Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
ERCOT Demand Response
Paul Wattles
Senior Analyst, Market Design & Development, ERCOT
Whitacre College of Engineering, Texas Tech University
April 6, 2012
North American Bulk Power Grids
• The ERCOT Region is one of 3 grid interconnections in USA-Canada
• The ERCOT grid:
– Covers 75% of Texas land
– Serves 85% of Texas
2 Texas Tech College of EngineeringApril 6, 2012
– Serves 85% of Texas load
– >40,000 miles of transmission lines
– >550 generation units
– Physical assets are owned by transmission providers and generators, including municipal utilities and cooperativesERCOT connections to other grids are limited to direct current (DC)
ties, which allow control over flow of electricity
North American ISOs and RTOs
Independent System
Operators and Regional Transmission Organizations
are the ‘air traffic
controllers’ of
3 Texas Tech College of EngineeringApril 6, 2012
controllers’ of the bulk
electric power grids
Peak Load Fuel Mix
40000
50000
60000
70000
80000
August 3, 2011 Generation by Fuel
Gas Wind Other* Coal Nuclear
*Other includes Hydro, Biomass, Solar, and unknown sources
>28 GW
4 Texas Tech College of EngineeringApril 6, 2012
0
10000
20000
30000
1:0
0
2:0
0
3:0
0
4:0
0
5:0
0
6:0
0
7:0
0
8:0
0
9:0
0
10
:00
11
:00
12
:00
13
:00
14
:00
15
:00
16
:00
17
:00
18
:00
19
:00
20
:00
21
:00
22
:00
23
:00
0:0
0
Off-peak vs. on-peak load by customer type
Residential
51.2%
(~35,000 MW)
Wed., Aug. 3, 20115:00 PMERCOT Load: 68,416 MWTemperature in Dallas: 109°
WednesdayMarch 9, 20115:15 PMERCOT Load: 31,262 MWTemperature in Dallas: 64°
5 Texas Tech College of EngineeringApril 6, 2012
• Customer class breakdown is for competitive choice areas; percentages are extrapolated for munis and co-ops to achieve region-wide estimate
• Large C&I are IDR Meter Required (>700kW)
8/3/2011 IE 17:003/9/2011 IE 17:15
Large C&I
23.7%
Residential 27.4%
(~8,500 MW) Small Commercial
25.2%
Small Commercial
28.9%
Large C&I
43.7%
One more way of looking at ERCOT
26.3%
NOIE Retail Choice
Competitive Choice vs. Muni/Co-op Load
Muni’s and co-ops are still vertically integrated
Many have existing and developing smart grid initiatives:
Dozens of REPs competing for residential and small commercial accounts
Mostly flat-price options including pre-paid, renewables
6 Texas Tech College of EngineeringApril 6, 2012
73.7%
smart grid initiatives:-- AMI-- Smart thermostats-- Other DLC
Possible triggers: Real-time prices, congestion management, 4CP response
paid, renewables
AMI deployment underway
‘Utility’ in the traditional sense in no longer a valid term
MWh 12/1/10 thru 8/11/11
Resource Adequacy
13.75% target
7 Texas Tech College of EngineeringApril 6, 2012
* Projects undergoing full Interconnection studies - may be cancelled or delayed beyond the commercial dates shown*** Monticello 1&2 – 1130MW (as a result of a federal court’s order to stay EPA’s CSAPR)
Definitions of Demand Response
• ‘The short-term adjustment of energy use by consumers in response to price changes or incentives.’ (FERC)
• ‘Changes in electric use by demand-side resources from their normal consumption patterns in response to changes in the price of electricity, or to incentive payments designed to induce lower electricity use at times of high wholesale market
8 Texas Tech College of EngineeringApril 6, 2012
induce lower electricity use at times of high wholesale market prices or when system reliability is jeopardized.’ (NERC)
• ‘A temporary change in electricity consumption by a Demand Resource in response to market or reliability conditions.’ (NAESB)
Definitions of Demand Response
• The common threads:
– Change in Load
– In response to a signal (economic or operational)
• 3 key questions relative to any DR:
1. What is the incentive?
9 Texas Tech College of EngineeringApril 6, 2012
1. What is the incentive?
2. What is the signal?
3. Who takes the action (pushes the button)?
DR potential in ERCOT
• FERC estimates >18 GW of DR potential in Texas by 2019
– Attributed to high peak demand
– This would represent 20-25% of total ERCOT peak
10 Texas Tech College of EngineeringApril 6, 2012
Source: FERC 2009 National Assessment of DR, page 42
ERCOT DR Summary
Load Type Service Requirements Description/Notes
Voluntary Load Response (VLR)
Load reduction in response to Market Price, 4CP signals or other factors
• Metering • Load reduction technology • Retail contract with demand response
incentives
• Economic: not dispatched by or reported to ERCOT ISO
• May include dynamic pricing (Time of Use, Critical Peak, Real-Time Pricing) and/or centrally dispatched load control
Load Resources (LRs)
Responsive Reserves
• Interval metering• Telemetry• Under-Frequency Relay• Load reduction technology• ERCOT Qualification
• Industrial Loads • 197 LRs with ~2400 MW of total
registered capacity• Limited to 50% of total RRS (1150 of
2300 MW) Dispatched during Energy Emergency Alert (EEA) or
11 Texas Tech College of EngineeringApril 6, 2012
• ERCOT Qualification Emergency Alert (EEA) or automatically due to frequency drop
Controllable Load Resources (CLRs)
Regulation ServiceResponsive Reserves
• Interval metering• Telemetry• Ability to receive AGC-type signals• Governor-type frequency response• ERCOT Qualification
• Industrial Loads with sophisticated control systems and ramping capability
• 1 CLR (~20 MW) currently enrolled
Emergency Interruptible Load Service (EILS)
10-minute special emergency DR service
• Interval metering• Load reduction technology• ERCOT Qualification
• Mid- to large commercial & industrial Loads
• Procured 3 times per year for 4-month Contract Periods
• Dispatched during EEA• ~450 MW enrolled depending on Time
Period
DR in ERCOT today
• Operational DR (these MW we know):
– 207 Load Resources with 2400+ MW of registered DR capacity• Participation in Responsive Reserves capped at 1400 MW
– Minimal participation by Controllable Load Resources in RRS (not capped) and Regulation service
– 475 MW of participating EILS from ~900 Load sites
– ~150 MW enrolled in TDSP Load Management SOPs• Summer peak hours only; some overlap with EILS
12
• Economic DR (these MW we don’t):
– Load curtailing in anticipation of 4CP intervals• 11,000+ IDR-metered Loads subject to tariffs
• Behavior is well-baked into ERCOT load forecasting
• Initiated by customer or thru direct load control by REP/third party
– Real-time & critical peak price response• Initiated by customer or thru direct load control by REP/third party
– Time of Use
Texas Tech College of EngineeringApril 6, 2012
DR incentives, signals and actions
DR Type Incentive Signal Action-taker
Load Resources providing RRS
Day-ahead Ancillary Service market clearing prices
ERCOT dispatch (EEA or frequency recovery), UFR trip
QSE (via SCADA or direct load control), UFR (automatic), Load personnel
EILS (ERS)4-month capacity payment
ERCOT dispatch (EEA)
QSE (via SCADA or direct load control), Load personnel
13 Texas Tech College of EngineeringApril 6, 2012
4CP Response TCOS avoidanceProbability of a 4CP interval
Load personnel (IDR-metered Loads, NOIEs) or 3rd party
Critical Peak Response
High price avoidance or rebatefrom LSE
Real-time Load Zone prices above a level specified in contract
Load personnel, or LSE or 3rd party via direct load control
DR incentives, signals and actions (continued)
DR Type Incentive Signal Action-taker
Real-TimePricing
Energy price avoidance
Probability of high Load Zone price -- signal provided by LSE or 3rd
party
Load personnel, or LSE or 3rd party via direct load control
TDSP Load Mgmt. Std.Offer Programs
Capacity paymentInstruction from TDSP(coordinated with ERCOT EEA)
Load personnel or aggregator
14 Texas Tech College of EngineeringApril 6, 2012
Time of Use
Potential to save by using lower priced off-peak power
Prices change at known time of day
Load (behavioralshift)
Load Resource deployments since 2006
Day Date Time Type of Deployment Season
1 Mon 4/17/2006 15:34 EECP Step 2 Systemwide Verbal Dispatch Instruction (VDI) Spring
2 Tue 10/3/2006 17:37 Systemwide VDI for frequency restoration Fall
3 Fri 12/22/2006 2:54 Under-frequency (UF) Event followed by VDI for frequency restoration Winter
4 Mon 7/2/2007 19:38 Systemwide VDI for frequency restoration Summer
5 Wed 9/5/2007 7:57 Systemwide VDI for frequency restoration Summer
6 Wed 12/12/2007 1:56 Systemwide VDI for frequency restoration Winter
7 Tue 2/26/2008 18:49 EECP Step 2 Systemwide VDI Winter
8 Sun 3/16/2008 11:37 UF Event, frequency < 59.7 Hz Spring
9 Mon 8/11/2008 17:14 Systemwide VDI for frequency restoration Summer
10 Tue 12/16/2008 15:49 Systemwide VDI for frequency restoration Winter
15 Texas Tech College of EngineeringApril 6, 2012
10 Tue 12/16/2008 15:49 Systemwide VDI for frequency restoration Winter
11 Sat 1/9/2010 10:32 Systemwide VDI for frequency restoration Winter
12 Sat 5/15/2010 16:14 UF Event, frequency < 59.7 Hz Spring
13 Wed 6/23/2010 15:20 UF Event followed by VDI to selected QSEs for frequency restoration Summer
14 Fri 8/20/2010 15:28 Systemwide VDI for frequency restoration Summer
15 Wed 11/3/2010 10:21 UF Event followed by VDI to selected QSEs for frequency restoration Fall
16 Wed 2/2/2011 5:20 EEA Level 2A Systemwide VDI Winter
17 Wed 3/23/2011 14:47 UF Event (partial), frequency dropped to near 59.7 Hz Spring
18 Thu 8/4/2011 14:32 EEA Level 2A Systemwide VDI Summer
19 Wed 8/24/2011 15:11 EEA Level 2A Systemwide VDI Summer
20 Tue 11/29/2011 3:29 UF Event, frequency < 59.7 Hz Fall
Peak vs. off-peak DR
• Unlike many ISO products, ERCOT DR is not designed specifically to reduce summer peak demand
– Responsive Reserves and EILS/ERS are procured 24/7/365
• Of 20 Load Resource deployments since 2006:
– 5 occurred during summer peak hours (3-7 PM weekdays, June through September)
– 6 occurred during winter months
16
– 9 occurred during non-business hours (overnights or weekends)
• However, 8 deployments were for frequency restoration
– Typically caused by generation unit trips
– Nodal market design (implemented 12/1/10) has reduced the need for deployments of this type
• Due to thinning reserve margins, ERCOT now views DR resources as important element for meeting peak load
Texas Tech College of EngineeringApril 6, 2012
EILS
• Emergency Interruptible Load Service has been an ERCOT operational tool since 2008
– Created through detailed PUC Rule with goal of enhancing reliability and opening DR to loads of all types
– Deployed only in late stages of grid emergency prior to firm load shedding (rotating outages)
– Open to individual Loads or aggregations of Loads
17
– Procured via RFP for four-month contract periods
– Offers selected by ERCOT management
– Participation has grown from 0 to >450 MW
• 2 EILS deployments in 2011
– One just prior to winter peak (Feb. 2), related to large number of unit trips
– One during summer peak (Aug. 4)
Texas Tech College of EngineeringApril 6, 2012
EILS → ERS
• Following summer of 2011, resource adequacy concerns prompted ERCOT and PUC to seek ways to increase EILS participation
– Goal: more capacity under contract by summer 2012
• Rule amendments adopted in March; follow-on Protocol revisions are scheduled for 6/1/12 implementation
• New name: Emergency Response Service (ERS)
18
– Now open to participation from unregistered generators (such as distributed generation) capable of injecting power to the grid
– Allows for contract period renewals to deal with sustained weather/grid reliability events
• Feb. 2-3 event in 2011 lasted 28 hours
– Minimum offer reduced from 1 MW to 0.1 MW – will allow more flexibility in assembling aggregations
Texas Tech College of EngineeringApril 6, 2012
Smart Grid
19
Smart grid at the bulk power level: ‘Macomber Map’
• ERCOT identified the need for situation awareness and improved user experience in control rooms
• Main issues:
– Need for wide-area and high-level views
– “Alarm storm”
– “Dust storm”
20 ERCOT InternalFeb. 22, 2012
– “Dust storm”
– Locations of equipment
– Navigation between one-lines
– Naming conventions for equipment
– Data integration across multiple systems
– Increased cognitive load as operators increasingly managing more tasks with more complex systems.
• This trend is only continuing (e.g., AMI)
The ‘Macomber Map’
• Visualization tool used by ERCOT operators for situational awareness
• Allows operators to see all data for single element (e.g., a generator), across applications and systems
• Used for many ERCOT applications:
– Control room
– Network modeling group
21
– Network modeling group
– Congestion Revenue Rights
– Market Information System (contour map of LMPs and SPPs)
– Training
– Engineering
• Developed in-house; named for one of the developers who passed away in 2008
ERCOT InternalFeb. 22, 2012
Smart grid at the bulk power level: Synchrophasors
• Joint project of ERCOT, Transmission Owners (AEP, Oncor, Sharyland) and the Center for the Commercialization of Electric Technologies (CCET)
• Initiated in Fall 2008 with 3 AEP PMUs
• Currently 14 Phasor Measurement Units (PMUs) installed at 12 TO locations
– Real-Time Dynamic Monitoring System (RTDMS) at ERCOT
• Expanding to 23 PMUs, provide TOs access to RTDMS
22
• Expanding to 23 PMUs, provide TOs access to RTDMS visualization applications
• GOALS:
– Provide real-time dynamic information
– Identify precursor conditions to undesirable grid performance and behavior
– Identify changes in operating procedures or actions to facilitate integration of phasor measurements, hence improving grid reliability
– Recalibrate dynamic models
ERCOT InternalFeb. 22, 2012
What the Phasors Do
• Phasor measurement technology offers a means of capturing wide-area snapshots of phase angle profiles and comparing them.
23Feb. 22, 2012 ERCOT Internal
ERCOT Phasor Locations
24Feb. 22, 2012 ERCOT Internal
Benefits of Phasor Measurements
• Direct measurement phasor angles
– In EMS, phasor (voltage or current) angles only can be calculated by State Estimator, which runs every 5 mins
– PMUs continuously measure voltage/current angles with high sampling rate (≥ 30 samples/s)
• High sampling rate
– Conventional SCADA measurements update in seconds
25
– Conventional SCADA measurements update in seconds
– PMUs typically take 30 observations per second
• Voltage phasor angle difference can be used as a good indicator of grid stress
Feb. 22, 2012 ERCOT Internal
Benefits of Phasor measurements
• Phasors are valuable tools for validation of stability studies and other real-time and after-the-fact evaluations of grid events
– Example: Real-time oscillations experienced at a unit during testing were monitored and reported back to ERCOT
– Max output level then established for the unit causing the oscillation
• Phasor data has also been helpful in ERCOT’s evaluations of Load Resource under-frequency relay trips
– Supplement to the high-speed frequency recorders present at
26
– Supplement to the high-speed frequency recorders present at ERCOT Control Centers
– Retrieving frequency data from multiple points across the System gives ERCOT a more complete view of an event
– Example: when a UFR event trips only part of the LR fleet, there may be more than one reason:
• Some LRs have their UFRs set slightly higher than 59.7 Hz requirement
• Frequency may not dip below 59.7Hz for full 20 cycles at all points on the System
Texas Tech College of EngineeringApril 6, 2012
Smart grid at the distribution level
• Improved elasticity of demand among small customers
– Advanced metering
– Home Area Networks
– Load control devices
– Dynamic price offerings
• A ‘self-healing’ grid that responds dynamically to shortages,
27
• A ‘self-healing’ grid that responds dynamically to shortages, disturbances or high price events
– Automatic response to signals at the device level
• ‘Demand Response 2.0’
ERCOT InternalFeb. 22, 2012
Smart Grid initiatives
• Municipals and Co-ops (24% of ERCOT Load) have a number of existing and developing smart grid initiatives
– AMI deployments
– Smart thermostats
– Other DLC
• Investor-owned TDSPs serving competitive-choice areas are over 75% of the way toward eventual deployment of 6 million-plus advanced meters
28
plus advanced meters
• As of early April:
– TDSPs have installed nearly 5 million advanced meters
– ERCOT keeping pace, settling on 15-minute data
• Robust TDSP features:
– Meter-reads-on-demand enhance retail switching
– Automatic outage detection
– Remote connect/disconnect
Texas Tech College of EngineeringApril 6, 2012
AMI and the retail market
• Smart Meter Texas portal has a low sign-up rate among customers with advanced meters
• Customer education initiatives and dynamic price offerings are scarce
– Vast majority of AMI customers are still buying flat-priced electricity
• Home Area Networks are still in pilot stage
29
• Low participation by REPs and aggregators in the AMIT process
• REP investments in customer smart-grid tools subject to:
– Risk of losing the customer to switching: potential stranded costs
– Lack of financial incentives due to low flat electricity prices
• $2 natural gas
Texas Tech College of EngineeringApril 6, 2012
Path to the Smart Grid
TDSP Back Office
15-minute Settlement
REP Back Office
Customer Education
Incentives
30
Office EducationIncentives
Home Area Network
Load Control Equipment
Texas Tech College of EngineeringApril 6, 2012
Relevant questions
• How many REPs will build AMI DR portfolios?
– Does DR capability = smarter energy consumers = enhanced customer loyalty?
• How much mass market DR potential is out there?
– FERC estimate is simply a calculation of peak demand, not a projection of adoption rates
• How fast will it develop?
31
• Who is most likely to push the button?
– Customer, REP, third party?
• Will shortage/scarcity conditions result in high prices?
• Will the signals all be economic?
– What energy prices (i.e., natural gas) are needed to stimulate DR?
– Will mass market AMI customers enroll in ERCOT DR services?
Texas Tech College of EngineeringApril 6, 2012
DR and intermittent renewables
• Wind integration is an ongoing initiative at ERCOT
– >10,000 MW of wind on a 68,000 MW system
– Wind blows better off peak, as you all know in West Texas
– Wind ramps tend to coincide with load ramps – creating ‘net load’ effect
• Feb. 2008 LR deployment helped recover from a net load event
• Recommendations from study of wind integration by GE (2009):
32
• Recommendations from study of wind integration by GE (2009):
– Improve wind forecasting
– Increase procurement of Regulation service and possibly Nonspin during hours with likely net load events
– Consider a new 15-minute Ancillary Service
• New products could potentially accommodate DR participation as well as generator participation
Texas Tech College of EngineeringApril 6, 2012
Challenges for ERCOT DR
• ERCOT energy-only market design
– DR in the eastern ISOs has grown much faster due to forward (long-term) capacity markets
– Energy-only market success depends on very high prices during scarcity condtions
• This creates opposition to DR due to tendency to suppress prices
• Price response hindered by:
– Inability for DR to compete with generation in the real-time energy
33
– Inability for DR to compete with generation in the real-time energy market, SCED (Security Constrained Economic Dispatch)
– Lack of advance notification of prices paid by Loads
– Both of these issues being addressed but will take years • ‘Nodal 2.0’
• Current ERCOT DR products require 10-minute response
– ERCOT will pilot a 30-minute DR product this summer, assuming PUC adoption of Rule enabling pilot projects
Texas Tech College of EngineeringApril 6, 2012
Lessons Learned
• Load Resource participation in Ancillary Services has matured into an extremely valuable operational tool
– However, participation drops off during high peaks, for a number of reasons including price/4CP response
• EILS growing pains:
– Harsh suspension provisions – Loads subject to 6-month suspensions for narrow failures of availability
34
• This has kept dozens of MWs sidelined at any time
– 28-hour deployment in Feb. 2011 has made marketing the product very difficult for aggregators
– Many issues addressed in new Rule/Protocols
• Low natural gas = cheap electricity = diluted incentives for retail DR/load management products
– Will policymakers need to jump-start the smart grid?
Texas Tech College of EngineeringApril 6, 2012
ON
Questions?
35 Texas Tech College of EngineeringApril 6, 2012
OFF