Panos Kotsampopoulos: Smart RUE-National Technical University of Athens

Felix Lehfuss: Austrian Institute of Technology - AIT

Georg Lauss: Austrian Institute of Technology - AIT

Benoit Bletterie: Austrian Institute of Technology - AIT

Nikos Hatziargyriou: Smart RUE-National Technical University of Athens

Limitations of Digital Simulation and Advantages of PHIL Testing

for DG providing Ancillary Services

DERlab General Assembly-Workshop, Paris, March 2015

Evolution of power system/component

analysis, testing and validation

Full System Hardware Testing

Component Hardware Testing Pure Simulation Methods

Hardware in the Loop (HIL): System Testing

Full system digital simulationNumeric analysis

Grid Simulator

InverterDuT

PV arraySimulator

Power Amplifier

InverterDuT

Power Amplifier

Network, Substation

PV arrays

AC side DC side

AC side DC side

Comm. Comm.

InverterDuT

RTS:PV arrays

RTS: Network,

Substation

PowerPower

PHIL Fault-Ride-Through tests

on PV inverters

Simulated network according to

Standards for Fault-Ride-Through

Fault in the simulated network in the RTDS

The hardware PV

inverter is disconnected

following the relevant

standard

The settings of the hardware-software relays are adapted to the network

conditions (e.g. DG on or off)

Adaptive protection testing via

HIL simulation

PHIL testing of ancillary services by

inverter-based DG

Inverter-based DG are particularly difficult to model

accurately

Therefore the “classic” pure digital simulation can face

limitations

PHIL makes use of the real DG and can therefore reveal the

actual system behaviour Software Hardware

grid-connected

PV inverter

DC sourceline impedance

transformer

programmable

voltage source

HV network MV line 10km

OLTC control

A

B

C

A

B

C

A

B

C

A

B

C

A

B

C

A

B

C

DC +

DC -

+ DC

- DC

Ssc=1 GVA

R/X=1/6110/30kV, 30 MVA

Ux=11.92%, Ur=0.37%

R=0.253 Ω/km

X=0.355 Ω/km

measurement: V

command:

tap change

New scaling and stability-accuracy method

PHIL allows scaling of the software and hardware. A 2kVA hardware LV PV

inverter is used to evaluate the integration of a 15MVA PV park connected

to the MV. The grid voltage control algorithm of the full and

reduced-scale DGs are exactly the same.

Shifting part of the software impedance on the hardware side is proposed.

Stability can be achieved without compromising accuracy (if necessary a

“smaller” feedback filter can be used)

e-sΤ1

Gfilter(s)

GLPF(s)

(optional)

1/α

RMV·b/α -

Rsh·α

ΧMV·b/α -

Χsh·α

Rsh Χsh

Inverter

filter

Shifted Hardware

Impedance

Equivalent part of

the MV line

Software: MV Hardware: LV

Vcn_s

+

-

A+

iHUT

Rtransf·b/α

Xtransf·b/α

e-sΤ2

Vcn_h

V se

cond

ary

V in

verte

r

+

-

+

-

HW

SW

V

Va

redHuT

fullHuT

S

Sb

_

_

.)(

)()(

sZa

sZa

bsZ

sh

SWSW

Accuracy Improvement with the proposed

method

0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 200010

-3

10-2

10-1

100

101

102

103

frequency (Hz)

Err

or

(%)

No shifting, LPF 500 Hz

Shifting, No LPF

GLPF(s)ZSW(s) 1/ZHW(s)

Vs e-sTd Gfilter(s)+VzHW

-

GLPF(s)ZSW(s)−a·Zsh(s) 1/ZHW(s)

Vs e-sTd Gfilter(s)+VzHW

1/a- ZHW(s)+Zsh(s)

ZHW(s)

(a)

(b)

1/a

No need for feedback filter

with the shifting impedances

→ Much higher accuracy

Formula for the minimum

shifted impedance to achieve

stablity was derived

Down-scaling of the voltage

improves the stability

Up-scaling of the current

deteriorates the stability

DG and OLTC interactions: cosφ(P)

DG operates with cosφ(P) control

DG active power increases, stays constant and then decreases

Recurring tap-changes occur

Good accuracy of the PHIL test due to the proposed method

Software Hardware

grid-connected

PV inverter

DC sourceline impedance

transformer

programmable

voltage source

HV network MV line 10km

OLTC control

A

B

C

A

B

C

A

B

C

A

B

C

A

B

C

A

B

C

DC +

DC -

+ DC

- DC

Ssc=1 GVA

R/X=1/6110/30kV, 30 MVA

Ux=11.92%, Ur=0.37%

R=0.253 Ω/km

X=0.355 Ω/km

measurement: V

command:

tap change

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80

0.97

0.98

0.99

1

1.01

1.02

1.03

1.04

1.05

1.06

time (s)

Vo

lta

ge

p.u

.

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80

-4

-3

-2

-1

0

1

time (s)

tap

po

sitio

n

tap PHIL

tap simulation

Vinverter simulation

Vsecondary simulation Vsecondary PHIL

Vinverter PHIL

DG and OLTC interactions: Q(U)

Active power of the DG increases → DG voltage increases → reactive

power absorption by the DG increases (Q(U)) → Voltage of the secondary of

the transformer decreases → tap-change occurs

Recurring tap-changes occur

Instability of the Q(U) controller (i.e. Oscillations): not visible at

the pure digital simulation

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80

0.99

1

1.01

1.02

1.03

time (s)

Vo

lta

ge

p.u

.

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80-3

-2

-1

0

1

2

time (s)

tap

po

sitio

n

tap PHIL

tap simulation

Vinverter simulation Vinverter PHIL

Vsecondary PHIL

Vsecondary simulation

DG and OLTC interactions: Q(U)

Additional reactive power flow

Instability of the Q(U) controller (i.e. Oscillations):

not visible at the pure digital simulation

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80-1.2

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

time (s)

P/P

no

m, Q

/Pn

om

Pinverter: PHIL & simulation

Qinverter simulation

Qinverter PHIL

DG and OLTC interactions: Q(U)

Voltage drop at the HV network

Similarly, the oscillations are not shown at the pure-digital

simulation

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

0.95

0.96

0.97

0.98

0.99

1

1.01

1.02

1.03

1.04

time (s)

Vo

lta

ge

p.u

.

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15-5

-4

-3

-2

-1

0

1

time (s)

tap

po

sitio

n tap PHIL

tap simulation

Vsecondary simulation Vsecondary PHIL

Vinverter simulation Vinverter PHIL

Q(U) controller instability

Closed-loop. Internal delays in the controller (e.g., signal

processing, average filtering of the voltage measurement,

delay in communication between processors) of the

commercial inverter are unknown. Could not be represented

in pure simulation.

PHIL was able to show the true behaviour.

Q(U)

Inverterdynamics

Grid

Measurement

QRef QInvURef U

Delay e-TsURMS

Conclusions

PHIL is a valuable tool that merges component testing

and power system simulation: Power system testing

New interfacing method was proposed

PHIL could reveal interactions of OLTC and DG that

were not visible in pure-digital simulation

Potential problems of OLTC and DG were identified

Thank you for your attention

P. Kotsampopoulos, F. Lehfuss, G. Lauss, B. Bletterie, N. Hatziargyriou,

"The limitations of digital simulation and the advantages of PHIL

testing in studying Distributed Generation provision of ancillary

services", Accepted for publication at the IEEE Transactions on Industrial

Electronics

www.smartrue.gr www.ait.ac.at

kotsa@power.ece.ntua.gr georg.lauss@ait.ac.at