-start resources directly dispatched in
real-time using the quick-start logic are not eligible for a make-whole payment, nor is the
minimum run time respected.31 This is a concern that revision request 116 is designed to
address.
In order to deal with the issue of uneconomic production, SPP staff presented a new quick-
start design proposal that was well-received by stakeholders in May 2015. Subsequently, this
proposal was submitted to the Market Working Group (MWG) by Golden Spread Electric
Cooperative32 and was approved in September 2015. The MMU agrees that incorporating
30 Quick-start resources are defined as those resources with a 10 minute start-up time and a minimum run time of one hour or less. Solar and wind resources are not considered quick-start resources. 31 Protocols 4.4.2.3.1 states that only the offer curves are used to dispatch. 32 Revision Request 116 (Quick-Start Real-Time Commitment) by the SPP board, but was placed on hold in lieu of FERC’s Order issued December 21, 2017.
0%
20%
40%
60%
80%
100%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 18Price above offer Price at or below offer Non-quick-start percent
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 73
commitment costs in the evaluation of the commitment will help reduce the incidence where
these resources are committed and prices are insufficient to cover costs.
While this change was pending FERC filing, FERC began a Section 206 proceeding33 in late-
December 2017 finding SPP’s current practices regarding pricing of quick-start resources
may be unjust and unreasonable. In response, the MMU filed an initial brief and reply
comments highlighting the following positions:
• Commitment costs should be evaluated as part of market optimization to ensure production costs are minimized.
• Quick-start resources should have a minimum run time of less than or equal to one hour.
• Economic minimum operating levels should not be reduced (relaxed) to zero because the approach would set prices based on resource flexibility and availability that is false (does not exist) and would likely require another optimization run that could put pressure on performance time and affect the real-time mitigation process.
• SPP’s screening run and subsequent pricing and scheduling runs are reflective of actual available resources and that the optimization logic uses appropriate operational and reliability constraints (i.e., relies on a resources’ physical offer parameters that determine how and when a resource should operate).
• The Commission’s proposal departs from the marginal pricing methodology upon which the SPP market is based and would likely result in average cost pricing.
• Uplift would likely be transferred under the Commission’s proposal and not necessarily reduced.
• Tariff language regarding the basis of parameters needs to be revised to ensure that information is accurate and based on physical and environmental factors.
FERC has yet to respond to comments and propose further direction on modeling of quick-
start resources. At this time, the MMU does not intend to take further action until FERC has
ruled on the Section 206 proceeding.
33 See FERC Docket No. EL18-35-000: https://www.ferc.gov/whats-new/comm-meet/2017/122117/E-5.pdf.
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 74
3.1.4 GENERATION SCHEDULING
The day-ahead market provides market participants with the ability to submit offers to sell
energy, regulation-up service, regulation-down service, spinning reserves, and supplemental
reserves, and/or to submit bids to purchase energy. The day-ahead market co-optimizes the
clearing of energy and operating reserve products out of the available capacity. All day-
ahead market products are traded and settled on an hourly basis.
In 2018, participation in the day-ahead market was robust for both generation and load.
Load-serving entities consistently offered generation into the day-ahead market at levels in
excess of the requirements of the limited day-ahead must-offer obligation. Participation by
merchant generation—for which no such obligation exists—was comparable to that of the
load-serving entities.
Figure 3—10 Day-ahead market offers by commitment status and participant type
Resource type Owner type Market Self Reliability Not participating Outage
Fossil fuel resources
Load-serving entity 51% 33% 2% 1% 13%
Merchant 74% 14% 0% 0% 12% Variable energy resources
Load-serving entity
53% 43% 0% 0% 4%
Merchant 48% 12% 0% 21% 19%
Figure 3—11 shows generation participation offers in the day-ahead market by commitment
status.
Figure 3—11 Day-ahead market offers by commitment status
0%
20%
40%
60%
80%
100%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec '16 '17 '18
Market Self Reliability Not participating Outage
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 75
The average percent of total offered capacity by commitment status remained consistent with
2017 levels with a slight decrease in the “self-commit” status and increase in the “outage”
status. The “market” commitment status averaged 53 percent while resources with
commitment statuses of “reliability” and “not participating” averaged one percent and four
percent, respectively. The “self-commit” status averaged 30 percent of total offered capacity
which was a slight decrease compared to 32 percent in 2017. 34 The “outage” commit status
averaged 12 percent up from 10 percent in 2017. While self-commits decreased from 2017,
they still constitute a large amount of the capacity offered into the market.
Compared with Figure 3—2 and Figure 3—3 in Section 3.1.1, which shows origins of only initial
starts, these values represent commitment status of all generation capacity offered including
those on-line. Initial starts show 11 percent being self-committed in 2018, whereas all
capacity offered averaged 30 percent. This can be attributed to desire to keep resources on-
line after its initial start even during low prices.
Figure 3—12 shows on-line capacity commitment as a percent of load.
Figure 3—12 On-line capacity as a percent of load
The capacity commitment as a percent of load has decreased significantly over the past few
years, declining from 122 percent in 2016 to just under 120 percent in 2018. Factors that
34 Of the self-committed resources, qualifying facilities (QF) account for three to four percent. Qualifying facilities often use self-commit status to exercise their rights under the Public Utilities Regulatory Policies Act of 1978 (PURPA).
110%
115%
120%
125%
130%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
On-
line
cap
acity
as
per
cent
of d
eman
d
2016 2017 2018
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 76
contribute to lower levels of on-line capacity are fewer self-committed coal plants and the
continued growth of wind capacity and generation.
Generation outages for the past three years are shown in Figure 3—13. Long-term outages
consist of outages with a duration beyond seven days, while short-term outages have a
duration of seven days or less. Additionally, outages are categorized for the reason for the
outage, which includes either maintenance, derates, or forced outages.
Figure 3—13 Generation outages
Generation outages are shown broken down into shoulder35 months and peak months for
each year. As can be observed from the chart above, the vast majority of outages are taken
during the shoulder months, when loads are typically lower. From 2017 to 2018, outages in
the shoulder months increased by 27 percent, after declining by seven percent from 2016 to
2017.
Figure 3—14 shows short-term outages for just the fall period for the past three years.
35 Shoulder months are March, April, May, September, October, and November.
0
20
40
60
80
100
2016 2017 2018 2016 2017 2018
Peak months Shoulder months
MW
h (m
illio
ns)
Long-term derate Long-term forced Long-term maintenanceShort-term derate Short-term forced Short-term maintenance
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 77
Figure 3—14 Short-term generation outages, fall months
For the fall period, short-term outages increased by 13 percent from 2017 to 2018, after
increasing by 11 percent from 2016 to 2017. This change was driven by an increase in short-
term maintenance and short-term derates. Short-term forced outages were about the same
in 2018 as in 2017. The large increase in outages in the fall can have a wide-ranging impact
to the market, including having fewer units able to dispatch when additional generation is
required. The increased outages in the fall period were one of the factors leading to an
increase in scarcity events in fall 2018.
3.2 DISPATCH
The real-time market co-optimizes the clearing of energy and operating reserve products out
of the available offered capacity based on the offer price for each product while respecting
physical parameters. The real-time market clears every five minutes for all products. The
settlement of the real-time market also occurs at the five-minute level, and the settlement is
based on market participants’ deviations from their day-ahead positions.
3.2.1 SCARCITY PRICING
When an insufficient amount of regulation-up service, regulation-down service, or
contingency reserve is cleared, a scarcity price is set by a demand curve. This scarcity price
informs market participants that the product was short and incentivizes future provision of
that product. The scarcity of these products can be caused by a lack of capacity or a lack of
0
2
4
6
8
10
2016 2017 2018
MW
h (m
illio
ns)
Short-term derate Short-term forced Short-term maintenance
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 78
ramp. These scarcities are due to capacity when a resource cannot provide the product
because it is limited by its minimum or maximum operating limit. These scarcities are due to
ramp when a resource cannot provide the product because of its ramp rate. Though scarcity
prices reflect the value of the scarce product, they can also raise the price of other products
when multiple products compete for the same, limited capability of the resource.
Scarcity events in 2018 were in line with 2017, with the exception of October. October 2018
had significant wind volatility, which along with other factors led to an uptick in operating
reserve scarcity events. In fact, 64 of the 167 intervals of operating reserve scarcity that
happened in 2018 happened in October.
While operating scarcity events happened evenly across the hour, regulation scarcity events
do not. Almost two-thirds of the regulation down reserve scarcity intervals in 2018 occurred
in the first interval of the hour. The same trend occurred with regulation up reserve scarcity
events, but to a lesser degree. Almost 30 percent of the intervals of regulation up scarcity
intervals occurred in the first interval of the hour.
The Integrated Marketplace uses scarcity pricing demand curves that administratively set
prices during periods of shortage. In an efficient market, the price of a product is the
marginal cost to produce the next increment of that product. When the next increment
cannot be produced, then the value of that product must be determined by something else.
Demand curves are used to determine the value of a product in this situation. Prices on a
demand curve are set administratively to send economic signals to alert market participants
that future capacity may be needed in this area.
FERC Order No. 825 was released in June 2016. The order stated that: “[W]e require each
RTO/ISO to trigger shortage pricing for any interval in which a shortage of energy or
operating reserves is indicated during the pricing of resources for that interval.” 36 At the
time of the order, SPP did not price ramp related shortages because those events were
considered transient in nature.
Revision request 175 “Ramp Shortage Compliance” was developed to bring SPP in
compliance with FERC Order No. 825 and was implemented in May 2017. Revision request
36 FERC order number 825 Settlement Intervals and Shortage Pricing in Markets Operated by Regional Transmission Organizations and Independent System Operators ,https://www.ferc.gov/whats-new/comm-meet/2016/061616/E-2.pdf ,citation 162 (June 16, 2016) , page 90.
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 79
198 was introduced to minimize the impacts of these increased scarcity events and was
implemented in August 2017. The revision request put variable price demand curves in
place of the fixed price demand curves. These variable price demand curves buffer the
impacts of scarcity pricing.
Prior to revision request 198, each products’ scarcity demand curves used one scarcity price.
Now regulation-up, regulation-down, and contingency reserves have multiple price points
depending on how short the cleared operating reserve was of the requirement. For
example, regulation-up and regulation-down had one scarcity price of $600/MW prior to
August implementation. This meant that being short just one megawatt of regulation
capacity would cause a $600/MW price for that product.37 Now there are six graduating
price points on the regulation up and down reserves’ scarcity demand curves and three for
operating reserves.
As an outcome of SPP pricing the ramp shortages, the MMU has difficulty discerning capacity
shortages from ramp shortages. The clearing engine does not record the reason for the
scarcity. The MMU suggests that SPP capture the appropriate information so that the reason
for the scarcity will be transparent.
Figure 3—15 displays the number of scarcity intervals for 2018 and compares the annual
scarcity numbers with 2017. The 2017 average scarcity prices in the chart only reflect the
periods from May 11 to the end of the year in 2017. This is because SPP did not start pricing
ramp scarcity events prior to that time.
37 The 2017 Market Monitoring Unit Annual State of the Market Report provides more details around the development of the ancillary service requirements and the structuring of the variable offer curves. https://www.spp.org/documents/57928/spp_mmu_asom_2017.pdf , page 110.
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 80
Figure 3—15 Scarcity intervals and marginal energy cost, after scarcity implementation
In 2018, there were 425 intervals with regulation-up reserve scarcity, 117 intervals with
regulation-down scarcity, and 167 intervals with operating-reserve scarcity. The average
respective scarcity prices for the events were $197/MW, $135/MW, and $362/MW. Sixty-four
of the 167 intervals of operating reserve scarcity happened in October. This month had
volatile wind fluctuations, along with baseload resources on outage, as is typical during the
shoulder months. When baseload resources are on outage, fast moving resources often run
up towards their maximum limits to make up for the lost capacity. This means that when
contingencies occur, like unexpected wind drops, the fast moving resources may not be
available to respond.
The frequency of regulation-down scarcity events decreased substantially from a monthly
average of 24 intervals in 2017 to just under 10 intervals per month in 2018. Regulation-up
scarcity was similar to 2017, averaging 35 intervals a month compared to 38 intervals in 2017.
The biggest increase happened with operating-reserve scarcity with an average of 14
intervals per month compared to three intervals per month in 2017. However, as stated
above, the 64 intervals in October were one of the leading reasons for the average monthly
increase.
Regulation up and down price spikes happed more frequently at the beginning of each hour.
Figure 3—16 below illustrates a count of the 2018 scarcity events in the real-time market by
the 12 intervals of each hour.
$0
$300
$600
$900
$1,200
0
50
100
150
200
Mar
gin
al e
nerg
y co
st
($/M
Wh)
Scar
city
inte
rval
s
Regulation-up scarcity Regulation-down scarcity Operating reserve scarcityRegulation-up scarcity price Regulation-down scarcity price Operating reserve scarcity price
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 81
Figure 3—16 Scarcity events by interval of the hour
Almost two-thirds of the 117 regulation down reserve scarcity intervals in 2018 occurred in
the first interval of the hour. The same trend occurred with regulation up reserve scarcity
events, but to a lesser degree. Almost 30 percent of the 425 intervals of regulation up
scarcity intervals in 2018 occurred in the first interval of the hour. Operating reserves appear
to be equally dispersed across the hour. The trend for regulation scarcity events occurring at
the beginning of the hour is not unique to 2018. This trend has occurred since the inception
of the marketplace. However, the pricing impacts were less severe prior to May 2017, as
ramp scarcity was not priced.
One potential reason for the trend is because SPP does not preposition regulating resources
to be at their regulating effective maximum and minimum limits prior to the period that the
resource is cleared for regulation. For example, assume a resource is currently being
dispatched to its economic minimum of 100 megawatts and it cleared for 20 megawatts of
regulation down reserves in the next hour. In order for the unit to be able to provide the
regulation down reserves it will need to move up to 120 megawatts. However, if the resource
moves there prior to the hour it will deviate from its current dispatch instruction, which has
0
25
50
75
100
125
:00 :05 :10 :15 :20 :25 :30 :35 :40 :45 :50 :55
Inte
rval
s
Interval beginningRegulation up scarcity Regulation down scarcity Operating reserve scarcity
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 82
negative financial penalties.38 This is why resources often follow dispatch until the hour their
regulation commitment begins, contributing to shortages in the first interval of the regulating
commitment.
The reason that regulation down has such a higher percentage in the first interval than
regulation up is likely because there is more ramp down offered into the market than ramp
up, as wind resources can only be ramped down. In either case, there is an argument for SPP
to preposition resources to their regulating ranges prior to the regulation period. However,
opportunity costs could occur for these resources during the preposition period and
compensation may be needed.
Scarcity prices apply to regulation up, regulation down, and operating reserves. Spinning
reserve scarcity is currently not priced through scarcity demand curves. Instead, SPP uses a
violation relaxation limit (VRL) during periods of spin scarcity. The VRL for spinning reserve
scarcity is set at $200/MW. This means that SPP will keep dispatching resources to meet the
spinning reserve requirement up to the point when the redispatch cost reaches $200/MW.
At this point, the market will relax the spinning reserve requirement to the quantity of
spinning reserve megawatts that can be obtained at a redispatch cost under the $200/MW
shadow price. The spinning reserve price will be the price of the marginal resource cleared
at the relaxed limit for that product, plus any scarcity pricing from an operating reserve
scarcity event. Since spinning reserves are a higher priority product than supplemental
reserves the market will never set the spinning reserve price lower than the supplemental
reserve price.
In most instances when the VRL shadow price is reached the marginal clearing price is set
near $200/MW. However, in circumstances where the spinning reserve is scarce because of
competition with other products, the spinning reserve shadow prices can be set at $0/MW.
Even though the shadow prices are $0/MW, the actual spinning reserve clearing prices are
typically set by an operating reserve scarcity event price. Figure 3—17 shows a scatter plot of
2018 spinning reserve shortages.
38 Resources that deviate from their dispatch signals can receive Uninstructed Deviated Charges for the deviated megawatts. The median price per megawatt for deviation in 2018 was $0.54. However, the maximum price charged for deviation was $11.01. In addition to this charge, resources do not receive cost reimbursement for any deviated megawatts in the event energy prices are lower than energy cost.
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 83
Figure 3—17 Spinning reserve shortages
This figure plots spinning reserve shadow prices against the percentage of spinning reserves
short, during periods of shortage. As can be seen in the chart, prices most frequently get set
near the $200/MW shadow price limit. However, there are often instances where the shadow
prices can be set near or at $0/MW. In 2018, there were three intervals where shadow prices
were set at $0/MW. During these intervals the spinning reserve clearing prices were set at, or
just above, the supplemental reserve price. In two of those three intervals that supplemental
reserve price was set by the operating reserve scarcity demand curve.
Energy shortages are currently priced using the VRL process. There are three VRL constraints
associated with energy shortages. The current energy shortage VRL shadow prices are
$5,000/MW for ramp shortages, $50,000/MW for balancing resources dispatch to load
consumption, and $100,000/MW for resource capacity shortages.
3.2.2 RAMPING
The increase or decrease of the resource’s output to achieve the next dispatch instruction is
called “ramp”. The number of megawatts a resource can ramp in one minute is the
resource’s “ramp rate”.
In real-time, resources are increasing and decreasing output to meet changes in both load
and generation. These changes can be measured as changes in net load. Net load is load
net of both variable energy generation and the combination of imports and exports.
$0
$50
$100
$150
$200
$250
0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50%
Spin
sha
do
w p
rice
s ($
/MW
)
Spinning reserve short as a percent of requirement
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 84
Figure 3—18 shows the frequency of how often the net load changes from one real-time
interval to the next.
Figure 3—18 Frequency of net load change, real-time
About 15 percent of intervals have a net load change of over 200 megawatts when both
positive and negative changes are considered. Some of these net load changes peaked over
1,000 megawatts. While the percentage of total intervals is low, these greater deviations
cause greater price volatility. This difference must be provided by resources with a flexible
dispatch range, or ramp capability.
Figure 3—19 below shows that the volatility in net load change increased significantly each
year from 2015 through 2017, but decreased slightly in 2018.
0%
10%
20%
30%
Perc
ent
of i
nter
vals
Interval to interval net load change
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 85
Figure 3—19 Volatility of interval-to-interval net load change
Net load volatility increased about 10 percent from 2015 to 2016 and about 19 percent from
2016 to 2017. Then in 2018, the net load volatility dropped by about five percent. Even
though the volatility in 2018 is slightly lower than in 2017, volatility is trending upward over a
longer period.
The dispatch typically assumes that variable energy resources remain at the same output
level at which they were last observed. Practically, this will not be the case. This discrepancy
can require more ramp in the next interval. Nearly one-quarter of total generation was
produced by wind and solar resources in 2018. This was a significant source of net load
changes. With over 55 gigawatts of wind-powered generation and nearly 25 gigawatts of
solar generation with an active generation interconnection request, this problem will likely
increase in the future as volatility from variable energy resources increases.
In any interval, resources typically have a range either above or below their current operating
point that they can move to in the next interval. The rate at which they can move is their ramp
0
40
80
120
160
200
2015 2016 2017 2018
Net
load
cha
nge
(MW
) vo
latil
ity
(sta
ndar
d d
evia
tion)
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 86
rate. The total amount they can ramp is their rampable capacity. SPP operators count on this
rampable capacity to meet future energy needs and to protect against uncertainty.39
Rampable capacity in the up direction is lowest in the spring and fall seasons when averaged
by month. These seasons are known for high wind and outages. Figure 3—20 shows the
average up-rampable capacity by month after energy and operating reserve obligations are
accounted for.40
Figure 3—20 Average up-rampable capacity by month
The amount of rampable capacity in the down direction is much larger than in the up
direction when averaged by month, as shown in Figure 3—21. The monthly average does not
change very significantly throughout the year.
39 This ramp in not currently being systematically procured by the software or being compensated as stand-by capacity. See section 3.2.3 on ramping capability product for further detail on the MMU’s recommendation to address these concerns. 40 The figures showing average rampable capacity are approximations. The market clearing engine allows for ramp sharing and also allows for some products to go short so that higher priority products can clear. There can be different amounts of rampable capacity available depending on the product. These graphs average the amount of load increase or decrease that would cause a shortage on the product that is nearest to a shortage.
0
250
500
750
1,000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 87
Figure 3—21 Average down-rampable capacity by month
Figure 3—22 shows the average rampable capacity in the up direction by hour of the day.
Rampable capacity in the up direction is lower following the morning ramp in hours
beginning 08, 09, and 10. From hour beginning 11 until hour beginning 21, the rampable
capacity increases slightly but remains lower. As resources move closer to their minimum
limits during the night, this rampable capacity increases.
Figure 3—22 Average up-rampable capacity by hour
-4,000
-3,000
-2,000
-1,000
0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
0
250
500
750
1,000
00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Ave
rag
e up
-ram
pab
le c
apac
ity
(MW
)
hour beginning
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 88
Figure 3—23 shows the average rampable capacity in the down direction by hour. As
previously mentioned, there is much more rampable capacity in the down direction than in
the up direction. Although this rampable capacity is less in the late evening hours, this
amount does not vary significantly throughout the day.
Figure 3—23 Average down-rampable capacity by hour
Ramp capability is needed to meet all of these changes in generation and load from interval
to interval. However, unlike capacity, ramp is not procured ahead of time. A resource’s ramp
rate is used to calculate its dispatch instruction. A resource will generally not be dispatched
beyond the capability of the resource’s ramp rate. However, the ramp rate is used only to
calculate the current dispatch instruction. Unlike capacity, the market clearing engine does
not specifically procure ramp ahead of time to meet these intra-hour net load changes,
although committed capacity usually comes with incidental ramp capability. Because the
committed capacity must have ramp capability, the market clearing engine should specifically
procure and incentivize rampable capacity.
3.2.3 RAMP CAPABILITY PRODUCT
In light of the scarcity events and the demonstrated need for ramp capability, a ramp
capability product would be beneficial to the market. A resource’s ability to ramp should be
planned for and should be valued by a price to the extent the ramp is beneficial to the
market. The market monitor prefers the use of a variable demand curve to accomplish this.
-4,000
-3,000
-2,000
-1,000
0
00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Ave
rag
e d
ow
n-ra
mp
able
cap
acity
(M
W)
hour beginning
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 89
3.2.3.1 Ramping limitations affect market outcomes
The real-time dispatch does not consider future intervals. It simply calculates one value: a
dispatch instruction for the next interval. While the real-time balancing market considers a
resource’s ramp capability for the purpose of calculating the dispatch instruction for the next
interval, ramp is not considered for any interval after that. Ramp is not currently accounted
for in terms of the subsequent dispatch instructions even though ramp is the very capability
that allows a resource to get to future dispatch instructions.
When ramp capability is not considered for future intervals, then the market clearing engine
may not be able to procure enough energy to serve the load or provide sufficient operating
reserves in those future intervals. Even when enough capacity is available, a lack of ramp
renders that capacity unreachable. This often leads to short-term transitory price spikes,41 as
seen in Figure 3—24.
Figure 3—24 Interval length of short-term price spikes
This figure shows that a scarcity pricing event in real-time was most likely to occur for only
one five-minute interval. Very few scarcity pricing events last more than two intervals. This
pattern was consistent in both 2017 and 2018.
Figure 3—25 shows the interval length for the different types of scarcity events.
41 This is essentially temporal, or time-based, congestion.
0
50
100
150
200
250
1 interval 2 intervals 3 intervals 4 intervals 5 intervals 6 or moreintervals
Scar
city
eve
nts
Event length in intervals
Regulation up Regulation down Operating reserve
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 90
Figure 3—25 Interval length of short-term price spikes, percentage
Of all the regulation-up scarcity events, about 70 percent lasted for only one interval and
about 15 percent lasted for two intervals. For regulation down scarcity events, about 90
percent lasted for only one interval. Operating reserve scarcity event lengths were more
diverse with about 40 percent lasting for one interval, about 20 percent lasting two intervals,
and about 20 percent lasting three intervals. Operating reserve scarcity events lasted up to
eight intervals, though there were very few events that lasted longer than three intervals.42
Scarcity events in the SPP market have historically been short in duration, and 2018 was no
exception.
Where sufficient capacity cannot be dispatched, scarcity prices are invoked. Scarcity prices
are economic signals alerting market participants to the insufficient supply of a product.
Almost all of these intervals with scarcity pricing were due to a lack of cleared ramp and not a
lack of capacity. If sufficient ramp were reserved in advance for these scarcity intervals, then
these scarcities likely could have been avoided.
In addition, marginal energy prices can be elevated even when energy is not scarce. When
ramp in the up direction is short, energy will always be given the highest priority. If there is
not sufficient ramp to meet both energy and regulation-up, for instance, then the regulation-
42 In the 2017 Annual State of the Market report, energy scarcity was shown in a similar graph. In 2018, a violation relaxation limit was used to mitigate energy shortages instead of a demand curve.
0%
20%
40%
60%
80%
100%
Regulation up Regulation down Operating reserve
Scar
city
eve
nts
Event length in intervals
1 interval 2 intervals 3 intervals 4 intervals 5 intervals 6 or more intervals
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 91
up scarcity price will be reflected in the marginal energy price. This causes a high marginal
energy price even though there is no energy scarcity because the two products are
competing for ramp. This makes prices more volatile. The regulation-up scarcity price raised
the marginal energy price in this way for about 30 percent of all regulation-up scarcity
intervals in 2018. If sufficient ramp had been available, then regulation-up scarcity prices
would not have raised the marginal energy price. A ramp capability product can ensure that
more ramping is available to meet energy so that regulation-up scarcity prices can be
avoided. This helps to better reflect system conditions and reduces dispatch volatility.
When prices are high in real-time, analysis of settlement data reveals that resources pay for
most of this cost to the market.43
Loads are typically well-hedged in the day-ahead market, so when prices spike in real-time, a
load’s settlement is typically not affected much. This is because the real-time settlement is
based on the difference between the withdrawal in real-time and in the day-ahead market.
Because loads are well-hedged, the difference between the two markets is small. This small
difference causes the real-time price to have little overall effect on load settlements.
Imports and exports have a small effect on the market overall, so they do not pay much when
real-time prices are high.
Though virtual offers pay more for high real-time prices than the aforementioned categories,
they do not pay the most. When real-time prices are high, the amount virtual offers pay is
less than when prices are low. This indicates that, overall, virtual offers adjust their market
behavior so that they are affected less by the high real-time prices.
Resources that produce less in real-time than day-ahead are mostly paying for the resources
that produce more in day-ahead than real-time. Although load pays for energy in the day-
ahead market, resources that buyback in the real-time market typically pay for the real-time
energy price above what load paid in the day-ahead.
43 Ramping Product, February 19-20, 2019 MWG, (https://www.spp.org/Documents/59502/mwg%20agenda%20&%20background%20materials%2020190219%2020.zip).
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 92
3.2.3.2 Clarification of preferences on ramp product design
In the 2017 annual report, the market monitor recommended that SPP create a ramp
capability product. Previously, the market monitor recommended that a ramp capability
product include the following features:
• Two products: ramp capability up and ramp capability down;
• Co-optimization with energy and other products to ensure the most economical
solution;
• Opportunity cost basis for pricing;
• No limitations on resource type as long as the resource can reliably provide ramp in
the direction for which it is cleared; and
• Consideration of both expected and unexpected ramping needs.
SPP has started on this design and has discussed progress with SPP’s Market Working Group.
At the time this report was written, the Market Working group has discussed design choices
such as:
• A violation relaxation limit versus a demand curve;
• Interaction between the ramp product and instantaneous load capacity; and
• Transmission constraints on ramp.
Though the role of the market monitor is to advise and not drive policy choices, some design
elements are more preferable to the market monitor than others. In light of this, we share
some preferences below. These preferences do not indicate that there is one design that the
market monitor will find acceptable.
3.2.3.2.1 Demand curves reflect the value of ramp in the price
The market monitor prefers demand curves over violation relaxation limits because demand
curves allow the product’s value to be reflected in the price. Though there can be many right
answers to the value of an unavailable product, it is inappropriate to say that there is no
value. Relaxing the ramp constraint without pricing it informs the market that the product
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 93
was not valued and is not beneficial. A requirement that is unfulfilled without cost or
consequence is not a requirement.
Furthermore, the resources that supplied the ramp may be under-incentivized to supply it in
the future because no economic signal is sent to the market. When resources have no
incentive to supply ramp that is necessary to achieve dispatch, that necessary ramp must be
acquired suboptimally and without payment. For these reasons, the market monitor prefers a
demand curve to reflect the value of ramp.
Additionally, a variable demand curve is preferable. A single-value (flat) demand curve
reflects merely that a product is short, but a variable demand curve reflects how short the
product is. This can be thought of as ramp insurance. Based on historical observations of
ramp needs, the probability can be calculated that a certain amount of ramp will be needed
in the future. The cost to the market that is avoided by supplying the ramp can be used to
price the certainty that a ramp scarcity will be avoided. These probabilities and avoided costs
can be represented much like the demand bid curves that are used in the market today. In
this way, the price for being 10 megawatts short of the ramp requirement should be cheaper
than the price for being 100 megawatts short.
3.2.3.2.2 Intra-hour and inter-hour ramp products may be appropriate
The process known as instantaneous load capacity is ramp procurement without ramp
payment. The instantaneous load capacity currently ensures that sufficient rampable capacity
is committed to ramp from one average hourly load to the next. To the extent that a resource
is being dispatched uneconomically so that its ramp can be used in this way, the resource
should be paid for that value. If it is determined that more than one ramping timeframe is
needed for more than one ramping purpose,44 then the market monitor could support
multiple ramp products.
It is clear that for resources to be dispatched so that future ramp can be provided, sufficient
rampable capacity must be committed. A ramp requirement should be included in the
commitment process.
44 For instance, it may be appropriate to have a ramp product similar to instantaneous load capacity to address inter-hour ramping needs in addition to an intra-hour ramp product to address short-term load variability.
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 94
3.2.3.2.3 Considering transmission congestion when clearing ramp is not necessary at this
time
Though transmission constraints could keep cleared ramp from being useful, the market
monitor does not see this as an issue that needs to be resolved in the initial implementation
of the ramp product. It is possible that ramp could be cleared on a resource that would not
be able to deliver the ramp because it is behind a congested transmission element.
Currently, regulation clearing does not account for congestion. Even though the SPP market
has delineated reserve zones, there is currently no zonal requirement for regulation clearing.
Ramp should be a lower priority than regulation. While transmission constraints could cause
inefficient ramp clearing, it is not necessary to clear the lower-quality ramp more precisely
than the higher quality regulation.
3.3 MUST-OFFER PROVISION
The Integrated Marketplace has a limited day-ahead must-offer provision that was intended
to incentivize load-serving entities to participate in the day-ahead market. Market
participants that are non-compliant are assessed a penalty based on the amount of available
capacity available in the day-ahead market relative to the market participant’s real-time load.
The requirement is limited in the sense that only market participants with generation assets
that serve load are subject to the rules. Load-serving market participants that offer enough
generation, and/or provide scheduling information indicating a firm power purchase to cover
at least 90 percent of their real-time load are not subject to a penalty. An alternative way to
satisfy the provision is to offer all generation that is not on outage. In 2018, one day-ahead
must-offer penalty, totaling less than $500, was assessed. No penalties were assessed in
2017.
In 2014, the MMU recommended that SPP simultaneously eliminate the limited day-ahead
must-offer provision and revise the physical withholding rules to include a penalty for non-
compliance based on the premise that the recommended penalty provision would be
sufficient to ensure an efficient level of participation in the day-ahead market. SPP
stakeholders then approved the removal of the day-ahead must-offer with no additional
physical withholding provisions, and SPP filed the tariff revision with FERC in the summer of
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 95
2017. FERC denied the removal of the limited must-offer requirement, as it did not include
physical withholding non-compliance penalties.45
The MMU continues to recommend updating the day-ahead must offer requirement and
addressing FERC’s concerns. However, given the status of other higher priority initiatives, the
MMU assigns a low priority to addressing the issue at this time. See further discussion in
Section 8.3.6.
3.4 VIRTUAL TRADING
Market participants in SPP’s Integrated Marketplace may submit virtual energy offers and bids
at any settlement location in the day-ahead market. Virtual offers represent energy sales to
the day-ahead market that the participant needs to buy back in the real-time market. These
are referred to as “increment offers”, which are like generation. Virtual bids represent energy
purchases in the day-ahead market that the participant needs to sell back in the real-time
market. These are referred to as “decrement bids”, which are like load. The value of virtual
trading lies in its potential to converge day-ahead and real-time market prices, and improve
day-ahead unit commitment decisions.
In order for virtual transactions to converge prices, there must be sufficient competition in
virtual trading; transparency in day-ahead market, reliability unit commitment, and real-time
market operating practices; and predictability of market events. Since the market began in
2014, there has been moderate, and increasing levels of virtual participation. Figure 3—26
displays the total volume of virtual transactions as a percentage of real-time market load
along with wind output levels.
45 See FERC ruling at https://www.ferc.gov/CalendarFiles/20171013130834-ER17-2312-000.pdf
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 96
Figure 3—26 Cleared virtual transactions as percent of real-time load
As shown in the figure, virtual transactions averaged 14.6 percent of real-time market load,
compared to 13.9 percent in 2016. Historically, the greatest increases in virtual transactions
as a percentage of load have been with cleared virtual offers. However, this trend stopped in
2018, as the percent of virtual offers to load was 8.1 percent, the same as 2017. This percent
was 5.4 percent in 2016 and 4.0 percent in 2015. Virtual cleared bids increased from 4.2
percent in 2016 to 5.8 percent in 2017 and further upward to 6.5 percent in 2018. Days with
high wind output see an increase in virtual offer activity. Virtual bids typically increase during
high load hours.
At just under 15 percent of load, the average hourly total volume of cleared virtuals ranged
from 1,947 MW of withdrawal, to 2,436 MW of injection. The net cleared virtual positions in
the market averaged about 489 MW of injection or supply each hour.
While the majority of virtual transactions occurred at wind resources in 2017, a trend that has
been increasing since mid-2015, fewer virtual transactions occurred at wind resources in
2018. Figure 3—27 illustrates the settlement location types where virtual offers clear.
0
2,000
4,000
6,000
8,000
0%
5%
10%
15%
20%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Win
d g
ener
atio
n (G
Wh)
Perc
ent
of l
oad
Cleared bids as a percent of load Cleared offers as a percent of load
Wind output (GWh)
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 97
Figure 3—27 Cleared virtual offers by settlement location type
This figure shows that over 43 percent of the virtual offers cleared at variable energy
resources during 2018.46 This is a decrease from 49 percent in 2017 and an increase from 36
percent in 2016. Even so, virtual offers at wind locations remain the largest volume of any
single location type. These large volumes highlight the possibility that participants of wind
resources may be missing financial opportunities by under-scheduling in the day-ahead
market.47
This is in contrast with the locational volumes of virtual bids. Cleared virtual bids were
primarily at resources other than variable energy resources, followed by load locations.
Figure 3—28, below, shows the cleared virtual bids by settlement location types.
46 This includes both dispatchable and non-dispatch variable energy locations. 47 Section 4.1.5 on price divergence discusses the effects of unscheduled wind in the SPP market.
0
500
1,000
1,500
2,000
2,500
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Cle
ared
vir
tual
offe
rs (G
Wh)
Non-dispatchable variable energy resource Dispatchable variable energy resource Non-variable energy resource Hub Interchange Load
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 98
Figure 3—28 Cleared virtual bids by settlement location type
Forty-one percent of cleared virtual bids in 2018 were transacted at non-variable energy
resource locations, which is down from 43 percent in 2017 and up from 39 percent in 2016.
Another 31 percent of the total cleared virtual bid volume was transacted at load locations in
2018, which is up slightly when compared to 2017 (28 percent) and 2016 (26 percent). The
associated quantities also increased 26 percent from 2017 to 2018 and 50 percent from 2016
to 2017. Virtual bids at wind locations declined both as a percent of total virtual bids and in
volumes in 2018.
Figure 3—29 shows cleared demand bids that offered more than $30/MWh over the cleared
day-ahead price, and the supply offers offered at less than $30/MWh under the cleared day-
ahead price.
0
500
1,000
1,500
2,000
2,500
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Cle
ared
vir
tual
bid
s (G
Wh)
Non-dispatchable variable energy resource Dispatchable variable energy resource Non-variable energy resource Hub Interchange Load
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 99
Figure 3—29 Virtual offers and bids, day-ahead market
These types of bids and offers are called “price-insensitive” and occurred more often with
bids up until 2017. Price-insensitive bids have remained steady with about 12 to 15 percent
of cleared bids since 2016, but that number has increased for offers from six percent in 2016
to 11 percent in 2018. Price-insensitive bids and offers are willing to buy/sell at a much
higher/lower price that could lead to price divergence rather than competitive, or price-
sensitive, bids and offers leading to price convergence between the day-ahead and real-time
markets. Price-insensitive bids and offers usually occur at locations with congestion and
arbitrage against the day-ahead and real-time price differences. Given that price-insensitive
bids and offers are likely to clear, these can be unprofitable if congestion around these
locations does not materialize, leading to divergence between the markets.
Financial information for virtual trades is shown monthly and on an annual basis for 2018 in
Figure 3—30.
-6,000
-4,000
-2,000
0
2,000
4,000
6,000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Ave
rag
e ho
urly
vo
lum
e (M
W)
Cleared bids < $30/MWh Cleared offers < $30/MWhCleared bids > $30/MWh Cleared offers > $30/MWhUncleared bids Uncleared offersNet cleared
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 100
Figure 3—30 Virtual profits with distribution charges
Month Raw profit Raw loss
Raw net profit
(prior to fees)
RNU charges/ credits
Day-ahead make-whole
payment charges
Real-time make-whole
payment charges
Virtual transaction
fee
Total net profit
January $ 22.3 $ (13.9) $ 8.4 $ 0.9 $ 0.3 $ 2.3 $ 0.0 $ 4.9 February 15.9 (14.3) 1.6 0.2 0.1 0.7 0.0 0.5 March 15.0 (12.5) 2.5 0.1 0.2 1.0 0.0 1.2 April 19.6 (15.2) 4.4 0.8 0.2 0.8 0.0 2.6 May 14.1 (13.9) 0.2 0.4 0.1 2.1 0.0 (2.4) June 15.5 (12.7) 2.8 0.3 0.1 1.2 0.0 1.1 July 9.4 (8.2) 1.2 0.2 0.1 1.9 0.0 (1.1) August 8.0 (6.0) 2.0 0.2 0.1 1.2 0.0 0.5 September 14.1 (10.0) 4.1 0.5 0.1 0.9 0.0 2.5 October 19.7 (16.0) 3.8 0.4 0.1 2.1 0.0 1.2 November 22.4 (15.6) 6.8 0.7 0.1 3.2 0.0 2.8 December 17.6 (11.1) 6.5 0.7 0.1 1.6 0.0 4.0 Total 193.5 (149.3) 44.2 5.4 1.7 18.7 0.5 17.8
All figures in $ millions
Virtual trades profited in aggregate for the year by about $44 million, a 19 percent decrease
year-over-year. Virtual bids can be charged distribution fees for day-ahead make-whole
payments and virtual offers are susceptible to real-time make-whole payment distribution
fees. In addition, both types of transactions can receive revenue neutrality uplift
charge/credits and are assessed a $0.05 per virtual bid or offer transaction fee for processing
virtual transactions. The average 2018 rates per megawatt for day-ahead make-whole
payments, real-time make-whole payments, and real-time revenue neutrality uplift
distributions are $0.04/MWh, $0.50/MWh, and $0.14/MWh, respectively. When factoring in
these charges and credits, the net virtual bidding profits for 2018 were $17.8 million, 60
percent lower after factoring in fees. Net profits in 2018 decreased 42 percent from $34.6
million in 2017.
Every month in 2018 was profitable in aggregate for virtual transactions before factoring in
transaction fees. However, after accounting for these fees, May and July were unprofitable in
aggregate. In the 58 months since the market began, only nine months have had a net loss
when factoring in fees. The highest payout months in 2018 happened in January and
December with net payouts just over $4.9 million and $4.0 million, respectively. These
months coincide with high wind periods and low load months when high price differences
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 101
can occur between day-ahead and real-time markets as a result of under-scheduled wind in
the day-ahead market.48
Net profits are typically small when assessed on a per megawatt basis. Figure 3—31 illustrates
the monthly average profit per megawatt for a cleared virtual in 2018.49
Figure 3—31 Profit and loss per cleared virtual, average
The chart shows that when factoring in all fees the average profit per megawatt for 2018 was
$0.46 per cleared megawatt, almost a 50 percent decrease from $0.91 in 2017, and a 28
percent decrease from the $0.59 in 2016.
There were 84 SPP participants who transacted virtuals in 2018. Figure 3—32 illustrates each
virtual participant’s virtual portfolio for the year by both net megawatts cleared and net
profits before adjusting for fees.
48 Section 4.1.4, price divergence, discusses the effects of unscheduled wind in the SPP market. 49 A change was made to the calculation of revenue neutrality uplift charges for virtuals. This correction resulted in a $0.06 per cleared megawatt reduction in the 2017 totals presented last year.
-$1.00
$0.00
$1.00
$2.00
$3.00
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Cle
ared
vir
tual
pro
fit/l
oss
per
MW
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 102
Figure 3—32 Virtual portfolios by market participant
Five participants accounted for over 46 percent of the virtual profits before fees, which can
also be referred to as gross profits. These participants also account for roughly 50 percent of
the cleared transactional volume in the market. In aggregate, virtual trading generated gross
profits for most participants. However, nineteen virtual participants lost money on a gross
profit basis. The total losses before fees amounted to roughly $1.2 million, and five entities
accounted for over $900,000 of that loss. However, total gross trading profits exceeded total
gross trading losses by more than forty-fold.
Five participants also accounted for over 46 percent of the virtual profits after fees, which can
also be referred to as net profits. These participants account for roughly twenty-one percent
of the transactional volume in the market. In aggregate, virtual trading generated net profits
for most participants. However, twenty-eight virtual participants lost money on a net profit
basis. The total losses after fees amounted to roughly $3.2 million, and five entities
accounted for over $2.1 million of that loss. However, total net trading profits exceeded total
net trading losses by almost seven-fold.
Additionally, Figure 3-28 highlights the disparity in the trading fees paid by each market
participant. These fees totaled more than $26.2 million in 2018, they include: virtual
transaction fees (two percent), real-time revenue neutrality uplift fees (21 percent), day-ahead
make-whole payment fees (six percent), and real-time make-whole payment fees (71
percent). Virtual bids are subject to virtual transaction fees, real-time revenue neutrality fees,
and day-ahead make-whole payment fees. Whereas virtual offers are subject to virtual
transaction fees, real-time revenue neutrality fees, and real-time make-whole payment fees.
-1
0
1
2
3
4
5
6
7
-$1
$0
$1
$2
$3
$4
$5
$6
$7
Cle
ared
vir
tual
GW
h (m
illio
ns)
Vir
tual
+p
rofit
/-lo
ss ($
mill
ions
)
Cleared virtual megawatts Virtual +profit/-loss before fees Virtual +profit/-loss after fees
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 103
Nearly three quarters of the total fees assessed to virtual transactions are assessed only to
virtual offers.
The discrepancy in virtual fees relates to the quantity calculation associated with payers of
real-time make-whole payments – specifically, the real-time net settlement location deviation
hourly amount. This determinant accounted for 78 percent of the real-time make-whole
payments in 2018, or roughly $35 million. Virtual offers paid 53 percent of this amount,
nearly $19 million. As the name implies, the quantity applied to applicable non-virtual
transactions includes only the incremental deviations from day-ahead, however the quantities
assessed to virtual offers include the full virtual offer quantity.
This calculation methodology when combined with the larger make-whole payments
normally associated with real-time, generally leads to higher fees associated with offers when
compared to virtual bids. In 2018, the fees associated with virtual offers amounted to $1.11
per megawatt compared to $0.25 per megawatt for virtual bids. This calculation
methodology and associated incentives could be part of the reason why virtual trading
offsets only part of the under scheduling of wind resources in the day-ahead market. The
market monitor will continue to evaluate these trends going forward.
Cross-product market manipulation has been a concern in other RTO/ISO markets, and
extensive monitoring is in place to detect potential cases in the SPP market. For example, a
market participant may submit a virtual transaction intended to create congestion that
benefits a transmission congestion right position. Generally, this behavior shows up as a loss
in one market, such as a virtual position, and a substantial associated benefit in another
market, such as a transmission congestion right position. In the SPP market, twelve market
participants lost more than $10,000 in 2018, which is only slightly more than in 2017.
Southwest Power Pool, Inc. Unit commitment and dispatch processes Market Monitoring Unit
State of the Market 2018 104
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 105
4 PRICES
This chapter covers prices in the SPP market, along with related metrics on negative prices,
make-whole payments, and long-run price signals for investment. Highlights of this chapter
include:
• Day-ahead market prices averaged $25.08/MWh in 2018, up about nine percent from
2017. The average real-time price for 2018 was $24.58/MWh, an increase of five
percent over 2017. Although gas prices declined from 2017 to 2018, higher loads in
2018 and declines in the frequency of negative price intervals contributed to higher
average prices in 2018.
• SPP hub prices converged in 2018, with only a $1/MWh difference between the North
and South hub prices, compared to a historical difference of around $5/MWh. This is
primarily attributed to reductions in congestion due to transmission expansion, as well
as higher prices in the northern portion of the SPP footprint.
• Price divergence (absolute) between the day-ahead and real-time markets has
increased by nearly 60 percent from 2016 to 2018, climbing from $5.56/MWh to
$9.42/MWh. Analysis has found that underscheduling of renewable resources and
short-term ramping limitations are drivers of this divergence.
• The frequency of negative price intervals declined in 2018 from 2017. However, the
MMU remains concerned about the frequency of negative price intervals. Negative
prices may not be a problem in and of themselves, however, they do indicate an
increase in surplus energy on the system.
• Sixteen resources received over $1 million in make-whole payments in 2018, up from
the twelve resources in 2017.
• One resource received over $3.6 million dollars in make-whole payments. The
resource was mainly used for transmission and voltage support.
• Roughly seventy-eight percent of real-time make-whole payments were paid for by
resources that withdrew more in real-time than their day-ahead market cleared
megawatts. Virtual offers were the largest source of payments, paying for forty-one
percent of the 2018 real-time make-whole payments.
• Revenues have been insufficient to support the cost of new entry of scrubbed coal,
advanced combined-cycle, and advanced combustion turbine generation since the
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 106
inception of the Integrated Marketplace, and 2018 was no exception. From 2015
through 2018, prices did support the going forward costs of combined-cycle and
combustion turbine units, though they did not support the cost of scrubbed coal units.
4.1 MARKET PRICES AND COSTS
This section reviews market prices and costs by focusing on the fuel prices, price volatility,
negative prices, operating reserve prices, and market settlement results including make-
whole payments. Overall, annual energy prices remained stable compared to previous years
with just a slight increase in both the day-ahead and real-time prices in 2018. However,
annual numbers may mask underlying issues related to market flexibility and efficiency. For
instance, we discuss increasing periods of price volatility and instances of negative prices
later in this chapter.
4.1.1 ENERGY MARKET PRICES AND FUEL PRICES
Figure 4—1 below compares day-ahead and real-time prices in SPP between 2007 and 201850
with natural gas prices.
Figure 4—1 Energy price versus natural gas price, annual
50 From 2007 to 2013, the average price from the Energy Imbalance Service market is shown. The 2014 real-time average includes two months of prices from the Energy Imbalance Service market and 10 months of prices from the Integrated Marketplace.
$0
$2
$4
$6
$8
$0
$20
$40
$60
$80
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
$/M
MB
tu
$/M
Wh
Real-time price Day-ahead price Panhandle Eastern Pipeline gas cost
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 107
Historically, electric market prices have followed the cost of natural gas. As natural gas prices
have remained low overall, so have SPP market prices. The average gas cost, using the price
at the Panhandle Eastern Pipeline (PEPL), decreased by two percent from 2017 to 2018.
Although electricity prices remained low in 2018, they were higher in comparison to previous
years. Day-ahead market prices averaged $25.08/MWh in 2018, up about nine percent from
2017. The average real-time price for 2018 was $24.58/MWh, an increase of five percent
over 2017. Although gas prices declined from 2017 to 2018, higher loads in 2018 driven by
weather, as discussed in Chapter 2, and declines in the frequency of negative prices
contributed to higher prices in 2018.
Figure 4—2 illustrates day-ahead energy prices, as well as gas costs, on a monthly basis for
2018, along with an annual comparison for the past three years.
Figure 4—2 Day-ahead energy price versus natural gas cost
On a monthly basis, natural gas prices have averaged around $2.50/MMBtu at the Panhandle
Eastern hub. The last three months of 2018 saw a steep climb in gas prices, peaking at
$3.67/MMBtu in November. Day-ahead prices were highest in November at around
$33/MWh, and real-time prices were highest in October at about $31/MWh. This is atypical
as prices in the SPP market tend to peak in the summer months, as temperatures increase
and loads peak. Gas prices were highest in November and contributed to the highest
electricity prices in November. Prices were lowest in March, with both day-ahead and real-
time electricity prices both below $19/MWh, as periods of high wind generation coincided
with low loads and low gas prices.
$0
$1
$2
$3
$4
$0
$10
$20
$30
$40
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 18
$/M
MB
tuDay-ahead price Panhandle Eastern Pipeline gas cost
$/M
Wh
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 108
Changes in gas prices have historically had the highest impact on electricity prices compared
to other fuels. This is because the short-run marginal costs of coal fired generation
historically have been cheaper than natural gas-fired generation. However, as natural gas
prices have fallen, the short-run marginal costs of natural gas fired generation have been
more competitive with coal fired generation, and in some instances displacing coal
generation. Figure 4—3 compares various fuel price indices with real-time prices.
Figure 4—3 Fuel price indices and wholesale power prices
This figure shows that regional natural gas prices trended up from 2016 to 2018, following
the national trend as represented by the Henry Hub, especially during the winter months
(November to December of 2018). The Henry Hub gas price averaged $3.17/MMBtu for
2018, while both Southern Star and Panhandle Eastern were around $2.60/MMBtu in 2018.51
The difference between Henry Hub prices and Panhandle Eastern and Southern Star prices
has grown over the last few years, with differences averaging $0.22/MMBtu in 2016,
$0.33/MMBtu in 2017, and $0.57/MMBtu in 2018. This difference is likely driven by pipeline
constraints in the Texas and Oklahoma area. Often, natural gas is a byproduct of oil drilling.
51 The relevant natural gas prices for the SPP market are those of the Henry Hub, the Panhandle Eastern Pipeline (PEPL), and Southern Star. These prices do not include transport costs.
$0
$10
$20
$30
$40
$0
$1
$2
$3
$4
$5
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
$/M
Wh
$/M
MB
tu
Real-time LMP Henry Hub Panhandle EasternSouthern Star Wyoming 8400 0.80 Rail Wyoming 8800 0.80 Rail
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 109
Natural gas production has continued to outpace takeaway capacity in this area, with
incremental production volumes quickly inundating any available space in the pipelines and
keeping supply-area prices at discounts compared to other trade hubs. Natural gas prices in
West Texas (Waha and El Paso) have been less than $1/MMBtu on a number of days, which
have helped high heat rate units be profitable.
Coal prices have remained relatively stable since 2016 with a slight increase overall from
2017 to 2018.52 Annual average coal prices at Powder River Basin for both types, 8,400
Btu/lb., and 8,800 Btu/lb., increased by about five percent. The price for 8,400 Btu/lb.
increased from $0.53/MMBtu in 2017 to $0.55/MMBtu in 2018, and the price for 8,800 Btu/lb.
increased from $0.67/MMBtu in 2017 to $0.71/MMBtu in 2018.
Controlling for changes in fuel prices helps to identify the underlying changes in electricity
prices from other factors.53 Figure 4—4 below adjusts the marginal energy cost for changes in
fuel costs.54
52 Platts’ coal prices are exclusive of transport costs. Transportation costs can have a significant impact on a coal resource’s short-run marginal costs, and may often exceed commodity costs 53 In addition to fuel, other variables also affect real-time prices. These variables include seasonal load levels, transmission congestion, scarcity pricing, and wind-powered generation. 54 The marginal energy component (MEC) indicates the system-wide marginal cost of energy (excluding congestion and losses). Fluctuations in marginal fuel prices can obscure the underlying trends and performance of the electricity markets. Fuel price-adjusted marginal energy costs is a metric to estimate the price effects of factors other than the change in fuel prices, such as changes in load or changes in supply. It is based on the marginal fuel in each real-time five-minute interval, when indexed to the three-year average of the price of the marginal fuel during the interval. If multiple fuels were marginal in an interval, weighted average marginal energy costs are based on the dispatched energy of different fuel types.
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 110
Figure 4—4 Fuel-adjusted marginal energy cost
As the figure shows, fuel-adjusted marginal energy costs were lower in 2018 compared to
nominal marginal energy costs55 both annually and in all months, except for May. While the
average nominal marginal energy cost in 2018 increased 15 percent from 2017, the fuel-
adjusted marginal energy cost also increased by 13 percent. Looking by month, the highest
nominal and fuel-adjusted marginal energy cost occurred in October. SPP experienced 26
price spikes in the $500/MWh to $1,000/MWh range in October. This was over three times
the historical monthly average for that price range. One of the main causes appears to be an
increase in mid-term and long-term wind forecast errors. When large wind dips are not
accurately forecasted, SPP operators may take manual action to bring more expensive
capacity on-line.
The largest differences between nominal and fuel-adjusted prices occurred in October,
November and December. During this period, there was a large jump in natural gas spot
prices. For instance, the spot index price for natural gas at the Panhandle hub increased 72
percent from September to November from $2.13/MMBtu to $3.67/MMBtu. The SPP market
normally experiences more units operating on oil and higher priced fuel during periods of
winter market conditions and limitations on natural gas pipelines.
55 Nominal marginal energy costs represent the non-fuel adjusted marginal energy costs.
$15
$20
$25
$30
$35
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
$/M
Wh
Average MEC Fuel-adjusted MEC
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 111
4.1.2 HUB PRICES
SPP has two hubs: the SPP North hub and the SPP South hub. The SPP North hub represents
pricing nodes in the northern part of the SPP footprint, generally in Nebraska. The SPP South
hub represents pricing nodes in the south-central portion of the footprint, generally in central
Oklahoma. Typically, the SPP South hub prices exceed the SPP North hub prices. This was
again true in 2018, but to a lesser degree than in previous years. The general pattern of
higher prices in the south and lower in the north is primarily due to fuel mix and congestion.
Coal, nuclear, and wind are the dominant fuels in the north and west. Gas generation
represents a much larger share of the fuel mix in the south and east.
Figure 4—5 shows the annual average day-ahead and real-time energy prices at the two SPP
market hubs.
Figure 4—5 Hub prices, annual average
On an annual basis, hub prices have converged with North hub prices averaging around
$24/MWh, and South hub prices averaging around $25/MWh for 2018. Historically, the
South hub price has been about $5/MWh higher than the North hub price. This reduction in
price spread can be attributed to reductions in congestion, which was primarily a result of the
addition of the second circuit of the Woodward to Mathewson 345kV line in early 2018 (see
Section 5.1.3 for further detail), as well as higher prices in the northern portion of the SPP
footprint.
Hub prices are shown on a monthly basis for 2018 in Figure 4—6 and Figure 4—7.
$0
$10
$20
$30
2016 2017 2018
$/M
Wh
North day-ahead North real-time South day-ahead South real-time
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 112
Figure 4—6 Hub prices, day-ahead, monthly
Figure 4—7 Hub prices, real-time, monthly
Since the beginning of the Integrated Marketplace in 2014, July 2017 was the first month
where the North hub real-time price exceeded the South hub real-time price. The North hub
again exceeded the South hub in the months of March, April, June, and August 2018. South
hub and North hub prices were nearly identical, particularly in the summer months, as
congestion that separated the regional prices declined.
Average on-peak, day-ahead prices for the SPP hubs, as well as other RTO hubs in the region
are shown in Figure 4—8.
$10
$20
$30
$40
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
$/M
Wh
North day-ahead South day-ahead
$10
$20
$30
$40
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
$/M
Wh
North real-time South real-time
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 113
Figure 4—8 Comparison of RTO/ISO average on-peak, day-ahead prices 2016 2017 2018
SPP North hub $ 24 $ 25 $ 30
SPP South hub $ 29 $ 31 $ 30
ERCOT North hub $ 26 $ 26 $ 42
ERCOT West hub $ 26 $ 26 $ 39
MISO Arkansas hub $ 27 $ 31 $ 34
MISO Louisiana hub $ 34 $ 34 $ 44
MISO Minnesota hub $ 25 $ 28 $ 32
MISO Texas hub $ 31 $ 38 $ 36
PJM West hub $ 35 $ 34 $ 42
Just as the average on-peak, day-ahead price at the SPP North hub increased from 2017 to
2018, all of the other hubs studied increased in price with the exception of the SPP South hub
and the MISO Texas hub. The average on-peak day-ahead SPP South and North hub prices
were similar as transmission expansion reduced congestion. The ERCOT hub price increases
were rather significant, increasing by over 50 percent from 2017 to 2018. These price
increases were likely driven by tight capacity conditions, particularly during the summer
months.
4.1.3 ENERGY PRICE VOLATILITY
Price volatility56,57 in the SPP market is shown in Figure 4—9 below. As expected, day-ahead
prices are much less volatile than those in real-time. The day-ahead market does not
experience the actual (unexpected) congestion and changes in load or generation found in
the real-time market. Real-time volatility tends to peak in the spring and fall, roughly
corresponding with times of higher wind and lower load, but can also peak during the
summer months because of peak load conditions.
56 Volatility is calculated as the standard deviation for load-serving entities in the SPP market. The standard deviation is calculated using hourly price in the day-ahead market and interval (five minute) price in the real-time market. 57 A measure of volatility is also shown earlier in this report at Figure 3—19. That volatility calculation is based on the interval-to-interval change in marginal energy cost. The volatility calculation in Figure 4-9 is based on the interval locational marginal prices for load-serving entities in the SPP market. Although the results are different, the magnitude of the annual change is very similar.
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 114
Figure 4—9 System price volatility
Volatility in the day-ahead market has climbed steadily from 2016 to 2018, while real-time
volatility decreased from 2017 to 2018. Even so, real-time price volatility in 2018 was almost
50 percent higher than in 2016. The high real-time volatility in 2017 can mostly be attributed
to high levels of wind generation in the spring and fall of that year. High real-time market
volatility during October and November 2018 can mostly be attributed to several short-lived
scarcity pricing events, as well as fluctuations in wind output.
Price volatility varies across the SPP market footprint for asset owners primarily because of
congestion on the system, which is based on the layout of the transmission system and the
distribution of the types of generation in the fleet. The volatility for the majority of asset
owners is grouped very closely to the SPP average in both the day-ahead and real-time
markets as shown in Figure 4—10.
0
20
40
60
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Vo
latil
ity (s
tand
ard
dev
iatio
n)
real-time volatility day-ahead volatility
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 115
Figure 4—10 Price volatility by market participant
Similar to 2017, the area of the footprint that experienced the most volatility in 2018 was in
southwest Missouri. This area was impacted by external flows and congestion during the
year, mostly on the Neosho-Riverton flowgate, which is discussed in more detail in Section
5.1.4.3.
4.1.4 DAY-AHEAD AND REAL-TIME PRICE CONVERGENCE
Price convergence between day-ahead and real-time prices is important, because the more
day-ahead prices reflect real-time prices, the better unit commitment and positioning of
resources occurs for real-time operations. Figure 4—11 shows day-ahead and real-time prices
monthly for 2018 and annually for the past three years.
12.67
44.19
0
10
20
30
40
50
60
70V
ola
tility
(sta
ndar
d d
evia
tion)
Day-ahead volatility SPP day-ahead average Real-time volatility SPP real-time average
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 116
Figure 4—11 Day-ahead and real-time prices
Figure 4—12 below shows the monthly difference between day-ahead and real-time prices for
2018, as well as annually for the past three years.
Figure 4—12 Difference between day-ahead and real-time prices
Day-ahead prices exceeded real-time prices in February, July, September, and October.
February had the highest spread of real-time prices above day-ahead at $4.50/MWh, while
November had the highest spread of day-ahead prices over real-time at $3.00/MWh. On an
annual basis, the day-ahead premium has returned in 2018, with day-ahead average energy
$0
$10
$20
$30
$40
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 18
Pric
e ($
/MW
h)
Day-ahead price Real-time price
–$5
–$4
–$3
–$2
–$1
+$0
+$1
+$2
+$3
+$4
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 18
Day
-ahe
ad p
rice
–re
al-t
ime
pri
ce
($/M
Wh)
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 117
price fifty cents higher than real-time. Going back to the beginning of the Integrated
Marketplace in 2014, the only year that had an annual premium for real-time prices was 2017.
Real-time price volatility drove the annual real-time price premium in 2017.
While average prices in the day-ahead and real-time markets have been close over the past
several years, average prices can mask real-time volatility and underlying price differences.
In 2018, real-time system volatility was down from 2017, but was still up nearly 50 percent
from 2016. These metrics indicate that while average prices are similar, the underlying prices
in the day-ahead and real-time markets were different. The averaging of price spikes, and in
particular, high prices during periods of scarcity, drove real-time average prices up, closer to
day-ahead prices. We attribute these short-term, transient price spikes with limitations in
ramping capability.58
In this section, we highlight underlying differences in prices after controlling for scarcity
events. This analysis shows that a significant volume of generation, particularly from wind
resources not accounted for in the day-ahead market, drives down real-time prices.
There are many factors that cause prices to diverge between the day-ahead and real-time
markets. Some of those factors may include, but are not limited to:
• Day-ahead offers may include premiums to account for uncertainty in real-time fuel
prices.59
• Load and wind forecast errors can cause differences in the real-time market results.
• Participants may not offer in all of their load or generation in the day-ahead market.
• Modeling differences including transmission outages between the two markets.
• Generation outages or derates that were different in real-time than was anticipated in
the day-ahead.
58 For further information on ramping issues, see Section 3.3.1. 59 Additionally, revision request 239 allowed historic fuel cost uncertainty to be considered in the development of mitigated energy offers.
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 118
• Impacts from other RTOs, that were not anticipated, affect the SPP real-time market.
• Changes in imports and exports from other systems in the real-time markets.
• Unanticipated weather changes affect the real-time markets.
Price divergence60 between the day-ahead and real-time markets at the system level is shown
in Figure 4—13 below. Market participants may be willing to pay a premium for more
certainty in prices. This can result in higher prices in the day-ahead market. A large
divergence between day-ahead and real-time prices may also indicate that actual conditions
in the market do not match expected conditions. An extended period of a large variance
between day-ahead and real-time prices may indicate a structural or design deficiency in the
market.
Figure 4—13 Price divergence
The absolute price divergence has increased by nearly 60 percent from 2016 to 2018,
climbing from $5.56/MWh to $9.42/MWh. Analysis by the MMU has found that
underscheduling of renewable resources and short-term ramping limitations are drivers of
this divergence. The MMU made recommendations in the 2017 Annual State of the Market
report to address the underscheduling of renewable resources and to implement a ramping
60 Price divergence is calculated as the difference between day-ahead and real-time prices, using system prices for each five-minute (real-time) or hour (day-ahead) interval. The absolute divergence is calculated by taking the absolute value of the divergence for each interval.
-30%
-20%
-10%
0%
10%
20%
-$30
-$20
-$10
$0
$10
$20
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Div
erg
ence
per
cent
age
Div
erg
ence
($/M
Wh)
Divergence (day-ahead – real-time) Divergence (absolute) Divergence percentage
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 119
product. The MMU anticipates that addressing these two recommendations would improve
absolute price divergence.
Figure 4—14, below, shows the marginal energy costs for both the day-ahead and real-time
markets during on-peak hours after controlling for scarcity events.61 Figure 4—15 shows the
same information, but for off-peak hours.
Figure 4—14 On-peak marginal energy prices, excluding scarcity hours
61 These numbers reflect only hours where scarcity demand curves where not applied for any interval during the hour. SPP uses scarcity demand curves for intervals when ramp or capacity requirements cannot be met through dispatch. Scarcity demand curves are discussed in detail in Section 4.2, below.
$0
$10
$20
$30
$40
$50
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
$/M
Wh
On-peak real-time marginal energy cost On-peak day-ahead marginal energy cost
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 120
Figure 4—15 Off-peak marginal energy prices, excluding scarcity hours
The marginal energy cost is one of three components that factor into location marginal prices
and represents the marginal cost to provide the next increment of dispatch absent losses and
congestion. Both charts clearly show that day-ahead prices are usually at a premium when
compared to real-time prices when controlling for scarcity, particularly in the off-peak hours.
In 2018, day-ahead marginal energy costs, for all hours, were just over three percent higher
than real-time prices.62 This is much lower than the ten percent price divergence in 2017 and
the twelve percent in 2016. The majority of the decrease can be attributed to the high
volume of price spikes seen in the final quarter of 2018.
The main contributors influencing the price differences are offered MW versus cleared MW of
wind resources in the day-ahead market, self-committing of units after the day-ahead market,
and economic reliability unit commitments. In fact, only 85 percent of the wind generation
was cleared in the 2018 day-ahead market, up three percent from 2017. This changes the
62 The MMU observed that 77 percent of the hours in 2017 had higher marginal energy cost in the day-ahead market than the real-time market. This is after removing any hours associated with scarcity pricing.
$0
$10
$20
$30
$40
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
$/M
Wh
Off-peak real-time marginal energy cost Off-peak day-ahead marginal energy cost
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 121
supply curve in real-time by shifting it outward and causes real-time prices to drop relative to
the day-ahead market. Furthermore, the market appropriately honors the minimum
submitted limits of all committed resources. With the unanticipated generation, many non-
wind units are dispatched down by the market to their minimum capacity limits, allowing
wind to set prices. When this happens prices often go negative as the energy offers for wind
units are typically negative to account for production tax credits.63
Figure 4—16 shows average hourly incremental differences in megawatts produced between
the real-time and day-ahead market in 2018.
Figure 4—16 Average hourly real-time generation incremental to day-ahead market
Wind generation had just under two-thirds of the 1,800 MW of incremental real-time
generation in 2018, with an hourly average of 1,118 MW of additional generation in real-time.
Self-committed generation accounted for an additional 255 megawatts and reliability unit
committed generation averaged about 275 MW. On average, SPP is a net exporter in both
the day-ahead and real-time markets and sees an average hourly decrease of 132 MW in the
real-time market compared to the day-ahead. This results in additional capacity committed in
day-ahead not needed in real-time. Averaging 490 MW, net virtual positions helped to offset
the additional generation, but only accounted for about 27 percent of the difference for the
year, which is down from about 650 MW (34 percent) in 2017.
63 Negative prices are discussed in detail in Section 4.1.5.
Wind
Net virtual supply (day-ahead)
RUC/manual committed
Self-committed after day-ahead market
Import increase
0
400
800
1,200
1,600
2,000
Additional energy Reductions
MW
h
Net additional real-time supply
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 122
4.1.5 NEGATIVE PRICES
With the prolific growth of wind generation in the SPP market, the incidence of intervals with
negative prices continues to be a concern. However, the frequency of negative price
intervals decreased from 2017 to 2018, as shown in Figure 4-17. This decline was primarily a
result of congestion reductions as a result of transmission expansion (see Section 5.1.3).
Higher loads were also a factor in lowering the frequency of negative priced intervals (see
2.2.3.
In 2018, less than one percent of all asset owner intervals64 in the day-ahead market had
prices below zero, as shown in Figure 4—17. This is down from three percent of all intervals in
2017.
Figure 4—17 Negative price intervals, day-ahead, monthly
While the same pattern holds in the real-time market (see Figure 4—18), negative price
intervals in the real-time market occurred almost five times more frequently than in the day-
ahead market.
64 Asset owner intervals are calculated as the number of asset owners serving load that are active in an interval. For example, if there 60 asset owners active in one five-minute interval throughout an entire 30 day month, the total asset owner intervals would be 518,400 for the month (60 asset owners * 288 intervals per day * 30 days).
0%
3%
6%
9%
12%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Perc
ent
neg
ativ
e p
rice
inte
rval
s
LMP −$0.01 to −$25.00 LMP −$25.01 to −$50.00 LMP less than −$50.00
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 123
Figure 4—18 Negative price intervals, real-time, monthly
The frequency of negative price intervals in the real-time market fell to just over three percent
of 2018 intervals, down from just over seven percent in 2017. Negative prices in the day-
ahead market were almost exclusively between −$0.01/MWh and −$25/MWh, with only two
percent of intervals with negative prices having prices lower than −$25/MW. However, in the
real-time market 15 percent of intervals with negative prices had prices lower than
−$25/MWh.
Additionally, occurrences of negative prices in the day-ahead market are most prevalent in
the overnight, low-load hours as shown in Figure 4—19.
0%
3%
6%
9%
12%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Perc
ent
neg
ativ
e p
rice
inte
rval
s
LMP −$0.01 to −$25.00 LMP −$25.01 to −$50.00 LMP less than −$50.00
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 124
Figure 4—19 Negative price intervals, day-ahead, by hour
This figure shows that the day-ahead negative price intervals in 2018 during overnight hours
are below both 2016 and 2017 levels. Also of note, during the on-peak hours, less than 0.2
percent of intervals in the day-ahead market were negative in 2018.
Negative price intervals in the real-time market (see Figure 4—20) follow the same pattern as
the day-ahead market with most negative price intervals occurring in the overnight, low-load
hours.
Figure 4—20 Negative price intervals, real time, by hour
0%
5%
10%
15%
20%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Perc
ent
neg
ativ
e p
rice
inte
rval
s
Hour ending2016 2017 2018
0%
5%
10%
15%
20%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Perc
ent
neg
ativ
e p
rice
inte
rval
s
Hour ending2016 2017 2018
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 125
These negative price intervals in the real-time market occur much more frequently than the
day-ahead market, with a 2018 peak of nine percent of intervals in real-time in the last hour of
the day, compared to a peak of just over 2 percent in day-ahead. During 2017, the first five
hours and last two hours of the day experienced negative prices in over 10 percent of all
intervals. However, in 2018 no intervals had negative prices over 10 percent of the time. In
2018, the real-time market had an average of about 3.5 percent of intervals with negative
prices during the on-peak hours.
At the asset owner level (for those serving load), the distribution of negative price intervals
during 2018 clustered around the footprint average, as shown in Figure 4—21.
Figure 4—21 Negative price intervals, real-time, by asset owner
In 2018, only one asset owner experienced negative prices in more than 10 percent of
intervals. This is in contrast to 2017, when four asset owners experienced negative prices in
excess of 15 percent of intervals. On the low end, all but six asset owners experienced
negative prices in fewer than five percent of all intervals. Contrasting to 2017, only 18 asset
owners experienced negative prices in fewer than five percent of intervals.
While the frequency of negative price intervals declined in 2018, the MMU remains
concerned about the frequency of negative price intervals. Negative prices may not be a
problem in and of themselves, however, they do indicate an increase in surplus energy on the
system. This may be exacerbated by the practice of self-committing after the day-ahead
market. In the SPP market where there is an abundance of capacity and significant levels of
0%
3%
6%
9%
12%
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61
Perc
ent
neg
ativ
e p
rice
inte
rval
s
Market participants (high to low)
LMP -$0.01 to -$25.00 LMP -$25.01 to -$50.00 LMP less than -$50.00 average
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 126
renewable resources, negative prices can occur when renewable resources need to be
backed down in order for traditional resources to meet their committed generation.
Moreover, unit commitment differences, due to wind resources not being in the day-ahead
market and then coming on-line for the real-time market, can create differences in the
frequency of negative price intervals between the day-ahead and real-time markets. This
disparity between the markets negatively impacts the efficient commitment of resources.
As more wind generation is anticipated to be added over the next several years, the
frequency of negative prices has the potential to increase. Negative price intervals highlight
the need for changes in market rules to address self-committing of resources in the day-
ahead market and the systematic absence of some forecasted variable energy resources in
the day-ahead market to improve market efficiency. These issues are discussed further in
Chapter 8.
4.1.6 OPERATING RESERVE MARKET PRICES
Operating reserve is made up of four products: (1) regulation-up, (2) regulation-down, (3)
spinning reserve, and (4) supplemental reserve. The regulation products are used to ensure
the amount of generation matches load on a subinterval basis. Generators respond to
regulation instructions in seconds. Spinning and supplemental products are reserved for
contingency situations and respond to instructions within ten minutes.
Average monthly real-time prices for operating reserve products are presented in Figure 4—
22.
Figure 4—22 Operating reserve product prices, real-time
$0
$4
$8
$12
$16
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Mar
ket c
lear
ing
pri
ce ($
/MW
)
Regulation up Regulation down Spinning reserves Supplemental reserves
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 127
Generally speaking, regulation-up and regulation-down usually have the highest market
clearing prices. Supplemental reserves always have the lowest average prices of the
operating reserve products, with prices averaging less than two dollars on an annual basis.
Spinning reserve prices typically fall between regulation-up and supplemental reserve prices.
However, 2018 saw three months with spinning reserve prices above regulation-down prices.
Day-ahead and real-time price patterns vary across the operating reserve products, see
Figure 4—23 through Figure 4-26.
Figure 4—23 Regulation-up service prices
From 2017 to 2018, the average real-time market clearing price for regulation-up decreased
from $10/MW to $9/MW. Day-ahead regulation-up market clearing price was virtually
unchanged from 2017 to 2018 at $8.50/MW. Monthly prices for regulation-up were highest
in the peak wind months during the spring and especially in the fall. The high prices during
these periods can mostly be attributed to higher wind penetration levels during these
periods.
$0
$4
$8
$12
$16
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Mar
ket c
lear
ing
pri
ce ($
/MW
)
day-ahead real-time
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 128
Figure 4—24 Regulation-down service prices
Regulation-down market clearing price in the real-time market averaged about $6.40/MW in
2018 after peaking in 2017 at nearly $10/MW. The higher 2017 prices can generally be
attributed to the high levels of wind production in that year, particularly in high wind and low
load months. During these months, many thermal units were operating at their economic
minimums, which are lower than regulation minimums. Costs are higher to move these units
up to the regulation range. Day-ahead regulation-down market clearing prices also have
returned to 2016 levels at around $4.70/MW after peaking in 2017 at about $6.60/MW.
Figure 4—25 Spinning reserve prices
$0
$4
$8
$12
$16
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Mar
ket c
lear
ing
pri
ce ($
/MW
)
day-ahead real-time
$0
$4
$8
$12
$16
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Mar
ket c
lear
ing
pri
ce ($
/MW
)
day-ahead real-time
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 129
The market clearing price for real-time spinning reserves actually dropped slightly from 2017
to 2018 at just over $5/MW. However, October 2018 experienced the highest monthly
market clearing price for real-time spinning reserves, at just over $10/MW, since the market
began in 2014. This can be mostly attributed to the increase in scarcity events in that month.
Day-ahead spinning reserves averaged almost $5.50/MW for the year and peaking in the fall
months as natural-gas prices increased.
Figure 4—26 Supplemental reserve prices
Supplemental reserve market clearing prices remained low in both markets, with prices
averaging about $1/MW. On a monthly basis, the October average real-time market clearing
price for supplemental reserves was just over $3/MW. This is the highest price since May
2014.
In March 2015, SPP introduced a new product paying regulating units for mileage costs
incurred when moving from one set point instruction to another. These mileage payments
are paid directly through the operating reserve prices shown for regulation-up and
regulation-down, as shown in Figure 4—27. The market calculates a mileage factor for both
products each month that represents the percentage a unit is expected to be deployed
compared to what it cleared. If a unit is deployed more than the expected percentage, then
the unit is entitled to reimbursement for the excess at the regulation mileage marginal
clearing price. If the unit is deployed less, it must buy back its position.
$0
$4
$8
$12
$16
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Mar
ket c
lear
ing
pri
ce ($
/MW
)
day-ahead real-time
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 130
Figure 4—27 Regulation mileage prices, real-time
On an annual basis, average monthly regulation-up mileage prices for 2018 rose slightly from
2017 at around $10/MW. Average monthly regulation-down mileage prices in 2018 returned
to 2016 levels after spiking in 2017, almost exactly matching the pattern in regulation-down
prices in those years. The decline in 2018 mileage prices is a result of changes in bidding
patterns compared to fall 2017.
Overall, the cost of ancillary services in 2018 was $76 million, a decrease of three percent
from 2017. This can be primarily attributed to a slight decrease in ancillary service prices,
and a slight increase in reserve requirements.
The MMU analyzed regulation mileage prices in 2017 and found a design inefficiency.
Mileage prices are not set by the marginal resource’s cost like other products. Instead,
resources are cleared for regulation based on their service offers. These service offers are
derived by taking the competitive offer for regulation and adding the mileage offer to it after
discounting the mileage offer by the applicable mileage factor. For instance, the service offer
of a resource with a competitive regulation-down offer of $1 and a regulation-down mileage
offer of $36 would be $10 if the mileage factor is 25 percent.65 If the $10 service offer is
economic, then the resource will clear for regulation-down and the regulation-down mileage
price will be set at $36 if this is the highest mileage price that cleared in the market.
65 $1 + $36 * 0.25 percent = $10
$0
$4
$8
$12
$16
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Mar
ket c
lear
ing
pri
ce ($
/MW
)
Regulation-up mileage Regulation-down mileage
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 131
The MMU has observed instances where resources cleared with regulation-down competitive
offers of $0 and mileage offers just under $50. These units consistently cleared with this offer
strategy because the service offer was near $10.50 (e.g. 21 percent*$50) which was lower
than the services offers of other resources offering in higher competitive offers. For instance,
another resource may offer in a $12 competitive offer and $0 mileage offer. This would make
that resource’s service offer $12 ($12+0*21 percent). In this circumstance, the resource with
the highest service offer will set the regulation-down price at $12, but the mileage offer will
be $50, set by the highest cleared mileage offer.
In addition, the MMU observed systematic overpayment of regulation mileage in the day-
ahead market, which appears to be the result of the mileage factor being set consistently too
high relative to actual mileage deployed. This occurred because the mileage factor is being
set on historical instructed regulation megawatts rather than deployed regulation. When
resources have to buy back their position, they typically have to buy back at the inflated
mileage offer. Using the example above, if a resource clears for 10 megawatts it will receive
the $12 clearing price for a total payment of $120, which was set using a $0 mileage offer.
However, if it does not get deployed for regulation it will have to buy back 2.1 megawatts at
the $50 mileage offer, because they performed less than expected. The unit was paid 2.1
megawatts at a $0 price for expected mileage at the clearing, but the buyback is now $105.
This makes the total payment to the resource for clearing regulation $15 or $1.50 per cleared
megawatt.
The instructed values for regulation are on average two and a half times what resources
perform. If the mileage factor was forecasted in the exact amount of what was performed
then the excess mileage payments should closely offset the unused mileage charges.
However, this is not the case. The reason for the difference is that regulation is deployed on
a 4 second basis but it is settled on a five minute basis. Resources can be directed to move
up 10 megawatts during at the beginning of the interval. However, 20 seconds later they
may be directed to hold off on providing that regulation. If their ramp rate is only 10
megawatts per minute, they will only have provided 3.3 megawatts of regulation. This is
generally what causes the instructed values to vary from the actual values.
Figure 4—28 below illustrates the differences between the unused mileage charges and the
excess payments.
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 132
Figure 4—28 Regulation mileage payments and charges
Negative values represent the payments made to resources that deployed for more
regulation megawatts than were expected and positive values represent charges made to
resources that deployed for less than what was expected. In 2018, roughly twice as much
was charged for mileage buyback as was paid out for excess mileage deployment, with $4.2
million being paid for excess mileage and $8 million being charged for unused mileage. This
is consistent with all prior years.
The MMU is concerned that participants with resources frequently deployed for regulation
will have an incentive to inflate the mileage prices by offering in $0 regulation offers and high
mileage offers and screens for manipulative behavior. The MMU also has concerns that the
inflated mileage factors are causing units to buy back megawatts at the inflated amounts
which can ultimately lead to higher uplift costs in the market. As such, the MMU has
recommended that SPP review and revise the regulation mileage pricing approach to send
more appropriate price signals.
4.1.7 MARKET SETTLEMENT RESULTS
The day-ahead market accounted for 98 percent of the energy consumed in the Integrated
Marketplace. This is consistent with the last two years. Figure 4—29 shows that
approximately 264 terawatt-hours of energy were purchased in the day-ahead market at load
settlement locations, of which only six terawatt hours were in excess of the real-time
consumption, requiring a sale back to the market.
-$8
-$4
$0
$4
$8
$12
2016 2017 2018
$ m
illio
ns
Regulation-down excess mileage payment Regulation-down unused mileage chargesRegulation-up excess mileage payment Regulation-up unused mileage charges
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 133
Figure 4—29 Energy settlements, load
Negative gigawatt hours denote withdrawals from the grid. Negative cash flows denote
charges to load-serving entities. Positive gigawatt hours represent sales of day-ahead
gigawatt hours back to the real-time market and negative cash flows represent payments to
load owners for those sales. As stated above, during several hours of 2019 load over
purchased in the day-ahead market causing a sale back in real-time. However, there were
also several hours where the load consumed in real-time was greater than the day-ahead
cleared quantities. Just over five terawatt hours of energy was purchased in the real-time
market because the real-time demand was higher than that of the day-ahead market. The
close relationship of day-ahead load consumption to real-time load consumption is a sign of
an efficient day-ahead market. Theses charts highlight the increase in load in 2018, noted in
Chapter 2, as well as the corresponding increase in market prices as described above.
Day-ahead generation accounted for 90 percent of generation settled in the market, a one
percent increase from last year and a two percent decrease from 2016. Figure 4—30 presents
the settlement of SPP generators.
-50,000
0
50,000
100,000
150,000
200,000
250,000
300,000
2016 '17 '18
Load
(GW
h)
-$2,000
$0
$2,000
$4,000
$6,000
$8,000
2016 '17 '18
Cas
h flo
w ($
mill
ions
)
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 134
Figure 4—30 Energy settlements, generation
Positive gigawatt hours denote injections into the grid. Positive cash flows denote payments
to generators. Negative gigawatt hours represent repurchases in the real-time market and
negative cash flows represent charges to generators for those repurchases. Ten percent of
the energy cleared in the day-ahead market was settled by purchasing energy in the real-
time market rather than generating the energy, which was about one percent more than what
was experienced in the prior three years.
SPP plays the role of the customer in the operating reserve market. At hour ending 6:00 AM
on the day before the operating day, SPP posts the forecasted amount of each operating
reserve product that is to be procured. This data sets the demand for the products for the
day-ahead market. SPP can change the demand levels after the clearing of the day-ahead
market, but this a rare occurrence. Even though the demand is essentially the same between
the day-ahead market and the real-time market, there is considerable activity with respect to
the operating reserve products in the real-time market. Figure 4—31 presents the settlements
data for operating reserves.
-50,000
0
50,000
100,000
150,000
200,000
250,000
300,000
350,000
2016 '17 '18
Gen
erat
ion
(GW
h)
-$2,000
$0
$2,000
$4,000
$6,000
$8,000
2016 '17 '18
Cas
h flo
w ($
mill
ions
)
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 135
Figure 4—31 Operating reserve product settlements
Positive numbers represent the cleared gigawatt hours for ancillary services. Purchases
represent the repurchase of day-ahead cleared ancillary service in the real-time market.
A large percentage of day-ahead sales (35 percent) are settled in the real-time market by
repurchasing the operating reserve product rather than supplying the service in the real-time
market. This is in contrast to 10 percent of the real-time energy generation in excess of day-
ahead cleared generation, which is settled at real-time prices. This 2018 repurchase rate is
relatively the same as the prior years. However, the repurchase is down five percent from the
40 percent that occurred in the first 12 months of the market.
Sixty-six percent of the 2018 real-time regulation-up service was settled at day-ahead prices,
exactly the same as the previous year. The corresponding percentages for regulation-down
service, spinning reserves, and supplemental reserves are 63 percent, 69 percent, and 61
percent respectively. These results were similar with the respective numbers in 2017 of 63
percent, 70 percent, and 62 percent. This essentially means that operating reserve products
are being moved around to different resources, due to their day-ahead and real-time
clearing, in about the same volumes as last year. This causes some resources to buy back
their day-ahead cleared megawatts at real-time prices, while other resources clear those
some quantities of megawatts and get paid real -time prices for their clearing.
0
3,000
6,000
9,000
12,000
2016 '17 '18 2016 '17 '18 2016 '17 '18 2016 '17 '18
Regulation-up Regulation-down Spinning reserves Supplementalreserves
GW
Day-ahead sales Real-time sales Real-time purchases
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 136
4.2 MAKE-WHOLE PAYMENTS
The Integrated Marketplace provides make-whole payments (MWPs) to generators to ensure
that the market provides sufficient revenue to cover the cleared offers providing energy and
operating reserves for a period in which the resource was committed. To preserve the
incentive for a resource to meet its market commitment and dispatch instruction, market
payments should cover the sum of the incremental energy cost, start-up cost, no-load cost,
transition cost, and cost of operating reserve products. The make-whole payment provides
additional market payments in cases where revenue is below a resource’s offer to make the
resource whole to its offers of operating reserve products, incremental energy, start-up,
transition, and no-load.
For the resources that are not combined-cycle, settlements separately evaluate: (1) day-
ahead market commitments based on day-ahead market prices, cleared offers and dispatch;
and (2) reliability unit commitments based on real-time market prices, cleared offers, and
dispatch. Combined-cycle resources can be cleared in both the day-ahead and real-time
markets at the same time, which is unique to combined-cycles. As a result, settlements must
evaluate the revenues and cost of both real-time and day-ahead commitments when
calculating real-time make-whole payments for combined-cycles.
For 2018, day-ahead market and reliability unit commitment make-whole payments totaled
approximately $72 million, up six percent from $68 million last year. Make-whole payments
averaged about $0.26/MWh for 2018, which is the same as 2017. In comparison to other
RTO/ISO markets, SPP’s make-whole payments were in the range of uplift costs, which varied
from $0.08/MWh to $0.29/MWh in 2017 and 2018.66
Figure 4—32 shows monthly and annual day-ahead make-whole payment totals by
technology type. Figure 4—33 shows the same make-whole payment information for
reliability unit commitment.
66 2017 ISO NE State of Market Report https://www.potomaceconomics.com/wp-content/uploads/2018/06/ISO-NE-2017-EMM-SOM-Report_6-17-2018_Final.pdf, page 39; 2018 State of the Market Report for PJM http://www.monitoringanalytics.com/reports/Presentations/2019/IMM_MC_SOM_20190321.pdf, page 9.
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 137
Figure 4—32 Make-whole payments by fuel type, day-ahead
Figure 4—33 Make-whole payments by fuel type, reliability unit commitment
Day-ahead make-whole payments constituted 38 percent of the total make-whole payments
in 2018. Gas-fired resources represent about 81 percent of all make-whole payments, with
69 percent of all make-whole payments to simple-cycle gas resources through reliability unit
commitment make-whole payments.
$0
$2
$4
$6
$8
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Mill
ions
Coal Gas, combined-cycle Gas, simple-cycle Hydro Wind Other
$0
$10
$20
$30
$40
$50
2016 '17 '18
Mill
ions
$0
$2
$4
$6
$8
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Mill
ions
Coal Gas, combined-cycle Gas, simple-cycle Hydro Wind Other
$0
$10
$20
$30
$40
$50
2016 '17 '18
Mill
ions
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 138
Make-whole payments occur for several reasons, which include some of the following: local
reliability commitments, uncaptured congestion in the day-ahead market, inflexibility of
resources to move in economic ranges between on-peak and off-peak hours, and excessive
transmission congestion not being solved by the market. Make-whole payments to resources
in the “other” category primarily represent payments to oil-fired resources. The increase in
real-time make-whole payments in 2018 can primarily be attributed to more resources being
brought on from the reliability unit-commitment processes for capacity needs.
Figure 4—34 shows the share of each cause of make-whole payments in the real-time and
day-ahead markets.
Figure 4—34 Make-whole payments, commitment reasons Real-time commitment reason
2016 2017 2018
Intra-day RUC 24.4%% 35.7% 26.1% Manual, SPP capacity 2.6% 4.5% 24.9% Short-term RUC 12.9% 6.1% 13.7% Manual, voltage 8.0% 6.4% 4.4% Day-ahead RUC 20.1% 7.5% 1.8% Manual, stagger 0.0% 0.0% 0.9% Other 0.0% 0.8% 0.3% Manual, intra-day RUC 8.4% 6.1%
Manual, day-ahead RUC 0.3% 1.2%
Manual, off-line supplemental 0.1% 0.0% 0.0%
Day-ahead commitment reason 2016 2017 2018
Day-ahead market 73.2% 93.7% 96.9% Manual, voltage support 26.8% 6.3% 3.1%
Voltage support commitments decreased in 2018 when compared to prior years. There was
a large increase in percentage of real-time make-whole payments paid to resources running
on for capacity. In fact, just under 25 percent of the real-make-whole payments were paid to
resources committed manually for capacity needs, which occurred over five times more than
in 2017. Manual, intra-day and day-ahead, RUC were removed as an option in 2018, thus
make-whole payments that were attributed to those categories in prior years are now
captured in the Manual, SPP capacity category.
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 139
There was an initiative in late 2018 to clean up the categorizing of manual commitments.67
Prior to the initiative operators often categorized capacity commitments under the intra-day
RUC and day-ahead RUC manual codes. However, after the initiative the manual RUC codes
were removed, manual commitments for capacity are all coded under the SPP capacity code.
This explains about half of the increase in make-whole payments for resources committed for
capacity. The residual increase can most likely be attributed to the increase in wind volatility
(see Section 4.1.3) and resource outages in October and November (see Figure 3—14).
These led to the increased need for manual commitments as rampable capacity was often
scarce.
Make-whole payments associated with voltage support commitments do not follow the same
uplift process outlined in Section 4.3.1. Instead, the cost of these make-whole payments are
distributed to the settlement areas that benefited from the commitment by way of a load ratio
share. Figure 4—35 illustrates the level of make-whole payments associated with voltage
support commitments.
Figure 4—35 Make-whole payments for voltage support
The make-whole payments stemming from voltage support commitments are down 18
percent from 2017 and down 75 percent from 2016. The high make-whole payments in 2016
67 SPP MMU, Quarterly State of the Market report, Summer 2018 special issues section, page 49-6: https://www.spp.org/documents/58811/spp_mmu_qsom_summer_2018_final.pdf.
$0.0
$0.2
$0.4
$0.6
$0.8
$1.0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Mak
e-w
hole
pay
men
ts ($
/mill
ions
)
Day-ahead Real-time
$0
$3
$6
$9
$12
$15
2016 '17 '18
Mak
e-w
hole
pay
men
ts ($
/mill
ions
)
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 140
were directly attributed to some resource outages in the south central section of the SPP
footprint.
Most SPP resources received modest total annual make-whole payments as highlighted in
Figure 4—36.
Figure 4—36 Concentration of make-whole payments by resource
The majority of resources in SPP received less than $250,000 in make-whole payments in
2018. Seven resources received over $1 million, compared to eight resources with $1 million
make-whole payments in 2017. One resource received $3.7 million dollars in make-whole
payments, with $800 thousand of those dollars coming voltage support commitments. The
vast majority of the make-whole payments to this resource occurred because of manual
reliability commitments made to control a temporary flowgate in the area.
Figure 4—37 reveals some concentration in the market participants that received the highest
levels of make-whole payments. The overall numbers have not changed significantly over the
past few years.
$0
$1
$2
$3
$4
0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340
Tota
l mak
e-w
hole
pay
men
ts($
mill
ions
)
Resources receiving make-whole payments
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 141
Figure 4—37 Market participants receiving make-whole payments 2016 2017 2018
> $1 million
> $5 million
> $10 million
> $1 million
> $5 million
> $10 million
> $1 million
> $5 million
> $10 million
Market participants receiving make-whole payments
12 4 2 12 5 1 16 5 1
% of make-whole payments by category
92% 61% 38% 90% 61% 18% 94% 57% 14%
In 2018, there were 16 market participants that each received annual make-whole payments
in amounts greater than $1 million. These 16 market participants accounted for 94 percent of
the total make- whole payments paid out in 2018. Twelve of those same 16 participants also
received over $1 million each in 2017. All but one of those 12 participants received over a $1
million in make-whole payments for each of the last three years. In 2018, there were five
participants that received over $5 million each in make-whole payments and out of that one
received $10 million. The participant with $10 million in total make-whole payments
accounted for 14 percent of all make-whole payments paid out in 2018. This was down from
18 percent in 2017.
4.2.1 MAKE-WHOLE PAYMENT ALLOCATION
The allocation of both day-ahead and real-time make-whole payments has important
consequences to the market. In principle, for market efficiency purposes uplift cost allocation
should be directed to those members that contributed to the need for the make-whole
payments (i.e., cost causation).
For the day-ahead market, make-whole payment costs are distributed to both physical and
virtual withdrawals on a per-MWh rate. The per-MWh rate is derived by dividing the sum of
all day-ahead make-whole payments for an operating day by the sum of all cleared day-
ahead market load megawatts, export megawatts, and virtual bids for the operating day. The
average per-MWh rate for withdrawing locations in the day-ahead market was just over
$0.09/MWh in 2018. This is approximately one and a half cents less than the 2016 and 2017
averages.
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 142
For the real-time market, make-whole payment costs are distributed through a per-MWh rate
that is assigned to all megawatt-hours of deviation in the real-time market. The average real-
time distribution rate was $0.98/MWh for 2018, up $0.08/MWh from $0.90/MWh in 2017.
There are eight categories of deviation and each category receives an equal amount per
megawatt when the cost of make-whole payments is applied. This can be seen in the
settlement location deviation charge, show in Figure 4—38.
Figure 4—38 Make-whole payments by market uplift allocation, real-time
Uplift type Deviation
MWs (thousands)
Uplift charge
(thousands)
Share of MWP
charges
Cost per MW of
deviation Settlement location deviation
37,685 $ 35,373 77.8% $ 0.94
Outage deviation 4,306 $ 5,064 11.1% $ 1.18
Status deviation 1,616 $ 1,721 3.8% $ 1.07
Maximum limit deviation 1,478 $ 1,598 3.5% $ 1.08 Reliability unit commitment self-commit deviation
590 $ 635 1.4% $ 1.08
Uninstructed resource deviation
566 $ 533 1.2% $ 0.94
Reliability unit commitment deviation 283 $ 266 0.6% $ 0.94
Minimum limit deviation 190 $ 231 0.5% $ 1.21
Even though each category of deviation is applied the same rate for deviation, approximately
78 percent of the real-time make-whole payment costs were paid by entities withdrawing
(physical or virtual) more megawatts in the real-time market than the day-ahead market.
Transactions susceptible to this charge are virtual offer megawatts, real-time load megawatts
in excess of the day-cleared megawatts for a unit, exporting megawatts in real-time in excess
of the export megawatts cleared in the day-ahead market, and units pulling substation power
in excess of any megawatts produced by the unit. However, virtual offers are the most
susceptible as 100 percent of their megawatts are considered incremental. Due to this fact
virtual offers alone paid for forty-one percent of all real-time make-whole payments.
4.2.2 REGULATION MILEAGE MAKE-WHOLE PAYMENTS
In March 2015, SPP introduced regulation compensation changes for units deployed for
regulation-up and regulation-down. One component of the regulation compensation
charges is regulation-up and regulation-down mileage make-whole payments for units that
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 143
are charged for unused regulation-up or regulation-down mileage at a rate that is in excess of
the regulation-up or regulation-down mileage offer.
SPP calculates mileage factors monthly for both regulation-up and regulation-down. These
mileage factors are ratios of historical averages of the percentage of each regulation product
deployed to the regulation product cleared in the prior month. The regulation-up mileage
factor and regulation-down mileage factors averaged 15 percent and 25 percent,
respectively, for 2018. The regulation-down mileage factor is up one percent from last year
and the regulation-up mileage factor is down three percent.
The mileage factor is a key component in the computation of mileage make-whole payments.
When the mileage factor is greater than the percentage of deployed regulating megawatts to
cleared regulating megawatts for each product, the resource must buy back the non-
deployed megawatts at the mileage marginal clearing price for the respective product. If the
mileage marginal clearing price used for the buyback is greater than the unit’s cost for the
product a make-whole payment may be granted.
Figure 4—39, below, illustrates the mileage make-whole payments for 2018 and the prior two
years.
Figure 4—39 Regulation mileage make-whole payments
The large increase in regulation-down make-whole payments for the second half of 2018 can
be directly attributed to the mileage clearing prices for regulation-down being nearly
$36/MW for some intervals in the latter half of 2018. The same trend also happened in 2017,
$0
$40
$80
$120
$160
$200
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Mak
e-w
hole
pay
men
ts ($
/tho
usan
ds)
Regulation-up mileage MWP Regulation-down mileage MWP
$0
$200
$400
$600
$800
$1,000
2016 '17 '18
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 144
but the prices during that period got as high as $50/MW. The highest regulation-down
mileage prices happened in March 2016, when they got as high as $60/MW. The MMU
identified a design inefficiency in the regulation-mileage compensation process. Details of
this issue are described in section 4.1.6.
4.2.3 POTENTIAL FOR MANIPULATION OF MAKE-WHOLE PAYMENTS
The 2014 Annual State of the Market Report, the MMU highlighted specific vulnerabilities that
market participants could potentially manipulate in SPP’s make-whole payment provisions.
The below vulnerabilities were directly associated with the FERC order regarding the make-
whole payments and related bidding strategies of JP Morgan Ventures Energy Corp:68
1) make-whole payments for generators committed across the midnight hour,
2) make-whole payments for regulation deployment, and
3) make-whole payments for out-of-merit energy.
In each case, a market participant has the ability to position its resource to receive a make-
whole payment without economic evaluation of its offers by the market.
A revision request was passed by the SPP board during the July meeting.69 The revision
request corrects the issue one by limiting make-whole payments for any resource with multi-
day minimum run times to the lower of the market offer or the mitigated offer. This limitation
only applies to offers falling in hours in which the resource’s offers were not accessed by one
of the security constrained unit commitment (SCUC) processes and the resource bid at or
above their mitigated offer on the first day. Subsequent to board approval of the proposal,
SPP legal staff identified tariff language that requires modification as it conflicts with the
language passed in the proposal. This has required that additional tariff modifications go
through the stakeholder process.
The MMU strongly recommends that SPP and its stakeholders clean up the tariff language
necessary to implement the changes to address the gaming issue. Addressing this matter in
a timely manner is a high priority for the MMU.
68 See 144 FERC 61,068. 69 Revision request 306 (2014 ASOM MWP MMU Recommendation [3-Day Minimum Run Time])
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 145
Two revision requests were implemented in 2018 concerning regulation adjustments. One
adds an assessment of economics when deploying units for regulation and the other caps the
offers used for the adjustment. The MMU feels that these two revision requests adequately
close the gaming opportunities present in the regulation adjustment.70
4.2.4 JOINTLY-OWNED UNIT MAKE-WHOLE PAYMENTS
Another make-whole payment concern existed related to jointly-owned resources and the
combined resource option. At the time the MMU made their original recommendations, the
market committed jointly-owned units as one unit, dispatched each separate owner on a
percentage of ownership, and paid make-whole payments for energy based on the individual
owners’ energy offers. This allowed a shareowner to benefit from a higher energy offer than
its co-owners through high minimum energy costs in the make-whole payment.
In August 2017, SPP implemented changes based on two revision requests to eliminate the
gaming opportunities present in the original design. The new design eliminates the potential
gaming opportunity by taking all owners’ pricing points of a combined resource option for
jointly-owned resources and aggregates those price points into a single energy offer curve
for the unit. This revision request also enforced all shares under a combined resource option
to have a minimum capacity limit of zero megawatts when being assessed for dispatch. Start-
up and no-load costs were still reimbursed on the same percentage of ownership method
used prior to the changes.
With the new method, individual shares only received a dispatch instruction greater than zero
megawatts based on the new aggregated energy offer curve, which eliminated the original
gaming opportunity. However, while the change eliminated the gaming opportunity outlined
in the 2014 annual state of the market report, the design introduced new issues. There are
multiple issues:
• The first issue is related to how the new design allocated costs when these resources
were self-committed and uneconomic. During these uneconomic periods, the lowest
offers of the jointly-owned unit will be cleared to meet the unit’s minimum physical
capacity operating limit. The participants with the lowest offers would pay the costs of
running uneconomically.
70 The revision requests were 242 and 243.
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 146
• The second issue occurs when a jointly-owned unit clears in any process after the day-
ahead market—such as the reliability unit commitment process. Market participants
can inflate their offers at the time of dispatch and be made whole to that inflated offer,
even though the jointly-owned unit may be running at its physical minimum limit. This
is because each jointly-owned unit owners’ minimum operating limits are considered
zero megawatts for settlement even though the unit is only running above zero
megawatts because of the physical minimum limits of the unit.
• The third issue is with regards to real-time make-whole payment distributions. These
distributions are not accurately applied because the co-owners of jointly owned units
each have zero minimum capacity limits in settlements. They are allocated costs
incorrectly for real-time make-whole payments for the difference in the megawatts
between the minimum limit of the resource and the desired energy level.
• Additionally, there are other deviation calculations that are based on the difference of
a unit in manual status output versus where they would be desired to run if the
resource were dispatchable. When this happens there is a potential for the combined
unit’s desired energy to be lower than the physical minimum limit of the unit. This is
because the settlement process calculates the desired energy quantities with the
assumption that the unit can run at zero megawatts, which is below the physical
minimum of the unit. As a result, the jointly-owned unit asset owners could have
distribution charges unduly applied.
Both SPP stakeholders and FERC approved a new design in 2018.71 The new design is
expected to be implemented after implementation of the new settlement system, which is
currently expected to take place in early 2020. The new design requires that jointly-owned
units offer in as one unit. The market system will dispatch the resource as one unit, and then
the settlements process will allocate the cost and revenues out by percentage of ownership
of the resource. The new design will address the concerns outlined above.
71 [Revision Request 266-JOU Combined Single Resource Modeling post Settlement Share Allocation, https://www.spp.org/spp-documents-filings/?id=21069..]
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 147
4.3 TOTAL WHOLESALE MARKET COSTS AND PRODUCTION COSTS
The average annual all-in price, which includes the costs of energy, day-ahead and real-time
reliability unit commitment make-whole payments, operating reserves,72 reserve sharing
group costs, and payments to demand response resources, was $27.65/MWh in 2018. This
compares to the average price of energy at load pricing nodes in SPP’s real-time market for
2018 of $27.08/MWh.73 The all-in price was just over 14 percent higher than the 2017
average all-in price, which is partially attributed to the increase in load and decline in
negative prices in 2018.74 Figure 4—40 plots the average all-in price of energy and the cost of
natural gas, measured at the Panhandle Eastern (PEPL) hub.
Figure 4—40 All-in price of electricity and natural gas cost
72 Operating reserves are resource capacity held in reserve for resource contingencies and NERC control performance compliance, which includes the following products: regulation-up service, regulation-down service, spinning reserve and supplemental reserve. 73 The cost of energy includes all of the shortage pricing components. 74 The Reserve Sharing Group costs and payments to demand response resources were negligible for both years.
$0
$1
$2
$3
$4
$0
$10
$20
$30
$40
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec '16 '17 '18
Nat
ural
gas
co
st ($
/mm
Btu
)
All-
in p
rice
($/M
Wh)
Energy (LMP) Reserves Day-ahead MWP Real-time MWP Gas PEPL
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 148
The figure shows that the vast majority of costs are from the day-ahead and real-time energy
markets.75 The graph also shows that the market cost of operating reserves and make-whole
payments constituted approximately two percent of the all-in price, with make-whole
payments and operating reserves amounting to $0.26/MWh and $0.31/MWh, respectively.
Production cost “is defined as the settlement cost for the market … for all resources.”76
Production cost, in this case, is the sum of four components:
• energy: cleared megawatts multiplied by locational marginal prices;
• ancillary service: cleared operating reserves multiplied by market clearing prices ;
• start-up: “…the out of pocket cost that a Market Participant incurs in starting up a
generating unit from an off-line state ….”;77 and
• no-load: “…the hourly fee for operating a synchronized Resource at zero … output.”78
Figure 4—41 shows the average daily production cost for the day-ahead market.
Figure 4—41 Production cost, daily average, day-ahead
75 Shortage pricing is included in the energy component and not easily separated out in the SPP settlement data. See Section 4.2 for a discussion of shortage pricing impacts. 76 Integrated Marketplace Protocols, Section 7.2.1 77 Integrated Marketplace Protocols, Start-Up Offer 78 Integrated Marketplace Protocols, No-Load Offer
$0
$10
$20
$30
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Mill
ions
Energy Ancillary service No-load Start-up
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 149
The daily average day-ahead market production cost was essentially unchanged from 2016
to 2017. However, the daily average production cost increased more than eighteen percent
from 2017 to 2018. The increase in production cost was almost fully attributed to the energy
component. The energy component is sensitive to numerous inputs, which include fuel cost,
ancillary service scarcity, and load levels. The range of the 2018 daily day-ahead production
costs exceeded $50 million ranging from $2 million to $52 million. Additionally, more than
eighty-one percent of the daily production costs ranged between $10 million and $30
million.
Figure 4—42 Production cost, daily average, real-time
The average daily real-time market production cost was also essentially unchanged from
2016 to 2017. However, real-time production cost increased more than 16 percent from
2017 to 2018. The increase in production cost, similar to day-ahead, was also almost fully
attributed to the energy component. The range of the 2018 daily real-time production costs
was almost $60 million ranging from $4 million to $63 million. Additionally, more than 87
percent of the daily production costs ranged between $10 million and $30 million.
4.4 LONG-RUN PRICE SIGNALS FOR INVESTMENT
In the long term, efficient market prices provide signals for any needed investment in new
transmission, generation, and ongoing maintenance of existing generation to meet load.
Given the very high amount of capacity in the SPP system at peak, which the MMU estimates
was about 35 percent in 2018 (see Section 6.1.3), the MMU does not expect market prices to
$0
$10
$20
$30
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Mill
ions
Energy Ancillary service No-load Start-up
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 150
support new entry of non-wind generation investments. While the SPP markets on their own
may offer low incentives for new growth in renewable generation and storage resources,
some additional reasons for these investments include federal and/or state incentives,
expansion of corporate goals, and emission reduction plans among others.
The MMU conducted an analysis to determine if the SPP market would support investments
in new thermal generation by analyzing the fixed costs, and annual fixed operating and
maintenance costs of three generation technologies relative to their potential net revenues79
at SPP market prices. The plants considered include a scrubbed coal plant, a natural gas
combined-cycle, and a combustion turbine. Figure 4—43 provides the cost assumptions and
Figure 4—44 shows the results of the net revenue analysis. The analysis assumes the market
dispatches the hypothetical resource when price exceeds the short-run marginal cost of
production. In addition to these assumptions a capital recovery factor of 12.6 percent was
used in the annual fixed operating and maintenance cost component.
Figure 4—43 Net revenue analysis assumptions
Scrubbed coal Advanced gas/oil combined-cycle
Advanced combustion Turbine
Size (MW) 650 429 237
Total overnight cost ($/kW-yr.)
$5,089 $1,108 $680
Variable overhead and maintenance ($/MWh)
$7.17 $2.02 $3.54
Fixed overhead and maintenance ($/kW-yr.)
$70.70 $10.10 $6.87
Heat rate (Btu/kWh) 9,750 6,300 9,600 Source: EIA Updated Capital Cost Estimates for Utility Scale Electricity Generating Plants, April 2018
79 Net revenue is equal to revenues minus estimated marginal cost.
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 151
Figure 4—44 Net revenue analysis results
Technology
Average marginal
cost ($/MWh)
Net revenue from SPP market
($/MW yr.)
Annual revenue
requirement ($/MW yr.)
Able to recover
new entry cost
Annual fixed O&M cost
($/MW yr.)
Able to recover
avoidable cost
Scrubbed coal $ 29.60 $ 35,845 $ 709,754 NO $ 70,700 NO
Advanced gas/oil combined-cycle
$ 18.27 $ 69,751 $ 149,238 NO $ 10,100 YES
Advanced combustion turbine
$ 28.31 $ 36,880 $ 92,261 NO $ 9,780 YES
Revenues have been insufficient to support the cost of new entry of generation for all three
technologies since the inception of the Integrated Marketplace, and 2018 was no exception.
In 2015, 2016 and 2017, prices did support the ongoing maintenance cost of combined-
cycle and combustion turbine units, though it did not support the cost of scrubbed coal units.
This is consistent with the 2018 results shown above. Declining gas prices, excess capacity,
and negative prices are the leading contributors to the decline in the profitability of coal
plants.
Figure 4—45 provides results by SPP resource zone, as indicated by the dominant utility in the
area.
Figure 4—45 Net revenue analysis by zone Scrubbed coal Gas/oil combined-cycle Combustion turbine
Resource Zone
Net revenue from SPP market
($/MW yr.)
Able to recover
net entry costs
Able to recover
avoidable cost
Net revenue from SPP market
($/MW yr.)
Able to recover
net entry costs
Able to recover
avoidable cost
Net revenue from SPP market
($/MW yr.)
Able to recover
net entry costs
Able to recover
avoidable cost
AEP $ 24,723 NO NO $98,566 NO YES $55,047 NO YES
KCPL $ 25,781 NO NO $72,891 NO YES $43,327 NO YES
NPPD $17,865 NO NO $45,537 NO YES $30,635 NO YES
OGE $23,438 NO NO $71,113 NO YES $40,395 NO YES
SPS $17,884 NO NO $76,809 NO YES $44,781 NO YES
WAUE $22,846 NO NO $53,536 NO YES $14,801 NO YES
WR $22,108 NO NO $67,004 NO YES $39,832 NO YES
This shows that the conclusions do not vary geographically, even with differing energy prices
and fuel costs. Other RTO/ISO markets have experienced a “missing money problem” in
their markets, where net revenues do not support needed new investments. The MMU
expects the market to signal the retirement of inefficient generation. Aging of the fleet will
Southwest Power Pool, Inc. Prices Market Monitoring Unit
State of the Market 2018 152
eventually change the peak available capacity such that price signals for higher net revenue
become increasingly important. The ability of market forces to provide these incentives and
long-run price signals is a strong benefit of the Integrated Marketplace.
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 153
5 CONGESTION AND TRANSMISSION CONGESTION RIGHTS MARKET
This chapter reviews transmission congestion in the SPP market, as well as the transmission
congestion rights market in SPP. Key points from this chapter include:
• The areas that experienced the highest congestion costs in 2018 were in central
Kansas, southeastern Oklahoma, and southwest Missouri. Western Nebraska
experienced the lowest congestion costs for the year.
• Overall, congestion is down across the SPP footprint, with intervals having no
congestion increasing steadily from 2016 to 2018, and intervals having a breached
constraint steadily decreasing in the same period.
• Frequently constrained areas identified in the 2018 study include central Kansas and
southwest Missouri. The Texas panhandle frequently constrained area was removed
as a result of the 2018 study.
• In aggregate, market participants were effective in hedging congestion in the SPP
market using SPP’s congestion hedging products. Load-serving entities covered
more than 120 percent and non-load-serving entities covered more than 125 percent
of their total congestion cost.
• Individual market participants hedged congestion with varying degrees of
effectiveness — more than 30 percent of load-serving-entities did not fully recover
their congestion cost.
• Transmission congestion right funding fell within the target range. The annual
funding percentage exceeded 94 percent, and the annual shortfall improved by more
than $6 million. Auction revenue right funding decreased from 160 percent to 145
percent; however, the ARR surplus increased more than $20 million and exceeded
$120 million.
• Participants can transfer congestion rights through use of a bulletin board or sell back
positions in the auction. However, most congestion rights are not transferred or sold.
No bulletin board trades cleared during 2018 and have averaged less than one-fourth
of one percent of the total auction quantity since the start of the Integrated
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 154
Marketplace. Intra-auction sales80 increased in volume, but continue to average only
about five percent of the total auction volume.
5.1 TRANSMISSION CONGESTION
The locational marginal price (LMP) for the over 950 settlement locations in the SPP market
reflects the sum of three components:
1) marginal energy component (MEC) - system-wide marginal cost of the energy
required to serve the market,
2) marginal congestion component (MCC) - the marginal cost of any increase or
decrease in energy at a location with respect to transmission constraints, and
3) marginal loss component (MLC) - the marginal cost of any increase or decrease in
energy to minimize system transmission losses.
𝐿𝐿𝐿𝐿𝐿𝐿 = 𝐿𝐿𝑀𝑀𝑀𝑀 + 𝐿𝐿𝑀𝑀𝑀𝑀 + 𝐿𝐿𝐿𝐿𝑀𝑀
LMPs are a key feature of electricity markets that ensure the efficient scheduling,
commitment, and dispatch of generation given the system load and reliability constraints.
LMPs also provide price signals for efficient incentives for future generation and transmission
investment and help guide retirement decisions.
This section focuses on the congestion and loss components of price and related items
including:
• geographic pattern of congestion and losses,
• changes in the transmission system that alter congestion patterns,
• congestion impacts on local market power,
• load-serving entities hedging congestion costs in the transmission congestion
rights market, and
• distribution of marginal congestion and loss amounts.
80 The sale of a previously acquired position in a subsequent auction.
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 155
5.1.1 PRICING PATTERNS AND CONGESTION
Figure 5—1 shows price contour maps representing the day-ahead and real-time average
prices in 2018.
Figure 5—1 Price map, day-ahead and real-time market
Annual average day-ahead market prices ranged from around $16/MWh on the western
edge of Nebraska, to around $36/MWh in the south-central section of Oklahoma. About 75
percent of this price variation can be attributed to congestion and 25 percent to marginal
losses, which is consistent with prior years. Because congestion is more volatile in the real-
time market, the average geographic price range is slightly larger, from $14/MWh to
$38/MWh.
In May 2017, upgrades were made at Woodward by adding a 138kV phase-shifting
transformer. This area was the most frequently constrained area in the SPP market before the
upgrade. However, since the phase-shifting transformer has been in service, the frequency of
Western Nebraska
New Mexico/ west Texas
Southwest Missouri
Central Kansas
Southeast Oklahoma
North Dakota/ Montana border
Woodward phaseshifter
2018 day-ahead locational marginal price
2018 real-time locational marginal price
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 156
congestion at Woodward has fallen significantly such that the MMU recommended no longer
designating Woodward as a frequently constrained area.81 Further upgrades in the area
completed in February 2018 included a second 345kV circuit of Woodward-Tatonga-
Mathewson. Buildout in this area has allowed higher levels of low-cost wind generation in the
western parts of the SPP footprint to serve load centers located on the eastern portion of SPP.
In addition, congestion has shifted eastward. The congestion pocket normally seen in the
Woodward area has shifted down towards southeastern Oklahoma.
As in prior years, the Hays, Kansas area was frequently congested with prices averaging
around $31/MWh in 2018. Lastly, while prices in the North Dakota averaged around
$22/MWh there is a pocket on the North Dakota and Montana border were prices averaged
around $28/MWh, which is consistent with prior years. The higher prices in this area can be
attributed to load associated with oil and gas exploration.
5.1.2 CONGESTION BY GEOGRAPHIC LOCATION
The major drivers of the congestion pattern in SPP are the physical characteristics of the
transmission grid and associated transfer capability, the geographic distribution of load, and
the geographic differences in fuel costs. The eastern side of the SPP footprint, with a higher
concentration of load, also has a higher concentration of high-voltage (345 kV) transmission
lines. Historically, high-voltage connections between the west and east have been limited, as
have high-voltage connections into the Texas Panhandle area.
The costs of coal-fired generation increases as transportation costs rise. For example,
transportation cost increases with distance from the Wyoming Powder River Basin near the
northwest corner of SPP’s footprint. This is important because coal is SPP’s predominant fuel
for energy generation at 42 percent in 2018.
Natural gas-fired generation, SPP’s largest fuel type by installed capacity (41 percent in
2018), resides predominantly in the southern portions of SPP. Wind-powered generation
generally lies in the western half of the footprint, and nuclear generation resides near the
center, while the majority of hydro is located in the north.
81 Frequently constrained areas are discussed further in Section 5.1.7.
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 157
These factors combine to create a general northwest to southeast split in prices. The
exception is slightly higher prices in the northern area of North Dakota resulting from the
growth of, and associated demand from, oil and gas exploration and production facilities.
Outside of the extreme northern part of North Dakota, the Integrated System typically sees
lower prices compared to the rest of the footprint.
Figure 5—2 depicts the average marginal congestion component for the day-ahead market
across the SPP footprint.
Figure 5—2 Marginal congestion cost map, day-ahead market
The lowest average marginal congestion costs occur near the Stegall tie-line in the western
Nebraska region, at −$7/MWh, and the highest marginal congestion costs lie in central
Kansas and southeast Oklahoma at around $5/MWh, and southwest Missouri at around
$7/MWh.
Hays
Neosho -Riverton
Woodward
Stegall tie-line
marginal congestion cost day-ahead market
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 158
The congestion in the Woodward area was persistent in 2017 until the phase-shifting
transformer upgrade went into place in May 2017. With the addition of the phase-shifting
transformer at Woodward, the congestion has shifted east to southeastern Oklahoma, but is
not as evident as the prior Woodward constraint. The Texas Panhandle congestion still
remains but has moved further south towards the Lubbock, Texas area.
5.1.3 TRANSMISSION EXPANSION
Several major transmission projects were completed during 2018 (shown on Figure 5—3
below) that will support the efficient transmission of energy across the SPP footprint.82
• Woodward-Tatonga-Mathewson 345 kV (circuit 2) o location: western Oklahoma o energized: February 2018
• Chisholm-Gracemont 345kV (circuit 1) o location: western Oklahoma o energized: December 2017
• Iatan-Stranger Creek 345kV voltage upgrade (circuit 1) o location: Kansas City area o energized: June 2018
• China Draw-North Loving-Kiowa-Hobbs 345k and terminal upgrades (circuit 1) o location: southeastern New Mexico o energized: June 2018
• Plant X-Tolk 230kV rebuild (circuit 1 and 2) o location: west Texas o energized: March 2018
82 See the 2019 SPP Transmission Expansion Plan Report at https://www.spp.org/documents/56611/2019_spp_transmission_expansion_plan_report.pdf.
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 159
Figure 5—3 SPP transmission expansion in service, 2018
The lines depicted on the map in Figure 5—4 below are projects that will further enhance the
SPP transmission grid in future years.
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 160
Figure 5—4 SPP transmission expansion plan
The Integrated Transmission Plan (ITP) projects shown are recommended upgrades to the
extra-high-voltage backbone (345kV and above) for a 20-year horizon. The ITP process seeks
to target a reasonable balance between long-term transmission investments and congestion
costs to customers. The notification to construct (NTC) and ITP projects shown have received
a written notice from SPP to construct a transmission project that was approved by the SPP
board of directors. Planned projects that may provide relief for the most congested areas in
SPP are listed in Figure 5—10.
5.1.4 TRANSMISSION CONSTRAINTS
Market congestion reflects the economic dispatch cost of honoring transmission constraints.
SPP uses these constraints to manage the flow of energy across the physical bottlenecks of
the grid in the least costly manner while ensuring reliability. In doing so, SPP calculates a
shadow price for each constraint, which indicates the potential reduction in the total market
production costs if the constraint limit could be increased by one megawatt for one hour.
Figure 5—5 provides the top ten flowgate constraints by shadow price for 2018.
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 161
Figure 5—5 Congestion by shadow price, top ten flowgates
Flowgate name Region Flowgate location
VINHAYPOSKNO Central Kansas Vine Tap-North Hays 115kV ftlo Post Rock-Knoll 230kV (MIDW)
NEORIVNEOBLC* SW Missouri Neosho-Riverton 161kV (WR-EDE) ftlo Neosho-Blackberry 345kV (WR-AECI)
TMP109_22593* Eastern Oklahoma Stonewall Tap-Tupelo Tap 138kV (WFEC) ftlo Seminole-Pittsburg 345kV (CSWS-OGE)
TMP270_23432 Eastern Oklahoma Cleveland-Cleveland AECI 138kV (GRDA-AECI) ftlo Cleveland-Tulsa North 345kV (GRDA-CSWS)
TMP151_23193 SW Missouri Oakland East Switch-Joplin Atlas Junction 161kV ftlo Asbury Plant-Purcell Southwest 161kV (EDE)
TAHH59MUSFTS* West Arkansas/East Oklahoma Tahlequah-Highway 59 161kV ftlo Muskogee-Fort Smith 345kV (GRDA-OKGE)
HALTUCHALSWI^ West Texas (Lubbock area) Hale County-Tuco 115kV ftlo Swisher County-Tuco 230kV (SPS)
TMP127_23359 Western Nebraska Scotsbluff-Victory Hill 115kV (NPPD) ftlo Stegall Xfmr 345kV/1 (WAUE)
TEMP72_22893 Eastern Kansas Wolf Creek 345kv/1 (WR) ftlo Wolf Creek-Waverly 345kV (WR-KCPL)
TMP322_23590 Eastern Oklahoma Stonewall Tap-Tupelo Tap 138kV (WFEC) ftlo Sunnyside-Terry 345kV (CSWS-OKGE)
* SPP market-to-market flowgate
^ also includes congestion from TMP228_22916, which became HALTUCHALSWI on July 2, 2018
Overall, the level of congestion in the SPP footprint has declined from 2017 to 2018. In 2017,
the most congested flowgate (Woodward-FPL Switch for the loss of Tatonga-Mathewson) had
a real-time shadow price of nearly $60/MWh, and was congested in almost 25 percent of
real-time intervals. But in 2018, the flowgate with the highest congestion (Vine Tap-North
Hays for the loss of Post Rock-Knoll) had an annual real-time shadow price of just $20/MWh,
and was congested just over eight percent of all real-time intervals. In fact, in 2017, three
flowgates were congested in over 15 percent of all real-time intervals, while in 2018 no
flowgate was congested in over nine percent of all real-time intervals.
0%
10%
20%
30%
40%
$0
$10
$20
$30
$40
Perc
ent
cong
este
d
Shad
ow
pri
ce ($
/MW
h)
Real-time average shadow price Day-ahead average shadow priceReal-time percent intervals congested Day-ahead percent intervals congested
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 162
Most of the highest congested corridors on the SPP system are significantly impacted by
inexpensive wind generation. Of the ten most congested flowgates, those affected the most
by wind generation are the west-to-east flows through the Hays, Kansas area, and west-to-
east flows in eastern Oklahoma. The southwest Missouri area is also impacted by wind and
external flows. Projects are planned throughout the SPP footprint that provide for more
transfer of wind generation from west to east and are listed in Figure 5—10.
5.1.4.1 Central Kansas constraints
The central Kansas area is also impacted by wind generation and several constraints have
appeared consistently since prior to the start of the SPP market. Figure 5—6 compares
congestion over the years for day-ahead and real-time in the central Kansas area since start of
the SPP market.
Figure 5—6 Central Kansas congestion83
The central Kansas area continued to see congestion in 2018. The Vine – Hays 115kV for the
loss of Post Rock – Knoll 230kV (VINHAYPOSKNO) constraint experienced congestion in 30
percent of all intervals in the day-ahead market and eight percent of all intervals in the real-
time market. The yearly average shadow price for the real-time market was $20/MWh for
83 Values combined for SHAHAYPOSKNO (South Hays – Hays 115kV) and VINHAYPOSKNO (Vine – Hays 115kV) constraints for 2017. South Hays – Hays 115kV terminated May 2017, and Vine – Hays 115kV activated September 2017.
0%
20%
40%
60%
$0
$10
$20
$30
2015 2016 2017 2018
Perc
ent
cong
este
d
Shad
ow
pri
ce ($
/MW
h)
Real-time average shadow price Day-ahead average shadow priceReal-time percent intervals congested Day-ahead percent intervals congested
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 163
2018 compared to $27/MWh in 2017. A second Post Rock – Knoll 230kV circuit was
energized in December 2018. This may provide relief to this area.
5.1.4.2 Southwest Missouri
The Neosho – Riverton 161kV constraint is a market-to-market flowgate that is impacted by
SPP and MISO wind, as well as flows from neighboring non-market areas.84 Congestion in
this area dates back to prior to the start of the Integrated Marketplace. However, continued
addition of wind in SPP and neighboring areas have contributed to the increased congestion.
Figure 5—7 compares congestion on the Neosho – Riverton 161kV constraint since 2015.
Figure 5—7 Southwest Missouri congestion
Congestion increased substantially on this constraint in 2017 when compared to previous
years continuing into 2018. This constraint has been a focus of seams discussions on the
amount of market-to-market payments, power swings or volatility, and transmission upgrades
that may benefit SPP, MISO, and other neighboring non-market entities. The market-to-
market process has settled over $27 million in payments for the Neosho-Riverton constraint
from MISO to SPP, and is discussed in Section 2.6.2.
84 Neighboring non-markets include; Tennessee Valley Authority, Associated Electric Cooperative Inc., and Southwestern Power Administration.
0%
10%
20%
30%
40%
$0
$10
$20
$30
$40
2015 2016 2017 2018
Perc
ent
cong
este
d
Shad
ow
pri
ce ($
/MW
h)
Real-time average shadow price Day-ahead average shadow priceReal-time percent intervals congested Day-ahead percent intervals congested
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 164
5.1.5 PLANNED TRANSMISSION PROJECTS
Figure 5—8 provides a list of projects that may alleviate congestion on the ten most
congested flowgates in the SPP system.
Figure 5—8 Top ten congested flowgates with projects Flowgate name Region Flowgate location Projects that may provide relief
VINHAYPOSKNO Western Kansas
Vine-Hays 115kV (WR) ftlo Post Rock-Knoll 230kV (WR)
Post Rock-Knoll 230kV, circuit 2 (energized December 2018, 2017 ITP10)
NEORIVNEOBLC * SW Missouri/ SE Kansas
Neosho-Riverton 161kv (EDE-WR) ftlo Neosho-Blackberry 345kV (WR-AECI)
Neosho – Riverton 161kV Terminal Upgrades (June 2018, 2017 ITP10) [Identified as an economic need in the 2019 ITP Assessment to be completed October 2019]
TMP109_22593* Eastern Oklahoma
Stonewall Tap-Tupelo Tap 138kV (WFEC) ftlo Seminole-Pittsburg 345kV (CSWS-OGE)
Stonewall Tap-Tupelo Tap 138kV Terminal Upgrades (July 2021, 2017 ITP10)
TMP270_23432 Eastern Oklahoma
Cleveland-Cleveland AECI 138kV (GRDA-AECI) ftlo Cleveland-Tulsa North 345kV (GRDA-CSWS)
None identified at this time. [Identified as an economic need in the 2019 ITP Assessment to be completed October 2019]
TMP151_23193 SW Missouri/ SE Kansas
Oakland East Switch-Joplin Atlas Junction 161kV ftlo Asbury Plant-Purcell Southwest 161kV (EDE)
None identified at this time. [Related to economic needs in the 2019 ITP Assessment to be completed October 2019]
TAHH59MUSFTS* West Arkansas/ East Oklahoma
Tahlequah-Highway 59 161kV ftlo Muskogee-Fort Smith 345kV (GRDA-OKGE)
None identified at this time
HALTUCHALSWI West Texas (Lubbock)
Hale County-Tuco 115kV ftlo Swisher County-Tuco 230kV (SPS)
None identified at this time. [Identified as an economic need in the 2019 ITP Assessment to be completed October 2019]
TMP127_23359 Western Nebraska
Scotsbluff-Victory Hill 115kV (NPPD) ftlo Stegall Xfmr 345kV/1 (WAUE)
None identified at this time
TEMP72_22893 Eastern Kansas Wolf Creek 345kv/1 (WR) ftlo Wolf Creek-Waverly 345kV (WR-KCPL)
None identified at this time. [Related to an economic and reliability need in the 2019 ITP Assessment to be completed October 2019]
TMP322_23590 Eastern Oklahoma
Stonewall Tap-Tupelo Tap 138kV (WFEC) ftlo Sunnyside-Terry 345kV (CSWS-OKGE)
Stonewall Tap-Tupelo Tap 138kV Terminal Upgrades (July 2021, 2017 ITP10)
* SPP market-to-market flowgate
5.1.6 GEOGRAPHY AND MARGINAL LOSSES
Variable transmission line losses decrease with increased line voltage or decreased line
length for the same amount of power moved. In the SPP footprint, much of the low-cost
generation resides at a distance from the load and with limited high-voltage interconnection.
The average variable losses on the SPP system for 2018 were 2.8 percent in the day-ahead
market. This is up slightly from 2.7 percent in 2017 and 2.3 percent in 2016. The marginal
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 165
loss component of the price captures the change in the total system cost of losses with an
additional increment of load at a particular location relative to the reference bus.
Figure 5—9 maps the annual average day-ahead market marginal loss components.
Figure 5—9 Marginal loss component map, day-ahead
The average day-ahead marginal loss component ranges from about −$4.45/MWh at the
Laramie River Station in eastern Wyoming, to −$3.32/MWh near North Platte, Nebraska, to
−$0.10/MWh in the Kansas City area, to $0.90/MWh in the Hobbs, New Mexico area, and up
to $2.67/MWh in the corners of Texas/Oklahoma/Arkansas. Negative values reduce prices
through the marginal loss component relative to the marginal energy cost. Positive values
increase prices as generation from these locations are more beneficial from a marginal loss
perspective. The $7.12/MWh spread between geographic prices is larger than the
$5.25/MWh spread in 2017.
day-ahead market marginal loss component
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 166
5.1.7 FREQUENTLY CONSTRAINED AREAS AND LOCAL MARKET POWER
Congestion in the market creates local areas where only a limited number of suppliers can
provide the energy to serve local load without overloading a constrained transmission
element. Under these circumstances, the pivotal suppliers have local market power and the
ability to raise prices above competitive levels thereby extracting higher than normal profits
from the market. SPP’s tariff provides provisions for mitigating the impact of local market
power on prices, and the effectiveness of market power mitigation is described in Section
7.2.2. Local market power can be either transitory, as is frequently the case with an outage,
or persistent, when a particular load pocket is frequently import-constrained.
Because the SPP tariff calls for more stringent market power mitigation for frequently
constrained areas, the MMU analyzes market data at least annually to assess the
appropriateness of the frequently constrained area designations. The 2017 frequently
constrained area study85 recommended the removal of the Woodward area and reduction of
the Texas Panhandle area, and these became effective on April 1, 2018.
The following year’s study (2018) recommended the removal of the reduced Texas
Panhandle (Lubbock) area and the addition of two new areas; southwest Missouri and central
Kansas. These recommendations began through the stakeholder process with a target of
being approved by the SPP board of directors in January 2019 followed by a tariff filing.
In the meantime, stakeholders improved this process by approving revision request 304
(Accelerate Frequently Constrained Area Process.) Under this improved process, the MMU is
to notify market participants of identified changes upon completion of the analysis and allow
a 14 day window for questions or concerns from stakeholders regarding the proposed
changes. After addressing any needed concerns or comment during this period, the
proposed changes will become effective. This improved process, approved by SPP
stakeholders, was accepted by FERC on December 22, 2018.86 After discussion with
stakeholders and implementation of the new process, these 2018 study changes became
effective on February 22, 2019.
85 The Frequently Constrained Area reports can be found on the SPP website: https://www.spp.org/spp-documents-filings/?id=25496. 86 http://elibrary.FERC.gov/idmws/file_list.asp?accession_num=20181220-3042
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 167
5.1.8 MARKET CONGESTION MANAGEMENT
In optimizing the flow of energy to serve the load at the least cost, the SPP market makes
extensive use of the available transmission up to constraint limits. When constraints reach
their limits, they are considered binding. The market occasionally allows transmission lines to
exceed their rating if the price to correct the overload becomes too high. This is considered
a breached constraint. Figure 5—10 highlights day-ahead market binding, breached, and
uncongested intervals.
Figure 5—10 Breached and binding intervals, day-ahead market
The figure shows that uncongested intervals and breached intervals are rare in day-ahead.
Historically in the Integrated Marketplace, less than one percent of day-ahead market
intervals incur a breached condition compared to nearly 25 percent for the real-time
market.87
In the more dynamic environment of the real-time market, uncongested intervals and
breached intervals occur much more frequently than in the day-ahead market. Real-time
congestion is shown in Figure 5—11.
87 SPP uses hourly intervals in the day-ahead Market and five-minute intervals in the real-time market for scheduling, dispatch, and settlement purposes.
0%
20%
40%
60%
80%
100%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Uncongested intervals Intervals with binding only Intervals with a breach
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 168
Figure 5—11 Breached and binding intervals, real-time
As shown above, uncongested intervals have been steadily increasing on an annual basis,
with 18 percent of intervals with no congestion in 2018, up from nine percent in 2016.
Conversely, real-time intervals with a breached constraint in steadily decreasing, with 27
percent of intervals with a breach in 2018, compared to 37 percent in 2016. This increase in
uncongested intervals, and accompanying decrease in breached intervals, indicates less
overall congestion in the SPP market footprint. This can primarily be attributed to the
transmission expansion that has occurred over the past few years.
Market-to-market coordination with MISO, as discussed in Section 2.6.2, was implemented in
March 2015, and the integration of the Integrated System occurred in October of that year.
Both of these events increased the number of constraint breaches. A market-to-market
breach of a MISO constraint could be an indicator that MISO has more efficient generation
than SPP to alleviate congestion on that constraint. Of the 27 percent of the real-time
intervals with a breached constraint, over half of these had a breached market-to-market
constraint.
5.1.9 CONGESTION PAYMENTS AND UPLIFTS
Market participants in the energy market incur congestion costs and receive congestion
payments based on their marginal impact on total market congestion cost through the
marginal congestion component of price. Most SPP market participants owning physical
assets are vertically integrated, so their net congestion cost depends on two things. The first
0%
20%
40%
60%
80%
100%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Uncongested intervals Intervals with binding only Intervals with a breach
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 169
is whether they are a net buyer or seller of energy. The second is the relative marginal cost
component at their generation and load. For financial market participants, congestion costs
reflect the impact of virtual positions on a binding or breached constraint in the day-ahead
and real-time markets.
Figure 5—12 shows the annual day-ahead and real-time market congestion payments for
load-serving market participants during 2018.
Figure 5—12 Annual congestion payment by load-serving entity
Most load-serving entities face congestion costs, depicted as negative payments (charges) in
the graph, because they are part of vertically-integrated entities with higher marginal
congestion components at load than at resources. Day-ahead congestion payments by
ranked load-serving entities ranged from about $73 million in charges to almost $4 million in
payments.88
Market participants also receive payments and incur costs for real-time market congestion,
which are charged and paid based on deviations between day-ahead and real-time market
positions. At an aggregate level, absent the additional revenue neutrality uplift costs, 97
percent of the SPP load-serving entities’ net congestion costs were hedged with day-ahead
88 Day-ahead congestion collections funds transmission congestion rights. These rights are described in greater detail in Section 5.2.
-$100
-$80
-$60
-$40
-$20
$0
$20
Net
co
nges
tion
pay
men
t ($
mill
ions
)
Day-ahead congestion payment Real-time congestion payment Net payment
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 170
prices. Figure 5—13 provides the aggregate congestion costs and hedging totals for load-
serving entities, non-load-serving entities, and financial only entities.
Figure 5—13 Total congestion payments
(in $ millions) Load-serving entities
Non-load-serving and financial only entities
2016 2017 2018 2016 2017 2018
DA congestion $ 259.6 $ 364.6 $ 390.8 $ 81.2 $ 219.0 $ 169.7
RT congestion 14.0 40.6 (10.0) (76.5) (157.2) (97.8)
Net congestion 273.7 405.3 380.9 4.6 61.8 71.9 RT congestion uplift (39.1) (68.7) (74.8) (4.3) (10.2) (11.9)
Real-time market congestion ranged from over $6 million in payments to over $12 million in
costs for load-serving entities. Real-time market congestion ranged from $36 million in
payments to $10 million in costs for non-load-serving entities. Many of the non-load-serving
entities incurring costs represent wind farms, which may sell at negative prices or buy back
day-ahead market positions. The real-time market congestion payments result in a net
benefit of $10 million for load-serving entities. Total real-time market congestion payments
for non-load-serving and financial only entities also resulted in a net benefit and amounted to
$98 million.
Unlike day-ahead congestion, which funds transmission congestion rights, real-time market
congestion costs are allocated to market participants through revenue neutrality uplift (RNU)
charges. In 2017, SPP allocated about 86 percent of revenue neutrality uplift charges to load-
serving entities, resulting in an additional $75 million in congestion-related charges for load-
serving entities.89
89 Real-time congestion uplift is not allocated in the same proportion in which it is collected.
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 171
5.1.10 DISTRIBUTION OF MARGINAL LOSS REVENUES (OVER-COLLECTED LOSSES)
Both the congestion and loss components of prices create excess revenues for SPP that must
be distributed to market participants in an economically efficient manner. In the case of
marginal loss revenues, this requires that the distribution does not alter market incentives.
This was not the case during the first year of SPP’s market, and SPP took steps that largely
corrected the incentive issues. SPP proposed changes to the method for distributing over-
collected losses in FERC Docket No. ER15-763.90 The Commission accepted these changes,
which went into effect in April 2015.
The enhanced design consolidates the distributions of day-ahead and over-collected loss
rebates into one distribution. Under this design, both day-ahead and real-time over-
collected loss rebates are distributed on just real-time withdrawing megawatts. This includes
loads, substation power, exports, wheel-throughs, pseudo-ties, and bilateral settlement
schedules (BSS). The only exception is that both day-ahead and real-time bilateral settlement
schedules are entitled to the rebate. In addition to consolidating the distributions to only
real-time withdrawing megawatts, changes were made to loss pool allocations. Under the
old method, virtual transactions drove up the SPP loss pool allocations of over-collected
losses rebates, even though virtual activity was not eligible for rebates. This caused real-time
exporters to get a large percentage of the over-collected losses rebates which, during that
time, were typically a charge. Virtual transactions are no longer considered in the loss pool
distributions. These design enhancements better allocate the over-collected losses to the
transactions that contributed to the over-collection while removing some of the adverse
incentives present under the former design.
Over-collected losses for the past three years are shown in Figure 5—14.
90 https://www.ferc.gov/CalendarFiles/20150331172533-ER15-763-000.pdf
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 172
Figure 5—14 Over-collected losses, real-time
A total of $145 million was paid out in over-collected losses rebates during 2018, with $133
million (89 percent) going to load. This is up from the $120 million in over-collected losses
rebates paid out in 2017 and $90 million paid in 2016.
The use of bilateral settlement schedules changes the distribution of over-collected losses.
The bilateral settlement schedules enable market participants to transfer energy from one
entity to another at a particular settlement location. It creates a financial withdrawal at the
settlement location for the seller and a financial injection at the settlement location for the
buyer. As long as the bilateral settlement schedules do not change the net withdrawal at the
location, the charges and credits for losses simply change hands between the entities owning
the bilateral settlement schedules. The MMU identified a design issue with the distribution of
loss rebates to bilateral settlement schedules. The issue was corrected in May 2018 with the
implementation of revision request 200.91
Where the bilateral settlement schedules create a net withdrawal that would not otherwise
exist, it creates credits and charges that would not otherwise exist. For example, if a bilateral
settlement schedule amount at a resource settlement location exceeds the cleared output of
the resource, it creates a net withdrawal, and the generation owner receives a loss
91 RR 200 (Design change for bilateral settlement schedule and over-collected losses distribution)
-$30
$0
$30
$60
$90
$120
$150
2016 2017 2018
Ove
r-co
llect
ed lo
ss p
aym
ents
($ m
illio
ns)
Load Resource Interface Hub Net payment
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 173
distribution credit for the excess megawatts of the bilateral settlement schedule. The same
occurs with a bilateral settlement schedule at hubs, where no energy is withdrawn, other than
a bilateral settlement schedule.
Prior to revision request 200, bilateral settlement schedules allowed participants to receive
loss rebates for withdrawing megawatts where losses were not paid into the market. Also,
the fact that bilateral settlement schedules can be entered subsequent to the initial
settlement statements provided information to participants on when to transact the bilateral
settlement schedules based on available loss rebates. This revision request addressed these
concerns.
A recommendation in the 2014 Annual State of the Market report was to remove bilateral
settlement schedule transactions from the over-collected losses distribution calculation.
Revision request 200 still allows for loss rebates to be paid to bilateral settlement schedules.
However, those rebates are now capped at the underlying energy flows of the bilateral
settlement schedules. The MMU believes that the adverse incentives given to bilateral
settlement schedules to transact in amounts that vary from the underlying flows of the
transaction have been removed.
5.2 CONGESTION HEDGING MARKET
In the Integrated Marketplace, the locational marginal prices assessed to load are generally
higher than the locational marginal prices assessed to generators. The largest portion of this
price difference is almost always attributed to congestion. This is an expected outcome and
central to the design of nodal electricity markets. The difference between what generators
are paid and what loads pay is often referred to as congestion rent. SPP remains revenue
neutral in all Integrated Marketplace transactions and therefore must allocate the congestion
rent back to the market participants. The congestion hedging market is the mechanism used
to allocate congestion overcollections.92
5.2.1 MARKET DESIGN
Market participants participate in the congestion hedging market by obtaining auction
revenue rights and/or transmission congestion rights. Auction revenue rights begin as
92 With respect to day-ahead congestion rent only.
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 174
entitlements associated with long-term, firm transmission service reservations. These
transmission service reservations are a revenue source for transmission owners and an
expense for transmission customers. More specifically, transmission owners receive revenues
from transmission customers for building and maintaining the transmission lines, and
transmission customers pay the transmission owners for the use of the lines by way of the
charges associated with transmission service reservations.93
Auction revenue rights link the transmission service, which provides physical rights, to the
Integrated Marketplace by converting these to financial rights. SPP verifies transmission
service entitlements, which become candidate auction revenue rights. To obtain auction
revenue rights, market participants nominate candidate auction revenue rights. If the
nominations pass the allocation’s simultaneous feasibility test, the candidate auction revenue
rights become auction revenue rights. The simultaneous feasibility test ensures that the
market’s aggregate nomination of auction revenue rights does not violate thermal constraint
limits under a single contingency.94 The test incorporates information from the network
model, which aids SPP in aligning the supply of auction revenue rights with the capacity of
the underlying transmission system. The simultaneous feasibility test aims to ensure the
revenues generated from the congestion right auction will sufficiently fund the quantity of
auction revenue rights nominated. If a candidate auction revenue right nomination fails the
simultaneous feasibility test, this reduces the quantity of auction revenue rights successfully
converted from candidate rights.95
Once market participants have successfully nominated their candidates into auction revenue
rights, they must choose to either hold their auction revenue right or convert it into a
transmission congestion right through a process known as self-conversion.96 If a market
participant holds their auction revenue right, they will receive, or pay, a stream of unchanged
cash flows over the life of the product. The size and direction of the cash flow depends on
the market’s collective assessment of the congestion rent along the auction revenue right
path as assessed by prices during the transmission congestion right auction. If a market
participant believes that the auction prices will undervalue the congestion rent associated
93 These charges are assessed through transmission settlements and include various tariff schedules. 94 SPP tariff, 5.3.3, Simultaneous Feasibility 95 2018 SPP MMU Spring Quarterly State of the Market Report, https://www.spp.org/documents/58275/spp_mmu_quarterly_spring_2018.pdf, page 50. 96 SPP tariff 5.4.1 (2)
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 175
with their auction revenue right, the market participant will likely self-convert. When a market
participant self-converts, their auction revenue right becomes a transmission congestion
right, which means their cash flow is subject to the fluctuations in day-ahead market
congestion rent.
Financial-only97 market participants participate alongside traditional utilities in the
transmission congestion right auctions to provide additional liquidity and price discovery. All
participants compete for the residual network capacity, on price. The auction software
attempts to maximize auction revenue without violating the simultaneous feasibility test. This
test helps align the supply of transmission congestion rights with the residual capacity of the
underlying transmission system. If a transmission congestion right bid fails the simultaneous
feasibility test, the quantity of transmission congestion rights successfully obtained will be
reduced to the point where the bid no longer violates the test. Once a market participant
obtains a transmission congestion right, they can hold it through to settlement, offer it for sale
in a subsequent auction, or transact on the bulletin board outside of the auction cycle. The
overwhelming majority of positions are held through to settlement.
5.2.2 MARKET TRANSPARENCY 5.2.2.1 Hedging effectiveness by classification
The transmission congestion right and auction revenue right net payments paid to entities in
the SPP market are shown in Figure 5—15.
97 Finacial-only market participants do not generate or serve load in the SPP footprint.
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 176
Figure 5—15 Total congestion and hedges
(in $ millions) Load-serving entities Non-load-serving and
financial only entities 2016 2017 2018 2016 2017 2018
DA congestion $ 259.6 $ 364.6 $ 390.8 $ 81.2 $ 219.0 $ 169.7
RT congestion 14.0 40.6 (10.0) (76.5) (157.2) (97.8)
Net congestion 273.7 405.3 380.9 4.6 61.8 71.9
TCR charges 51.0 122.5 214.8 63.5 138.0 174.8
TCR payments (212.5) (308.8) (331.5) (158.7) (314.3) (262.8)
TCR uplift 21.1 24.3 26.1 18.1 31.8 32.2
TCR surplus * (4.2) (4.5) (5.4) (4.1) (5.9) (7.5)
ARR payments (74.8) (147.2) (249.1) (6.5) (11.7) (17.8)
ARR surplus (29.2) (94.1) (113.2) (1.9) (7.5) (9.6)
Net TCR/ARR (248.6) (407.9) (458.2) (89.6) (169.5) (90.7)
* remaining at year end
Payments to load-serving entities of $458 million exceeded their day-ahead congestion costs
of $391 million in 2018. Additionally, real-time congestion costs aided load-serving entities
by $10 million, thereby reducing total congestion cost to $381 million. This shows that
overall, load-serving entities did a good job at hedging congestion through the transmission
congestion right market. In aggregate for 2018, non-load-serving and financial only entities
collected transmission congestion right and auction revenue right net revenues of $91
million, which exceeded their day-ahead and real-time market congestions costs of $72
million. Overall, day-ahead congestion cost decreased four percent year-over-year, but
increased more than 40 percent over 2016.
5.2.2.2 Bidding behaviors
The SPP working groups continued dialogue over the past year with respect to obtaining
auction revenue rights, and by extension self-converted transmission congestion rights. The
market monitor reported on this topic in the 2018 spring quarterly state of the market
report,98 and is actively engaged in efforts to improve this issue. As noted above, in
aggregate, load-serving entities received more revenue from their congestion hedges than
98 2018 SPP MMU Spring Quarterly State of the Market Report, https://www.spp.org/documents/58275/spp_mmu_quarterly_spring_2018.pdf, page 50.
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 177
they paid in day-ahead and real-time congestion cost. However, on the individual participant
level, some load-serving entities were under-hedged while others were over-hedged.
Figure 5—16 shows, by load-serving entity, the day-ahead congestion exposure along with
the value of auction revenue rights and transmission congestion rights as well as the net
overall position.
Figure 5—16 Net congestion revenue by market participant
The figure highlights that the majority of participants received positive net revenues, while a
handful of participants held portfolios that did not cover their day-ahead congestion costs.
For instance, the bottom five participants paid over $26 million more in congestion costs than
was offset by their auction revenue right and transmission congestion right positions. This is
substantially less than the $125 million for the bottom five participants in the 2017 calendar
year.
The range of participant outcomes is influenced by three main factors: hedging need,
individual participant bidding behavior, and the market’s collective bidding behavior.
With respect to hedging need, each participant experiences varying levels of congestion
exposure mostly related to geographic location and type of physical interconnection. The
various levels of congestion exposure lead to different hedging needs among market
participants.
-$100
-$75
-$50
-$25
$0
$25
$50
$75
$100
Mill
ions
Day-ahead congestion rent ARR + TCR settlement Net
Net gain
Net loss
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 178
The bidding behavior of the individual participant affects the auction revenue rights they
receive through the allocation. Participants can, and do, employ numerous strategies with
varying degrees of success.
The bidding behavior of the other market participants, as a whole, affects the ability of each
and every other market participant to obtain hedges through the auction revenue right
allocation. More specifically, if a transaction is physically feasible with respect to transmission
service, and by extension the day-ahead market, it does not necessarily mean the transaction
will also be feasible in the auction revenue right allocation. The issue arises because the
transmission system’s capacity, represented by transmission service requests,99 includes both
prevailing flow and counter-flow transactions. However, participants often choose not to
nominate their counter-flow candidate auction revenue rights in the allocation process, in
part, because these positions tend to carry negative cash flows. These incentives motivate
individual participants to abstain from counter-flow positions.100 By not nominating all
candidate auction revenue rights, the capacity in the allocation will not match the capacity of
the physical system. Because the basis of the auction revenue right is transmission service, if
the two capacities do not align, a participant’s auction revenue right may not perfectly hedge
their transmission service and their related day-ahead market activity.
Differences between outages modeled in the auction processes and day-ahead market can
also affect market participants’ ability to obtain hedges. Details on outage modeling are
discussed below.
5.2.2.3 Transmission outage modeling
When outages appear in the day-ahead market that were not in the transmission congestion
rights auction, they reduce system capacity and can cause underfunding. Figure 5—17 shows
transmission outages by reported lead time.
99 Long-term, firm, transmission service requests 100 SPP MWG Meeting Materials, 1/22/2019, https://www.spp.org/spp-documents-filings/?id=18437, 11a.SPP MMU_ARR Observation.pdf
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 179
Figure 5—17 Transmission outages by reporting lead time
SPP models only transmission outages that were reported at least 45 days prior to the first of
the month in the transmission congestion rights auction.101 However, SPP only requires
transmission owners to submit planned outages 14 days in advance.102 The above figure
shows that the majority of outages are not considered in the transmission congestion rights
markets solely due to the submission lead time. Roughly 86 percent of outages are ruled out
of the transmission congestion right model by this phase alone. This is a similar level as prior
years.
Figure 5—18 shows the duration in days for the different types of outages.
101 SPP Integrated Marketplace Protocols, Section 6.6 102 SPP Operating Criteria Appendix OP-2
0
500
1,000
1,500
2,000
2,500
Less than 7 7 to 15 15 to 30 30 to 45 45 to 60 60 to 75 More than75
Num
ber
of o
utag
es
Reporting lead time (days)
Planned Opportunity Forced Discretionary Operational Emergency Urgent
included in auction
not included in auction
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 180
Figure 5—18 Transmission outages by duration
Outages shorter than five days are excluded from auction revenue right/transmission
congestion right processes. This means the vast majority of outages (84 percent) are
excluded from the transmission congestion rights models because they are less than five
days.
Figure 5—19 shows outages by real-time, day-ahead, and transmission congestion right
markets.
Figure 5—19 Transmission outages by market
0
1,000
2,000
3,000
4,000
Less than 1 1 to 3 3 to 5 5 to 7 More than 7
Num
ber
of O
utag
es
Duration (days)
Planned Opportunity Forced Discretionary Operational Emergency Urgent
included in auction
not included in auction
0
1,000
2,000
3,000
4,000
Planned Opportunity Forced Discretionary Operational Emergency Urgent
Num
ber
of o
utag
es
Real-time Day-ahead Transmission congestion right
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 181
While the number of outages in the day-ahead and real-time are similar, the outages
represented in the transmission congestion rights market are only a fraction of the total
number of outages. The transmission congestion market only includes outages that are
longer than five days, and are submitted at least 45 days in advance of the first of the month.
This represented only about five percent of the total outages in the day-ahead market. These
differences in outages can create underfunding of transmission congestion rights.
Ideally, outages in the transmission congestion rights markets would be perfectly aligned
with the day-ahead market. However, the MMU understands the challenges associated with
accounting for outages in the transmission congestion rights market and recognizes that
there can never be an exact match among the markets. Even so, improving how outages are
handled and accounted for in the auction processes could help to improve underfunding.
We encourage stakeholders to consider improving how outages are accounted for in the
transmission congestion right auction process to improve the congestion hedging market
results.
5.2.3 FUNDING
Funding for transmission congestion rights increased slightly in 2018. Conversely, funding
percentage for auction revenue rights decreased in 2018. While the auction revenue right
funding percentage decreased, the auction revenue right surplus increased. As mentioned
in previous reports,103 the overfunding of auction revenue rights could be cause for concern.
The market monitor continues to encourage SPP to review and address the reasons for this
overfunding.
In the 2014 and 2015 Annual State of the Market reports, the MMU discussed a
recommendation and corresponding revision request104 that was intended to reduce over-
allocation of auction revenue rights and the resultant over-selling of transmission congestion
rights by reducing the transmission capability available in the annual auction revenue right
allocation. This revision request made its way through the SPP stakeholder process in 2015
and 2016, and FERC gave final approval in time for the 2016 annual auction revenue
right/transmission congestion right processes. The “TCR year” runs from June 1 through May
31 of each year, so it is not in-sync with the annual report year. Looking at calendar year data,
103 2017 SPP Annual State of the Market Report 104 Revision request 91
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 182
2016 included the revisions for half the year, 2017 and 2018 included the revisions for the
entire year.
Figure 5—20 Transmission congestion right funding levels, annual
The 2016 calendar year, which included the revisions for about half of the year, yielded 92
percent transmission congestion right funding. The 2017 calendar year, which included the
revisions for the entire year, produced 94 percent transmission congestion right funding.
The 2018 calendar year, which also included the revisions for the entire year, also produced
94 percent transmission congestion right funding. While the funding percentage very
marginally improved year-over-year, the shortfall improved by more than $6 million during
2018.
Monthly transmission congestion right funding levels and revenue are shown in Figure 5—21.
0%
20%
40%
60%
80%
100%
-$200
$0
$200
$400
$600
$800
2016 2017 2018
Fund
ing
per
cent
Mill
ions
Day-ahead revenue Transmission congestion right fundingSurplus/shortfall Funding percent
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 183
Figure 5—21 Transmission congestion right funding levels, monthly
The monthly averages appear fairly consistent, with the majority of the months hovering near
the 90 to 100 percent target range. The lowest funding percentage was 86 percent in
February and the highest was at 112 percent in July.105
Daily observations of transmission congestion right funding for the past three years are
shown in Figure 5—22.
105 Target range is specified in the Protocols section 5.3.3. “In the event the cumulative funding is at or below 90 percent or above 100 percent, MWG may approve an additional adjustment…”
0%
20%
40%
60%
80%
100%
120%
-$20
$0
$20
$40
$60
$80
$100
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Fund
ing
per
cent
Mill
ions
Day-ahead revenue Transmission congestion right fundingSurplus/shortfall Funding percent
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 184
Figure 5—22 Transmission congestion right funding, daily
Most daily observations of transmission congestion right funding percent fall between 80
percent and 120 percent, as seen in Figure 5—22. While variation in funding can be expected
as a result of factors including transmission outages and derates, the fact that the majority of
funding falls in this range indicates that the overall process is generally effective. It is
important to note that there was an increase in funding events in excess of 155 percent
increased from one observation in both 2016 and 2017 to seven observations in 2018. This
data suggests that while the annual funding percentage is relatively stable year-over-year, the
upside volatility in the daily observations increased in 2018.
Another way to look at funding is through the funding level and not just percentage. Most
daily observations of transmission congestion right funding fell between plus and minus
$250,000 in 2018. However, the frequency of events outside this range continues to
increase. This data also suggests that while the annual funding shortfall has been relatively
consistent year-over-year, the volatility in the daily surpluses and shortfalls has increased.
The market monitor will continue to evaluate these trends going forward.
Figure 5—23 shows that the auction revenue right funding percentage has increased against
2016 levels in both 2017 and 2018.
0
10
20
30
40
50
60
Day
s
2016 2017 2018
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 185
Figure 5—23 Auction revenue right funding levels, annual
In 2016, auction revenue rights were 138 percent funded. In 2017, auction revenue rights
were 160 percent funded. In 2018, auction revenue rights were 145 percent funded.
Auction revenue right surpluses increased from $27 million in 2016 to $101 million in 2017
and increased again in 2018 to almost $123 million. While the decrease in funding
percentage is encouraging, the substantial and growing surplus presents a potential concern,
as it could be an indicator of inefficiency. The market monitor urges SPP, along with the
stakeholders, to determine the root cause of the overfunding, and analyze the surplus
distribution methodology to ensure it is equitably allocated.
Figure 5—24 shows the 2018 monthly funding levels and revenues for auction revenue rights.
100%
120%
140%
160%
180%
200%
$0
$100
$200
$300
$400
$500
2016 2017 2018
Fund
ing
per
cent
Mill
ions
Transmission congestion right revenue Auction revenue right funding Surplus/shortfall Funding percent
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 186
Figure 5—24 Auction revenue right funding levels, monthly
The shift in auction revenue rights funding beginning in June reflects the change in the TCR
year. The figure also shows the auction revenue right funding was elevated during the last
five months of the 2017 TCR year when compared to the first seven months of the 2018 TCR
year. This general trend was also observed in 2017.
5.2.4 INTRA-AUCTION SALES AND BULLETIN BOARD TRANSACTIONS
Intra-auction sales refer to the sale of a previously acquired transmission congestion right
position in a subsequent auction. Bulletin board transactions refer to trades where a market
participant buys or sells a position outside of the auction cycle through the SPP bulletin
board. Overall, both inter-auction sales and bulletin board transactions remain low.
Figure 5—25 shows the transaction volume by type as a percentage of all transmission
congestion right purchase volume.
0%
50%
100%
150%
200%
250%
$-
$10
$20
$30
$40
$50
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Fund
ing
per
cent
Mill
ions
Transmission congestion right revenue Auction revenue right funding Surplus/shortfall Funding percent
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 187
Figure 5—25 Intra-auction sales and bulletin board transactions
The increase in bulletin board transactions from 2016 to 2017 was due, in large part, to one
market participant selling all of their positions to another market participant. The acquirer
also subsequently sold some of their newly acquired positions to third party market
participants in 2017. However this was unusual, and zero bulletin board transactions cleared
in 2018. Additionally, intra-auction sale volume remained relatively stable around six percent
of the total transmission congestion right volume.
Outside their relationship to the auction cycle, these transactions also differ from each other
in another material way. Bulletin board transactions are similar to the secondary equity
market, where a share of stock is offered for sale and that same share is later purchased by
another market participant. As such, the bulletin board transactions do not affect total
supply; they only affect ownership of the existing supply.
However, intra-auction sales can affect supply in addition to ownership. When market
participants offer their prevailing flow positions for sale intra-auction, the capacity of those
positions once used is available to the market again. But, the newly available system capacity
can be taken up by any path, not just the path sold intra-auction. Furthermore, to sell
counter-flow positions intra-auction, unclaimed prevailing flow capacity must exist for the
transaction to clear.106 This is because counter-flow intra-auction sales reduce the total
106 The additional capacity could also be provided by another counter-flow intra-auction bid.
0%
1%
2%
3%
4%
5%
6%
Intra-auction prevailingflow sales
Intra-auction counterflow sales
Bulletin boardprevailing flow
transactions
Bulletin board counterflow transactions
Perc
enta
ge
of a
ll tr
ansm
issi
on
cong
estio
n ri
ght
tran
sact
ions
2016 2017 2018
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 188
capacity available. The counter-flow now offered for sale, previously facilitated other
prevailing flow positions. If the counter-flow sale were to clear without considering supply,
the prevailing flow once facilitated by this counter-flow would no longer be feasible. So in
order for these existing prevailing flow transactions to remain feasible, additional prevailing
flow capacity must exist. Practically, the additional prevailing flow capacity required plays the
same role once played by the counter-flow being offered for sale. These circumstances likely
also incent market participant abstentions from counter-flow. Generally, if a market
participant holds a counter-flow position, it could be very difficult to sell the position intra-
auction, which is the main source of intra-marketplace liquidity.107
5.2.5 ISSUES, PROGRESS, AND NEXT STEPS
While the congestion hedging market is workably effective overall, the market monitor
highlights the following five areas where continued progress could bring about improved
market outcomes, risk reduction, and efficiency gains.
1. Obtaining auction revenue rights
As previously mentioned market participants experience varying levels of success in
obtaining auction revenue rights and by extension self-converted transmission
congestion rights. SPP and its stakeholders have presented various proposals
regarding the auction revenue rights allocation. The various working groups have yet
to reach consensus, and await the Holistic Integrated Tariff Team’s recommendations,
which are expected in April 2019. The market monitor encourages the continued
development of solutions that have the potential to reduce the magnitude of this
issue.
2. Credit policy
The SPP credit policy’s financial security requirements, associated with congestion
hedges, may not fully reflect the risks inherent in these products. The Credit Practices
Working Group is reviewing SPP’s credit policy and working toward a solution, which
more fully addresses the various risks associated with these products. Given the
gravity and magnitude of the GreenHat Energy default in PJM, the market monitor
107 Intra-market refers to inside the SPP Integrated Marketplace.
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 189
encourages the Credit Practices Working Group to make this particular issue its
highest priority going forward.
3. Allocation of the auction revenue right surplus
The auction revenue right surplus continues to increase and exceeded $120 million
dollars in 2018. SPP should determine the root cause of the surplus and further
determine if the current allocation methodology could be enhanced. The market
monitor views the enhancement of the allocation methodology to be a meaningful
potential source of efficiency gains.
4. Secondary intra-market liquidity
Bulletin board transaction volume ceased in 2018 and intra-auction sales, while stable,
continue to represent modest transaction volume. The limited liquidity associated
with these products is not unique to the SPP markets; however, improved liquidity
would likely prove beneficial for all market participants and enhance efficient auction
price formation. The market monitor encourages SPP to adopt policies and
procedures that deepen liquidity by incenting market participants to transact in SPP’s
secondary market.
5. Modeling inconsistencies
Modeling inconsistencies and outage discrepancies between the congestion hedging
and day-ahead models remain significant. The process and rules surrounding the
modeling of congestion hedging outages should be reviewed, and during the review
SPP and its stakeholders should determine if the current practice is appropriate. The
market monitor recognizes the significant challenges associated with model
convergence and encourages SPP’s continued focus in this area.
Southwest Power Pool, Inc. Congestion and transmission congestion rights market Market Monitoring Unit
State of the Market 2018 190
Southwest Power Pool, Inc. Planning process Market Monitoring Unit
State of the Market 2018 191
6 PLANNING PROCESS
The scope of market monitoring work covers all aspects of the SPP market including activities
that could impact short and long-term operation of the market. The planning process and its
outcomes, in particular, affect congestion patterns, operational effectiveness, and costs, as
well as reliability.
The MMU’s capability to possess and analyze comprehensive market information positions it
to be able to provide valuable input in the planning process. For this reason, starting in 2018
the MMU, in its advisory capacity, has been involved in SPP’s planning process, primarily with
meetings and the discussions of the Economic Studies Working Group (ESWG).108 The MMU
believes its feedback in the planning process on relevant topics will improve planning
outcomes, and benefit the market as a whole.
Key highlights from this chapter include:
• Modified the Future 1 and Future 2 assumptions of the 2020 ITP—carried forward from
the 2019 ITP—on coal and gas generators’ retirement age from 60 years to 56 years for
coal and 50 years for gas.
• The proposed battery accreditation was set to 100 percent as opposed to 48 percent
following a mutual agreement with the SPP staff.
• In addition, the MMU supported the SPP staff’s recommendation for lowering the
modeling assumption of variable operations and maintenance cost (VOM) for
renewable generation in the ITP manual from $8/MWh for wind and solar units to
$0/MWh, consistent with generally observed VOM levels in mitigated offers. The
proposal was approved by the SPP stakeholder process.
• The MMU’s peak available capacity metric was 35 percent in 2018, up slightly from 33
percent in 2017.
108 The ESWG advises and assists SPP staff, various working groups, and task forces in the development and evaluation principles for economic studies. The group provides technical support for the development and application of economic studies. The group also reviews the economic planning processes for adherence to sound economic metrics methods and provides recommendations for improvement of the economic evaluations (see https://www.spp.org/organizational-groups/board-of-directorsmembers-committee/markets-and-operations-policy-committee/economic-studies-working-group/).
Southwest Power Pool, Inc. Planning process Market Monitoring Unit
State of the Market 2018 192
Addressing congestion related issues requires consideration of both transmission and
generation investment options from an overall market view. For instance, the current excess
generation capacity in the SPP market and its implication on market outcomes is a topic that
needs to be evaluated in the planning process. For the last several years, the persistently
high amount of available generation capacity has been a contributing factor to the relatively
low market prices in the SPP market. This affects the financial viability of generators as low
prices that are not supportive of the existing generation capacity makes future retirements
more likely.
The MMU outlined its first recommendations to improve planning processes in the 2017
annual report.109 In the 2017 report, the MMU pointed out that enhancing the accuracy of
planning processes with operational realities would enable SPP and its members to more
effectively plan for future system needs and conditions. The MMU observed that while SPP
and stakeholders have progressed in improving and aligning the planning and operational
processes,110 there are additional areas that could be improved upon. Specifically, we noted
that the economic studies and the resource adequacy processes would benefit from further
alignment with operational conditions.
The MMU was involved in helping stakeholders develop key inputs to the 2020 Integrated
Transmission Planning (ITP) scope and futures case studies. In particular, MMU feedback on
the following items helped shape planning assumptions:
• Coal and natural gas generators’ retirement age were modified from 60 years to 56
years for coal and 50 years for natural gas resources. These changes were based on
MMU analysis of information posted by the Energy Information Administration (EIA)
and were included in both the Future 1 and Future 2 assumptions of the 2020 ITP.
• The MMU assisted SPP staff in assessing how to consider the capacity value of battery
storage resources. The MMU recommended that the capacity of batteries be counted
at 100 percent of capacity as opposed to 48 percent. SPP staff presented and
stakeholders approved this value.
109 SPP Market Monitoring Unit Annual State of the Market 2017, published May 8, 2018 at 198-199. The following is based on that recommendation. 110 The report cited the introduction of an assessment for persistent operational needs and the criteria for identifying these needs in SPP Integrated Transmission Planning Manual (including sections 4.4, 5.3.4, and 6.1.4).
Southwest Power Pool, Inc. Planning process Market Monitoring Unit
State of the Market 2018 193
Section 6.1 of this chapter covers the resource adequacy process. The remainder of this
chapter covers in more detail on the transmission planning process and the input provided
by the MMU.
6.1 RESOURCE ADEQUACY
6.1.1 CAPACITY ADDITIONS AND RETIREMENTS
Nearly 48 percent of SPP’s fleet is more than 30 years old. In particular, nearly 87 percent of
coal capacity and just under 38 percent of gas capacity is older than 30 years. According to
the U.S. Energy Information Administration (EIA), the national average retirement age of coal-
fired generation that retired between 2008 and 2017 was 52 years.111 Aside from the
resources that joined SPP from Nebraska in 2009 and the Integrated System in 2015, the
greatest majority of significant new capacity in the SPP footprint over the last 10 years has
been wind capacity.
Figure 6—1 illustrates that certain segments of the SPP generation fleet are aging. The chart
highlights that almost 30 GW of generation is over 40 years old. Almost half of which is gas,
and about a third is coal.
Figure 6—1 Capacity by age of resource
111 See https://www.eia.gov/todayinenergy/detail.php?id=34452. The EIA analysis includes retirements from not just the electricity sector.
0
10
20
30
40
Coal Gas Hydro Wind Other
Nam
epla
te c
apac
ity (G
W)
0 to 30 years 30 to 40 years 41 to 45 years 46 to 50 years Over 50 years
0
20
40
60
80
100
Total
Southwest Power Pool, Inc. Planning process Market Monitoring Unit
State of the Market 2018 194
Figure 6—2 shows the annual trend of capacity additions and retirements over the past three
years.
Figure 6—2 Capacity additions and retirements by year
Almost all of the coal and gas capacity retired since 2016 has been 1950s era plants. The
nuclear unit retired in 2016 was commissioned in the early 1970s. The wind units that retired
in 2017 were first-generation wind resources with very low capacity.
For capacity additions, wind generation has accounted for 85 percent of the additions over
the last three years and 97 percent of the additions in 2018, totaling over 2,300 MW of
nameplate capacity additions in 2018. Even with the increased amount of solar generation in
the generation interconnection queue, the last solar generation added to the SPP market was
in 2016 at 165 MW.
6.1.2 GENERATION CAPACITY COMPARED TO PEAK LOAD
In the 2017 annual report, the MMU introduced a new peak available capacity metric to
replace the previous reserve margin metric112 used in prior reports. The new metric uses a
percentage of the average maximum capacity for each resource during July and August, and
divides that figure by the nameplate capacity for the resource. This method essentially
112 The previous reserve margin metric used unit registration ratings (i.e. nameplate capacity) to determine system capacity, while wind counted at only five percent of registered capacity.
0
1,200
2,400
3,600
2016 2017 2018 2016 2017 2018
Nam
epla
te c
apac
ity (M
W)
Nuclear Coal Wind Solar Hydro Gas Oilretirements additions
Southwest Power Pool, Inc. Planning process Market Monitoring Unit
State of the Market 2018 195
creates a derated capacity value due to ambient temperatures for each resource and is a
more conservative measure of capacity when compared to nameplate, or even summer rated
capacity. A percentage is then calculated for each fuel type of resource and that percentage
is applied to the total nameplate capacity. Based on capacity factors during the afternoon
hours in July and August, wind and solar resources are derated accordingly. Solar resources
are derated to 50 percent of nameplate capacity. For wind resources, 24 percent of
nameplate capacity is included in the calculation.
The peak available capacity percent is the amount of extra system capacity available after
serving system peak load, and is shown in Figure 6—3.
Figure 6—3 Peak available capacity percent
Year Peak available capacity (MW) Peak load (MWh)
Peak available capacity percent
2014 62,332 45,301 38% 2015113 62,958 45,279 39%
2016 68,839 50,622 36% 2017 67,950 51,181 33% 2018 67,475 49,926 35%
For 2018, the peak available capacity percent was 35 percent, up from 33 percent in 2017.
At 35 percent, peak available capacity is still nearly three times higher than SPP’s minimum
required planning reserve margin of 12 percent.114 Also note that the peak availability
capacity metric will differ from the reserve margin calculated by SPP because of differences in
methodology. Most notably, the SPP methodology only includes capacity with firm
transmission, whereas our metric includes all system resources interconnected with the SPP
grid using derate factors.
A relatively high peak available capacity percentage such as this has positive implications for
both reliability and mitigation of the potential exercise of market power within the market.
However, it also contributes to downward pressure on market prices, negatively affects
113 2015 uses the total capacity on September 30, prior to the addition of the Integrated System. 114 SPP Planning Criteria 4.1.9.
Southwest Power Pool, Inc. Planning process Market Monitoring Unit
State of the Market 2018 196
revenue adequacy, and can burden ratepayers with additional and potentially unnecessary
costs.115
6.2 INTEGRATED TRANSMISSION PLAN
The SPP tariff116 requires SPP to conduct an annual Integrated Transmission Planning
Assessment to evaluate the transmission system upgrades for a ten-year planning horizon.
The planning assessment serves as a regional planning process. This process involves many
aspects of transmission planning including the considerations for reliability, public policy,
operational, and economic needs, and generator interconnection to develop a cost-effective
transmission portfolio for a ten-year planning horizon.117,118 The assessment employs a set of
modeling assumptions that are used as the starting point for planning studies and identifies
system needs from interconnection and transmission service requests within the limits of the
process timelines established. The process coordinates the evaluation of transmission
service needs and associated projects with those identified in the ITP assessment. This way
the current approach targets to facilitate continuity in the SPP’s overall transmission
expansion plan.119
Each integrated transmission plan has a study scope that receives input from stakeholders in
conjunction with the assumptions and parameters that are not standardized in the integrated
transmission process manual.120, The SPP transmission planning studies coordinate the
evaluation of transmission service needs—resulting from generator interconnection and
transmission service requests—and associated projects with those identified in the ITP
assessment to conduct an overall transmission expansion plan.
115 We recognize that grid resiliency is a topic of concern and discussion. However, we feel that resiliency must be properly measured and evaluated in the SPP market to ensure that ratepayers are not burdened by unnecessary costs of excess capacity. 116 See SPP OATT Sixth Revised Vol. No. 1, Attachment O Transmission Planning Process, Section III. 117 The latest version of the ITP manual published on October 17, 2018 is available at the SPP website. 118 SPP also performs a 20-year assessment. We have not covered this assessment in this report as this study is currently not under development. 119 Ibid. at 1. 120 See SPP Tariff Attachment O at Section III.1.c.
Southwest Power Pool, Inc. Planning process Market Monitoring Unit
State of the Market 2018 197
SPP, along with the ESWG and the Transmission Working Group (TWG), develops an
assessment scope for each ITP assessment for items that require SPP stakeholder review and
approval with each new study.121 The study scope document122 describes the assumptions
and methodologies that need to be updated at each study cycle so that the future
performance of the existing transmission system and any needed improvements could be
assessed.123
In 2018, the MMU participated in the development discussions of the 2020 ITP scope. The
MMU’s input covered an array of topics including the retirement age assumptions for coal
and gas-fired generators and accreditation level assumptions for new wind, solar, and battery
resources.
6.2.1 2020 ITP GENERATOR RETIREMENT ASSUMPTIONS
The initial 2020 ITP scope proposal carried forward the 2019 ITP assumptions which assumed
60 years as the retirement age of for coal and natural gas-fired generators. Given the
importance of such assumptions, the MMU recommended updating the 2020 ITP retirement
assumptions to no more than the Energy Information Administration (EIA) average for the last
five years (2013 through 2017) as the existing assumptions appeared to underestimate the
actual retirement trend in the SPP market. The averages were 56 years for coal and 50 years
for natural gas-fired generators.124 The MMU based its analysis on the historic retirement
data compiled by the EIA.125
121 Ibid. at 1-2. The assessment report requires approvals from the MOPC and the board. 122 The most recent scope document is the “2020 Integrated Transmission Planning Assessment Scope,” published on January 18, 2019. The document was approved by the SPP stakeholder process. 123 The ITP manual at 5. 124 The last five year averages through November 2018 (2014-2018) are even lower, 55 years for coal and 49 years for gas-fired resources. 125 The MMU used the EIA’s monthly generator retirement data for the entire U.S. to calculate historic averages for each year. The average retirement years were calculated for the three five-year sub periods to see historical trends for coal and gas-fired generators, and the most recent sub period was used for the recommendation. Only the electric utility generators were included in this calculation (See EIA-860M Preliminary Monthly Electric Generator Inventory (based on Form EIA-860M as a supplement to Form EIA-860 available at https://www.eia.gov/electricity/data/eia860M/).
Southwest Power Pool, Inc. Planning process Market Monitoring Unit
State of the Market 2018 198
The MMU further argued that its net revenue analysis indicates a new coal resource cannot
recover fixed operating and maintenance (O&M) costs at current market prices (see Section
4.4 for 2018 results). In addition, the trend for coal and natural gas resources’ retirements in
2018 was below the five-year EIA averages. As such, even lower retirement ages for the next
10 years could be expected. The MMU’s argument is also supported by the recent
announcements of several market participants redirecting their investment policies towards
more renewable resources and less carbon intensity.126
Figure 6—4 shows the data for average age of retirements for coal and natural gas-fired
resources for the U.S.127
Figure 6—4 Average age of retirements for coal and gas resources 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
Coal 53 48 54 64 54 54 56 58 56 56 49
Gas 43 52 43 49 48 51 52 48 46 52 46
2002-17 2008-17 2013-17
Coal 55 55 56
Gas 47 49 50
Source: EIA-860 Electric Utility Retirements (data through December 2018).
126 For instance, Omaha Public Power District (OPPD) set a long-term goal of providing at least 50 percent of its retail sales from renewable sources and reducing its carbon intensity by 20 percent from 2010 to 2030 (S&P Global Market Intelligence published on Nov., 5, 2018). Xcel Energy Inc. announced that it plans to completely decarbonize its power supply portfolio serving eight states by midcentury, first cutting carbon emission by 80 percent by 2030, from 2005 levels in its utility service territories in Colorado, Michigan, Minnesota, New Mexico, North Dakota, South Dakota, Texas and Wisconsin (S&P Global Market Intelligence published on Dec., 4, 2018). AEP announced in February 6, 2018 that it will pursue a goal to reduce its CO2 emissions of its generating units 60 percent by 2030 and 80 percent by 2050 from 2000 levels (https://www.aep.com/news/releases/read/1503/AEPs-Clean-Energy-Strategy-Will-Achieve-Significant-Future-Carbon-Dioxide-Reductions-). 127 SPP staff did an analysis that compared the SPP region to the total U.S. and identified that the U.S. results were similar to the SPP regional results. See September 25, 2018 meeting materials, Item 3 – 2020 ITP Futures: https://spp.org/Documents/58652/eswg%20twg%20agenda%20&%20background%20materials%2020180925.zip.
Southwest Power Pool, Inc. Planning process Market Monitoring Unit
State of the Market 2018 199
Figure 6—5 below presents the age of all existing resources along with the original proposed
retirements and with the change associated with using EIA averages.
Figure 6—5 Average age of generators in the SPP market
The ESWG modified generator retirement assumptions used in the 2020 ITP to 56 years for
coal and 50 year for natural gas-fired resources. However, participants included an exception
to add back resources that repowered, even if the age limit was reached. This has resulted in
over 4,000 MW of resources that were set to retire based on age back into the study.128
6.2.2 2020 ITP RENEWABLE CAPACITY ACCREDITATION ASSUMPTIONS
In addition to generator retirements assumptions, the 2020 ITP scope includes assumptions
on how to value—or provide credit for—new wind and solar generation capacity. The value of
128 This was approved in the February 2019 ESWG meeting.
0
10
20
30
40
Coal Gas Hydro Wind Other
Nam
epla
te c
apac
ity (G
W)
0 to 30 years 30 to 40 years 41 to 45 years 46 to 50 years Over 50 years
set to retire under original assumption
additional retirement using EIA average
Southwest Power Pool, Inc. Planning process Market Monitoring Unit
State of the Market 2018 200
capacity is called accreditation.129 Each RTO/ISO market has its own assessment of the value
of new wind and solar resources in its region. The original 2020 ITP proposal was to carry
over the assumptions from the 2019 ITP, which valued new wind at 20 percent and new solar
at 70 percent.130,131
The MMU reviewed various accreditation calculations made by other RTOs/ISOs, and no
consistent method was found across the markets. In fact, while some value new wind and
solar capacity during peak or super-peak hours, others use performance during a small set of
the highest load hours. In some cases capacity is differentiated by location (such as coastal
versus non-costal), in other cases capacity is differentiated by technology (such as fixed
versus tracking solar).
As stated in Section 6.1.2, the MMU modified the reserve margin calculation methodology in
the 2017 annual report to make the reserve margin estimation more representative of
operational conditions. Using the new methodology for calculating the peak available
capacity metric, shown in Figure 6—3 above, the MMU developed a set of wind and solar
129 These assumptions for capacity contributions are used as inputs in development of the market economic models and reflect resources’ expected supply availability given a particular future. The final target in the process is to develop and recommend a final portfolio after considering solutions and portfolios addressing varying goals such as mitigating economic needs (cost effectiveness), reliability needs, public policy needs, and persistent operational needs. 130 These numbers were approved by the Supply Adequacy Working Group (SAWG) and used in the ITP. 131 The resource adequacy requirement in SPP is for load to procure generation equal to 112 percent of expected load. The accredited renewable capacity is capped at 12 percent of load-serving entity’s total load only for non-designated resources. The 12 percent cap for renewable capacity does not apply to designated resources. Designated resources are those that can be called upon at any time for the purpose of serving network load on a non-interruptible basis. The designated generation resources must be owned, purchased or leased by the owner of the network load. (See SPP Planning Criteria, Revision 1.6, January 31, 2019).
Southwest Power Pool, Inc. Planning process Market Monitoring Unit
State of the Market 2018 201
capacity factors to aid in decision-making on capacity accreditation assumptions.132 The
results are shown in Figure 6—6 below.
Figure 6—6 MMU calculation for wind and solar capacity factors133 Peak type Capacity factor Wind, peak hour 24%
Wind, super-peak hour 23%
Solar, peak hour 38-45%
Solar, super-peak hour 53-64%
Comparing the SPP accreditation numbers with historical performance, the MMU observed
that the accreditation numbers for new wind resources may be understated and accreditation
numbers for new solar resources may be overstated in the 2020 ITP scope. Consequently,
the MMU recommended that SPP stakeholders modify their accreditation numbers for new
wind and solar resources consistent with historic performance. SPP stakeholders, however,
decided to leave the existing accreditation values unchanged.134
132 The MMU used average historic peak hour performance in July and August over three years, and defined two sets of peak hours: peak hours and super-peak hours. Peak hours reflect the 16 hour peak load period (7 am to 11 pm, non-holiday weekdays) and the super-peak hours reflect the hours beginning 3 pm through 6 pm (four-hour period on all days). Weights were assigned to capacity factors observed in three years giving more weight to a more recent year assuming that a more recent capacity factor would better represent the current situation. As a result, wind peak hour and wind super-peak hours rounded to the same value for the 2015-2017 and 2016-2018 periods; solar peak hour was 38 percent for the 2015-2017 period and 45 percent for the 2016-2018 period; and solar super-peak hour was 53 percent for the 2015-2017 period and 64 percent for the 2016-2018 period. 133 These numbers are calculated from peak and super-peak hours in July and August. Peak hours are 16-hour blocks from 7am to 11pm on non-holiday weekdays. Super-peak hours are calculated from 3pm through 6pm. 134 The ESWG expressed that the Supply Adequacy Working Group (SAWG) has the expert knowledge and opinion on the subject and decided to keep the existing accreditation numbers.
Southwest Power Pool, Inc. Planning process Market Monitoring Unit
State of the Market 2018 202
Finally, the MMU recommended SPP develop a documented methodological approach
going forward for calculating renewable accreditation levels similar to other RTOs/ISOs and
stakeholders to provide input on that approach.135
6.2.3 OTHER TOPICS
In revision request 276, SPP staff recommended updating the pricing of variable operations
and maintenance cost (VOM) for renewable generation in the ITP manual. Specifically, SPP
staff recommended lowering the renewable modeling assumption of $8/MWh for wind and
solar units to $0/MWh in production cost calculations as an interim measure. The MMU
concurred with SPP staff’s assessment as $8/MWh for renewable VOM level was not reflective
of generally observed variable operations and maintenance cost levels in mitigated offers.
Thus, relative to the $8/MWh level, the MMU supported the SPP staff’s $0/MWh proposal
because it more closely reflects operational realities observed in the SPP market, and filed
comments accordingly.136
SPP staff also recommended 48 percent accreditation for battery storage to be used in the
2020 ITP scope. The MMU conducted an analysis by reviewing technical and market
information, and recommended a 100 percent accreditation assumption for this study. The
MMU’s recommendation was based primarily on the following: the general market practice
of capacity contracts for batteries enhances 100 percent accreditation; the physical capability
of the resource should be the main factor for such calculation; and there is no justification for
an accreditation number other than the 100 percent level at this time. After further
135 The Supply Adequacy Working Group (SAWG) is currently considering to adopt a specific methodology for determining accreditation figures as used in other markets. In its March 27, 2019 meeting, the SAWG approved a motion to use Effective Load Carrying Capability (ELCC) method as the guiding principle for the accreditation of solar, wind and storage resources to replace the current accreditation methodology found in section 7.1.5.3 (7) of the SPP Planning Criteria once new criteria language is approved. 136 In the March 2018 ESWG meeting, the stakeholders approved the $8/MWh number and rejected the $0/MWh recommendation. The Market Operations and Policy Committee overruled the ESWG decision and approved renewable VOM costs at $0/MWh at its April 2018 meeting.
Southwest Power Pool, Inc. Planning process Market Monitoring Unit
State of the Market 2018 203
consideration, SPP staff concurred with the MMU arguments and the ESWG approved the
accreditation for battery storage at 100 percent in December.137
In addition, the ESWG evaluated the potential of adding a third study case as part of the 2020
ITP. The third future would have considered a shift in environmental regulations—such as a
carbon tax adder—and a change in technologies/market trends including accelerated
deployment of storage devices or electric vehicles. The MMU supported evaluation of this
third scenario as it would help SPP and stakeholders plan for a wider range of future
possibilities than was captured in the other study cases. Moreover, the third study case was
similar to a study case in the MISO transmission plan.138 SPP stakeholders rejected the third
study case as they considered it too costly to include another case study as part of the
evaluation process.
6.3 ASSESSMENT
The MMU has a unique position to evaluate SPP data and market outcomes and provide
feedback in the planning process to improve efficiencies and outcomes. This involvement
has already proven beneficial as stakeholders approved some MMU recommendations for
the 2020 ITP.
The MMU believes the accuracy of (planning) projections for needed upgrades in
transmission infrastructure heavily depends on market data and analysis. When underlying
assumptions (or inputs) for economic modeling are not reflective of operational realities or
historical trends—such as the retirement ages of generators, the role of renewables, or
emerging new technologies—both the location and magnitude of need for new transmission
might be inappropriate. Consequently, the market signals for investment in transmission and
generation could be distorted resulting in unexpected congestion patterns and suboptimal
levels of transmission. This in turn would inevitably translate into suboptimal costs and
market prices.
137 The MMU raised the issue in ESWG’s November 2018 meeting, and the stakeholders approved the 100 percent recommendation in their December 3-4, 2018 meeting. 138 MTEP19 Futures: Summary of definitions, uncertainty variables, resource forecasts, siting process and siting results, undated (available at https://cdn.misoenergy.org/MTEP19%20Futures%20Summary291183.pdf).
Southwest Power Pool, Inc. Planning process Market Monitoring Unit
State of the Market 2018 204
For instance, planning for generation and resulting congestion patterns that would not likely
exist given prevailing trends would impose an unnecessary cost of transmission on market
participants in particular and ratepayers in general. Alternatively, a robust planning
projection could serve as a preemptive measure for eliminating or diverting congestion. A
particularly useful example is the case of addition of the phase-shifting transformer at the
Woodward substation in 2017 to address congestion. This upgrade substantially reduced
the frequency of congestion at Woodward such that the MMU recommended no longer
designating Woodward as a frequently constrained area. Furthermore, as highlighted earlier
in this report (see Section 5.1.3), further expansions at Woodward in 2018 were a significant
factor in reducing the incidence of negative prices. The MMU believes that SPP can further
align planning processes with market and operational realities to achieve full benefits of the
process.
Finally, the MMU has the most up-to-date cost data for all resources and can provide valuable
feedback in the planning process when financial assumptions are made. As in the case of
renewable variable operations and maintenance modeling assumptions, the MMU is
prepared to make directional guidance on the level of such costs without compromising
confidential market participant data.
6.4 RECOMMENDATIONS GOING FORWARD
Based on its experience with the planning process in 2018, the MMU recommends the
planning process continue to use data gathering and analysis methods to determine
economic modeling inputs so that a more accurate assessment can be made for the SPP’s
system needs and conditions. One way to accomplish this is to review best practices across
other organized markets.
For instance, the accreditation calculations for new solar and wind resources in other markets
encompass varying methods that differ by peak and super-peak hours, location (coastal
versus non-coastal) or type (fixed versus tracking). In some markets, the incremental effect of
the loss of load without the (renewable) resource is used in calculating accreditation levels.
Accreditation calculations for battery storage also require equally careful approaches. A
better way to accomplish this going forward is to develop a documented and dynamic
approach along with input from stakeholders.
Southwest Power Pool, Inc. Planning process Market Monitoring Unit
State of the Market 2018 205
Another important issue is the generator retirement age assumptions used in economic
modeling. In addition to market participant input, substantial consideration needs to be
given to the operational realities and historical trends when the assumptions for retirement
ages are made.
The futures used in the planning process allow for the assessment of likely deviations from
the (base) reference case where the impact could be significant. When including scenarios,
the ITP scope should consider developments occurring in the power sector in general with
due considerations given for probable shifts in environmental regulations, market trends, or
technological advancements. In addition, the planning processes should factor in
announcements with regard to major shifts in corporate policies given the evolving market
trends and technologies towards more renewable generation and reduced emissions.
The planning process should have the ability to assess a range of potential outcomes in
transmission planning. Following market trends helped shape the MMU’s thinking in terms of
how to consider the future generation mix and the potential need for an additional scenario.
The MMU believed that a third scenario would provide a bookend scenario for the 2020
analysis but was rejected by stakeholders as it was considered too costly to study.
While the transmission planning process theoretically can include a third scenario, in practice
it does not have the flexibility to include one given that cost of including a third scenario was
not unique to the 2020 planning process. This appears to be a significant shortcoming in the
planning process as the study process is limited to only two potential scenarios in its 5- and
10-year look ahead. This limits the range of potential outcomes that SPP could study139 or
alternatively requires the cases to be very broad and different from each other.
We recommend that SPP enhance their study process to allow the ability to study a range of
potential outcomes. If such range of potential outcomes are not captured in a third case
study, the MMU recommends to factor them into either an existing case study or potentially
as part of the 20 year assessment. The wider the range of possibilities studied, the more
robust the results will be.
139 The MISO transmission planning process studies four cases, see https://cdn.misoenergy.org/MTEP19%20Futures%20Summary291183.pdf.
Southwest Power Pool, Inc. Planning process Market Monitoring Unit
State of the Market 2018 206
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 207
7 COMPETITIVE ASSESSMENT
Chapter 7 of this report provides a competitive assessment of the SPP market. Key points
from this chapter include:
• Structural and behavioral metrics indicate that the SPP markets have been very
competitive over the last several years. The market share, HHI, and pivotal supplier
analyses all indicate minimal to moderate potential structural market power in SPP
markets outside of areas that are frequently congested.
• There were 35 percent of the hours in 2018 having the market share exceeding the 20
percent threshold, which was a significant increase from 2017. The data shows that
the increase in highest market share hours occurred following May 2018, which
coincides with the merger between Great Plains and Westar.
• The HHI index analysis show that 12 percent of hours were considered moderately
concentrated in 2018, with all of those hours occurring after the merger, while the
market was unconcentrated in all hours of 2016 and 2017.
• While the Great Plains and Westar merger has increased market concentration
compared to the last two years, the market was even more concentrated before 2016.
The MMU does not have any concerns with the market concentration increase with the
merger.
• The results of pivotal supplier analyses indicate that the percent of hours with pivotal
supplier is the highest (98 to 100 percent) in the New Mexico and West Texas region,
increasing with demand level. This is where one of the SPP’s frequently constrained
areas in 2018 was located.
• The monthly offer price markups of 2018 ranged from −$5.17 to $2.64/MWh for off-
peak periods and from −$6.01 to $3.07/MWh for on-peak periods. The lowest
markups occurred in spring and winter in off-peak and on-peak hours, when wind
generation was generally the highest.
• Both off-peak and on-peak average markups were at the lowest levels since
implementation of the Integrated Marketplace at around −$2.05/MWh and
−$2.41/MWh, respectively.
• The incremental energy mitigation was extremely low, at less than 0.01 percent in
both the day-ahead real-time markets in 2018. The overall mitigation frequency of
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 208
start-up offers in 2018 were at the lowest levels since the market started in 2014 as the
combined frequency of mitigation of start-up offers for day-ahead, reliability unit
commitment, and manual commitments decreased to 2.8 percent in 2018 from 3.2
percent in 2017.
• Similar to 2017, economic withholding mitigation remained infrequent. The results of
the output gap analysis indicate still a very low levels of (economic) output withheld in
the SPP footprint, less than 0.01 percent on annual average level. These outcomes are
consistent with the expectations of competitive market conduct.
• The unoffered economic capacity results show that on an annual average basis the
total unoffered capacity equaled 2.0 percent in 2016, 1.9 percent in 2017, and 3.2
percent in 2018. The majority of the outages were long-term, and were primarily the
result of maintenance in the spring and fall shoulder months.
• The start-up mitigation only occurred in less than three percent of intervals in 2018
and day-ahead mitigation accounted for 84 percent of the total start-up cost
mitigation. These figures were similar to results in 2017.
• Major maintenance costs will begin being included in mitigated start-up and no-load
offers in April 2019.
• Starting from 2017, the MMU has been involved with the SPP’s initiative to develop a
new process for evaluating generator retirement applications. Given the expected
market developments and increased rate of retirements, a streamlined process of
collecting generator specific (financial) data would greatly enhance the MMU’s review
of retirements and would provide sufficient time to recommend and develop
appropriate mitigation measures.
The SPP Integrated Marketplace provides sufficient market incentives to produce competitive
market outcomes in regions and periods when there are no concerns with regard to local
market power. The MMU’s competitive assessment provides evidence that in 2018 market
outcomes were workably competitive and that the market required mitigation of local market
power infrequently to achieve those outcomes.
The market power analysis in this report considers both the structural and behavioral aspects
of market power concerns. The structural aspects can be detected by various techniques
such as market share analysis, (market-wide) concentration indices, and pivotal supplier
analysis. The structural indicators are used to look for the potential for market power without
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 209
regard to the actual exercise of market power. Behavioral aspects, on the other hand, assess
the actual offer or bid behavior (i.e., conduct) of the market participants, and the impact of
such behavior on market prices by looking for the exercise of market power. These
behavioral indicators include offer price markup;140 economic withholding analysis,
addressed through automated mitigation; output gap analysis; and physical withholding.141
This chapter evaluates the SPP market’s competitive environment by establishing the level of
structural market power and then examining market prices for indications of market power
impact. The level of structural marker power is assessed both at the general level through
concentration indices and at the local—transmission constraint—level through pivotal supplier
analysis. Mitigation of economic withholding is accomplished ex-ante through automatic
market power mitigation processes that limit the ability of generators with local market power
to raise prices above competitive levels. The mitigation program is monitored and evaluated
to ensure it is efficient and effective. Accordingly, the following subsections examine the
significance of market power and the effectiveness of local market power mitigation in the
SPP markets.
7.1 STRUCTURAL ASPECTS OF THE MARKET
Three core metrics of structural market power are the market share analysis, the Herfindahl-
Hirschman Index (HHI), and pivotal supplier analysis. The first two of these indicators
measure concentration in the market and are of a static nature. Pivotal supplier analysis, on
the other hand, takes into account the dynamic nature of power markets and considers
changing demand conditions and locational transmission constraints in assessing potential
market power.
Figure 7—1 displays the market share of the largest on-line supplier (i.e., market participant)
in terms of energy output in the real-time market by hour for 2018.
140 While the SPP MMU uses offer price markup, other market monitors may use price cost markup as a behavioral indicator. 141 The uneconomic production analysis, as another behavioral indicator, is addressed through FERC referrals.
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 210
Figure 7—1 Market share of largest supplier
The market share ranged from 9.1 percent to 28.4 percent, exceeding the 20 percent
threshold142 in 3,093 of the hours (35 percent) for the year. This represents a significant
increase from the 9.3 percent to 16.9 percent range in 2017. The data show that the above
mentioned increase in highest market share hours occurred beginning June 2018 which
coincides with the merger between Great Plains (the parent company of KCP&L) and Westar
Energy that formed Evergy, Inc. Furthermore, these 3,093 hours covers both summer peak
and the following shoulder months indicating a consistent pattern. Note that although a
mere increase in market share in itself does not pose a threat to the structural
competitiveness of the SPP market, other relevant market data including pivotal supplier
hours and local market power mitigation must also be evaluated for competitive assessment
(see below).
142 The 20 percent threshold is one of the generally accepted metrics that would indicate structural market power. Note, however, that neither market share nor the HHI metric alone would be sufficient for the assessment of market power particularly in today’s spot electricity markets where load pockets formed by transmission congestion may lead to market power with much smaller market shares and/or HHI values.
0
5
10
15
20
25
30
Perc
ent
mar
ket s
hare
of
larg
est m
arke
t p
artic
ipan
t
20% threshold
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 211
The Herfindahl-Hirschman Index (HHI) is another general measure of structural market power,
analyzing overall supplier concentration in the market. It is calculated by using the sum of the
squares of the market shares of all suppliers in a market as follows:
𝐻𝐻𝐻𝐻𝐻𝐻 = ��𝐿𝐿𝑀𝑀𝑖𝑖∑ 𝐿𝐿𝑀𝑀𝑖𝑖𝑖𝑖
∗ 100�2
𝑖𝑖
According to FERC’s “Merger Policy Statement,”143 which is similar to Department of Justice
merger guidelines, an HHI less than 1,000 is an indication of an unconcentrated market, an
HHI of 1,000 to 1,800 indicates a moderately concentrated market, and an HHI over 1,800
indicates a highly concentrated market.
Figure 7—2 provides the number of hours for each concentration category in terms of actual
generation over the last three years.144
Figure 7—2 Market concentration level, real-time
Concentration HHI Level
2016 2017 2018
Hours % of Hours Hours % of
Hours Hours % of Hours
Unconcentrated Below 1,000 8,784 100% 8,760 100% 7,692 88%
Moderately concentrated
1,000 to 1,800 0 0% 0 0% 1,067 12%
Highly concentrated
Above 1,800 0 0% 0 0% 0 0%
The SPP market was unconcentrated in all hours of 2016 and 2017. However, 12 percent of
hours were considered moderately concentrated in 2018. The SPP market has never risen
above the highly concentrated threshold of 1,800 since the start of the Integrated
Marketplace in 2014. Figure 7—3 depicts the hourly real-time market HHI in terms of
generation for 2018.
143 Inquiry Concerning the Commission’s Merger Policy Under the Federal Power Act: Policy Statement, Order No. 592, Issued December 18, 1996 (Docket No. RM96-6-000). 144 The SPP MMU calculates HHI by actual sales. The FERC merger guidelines uses capacity owned.
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 212
Figure 7—3 Herfindahl-Hirschman Index
Hourly HHI values ranged from 502 to 1,283 during 2018, pointing to a higher trend relative
to 2017.
Figure 7—4 shows a graphical breakdown of the HHI for all hours since the start of the
Integrated Marketplace in March 2014. For the years with significant events impacting the
make-up of the market – 2015, for the addition of the Integrated System; and 2018, with the
creation of Evergy – the years are divided on the chart showing the HHI before and after the
event.
0
500
1,000
1,500
2,000
Her
find
ahl-H
irsc
hman
Ind
ex
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Unconcentrated
Moderately concentrated
Highly concentrated
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 213
Figure 7—4 Herfindahl-Hirschman Index, annual
As shown above, the market remained unconcentrated from the date of the addition of the
Integrated System to the date of the KCPL/Westar merger creating Evergy.145 Even though
moderately concentrated hours increased following the creation of Evergy, just under 20
percent of hours were considered moderately concentrated. In contrast, in 2014, just over 50
percent of hours were considered moderately concentrated, and in 2015, prior to the
addition of the Integrated System, nearly 40 percent of hours were moderately concentrated.
While moderately concentrated hours increased following the creation of Evergy, it is not
necessarily a cause for alarm. Given the HHI results in conjunction with the elevated market
shares, a closer monitoring of market behavior by market participants in terms of the exercise
of local market power may be required.146
145 Note that the HHI analysis is performed at the market participant level. There may be asset owners under market participants on a contractual basis, where bidding control is not under the purview of the market participant. 146 Section 7.2 analyzes behavioral aspect of market power.
0%
20%
40%
60%
80%
100%
2014 2015pre-IS
2015post-IS
2016 2017 2018pre-merger
2018post-merger
Unconcentrated Moderately concentrated Highly concentrated
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 214
SPP market participants with generation spanning all supply segments,147 reflecting the
technology and fuel mix of resources, have the greatest ability to benefit from structural
market power. These market participants may frequently set prices regardless of the
technology type on the margin. Figure 7—5 provides the percent of real-time market intervals
that each market participant had a resource on the margin.148
Figure 7—5 Market participants with a marginal resource, real-time
The chart shows that the top three market participants each set price in more than 10
percent, but less than 15 percent, of all real-time market time intervals. Conversely, well over
half of all participants set price in less than one percent of all intervals.149
The MMU’s market share analysis and calculated HHI both indicate a moderate potential for
general structural market power in SPP markets outside of areas that are frequently
congested. Structural market power is also assessed at a more localized level and in the
context of locational transmission constraints by reevaluating frequently constrained areas
periodically and (re)defining them accordingly as was discussed in Section 5.1.7.
147 Supply segments include baseload, intermediate, and peak resources. 148 Kansas City Power & Light and Westar Energy were treated as separate entities for the purpose of this analysis. 149 The percentages on this chart are not additive because multiple market participants may have a resource on the margin during any given interval.
0%
5%
10%
15%
20%
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63
Perc
ent
of r
eal-t
ime
inte
rval
s w
ith a
m
arg
inal
res
our
ce
Market participants
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 215
Pivotal supplier analysis takes into account the dynamic nature of the power market,
particularly demand conditions, and evaluates the potential for market power in the presence
of “pivotal” suppliers. A supplier is pivotal when its resources are needed to meet demand.
There may be one or more pivotal suppliers in a particular market defined by transmission
constraints and load conditions, and a supplier’s status of being pivotal may vary between
time periods irrespective of its size. In the market clearing process, market power is
evaluated locally through a local market power evaluation because the exercise of local
market power is relevant for determining prices.
The following analysis identifies the frequency with which at least one supplier was pivotal in
the five different reserve zones (regions) of the SPP footprint in 2018.150 One condition for a
supplier to have an ability to raise prices above competitive levels, is the frequency it
becomes pivotal. Another market condition is during times of shortage or high demand.
The mere size of a supplier has no link to being pivotal; however, suppliers with a high
frequency of being pivotal in tight supply periods have an even greater ability to exercise
market power. For this reason, the frequency of being a pivotal supplier is also analyzed at
various levels of demand across these five regions.
Figure 7—6 shows how frequently a supplier is pivotal at varying load levels in the five reserve
zones in the SPP market footprint.
150 SPP divides market resources (generation) into five reserve zones. For the purpose of this report, these reserve zones are named as “Nebraska”, “Western Kansas and Panhandles”, “New Mexico and West Texas”, “Kan., Mo., Okla., Ark., East Texas, La.”, and “Iowa, Dakotas, Montana”. Thus, each generation resource is mapped to one of these reserve zones. To define a load zone to match with a resource zone, each load settlement location was mapped to a reserve zone to approximate demand within a particular zone. Additionally, import limits are approximated by the average of the reserve zone limits for the times they were activated in 2018.
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 216
Figure 7—6 Hours with at least one pivotal supplier
The results indicate that the percent of hours with pivotal supplier is the highest (98 to 100
percent) in the New Mexico and West Texas region, increasing with demand level. This is
where one of the SPP’s frequently constrained areas in 2018 was located. This region is
followed by the Iowa, Dakotas, Montana region where, depending on the level of load, one
to 52 percent of the hours exhibit at least one pivotal supplier. Compared to previous years,
the Nebraska region increased from seven to 12 percent of high demand hours having a
pivotal supplier. The remaining regions experience pivotal supplier conditions for only
negligible periods.
7.2 BEHAVIORAL ASPECTS OF THE MARKET
7.2.1 OFFER PRICE MARKUP
In a competitive market, prices should reflect the short-run marginal cost of production of the
marginal unit. In SPP’s Integrated Marketplace, market participants submit hourly mitigated
energy offer curves that represent their short-run marginal cost of energy. Market
participants also submit their market-based offers, which may differ from their mitigated
offers. To assess market performance, a comparison is made between the market offer and
the mitigated offer for the marginal resources for each real-time market interval. Figure 7—7
0%
20%
40%
60%
80%
100%
Iowa, Dakotas,Montana
Nebraska New Mexico andWest Texas
Kan., Mo., Okla.,Ark., East Texas, La.
Western Kansasand Panhandles
Perc
ent
of h
our
s
< 20th between 20-80th > 80thDemand levels
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 217
provides the average marginal resource offer price markups151 by month for on-peak and off-
peak periods. Figure 7—8 shows the average on-peak offer price markup by fuel type of
resource. The 2018 data show that we see a growing trend of negative markups in both on-
and off-peak periods, but also in most months as well (see Figure 7—9 for annual ranges).
This implies that there is significant competition in the SPP market.
Figure 7—7 Average offer price markup of marginal resource, monthly
In 2018 the average monthly markups ranged from −$5.17 to $2.64/MWh for off-peak
periods and from −$6.01 to $3.07/MWh for on-peak periods. The lowest markups occurred
in spring and winter in off-peak and on-peak hours, when wind generation was generally the
highest.
151 Offer price markup is calculated as the difference between market-based offer and the mitigated offer where the market-based offer may or may not be equal to the mitigated offer. The MMU calculates a simple average over all marginal resources for an interval. The markups are weighted to reflect each marginal resource’s proportional impact on price.
-$8
-$6
-$4
-$2
$0
$2
$4
Jan 18 Feb 18 Mar 18 Apr 18 May 18 Jun 18 Jul 18 Aug 18 Sep 18 Oct 18 Nov 18 Dec 18
Offe
r p
rice
mar
gin
al c
ost
mar
kup
($
/MW
h)
Off-peak On-peak
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 218
Figure 7—8 Average on-peak offer price markup by fuel type of resource, monthly
The observed levels of negative markups indicate that many market participants’ real-time
market offers were below their mitigated offers.152 This could occur where generators decide
to offer below their marginal cost to maintain commitments in a very competitive
marketplace. For instance, wind resources at the margin increased from 10.4 percent of all
resource hours in 2017 to 12.2 percent in 2018. Figure 7—9 below points to a continued
declining annual trend of off-peak and on-peak average markups.
152 Wind units may have negative mitigated offers because of subsidies related to production tax credits.
-$30
-$20
-$10
$0
$10
Jan 18 Feb 18 Mar 18 Apr 18 May 18 Jun 18 Jul 18 Aug 18 Sep 18 Oct 18 Nov 18 Dec 18
On-peak
Offe
r p
rice
mar
gin
al c
ost
mar
kup
($
/MW
h)
Coal Gas Hydro Wind
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 219
Figure 7—9 Average offer price markup, annual
Both off-peak and on-peak average markups were at the lowest levels since implementation
of the Integrated Marketplace at around −$2.05/MWh and −$2.41/MWh, respectively,
showing lower levels compared to 2017 figures. Although a lower offer price markup level in
itself would indicate competitive pressure on suppliers in the SPP market, the observed
continuous and deepening downward trend may raise questions about the commercial
viability of generating units and the possibility of generation retirements. However, even
though these pressures have continued to mount over the last several years, SPP continues to
add new generation (mostly wind), and generation capacity relative to peak loads remains
high (see Section 6.1.2). More renewable generation is anticipated to be added over the
next few years, which is expected to outweigh planned retirements. As such the trend in
negative offer-price mark ups will likely continue, further putting pressures on the economic
viability of many generating resources.
7.2.2 MITIGATION PERFORMANCE AND FREQUENCY
SPP employs an automated conduct and impact mitigation process to address potential
market power abuse through economic withholding. The mitigation applies to resources that
exercise local market power in areas of transmission congestion, reserve zone shortages, and
manual commitments.
-$8
-$4
$0
$4
2016 2017 2018 2016 2017 2018
Off-peak On-peak
Offe
r p
rice
mar
gin
al c
ost
mar
kup
($
/MW
h)
High Average Low
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 220
SPP resources’ incremental energy, start-up, no-load, and operating reserve offers are subject
to mitigation for economic withholding when the following three circumstances occur
simultaneously in a market solution:
1) The resource has local market power;
2) The offer has failed the conduct test. Resources submit two offers for each product: a
mitigated offer representing the competitive baseline costs that must adhere to the
mitigated offer development guidelines153 and a second offer generally referred to as
a market offer. An offer fails the conduct test when the market offer exceeds the
mitigated offer by more than the allowed threshold; and
3) The resource either:
a) Is manually committed by SPP for capacity, transmission constraint, or voltage
support; or by a local transmission operator for local transmission problems or
voltage support; or
b) The application of mitigation impacts market prices or make-whole payments by
more than the allowed $25/MWh threshold.
Mitigation was very low overall, with some variation across products and markets. Figure 7—
10 shows that the mitigation of incremental energy, operating reserves, and no-load was
generally infrequent in the day-ahead market in 2018.
153 As indicated in Appendix G of the SPP’s Market Protocols.
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 221
Figure 7—10 Mitigation frequency, day-ahead market
Overall patterns for mitigation frequency for the three products in the day-ahead market
were similar to those in 2017. Mitigation frequency increased slightly in September and
October, particularly for no-load offers. The mitigation frequency for operating reserves
increased slightly starting in October through December. The application of mitigation in the
day-ahead market occurred at levels of 0.05 percent for operating reserves, 0.16 percent for
no-load, less than 0.01 percent for incremental energy, and about 2.4 percent of start-up
offers.
Mitigation of incremental energy in the real-time market is shown in Figure 7—11 below.
0.0%
0.1%
0.2%
0.3%
0.4%
0.5%
0.6%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Perc
ent
of r
eso
urce
ho
urs
miti
gat
ed
Energy Operating reserves No-load
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 222
Figure 7—11 Mitigation frequency, real-time market
For the real-time market, the mitigation of incremental energy has been at very low levels
since the start of the market with an annual average around 0.01 percent in 2018, which is
about the same as 2017 levels.
Figure 7—12 shows the mitigation frequency for start-up offers for the various commitment
types.
Figure 7—12 Mitigation frequency, start-up offers
0.00%
0.02%
0.04%
0.06%
0.08%
0.10%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Perc
ent
of r
eso
urce
ho
urs
miti
gat
ed
0%
2%
4%
6%
8%
10%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Perc
ent
of s
tart
-up
offe
rs m
itig
ated
Day-ahead Reliability unit commitment Manual
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 223
The overall mitigation frequency of start-up offers was the lowest since the market began in
2014, as it decreased in 2018 relative to 2017 levels, to under three percent. While the
frequency of manual and real-time mitigation dropped in 2018 compared to 2016 and 2017
levels, day-ahead mitigation remained less than 3 percent from 2016 to 2018. Day-ahead
mitigation accounted for 84 percent of the total start-up cost mitigation. The highest level of
mitigation of start-up offers was at 6.5 percent in June and fell to around 1.5 percent in
December.
7.2.3 OUTPUT GAP (MEASURE FOR POTENTIAL ECONOMIC WITHHOLDING)
Economic withholding by a resource is defined as submitting an offer that is unjustifiably high
such that either the resource will not be scheduled or dispatched, or if scheduled or
dispatched the offer will set a higher than competitive market clearing price. Accordingly,
the output gap metric aims to measure the economic (or competitive) amount of output that
was withheld from the market by submitting offers in excess of competitive levels. The
output gap is the amount of generation not produced as a result of offers exceeding the
mitigated offer above an appropriate conduct threshold. The conduct threshold is employed
to compensate for any inaccuracies or uncertainties in estimating the cost, similarly to the one
used in economic withholding mitigation. In this report, the output gap is calculated as the
difference between the resource’s economic level of output at the prevailing market clearing
price and the actual amount of production. The economic level of output is produced by a
generator between its minimum and maximum economic capacity.154
The output gap calculation used the same method as last year.155,156 We
implemented a multi-stage process to determine the economic output level for a unit
for output gap evaluation.
In the first stage, we determined if the unit would have recovered its start-up, no-load,
and incremental costs if ran for its minimum run time at the dispatch point dictated by
154 The MMU calculates this metric by including all resources’ total (reference level) capacity when calculating output gap percentages. 155 The metric is based on the approach used by Potomac Economics. 156 Under this method, units were grouped into two categories: economic units that are not committed, and committed units that are dispatched at lower levels than their economic level.
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 224
the prevailing day-ahead energy price. In the second stage, if a unit was economic
for commitment, we then identified the economic level of incremental output during
hours when it was economic to produce energy based on market prices. To reflect
the timeframe in which commitment decisions are actually made, we used day-ahead
prices for non-quick-start resources and real-time prices for quick-start resources in
assessing the output gap.
This year the MMU still considered the 17.5 percent conduct threshold for the
frequently constrained areas and the 25 percent conduct threshold for the rest of the
footprint to reflect the actual thresholds used in economic withholding mitigation.157
In order to account for the discrepancy between a resource’s offered capacity and the
dispatched amount (because of possible limitations in real-time market conditions
such as transmission constraints, operator actions or ramp limitations, virtual
participants), an upward adjustment is made by taking the greater of the day-ahead
scheduled or the real-time dispatched amount to reflect the actual amount of
production.
Note that certain market conditions such as congestion (supplier location), supplier size, or
high demand can create market power and facilitate economic withholding behavior. For
this reason, the output gap is calculated as percentages of total economic output withheld
compared to total reference capacity for the SPP footprint and for the two frequently
constrained areas. In addition, the output gap is calculated for the largest three suppliers
(market participant portfolios) in each area comparing the levels to those of the remaining
suppliers. Similar to the last year’s report, the annual calculations were run at varying levels of
demand as a potential market condition that can affect the withholding outcome.
The results in Figure 7—13 below show the SPP footprint-wide monthly levels of the output
gap from 2016 to 2018.
157 The two frequently constrained areas Texas Panhandle and Woodward stayed as such until April 1st after which Lubbock was designated as the only frequently constrained area in the SPP footprint. The following charts reflect this situation.
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 225
Figure 7—13 Output gap, monthly
Though slightly higher in 2018 compared to previous years, the output gap remained low
overall averaging less than 0.03 percent in all months, reflecting a high level of participation
in the market, overall. The increase in May was due to two resources including opportunity
cost adders only in their market offers without reflecting them in their mitigated offers. This
was also the reason for the slight increases in September and October. These impacts are
shown as the light purple bars in Figure 7—12. When controlling for these impacts, the
output gap remained low overall averaging less than 0.01 percent in all months.158
Figure 7—14 displays the output gap calculated by demand level and participant size for the
entire SPP market footprint, while Figure 7—15 shows the output gap for the Lubbock
frequently constrained area.159 The Woodward and Texas Panhandle frequently constrained
areas were only active for the first three months of 2018, with results of those frequently
constrained areas being insignificant, they are not included as a chart in this section. In
general more output is expected to be withheld at higher demand levels or by larger
suppliers. However, at times output may also be withheld in low load periods, as prices are
often negative during the lowest 20 percent of load hours.
158 These impacts have been excluded from the following charts in this section. 159 The Lubbock frequently constrained area was activated on April 1, 2018, therefore data shown only covers April through December 2018.
0.00%
0.01%
0.02%
0.03%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Out
put
with
held
Actual output gap OCC impacted gap
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 226
Figure 7—14 Output gap, SPP footprint
Figure 7—15 Output gap, Lubbock frequently constrained area
There was no output gap in the new Lubbock frequently constrained area for any demand
and supplier levels. In general, the results indicate a very low level of (economic) output
withheld in the SPP footprint, less than 0.01 percent on annual average level. These
outcomes are consistent with expectations of competitive market conduct.
0.00%
0.05%
0.10%
0.15%
0.20%
Top 3 suppliers Other suppliers Top 3 suppliers Other suppliers Top 3 suppliers Other suppliers
< 20th percentile between 20-80th percentiles > 80th percentile
Cap
acity
with
held
2016 2017 2018
0.00%
0.05%
0.10%
0.15%
0.20%
Other suppliers Top 3 suppliers Other suppliers Top 3 suppliers Other suppliers Top 3 suppliers
< 20th percentile between 20-80th percentiles > 80th percentile
Cap
acity
with
held
Total2018
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 227
7.2.4 UNOFFERED GENERATION CAPACITY (MEASURE FOR POTENTIAL PHYSICAL WITHHOLDING)
As part of the competitive assessment, we also looked into the potential physical withholding
behavior by generators throughout the 2016 to 2018 period. Physical withholding refers to a
conduct where a supplier derates a resource or otherwise does not offer it into the market.
Physical withholding may include intentionally not following dispatch instructions, declaring
false derates or outages, refusing to provide offers, or providing inaccurate capability
limitations. Any economic generation capacity that is not made available to the market
through a derate, outage, or otherwise not offered to the market is considered for this
analysis.160,161
We classified total economic capacity that was derated from respective reference levels by
reason and duration. Deratings can take the form of planned outages approved in SPP’s
outage scheduling system, forced outages, or any undesignated unoffered capacity.162 Any
deratings from reference levels including partial deratings are considered in this analysis.
Derates were divided into short-term and long-term. Those with less than seven days
duration were classified as short-term and the rest as long-term. This is because the
economic capacity that was not offered short-term has more potential for physical
withholding relative to long-term derates as it would be less costly—because of loss of sales—
for a supplier to withhold capacity for a short duration of time.
As in the case for economic withholding, potential for physical withholding is also affected by
various market conditions at the time offers are made including location (congestion),
supplier size, or demand levels. Larger suppliers would be in a more advantageous position
160 This analysis, in part, draws on “Assessment of the Market Monitoring Metrics for the SPP Energy Imbalance Service (EIS) Market,” Potomac Economics, December 2010 and “2016 State of the Market Report For the New York ISO Markets,” Potomac Economics, May 2017. 161 Economic capacity is determined in a similar way as in the output gap analysis in Section 7.2.3 by comparing resource’s (cost-based) mitigated offer to the prevailing locational price. 162 The planned maintenance outages by nuclear generation and unoffered capacity by hydro, wind and solar is excluded in this analysis.
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 228
to exercise market power. During tight market conditions, suppliers have more incentive and
opportunity to physically withhold capacity for strategic reasons. In addition, scheduling
maintenance outages in high demand periods may indicate a strategic behavior to create
artificial shortages.
In the assessment, we considered derated and unoffered economic capacity both in day-
ahead and real-time. Similar to the output gap analysis, the commitment decisions were
made based on day-ahead market outcomes for non-quick-start units and real-time market
outcomes for quick-start units.163 The unoffered capacity is calculated as the difference
between the unit’s economic capacity164 and its offered maximum economic capacity
operating limit during intervals when the unit was deemed economic (i.e., covering its costs
given the clearing price).
The following figures show unoffered economic capacity as percent of total resource
reference levels by month for the SPP footprint, for the Lubbock frequently constrained area,
and by supplier (participant) size against varying load levels.165
Figure 7—16 Unoffered economic capacity
163 See Section 7.2.3 for explanation of this method. 164 Bounded by a resource’s reference level. 165 Unoffered capacity percentages are calculated out of the total reference levels of the corresponding area (i.e., SPP footprint or each of the frequently constrained areas).
0%
1%
2%
3%
4%
5%
6%
7%
8%
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 2016 '17 '18
Perc
ent
of c
apac
ity
Short-term outage Long-term outage All other unoffered Gas, internal combustion unoffered
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 229
Figure 7—16 shows that on an annual average basis the total unoffered capacity equaled 2.0
percent in 2016, 1.9 percent in 2017, and 3.2 percent in 2018.
The figure also shows that the majority of the outages were long-term and concentrated
during the fall and spring shoulder months. When short and long-term outages were
excluded from the averages, the remaining unoffered capacity amounts to 0.22 percent, 0.23
percent, and 0.39 percent for 2016 through 2018, respectively. From a competitive market
perspective, the results generally indicate reasonable levels of total unoffered economic
capacity. The latter (net of outages) results, which are very low, could also be interpreted to
indicate pressure on market participants, particularly on coal-fired resources, to offer—and
maintain commitments—given their long-term coal contracts.166 The general high levels of
self-scheduling of supply offers in the SPP market could be another contributing factor in this
outcome.
Figure 7—17 shows that short-term outages by either large suppliers or others do not rise with
increasing load across the SPP footprint.
Figure 7—17 Unoffered economic capacity at various load levels, SPP footprint
On the other hand, unoffered economic capacity of gas (peaker) units by large suppliers rises
with increased load albeit at very low levels of 0.03 percent, 0.04 percent, and 0.06 percent
166 Coal-fired resources may prefer to offer at times even below their mitigated offers to guarantee to be scheduled or dispatched. See also the offer price markup analysis earlier in Section 7.2.1 and the negative markup results reported therein.
0%
1%
2%
3%
4%
5%
Top 3 suppliers Other suppliers Top 3 suppliers Other suppliers Top 3 suppliers Other suppliers
< 20th percentile between 20-80th percentiles >80th percentile
Uno
ffere
d e
cono
mic
cap
acity
Short-term outage Long-term outage All other unoffered Gas, internal combustion unoffered
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 230
of respective load levels. The unoffered economic capacity of gas units by other suppliers
increases even at lower levels against load. Unoffered economic capacity compared to load
is more apparent for the remaining resource types, however, it does not exceed 0.22 percent
for either of the supplier groups which was the same as in 2017.
Another take away from the results is that while long-term outages constitute the majority of
total outages in the SPP footprint at higher load levels, short-term outages play a larger role
in the Lubbock frequently constrained area167, as shown in Figure 7—18. As with the output
gap metrics, the Woodward and Texas Panhandle frequently constrained areas were only
active for the first three months of 2018, with results of those frequently constrained areas
being insignificant, they are not included as a chart in this section.
Figure 7—18 Unoffered economic capacity at various load levels, Lubbock area
The Lubbock frequently constrained area shows the dominance of larger participants in
declaring long-term and short-term outages.
The SPP-wide generation outage data168 (see Section 3.1.4) show that most long-term
outages were for maintenance (70 percent). Of those outages, 77 percent were scheduled
during shoulder months. Out of the short-term outages, approximately 33 percent were
167 The Lubbock frequently constrained area was activated on April 1, 2018, therefore data shown only covers April through December 2018. 168 Covering all resources in the SPP market including nuclear, hydro, wind and solar generation.
0%
5%
10%
15%
20%
25%
Top 3 suppliers Other suppliers Top 3 suppliers Other suppliers Top 3 suppliers Other suppliers
< 20th percentile between 20-80th percentiles >80th percentile
Uno
ffere
d e
cono
mic
cap
acity
Short-term outage Long-term outage All other unoffered Gas, internal combustion unoffered
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 231
forced outages, with close to half of them in shoulder months. The results are generally
consistent with competitive market conduct.
7.3 OFFER BEHAVIOR DUE TO MITIGATION THRESHOLD
As discussed in the 2016 State of the Market report, the MMU observed inefficient market
behavior with regard to the mitigated threshold. The MMU recommended that mitigation
measures for resources committed for a local reliability issue be treated separately from the
mitigation measures for economic withholding.169 Resources that fall into this category are
not subject to the three tests associated with economic withholding, which is appropriate.
The MMU submitted revision request 231170 to the Market Working Group in May 2017 to
address this issue. In that revision request, the MMU proposed converting the 10 percent
threshold for local reliability mitigation to a 10 percent cap. The change was implemented in
December 2018 and caps offers at the mitigated offer plus ten percent for local
commitments.
7.4 START-UP AND NO-LOAD BEHAVIOR
Analysis of no-load and start-up offers showed that many market participants made start-up
and no-load offers considerably above their mitigated offer levels (see Figure 7—19).
Nonetheless, start-up mitigation only occurred in less than three percent of intervals in 2018
and day-ahead mitigation accounted for 84 percent of the total start-up cost mitigation.
These figures were similar to results in 2017.
169 In its 2016 report, the MMU recommended converting the 10 percent mitigation threshold for local reliability commitments to a 10 percent cap. 170 Revision request 231 mitigation of locally committed resources.
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 232
Figure 7—19 Mitigation start-up offer mark-up by fuel category
MMU’s further analysis and discussion starting in 2017 indicated that start-up mitigation was
most frequently the result of major maintenance costs inherent in the repeated heating and
cooling process of starting-up and shutting down units that were included in market based
offers but not mitigated offers. As the purpose of the mitigated offer is to prevent undue
costs not directly tied to a commitment or dispatch decision from being imposed on the
market, revision request 245 was developed and submitted in late 2017.171 This revision
request allows for the inclusion of major maintenance costs that can be directly tied to the
number of run hours or starts to be included in the mitigated offers for start-up and no-load
and properly evaluated by the market clearing engine in determining commitments.
Revision request 245 was approved by the FERC in October 2018.172 Major maintenance
costs can be included in mitigated start-up and no-load offers in April 2019.
7.5 GENERATOR RETIREMENTS
As discussed earlier in this report, the SPP market has currently a significant amount of excess
generation capacity. Furthermore, the substantial incoming—and projected—new capacity,
primarily wind, is likely to exacerbate this situation absent significant amount of generator
retirements. While the SPP tariff provides for a 12 percent capacity (reserve) margin for the
171 RR245 Mitigated Start-Up and No-Load Offer Maintenance Cost. 172 FERC Order Accepting Tariff Revisions Docket No. ER18-1632-001, October 18, 2018.
0%
100%
200%
300%
400%
500%
600%
700%
800%
900%
1000%
Coal Gas, CC Gas, steam Gas, turbine/engine Oil
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 233
footprint, the MMU’s calculated peak available capacity metric discussed in Section 6.1.3
point to levels well above the required level, at 35 percent. Calculating reserve margin
requirements is based on long established standards and provide sufficient cushion for long
term reliability whereas significant excess capacity impose additional costs on the market.
When such excess capacity exists, functioning competitive markets send signals mainly
through (low) energy prices173 indicating a need for the exit of that excess capacity.
A recent wave of generator retirements, particularly of coal-fired generation, has been widely
observed throughout the country. The SPP market should be expected to follow this trend
because of excess capacity, aging fleet, and cost disadvantages of certain types of
generation technologies vis-à-vis the prevailing market prices.
Considering these developments, and starting from 2017, the MMU has been involved with
the SPP’s initiative to develop a new process for evaluating generator retirement applications.
The MMU has approached this primarily from market economics and market power
perspectives. In fact, a strategically motivated generator retirement particularly in a
congested area may create structural and behavioral market power issues that may
potentially amount to physical withholding. Creating a shortage through a retirement may
lead to sustained price spikes and/or may benefit generators affiliated with the retiring
generator(s). The MMU is already evaluating generator retirements from reasonableness,174
market impact, and physical withholding perspectives. However, given the expected market
developments and increased rate of retirements, a streamlined process of collecting
generator specific (financial) data would greatly enhance the MMU’s review of retirements
and would provide sufficient time to recommend and develop appropriate mitigation
measures. This way, based on the MMU’s impact assessment, measures can be identified to
alleviate or preempt any detrimental impact such as transmission upgrades, frequently
constrained area designation, or stricter mitigation rules.
The MMU has already informed SPP staff and the market participants about its proposed
review process through internal meetings and presentations at the Market Working Group,
and will continue to provide feedback as necessary.
173 As highlighted earlier in Section 7.2.1, market offers are less than mitigated offers, which contributes to price formation. 174 Through both technical and economic justifications.
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 234
7.6 COMPETITIVE ASSESSMENT SUMMARY
Overall, structural and behavioral metrics indicate that the SPP markets have been very
competitive over the last several years. The market share, HHI, and pivotal supplier analyses
all indicate minimal to moderate potential structural market power in SPP markets outside of
areas that are frequently congested.
Market share indicators increased from minimal to lower moderate levels after Great Plains
and Westar merged to form Evergy, Inc. in June 2018. Furthermore, the percent of the hours
where the SPP market was unconcentrated declined from 100 percent to 88 percent. These
analyses both indicate a moderate potential for general structural market power in SPP
markets outside of areas that are frequently congested. The MMU will continue to evaluate
structural market power concerns going forward.
Meanwhile, there was one frequently constrained area in 2018 where the potential concerns
of local market power were the highest. 175 Ongoing analysis shows existing mitigation
measures have been an effective deterrent in preventing pivotal suppliers from unilaterally
raising prices.
Behavioral indicators were also assessed through the analysis of actual offer or bid behavior
(i.e., conduct) of the market participants and the impact of such behavior on market prices to
look for the exercise of market power. One such indicator, the negative offer price mark-ups
in 2018 continued to show increased negative offer levels similar to those in 2017. This also
points to a continued declining annual trend of off-peak and on-peak average offer price
markups. What is striking for 2018 is that not only the relative and absolute amounts with
respect to 2017 have increased but the pattern has changed to include winter and summer
peak period months as well. This could occur where—particularly coal—generators decide to
offer below their marginal cost to maintain commitments or when wind units—that may have
negative mitigated offers because of subsidies related to production tax credits—constitute a
sizeable share in total resource portfolio. As stated earlier, a lower offer price markup level in
itself would indicate a competitive pressure on suppliers in the SPP market, the observed
175 Texas Panhandle and Woodward maintained their frequently constrained area status until April 1, 2018 after which Lubbock was designated as the only such area in the SPP footprint.
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 235
continuous and deepening downward trend may raise questions about the commercial
viability of generating units and the possibility of generation retirements.
Similar to 2017, economic withholding mitigation remained infrequent. In particular, the
incremental energy mitigation was extremely low in both markets, at less than 0.01 percent in
both the day-ahead real-time markets in 2018. The overall mitigation frequency of start-up
offers in 2018 were at the lowest levels since market started in 2014 as the combined
frequency of mitigation of start-up offers for day-ahead, reliability unit commitment, and
manual commitments decreased to 2.8 percent in 2018 from 3.2 percent in 2017.
Meanwhile, the combined frequency of no-load, operating reserve, and incremental energy
mitigation decreased from their in 2017 levels and remained at negligible levels.
The system wide monthly output gap results show a very low-level—less than 0.03 percent—of
economic withholding in all months in 2016 to 2018 across the SPP footprint. These low
overall levels of economic output withheld are consistent with competitive market conduct.
The new metric introduced last year, the average unoffered economic capacity was around
two percent in the 2016 to 2017 period and was just over three percent in 2018. The majority
of the outages were long-term, and were primarily the result of maintenance in the spring
and fall shoulder months. Furthermore, the short-term outages by either large suppliers or
others do not rise with increasing load across the SPP footprint. The results are generally
consistent with the workings of a typical competitive market.
Meanwhile, the very low level—less than one percent—of unoffered capacity net of outages
could indicate a pressure on market participants to offer—and maintain commitments—given
their longer term agreements. The general high levels of self-committing supply in the
market could be another factor in the low levels of unoffered capacity.
Overall, the SPP Integrated Marketplace provides effective market incentives and mitigation
measures to produce competitive market outcomes particularly during market intervals
where exercise of local market power is a concern. The competitive assessment in this report
provides evidence that market results in 2018 were workably competitive and that the market
required mitigation of local market power infrequently to achieve those outcomes.
Nonetheless, mitigation remains an essential tool in ensuring that market results are
competitive during periods when such market conditions offer suppliers the potential to
abuse local market power.
Southwest Power Pool, Inc. Competitive assessment Market Monitoring Unit
State of the Market 2018 236
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 237
8 RECOMMENDATIONS
One of the core functions of a market monitor is “to advise the Commission, the RTO or ISO,
and other interested entities of its views regarding any needed rule and tariff changes.”176
The MMU accomplishes this responsibility through many forums, including but not limited to
active participation in the SPP stakeholder meetings process, commenting on FERC notices
of proposed rulemakings, submitting comments at FERC on SPP filings, and making
recommendations in the Annual State of the Market report. This section outlines the MMU
recommendations to SPP and stakeholders to address our concerns with the current design,
rules, and processes.
This section highlights new recommendations, clarifies existing recommendations, and
outlines recommendations made in prior reports. For each recommendation, we present the
reasons for our recommendation and assign a priority. We also identify the current status of
previous recommendations. Overall, SPP and its stakeholders have made significant
progress on most outstanding MMU recommendations. Section 8.4 lists the status of past
and current annual report recommendations.
8.1 NEW RECOMMENDATIONS
The following recommendations are new for 2018. We have five new recommendations
including recommendations on parameter values and market power, on how SPP calculates
credit requirements in light of a credit situation that occurred in the PJM markets in 2018,
pricing capacity for uncertainties, improving results in the transmission planning processes,
and improving mileage price formation
8.1.1 LIMIT THE EXERCISE OF MARKET POWER BY CREATING A BACKSTOP FOR PARAMETER CHANGES
The market is currently vulnerable to the exercise of market power by the manipulation of
their resources’ non-dollar based parameters. Where there is no market power, a seller
cannot control price because other sellers are competing for revenue. Market power is a
market participant’s ability to manipulate price by manipulating either supply, demand, or
176 As defined by FERC in Order No. 719.
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 238
some amount of both through either dollar or non-dollar based offers. Where a seller has
market power, the seller can control price.
Market participants can exercise market power by manipulating a resource’s non-dollar-
based parameters. For instance, a resource can manipulate energy price by reducing its
maximum operating limit from its actual limitation so that it reduces the supply of energy and
raises the price. Non-dollar-based parameters should not be manipulated to affect market
clearing. Although SPP’s tariff and protocols have well-defined expectations and precise
limitations for the basis of dollar-based offer components in the presence of local market
power, the expectations for the basis of non-dollar-based offer components are much less
clearly defined and the limitations are much less precise.
The MMU recommends that SPP strengthen the language regarding non-dollar-based
parameters so that the expectation for the basis of these values is clear and the potential
exercise of market power is much more limited. The expectations for the basis of the
parameters should be clear and well-defined. Changes to these parameters should be
limited to actual capability and should be applied, at a minimum, in the presence of market
power. One option would be to require parameters to reflect actual limitations always.
Actual limitations could include physical and environmental limitations and potentially other
true and verifiable limitations. Another option could be to automatically apply parameter
mitigation in the presence of market power similar to the automatic mitigation of dollar-
based offer components.
8.1.2 ENHANCE CREDIT RULES TO ACCOUNT FOR KNOWN INFORMATION IN ASSESSMENTS
In 2018, GreenHat Energy, a financial-only market participant in PJM, defaulted on its
portfolio of congestion hedging products in the PJM markets. Current estimates of the
default exposure exceed $400 million.177 GreenHat Energy acquired its portfolio in
compliance with the PJM credit policy, which due to its design, required GreenHat Energy to
post less than $1 million in financial security. The disconnect between the projected loss and
the required financial security stems largely from PJM’s credit policy’s assessment of the
historic congestion patterns associated with GreenHat Energy’s positions. Planned
177 https://pjm.com/-/media/committees-groups/committees/mic/20190206/20190206-item-01a-informational-update-ferc-order-denying-waiver-request.ashx, page 5.
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 239
transmission expansion changed the historic congestion patterns in the PJM markets, which
caused GreenHat Energy’s congestion hedging portfolio to become unprofitable. Ultimately,
GreenHat Energy defaulted and has forced the rest of the PJM market to absorb the costs.
While the SPP market is different from the PJM market, SPP’s credit policy is similar to PJM’s
in some respects. For instance, SPP’s financial security requirements for transmission
congestion rights are based only on historic congestion patterns even when significant
transmission upgrades are planned or have occurred. For instance, as noted in the 2017
Annual State of the Markets report, congestion patterns in SPP shifted significantly after
installation of the phase shifting transformer at the Woodward substation, which changed
congestion patterns throughout the footprint.178 This known event was not factored into
SPP’s financial security requirements for transmission congestion rights even though it was
expected to have a significant effect on outcomes. Furthermore, the congestion pattern was
so noticeably shifted that after only a few months the MMU recommended eliminating a
frequently constrained area as a result of the expansion.179 However, SPP’s financial security
requirements only factored in the historic data.180
The MMU has engaged SPP’s Credit Practice Working Group and has contributed to the
dialogue about next steps. SPP and its stakeholders generally agree that updating the SPP
credit policy to protect from exposure such as that experienced in PJM is a priority. The
MMU concurs with their assessment and recommends that changes to the credit policy be
developed to protect stakeholders from risks of default, particularly with respect to
information that is known such as transmission expansion.
8.1.3 DEVELOP COMPENSATION MECHANISM TO PAY FOR CAPACITY TO COVER UNCERTAINTIES
Because of unexpected variations in wind output, SPP operators often needs to manually
commit resources (often in excess of 50 units) in order to meet instantaneous load capacity
178 https://www.spp.org/documents/57928/spp_mmu_asom_2017.pdf, page 141. 179 https://www.spp.org/documents/56330/fca%202017%20report%20-%20final.pdf 180 Per SPP Tariff Attachment X, Section 5A.2.
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 240
requirements. Often, however, these resources do not earn enough revenue to cover their
offered costs181 which contributed to an increase in real-time make-whole payments.182
Moreover, resources providing this reliability service are not compensated specifically for the
need. Even when resources are needed so much that they are committed for capacity, the
supplemental reserve price is still very low. While the MMU recognizes that SPP operators
may need to commit units to account for unforeseen circumstances, manual capacity
commitments occur often enough that systematic solutions should be developed. An ideal
design of a ramping product should allow for most of the capacity and stagger needs to be
met with the compensated ramping product instead of the uncompensated headroom. The
large number of manual capacity commitments may indicate that there is a need for a new
uncertainty product beyond just a short-term ramping product that might better address
problems in the 30-minute to 3-hour time frame.183 Developing products to reduce the need
for the large number of manual capacity commitments would help ensure appropriate
compensation is being provided for the reliability services provided.
The MMU has engaged SPP’s Market Working Group and has participated in a high level
discussion on the potential need for an uncertainty or standby reserve product. SPP and
stakeholders are generally interested in exploring this topic further. The MMU concurs and
recommends that further work be done to develop a compensation mechanism or product to
compensate capacity used to cover uncertainty of generation and load.
8.1.4 ENHANCE ABILITY TO ASSESS A RANGE OF POTENTIAL OUTCOMES IN TRANSMISSION PLANNING
SPP’s transmission planning process develops an annual look-ahead plan. This plan
evaluates transmission needs over a 5- and 10-year time horizon. The plan typically evaluates
181 Offered costs included their incremental energy, no load, start-up and reserve offers. When mitigated, the resource offers are replaced with the mitigated offers. 182 See Section 4.2. 183 The exact time periods of potential uncertainty products should be determined through the technical expertise of the stakeholder process.
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 241
two scenarios.184 The first case is a base case, and the second case is an emerging
technology case. The process could consider a third scenario. In 2018, stakeholders
requested that SPP staff consider an additional case. The third case would have considered a
shift in environmental regulations—such as a carbon tax or adder—and a change in
technologies/market trends including accelerated deployment of storage devices or electric
vehicles, higher levels of generation retirement, and higher penetration of renewables.185
MMU staff follows market trends to keep abreast of developments, including corporate
announcements on carbon reduction and renewable energy targets. Following these trends
helped shape the MMU’s thinking in terms of how to consider the future generation mix and
the potential need for an additional scenario. The MMU believed that a third scenario would
provide a bookend scenario for the 2020 analysis and supported inclusion of the scenario as
part of the 2020 Integrated Transmission Plan. Ultimately, stakeholders rejected the third
scenario at the Economic Studies Working Group, at the Market Operations and Policy
Committee, and at the Strategic Planning Committee. The primary reason cited was the cost
to perform the study.
While the transmission planning process theoretically can include a third scenario, in practice
it does not have the flexibility to include one given that cost of including a third scenario was
not unique to the 2020 planning process. This appears to be a significant shortcoming in the
planning process as the study process is only limited to two potential scenarios in its 10-year
look ahead. This limits the range of potential outcomes that SPP could study.186 We
recommend that SPP enhance their study process to allow the ability to study a range of
potential outcomes. If such range of potential outcomes are not captured in a third case
study, the MMU recommends to factor them into either an existing case study or potentially
as part of the 20-year assessment. The wider the range of possibilities studied, the more
robust the results will be.
8.1.5 IMPROVE REGULATION MILEAGE PRICE FORMATION
In addition to regulation capacity payments, resources that are deployed for regulation also
receive payments for costs incurred when moving from one set point instruction to another.
184 These scenarios are also known as futures cases. 185 This scenario was similar to a scenario used in the MISO transmission plan. https://cdn.misoenergy.org/MTEP19%20Futures%20Summary291183.pdf. 186 The MISO transmission planning process studies four cases.
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 242
These mileage payments are paid directly through regulation-up and regulation down
payments in the day-ahead market. The market calculates a mileage factor for both products
each month that represents the percentage a unit is expected to be deployed compared to
what it cleared. If a unit is deployed more than the expected percentage, then the unit is
entitled to reimbursement for the excess at the regulation mileage marginal clearing price. If
the unit is deployed less, the position must be bought back.
The MMU has identified a mileage price inefficiency. Mileage prices are not set by the
marginal cost of mileage like other products. Instead, units are cleared for regulation based
on what are known as regulation service offers. These service offers are calculated by taking
the competitive offer for regulation and adding the mileage offer to it after discounting the
mileage offer by the mileage factor.
The MMU has observed instances where resources cleared with regulation-down competitive
offers of $0 and mileage offers just under $50. These units consistently cleared with this offer
strategy because the service offer was near $10.50 (e.g. 21 percent*$50) which was lower
than the services offers of other resources offering in higher competitive offers. For instance,
another resource may offer in a $12 competitive offer and $0 mileage offer. This would make
that resource’s service offer $12 ($12+0*21 percent). In this circumstance, the resource with
the highest service offer will set the regulation-down price at $12, but the mileage offer will
be $50, which would set the mileage clearing price.
The MMU is also concerned that the mileage clearing price does not correctly reflect price
formation of mileage given that this price does not represent the expected or realized
mileage deployment. Furthermore, the MMU is concerned that participants with resources
frequently deployed for regulation will have an incentive to inflate the mileage prices by
offering in $0 regulation offers and high mileage offers. The MMU also identified systematic
overpayment of regulation mileage in the day-ahead market, which appears to be the result
of the mileage factor being set consistently too high relative to actual mileage deployed.
The MMU discussed our concerns with the Market Working Group at its August meeting.
While an action item was developed requesting SPP staff and the MMU to review the
effectiveness of the regulation mileage pricing process and present further options, no
additional work has been done since that time. We recommend that SPP staff review the
performance of regulation mileage, and develop potential approaches to improve regulation
mileage price formation. Furthermore, we recommend that SPP staff consider adjusting the
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 243
mileage factor. We believe that SPP staff and stakeholders should include these items as part
of its analysis and change development processes for moving forward.
8.2 CLARIFICATION OF PREVIOUS RECOMMENDATIONS
8.2.1 ADDRESS INEFFICIENCY CAUSED BY SELF-COMMITTED RESOURCES
The MMU and market participants have identified several reasons why they self-commit
resources in the market. Some of these reasons include contract terms for coal plants, low
gas prices that reduce the opportunity for coal units to be economically cleared in the day-
ahead market, long startup times, overtime costs, increased major maintenance costs, and a
risk-averse business practice approach. However, it is imperative to minimize the need to
self-commit resources to realize the full benefits of SPP’s market. While there may not be a
single reason causing market participants to self-commit resources, there can be ways that
SPP and its stakeholders can work to minimize the need to self-commit.
For instance, long lead-time and long run-time resources are often self-committed and
contribute to depressing prices in the SPP market. These resources are not appropriately
evaluated in the current market structure and can be committed by market participants
during uneconomic periods. The current clearing engine logic does not provide
commitments beyond the 24-hour period of the next operating day. The creation of a market
process that economically evaluates resources over a longer period will allow for more
efficient market solutions, as well as decreased production costs.187
In the current design, a resource that is required to run for multiple days is not evaluated by
the day-ahead market to see if the resource is economic over its minimum run-time. The
clearing engine may see that it is economic on the first day and issue the commitment, and
then in future days the resource will stay on until its minimum run-time is met even if it is
uneconomic. As such, many resources that have multi-day minimum run times avoid the
market clearing process and instead self-commit in the market based not on an evaluation by
the market, but on their own evaluation of market conditions. This is not the optimal solution
187 This would be different from the current multi-day reliability unit commitment process.
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 244
for the SPP market as it removes the ability for the SPP market software to evaluate and
commit the resources economically relative to all other resources in the market.
Adding multi-day unit commitment logic is at the top of the current stakeholder market
design initiative list and has been discussed in 2018 and 2019 in the stakeholder process.
SPP staff has proposed a multi-part approach to address multi-day unit commitment. First,
they have indicated that they prefer to provide a multi-day forecast of prices or schedules as
it would be quicker, easier, and less expensive to implement. The multi-day forecast would
serve to provide information to aid market participants that self-schedule to do so in periods
that would be more favorable for their resources and for the market. Second, SPP staff has
indicated that after the multi-day forecast was made available, they would consider
developing a multi-day unit commitment process.
The MMU is currently in the process of reviewing the SPP staff proposal to provide multi-day
forecast information. At this time, it is not clear if the benefits of this approach outweigh costs
and concerns. For instance, the accuracy of the forecast significantly declines the further out
the forecast goes. Thus, it is not clear how useful the information will be to market
participants if the accuracy of the multi-day forecast is low. Furthermore, there have been
several recent statements by multiple SPP market participants indicating corporate goals to
increase renewable generation targets or to set carbon reduction goals over the coming
years.188 This has the potential to change the resource mix such that many of the longer run
resources may no longer participate in the market in future years.
Given these concerns, the MMU clarifies its recommendation. Progress has been made in
addressing self-commitment reasons by allowing the inclusion of major maintenance in
mitigated offers effective April 18, 2019 and changing local mitigation to a threshold of 10
percent effective December 2018. Broadly, the MMU recommends that SPP and its
stakeholders continue to explore and develop ways to reduce the incidence of self-
commitment of resources outside of the market solution. With regards to the development
188 For instance, Omaha Public Power District (OPPD) set a long-term goal of providing at least 50 percent of its retail sales from renewable sources and reducing its carbon intensity by 20 percent from 2010 to 2030. Xcel Energy Inc. announced that it plans to completely decarbonize its power supply portfolio serving eight states by midcentury, first cutting carbon emission by 80 percent by 2030, from 2005 levels in its utility service territories in Colorado, Michigan, Minnesota, New Mexico, North Dakota, South Dakota, Texas and Wisconsin. AEP announced in February 6, 2018 that it will pursue a goal to reduce its CO2 emissions of its generating units 60 percent by 2030 and 80 percent by 2050 from 2000 levels. (See Section 6.2.1 for discussion).
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 245
of multi-day forecasting of prices or schedules or a multi-day unit commitment process, the
MMU supports these as attempts to reduce self-commitment of generation, but we will
carefully weigh the benefits and costs before we support any specific proposal especially in
light of the accuracy of forecasted information and the potential changing generation mix.
We continue, however, to view reducing self-commitment of generation as a high priority for
SPP and its stakeholders as this will enhance market efficiency and improve price signals.
8.2.2 ENHANCE COMMITMENT OF RESOURCES TO INCREASE RAMPING FLEXIBILITY
As noted in last year’s annual report, increasing ramping flexibility is becoming increasingly
important to integrate higher levels of renewable generation.189 For instance, one way to
address ramping flexibility is to ensure sufficient ramping resources are on-line to meet
changing conditions (see Section 8.3.1 below). Over-commitment of resources in real-time
can reduce ramping flexibility, suppresses prices, and leads to increased make-whole
payments. This can be caused by changing conditions between the time a resource is locked
into a commitment by the market software and the time the resource actually comes on-line.
The MMU recommends that SPP and its stakeholders address this issue by enhancing its
markets rules to enhance the commitment of resources to increase rampoing flexibility. Last
year, the MMU described this as enhancing decommitment of resources. However, having
explored this issue further, the issue is not just about decommitment of resources, but is also
about improving how resources are committed.
For instance, the market software frequently commits resources well in advance of when they
are actually required to start. This commitment is based on the known assumptions and
information available at the time the market engine clears the market. However, conditions
change over time. For example, load forecasts, wind forecasts, and outages change, and
resources trip off-line. Resources are committed because the market software evaluates it to
be profitable over that study period. When conditions change, the resource may no longer
be profitable; however, resources may continue to start-up or to run as the commitment is
189 https://www.spp.org/documents/57928/spp_mmu_asom_2017.pdf, page 193.
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 246
not changed or the start-up is not cancelled. This is due, in part, to the current tariff language
that only allows the decommitment of day-ahead market committed resources to prevent
anticipated excess supply conditions or other emergency conditions.
The MMU evaluated a day in June where wind output increased by several thousand
megawatts.190 We identified several quick-start and other short lead time resources that had
day-ahead schedules that ran in real-time.191 Real-time system marginal energy costs were
about $35/MWh less than day-ahead costs during the peak hours. Other markets have
different commitment rules that would help increase market flexibility. For instance, PJM and
California ISO delay the commitment of short start resources and quick-start resources until
the real-time market, making the day-ahead commitment instruction advisory. Furthermore,
the California ISO can also decommit resources after a resource’s minimum run time has
been met, ignoring the day-ahead commitment period. These commitment rules can help
increase the flexibility of resources in SPPs markets and help improve pricing outcomes.
The MMU recommends that SPP and stakeholders explore options, such as those noted
above, to enhance commitment of resources to increase flexibility. We view this as a high
priority item.
8.2.3 FURTHER ENHANCE ALIGNMENT OF PLANNING PROCESSES WITH OPERATIONAL CONDITIONS
Enhancing the accuracy of planning processes with operational realities enables SPP and its
members to more effectively plan for future system needs and conditions. Many of the
challenges outlined in this report—including congestion, negative prices, and low generator
net revenues—as well as improvements—such as transmission additions—are, in part, a
reflection of planning decisions. The more the planning process can learn from and
incorporate operational information, the more planning can identify and address concerns in
190 The MMU evaluated June 12, 2018. Wind was approximately 4,000 MW higher in real-time than expected in the day-ahead market. 191 The day-ahead market committed 25 quick start resources that could be started within 10 minutes or less and an additional 18 resources with start-up times less than four hours.
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 247
advance of market operations. SPP and stakeholders over the past few years have worked to
improve and align the planning and operational processes. For instance, the Integrated
Transmission Planning manual includes an assessment for persistent operational needs and
the criteria for identifying these needs.192 In 2017, the MMU identified two areas where
further alignment would be beneficial. We update the status of these recommends and
clarify this recommendation by adding a new item below.
First, in early 2018, SPP proposed to set the interim default variable operations and
maintenance costs for wind resources to $0/MWh in revision request 276. While the
Economic Studies Working Group rejected this revision request and narrowly passed an
amended version of this request, the Market Operations and Policy Committee reversed
direction and accepted the original SPP proposal. Based on the MMU’s review of operational
information, variable operations and maintenance costs for wind resources are closer to
$0/MWh. The 2019 Integrated Transmission Plan used the $0/MWh value. After reviewing
the results, stakeholders have agreed to continue to use the $0/MWh value going forward.
As such, the MMU believes that this recommendation is complete.
Second, it is important that the full set of generation that exits be considered at peak load
conditions. SPP’s deliverability study evaluates deliverable capacity purchases by a load
responsible entity and does not require firm transmission service to support the
capacity.193,194 Thus, we consider SPP to be accounting for this capacity and not
recommending any further action on this item on the resource adequacy
process. Accordingly, we withdraw this recommendation on this basis.
Third, after engaging in the development of the 2020 Integrated Transmission Plan scope in
2018, one of the key takeaways from our involvement in this process was that some of the
proposed economic modeling assumptions, particularly around generator retirement age
and renewable resource accreditation, differed from prevailing data trends. With regards to
retirement age, the MMU performed an analysis of public Energy Information Administration
192 See SPP Integrated Transmission Planning Manual (dated October 17, 2018), including sections 4.4, 5.3.4, and 6.1.4 (available at https://www.spp.org/Documents/22887/ITP%20Manual%20version%202.3.docx). 193 See SPP Open Access Transmission Tariff, Sixth Revised Volume, Attachment AA - Resource Adequacy, Section 10. 194 Hence, based on the information brought to our attention with regard to SPP’s deliverability study we clarify our earlier point made in last year’s report such that SPP excludes resources without firm transmission for resource adequacy calculation.
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 248
(EIA) data and presented on prevailing trends. With regards to renewable accreditation, we
reviewed other market practices, and SPP’s actual wind and solar production data. This
information can be accessed by SPP staff and the MMU. Notably, SPP staff requested access
to the retirement information used by the MMU and performed an even more detailed
analysis of the EIA data, breaking down natural gas retirements by technology. The MMU
supported this SPP analysis.
Based on our experience in this process, we recommend that SPP staff continue to
collaborate with the MMU to build upon the work in 2018, by reviewing data and best
practices in other markets, factoring in market conditions and trends in decision-making
process, and performing analysis to support and inform assumptions. Doing so will help to
improve assumptions and ultimately study results.
8.3 ADDRESS PREVIOUS RECOMMENDATIONS
The MMU has provided recommendations to improve market design in each of our previous
Annual State of the Market reports since the launch of the integrated marketplace in 2014.
Overall, SPP and its stakeholders have found ways to effectively address many of our
concerns. However, there remain outstanding recommendations. A description of each of
these recommendations, their currents status, and our assessment of their priority are
outlined below.
8.3.1 DEVELOP A RAMPING PRODUCT
A ramping product that incents actual, deliverable flexibility can send appropriate price
signals to value resource flexibility. This resource flexibility can help prepare the system for
fluctuations in both demand and supply that result in transient short-term positive and
negative price spikes.
Today, the SPP real-time dispatch engine solves for only the current interval and has no look-
ahead logic to ensure that there is enough rampable capability to meet the needs of future
intervals. This can cause quick-ramping resources to be dispatched to their maximum limits
in one interval, and then, in the next interval, there is a shortage of ramp because the only
resources able to move have slower ramp rates. In these cases, SPP can have plenty of
capacity on-line, but not enough rampable capacity, which can result in scarcity pricing.
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 249
SPP’s scarcity events are very short-term, transient events that are frequently a result of a
shortage of rampable capacity. A ramping product will compensate resources for holding
back capability in one interval so it can be used as energy in a future interval. This will reduce
the frequency of scarcity events and provide value to the resources providing ramping
capability.
Both the California ISO and Midcontinent ISO have designed and successfully implemented
ramping products. In 2018 and into 2019, SPP, stakeholders, and the MMU have discussed
design elements of a ramping product. SPP staff hopes to have a design approach approved
by mid-year.
We agree with the assessment of SPP and its stakeholders that development of a ramping
product is a high priority initiative and continue to recommend that a ramping product
design be completed in 2019.
8.3.2 ENHANCE MARKET RULES FOR ENERGY STORAGE RESOURCES
With the increase in wind penetration in the SPP market, there is not only a need for resource
flexibility, but also for storage where the frequency of negative prices has increased, as
discussed in Section 4.1.5. Stored energy resources have the potential to address both the
need for flexibility and reduce the incidence of negative prices. However, SPP’s current tariff
does not easily allow these resources to integrate in our market. In order to capture the
benefits of these new technologies, a new market design needs to be developed.
FERC issued Order No. 841 in February 2018 to reduce barriers to participation and to
develop a participation model for electric storage resources.195 Over the course of 2018,
SPP, stakeholders, and the MMU worked on changes that would comply with FERC Order No.
841. These changes passed the October SPP Board meeting and SPP filed the changes with
FERC in December. The MMU filed supportive comments shortly after. In the MMU
195 https://www.ferc.gov/whats-new/comm-meet/2018/021518/E-1.pdf
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 250
comments, we identified multiple areas of further work. These areas include further
enhancements to electric storage integration include addressing the potential for storage
resources to exercise downward market power, the potential for market storage resources
(MSR) to manipulate the transmission market, possible market design gaps regarding major
maintenance and quick-start resource requirements, and the inefficient commitment of non-
continuously dispatchable resource requirements in relation to market storage resources.196
The MMU views integration of storage resources in the SPP markets as an ongoing high
priority as several outstanding items beyond compliance with FERC Order No. 841 need to
be addressed in order to fully integrate electric storage resources in the SPP markets.
8.3.3 ADDRESS INEFFICIENCY WHEN FORECASTED RESOURCES UNDER-SCHEDULE DAY-AHEAD
Our analysis shows that, on average, 85 percent of forecasted wind generation was cleared in
the day-ahead market in 2018, compared to 82 percent in 2017, and 89 percent in 2016. On
average for the year, over 1,000 MW of real-time wind generation was not included in the
day-ahead market in each hour.197 When this happens, we frequently observe day-ahead
prices exceeding real-time prices. While we also observe participants virtual transactions at
wind locations during these times, we find that price convergence in absolute terms is not
improving.198 As a result, under-scheduling of wind reduces the efficiency of the day-ahead
unit commitment process. In this case, other non-wind resources may be overcommitted in
196 Some market storage resources have a non-dispatchable range between their charging range and discharging range. The dispatch calculation for this type of non-continuous dispatch range is much more complicated than the typical linear dispatch calculation. SPP’s current proposal is to commit this type of market storage resource for either charging or discharging. This type of commitment is inefficient because it does not make the whole dispatch range available. For more detail, see Motion to Intervene and Comments of the Southwest Power Pool Market Monitoring Unit, Section I.B.5, Docket No. ER19-460, December 7, 2018. 197 From a reliability standpoint, the reliability unit commitment assesses wind resources at forecasted levels. However, the reliability unit commitment process cannot economically decommit resources scheduled by the day-ahead market. 198 In many cases, we find that virtual transactions are setting prices at wind locations with positive offer prices. In most cases, market participants do not offer wind resources at positive values. As a result, while the virtual transactions may be reflecting missing quantity values at a resource location, they are not reflecting the offer a wind resource would likely have. Moreover, virtual participants may pair virtual offers at resource locations with offsetting virtual bids at other locations. This effectively works to capture congestion differences between the day-ahead and real-time markets, but does not work to converge the energy price component.
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 251
the day-ahead market, which results in real-time prices significantly lower than day-ahead
prices when wind generation increases in the real-time market.
Systematic under-scheduling of wind resources in the day-ahead market can contribute to
distorting market price signals, suppressing real-time prices, and affecting revenue adequacy
for all resources. Therefore, the MMU continues to recommend that SPP and its stakeholders
address this issue through market rules changes that address the consequences of the under-
scheduling of forecasted supply resources in the day-ahead market. We consider this a high
priority as it helps to enhance market efficiency and improve price signals.
8.3.4 CONVERT NON-DISPATCHABLE VARIABLE ENERGY RESOURCES TO DISPATCHABLE
In the 2015 Annual State of the Market report, the MMU identified non-dispatchable variable
energy resources as a concern because of their adverse impact on market price and system
operations. These resources exacerbate congestion, reduce prices for other resources,
increase the magnitude of negative prices, cause the need for market-to-market payments,
and force manual commitments of resources that can increase uplifts. Going forward,
resource flexibility is essential to integrate an increasing volume of wind generation in the
SPP market. FERC demonstrated strong support for the elimination of most instances of non-
dispatchable resources with the approval of a rule change for the New England market in
December 2016.199 Furthermore, FERC also rejected California ISO’s proposal to extend the
transition period for protective measures related to non-dispatchable variable energy, thus
requiring dispatchability.200
The Market Working Group passed a proposal by SPP at the February 2018 meeting that
required full conversion of non-dispatchable variable energy resources. The proposal failed
to pass the Market Operations and Policy Committee meeting in April and was appealed to
the SPP board. The SPP board tabled discussion in April. In July, the Market Worked Group
passed a modified version of the proposal that allowed exemptions for resources exercising
rights under PURPA201 and for run-of-river hydro that could not be dispatchable. This
199 See FERC Docket Nos. ER17-68-000 and ER17-68-001. 200 See FERC Docket No. ER17-1337-000. 201 Public Utilities Regulatory Policies Act of 1978
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 252
subsequently passed the July Market Operations and Policy Committee meeting and SPP
board.
While the MMU preferred the original version of the proposal that was appealed to the April
SPP Board meeting, the MMU found the modified proposal acceptable and supported it at
the July SPP Board meeting and in the MMU comments with FERC.202 FERC approved the
filing in April 2019.
8.3.5 ADDRESS GAMING OPPORTUNITY FOR MULTI-DAY MINIMUM RUN TIME RESOURCES
Resources with minimum run times greater than two days have the opportunity to game the
market. The current implementation of the market rules limit make-whole payments to the
as-committed market offers for the first two days of a resource’s minimum run time. However,
after the second day, no rule exists to limit make-whole payments for a resource that
increases its offers from the third day onward until the resource’s minimum run time is
satisfied. For resources with minimum run times greater than two days, the market
participant knows that the resource is required to run and can increase their market offers
after the second day to increase make-whole payments.
The SPP board passed a proposal203 at the July meeting that would limit make-whole
payments for any resource with multi-day minimum run times to the lower of the market offer
or the mitigated offer. This limitation only applies for offers falling in hours not accessed by
one of the security constrained unit commitment (SCUC) processes and the resource bid at or
above their mitigated offer on the first day. The MMU supported the proposal. Subsequent
to board approval of the proposal, SPP legal staff identified internally inconsistent tariff
language that the revisions revealed by but did not address. This has required that
additional tariff modifications go through the stakeholder process.
The MMU strongly recommends that SPP and its stakeholders clean up the tariff language
necessary to implement the changes to address the gaming issue and provide an anticipated
202 SPP Market Monitoring Unit comments on ER19-356, https://elibrary.ferc.gov/idmws/common/opennat.asp?fileID=15111404. 203 Revision request 306 (2014 ASOM MWP MMU Recommendation [3-Day Minimum Run Time])
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 253
implementation date for the system changes. Addressing this matter in a timely manner is a
high priority for the MMU.
8.3.6 REPLACE DAY-AHEAD MUST OFFER, ADD PHYSICAL WITHHOLDING PROVISION
FERC rejected in 2017 SPP’s proposal to remove the day-ahead must offer requirement and
indicated that it would consider removal of the requirement if it were paired with additional
physical withholding provisions.
While the MMU remains concerned with the current day-ahead must offer requirement, we
recommend that further consideration of this issue be a low priority. The MMU will continue
to track market performance concerns related to this provision and will consider raising the
priority on this matter if further issues are identified or current issues are exacerbated.
Otherwise, we regard other matters as having higher impacts to the market and priority for
development at this time.
8.4 RECOMMENDATIONS UPDATE
The table below lists the status of Annual State of the Market recommendations included in
the 2017 report and those that are new to this report. Recommendations closed prior to the
completion of the previous year’s report do not appear in this table. To review closed
recommendations that are not covered in this report, please review earlier reports. All
previous annual reports can be found at https://www.spp.org/spp-documents-
filings/?id=18512.
Figure 8—1 Annual State of the Market recommendations update
Recommendation Report
year Current status
1. Limit market power by backstopping parameter changes
2018 Awaiting stakeholder engagement
2. Enhance credit process to account for known information
2018 Awaiting stakeholder engagement
3. Develop compensation or product for capacity used for uncertainties
2018 Awaiting stakeholder engagement
4. Enhance ability for transmission planning to cover range of outcomes
2018 Awaiting stakeholder engagement
Southwest Power Pool, Inc. Recommendations Market Monitoring Unit
State of the Market 2018 254
5. Improve regulation mileage price formation
2018 Outstanding Market Working Group action item
6. Reduce self-scheduling in market
2017 Multi-day commitment process – under stakeholder consideration
7. Address under-scheduling of wind
2017 Under consideration by Holistic Integrated Tariff Team
8. Develop ramping product 2017 Under stakeholder consideration 9. Enhance unit commitment
logic 2017 Under stakeholder consideration
10. Develop energy storage design
2017 Awaiting FERC order and follow-up
11. Continue alignment of planning processes with operational conditions (multiple items)
2017
Wind VOM – procedure updated with MMU recommended value, recommendation complete Counting generator capacity – Deliverability study addresses MMU’s concern
12. Local reliability commitment mitigation threshold conversion to a cap
2016 Complete - implemented in December 2018
13. Non-dispatchable variable energy resource transition to dispatchable variable energy resource status
2015 FERC accepted filing in April 2019
14. Improved quick-start logic 2014 Awaiting FERC order
15. Manipulation of make-whole payment provisions (multiple items)
2014
Across midnight hour – awaiting FERC filing Out-of-merit payments – withdrawn by MMU Jointly-owned units – awaiting implementation Regulation – implemented in 2018
16. Day-ahead must offer requirement and physical withholding
2014 Assigned low priority, awaiting stakeholder engagement
17. Allocation of over-collected losses
2014 Awaiting implementation
Southwest Power Pool, Inc. Market Monitoring Unit
State of the Market 2018 255
COMMON ACRONYMS
1073 City of Malden (Mo.) – Board of Public Works
AECC Arkansas Electric Cooperative Corporation
AECI Associated Electric Cooperative, Inc.
AEP/AEPM American Electric Power
ARR auction revenue rights
BEPM Basin Electric Power Cooperative
BRPS Big Rivers Electric Corporation
BSS bilateral settlement schedules
BTU British thermal unit
CC combined-cycle
CDD cooling degree days
CHAN City of Chanute (Kan.)
COSN City of Superior (Nebr.)
COWP City of West Plains (Mo.) – Board of Public Works
CT combustion turbine
CWEP Carthage (Mo.) Water and Electric Plant
DA day-ahead
DAMKT day-ahead market
DA RUC day-ahead reliability unit commitment
DISIS definitive interconnection system impact study
EDE/EDEP Empire District Electric Co.
EHV extra high voltage
EIA Energy Information Administration
EIS energy imbalance service
ERCOT Electric Reliability Council of Texas
ETEC East Texas Electric Cooperative, Inc.
EUDO_X City of Eudora (Kan.)
FCA frequently constrained area
FCU_X Falls City (Nebr.) Utilities
FERC Federal Energy Regulatory Commission
FREM City of Fremont (Nebr.)
GATE_X Gateway Plant
GI generation interconnection
Southwest Power Pool, Inc. Common acronyms Market Monitoring Unit
State of the Market 2018 256
GLDF generator to load distribution factor
GMOC/UCU Greater Missouri Operations Company (KCPL)
GRDA/GRDX Grand River Dam Authority
GSEC Golden Spread Electric Cooperative, Inc.
GW gigawatt
GWh gigawatt hour
HDD heating degree days
HHI Herfindahl-Hirschman Index
HMMU Harlan (Iowa) Municipal Utilities
HVDC high-voltage direct current
IA interconnection agreement
ID RUC intra-day reliability unit commitment
IDC interchange distribution calculator
INDN City of Independence (Mo.)
IOU investor owned utility
IPP independent power producer
IS Integrated System
ISO independent system operator
ITP Integrated Transmission Plan
JOU jointly-owned unit
KBPU Kansas City (Kan.) Board of Public Utilities
KCPL/KCPS Kansas City Power & Light
KMEA Kansas Municipal Energy Agency
KN01 Kennett (Mo.) Board of Public Works
KPP Kansas Power Pool
kV kilovolt (1,000 volts)
LES/LESM Lincoln (Nebr.) Electric System
LIP locational imbalance price
LMP locational marginal price
MCC marginal congestion component
MEAN Municipal Energy Agency of Nebraska
MEC/MECB MidAmerican Energy Company
MEUC Missouri Joint Municipal Electric Utility Commission
MIDW Midwest Energy Inc.
Southwest Power Pool, Inc. Common acronyms Market Monitoring Unit
State of the Market 2018 257
MISO Midcontinent Independent Transmission System Operator
MLC marginal loss component
MM million
MMBtu million British thermal units (1,000,000 Btu)
MMPA_X Minnesota Municipal Power Agency
MMU Market Monitoring Unit
MW megawatt (1,000,000 watts)
MWh megawatt hour
MWP make-whole payment
MRES Missouri River Energy Services
MUMZ_X Missouri River Energy Services
NCU_X Nebraska City (Nebr.) Utilities
NDVER non-dispatchable variable energy resource
NELI_X City of Neligh (Nebr.)
NERC North American Electric Reliability Corporation
NOAA National Oceanic and Atmospheric Administration
NPPD/NPPM Nebraska Public Power District
NSP/NSPP Northern States Power Energy
NWMT_X Northwestern Energy
NWPS Northwestern Energy
O&M operation and maintenance
OGE Oklahoma Gas & Electric
OMPA Oklahoma Municipal Power Authority
OOME out-of-merit energy
OPPD/OPPM Omaha Public Power District
OTPW Otter Tail Power Company
OTP_X Otter Tail Power Company
OTPR_X Otter Tail Power Company
PARL City of Piggott (Ark.) Municipal Light, Water, and Sewer
PBEL City of Poplar Bluff (Mo.) Municipal Utilities
PEOP_X Peoples Electric Cooperative
PEPL Panhandle Eastern Pipe Line
PISIS preliminary interconnection system impact study
PJM PJM Interconnection, LLC
Southwest Power Pool, Inc. Common acronyms Market Monitoring Unit
State of the Market 2018 258
PLWC Paragould (Ark.) Light & Water Commission
RC reliability coordinator
REMC Rainbow Energy Marketing Corporation
RNU revenue neutrality uplift
RR revision request
RSG reserve sharing group
RT real-time
RTBM real-time balancing market
RTO regional transmission organization
RUC reliability unit commitment
SC simple-cycle
SECI/SEPC Sunflower Electric Power Corporation
SPA Southwestern Power Administration
SPP Southwest Power Pool, Inc.
SPRM City Utilities of Springfield (Mo.)
SPS Southwestern Public Service Company
ST steam turbine
ST RUC short-term reliability unit commitment
TCR transmission congestion right
TEA The Energy Authority
TNGI_X City of Grand Island (Nebr.)
TNHP_X Heartland Consumers Power District
TNHU_X City of Hastings (Nebr.) Utilities
TNSK Tenaska Power Service Company
UGPM Western Area Power Administration, Upper Great Plains
WAPA Western Area Power Administration
WECC Western Electricity Coordinating Council
WFEC/WFES Western Farmers Electric Cooperative
WR/WRGS Westar Energy, Incorporated