Distribution grid dynamics will become increasingly complex due to the transition from passive to active networks arising from the increase of renewable energy sources at medium and voltage level. A successful transition requires to increase the observability and awareness of the interactions between Transmission and Distribution (T&D) grids, particularly to guarantee adequate operational security. This presentation explores how different technical means can facilitate interactions between TSOs and DSOs with the utilization of GPS-time-synchronized phasor measurements (aka Phasor Measurement Units (PMUs)) with millisecond resolution. If made available in actual T&D networks, such high-sampled data across operational boundaries allows an opportunity to extract information related to different time-scales. As part of the work carried out in the EU-funded FP7 IDE4L project (http://ide4l.eu/), a specific use case, containing PMU-based monitoring functions, has been defined to support the architecture design of future distribution grid automation systems. As a result, the architecture can accommodate for key dynamic information extraction and exchange between DSO and TSO. This presentation presents the use case and focuses on the technical aspects related to the development and implementation of the PMU-based monitoring functionalities that can provide means to facilitate technical co-operation between transmission and distribution operations.

1. ideal grid for all
Synchrophasor
Applications
Facilitating
Interactions
between
Transmission
and
Distribution
Operations
Luigi
Vanfretti
Associate
Professor,
Docent
KTH
Royal
Institute
of
Technology,
Stockholm,
Sweden
https://www.kth.se/profile/luigiv
Thursday
February
9th 2017

2. ideal grid for all
Acknowledgements
• The
work
in
this
presentation
is
result
of
the
research
carried
@KTH
SmarTS_Lab
in
the
FP7
IDE4L
project
(2014-­‐2016)
and
other
projects
form
2011-­‐2016.
• Special
thanks
to
Dr.
Hossein
Hooshyar,
who
undertook
the
day-­‐to-­‐day
project
management
of
this
project
at
KTH
SmarTS
Lab.
with
great
dedication.
• All
of
the
following
students
contributed
to
this
presentation
with
their
hard
work!
29/11/2016 WWW.IDE4L.EUSLIDE 2
Hossein Reza AliFarhan Ravi Narender Maxime Jan

3. ideal grid for all
Outline
• Introduction:
PMUs
in
distribution
networks?
• PMU
Application
Use
Case
Definition
using
SGAM
• Methodology
and
Experimental
Lab
Testing
• Development,
Implementation
and
Testing
using
RT-­‐HIL
Simulation
• The
Reference
Distribution
Grid
Model
• Handling
PMU
Data:
• Getting
the
data:
The
Khorjin Gateway
and
The
S3DK
Toolkit
• Cleaning
the
data:
PMU
Data
Feature
Extraction
• Using
the
Data:
Applications
• Steady
State
Model
Synthesis
• Dynamic
Line
Rating
for
Distribution
Feeders
• Small-­‐Signal
Dynamic
Analysis
• Conclusions
29/11/2016 WWW.IDE4L.EUSLIDE 3

4. ideal grid for all
Introduction
(1/2)
• It
is
becoming
necessary
to
increase
observability
between
T&D
grids,
emerging
dynamics
active
distribution
networks
due
to
renewables
will
lead
to
• Fast
changing
conditions
in
the
network
• Fast
behavior
of
components
• Traditional
monitoring
technology
not
capable
of
satisfying
the
technical
requirements
to
meet
these
conditions:types
of
signals,
time-­‐
synchonization and
speed
of
data
acquisition
• There
is
great
potential
of
utilizing
real-­‐time
Synchrophasor
data
from
PMUs
(Phasor
Measurement
Unit(s))
• From
different
actors,
voltage
levels and
across
conventional
operation
boundaries.
• to
extract key
information related
to
fast changing
conditions
&
dynamic
behavior,
• To
use
such
information
adequately,
a
properly
designed
and
implemented
architecture
needs
to
be
considered.
SLIDE
4 29/11/2016 WWW.IDE4L.EU

5. ideal grid for all
Introduction
(2/2)
• The
IDE4L
architecture
is
built
upon
the
5-­‐layer
Smart
Grid
Architecture
Model
(SGAM)
framework.
• The
main
inputs
to
the
architecture
design
are
the
use
cases.
• The
IDE4L
use
case,
including
the
PMU-­‐based
functions
for
dynamic
information
extraction
and
exchange,
is
presented
in
this
presentation.
SLIDE
5 29/11/2016 WWW.IDE4L.EU

6. ideal grid for all
DistributionTrans.
ProcessFieldStationOperationEnterprise
PMU
(distributed)
Transducer
(PS/SS/distributed)
Electrical
conversion
(PS/SS/distributed)
Synchrophasor
calculation
(distributed)
Synchrophasor
Calculation
(PS/SS)
PMU
(PS/SS)
Measurement
acquisition
(PS/SS)
Communication
interface
(PS/SS)
Data
concentration
(PS/SS)
PDC
(PS/SS)
DMS
at
DSO
Data
curation
and
extraction
of
components
Partial
derivation
of
key
information
Data
transfer
TSO
Computation
unit
(PS/SS)
Measurement
acquisition
Communication
interface
Data
curation
and
extraction
of
components
Derivation
of
key
information
Data
export
Gen. DER Cost.
Prem.
• Data
flows from
PMUs
=>
PDC
=>
Super
PDC
=>
DMS
computers
(where
dynamic
information
is
extracted)
and
exchanged
with
the
TSO
(possible
also
w.
DSOs).
• Data
processing
and
information
extraction
can
occur
at
both
the
station
(partially)
and
the
operation
(i.e.
DMS
computers)
zones.
• Actors:
Instrumentations,
PMU,
PDC,
communication
interface,
DMS,
and
TSO
• Functions: electrical
conversion,
synchrophasor
calculation,
data
acquisition,
data
concentration
and
time-­‐
alignment,
data
exporting,
data
curation,
and
derivation
of
key
information
out
of
the
data
(aka
PMU
applications).
SLIDE
6 29/11/2016 WWW.IDE4L.EU
Use
Case
Definition

7. ideal grid for all
• The
development
of
the
applications
has
been
carried
out
using
real-­‐time
hardware-­‐
in-­‐the-­‐loop
simulation
consisting
of:
• A
real-­‐time
simulation
model
of
active
distribution
networks.
• The
real-­‐time
simulation
model
is
interfaced
with
PMUs
in
HIL.
• PMU
data
is
streamed
into
a
computer
workstation
through
PDC.
SLIDE
7 29/11/2016 WWW.IDE4L.EU
Development,
Implementation
and
Testing
using
RT-­‐HIL
Simulation
• The
computer
makes
available
to
specialized
software
development
tools
within
the
LabVIEW
environment.
• All
data
acquisition
chain
is
carried
out
using
the
corresponding
PMU
standards.

8. ideal grid for all
SLIDE 8 29/11/2016 WWW.IDE4L.EU
The
Reference
Distribution
Grid
Model
(1/2)
101
100
102 103
104
105
106 800 802
806 808
810
812
814 850 816
818
820
822
824 826
828
830 854
856
852
832
858
864
834
842
844
846
848
860 836 840
862
838
888
890/799
701
+
-
+
-
702 713 704
714
718 725
706
720
707
722
724
705
712
742
703
727744
728
729
730
709 731708732
775
733
734
736
710
735 737
738 711 741
740
0.4kV
220 kV
36 kV
36kV
36kV
6.6 kV
6.6 kV
6.6kV
Electric power system
ZIP load
Motor load
Voltage regulator
Wind farm
PV farm
Residential PV system
Battery storage
+
-
Capacitor bank
Circuit breaker
Overcurrent protection
Recloser
Legend
Component Number of units
1 three-phase unit
11 three-phase units, 56 single-phase
units
2 three-phase units, 3 single-phase units
3 sets each consisting of 3 single-phase
units
3 three-phase units
2 three-phase units
3 single-phase units
2 three-phase units
11 single-phase units
1 three-phase units
1 set consisting of 3 phase relays and 1
ground relay
1 three-phase unit, 2 sets each consisting
of 3 single-phase units
Rated Voltage Number of branches
220 kV 7 three-phase
36 kV 28 three-phase, 8 single-phase
6.6 kV 39 three-phase
HV
MV
LV
0.4 kV 1 three-phaseRLV
No. of buses
6
34
37
2
Roy Billinton
Transmission
Test System IEEE 34
bus test
feeder
IEEE 37
bus test
feeder
RLV DG
and Load
Steam
turbine
Governor
ωm
PSS
Excitation
System
V
dω
System
load
Output
feeder

9. ideal grid for all
The
Reference
Distribution
Grid
Model
(2/2)
• Runs
on
4
cores
using
the
State-­‐Space-­‐Nodal
(SSN)
solver
with
time-­‐step
of
100μs.
• The
two
different
implementations
of
the
grid
model
can
be
downloaded
from
the
KTH
SmarTS
Lab
GitHub
repository
at
• The
SSN
implementation
is
also
available
in
the
ARTEMiS 7.0.5
demo
section
of
Opal-­‐RT.
SLIDE
9 29/11/2016 WWW.IDE4L.EU
https://github.com/SmarTS-­‐
Lab/FP7-­‐IDE4L-­‐KTHSmarTSLab-­‐
ADN-­‐RTModel

10. ideal grid for all
Outline
• Introduction:
PMUs
in
distribution
networks?
• PMU
Application
Use
Case
Definition
using
SGAM
• Methodology
and
Experimental
Lab
Testing
• Development,
Implementation
and
Testing
using
RT-­‐HIL
Simulation
• The
Reference
Distribution
Grid
Model
• Handling
PMU
Data:
• Getting
the
data:
The
Khorjin Gateway
and
The
S3DK
Toolkit
• Cleaning
the
data:
PMU
Data
Feature
Extraction
• Using
the
Data:
Applications
• Steady
State
Model
Synthesis
• Dynamic
Line
Rating
for
Distribution
Feeders
• Small-­‐Signal
Dynamic
Analysis
• Conclusions
29/11/2016 WWW.IDE4L.EUSLIDE 10

11. ideal grid for all
IEEE
Std
C37.118.2
Today’s
Architecture
Deployment
Time
Guesstimate:
~15-­‐20
Years
Transition
Likely
Future
Scenario
Ø Security
Not
Addressed
IEEE
Std
C37.118.2
Application
System
PDC
PMU
1
PMU
2
PMU
n
PDC PDC PDC
Communication
Network
PMU
1
PMU
2
PMU
n
PDC PDC PDC
IEC
61850-­‐90-­‐5
Application
System
PDC
Ø Fulfills
the
Gaps
Not
addressed
in
the
IEEE
Std C37.118
e.g.
Security
Enhancement
Ø Harmonization
with
IEC
61850
Ø Security
Not
Addressed
v Two
Segregated
Systems
ßà Two
Protocols
(Even
in
the
same
substation)
v It
will
be
a
huge
CHALLENGE to
adopt
IEC
61850-­‐90-­‐5
Standard
v Need
of
Interfaces
v @PMUs,
@PDCs,
@App
Sys,…
à How
to
maintain
this
?
Getting
the
Data… in
the
Advent
of
IEC
TR
61850-­‐90-­‐5
Standard
Specification
Application
System
PDC
PMU
1
PMU
2
PMU
n
PDC PDC PDC
Communication
Network

12. ideal grid for all One
possible
approach
and
our
contribution:
Khorjin
• A
Gateway:
• Can
act
as
the
IEEE
C37.118.2
to
IEC
61850-­‐90-­‐5
protocol
converter.
• Providing
future
compatibility
for
legacy devices
already
in
the
field
• Capable
of
being
used
at
various
levels:
• @PMU
Level
• @PDC
Level
• @Application
Level
• …
• Gateway
Application
and
Implementation
Test:
Wide-­‐Area
Controller
Application
System
Application
System
PDC
PMU
PDC
Communication
Network
IEEE
C37.118.2
Compatible
IEDs
and
APPs
IEC
61850-­‐90-­‐5
Gateway

13. ideal grid for all
Seyed
Reza
Firouzi srfi@kth.se
Receiver
Gateway
National Instrument
CompactRIO à PMU
Wireshark captures
IEC 61850-­90-­5
R-­SV / R-­GOOSE
Khorjin
Gateway Testing
• Test
of
the
gateway
implementation
was
tested
using
real-­‐time
HIL
simulation
• Data
is
transformed
from
IEEE
C37.118.2
from
actual
PMUs,
and
published
using
the
IEC
61850-­‐90-­‐5
protocol

14. ideal grid for all Khorjin
Gateway Architecture
• The
Gateway
functionality
of
“Khorjin”
library
is
getting
use
of
its
modular
architecture:
à Easier
for
future
development
• The
Khorjin Gateway
is
designed
and
implemented
in
three
main
components
of:
1)
IEEE
C37.118.2
Module,
2)
IEC
61850
Mapping
Module,
and
3)
IEC
61850-­‐90-­‐5
R-­‐SV
/
R-­‐GOOSE
Publisher
Module.
14
• In
order
to
be
platform-­‐independent:
à A
Platform
Abstraction
Layer
is
Implemented.
à Depending
on
the
platform,
the
relevant
platform-­‐dependent
functions
are
utilized
(i.e.
Socket,
Thread,
Time
and
…).
• The
Khorjin Gateway
is
designed
and
implemented
in
three
main
components
of:
1)
IEEE
C37.118.2
Module,
2)
IEC
61850
Mapping
Module,
and
3)
IEC
61850-­‐90-­‐5
R-­‐SV
/
R-­‐GOOSE
Publisher
Module.

15. ideal grid for all
• The
Khorjin library
is
an
Open
Source
code
providing
following
functionalities:
à IEEE
C37.118.2
Traffic
Parser
à IEEE
C37.118.2
to
IEC
61850-­‐90-­‐5
Protocol
Converter
à IEC
61850-­‐90-­‐5
Traffic
Generation
à Routed-­‐Sampled
Value
à Routed-­‐GOOSE
à IEC
61850-­‐90-­‐5
Traffic
Parser
à Routed-­‐Sampled
Value
à Routed-­‐GOOSE
• The
Khorjin library
supports
different
platforms.
à Will
be
publicly
available
sometime
in
2017,
keep
an
eye
on
our
Github repo:
https://github.com/SmarTS-­‐Lab-­‐Parapluie
Khorjin
Windows
Linux
Mac
NI
cRIO
Raspberry
Pi
15 Khorjin
Functionalities – it’s more than a
gateway!

16. ideal grid for all
• PMU
SW
Apps
require
real-­‐time
data
acquisition.
• PMU
data
is
sent
to
these
SW
Apps
using
many
different
comm.
protocols.
• For
fast
software
prototyping
and
testing,
communication
protocol
parsing
is
required.
• Low
level
data
management
routines
(windowing,
etc.)
do
not
need
to
be
reinvented
N times.
• A
tool-­‐set
is
needed
to
assist
students
and
researchers
with
a
background
in
power
systems,
but
lackingproficient
software
development
and
programming
skills.
PMU
1
PMU
2
PMU
n
PDCCommunication
Network
Infrastructure
Data
in
IEEE
C37.118
Protocol
Real-­‐time
data
locked
into
vendor
specific
software
system
Historical
Data
in
Proprietary
Database
Interfaces
using
standard
protocols
and
a
flexible
development
environment
are
needed
Last
Mile
in
PMU
App
Development
16
The
next
step…
Now
you
got
the
data…
how
do
you
implement
an
application?

17. ideal grid for all S3DK
Smart
grid
Synchrophasor
Software
Development
tool-­‐Kit
(SDK)
• Infrastructure
(external).
• Software
Development
Kit
(SDK):
a
set
of
different
computer
software
that
allows
a
user
to
develop
other
derived
software
applications.
• Our
SDK
is
composed
by
two
main
pieces:
• Real-­‐Time
Data
Mediator
(DLL)
• PMU
Recorder
Light
(PRL)
Computer/Server
S3DK
Real-­‐Time
Data
Mediator
(RTDM)
“DLL”
LabView
PMU
Recorder
Light
(PRL)
PMU
1
PMU
2
PMU
n
PDCCommunication
Network
Infrastructure
Data
in
IEEE
C37.118
Protocol

18. ideal grid for all S3DK’s
Real-­‐Time
Data
Mediator
a.k.a.
the
“DLL”
• A
library
in
C++
was
implemented
with
the
following
architecture
design
in
mind
Client
Applications
(PRL)
Synchronization
Layer
IEEE
C37.118
Server
IEC
61850
Server
Other
protocols
Only
prepared,
not
fully
implemented.
Future
Support
/
Integration
IEEE
C37.118
Client
IEC
61850
Client
Other
protocols
Console
Test
Tool
PMU
or
PDC
1 PMU
or
PDC
n
Interface
Implemented
and
Tested
Other
Protocol
Servers
Clients
Servers
IEEE
C37.118
DLL

19. ideal grid for all 19
GUI
Many blocks
to
access
and
handle
RT
data!
Graphical
User
Interfaces
Communication
Configuration
Development Tooling
S3DK
– LabView API,
UI
and
Tools

20. ideal grid for all
S3DK’s
LabVIEW
Function Library/Toolbox/Pallete (aka PRL)

21. ideal grid for allRepositories Currently Available at
GitHub!
• Main
repo:
https://github.com/SmarTS-­‐Lab-­‐Parapluie/
• S3DK:
https://github.com/SmarTS-­‐Lab-­‐Parapluie/S3DK
• BabelFish:
https://github.com/SmarTS-­‐Lab-­‐Parapluie/BabelFish
• Khorjin:
Will
be
available at
GitHub sometime in
2017.
L.
Vanfretti,
V.
H.
Aarstrand,
M.
S.
Almas,
V.
S.
Perić and
J.
O.
Gjerde,
"A
software
development
toolkit
for
real-­‐time
synchrophasor
applications," PowerTech (POWERTECH),
2013
IEEE
Grenoble,
Grenoble,
2013,
pp.
1-­‐6.
doi:
10.1109/PTC.2013.6652191
L.
Vanfretti,
I.
A.
Khatib and
M.
S.
Almas,
"Real-­‐time
data
mediation
for
synchrophasor application
development
compliant
with
IEEE
C37.118.2," Innovative
Smart
Grid
Technologies
Conference
(ISGT),
2015
IEEE
Power
&
Energy
Society,
Washington,
DC,
2015,
pp.
1-­‐5.
doi:
10.1109/ISGT.2015.7131910
L.
Vanfretti,
M.S.
Almas
and
M.
Baudette,
“BabelFish– Tools
for
IEEE
C37.118.2-­‐compliant
Real-­‐Time
Synchrophasor Data
Mediation,”
SoftwareX,
submitted,
June
2016.
S.R.
Firouzi,
L.
Vanfretti,
A.
Ruiz-­‐Alvarez,
F.
Mahmood,
H.
Hooshyar,
I.
Cairo,
“An
IEC
61850-­‐90-­‐5
Gateway
for
IEEE
C37.118.2
Synchrophasor Data
Transfer,”
IEEE
PES
General
Meeting
2016,
Boston,
MA,
USA.
Pre-­‐print:
link.
S.R.
Firouzi,
L.
Vanfretti,
A.
Ruiz-­‐Alvarez,
H.
Hooshyar and
F.
Mahmood,
“Interpreting
and
Implementing
IEC
61850-­‐90-­‐5
Routed-­‐Sampled
Value
and
Routed-­‐GOOSE
Protocols
for
IEEE
C37.118.2
Compliant
Wide-­‐Area
Synchrophasor
Data
Transfer,”
Electric
Power
Systems
Research,
2017.
G.M.
Jonsdottir,
E.
Rebello,
S.R.
Firouzi,
M.S.
Almas,
M.
Baudette
and
L.
Vanfretti,
“Audur – Templates
for
Custom
Synchrophasor-­‐Based
Wide-­‐Area
Control
System
Implementations,”
SoftwareX,
in
preparation,
2016.

22. ideal grid for all SLIDE
22 29/11/2016 WWW.IDE4L.EU
Cleaning
the
Data!
PMU
Data
Curation,
Extraction
of
Different
Time-­‐Scale
Components

23. ideal grid for all
PMU
Data
Processing
(1/3)
• PMU
measurements
are
normally
polluted
with
• Undesirable
noise
• Outliers
• Missing
data
• Measurements
obtained
from
PMUs
during
different
events
in
power
systems
contain
different
signal
features
at
different
time
scales,
i.e.
features
of
different
types
of
power
system
dynamics.
• So
before
being
fed
to
the
applications:
• PMU
measurements
should
be
curated.
• Signals
required
for
each
application
should
be
extracted
from
the
PMU
measurements.
SLIDE
23 29/11/2016 WWW.IDE4L.EU

24. ideal grid for all
PMU
Data
Processing
(2/3)
• An
enhanced
Kalman filtering
technique
has
been
utilized
for
both
bad
data
removal,
i.e.
data
curation,
and
to
extract
different
time-­‐scale
dynamics
from
the
PMU
data
(ie used
as
a
real-­‐time
filter)
• R and
Q (the
measurement
and
process
noise
covariance
matrices),
can
be
updated
in
real-­‐time
to
perform
the
data
processing.
SLIDE
24 29/11/2016 WWW.IDE4L.EU
Step Equation Significance
1 Prediction 𝑥" 𝑘
−
= 𝐴𝑥" 𝑘−1
−
+ 𝐵𝑢 𝑘 Project the state ahead.
2 𝑃𝑘
−
= 𝐴𝑃𝑘−1 𝐴 𝑇
+ 𝑄 Project the error covariance ahead.
3 Correction 𝐾𝑘 = 𝑃𝑘
−
𝐻 𝑇( 𝐻𝑃𝑘
−
𝐻 𝑇
+ 𝑅)−1
Compute the Kalman gain.
4 𝑥" 𝑘 = 𝑥" 𝑘
−
+ 𝐾𝑘(𝑧 𝑘 − 𝐻𝑥" 𝑘
−
) Update estimate with measurement zk.
5 𝑃𝑘 = ( 𝐼 − 𝐾𝑘 𝐻) 𝑃𝑘
−
Update the error covariance.

25. ideal grid for all
PMU
Data
Processing
(3/3)
• Sample
results.
SLIDE
25 29/11/2016 WWW.IDE4L.EU
0 20 40 60 80 100 120 140 160
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
Time (sec)
Vmag(p.u.)
Raw data
Curated data
Steady state component
Dynamics
0 20 40 60 80 100 120 140 160
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
Time (sec)
Vang(rad.)
Raw data
Curated data
Steady state component
Dynamics
voltage
magnitude
(per
unit) voltage
angle
(radian)

26. ideal grid for all
Outline
• Introduction:
PMUs
in
distribution
networks?
• PMU
Application
Use
Case
Definition
using
SGAM
• Methodology
and
Experimental
Lab
Testing
• Development,
Implementation
and
Testing
using
RT-­‐HIL
Simulation
• The
Reference
Distribution
Grid
Model
• Handling
PMU
Data:
• Getting
the
data:
The
Khorjin Gateway
and
The
S3DK
Toolkit
• Cleaning
the
data:
PMU
Data
Feature
Extraction
• Using
the
Data:
Applications
• Steady
State
Model
Synthesis
• Dynamic
Line
Rating
for
Distribution
Feeders
• Small-­‐Signal
Dynamic
Analysis
• Conclusions
29/11/2016 WWW.IDE4L.EUSLIDE 26

27. ideal grid for all
Real-­‐Time
Steady
State
Model
Synthesis
of
Active
Distribution
Networks
SLIDE
27 29/11/2016 WWW.IDE4L.EU

28. ideal grid for all
Background
• Currently,
TSOs
maintain
reduced
models
of
portions
of
the
distribution
networks,
however:
• The
models
covers
limited
portions
of
the
distribution
network
due
to
the
lack
of
network
“observability”
(measurement
points)
and
computational
burden
associated
with
simulating
large
joint
T&D
models.
• The
models
are
not
updated
frequently.
• The
reduction
methods,
used
by
TSOs,
often
make
assumptions
that
are
no
longer
valid
for
active
distribution
networks.
SLIDE
28 29/11/2016 WWW.IDE4L.EU

29. ideal grid for all SLIDE
29 29/11/2016 WWW.IDE4L.EU
Methodology
and
Application
V1
abc
<δ1
abc
PMU1
I1
abc
<φ1
abc
Any feeder
configuration with an
arbitrary combination
of load and DG
PMU2
V1
a
<δ1
a
I1
a
<φ1
a
I2
a
<φ2
a
V2
a
<δ2
a
R a
X a
R a
X a
Ra
Xa
E a
<δ a
V0
a
<δ0
a
I0
a
<φ0
a
The reduced steady state
equivalent model for
phase ‘a’
3
3
3 ... V3
abc
<δ3
abc
I3
abc
<φ3
abc
3
PMU3
..
.
PMU N
VT
abc
<δT
abc
PMU at
Transmission
Level
IT
abc
<φT
abc
Any feeder
configuration with an
arbitrary combination
of load and DG
3
3
3
V1
abc
<δ1
abc
I1
abc
<φ1
abc
3
PMU at
Distribution Level
VT
a
<δT
a
IT
a
<φT
a
I1
a
<φ1
a
V1
a
<δ1
aR a
X a
R a
X a
V0
a
<δ0
a
E a
<δ a
Ra
XaI0
a
<φ0
a
The reduced steady
state equivalent
model for phase ’a’
2R a
2X a
The reduced steady state equivalent
model for phase ’a’
E a
<δ a
I2=0
Open-circuited
VT
a
<δT
a
IT
a
<φT
a
With
N available
PMUs With
1 available
PMU
• Model
parameters
are
obtained
by
writing
KVL
equations
across
the
model
branches
and
equate
Vi’s
and
Ii’s to
PMU
measurements.
Event1
Event2
0 80 1600 80 160
Event1
Event2
Event1
Event2
Event1
Event2
R(p.u.)X(p.u.)
Ea
(p.u.)δa
(rad)
Estimated
parameters
of
equivalent
model
LabVIEW
Application

30. ideal grid for all SLIDE
30 29/11/2016 WWW.IDE4L.EU
Illustration
Example
(1/2)
• Steady
state
model
of
a
portion
of
the
reference
grid
is
estimated
during
wind
curtailment
at
different
dispatch
levels.
R
X
E
δ
E<δ
R X

31. ideal grid for all SLIDE
31 29/11/2016 WWW.IDE4L.EU
Illustration
Example
(2/2)
• True
and
reproduced
current
and
voltage
phasors
are
compared
using
TVE.
• The
mean
TVE
is
less
than
2.5%.
1.5 1.55 1.6 1.65 1.7 1.75 1.8
0
5
10
15
20
25
30
35
40
TVE for V2 (%)
Probabilitydistribution
TVE for Voltage Phasor at PMU2
empirical
generalized pareto
generalized extreme value
inversegaussian
birnbaumsaunders
1 1.5 2 2.5 3 3.5 4 4.5
0
0.5
1
1.5
2
2.5
3
3.5
4
TVE for I1 (%)
Probabilitydistribution
TVE for Current Phasor at PMU1
empirical
generalized pareto
generalized extreme value
inversegaussian
birnbaumsaunders
1.3 1.35 1.4 1.45 1.5 1.55 1.6 1.65 1.7
0
5
10
15
20
25
30
35
40
TVE for I2 (%)
Probabilitydistribution
TVE for Current Phasor at PMU2
empirical
generalized extreme value
generalized pareto
inversegaussian
birnbaumsaunders
0.08 0.09 0.1 0.11 0.12 0.13 0.14 0.15 0.16 0.17 0.18
0
10
20
30
40
50
60
70
80
90
TVE for V1 (%)
Probabilitydistribution
TVE for Voltage Phasor at PMU1
empirical
generalized pareto
generalized extreme value
beta
gamma
Variables
with
hat
are
reproduced
ones.

32. ideal grid for all
Dynamic
Line
Rating
for
Aerial
Distribution
Feeders
SLIDE
32 29/11/2016 WWW.IDE4L.EU

33. ideal grid for all SLIDE
33 29/11/2016 WWW.IDE4L.EU
Background
• Dynamic
line
rating
(DLR)
is
a
way
to
optimize
the
ampacity
of
transmission
and
distribution
lines
by
measuring
the
effects
of
weather
and
actual
line
current.
• The
inputs
needed
for
the
method:
• Weather
data
=> provided
by
a
close-­‐by
weather
station.
• Line
loading
=> provided
by
PMU!
• Real-­‐time
sag
=> provided
by
a
GPS-­‐based
measurement
device.

34. ideal grid for all SLIDE
34 29/11/2016 WWW.IDE4L.EU
Methodology
and
Application
LabVIEW
Application

35. ideal grid for all SLIDE
35 29/11/2016 WWW.IDE4L.EU
• Results
are
obtained
by
applying
the
method
on
data
from
a
real
feeder.
• Output
shows
accurate
correlation
with
different
inputs.
Sample
Results

36. ideal grid for all
Small
Signal
Dynamic
Analysis
of
Active
Distribution
Networks
SLIDE
36 29/11/2016 WWW.IDE4L.EU

37. ideal grid for all SLIDE
37 29/11/2016 WWW.IDE4L.EU
Background
• For
reliable
operation
of
the
power
system,
oscillations
are
required
to
decay.
• Accurate
and
real-­‐time
estimates
of
active
distribution
networks
oscillatory
modes
has
become
ever
so
important.
• Timely
extraction
of
these
modes
and
related
parameters
from
network
measurements
has
considerable
potential
for
near
real-­‐time
dynamic
security
assessment.

38. ideal grid for all
Extracting
and
Exchanging
Small
Signal
Stability
Information
38
Data
curation,
fusion
and
extraction
of
steady
state
component
• Ambient
Data
Analysis
(for
stochastic
variations)
• Ringdown
Data
Analysis
(for
transients)
TSO
Calculation
of
stability
indices
Data
process
and
analysis
Data
process
and
analysis
Data
process
and
analysis
Centralized
Architecture
Decentralized
Architecture

39. ideal grid for all SLIDE
39 29/11/2016 WWW.IDE4L.EU
Implementation
Architectures
Centralized
Decentralized

40. ideal grid for all SLIDE
40 29/11/2016 WWW.IDE4L.EU
PMU
data
collection
and
transmission
Dynamic
component
extraction
by
Kalman
filter
Data
pre-­‐processing
(detrending
and
downsampling)
Ringdown
detected?
Ambient
analysis
(ARMA)
Ringdown
analysis
(ERA)
Modal
parameters
Modal
parameters
LabVIEW
Application
Methodology
and
Application

41. ideal grid for all SLIDE
41 29/11/2016 WWW.IDE4L.EU
Illustration
Example
(1/2)
• 4
PMUs
were
deployed
in
the
reference
grid
to
be
used
for
mode
estimation
under
two
different
architectures
of
centralized
and
decentralized

42. ideal grid for all SLIDE
42 29/11/2016 WWW.IDE4L.EU
Illustration
Example
(2/2)
Centralized
Decentralized
Inter-­‐area
oscillatory
mode
between
HV
section
and
rest
of
the
grid
Forced
local
oscillatory
mode
detectable
in
decentralized
architecture

43. ideal grid for all
References
and
Resources
of
work
presented
and
related
to
this
presentation
• Github Repositories
(Open
Source
Software
Resources)
• https://github.com/SmarTS-­‐Lab/
• https://github.com/SmarTS-­‐Lab-­‐Parapluie
(PMU
applications
and
software
tools)
• Under
Review:
• N.
Singh,
H.
Hooshyar and
L.
Vanfretti,
“Feeder
Dynamic
Rating
Application
for
Active
Distribution
Networks
using
Synchrophasors,”
Sustainable
Energy,
Grids
and
Networks.
(under
third
review)
• F.
Mahmood,
H.
Hooshyar,
and
L.
Vanfretti,
“Sensitivity
Analysis
of
a
PMU-­‐Based
Steady
State
Model
Synthesis
Method,”
IEEE
PES
GM
2017.
• R.S.
Singh,
H.
Hooshyar,
and
L.
Vanfretti,
“Real-­‐Time
Testing
of
a
Decentralized
PMU
Data-­‐Based
Power
System
Mode
Estimator,”
IEEE
PES
GM
2017.
29/11/2016 WWW.IDE4L.EUSLIDE 43

44. ideal grid for all
References
and
Resources
of
work
presented
and
related
to
this
presentation
• Published/Accepted
for
Publication:
• S.R.
Firouzi,
L.
Vanfretti,
A.
Ruiz-­‐Alvarez,
H.
Hooshyar and
F.
Mahmood,
“Interpretation
and
Implementation
of
IEC
61850-­‐90-­‐5
Routed-­‐
Sampled
Value
and
Routed-­‐GOOSE
Protocols
for
IEEE
C37.118.2
Compliant
Wide-­‐Area
Synchrophasor
Data
Transfer,”
Electric
Power
Systems
Research,
2017.
• F.
Mahmood,
H.
Hooshyar,
J.
Lavenius,
P.
Lund
and
L.
Vanfretti,
“Real-­‐Time
Reduced
Steady
State
Model
Synthesis
of
Active
Distribution
Networks
using
PMU
Measurements,”
IEEE
Transactions
on
Power
Delivery,
2017.
http://dx.doi.org/10.1109/TPWRD.2016.2602302
• F.
Mahmood,
H.
Hooshyar and
L.
Vanfretti,
"Extracting
Steady
State
Components
from
Synchrophasor
Data
Using
Kalman Filters,"
Energies,
vol.
9,
2016.
• H.
Hooshyar,
F.
Mahmood,
L.
Vanfretti,
and
M.
Baudette
“Specification,
Implementation
and
Hardware-­‐in-­‐the-­‐Loop
Real-­‐Time
Simulation
of
an
Active
Distribution
Grid,”
Sustainable
Energy,
Grids
and
Networks
(SEGAN),
Elsevier,
vol.
3,
Sept.
2015,
pp.
36-­‐51.
Available
in
open-­‐
access:
https://dx.doi.org/doi:10.1016/j.segan.2015.06.002
• H.
Hooshyar,
L.
Vanfretti
and
C.
Dufour,
“Delay-­‐Free
Parallelization
for
Real-­‐Time
Simulation
of
a
Large
Active
Distribution
Grid
Model,”
IEEE
IECON
2016,
Firenze,
October
2016.
• R.S.
Singh,
M.
Baudette,
H.
Hooshyar,
M.S.
Almas,
Stig Løvlund,
L.
Vanfretti,
“‘In
Silico’
Testing
of
a
Decentralized
PMU
Data-­‐Based
Power
Systems
Mode
Estimator,”
IEEE
PES
General
Meeting
2016,
Boston,
MA,
USA.
• A.
Bidadfar,
H.
Hooshyar,
M.
Monadi,
L.
Vanfretti,
Decoupled
Voltage
Stability
Assessment
of
Distribution
Networks
using
Synchrophasors,”
IEEE
PES
General
Meeting
2016,
Boston,
MA,
USA.
• S.R.
Firouzi,
L.
Vanfretti,
A.
Ruiz-­‐Alvarez,
F.
Mahmood,
H.
Hooshyar,
I.
Cairo,
“An
IEC
61850-­‐90-­‐5
Gateway
for
IEEE
C37.118.2
Synchrophasor
Data
Transfer,”
IEEE
PES
General
Meeting
2016,
Boston,
MA,
USA.
• R.S.
Singh,
H.
Hooshyar and
L.
Vanfretti,
“Laboratory
Test
Set-­‐Up
for
the
Assessment
of
PMU
Time
Synchronization
Requirements,”
IEEE
PowerTech 2015,
The
Netherlands,
2015.
• R.S.
Singh,
H.
Hooshyar and
L.
Vanfretti,
“Assessment
of
Time
Synchronization
Requirements
for
Phasor
Measurement
Units,”
IEEE
PES
PowerTech 2015,
The
Netherlands,
2015.
• H.
Hooshyar,
F.
Mahmood,
and
L.
Vanfretti,
“HIL
Simulation
of
a
Distribution
System
Reference
Model,”
NORDAC
2014.
• H.
Hooshyar and
L.
Vanfretti,
“Specification
and
Implementation
of
a
Reference
Grid
for
Distribution
Network
Dynamics
Studies,”
IEEE
PES
General
Meeting,
2014,
National
Harbor,
MD
(Washington,
DC
Metro
Area).
• L.
Vanfretti,
V.
H.
Aarstrand,
M.
Shoaib Almas,
V.
Peric,
and
J.
O.
Gjerde,
”A
software
development
toolkit
for
real-­‐time
synchrophasor
applications”,
2013
IEEE
Grenoble
Conference
PowerTech,
POWERTECH
2013;
Grenoble,
France,
16-­‐20
June
2013.
29/11/2016 WWW.IDE4L.EUSLIDE 44

45. ideal grid for all
Conclusions
• The
increase
of
intermittent
renewable
sources
at
the
distribution
network
bring
technical
challenges into
DSO
operations.
• Synchrophasor
measurements
have
a
great
potential
to
support
the
technical
operation
of
distribution
networks
• Applications
can
play
an
important
role
of
extracting
key
information
about
the
operation
of
the
grid
• TSOs-­‐to-­‐DSOs
interaction
in
operations
through
PMU
Apps.
• DSOs
can
enhance
the
way
they
operate
by
having
better
knowledge
of
the
system’s
performance
in
near
real-­‐time
• TSOs
can
gain
visibility
of
the
phenomena
at
lower
voltage
levels, and
device
actions
• Real-­‐time
automatic
control
and
protection
is
the
next
big
step
• Existing
architecture,
automation
and
system
level
technology
not
up
to
the
task
• Interoperability
and
Standardization:
• Too
slow
and
maybe
even
useless
without
Open
Source
Software
and
Hardware
platforms
and
building
blocks
• Need
to
develop
and
support
a
truly
open
market
of
products
and
services
29/11/2016 WWW.IDE4L.EUSLIDE 45

46. ideal grid for all
Thank
you!
www.ide4l.eu
https://www.kth.se/profile/luigiv
luigiv@kth.se