mod-33 principle, criteria and methodology: pacific northwest’s principle_pacific northwest... ·...

102
MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo Gong ColumbiaGrid WECC MVWG Workshop Nov. 2018, Salt Lake City 1

Upload: others

Post on 22-Mar-2020

2 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s

Experience

Bo GongColumbiaGrid

WECC MVWG WorkshopNov. 2018, Salt Lake City

1

Page 2: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

This work is jointly contributed by 11 utilities in PNW:

Avista: Tracy Rolstad, Richard Maguire

BPA: Dmitry Kosterev, Dave Carthcart, Kalin Lee, Hamodi Hindi

Chelan County PUD: Zachary Zornes

Cowlitz County PUD: Jerod Vandehey

Douglas County PUD: Jeff Heminger, Leslie Corson

Grant County PUD: Ken Che, May Le

PacifiCorp: Song Wang

Puget Sound Energy: Shengli Huang, Sarah Davis

Seattle City Light: Desmond Chan, Dejene Mersha

Snohomish County PUD: Long Duong

Tacoma Power: Khanh Thai, Mark Pigman

Acknowledgement

2

Page 3: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

NERC compliance became effective in 7/1/2017

Each Planning Coordinator performs power flow and dynamic validation every 24 months Comparing planning models with real time measurements

Steady-state measurements in western interconnection mostly comes from node-breaker WSM case

Real time dynamic measurements comes from PMU/DFR/EMS/SCADA

A guideline for unacceptable comparison discrepancy

A guideline to resolve unacceptable discrepancy

NERC MOD-33 Standard

3

Page 4: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Tightly coupled transmission system in ColumbiaGrid footprint

9 PCs (Federal, Municipal, IOU)

Backbone 500kV and 230kV owned and operated by BPA, other utilities coupled to each other through 230kV and 115kV

Many small municipals GO and LSE

Hardly can shield the influence from neighbors’ models when utility evaluates their own performance

Joint efforts on system model validation led by ColumbiaGrid

8 CG members and 2 participants

PAC joined in August, 2018

System Model Validation in Pacific NW

4

Page 5: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Extensive experience and better calibrated models serves as foundations to system wide model validation in PNW

Dynamic stability issues due to geographic factors Distribution of hydro generation

Concentrated load centers in coastal metropolitan area (Seattle and Portland)

High power transfer in long transmission lines

Power exchange with neighboring regions (California, British Columbia, and Montana) varied largely in direction and amount

Extensive model validation experience in the past 2 decades

The failure of reproducing Aug 10, 1996 WSCC outage by simulation

Improving load and generator models through MVWG

An online tool using PMU data developed by BPA to constantly improves dynamic models whenever a disturbance happened

High resolution measurement of PMU, DFR (BPA, PacifiCorp and PSE) to monitor major generation hubs, interties and load centers

SCADA and EMS measurements covers every corners of the footprint.

System Model Validation in Pacific NW

5

Page 6: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

In 2016, ColumbiaGrid and member utilities started to develop a process document satisfying NERC MOD-033 R1

Establish a documented process includes:

system condition/event selection

data acquisition

base case development

data review and correction

methodology, etc.

Include a guideline for determining “acceptable/unacceptable” power flow and dynamic comparison

Updating the guideline for “resolving unacceptable discrepancy”

https://www.columbiagrid.org/download.cfm?DVID=4747

CG MOD-33 Process Document

6

Page 7: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

CG’s MOD-33 Principles

7

Page 8: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

It is a system wide validation effort

Other NERC compliance (MOD-25/26/27) exists for plant level validation

A system wide efforts cover a large area, validate system performance for the whole region

1. Do not isolate good and bad models’ performance

Let them interact and contribute together to the system performance.

2. Do not tune a part of system to get other parts comparison good.

Comparing measurements all at the same time

3. As a bonus, you will see devices interact with each other during a complicated events

Normally not captured by staged tests or individual plant validation.

CG’s MOD-33 Principles

8

Page 9: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Objective to improve system modeling

First we admit there are bad model/data

Located “bad” models/data is a very challenging task

“Unacceptable” criterion is a tool to help us distinguish good & bad

Do not set the bar too low so that everything become “good”

Accommodate discrepancies not coming from modeling aspects (measurement, software, etc)

Before chasing “bad”, started with a good case so that we are truly comparing “apple to apple”

A “bad” comparison should be truly caused by “bad” models, not other factors

90% of time spent in our study is to create a good case

It is an iterative process and we expect more and more stringent criterion when we make models better and better.

CG’s MOD-33 Principles

9

Page 10: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Ensure equal benefit to all participants

There are members and non-members whose transmission systems are tightly coupled.

Some utilities who own mostly 230kV~500kV transmission lines, some only have 115kV and below.

Some utilities owns a large amount of generation (hydro, gas, renewable, etc.), some are primarily a LSE.

A uniformed model validation process helps to bring all utilities modeling practice to a consistent standard, proving long term benefit

We validate all transmission system 115kV and above, treat equally transmission, load and generation.

CG’s MOD-33 Principles

10

Page 11: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

CG’s MOD-33 Criterion

11

Page 12: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Power Flow Validation Criterion

12

Parameters for

Comparison

Bandwidth for Comparison

115kV 230kV 300kV 345kV 500kV

Real Power flows

• Lines and transformers +/-10% (AC) +/-10% (AC) +/-10% (AC) +/-10% (AC) +/-10% (AC)

• Generators or 10 MW, or 20 MW, or 30 MW, or 50 MW, or 100 MW,

• HVDC, series cap, etc whichever is

larger

whichever is

larger

whichever is

larger

whichever is

larger

whichever is

larger

Voltages

• Transmission buses,

generator terminal, DC

lines’ terminals, reactive

device

+/-4% +/-3% +/-3% +/-3% +/-2%

Reference to several existing MOD 33 criterions and our own system validation studies

All 115 kV and above, different voltage use different acceptable band

MW Line flow: Max (a fix band, percentage actual flow)

Voltage: Percentage of actual measured voltage

Page 13: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Examples: A 230 kV line was measured at 600 MW flow, if the planning case flow is 535 MW, it

is exceed both 10% of the 600MW and 20 MW fixed band, it is “unacceptable”

A 115 kV line was measured at 150 MW flow, if the planning case flow is 135 MW, it is within 10% of the 150MW even though it exceed the 10MW fixed band, it is still “acceptable”

A 115kV line was measured at 50 MW flow, if the planning case flow is 58 MW. Even though it exceed the 10% of the 50 MW flow, it is still within 10MW fixed band, so it is still “acceptable”

A 115/500kV transformer was measured at 500 MW flow, if the planning case flow is 445 MW, it exceed both 115kV criterion (the 10% of 500 MW and 10MW fixed band for 115kV), regardless of the 500kV criterion, it is “unacceptable”

An HVDC line was measured at 1500 MW flow, the planning case has the flow at 1620MW. It exceed the 100MW fixed band (there is no 10% band for DC line), so it is “unacceptable”.

A 230 kV bus voltage was measured at 1.0 p.u., the planning case voltage is 1.025pu. It is within the 3% band, so it is “acceptable”.

Power Flow Validation Criterion

13

Page 14: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Power Flow Validation Criterion

14

Why we don’t use reactive power flow but real power flow

Real power flow can be done much efficiently as line loss is small, get source (generator), sink (load) and path (line status) correct, you will get it matched

Reactive power flow, each transmission line can individually generate or absorb reactive power, and it doesn’t travel far, not practical in a large area.

Why we don’t use reactive line flow but voltage

After having voltage matched everywhere in large area, system condition matches well in many other comparison (dynamics, real power, etc. )

Solving power flow, match voltage set points help you align voltage efficiently

Voltage difference are much easier to trace in transmission system than MVar flow

Page 15: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

ColumbiaGrid will primarily adopt visual inspection to compare the real-time measurements and simulation.

As we use power flow validated case for dynamic validation, power flow criteria play an important role in validation results.

Dynamic Validation Criterion

15

Page 16: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Dynamic comparison is different from steady state ones A band all along the trajectory, at final time, or after oscillation damped out?

Can we allow exception at the time of switching? If yes, then what action is eligible as a switching?

Percentage vs Absolute metrics? Percentage value are depends on the base value, what if the base value

may become zero, or come close to zero?

Using an absolute error, on the other hand, may fail to disclose similarity and neglecting the severity of the disturbance

Problems of using Metrics for Dynamics

16

-2500

-2000

-1500

-1000

-500

0

500

0 50 100 150 200

-500.00

0.00

500.00

1000.00

1500.00

2000.00

2500.00

0 20 40 60 80 100 120

Page 17: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Metrics varied for event types Frequency may be less important for a voltage

event and vice versa

Metrics varied on dominant trends and looking forward horizon Criterion evaluating short horizon may fail to

disclose a dominant trend in long horizon and vice versa

Metrics varied on dynamic behaviors involved There is no metrics can capture all dynamic

behaviors, and in most time it can only be picked when you saw it

Example: (on the right) oscillations frequency and magnitude can

be more important than point wise difference along trajectories.

Dynamic Metrics v.s. Events and Variables

17

-200

-150

-100

-50

0

50

100

150

200

0 50 100 150 200

-60

-40

-20

0

20

40

60

80

100

0 50 100 150 200

Page 18: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Some Hints of Choosing Criteria

18

Making PF and dynamic comparison first

May start with a WECC mapped case

Fixing model problems, improving comparison until it becomes satisfactory

Considering both the time you want to spend and accuracy to achieve

After satisfied with comparison results, reversely select the criteria that you can achieved for comparison

Leave some space for possible variations

As model improves, you should expect a tighter criteria

Page 19: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Power Flow Validation(8/8/2017 Event Example)

19

Page 20: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

MOD 33 Power Flow Study Process

20

Page 21: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Both a large frequency and a large voltage event

2017 8/8 3:08 am PCT

A large frequency event: three Colstrip units tripped

A large voltage event: gen/line tripping, reactor switching, RAS action

CG’s 1st MOD 33 Event

21

Time (MDT) Time From Fault Cycles From Fault Event

04:08:18.279 00:00:00.000 0.00 A-Phase fault on Colstrip - Broadview 500 kV Line B; Z=0+ j0.0470 pu 35% from Broadview

04:08:18.329 00:00:00.050 3.00 Fault cleared by opening Colstrip - Broadview 500 kV Line B

04:08:18.647 00:00:00.368 22.08 Series capacitors on the Colstrip – Broadview A line bypassed

04:08:18.687 00:00:00.408 24.48 Colstrip - Broadview 500 kV Line A opens

04:08:18.833 00:00:00.544 33.24 Colstrip Unit 3 Tripped

04:08:18.833 00:00:00.544 33.24 Colstrip Unit 4 Tripped

04:08:18.917 00:00:00.638 38.28 Colstrip Unit 1 Tripped

04:08:23.932 00:00:05.653 339.18 Broadview 500/230 kV Bank 3 tertiary reactors inserted

04:08:28.943 00:00:10.664 639.84 Broadview 500/230 kV Bank 4 tertiary reactors inserted

Page 22: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Data were collected (as needed) along the process, not all at once

Peak RC provided system snapshots in WSM format pre/post event at 03:04 am and 03:10am

NWE provided event sequence

BPA provided PMU data on 43 substations (43 voltage, 43 frequency, 247 line flows)

PSE provided DFR data on 5 substations (5 voltage and 5 frequency)

Snohomish County PUD provided SCADA continuous measurements on one hydro unit (Jackson #2)

Chelan County PUD provided PI continuous measurements on two substations (2 voltage and 2 frequency)

Other utilities provided additional RAS and switching records

MOD 33 Data Collection

22

Page 23: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

As the first step, WECC helped to develop a mapping planning case from 17HS operation cases, available Dec. 2017

Individual Generation Mapping

Scaled load based on BAs

Major transmission path/branches status mapping

Major switching shunts mapping

Major generator voltage set-point mapping

System wide generation and interchange level matching are hard to be addressed by our regional efforts

MOD 33 Case Preparation 1: WECC efforts

23

Page 24: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

This is the key step to make sure we are comparing “Apple to Apple”

ColumbiaGrid and member utilities work together to refine the WECC case considering the following aspects:

Generation Mapping

Load Adjustments

Transmission Impedance/Rating/Status Mapping

Reactive Power Devices Mapping

Generator Voltage Set point Mapping

Topology Difference

MOD 33 Case Preparation 2: CG’s efforts

24

Page 25: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

For refining, we need to identify the difference exists between WECC mapped case and WSM, and improve it

The difference can come from mismatches/errors either in WECC mapped planning case, or WSM case

This is achieved by establishing branch/bus mapping relations between two cases for our region

Using both cases and extra information collected from utilities, an in house software tool automatically setup mappings for 3126 (out of 4989) branches between WSM/planning cases: 0.6kV ~ 500kV

Set up bus mapping for most buses (4086)

Identify topology difference

CG’s Methodology

25

Page 26: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

The auto-generated table by CG’s tool looks like:

CG’s Methods

26

Page 27: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

What’s the Difference?

Real Power Comparison

27

Page 28: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

MW Generation: Planning vs WSM

Jim Bridger Service Load netted in WSM case

Swing changed to Palo Verde Unit 1, Coulee #22 restored to its WSM value

Stability model changes for synchronous condenser mode of Coulee units

Case Comparison WSM vs planning case

28-50

150

350

550

750

950

1150

1350

1550

1750

-50 150 350 550 750 950 1150 1350 1550 1750

Gen MW comparison

Jim Bridger

Grand Coulee

Page 29: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

MW Load: Planning vs WSM

Case Comparison Results

29

y = 0.8727x + 0.4804R² = 0.7008

-50

0

50

100

150

200

250

300

350

-50 0 50 100 150 200 250 300 350

MW Load comparison

Grand Coulee

Jim Bridger

Page 30: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Pump is treated as load in WSM, negative generation in Planning case

WECC doesn’t map pump in planning case, treat as normal load to scale with other loads

Same case for several paper mills

Adopted the planning case model for stability simulations

Gen & Load Type Mismatch Example: Coulee Pump

30

Page 31: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

MW branch flows are primarily impacted by the following factors:

Generation + load + branch losses + mismatch + topology

Good + ok + negligible + small + ok = ok

3126 (0.6kV ~ 500 kV) branches comparison between planning case and WSM

MW Branch Flow Difference

31

y = 0.9624x - 0.1889R² = 0.9776

-200

-150

-100

-50

0

50

100

150

200

-200 -150 -100 -50 0 50 100 150 200WS

M

Planning

Page 32: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

What’s the Difference?

Reactive Power Comparison

32

Page 33: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Mvar Load: Planning vs WSM

6015 Mvar vs 2092 Mvar

Case Comparison Results

33

y = 0.4817x - 0.6891R² = 0.3551

-40

-20

0

20

40

60

80

100

120

140

-40 -20 0 20 40 60 80 100 120 140

MVar Load comparison

Page 34: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Shunt: Planning vs WSM

-3190 MVar vs. -7203 Mvar

Case Comparison Results

34

y = 0.8794x - 16.681R² = 0.7569

-800

-600

-400

-200

0

200

-800 -600 -400 -200 0 200

Shunt comparison

Marion

Captain Jack

Maple VL

Malin

Celilo

Longview

Page 35: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

MVar Generation: Planning vs WSM

-1507 Mvar vs. -2628 Mvar

Case Comparison Results

35

y = 0.8375x - 5.4196R² = 0.5394

-650

-550

-450

-350

-250

-150

-50

50

150

250

350

-650 -550 -450 -350 -250 -150 -50 50 150 250 350

Gen MVar comparison

Page 36: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

MVar branch flows are primarily impacted by the following factors:

Generation + load + branch charging (and topology) + shunt + mismatch

Not Good + Not Good + Significant + Not Good + Significant = Not Good

3126 branch comparison between planning case and WSM

MVar Branch Flow

36

y = 0.5541x - 0.865R² = 0.4204

-100

-80

-60

-40

-20

0

20

40

60

80

100

-100 -50 0 50 100WS

M

Planning

Mvar Branch Flow

Page 37: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Improvements:

MW Line Flow Adjustments

37

Page 38: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Step 1, each utility reviews their branch/bus mapping table

Large amount of impedance/rating mismatch found

About 50% of mismatch were errors in WSM case

Some generator/line/breaker status errors in WSM case, as well

Step 2, individual loads mapping (change + move)

MW Line Flow Adjustments

38

y = 1.0001x + 0.006R² = 0.9936

-40

-20

0

20

40

60

80

100

120

140

160

-50 0 50 100 150

MVar Load comparison

y = 0.9998x + 0.319R² = 0.871

-50

0

50

100

150

200

250

300

350

400

-100 0 100 200 300 400

Substation MW Load comparison

Page 39: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Step 3, topology difference review and correction

Step 4, maps outage status, phase shifter angle, transformer tap, series cap bypass status, etc.

Step 5, remove Peak RC’s artificially inserted mismatches in WSM cases, confirmed with utilities’ own measurement data.

MW Line Flow Adjustments

39

Page 40: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

PF Comparison Improvements:

Hand-on Example

(no slides)

40

Page 41: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

All 3126 branches, 0.6kV ~ 500kV in area 40

MW Flow Comparison WSM vs. planning

41

y = 0.997x - 0.0036R² = 0.9988

-1500

-1000

-500

0

500

1000

1500

-1500 -1000 -500 0 500 1000 1500WS

M

Planning

PLN_vs_WSM

PLN_vs_WSM

Page 42: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Mapping before and after the adjustments

MW Flow Comparison WSM vs. planning

42

y = 0.9624x - 0.1889R² = 0.9776

-200

-150

-100

-50

0

50

100

150

200

-200 -150 -100 -50 0 50 100 150 200WS

M

Planning

MW Branch Flow

y = 0.997x - 0.0036R² = 0.9988

-200

-150

-100

-50

0

50

100

150

200

-200 -150 -100 -50 0 50 100 150 200WS

M

Planning

PLN_vs_WSM

PLN_vs_WSM

Page 43: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

500 kV branches

43

y = 0.9977x + 0.5159R² = 0.9993

-2000

-1500

-1000

-500

0

500

1000

1500

2000

-2000 -1500 -1000 -500 0 500 1000 1500 2000WS

M

Planning

PLN_vs_WSM

500_low

500_high

Linear (PLN_vs_WSM)

Page 44: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

345 kV branches

44

y = 1.0242x + 0.4409R² = 0.9999

-400

-300

-200

-100

0

100

200

300

400

-400 -300 -200 -100 0 100 200 300 400WS

M

Planning

PLN_vs_WSM

345_low

345_high

Linear (PLN_vs_WSM)

Page 45: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

230kV branches

45

y = 0.9929x + 0.073R² = 0.9971

-1000

-800

-600

-400

-200

0

200

400

600

800

1000

-800 -600 -400 -200 0 200 400 600 800WS

M

Planning

PLN_vs_WSM

230_low

230_high

Linear(PLN_vs_WSM)

Page 46: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

115kV branches

46

y = 0.9891x - 0.0269R² = 0.9929

-250

-200

-150

-100

-50

0

50

100

150

200

250

-250 -200 -150 -100 -50 0 50 100 150 200 250WS

M

Planning

PLN_vs_WSM

115_low

115_high

Linear(PLN_vs_WSM)

Page 47: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Improvements:

Voltage Adjustments

47

Page 48: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Voltage are much more challenging to match than MW line flow, most of them need to be adjusted manually

Line impedance, much more errors in B values than R, X

Shunts come with blocks, not a continuous value to map between cases

Generator voltage set points and Mvar output, most time you can only choose one to match, and impact other parts of the system

Difference in distribution network and topology gives different Var flow

Reactive power doesn’t travel far, you need to match voltage almost piece by piece

Type mismatches exist everywhere. It can take any forms of generator, load, switched shunt, line shunt, SVC. You need to identify them before matching them.

Challenge for Improving Voltage

48

Page 49: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Mapped the individual load Mvar value

Align key generator voltage set points, they provides supporting points throughout the network

Review branch mapping table to correct impedance difference

Align Shunts Manually

May neglect some smaller one nested in low voltage level

Starting from 500 kV, tracking down the voltage difference points and fix any model difference causing it

Pay attention to the type mismatches

Load, Gen, SVC, Switched Shunt, Line Shunt, etc.

Steps to Improve Voltage

49

Page 50: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

After load scaling and adjustment, type mismatch may show as points off.

MVar Load Improvement

50

y = 0.9845x + 0.1406R² = 0.5281-50

0

50

100

150

200

-50 0 50 100 150 200

Substation Load MVar comparison

Page 51: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Some type mismatches, shunt vs load (gen, line shunt, etc)

Smaller shunts in distribution feeder are not exactly matched

Shunt Improvements

51

y = 1.0121x + 0.0363R² = 0.9579

-800

-600

-400

-200

0

200

-800 -600 -400 -200 0 200

Substation Shunt comparison

Page 52: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Most our members chose to align voltage set points, leave MVar output to whatever values from power flow solution

Generator Mvar Output Improvements

52

y = 0.795x - 0.0946R² = 0.7505

-540

-440

-340

-240

-140

-40

60

160

-600 -500 -400 -300 -200 -100 0 100 200

Substation MVar Gen comparison

Page 53: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

All buses in Area 40 (0.6 kV ~ 500 kV)

No correction for voltage discrepancy bellow 115 kV

Voltage Improvements WSM vs planning

53

y = 0.9361x + 0.0676R² = 0.7639

0.9

0.95

1

1.05

1.1

1.15

0.9 0.95 1 1.05 1.1 1.15

WS

M

Planning

PLN_vs_WSM

Linear (PLN_vs_WSM)

Page 54: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

500 kV

54

y = 0.9066x + 0.0981R² = 0.7966

0.9

0.95

1

1.05

1.1

1.15

0.9 0.95 1 1.05 1.1 1.15

WS

M

Planning

PLN_vs_WSM

500_low

500_high

Linear (PLN_vs_WSM)

Page 55: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

230 ~ 345 kV

55

y = 0.9928x + 0.0056R² = 0.9764

0.9

0.95

1

1.05

1.1

1.15

0.9 0.95 1 1.05 1.1 1.15

WS

M

Planning

PLN_vs_WSM

230~345_low

230~345_high

Linear (PLN_vs_WSM)

Page 56: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

115 kV

56

y = 0.9395x + 0.0657R² = 0.9328

0.9

0.95

1

1.05

1.1

1.15

0.9 0.95 1 1.05 1.1 1.15

WS

M

Planning

PLN_vs_WSM

115_low

115_high

Linear (PLN_vs_WSM)

Page 57: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

MW and Mvar improvements sometimes need to be done iteratively

Linear Regression is a good way to show the comparison for a large amount of points y = ax + b

R2 = c

For line flow: 500kV +: a>0.995, c>0.999

230kV ~ 499kV: a>0.990, c>0.995

115kV ~ 229kV: a>0.985, c>0.990

a+b as close as possible to 1.0

For Voltage

a > 0.9, c > 0.9

Power Flow Experience Summary

57

y = 0.9395x + 0.0657R² = 0.9328

0.9

0.95

1

1.05

1.1

1.15

0.9 0.95 1 1.05 1.1 1.15

WS

M

Planning

PLN_vs_WSM

115_low

115_high

Page 58: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Dynamic Model Validation(8/8/2017 Event Example)

58

Page 59: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

MOD 33 Dynamic Model Validation Process

59

Page 60: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Case Preparation

60

Page 61: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Master Dynamic File

Most up to date models submitted by utilities in their original form, no modification or changes from WECC

A large amount of errors

No composite load model

DYD from WECC planning case (suggested)

WECC fixed many errors, while removed/netted some models

Models only in Peak/Light conditions, some may need to be adjusted

Composite load model need to be replaced with data at the event time

DYD from WECC mapped MOD 33 case (used for this MOD 33 study)

Models are further fixed/improved for the event time

A large amount of models being removed

Dynamic Database to Choose

61

Page 62: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

In the DYD, WECC has made some adjustments for the dynamic data:

Netted 597 generators for the reason of: renewable, bad models, low loading, etc

84 units were in area 40, sent to utility for confirmation, feedback indicates most units should not be netted

Many other generators with bad models are switched off

A concerns of accuracy as loss of governor response and inertia.

Extra attention needed for events closed to the boarder, where we don’t have other regions to review their generator being netted.

WECC Dynamic Data for 8/8 Event Case

62

Page 63: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Replace the composite load model with 3AM data

Model change for different operating mode (generator/condenser/pump …)

Fixed some errors in stability models

Adopt latest model changes from members

CG’s Further Improvement of Dynamic Data

63

Page 64: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

WSM is measurement based, Planning case can adopt other modeling assumptions. The difference may give initialization issues

Service load

Wrong MOD 32 data

Measurement error

Measurement taken in transients

etc.

Flat run may not be very flat

Initialization Issues

64

Page 65: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Two hydro units in Arizona went unstable in our event simulation after 6 seconds:

Model Changes for Operating Modes

65

109.89.69.49.298.88.68.48.287.87.67.47.276.86.66.46.265.85.65.45.254.84.64.44.243.83.63.43.232.82.62.42.221.81.61.41.210.80.60.40.20

1.01

1.01

1.01

1.009

1.009

1.008

1.008

1.007

1.007

1.006

1.006

1.005

1.005

1.004

1.004

1.003

1.003

1.002

1.002

1.001

1.001

1

1

0.999

0.999

0.998

0.998

0.997

0.997

0.996

0.996

0.995

0.995

0.994

0.994

0.993

0.993

0.992

0.992

0.991

0.991

0.99

0.99

0.989

0.989

0.988

0.988

0.987

0.987

0.986

Speed, Gen HRSMS4 (15934) #1

gfedcb

Speed, Gen MRMFLT12 (15941) #2

gfedcb

Page 66: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Both units are mapped to negative MW outputs from WSM. WSM capability allows +/- MW output, while planning case only has positive MW capability.

Discussed with Peak RC and USBR, it was confirmed as pumping mode, stability models were updated for the pumping mode.

Model Changes for Operating Modes

66

Page 67: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Model Changes for Operating Modes

67

109.89.69.49.298.88.68.48.287.87.67.47.276.86.66.46.265.85.65.45.254.84.64.44.243.83.63.43.232.82.62.42.221.81.61.41.210.80.60.40.20

1.0000

1.0000

0.9999

0.9999

0.9998

0.9998

0.9997

0.9997

0.9996

0.9996

0.9995

0.9995

0.9994

0.9994

0.9993

0.9993

0.9992

0.9992

0.9991

0.9991

0.9990

0.9990

0.9989

0.9989

0.9988

0.9988

0.9987

0.9987

0.9986

0.9986

0.9985

0.9985

0.9984

0.9984

0.9983

0.9983

0.9982

0.9982

0.9981

0.9981

0.9980

0.9980

0.9979

0.9979

0.9978

0.9978

0.9977

0.9977

0.9976

0.9976

0.9975

0.9975

0.9974

Speed, Gen HRSMS4 (15934) #1

gfedcb

Speed, Gen MRMFLT12 (15941) #2

gfedcb

Page 68: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Event Analysis

68

Page 69: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Why We Need Event Analysis

69

59.8

59.85

59.9

59.95

60

60.05

0 20 40 60 80 100 120 140

WSM snapshot is saved every 5~6 minutes

When an event happened between 2 snapshots, how confident for you to use the system condition of pre-fault snapshot as the time event happened?

If no major changes (fault, switching of 500 kV line, outage of large power plant, etc. )

If the snapshot is quite close to the time fault happened

Then what if

Event is minutes after the pre-fault snapshot?

Your system is experiencing some changes while

event happens

Using a frequency comparison as example

Simulation nadir < measurement nadir

Load tripping & stall too much?

Other generation tripping elsewhere?

Inertia difference?

Page 70: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Event Analysis

(Steady State)

70

Page 71: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Comparison between 2 WSM snapshots at 3:04am and 3:10am

Event Analysis: Generation

71

-200

0

200

400

600

800

1000

1200

1400

-200 0 200 400 600 800 1000 1200 1400

3:1

0a

m D

isp

atc

h

3:04am Dispatch

Colstrip 1P. Rapids 1

Navajo 1&3

RockyR 7Castic 3

Mica 5

Mica 3

Colstrip 3&4

Page 72: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Generation switched off

Colstrip #1,3,4 (confirmed triggered by events)

Rocky Reach #7 (34.4 MW, out for maintenance at 3:07am, confirmed by Chelan PUD)

Significant generation reduction

Navajo #1&3 (Arizona, -122 MW)

Castaic #3 (pumped storage in LADWP, pre-condensing, post-pumping, -34.7 MW)

Mica #3, 5, 6 (in BCH, -146MW)

Priest Rapids #1 (-21.5 MW, while PR2 + 6.7MW)

Generation increasing

No abnormal generator increasing observed

3:10am was 2 minutes after the events, operators may adjust generators output manually

Using the comparison table can help to track down & add back governor response for units being netted by WECC

Event Analysis: Generation

72

Page 73: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Comparison MW flow between 2 WSM snapshots at 3:04am and 3:10am

Line flow difference mostly related to the events and dispatch changes

Event Analysis: Branch

73

Page 74: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Line Status changes except already involved in the events

Rocky Reach #7 breaker (out for maintenance at 3:07am, confirmed with Chelan PUD)

Switched on Noxon reactor (-51.3Mvar), when ramp up Noxon #5 from condenser mode to 42MW

Switched on Willis capacitor 7.4 Mvar

Switched on Cougar (EWEB) generator unit #2

Switched off Dillons (NWE) capacitor 7.5 Mvar

No significant branch status change in our region between 3:04am ~ 3:10am

Event Analysis: Branch

74

Page 75: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Event Analysis

(Dynamics)

75

Page 76: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Revisit the event sequence where Colstrip units trip

Colstrip ATR Model

76

Time (MDT) Time From

Fault

Cycles From

Fault

Event

04:08:18.279 00:00:00.00

0

0.00 A-Phase fault on Colstrip - Broadview 500 kV Line B; Z=0+ j0.0470

pu 35% from Broadview

04:08:18.329 00:00:00.050 3.00 Fault cleared by opening Colstrip - Broadview 500 kV Line B

04:08:18.647 00:00:00.368 22.08 Series capacitors on the Colstrip – Broadview A line bypassed

04:08:18.687 00:00:00.408 24.48 Colstrip - Broadview 500 kV Line A opens

04:08:18.833 00:00:00.544 32.64 Colstrip Unit 3 Tripped

04:08:18.833 00:00:00.544 32.64 Colstrip Unit 4 Tripped

04:08:18.917 00:00:00.638 38.28 Colstrip Unit 1 Tripped

04:08:23.932 00:00:05.653 339.18 Broadview 500/230 kV Bank 3 tertiary reactors inserted

04:08:28.943 00:00:10.664 639.84 Broadview 500/230 kV Bank 4 tertiary reactors inserted

Page 77: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Simulation of ATR model

77

Contingency Name Cycles From Fault Event

August_8_events 37 Colstrip Unit 2 Tripped

August_8_events 38 Colstrip Unit 3 Tripped

August_8_events 41 Colstrip Unit 4 Tripped

Page 78: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Remove ATR, manually Add Tripping

78

Page 79: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Keeler SVC connected 230 kV and 500 kV buses simulation shows a much higher post fault voltage

Keeler SVC Post Fault Voltage

79

1.018

1.02

1.022

1.024

1.026

1.028

1.03

1.032

-100 -50 0 50 100 150 200 250 300 350

Time

1.06

1.062

1.064

1.066

1.068

1.07

1.072

1.074

1.076

-100 0 100 200 300 400

Time

Page 80: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

All nearby reactors have already on pre-fault, and the SVC is at 70 Mvar capacitive, selected to switched off during contingency

230 kV voltage drop too low, all buses voltage show a much slower ramp rate (not right)

It implied the SVC should be on, but the dynamic performance may not be accurate in simulation

Is the SVC tripped?

80

1.005

1.01

1.015

1.02

1.025

1.03

1.035

-100 -50 0 50 100 150 200 250 300 350

Time

1.056

1.058

1.06

1.062

1.064

1.066

1.068

1.07

1.072

1.074

1.076

-100 0 100 200 300 400

Time

Page 81: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Adding a dummy reactor (-30Mvar) and switched in during the fault

The voltage/ramp rate matches well, may suggested SVC post-fault Mvar should be 30Mvar less

Post fault SVC Output

81

1.016

1.018

1.02

1.022

1.024

1.026

1.028

1.03

-100 -50 0 50 100 150 200 250 300 350

Time

Page 82: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Switching and voltage discrepancy (low simulated voltage) observed in east side

Reactive Switching in East Side

82

1

1.01

1.02

1.03

1.04

1.05

1.06

1.07

1.08

-100 0 100 200 300 400

Time

Bell 230kV

0.9

0.95

1

1.05

1.1

1.15

1.2

-100 0 100 200 300 400

Time

Garrison 500kV

1

1.02

1.04

1.06

1.08

1.1

1.12

1.14

1.16

-100 0 100 200 300 400

Time

Taft 500kV

Page 83: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

A RAS action should be modeled to switch out Garrison Reactor (-218.3Mvar) after 5 cycles of Colstrip line tripping

A much better post transient voltage match for 4 second

Two additional steps for switching observed (reverse direction), using the Noxon reactors (2 step@ -50 Mvar), no stability model but switched in through contingency

Reactive Switching in East Side

83

0.9

0.95

1

1.05

1.1

1.15

1.2

-80 -60 -40 -20 0 20 40

Time

Garrison 500kV

GARR 500 B1 SA .B500EAST_____1VP VPM kV GARR 500 B1 SA

GARRISON_500.0_TSBusVinKV

Page 84: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Bell 230 kV voltage now matched perfectly, not enough for Garrison and Taft.

AVA confirmed Noxon reactors switching using the recorded data

Reactive Switching in East Side

84

1

1.02

1.04

1.06

1.08

-100 0 100 200 300 400

Time

Bell 230 kV

BELL 230 B1 SA .B230SECT1____1VP VPM kV BELL230 B1 SA

'BELL S0_230.0'_TSBusVinKV

0.9

0.95

1

1.05

1.1

1.15

1.2

-80 -60 -40 -20 0 20 40

Time

Garrison 500kV

GARR 500 B1 SA .B500EAST_____1VP VPM kV GARR500 B1 SA

GARRISON_500.0_TSBusVinKV

Page 85: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Dynamic Comparison in PMU/DFR/PI/SCADA

(in pdf and Excel)

85

Page 86: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Dynamic Comparison Discrepancy

86

Page 87: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Frequency nadir: simulation higher than PMU

Lower inertia in simulation, should results in lower nadir of simulation

Confirmed no other related generator tripping during the events except Colstrip

Load model becomes the most likely reason, utilities is working on improving load composition

Frequency Dip

8759.8

59.85

59.9

59.95

60

60.05

-100 -50 0 50 100 150 200 250 300 350

Fre

q

Time

Page 88: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

At the time of fault, PDCI was ramping down its MW flow

A coincidental change that help to relief impact from Colstrip tripping

A more accurate MOD 33 simulation should include a model to mimic the flow ramping.

PDCI Flow Change During the Fault

88

3.7

3.75

3.8

3.85

3.9

3.95

4

4.05

4.1

-100 -50 0 50 100 150 200 250 300 350

'BIG EDDY_500.0' CELILO1_500.0 1_TSACLineFromP

BGED 500 B1 SA .A500CELILO___1MW

'BIG EDDY_500.0' CELILO1_500.0 1_TSACLineFromP

Page 89: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

PMU shows a flow increase on Longvanx to Mintfarm 230kV line

PSE’s Mint Farm’s gas turbine was set as “base load” at Pmax. PSE provided a new Gas governor model validated in March, it shows a significantly increased capability. This governor model will be updated by PSE in their future modeling data submission

Line Flow around Longview

89

-2.75

-2.7

-2.65

-2.6

-2.55

-2.5-100 -50 0 50 100 150 200 250 300 350

LONGVANX_230.0 MINTFARM_230.0 1_TSACLineFromP

LONG 230 B2 SA .A230MINTFARM_1MW

LONGVANX_230.0 MINTFARM_230.0 1_TSACLineFromP

Page 90: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

PMU shows a flow increase on McNary S1 to Ph2 line

MCN#1 is off, MCN#2 is on with a IEEE_G3 governor model. May either switching on MCN#1 and ramp up the power, or imply a wrong governor response for #2.

USACE confirmed a recent updated the governor to H6E model, provide to BPA in August.

Line Flow around McNary

90

-0.7

-0.6

-0.5

-0.4

-0.3

-0.2

-0.1

0-100 -50 0 50 100 150 200 250 300 350

'MCNRY S1_230.0' 'MCN PH2_230.0' 2_TSACLineFromP

MCNY 230 B1 SA .A230MCNARYPH_2MW

'MCNRY S1_230.0' 'MCN PH2_230.0' 2_TSACLineFromP

Page 91: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Coulee # 19 switched from condenser model to generator mode due to low frequency ~ 270 seconds later

Anybody knows why it absorbing more power (-15MW -> -30MW) before it starts to generate? (Gate control? Motor starting? … )

Switching from Condenser to Generator

91

-0.2

-0.1

0

0.1

0.2

0.3

0.4

-100 -50 0 50 100 150 200 250 300 350

COULEE_500.0 COULEE19_500.0 1_TSACLineFromP

GCFI 500 B1 SA .A500GEN19____1MW

COULEE_500.0 COULEE19_500.0 1_TSACLineFromP

Page 92: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

R1.3: Which dynamic comparison are acceptable, or not?

R1.4: If not, how can we resolve them?

MOD-33 Compliance Requirements

92

Page 93: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Load model improvement (very very long time)

Gen/Excitation model validation (MOD-26, 0~5 years)

Governor model validation (MOD-27, 0~5 years)

New model development (PDCI ramping, condenser control mode switching etc., several months)

Adjustments to existing model (parameters, settings, mode, etc., days ~ months)

+

Software issues (random time)

Measurement issues (random time)

So far, to resolve “Unacceptable Discrepancy”

93

Page 94: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Other Usage for MOD 33 Case: MOD 26 Example

94

Page 95: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

SCADA data sampled 1Hz for Ifd, Efd, Vt, P & Q,

Snohomish Jackson U2

95

560

570

580

590

600

610

620

630

-150.00 -100.00 -50.00 0.00 50.00 100.00 150.00 200.00

Jackson U2 Ifd

U2_AVR.GeneratorFieldCurrent

JACKSN2_13.80 1_TSGenIfd

60

80

100

120

140

160

-200.00 -100.00 0.00 100.00 200.00

Jackson U2 Efd

U2_AVR.GeneratorFieldVoltage

JACKSN2_13.80 1_TSGenFieldV

-4

-3

-2

-1

0

1

-200.00 -100.00 0.00 100.00 200.00

Jackson U2 Mvar

U2_AVR.GeneratorMegavars

JACKSN2_13.80 1_TSGenQ

13.75

13.8

13.85

13.9

13.95

14

14.05

14.1

14.15

-200.00 -100.00 0.00 100.00 200.00

Jackson U2 Vterm

U2_AVR.GeneratorTerminalVoltage

JACKSN2_13.80 1_TSGenTermVPU

Page 96: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Other Usage for MOD 33 Case: MOD 27 Example

96

Page 97: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

4 co-gen units connected to a 115kV substation in Puget Sound Energy. DFR installed in 115kV substation (high side of GSU), no direct measurements on plant terminal

Current from Unit 1 (lower side of GSU) can be derived from current (vector summation) on breaker 3192 and 3194.

DFR in 115kV substation

97

Page 98: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Phase current magnitudes in Amps are measured by DFR, no phase angle available:

3192 current >> 3194 current; 3194 current is constant before & after fault

Use 3192 current to approximate generator current from unit #1

DFR Current

98

185

190

195

200

205

210

215

0 20 40 60 80 100 120 140

3192 IA

23.5

24

24.5

25

25.5

26

26.5

27

27.5

28

0 20 40 60 80 100 120 140

3194 IA

Page 99: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Simulation vs DFR

A good match for pre-fault and first spike current in magnitude

DFR shows unit #1 settled to about 8 amps phase current higher, while simulation keep the current flat

DFR Current

99

180

185

190

195

200

205

210

215

0 10 20 30 40 50 60

Gen 1 Current comparison

3192 IA

Simulation from Gen 1

Page 100: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

CG and PSE reviewed the modeling & test reports

2011 original modeling report, 2013 model validation report, 2017 MOD26&27 report

Confirmed the stability models in used are consistent with the reports

Believe a wrong trate value is used. Set trate = 39.5, and governor response limit to “normal”, re-run the simulation

Existing Modeling Document Review

100

180

185

190

195

200

205

210

215

0 20 40 60 80 100 120 140

Gen 1 Current comparison

Page 101: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

GO needs to validated their generators every 5 years

Can use either system events or staged tests

MOD 33 case is a system event case

Some reasons you may want to check MOD 33 case for exciter/governor modeling verification

Staged tests doesn’t verify you a full list of parameters

Model parameters (performance) may vary under different conditions (temperature, loading, operating mode, etc.)

Always good to check periodically for your machine models

TO may raise the question to GO if some discrepancy found

Verify the models from consultants

System event validation complement but not fully replace staged tests

NERC MOD 26/27

101

Page 102: MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Principle_Pacific Northwest... · MOD-33 Principle, Criteria and Methodology: Pacific Northwest’s Experience Bo

Thank you

Bo Gong, [email protected]

102