TSAT-RTDS Interface - The Development of a Hybrid Simulation Tool

Xi Lin, Pouya Zadehkhost

Powertech Labs Inc.

1

Presented at panel session “Challenges and solutions of interfacing techniques for EMT/TSA hybrid simulation - Industry perspectives”

2017 IEEE PES General Meeting

Jaegul Lee, Jiyoung Song, Baekkyeong Ko

KEPCO

Kyeon Hur

Yonsei University

Feature Electro-Magnetic Transients

(EMT) Phasor Domain

(TSA)

Sample Programs PSCAD, RTDS TSAT, PSS/E, PSLF

Level of details Three-phase instantaneous values

Detailed models

Phasor-domain positive sequence

Simplified dynamic models Network dynamics ignored

Size of modeled system

Varies between a few to several hundreds of buses

Often used for simulating systems with tens of thousands of buses

Common Application Any types of studies that need detailed modeling

Hardware in Loop (HIL) simulation

Bulk power system planning and operation

On-line Dynamic Security Assessment

Power System Simulation Methods

• Power system dynamics are conventionally categorized into low- and high-frequency transients

• Two groups of industrial-grade tools have been developed based on this categorization

2

Conventional Simulation Methods -Challenges

• Can focus on either detailed models in small system or simplified models in large system – Increasing level of details without reducing system size can be costly

• Study interactions between system-wide events and detailed devices can be challenging, e.g. – Fault analysis in HVDC systems

– Sub-synchronous resonance studies

• A detailed model might be available only in an EMT package, e.g. – HVDC systems, renewable generators, FACTS devices, etc.

• To built a full system model for EMT simulation is challenging – While this is a common practice in TSA studies

3

Hybrid Simulation

• Hybrid simulation approach addresses these challenges by using both EMT and phasor-domain simulation methods

• Advantages – Effective in analyzing impact of low-frequency oscillations on specific components and

vice-versa

– A cheaper solution for studying large systems compared to full-EMT simulation

– Takes advantage of rich modeling library available in EMT and phasor-domain simulation packages

– Perform Hardware-In-Loop simulation with a large system model

4

TSAT-RTDS Interface (TRI)

• A tool for performing hybrid simulation studies – Using both TSAT from Powertech Labs and RTDS from RTDS Technologies

• TRI is developed with special focus on practical aspects – User-friendly, minimizing case setup efforts, simplifying results analysis steps etc.

• How does TRI work – TSAT simulates external system at normal time-step (e.g. 4ms)

– RTDS simulates internal system at normal time-step (e.g. 50us)

– Boundary injections are exchanged at the end of every TSAT time-step

5

RTDS Time Steps

IB(t)VB(t) VB(t+2ΔT)IB(t+ΔT) IB(t+2ΔT)VB(t+ΔT)

TSAT

RTDS

TSAT Time Steps

TRI Structure

6

FPGA Board TSATRTDS

Xilinx VC707 FPGA Board

(mounted on PCI Express slot of PC which runs TSAT)

RTDS Case

Internal System

Boundary 2

Boundary 3

Boundary 1

TSAT Case

External System

Boundary 2

Boundary 3

Boundary 1

Representation of External System in EMT

• Approach 1 – simple Norton (or Thevenin) Equivalent – External system is modeled as a Norton

equivalent

– Easy-to-use since TSAT automatically • calculates Thevenin impedance

• updates Norton source current

– High frequency transients of the external system are ignored • may fail when fault is applied at boundary

– A buffer zone between internal and external systems is recommended

7

Z th2

Ith1

Ith2

Z th1

...

...

Inte

rnal

Sy

stem

Bu

ffe

r Zo

ne

Representation of External System in EMT (cont’d)

• Approach 2 – Frequency Dependent Network Equivalent (FDNE) – External system is modeled as a

frequency-dependent mathematical model plus Norton current source(s)

– More accurate than Norton (Thevenin) equivalent

– May not need buffer zone

– Difficult to build the FDNE • Numerical stability

• Computation burden on the EMT

• Especially difficult for multi-port (2+)

– Handling network changes in external system is challenging

8

...

...Inte

rna

l Sys

tem

FDN

E M

od

el Ith1

Ith2

TRI Features

• Supports single-port and multi-port boundaries

9

G

...

...

...

G

...

...

...

G

...

...

...

One-port

system

Two-port

system

TRI Features (cont’d)

• Supports both (simple) Norton Equivalent and FDNE

• Potential TSAT-RTDS Configurations

10

RTDS Case

TSAT Case 1

TSAT Case 2

TSAT Case 3

TSAT Case

RTDS Case 1 RTDS Case 2

Simulation Case Setup – RTDS Side

• RTDS case is being setup as normal

• With addition of GTFPGA and TSA-Interface

11

Handling data exchange

with FPGA board

Custom model

representing one

TSAT boundary

Simulation Case Setup – TSAT Side

• TSAT case is being setup as normal – Commonly used planning data

– With addition of the Hybrid Simulation data

12

Provided in a

typical TSA study

Provided in hybrid

simulation study

Simulation Case Setup – TSAT Side (cont’d)

13

Defining boundaries between

internal and external systems

RTDS quantities may be

monitored on TSAT side

(optional)

System Setup – Hardware

14

FPGA Board mounted on PCI

Express slot

RTDS PB5 card connected to FPGA

Board through an optical fiber

Starting Hybrid Simulation

15

TSAT waits during

RTDS start-up

Boundary mismatch during

RTDS start-up can be monitored

User notifies TSAT once RTDS starts-

up (may automate this step in future)

TSAT window

RTDS window

Running Hybrid Simulation

16

TSAT and RTDS run simultaneously

A disturbance applied in

one tool affects the

other

TSAT window RTDS window

Case Study #1

17

• IEEE 39-bus test system

• Fault applied in internal system (bus 28)

• Generator at bus 38 is monitored

26

0.00

139.00

26

0.00

139.00

-107.72

-57.70

29

24.84

261.85

28

25.26

188.5028

29

6.80

309.40

7.36

-308.08

24.84

261.85

6.80

25.26

188.50

309.40

7.36

-308.08

-0.85

293.31

38

103.67

830.001

38

41

103.67

830.001

41

-118.71

-32.62

119.58

42.01

-33.40

258.76

64.97

-255.88

108.55

-374.61

-106.95

374.61

-118.71

-32.62

-33.40

108.55

-374.61

-106.95

374.61

258.76

64.97

-255.88

119.58

42.01

-11.99

-292.16

55.63

58.01

Internal system

modeled in RTDS

Case Study #1 – Generator Rotor Speed

18

Case Study #1 – Generator Terminal Voltage

19

Case Study #2

20

• IEEE 39-bus test system with 4 ports

Case Study #2 – Generator Rotor Speed

21

Case Study #1 – Generator Terminal Voltage

22

Case Study #3

23

• A practical case with 2189 buses and 459 generators – Two generators modeled in RTDS

– Rest of system is modeled in TSAT (2 ports)

• Contingency description – Fault is applied on TSAT-side (2 buses away from one of boundaries)

– Cleared after 0.1 seconds

• A generator close to fault is monitored

• Long time simulation test – The fault is applied and cleared every ~200 seconds

– Simulation ran for 3 hours

– TSAT simulation keeps synchronization with RTDS

• Hard real time

• Using a Intel Core i7 7700K 4.2 GHz CPU (one core is used) and 32 GB RAM

• 4ms time-step on TSAT side (50us on RTDS side)

Case Study #3 – Generator Rotor Speed

24

Case Study #3 – Generator Terminal Voltage

25

Summaries

• Why using hybrid simulation? – Takes advantage of both EMT and phasor-domain simulation packages

– Facilitates analyzing interactions between low- and high-frequency transients

• TSAT-RTDS Interface – Performs hybrid simulation studies using TSAT and RTDS

– Practical aspects have been one of the main objectives

– Preliminary testing demonstrated that the tool is promising

– Allows monitoring interactions that may be missed in pure EMT or pure phasor-domain simulations

26