Modeling Renewables as IBR

Penetration Rises

MIPSYCON 2019

November 13, 2019

Andrew Isaacs

Open Distribution
Overview: Too many topics…
 What is EMT? What is it good for?
 Screening Approaches
 Strategies to verify EMT and RMS model quality
 Study Techniques and Tools
 Potential limits to IBR penetration

2
 Elecromagnetic Transient Output (Instantaneous Quantities)
1.50

1.00

What is EMT?

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 50/60 Hz phasor

y (Per Unit Volts)

0.80

domain
0.60

0.40

 Electromagnetic Transient (EMT) software 0.20

solves systems of differential equations Time

0.00

which describe the three-phase electrical

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

3
Where might you use EMT today?
 Classical studies are still done:
 Lightning evaluation
 TOV/TRV/Insulation coordination
 Switching/line energization
 Transformer energization
 Harmonic analysis
 Sub-Synchronous Resonance (SSR)
 HVDC/FACTS control design and analysis

4
Some newer applications for EMT…
 Weak interconnections
 Very high renewable penetration
 Regions with high chance of power electronic device
interactions (eg. dynamic performance, FRT tests, etc)
 SSO phenomena (SSR, SSCI, SSTI, SSTA)
 Black start analysis
 Detailed fault current / protection analysis.
 Model Validation
 Support for Factory Acceptance Testing

5
A few notes on Screening…
When should you do EMT studies?

6
Screening Approaches
Weak grid/High Penetration metrics:
 Weak grid studies are more common now.
 Looking for ride-through failure, control instability, device
interactions
 Engineers continue to misunderstand SCR based metrics.
Key points to understand:
 There is no single threshold which is “weak”, particularly
when there is more than one plant close together
 SCR metrics become more and more problematic the more
plants are included, and the larger the network being studied.
For some systems, they are completely inappropriate!

7
SCR Metric
 SCR is the most basic and easily applied metric to
determine relative system strength.
 Calculations should be applied after worst realistic
contingency
 Assume no contribution from nearby IBR

𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑃𝑃𝑃𝑃𝑃𝑃
𝑆𝑆𝑆𝑆𝑆𝑆𝑃𝑃𝑃𝑃𝑃𝑃 =
𝑀𝑀𝑀𝑀𝑉𝑉𝑉𝑉𝑉𝑉
 Doesn’t account for load
 If there are multiple IBR plants, SCR doesn’t work well

8
Screening Approaches
 WSCR and CSCR can be used for multiple IBR
plants
 Requires well-delineated renewable regions
with low load
 Acceptable threshold is system dependent
 IPSCR can be used for screening wider areas
 New metric, good for very wide area screening
 Requires more experience and validation
 Weak grid issues are system and
equipment specific and it is difficult to
define a “minimum system strength”
criteria that can be applied uniformly.
 Threshold selection involves detailed
study first!!
9
Strategies to verify EMT and RMS
model quality

10
If you are doing EMT system
studies, you need good IBR models!
 IBR model adequacy should be tested against quality criteria:
 Criteria examples:
 Electranix PSCAD model submission guidelines V6
 Electranix Model Compliance Checklist (NEW, being tested…)
 ERCOT PSCAD Model Submittal Guidelines (Draft Jan 2018)
 ISONE Planning Procedure PP5-6
 NERC IRPTF guideline for IBR connections
 IEEE 2800 draft wording on model requirements
 New ATC model submission criteria
 Many AEMO (Australia) requirements
 Adequacy tests include a statement of compliance, visual
inspection, review of supplied documentation, and actual spot-
testing in PSCAD.
11
Strategies to verify model quality:
Benchmark RMS against EMT
 Provides check on parameters, adds to validation not covered by MOD-
27, and identifies shortcomings in both models.
 Note: Overly stringent benchmarking requirements can be self-
defeating!
 Can be expensive and time consuming
 If modeling experts and OEMs are overwhelmed, the care in modeling
declines and then you’re worse off!
 Needs compromise to accommodate existing resources.
 Note: Overly relaxed benchmarking requirements (or no
benchmarking requirements) can result in bad models, leading to
reliability risks and study churn.

12
What is the consequence of
accepting poor quality models?
 You will waste expensive and time-consuming EMT study effort.
 You will get behind on meeting your tariff study timeline requirements
 You may never get another chance to ask for better models!
 You may spend a lot of effort mitigating threats which are not real.
 You may ultimately miss reliability threats

(Potential dumpster fire)

13
(Aside)
If you’re doing EMT system studies, you also need good
EMT system models with appropriate size and detail!

14
Long term: Model Standardization
 DLL common modelling standards
 Existing IEC-61400-27-1 standard
 New IEEE PES Task Force "Use of Real-Code in EMT
Models for Power Systems" (this task force is organized
under the AMPS (Analytical Methods for Power
Systems) Committee, and TASS (Transient Analysis and
Simulation) working group.
 Cigre C4.82 Working Group “Guidelines for Use of Real-
Code in EMT Models for HVDC, FACTS and Inverter
based generators in Power Systems Analysis”

15
Study Techniques and Tools

16
Step 3: System Model Development

Net
Equiv
Page

17
System Model Development
 For the “kept” network, the E-Tran tool (= PTI data
conversion module) will:
 Initialize all voltages, angles, taps, switched shunts, etc to
solved values.
 Create Bergeron or PI section line models where only simple
R,X,B data is available
 Use generic .dyr machine data where available (including
generators, exciters, governors, CLOD, and other)
 Create multi-port passive system equivalents at the
boundaries of the “kept” network.
 Add fault and line monitoring automation to the case for the
case as required.
 Replace powerflow data with the detailed data in the library
(IBRs, lines, loads, faults, SVCs, HVDC, SPS, etc)

18
Data management
Model library development:
 Some interconnections have developed their EMT data
libraries and automation to a degree that the
interconnection EMT study cycle is very fast (as short
as 1 month beginning to end), depending on
requirements.

19
Simulation and Analysis
System Performance:

 Eg: interactions between different

controls, or between a control system
and the bulk system

(Plot is system voltage, 6 cycles per

division)

Eg: TOV/FRT Failure ->

20
Simulation and Analysis
Validation:

21
As models get bigger, more tools
are required!
Parallelization and co-simulation tools:
 E-Tran Plus for PSCAD -
parallel simulation for:
 Speed
 Overcoming compatibility
errors (eg. FORTRAN,
timesteps, conflict between
models, multiple instances)
 PSCAD ENI: parallel
simulation built-in to PSCAD
 E-Tran Plus for PSS/E (= PTI
Co-simulation module): hybrid
simulation between transient
stability and EMT

22
23
Simulation and Analysis
 Massive amount of output makes engineering challenging. There are
many things to see…
 We use sensitive high level pass-fail criteria, then dig deeper when
something fails. We test for:
 Ride-through failure, momentary cessation
 Voltage recovery and stability
 Very high or extended transient voltages
 High or low steady state voltages
 Undamped oscillations in P, Q, V
 Harmonic content
 Loss of synchronism or slipped poles
 Custom software is used to help evaluate and summarize key output,
and then engineers can zoom in to key issues as required.
24
 Example summary tables…
WF1 WF2 WF3 WF4 WF5 WF6 WF7 WF8 WF9 WF10 WF11 WF12 WF13 WF14 WF15 WF16 WF17 WF18 WF19 WF20 WF21 WF22 WF24 WF27 WF28
Fault 1
Fault 2
Fault 3
Fault 4
Fault 5
Fault 6
Fault 7
Fault 8
Fault 9
Fault 10
Fault 11
Fault 12
Fault 13
Fault 14
Fault 15
Fault 16
Fault 17
Fault 18
Fault 19
Fault 20
Fault 21
Fault 22
Fault 23
Fault 24
Fault 25
Fault 26
Fault 27
Fault 28
Fault 29
Fault 30
Fault 31

25
Example individual fault table…
Pass 1 Pass 2 Pass 3 Pass 4 Pass 5 Pass 6 Pass 7 Pass 8 Pass 9 Pass 10 Pass 11 Pass 12 Pass 13
WF1
WF2
WF3
WF4
WF5
WF6
WF7
WF8
WF9
WF10
WF11
WF12
WF13
WF14
WF15
WF16
WF17
WF18
WF19
WF20
WF21
WF22
WF24
WF27
WF28

26
Fancy tools don’t equal good
decisions!
 Beware precision without accuracy
 Don’t be afraid to use judgement!

27
What about the Future?

28
Potential limits to IBR penetration
 Should politicians “Do MORE!!”?

 Can we get to 100% renewables? Yes! But…

 What would happen if we suddenly retired all oil, gas,
and coal based generation?
 There is a technology gap, not just a money/policy gap.
 There is a path forward, but it isn’t easy, and it isn’t ready
yet.

29
Nevermind the politicians… what
should we be doing?
 What is our role between now and 100% renewables?
 Continue to develop our simulation capabilities
 “Grid-forming” (Grid-firming?) will require a big
industry/research push. We need to understand how to
specify this so we can close the technology gap.
 Batteries can play a big role, but we need to uncap their
technology potential.
 We need to keep our minds open, but also push back on
bad policy (Ethics, and engineering judgement)

30
Nevermind the politicians… what
should we be doing?
 What is our role between now and 100% renewables?
 Study timeframes are decreasing, and complexity is
increasing.
 Engineers are stressed (on both TSO/ISO and GO sides)
We should be investing in engineering (human)
resources. Value what we have, and prioritize power
systems engineering in our organizations.

31
Questions?

32