NERC Load Modeling Activities Load... · 4 RELIABILITY | ACCOUNTABILITY Why Dynamic Load Modeling? • Validation of power‐voltage oscillations WECC: July 2 and August 10 1996,

NERC Load Modeling ActivitiesRyan D. Quint, PhD, PESenior Engineer, System Analysis, NERCMRO Fall Reliability ConferenceNovember 2016

RELIABILITY | ACCOUNTABILITY2

Introduction

• Kickoff January 2016• LMTF webpage• Chair: Dmitry Kosterev, BPA• Focus Areas

Coordination with Regional load modeling groups Broad expertise involvement Full utility industry coverage and participation Consolidate and share load modeling practices across industry

• Tasks Support industry‐wide advancement of dynamic load modeling Develop guidelines, technical references, industry webinars, etc. Help ensure robust software implementation Share lessons learned and study approaches


History & Background

Why is this important?How did we get to where we are today?


Why Dynamic Load Modeling?

• Validation of power‐voltage oscillations WECC: July 2 and August 10 1996,

August 4 2000

• Fault‐Induced Delayed Voltage Recovery (FIDVR) Observed as early as 1980s in Southern

California, Florida, Georgia, mid‐West Related to stalling of residential air‐

conditioners

• Distributed energy resources Emerging need in early 2010s

400

420

440

460

480

500

520

540

560

‐10 0 10 20 30 40


Air Conditioner Testing Round 1

• SCE, EPRI, and BPA tested single‐phase residential A/C units Voltage sags, ramps, oscillations,

frequency excursions

• Stall for sudden ΔV to 50‐60% of nominal in less than 3 cycles

• Once stalled, they remain stalled Cannot overcome load torque ‐

coolant pressure must equalize

• Reactive power up to ~ 7x rated• Thermal protection trips in 2‐30

seconds


Air Conditioner Testing Round 1

80 85 90 95 100 105 110 1152.6

2.8

3

3.2

3.4

3.6

Pow

er (k

W)

Ambient Temperature (F)80 85 90 95 100 105 110 115

0.56

0.58

0.6

0.62

0.64

0.66

Sta

ll V

olta

ge (p

er u

nit)


Development of Initial Performance Model

• Motor stalls when voltage drops below Vstall for duration Tstall Vstall ~ 0.52‐0.6 pu Tstall ~ 0.033 sec

0 50 100 150 2000

2000

4000

6000

8000

10000

12000

Voltage [V]

Com

pres

sor R

eal P

ower

[W]

Real Power

RUN

STALL

STALL

115F110F105F100F95F90F85F80F

0 50 100 150 2000

2000

4000

6000

8000

10000

12000

Voltage [V]

Com

pres

sor R

eact

ive

Pow

er [V

AR

]

Reactive Power

RUN

STALL

STALL

115F110F105F100F95F90F85F80F

0 0.2 0.4 0.6 0.8 1 1.20

1

2

3

4

5

6Real Power

Rea

l Pow

er (p

er u

nit)

Voltage (per unit)

RUNSTALL

STALL

Source: BPA


Electromagnetic Transient Simulations

Negative peak of electrical torque ~8x rated torque

Speed is pulled down very strongly by negative Telec

Current by stalled motor 5x rated current

5 kW 1‐ph A/CH = 0.048 sec

Source: J. Undrill


Electromagnetic Transient Simulations

Negative peak of electrical torque ~1‐2x rated torque

Speed minimally pulled down by negative Telec

Current returns to near rated current

100 kVA 3‐ph MotorH = 0.3 sec

Source: J. Undrill


Point-on Wave Simulations

• “Common mode failure” of testing Applied voltage sag at point‐on‐wave

zero crossing in every test

• Instantaneous voltage drop to 0.62 pu for 3 cycles Voltage zero crossing Voltage peak Voltage 45 deg point

• Worst case – zero crossing

Source: Univ. Wisconsin


Stall Voltage vs. Point on WaveSimulation

Source: Univ. Wisconsin


Stall Voltage vs. Point on WaveTesting

Source: Univ. Wisconsin, BPA


Scroll vs. Reciprocating

Scroll Compressors• Vast majority of new compressors• Better fault ride‐through ability• May run backwards after fault

~1‐1.25x rated current – not locked rotor – up to tens of minutes

• Estimated to be ~ 50% of A/C fleet today (2015 NERC FIDVR Workshop)

Reciprocating Compressors• Majority of fleet until 2000s• Disappearing due to energy

efficiency requirements

Source: BPA


Meeting the Needs of TODAYfor Dynamic Load Modeling

Addressing issues and practiceswith the existing dynamic load models


Technical Reference Document

• Documents current state of dynamic load modeling• Follow‐up to the NERC FIDVR Workshop in September 2015• Final NERC PC approval and publication in December 2016


Network Boundary Equations & Initialization

Source: PowerWorld

• Standardized procedures for software initialization• Familiarize and standardize practices for dealing with current

sources in dynamics – motor model numerical issues• Overcome “crashing” issues


Common Initialization & Network Boundary Equations

Source: PowerWorld


Load Model Benchmarking

• Initial LMTF member and vendor benchmarking• NERC LMTF default data set tested using standard test events• Results being compiled to identify any discrepancies• Fixing any software implementation issues identified

Voltage Flow (Measured at From End)Bus Vmag [pu] From Bus To Bus Flow MW Flow MVar1 1.020 102 101 165.0 82

101 1.020 101 1 165.3 90.1102 0.999


Load Model Benchmarking

-20 0 20 40 60 80 100 1203.545

3.55

3.555

3.56

3.565

3.57

3.575

3.58

3.585

3.59

P-M1----

MOTOR A P (PU ON SYSTEM MVA BAS

PTIGE

Source: PacifiCorp

Source: PowerWorld4 6 8 10 12 14 16 18 20

Time [s]

159

160

161

162

163

164

165

166

P lo

ad

[MW

]

P load[MW]

PSLFPSSEPowerWorldTSAT


Robust Default Data Sets

• Developed robust default data sets for use across regions as starting point Suitable and reasonable parameters for tripping Can be modified with regional data, composition data


Reliability Guideline: Load Composition

• Guideline on developing representative load composition data for dynamic load models

Source: WECC


Reliability Guideline: Load Composition

• Guideline on developing representative load composition data for dynamic load models


Reliability Guideline:DER in Dynamic Load Models

• Guidance on modeling DER in dynamic load models and powerflow models

• Coordination with DERTF efforts



• Utility‐Scale Distributed Energy Resources (U‐DER): distributed energy resources directly connected to the distribution bus or connected to the distribution bus through a dedicated, non‐load serving feeder. These resources are specifically three‐phase interconnections, and can range in capacity, for example, from 0.5 to 20 MW although facility ratings can differ.

• Retail‐Scale Distributed Energy Resources (R‐DER): distributed energy resources that offset customer load. These DER include residential, commercial, and industrial customers. Typically, the residential units are single‐phase while the commercial and industrial units can be single‐ or three‐phase facilities.





Source: PowerWorld


Industry-Wide Webinar

• Broader industry engagement• Topics to cover

Fundamentals of end‐use load Composite load model Parameters and their meaning Load composition data Distributed energy resource modeling Sensitivity analysis

• December 9 – announcement to be made shortly


Utility Forum: Dynamic Load Modeling


Meeting the Needs of TOMORROWfor Dynamic Load Modeling

Addressing issues and practiceswith future dynamic load models


Progressive Stalling and Tripping

Source: PowerWorld

Source: J. Undrill


Efficient Data Format & Model Management

• Modularized load model structure• Plug ‘n play concept• Distinction between network and load components• Explicit DER representation• Data input groups


Dynamic Load Modeling in Real-Time Stability Analysis

• Why do we ignore real‐time modeling practices?

• Why do we require induction motor load in planning studies but not in real‐time studies?

• What are the limitations in moving towards inclusion of induction motor load in real‐time models?

• How do we proceed cautiously?
