Webinar 4 - Aggregated Slides

Download as pdf or txt
Download as pdf or txt
You are on page 1of 78

Inverter Based Resources

Short-Circuit Modeling and


Their Impact on System
Protection

Evangelos Farantatos, Ph.D.


Technical Leader, Grid Operations & Planning
EPRI

Inverter-Based Resources Webinar Series


Session IV
May 17, 2018
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Outline
1. Motivation, Challenges and Needs
2. Wind/PV Short Circuit Models
3. Protection Guidelines for Systems
with High Levels of Inverter Based
Resources
4. Related Industry Activities

2
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Motivation, Challenges & Needs
• Continuously increasing penetration level of inverter
interfaced resources, predominantly renewables (Type
III, Type IV WTGs & PVs)

Challenge

• Complex fault response


• Differs significantly from synchronous short-circuit
current contribution (SCC)

Impact on System Protection


• Accurate short-circuit models for protection/planning
studies
• Performance of legacy protection schemes (distance
protection etc)
3
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Inverter Based Resources Fault Response Characteristics
Synchronous Generator

• SCC close to nominal load


current (typically 1.1-1.5
pu)
• Typically low/zero
negative sequence
contribution
Type IV WTG • No zero sequence
contribution
• Fault response depends
on WTG/PV inverter
control scheme

4
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Wind/PV Short Circuit Models

5
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Inverter Based Resources Short-Circuit Modeling

Synchronous generator classical short circuit model


(voltage source behind an impedance) is not applicable

IEEE PSRC WG C24 “Modification of Commercial Fault


Calculation Programs for Wind Turbine Generators”

6
© 2018 Electric Power Research Institute, Inc. All rights reserved.
EPRI Phasor Domain Model for SC Calculations
Type IV WTG Type III WTG Solar PV
Wind Wind
turbine Stator power
turbine Grid
Stator-Side Grid-Side Gearbox
Slip rings
Gearbox Converter Converter IL , P L Transformer

iPMSG ig Chopper

Grid
Rotor power
Step down Crowbar
Rotor -Side Grid -Side
transformer Converter Converter

Phasor Domain Short Circuit Model

WTG (Type III & Type IV) and Solar/PV phasor domain short circuit model:
• Voltage controlled current source
• Iterative solution (nonlinear behavior)
• considers the impact of controls (reactive power/voltage control) on the short circuit response
• respects converter current limits
7
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Inverter Control Mode Options
Function Control Mode Performance/Description
Constant power factor Allows for inverter injection/absorption of
reactive power based on a desired power
factor
Constant Q Allows for inverter fixed desired value of
Reactive power/voltage
reactive power injection/absorption
control during ride-
through V Control Allows for inverter control of voltage to desired
value
Dynamic reactive current control based on Allows for reactive current injection based on a
reference curve reference curve (e.g. grid code)

1) Control mode defines desired active &


reactive current
2) Then current limiter is applied (P or Q
priority)

8
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Current Limiter - PQ Priority - Examples
Example 1:
Assume: Desired Currents: Upon current limiter:
Active Power: 1 p.u. Iactive= 1/0.7=1.43 p.u Iactive= 0.92 (reduced to
Post fault voltage: 0.7 pu satisfy limit)
Ireactive=2(1-0.7) = 0.6 p.u
Control mode: Reference curve with slope 2 Ireactive= 0.6 p.u
Itotal=1.55 pu (exceeds
Q priority Itotal= 1.1 pu
limit)
Ilimit=1.1 pu

Example 2:
Upon current limiter:
Assume: Desired Currents:
Iactive= 0 (reduced to satisfy
Active Power: 1 p.u. Iactive= 1/0.4=2.5 p.u
limit)
Post fault voltage: 0.4 pu Ireactive=2(1-0.4) = 1.2 p.u
Ireactive= 1.1 p.u (reduced
Control mode: Reference curve with slope 2
Itotal=2.77 pu (exceeds to satisfy limit)
Q priority limit)
Itotal= 1.1 pu
Ilimit=1.1 pu
9
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Demonstrating Results Type IV WTG - LLG fault (AB) - BUS 1

Type III WTG - LL fault (AB) - BUS 4

• Type IV WTG/Solar model assumes


zero negative sequence current
contribution
• Type III WTG has negative sequence
current contribution due to the DFIG
stator connection to the grid
10
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Vendor Engagement

• Goal: Vendor engagement and implementation of the


models in commercial platforms (CAPE, ASPEN,
CYME, PSS/E, Powerfactory, etc).

• ASPEN & Electrocon have started implementing a


beta version of the models and EPRI is providing
technical support

• Present Status: Testing and benchmarking of the


models with vendors using benchmark systems and
databases provided by EPRI members.

11
© 2018 Electric Power Research Institute, Inc. All rights reserved.
CAPE Implementation - Electrocon
• Electrocon has implemented so far the Type IV WTG/
Solar model
• Type III WTG model is under development and will be
presented at the CAPE UGM in June 2018
• Technical paper presented at the CAPE UGM - June 2017
• EPRI and Electrocon have benchmarked the Type IV
WTG model using a 9 bus test system

12
© 2018 Electric Power Research Institute, Inc. All rights reserved.
OneLiner Implementation - ASPEN ASPEN OneLiner v14

• ASPEN has implemented the Type IV WTG/


Solar model with FRT control mode (v14)
• Voltage Controlled Current Source model
• Tentative implementation as a V-I-pf table (v14)
• Work in progress: Replace tabular input with
FRT function settings. No need for user to enter
data. (v15)
• Type III WTG model has been also implemented
and will be available in OneLiner v15 ASPEN OneLiner v15
• ASPEN has prepared a technical bulletin
• EPRI and ASPEN have benchmarked both Type
III and Type IV WTG model using a 9 bus test
system

13
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Webcast: EPRI Wind/Solar Phasor Domain Short-Circuit Models &
Implementation Status in Commercial Tools

 EPRI Webcast Recording (public)


‒ EPRI Wind/Solar Phasor Domain
Short-Circuit Models Description
‒ Status of Model Implementation
in CAPE (Donald McGregor,
Electrocon)
‒ Status of Model Implementation
in ASPEN OneLiner (Sherman
Chan, ASPEN)
 Product ID: 3002010940

Link https://www.epri.com/#/pages/product/000000003002010940/

14
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Model Validation – 3 Approaches
2. Manufacturer EMT Models
1. Generic EMT Models

3. Fault Recorded Measurements

15
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Type-III WTG Wind Park Connected to a 230-kV Substation
EMTP Model
Slack: 230kVRMSLL/_0
Vsine_z:VwZ1 V1:1.00/_-0.00
V2:0.00/_102.09
LF V0:0.00/_45.00
LF1 Va:1.00/_0.00
Vb:1.00/_-120.00
VwZ1 Vc:1.00/_120.00
+
6604_LATIGO
230kVRMSLL /_0 + Relay_Transmission

+
Line_LATIGO_3BUTTES

PI
WP_DFIG1

Relay_Wind
+
11847_THREE_BUTTES
DFIG AVM V1:1.00/_0.2
110.022MVA V2:0.00/_-89.8
230kV V0:0.00/_-89.8
Q-control Va:1.00/_0.2
Vb:1.00/_-119.8
Vc:1.00/_120.2

Phasor Model Results


POI - pu

Variable
EMTP-RV Phasor Model

I+ 0.825 (-39.7) 0.810 (-56.4)


 Wind farm with 66x1.5MW
V+ 0.509 (1.5) 0.509 (0.6)
type-III wind turbine generators
I− 0.858 (105.8) 0.862 (98.4)
 B-C phase to phase fault on the
V− 0.488 (0.4) 0.486 (0.1) tie line to the POI substation

16
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Solar Model Validation with Recorded Data

 Three-phase fault in adjacent line


 Close match between simulation results and recorded data

17
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Validation of Type III Wind Turbine with Manufacturer EMTP Model

18
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Protection Guidelines for Systems with High Levels of
Inverter Based Resources

19
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Protection System Performance Evaluation – EPRI Guidelines Document
• Protection System Performance Evaluation: Study relays response & identify relay misoperation
scenarios on benchmark systems with high renewable penetration.
• EPRI Guidelines Document: Provide recommendations and study practices to protection engineers
when conduction protection studies to prevent relay misoperation/miscoordination

20
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Impact of Inverter-Based Generation on Negative Sequence Based Protection
 Typically inverter-based resources are designed Negative Sequence Directional
to suppress negative sequence current partially Reverse fault misinterpreted as forward
or entirely.
 Negative-sequence current suppression may
result in the misoperation of legacy negative-
sequence-based protection schemes.
– Negative-sequence overcurrent element
– Negative-sequence-based directional element
 Inverter controls define the phase angle
relationship between negative sequence voltage
and current

Negative Sequence Overcurrent

Source: M. Nagpal and C. Henville, “Impact of power


electronic sources on transmission line ground fault
protection,” IEEE Trans. Power Del, 2018
21
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Impact of Inverter-Based Generation on Power Swing Protection

• Large levels of inverter-based


resources might impact the rate of
change of the impedance (due to fast
controls) and the impedance trajectory
measured by power swing relays
potentially resulting in:
• Power Swing Blocking (PSB) function
potential misoperation due to faster
power swings under high levels of
IBR – swings misinterpreted as faults
• Out of Step Tripping (OST) function
misoperation due to modified
impedance trajectories under high
levels of IBR
22
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Related Industry Activities

23
© 2018 Electric Power Research Institute, Inc. All rights reserved.
PSRC WG C24 “Modification of Commercial Fault Calculation Programs for
Wind Turbine Generators”
• Chair: Dr. Sukumar Brahma (NMSU), Vice-Chair: Evangelos Farantatos (EPRI)
• Scope:
• 1) To survey WTG manufacturers to determine what parameters they could provide that could be used by steady state short
circuit program developers in various time frames.
• 2) Use the result of this survey to prepare a report that can be used by steady state program developers to refine their models.
• EPRI has a leading role to the WG. Members include WTG manufacturers (Siemens, GE) and software vendors
(Electrocon, ASPEN, ETAP)
• WG has proposed a voltage controlled current source model with iterative solution
• Input model data:
• Algorithms for generic converter control schemes (EPRI proposal)
• Tabular format (suggested to be provided by manufacturers with non generic converter control scheme)

Data to be requested data from manufacturers


Time Frame (cycles)
Siemens Type IV WTG Model – Information provided to the WG
Positive Positive Negative Power factor of • Controlled fault response starting at 3rd cycle
sequence sequence sequence positive sequence • Ireactive = K*(1-Vt) with K(slope)=2
voltage (p.u) current (p.u) current (p.u) current • Iactive=P*(1/Vt)
1.0
0.8
• I_limit=1.1 p.u
0.6 • Deadband low: 0.9 p.u
0.4 • Q Priority
0.2
• No negative sequence current
0
24
© 2018 Electric Power Research Institute, Inc. All rights reserved.
PSRC WG C32 “Protection Challenges and Practices for the Interconnection of
Inverter Based Generation to Utility Transmission Systems.”

• Chair: Jimmy Deaton (Southern


Company), Vice-Chair: Mukesh Nagpal
(BC Hydro)
• Scope: Write a report that addresses
protection challenges and practices for
the interconnection of inverter based
generation to utility transmission
systems
• EPRI is participating to the WG and is
contributing to several sections of the
report

25
© 2018 Electric Power Research Institute, Inc. All rights reserved.
IEEE PES - NERC Technical Report “Impact of Inverter Based Generation on
Bulk Power System Dynamics and Short-Circuit Performance”
• Joint IEEE PES and NERC TF
• Co-chairs: Kevin Jones & Pouyan Pourbeik
• Target completion: Summer 2018
• Chapter 2: “Large System Impact Issues Related to Large
Penetration of Inverter Based Resources”
• Chapter 3: “Protective Relay Issues Related to Large Penetration
of Inverter Based Resources”
• PSRC CTF34 “Inverter Based Generation Short Circuit Current
Impacts”
• Assignment: “Coordinate/communicate the efforts of the
PES/NERC Low Short Circuit Current Impacts Task Force and
PSRC working groups addressing the issues of inverter-based
resources”
• Chair: Kevin Jones, Vice chair: Gary Kobet
• EPRI is participating in CTF34
26
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Together…Shaping the Future of Electricity

27
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Inverters and Harmonics
17-May, 2018

Sean McGuinness
Principal Technical Leader,
EPRI Intl.
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Introduction

Emissions from inverters


• Wind turbines: Type 3 DFIG, Type 4 full
converter
• Solar PV/Battery Inverters

Harmonic resonance due to cable


capacitance
• Amplification of grid harmonics
• Acting as sink for grid harmonics
2
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Type 4 Full Converter Wind Turbine/Solar PV Inverter
 Switch at relatively high frequency (>1 kHz)
 Voltage sine wave is imperfect
 Harmonic emissions vary with power output

 Inverter output filter attenuates harmonics at and


above switching frequency
 Distortion below switching frequency tends to be
low

 Note: Voltage unbalance tends to increase


harmonic distortion

Clean Sine Wave needs high switching frequency


High switching frequency increases inverter losses
Outcome: switching algorithms usually optimized to to maximize
efficiency while keeping harmonic distortion within limits

3
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Solar PV Inverters

 Distortion is dependent on Filter design, Inverter switching


algorithm, power output
0.40%
20% Rated Output

0.35% 60% Rated Output


100% Rated Output
(Percentage of Fundamental Currnet)

0.30%
Harmonic Current

0.25%

0.20%

0.15%

0.10%

0.05%

0.00%
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Harmonic Index

4
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Type 3 Wind Turbine - DFIG
 Harmonics emit from:
– Grid Side Converter
– Rotor Side Converter
– Induction Generator air-gaps

H=11 Measured harmonic current


 Induction generator may frequency and magnitude variation
with increasing rotor speed
operate with up to 30% slip
 Inter-harmonics dominate
 Harmonic spectrum and
magnitude vary with rotor
speed
5
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Aggregating distortion from multiple inverters
Example Spectrum of a 2.5 MW Type 4 WTG

 Harmonics from parallel inverters


0.45%

0.40%

Current MAgnitude (A)


0.35%

tend to destructively interfere 0.30%

0.25%

 10 inverters ≠ 10 times 1 inverter! 0.20%

0.15%

0.10%

 IEC Std has method to aggregate 0.05%

0.00%
5 7 11 13
Harmonic Coherency of Harmonic Voltage Phase Impact of harmonic Harmonic Index
Order Angle distortion
Example Spectrum of 10x2.5 MW Type 4 WTG
>10th Voltage phase angle variation is pretty Destructive interference 0.18%

random between nearby inverters 0.16%

Current MAgnitude (A)


0.14%
5th-10th Some coherency of voltage phase angle Some destructive 0.12%
between nearby inverters interference 0.10%
0.08%
0.06%
<5th Voltage phase angle between electrically A little destructive 0.04%
connected inverters assumed to be in- interference 0.02%
phase 0.00%
5 7 11 13
Harmonic Index
IEC 61000-3-6, “Harmonic Emission Limits for Customers Connected to MV, HV, and EHV”,
C. Larose,
6 R. Gagnon et al. “Type-III Wind Power Plant Harmonic Emissions: Field Measurements and Aggregation Guidelines for Adequate
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Representation of Harmonics”, IEEE Transactions on Sustainable Energy, Vol. 4, No. 3, July 2013
Inverter Models

 Can use frequency domain or time-


domain simulations

 Level of detail is dependent on study


 Individual inverters + collector
network vs single lumped equivalent

 Study multiple inverter operating


points

7
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Impact of Cable Network
 Wind and Solar Power Plants tend to have large
collector networks, mostly underground cable

 What does cable impedance compare to


overhead line?
– Lower series reactance per unit length (less than half)
– Much higher shunt capacitance per unit length (30-40
times greater)

 Behaves much large a large capacitor bank


 Can introduce a low order harmonic resonance
and also sink harmonics from the grid

8
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Impact of Cable Network 2

Amplification of existing background distortion


• Increases voltage distortion
• Occurs even if inverters are off
• Noticeable when grid distortion increases after energizing
wind or solar power park

Sink existing grid distortion


• Results in flow of harmonic current from the grid into the cable
• Noticeable when grid distortion decreases after energizing
wind or solar power park
• Many PQ recorders only record harmonic current magnitude,
not direction
9
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Summary
 DFIGs emit inter-harmonics and distortion varies with wind speed
and power output

 Full converter wind turbines and solar PV inverters harmonic


distortion changes with operating point and is influenced by
voltage unbalance

 Distortion above 5th harmonic tends to destructively interfere


when there are parallel inverters due to random variations in
harmonic phase angle

 Cables can introduce low-order harmonic resonance which may


amplify or sink grid harmonics
10
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Questions or Comments?

11
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Together…Shaping the Future of Electricity

12
© 2018 Electric Power Research Institute, Inc. All rights reserved.
Modeling for EMT studies
Inverter-Based Resources Webinar Series

Webinar #4 – May 17, 2018


Andrew Isaacs

Open Distribution
Elecromagnetic Transient Output (Instantaneous Quantities)
1.50

What is EMT?
1.00

y (Per Unit Volts)


0.50

0.00

-0.50

-1.00

-1.50

 Power-flow and transient stability


Transient Stability Output (RMS Quantities)
1.20

programs iteratively solve a systems of 1.00

equations in the RMS 60 Hz phasor

y (Per Unit Volts)


0.80

domain
0.60

0.40

 Electromagnetic Transient (EMT) software 0.20

solves systems of differential equations 0.00

which describe the three-phase electrical


Time 1.000 1.020 1.040 1.060 1.080 1.100 1.120

network in the time domain.


 This allows EMT simulations to represent the power system behavior at all
frequencies, limited only by the period of time between solutions
(simulation time-step). Every individual instantaneous phase quantity is
represented, allowing unbalanced faults, harmonics, transients and other
effects to be modeled.

 A few software brands (alphabetical order): ATP, DigSilent (EMT mode),


EMTP-RV, PSCAD/EMTDC

2
These are not your Grandfather’s
EMT Studies!

1975 Prime 400: Image courtesy chilton-computing.org.uk

3
Where might you use EMT today?
 Classical studies are still done:
 Lightning evaluation
 TOV/Insulation coordination
 Switching/line energization
 Transformer energization
 Harmonic analysis
 SSR (Remember Mojave?)
 HVDC/FACTS control design and analysis
 Many other fascinating things done with EMT since
1970!
4
Where might you use EMT for IBR
connections today?
 Weak Systems (See webinar #3!)
 Regions with high chance of power electronic device
interactions (eg. dynamic performance, FRT tests, etc)
 SSO phenomena (SSR, SSCI, SSTI, SSTA, SSTMTLA)
 Blackstart analysis
 Detailed fault current / protection analysis.
 Model Validation
 Support for Factory Acceptance Testing
 Others….

5
Clarifications on Weak Grids
 (From the last webinar) “It is very difficult to use SCR to set
planning guidelines and thresholds… a lot of study is
required!”
 1. Difficult ≠ Impossible…
 2. There are limits for conventionally controlled IBR plants…
Planners need to plan accordingly, use all available screening
and analysis tools, and make judgement calls in absence of
complete information.
 3. Mitigation of weak grid issues (incl. using Sync.
Condensers) requires care. You can make new problems for
yourself!
 4. Distributed rooftop PV is a whole new world of weak grid
uncertainty… Research is required and is ongoing, but can’t
happen quickly enough!

6
SSCI Phenomena
 Currently, any utility connecting IBR near series
compensated circuits is (or should be!!) performing EMT
studies to verify stability.
 Examples: PacifiCorp, any ERCOT utility, any Australian utility,
ISONE, SCE, most Canadian provinces, many European
utilities etc…

7
Model Validation
 Transient
stability models
may be tested
against highly
detailed (real
code) EMT
models to verify
performance
that can’t be
validated in the
field.
 Example from
ISONE:

8
Trend: Large composite EMT models
 Increasingly common as tools allow the capability. Current study
examples:
 Electranet, South Australia: 30 IBR plants, 700 busses (Entire South
Australian grid) – Sync Condenser sizing study.
 South Texas: 22 IBR plants, 7 SVCs, 469 busses – SSCI studies.
 ERCOT Panhandle: 25 IBR plants, 435 busses – “weak grid” studies.
 (planned) ATC Southwest Wisconsin: 12 Windfarms – dynamic
performance tests
 PacifiCorp: 21 IBR models, 50 complex load models – high transfers
with complex RAS, IBR performance critical for new SVC sizing.
 HECO: 8 IBR plants and 18 DG PV proxy models – High penetration
dynamic performance evaluation
 ISONE: 5 IBR plants and multiple new HVDC ties – High chance of
interaction with very large planned offshore and DC projects.
 AEMO (Australian ISO): Long blackstart sequences, very large models.

9
Parallel/Hybrid EMT Simulation
 Models can be
parallelized to allow
increasingly large
networks to be
represented:

10
So you need EMT. What should an EMT
generator model look like? Accurate!
 Represent full detailed inner control loops of the
converters
 Eg. current control, PLL, etc.
 Represent all pertinent control features
 Eg. FRT, SSCI damping, voltage control, plant control etc.
 Represent all pertinent electrical and mechanical
configurations
 Eg. transformers, filters, gearboxes, etc.
 Represent all pertinent protections
 Eg. OV, UV, DC bus, overcurrent, and many others

11
What should an EMT generator model look
like? Usable!
 Accessible user control for settings which are adjustable
 Reasonable time-step (>10 – 20 us)
 Documentation included, and a test case
 Self initializing
 Multiple instance support (and snapshots?)
 Accepts reference values (no hardcoded setpoints!)
 Stable interfacing/scaling methods (injection models often not
suitable)
 Others… see:

http://www.electranix.com/publication/technical-memo-pscad-model-requirements/

12
Electranix Corporation
But Andrew, that’s old material!
 Yes, and manufacturers have (mostly) listened!
 Current model quality is getting better
 New IEEE working group (“Use of Real-Code in
EMT Models for Power System Analysis”) on “real
code” model development. All plusses!!
 State of the art “real code” models are now
becoming widespread.
 You still need to check model quality.
 BUT… Some manufacturers are now complaining
that their beautiful generator models are being
used with overly simplistic system models.

13
Doing an EMT study?
 What can you tell from this
model?
 Small Signal control stability
(SCR Tests) (yes, sometimes) P = 99.43
Vsys_LGpk

Q = 32.75
 SSCI? (no) V = 1.051
+ +
V
R=0

FRT? (no)
Beautiful A
 Detailed Model
V Rsource Lsource

 Control interactions? (no)


 Voltage overshoot? (not

Fault
necessarily)
 Insulation coordination? (no)
 Harmonics? (no)
 Lightning/TOV/TRV/etc? (no)

14
Dynamic performance models
 If you want to correctly evaluate these things, you need:
 Sufficiently big system model (How big? Good question, but maybe
bigger than you think!)
 Frequency dependent representation of transmission lines
 Series Capacitors (and sometimes series capacitor arresters and
protection)
 Transformer saturation, tap settings, winding configurations, etc.
 Detail for nearby power electronics (HVDC, wind, SVC, etc.)
 Suitable load representation
 Synchronous machine dynamic representation
 Shunt capacitors and reactors, neutral reactors, etc.
 Suitable system equivalent (multi-port or dynamic if necessary)
 Depending on the study, you may also need:
 Station and line surge arresters
 Protection, SPAR/TPAR schemes, RAS, SPS data, other stuff…

15
Dynamic performance models
 The specific system model you need depends on what
problems you’re looking for…
 Tools are available to simplify model development and extract
available data from the powerflow and transient stability tools
(full disclosure… Electranix sells some of these – E-Tran suite of tools, some of which
are licensed as PSS/E modules)
 Initialization, data and library management is required…
 Even with these tools, there is a lot of extra data to obtain and
enter. There is a lot of work to do!

16
Example
System
Model: P = 99.43
Vsys_LGpk
Q = 32.75
V = 1.051 V
+ + R=0
Beautiful A
Detailed Model
V Rsource Lsource
Fault

17
(example system equivalent)

18
What is the future for EMT?
 Interconnected/harmonized system models
(cross-platform, cross-genre models)
 Real code models
 Very large parallelized EMT models for
dynamic performance evaluation
 Regional library development for EMT
system models
 As IBR and DG penetration continues to
rise and fossil fuel based generation
declines, need for new control concepts
(and detailed models) will rise too.
 Increasing adoption of EMT as part of
routine planning

…Thanks for listening!!!


19
Subsynchronous Control
Interaction and Inverter Based
Resources:
ERCOT Perspective

NERC-NATF-EPRI-UVIG

ERCOT Dynamic Studies, Transmission Planning

May 17, 2018


Outline

ERCOT SSR Studies for IBRs


• 2009 SSCI (subsynchronous control interaction) event involving
ERCOT wind farms

• Frequency Scanning method for renewables

• Detailed time-domain studies

• Observations from past studies and operational experience

PUBLIC 2
Series Capacitors in ERCOT
Cross (x2)

Gauss (x2)

Kirchhoff (x2)

Romney (x2) & Kopperl (x2)

Oersted (x2) & Edison (x2)

Cenizo (x1) / Del Sol (x1) /


Rio Hondo / (x1)
North Edinburg (x1)

PUBLIC 3
South Texas SSR Event (2009)

• Series capacitors installed on


long 345 kV line to South Texas.
• 1,000 MW of wind farms
Lon Hill
connected to Ajo. (Type III wind.)
• In 2009, fault caused north line to
trip, leaving wind radially
connected to series caps. Ajo

• Very high currents resulted in


damage.
Rio
Hondo

PUBLIC
Induction Gen Effect and Control Interaction

• The 2009 Event was a different type of SSR:


– Induction Generator Effect & Control Interaction
– Not mechanical. Purely an electrical resonance.
– Very fast (~150 ms).
• Solved by wind turbine control modifications

PUBLIC 5
Types of SSR

Type of SSR Affects SSR if Oscillations Are


Torsional Interaction (TI) Mechanical Growing

Torque Amplification (TA) Mechanical Damped but high energy

Induction Generator Electrical Growing


Effect (IGE)

Control Interaction (SSCI) Electrical Controls Growing

• IGE can occur for either synchronous or IBR


• SSCI generally only occurs for IBR
PUBLIC 6
SSR Studies in ERCOT
To handle all these
• New Generators through studies, ERCOT uses
Interconnection Process layers of screening
– ~150 SSR screening studies / yr
– ~20 detailed studies / yr

• Transmission Project-Related
Studies

• Annual Assessment

PUBLIC 7
ERCOT Study Process
(Same process for all generators)

TOPOLOGY CHECK

FREQUENCY SCAN 0.30

0.25

Impedance (ohms)
0.20

0.15
Resistance
0.10
Reactance
0.05

TIME DOMAIN 0.00

‐0.05
5 9 13 17 21 25 29 33 37 41 45 49 53 57

Frequency (Hz)
STUDY P1 : Graphs
PWF (MW) QWF (MVAr)
250

200

150

100

50

-50

-100
x 0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0

PUBLIC 8
Topology Test: Counting Contingencies
345 kV B
Outages “A”
and “B” make
138 kV
the generator
A “radial” to the
series capacitor
GenY ~

Discussion: Test is limited because generators can be vulnerable


under non-radial conditions. Thus a margin is necessary.

PUBLIC
Study Process

TOPOLOGY TEST
0.30

0.25

Impedance (ohms)
0.20

FREQUENCY SCAN
0.15
Resistance
0.10
Reactance
0.05

0.00
5 9 13 17 21 25 29 33 37 41 45 49 53 57
‐0.05
Frequency (Hz)
TIME DOMAIN STUDY
P1 : Graphs
PWF (MW) QWF (MVAr)
250

200

150

100

50

“ERCOT subsynchronous resonance topology and -50


0

frequency scan tool development”, Yunzhi Cheng, x


-100
0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0

Jonathan Rose, IEEE PES General Meeting, July 2016.

PUBLIC 10
How Do We Run Frequency Scans?
Measure Thevenin impedance at
different frequencies 5 ~ 55 Hz

Network

System-Side Generator-Side
Frequency Scan Frequency Scan
Don’t forget to
simulate
contingencies!

11

PUBLIC
How Do We Run Frequency Scans?

+
Total System+Turbine Scan Total System+Turbine Scan
140 60

120
R X 40 R X
100
80

=
20
60
40
0
20 25 27 29 31 33 35
0
‐20
0 10 20 30 40 50 60
‐20
‐40 ‐40
‐60

‐60

12
Here, resistance is positive where reactance crosses from negative to
positive, so anticipate no SSR concern
PUBLIC
Frequency Scans have limitations
(Small Signal vs Large Signal)
• Observed during SSR study:
– Frequency scan indicated positive resistance  GOOD
– Non-fault EMT indicated stable response  GOOD
– Fault EMT indicated SSR and tripping  BAD!

• What happened?
– Frequency scans are linearized around an operating point. Thus
good for small signal but not for large signal.
– IBR are much more non-linear than synchronous generation
• Different operating modes and switches in controls

• Thus, EMT (time-domain simulation) may be needed for


confirmation
PUBLIC 13
Why bother with frequency scans?
Why not just jump to time-domain simulations?

• Time-domain is fine if you only have a few scenarios.

• Frequency scans are preferable if you have dozens of sensitivities and


contingency combinations. They can help identify the worst scenario
for time-domain simulation.

• Frequency scans provide more intuitive understanding. You can


visually see how close a system is to being unstable and having SSR.
(Stability margin.)

PUBLIC
What about sensitivities? (Some observations…)

• Large impact:
– Transmission Outages
– Nearby generators on/offline

• Moderate impact:
– Number of turbines online and real / reactive dispatch
– Fault / faultless outages
– Series capacitor bypass configurations

• Lesser impact:
– Switched reactive shunts

PUBLIC 15
Industry Progress

• Solar and Type IV Wind are frequently not susceptible


– ERCOT requires SSR studies anyway because controls could
inadvertently cause SSCI

• Type III (DFIG) wind tends to be highly vulnerable


• Most Type III DFIG turbine manufacturers have SSR
mitigation packages (damping controllers)
– Good progress since 2009
– However not all packages equally effective. Some require tuning.
• Potential challenges as system changes
– Most solutions implemented via firmware upgrade

PUBLIC 16
Protection vs Mitigation – ERCOT’s Practice

• ERCOT generally requires mitigation, meaning that


the generator and series capacitor should operate fine
together, even under X simultaneous transmission
outages (e.g. N-4: Four outages)
• ERCOT protocols generally do not allow Protection to
be the primary means of defense (i.e. tripping upon
detection of SSR)

• Mitigation required for new resources (new resources


must adapt to existing grid and be able to operate
without intervention)

PUBLIC 17
Issues Locating Very Close to Capacitors

• Recently,
Renewable Resources are connecting
closer to series capacitors, even tapping
series-compensated lines themselves

– Creates complex engineering challenges

– Will require extra time and extra capital


to resolve

PUBLIC
“Energy Accumulation” and VRT Tripping Issues

• Energy accumulates in capacitor during fault


• Fault clears, and energy discharges into IRR
• May reverse power flow leading to generator
tripping

• ERCOT expects generators to ride through


(LVRT)

• May require hardware changes to fix


– Custom engineering. No “off-the-shelf” fix.

PUBLIC 19
Questions?

PUBLIC 20

You might also like