1 of 11

Stability Assessment Using Composite Load Model in Transmission Systems

​

​

November 27th 2024

Daniel Santos

GERS

Di Wu

NDSU

Andrés Zapata

GERS

Diego Rodríguez

GERS

2 of 11

Methodology�Composite Load Model

Develop load models using the output from the PNNL (NERC) LMDT software. Create baseline cases with all network upgrades, the CMLD model, and simulation contingencies.

Analyze results, define a set of CMLD parameters to be modified (with the SAS members), and prepare the sensitivity cases.

Repeat simulations with the sensitivity cases and analyze differences from baseline case (~10,200 simulations)

3 of 11

Sensitivity Analysis �Key Performance Indicators (KPI)

The composite load model is considered the state-of-the-art in dynamic load modeling, enabling representation of three phase and single-phase induction motors, power electronic load, static load, distributed generation, and the distribution equivalent network.

​

For each parameter variation, the same set of 162 contingencies was run, and the dynamic response of the system was evaluated considering the following criteria:

​

    • Total load loss per motor type (3Φ and 1Φ)
    • Total generation loss (synchronous and IBRs)
    • Total generation redispatch
    • Overvoltage to 120% for more than 20 s for monitored buses
    • Voltage recovery to 80% for more than 20 s for all buses

x20

x95

x18

x20

x20

x35

x4

x11

x15

4 of 11

Study Case and Assumptions�Composite Load Model (CMLD)

LoadComp

FmA

FmB

FmC

FmD

Fel

COMERCIAL

0.09

0.273

0.059

0.022

0.355

RESIDENTIAL

0.104

0.087

0.085

0.435

0.072

RURAL AND AGRICULTURAL

0.246

0.095

0.076

0.262

0.095

MIX (COM AND RES)

0.241

0.114

0.072

0.239

0.109

Voltage level

[kV]

Buses

Lines

Loads

Shunt

Generators

69

1182

1353

730

154

-

115

667

784

427

74

-

138

600

701

417

55

-

230

746

977

301

54

-

Total

3195

3815

1875

337

580

  • The stability evaluation was for 2024 using the light load (50%) and peak load (100%) scenarios.

​

  • The total load for the peak load (100%) case is 51.7 GW. Residential loads play a relevant role accounting for 59% of the total load in the system.

 

Geographical load location. Purple: JAX, blue: TPA, and green: MIA

5 of 11

Sensitivity Analysis Report �Values

Note:

  • Contactor settings (Vc1off, Vc2off, Vc1on, Vc2on) are changed simultaneously. The rest of the sensitives only changed one parameter at the time
  • (*) vcmax2 considers Vstall = 0.6

Based on NERC recommendations, these are the list of parameter modified in the study from a total of 133:

Simulation Settings:

​

  • Simulation time = 60 s
  • Time Step = 0.004167 s
  • Voltage recovery function:

​

    • 0.8 pu for more than 20 s
    • 1.2 pu for more than 20 s

​

Parameter

Base

(p.u. or s)

Sensitivity 1

Sensitivity 2

Sensitivity 3

Motor D Vstall - compressor stall voltage (pu)

0.45

0.4

0.425

0.45

Motor D Tstall - stall time delay (sec)

0.033

0.033

0.05

999

FmD - Motor D Fraction

FmD

0.85*FmD

1.15*FmD

 

Tth - compressor motor heating time constant (sec)

15

5

10

15

Frst - fraction of motors capable of restart

0.2

0.2

0.4

0.6

Fuvr - fraction of motors with U/V relays

0.1

0

0.1

0.5

Contactor settings:

Base

Sensitivity 1

vcmin

Sensitivity 2

vcmax

Sensitivity 3

vcmax2 (*)

VC1off - Voltage 1 at which contactors start dropping out (pu)

VC2off - Voltage 2 at which all contactors drop out (pu)

VC1on- Voltage 1 at which all contactors reclose (pu)

VC2on - Voltage 2 at which contactors start reclosing (pu)

 0.5

0.4

0.6

0.5

0.3

0.2

0.4

0.3

0.65

0.55

0.75

0.65

0.65

0.55

0.75

0.65

Around 10,200 simulations

6 of 11

Transient Stability Results �Composite Load Model – General Impacts

 

The values shown in this table are the total sum for the 162 contingencies under study.

*The values in the generation lost column do not reflect the generation tripped by clearing sequence.

Voltage response using CMLD and CLOD base cases

Case

Load Model

Generation Lost

SyncGen Lost

PVGen Lost

Gen. Redispatch

3Ph Motor disconnection

1Ph Motor disconnection

Overvoltage

[GW]

Delta (%)

[GW]

Delta (%)

[GW]

Delta (%)

[GW]

Delta (%)

[GW]

[GW]

[N]

50%

CLOD

17.4

​

17.4

​

0

​

0.7

​

-

-

0

CMLD

28.8

66%

28.8

66%

0

-

2.5

277%

9.8

13.2

31

FI

CLOD

26.0

 

24.7

 

1.3

 

0.0

 

-

-

0

CMLD

142.7

450%

127.7

417%

15.0

1081%

25.2

-

29.2

27.6

0

HI

CLOD

27.3

 

25.9

 

1.4

 

0.9

 

-

-

0

CMLD

74.9

174%

66.5

157%

8.4

496%

19.0

2001%

38.8

40.0

0

7 of 11

Transient Stability Results �Composite Load Model – General Impacts - Peak Firm Import

Generally, Vcmin, Vcmax, FmD-15, FmD+15, and Fuvr (> 0) have the highest impacts on the system

 

  • The values shown in this table are the total sum for the 162 contingencies under study.
  • The same analysis is made for the 50% off-peak and Peak HI cases

*The values in the generation lost column do not reflect the generation tripped by clearing sequence.

Case

Generation Lost

SyncGen Lost

PVGen Lost

Generation Redispatch

3Ph Motor disconnection

1Ph Motor disconnection

[GW]

Delta (%)

[GW]

Delta (%)

[GW]

Delta (%)

[GW]

Delta (%)

[GW]

Delta (%)

[GW]

Delta (%)

Base-CMLD

142.7

​

127.7

​

15.0

​

25.2

​

29.2

​

27.6

​

vstall04

105.8

-26%

97.8

-23%

8.0

-46%

11.4

-55%

23.0

-21%

26.0

-6%

vstall0425

87.8

-38%

79.6

-38%

8.2

-45%

17.0

-33%

21.7

-26%

24.4

-12%

tstall999

78.6

-45%

74.1

-42%

4.4

-71%

5.8

-77%

20.3

-30%

26.8

-3%

tstall005

131.8

-8%

119.2

-7%

12.6

-16%

12.4

-51%

24.2

-17%

26.2

-5%

frst04

141.7

-1%

126.8

-1%

15.0

0%

24.5

-3%

29.3

0%

27.6

0%

frst06

137.7

-3%

122.8

-4%

14.9

-1%

21.4

-15%

29.2

0%

27.6

0%

vcmin

814.1

471%

711.8

457%

102.3

583%

20.0

-21%

105.9

263%

39.7

44%

vcmax

31.2

-78%

29.4

-77%

1.9

-88%

0.7

-97%

7.4

-75%

16.4

-41%

tth5

127.6

-11%

116.3

-9%

11.3

-25%

40.0

59%

27.8

-5%

26.8

-3%

tth10

138.8

-3%

124.6

-2%

14.2

-5%

22.8

-9%

28.9

-1%

27.2

-2%

fuvr0

155.6

9%

138.4

8%

17.2

15%

37.4

48%

34.2

17%

3.1

-89%

fuvr05

45.3

-68%

42.9

-66%

2.4

-84%

16.3

-35%

10.1

-65%

99.9

262%

FmD-15

42.6

-70%

40.9

-68%

1.7

-89%

0.3

-99%

9.9

-66%

8.4

-70%

FmD+15

208.4

46%

183.3

44%

25.1

68%

53.1

110%

41.5

42%

39.2

42%

8 of 11

Transient Stability Results �Composite Load Model

Sensitivity for Fraction of Motor D (FmD)

Case: FI Contingency: JEA_E2-D_89

Fault Location

Fault Description: 3Φ fault on SE JAX 230kV w/ non-red relay failure

FmD fraction base: 44% in Residential load and 25% in Mix and Rural Agricultural

Generation disconnected (3Φ flt): CLOD 1131MW and CMLD 1783MW

CMLD load disconnected (3Φ flt): 145MW of 3Φ motors and 146MW of Motor D

Blue Line: 500kV Duval Line

Red points: Contingencies location

Green points: Measurement Points

Voltage Criteria:

  • dVb0
  • dTr
  • dVm

Variable

Description

dVo (%)

First sample voltage dip

dVa (%)

Persistence voltage dip

dTa (cyc)

Persistence time

dVb (kV)

dVb in kV

dVb0 (%)

Lowest value of dVb in %

dVb1 (%)

Highest value of dVb in %

dTb (cyc)

Time from lowest to highest dVb

dVm (%)

Peak voltage overshooting

dTr (s)

Time to reach initial volt level

dTm (s)

Time to reach maximum voltage

9 of 11

Impact of Parameter Sensitivities �Compared to Default NERC Data Set for Motor D

9

Pmt Name

Generation Lost

Generation Redispatch

3Ph Motor disconnection

1Ph Motor disconnection

NDDS (+)

NDDS (-)

NDDS (+)

NDDS (-)

NDDS (+)

NDDS (-)

NDDS (+)

NDDS (-)

Vstall

NA

 ↓ in observations

NA

Large ↓ in observations

NA

 ↓ in observations

NA

Small ↓ in observations

Tstall

 ↓ in observations

NA

Large ↓ in observations

NA

 ↓ in observations

NA

Small ↓ in observations

NA

FmD

 Large ↑ in observations

 Large ↓ in observations

 Large ↑ in observations

 Large ↓ in observations

 Large ↑ in observations

 Large ↓ in observations

 Large ↑ in observations

 Large ↓ in observations

Tth

NA

Small ↓ in observations

NA

 Large ↑ in observations

NA

Small ↓ in observations

NA

Small ↓ in observations

VC

 Large ↑ in observations

 Large ↓ in observations

 ↓ in observations

 Large ↓ in observations

 Large ↑ in observations

 Large ↓ in observations

 Large ↑ in observations

 Large ↓ in observations

Frst

Small ↓ in observations

NA

Small ↓ in observations

NA

Not change

NA

Not change

NA

Fuvr

 Large ↓ in observations

 Small ↑ in observations

 Large ↓ in observations

 Large ↑ in observations

 Large ↓ in observations

 Small ↑ in observations

 Large ↑ in observations

 Large ↓ in observations

NDDS (+): indicates a parameter increase from the default value

NDDS (-): indicates a parameter decrease from the default value

10 of 11

Assessment Recommendations �Further Studies

�

�

    • In this paper, we presented a methodology for transient stability evaluation using the Composite Load Model - CMLD.

​

    • This methodology is recommended for transmission planners and operators that will migrate their dynamic load models to more sophisticated mathematical representations. This methodology was applied to and verified with one of the biggest interconnected systems in the US.

​

    • Based on the simulation results, it is advisable to use the current accepted model while including the FIDVR from the single-phase motor components. Some parameter adjustments are recommended for the transitional model’s migration to include the FIDVR and to avoid over-pessimistic or optimistic network responses in the evaluation. Over-pessimistic results may generate transmission network upgrades not required by the National Grid (Sync condensers, new lines, etc.)

​

    • Additional model validation, including real responses in the system, is recommended to fine tune the model and avoid unnecessary network upgrades.

​

11 of 11