Campus da FEUP
Rua Dr. Roberto Frias, 378
4200 - 465 Porto
Portugal
T +351 222 094 000
F +351 222 094 050
www.inescporto.pt
© 2009
Multi-Microgrids
coordinated voltage and frequency control,emergency functions and ancillary services
J. A. Peças Lopes
(Director of INESC Porto)
2010 Jan 29
© 2009 2
PV
Wind Gen
MV
LV
MicroGrid: A Flexible Cell of the Electric Power System
Microturbine
Fuel Cell
Storage
DeviceMGCC
MC
MC
MC
MC
MC
LC
LC
LC
LC
LC
MG Hierarchical Control:
• MGCC, LC, MC
• Communication
infrastructure
© 2009
Multi-microgrids – MV distribution network of the future
• Microgrids
• Storage device-VSI
• Large DFIM
• Mini-Hydro
• CHP
• Small Diesel
• Sheddable Loads
HV Network
VSI
Diesel
Wind
generator
MicroGrid
MicroGrid
MicroGrid
Capacitor
Bank
Miini-
Hydro
CHP
Sheddable
Loads
Storage
3
© 2009 4
The concept: Multi Microgrids
• Integration of several Microgrids in MV networks
• Active Management of microgrids, DG units and loads for
– Normal Steady-state Operation or
– Emergency Conditions
DMS
CAMC
RTU
MGCC
~
~
MGCC
MGCC
CAMCRTU
HV
MV
HV
MV
LV
LV
LV
DG
DG
© 2009 5
Multi-MicroGrids: Interaction of MGCC and DMS
DMS
CAMC
MCOLTCSVC Load LC
MGCC
DG
Control Level 1
Control Level 2
Control Level 3
Control Scheme of a Multi-MicroGrid System
DMS
CAMC
RTU
MGCC
~
~
MGCC
MGCC
CAMCRTU
HV
MV
HV
MV
LV
LV
LV
DG
DG
© 2009 6
Multi-MicroGrids: Definition of Ancillary Services offered to the MV
Network
• Energy / Services… to be provided in each operating mode
(Normal and Emergency)
• Different services can be provided by these agents regarding
MicroGrids and Distribution / Transmission Network Operators
• A list of new DMS functionalities
• List of needed data was identified
© 2009 7
Management of the Multi-MicroGrid
• Existing DMS functionalities need to be adapted to Multi-MicroGrid operation
• The management of the Multi-MicroGrid will be performed through the CAMC
using an Hierarchical Control Architecture, which requires new functionalities
like:
– Local State Estimation
– Coordinated Voltage Support and Flow Control
– Peak shaving
– Coordinated Frequency Support
CAMCLocal
Measurements
Remote Terminal
Units
State EstimationEmergency
Functions
Control Scheduling
(Markets)
Voltage VAR
Support
Coordinated
Frequency Support
Network Data,
Pseudo
Measurements
Constraints
Contracts and Bids
MGCCs
OLTCs
Capacitors
SVCs
DG Units
Loads
© 2009
Control and Management Architecture
• Communication Scheme of a Multi-MicroGrid System
DMS
CAMC
MC
OLTCSVC Load
RTU
LC
MGCC
DG
8
© 2009 9
Local State Estimation
• Traditional State Estimation routines for DMS can be adapted to the new paradigm of Multi-MicroGrid systems
– New algorithms for Distributed State Estimation in order to minimize the amount of data required
– Fuzzy State Estimation approaches
© 2009 10
Local State Estimation
Estimation of bus voltages (both
module and phase), power injections
for each bus and power flows and
current in each branch
FUZZY STATE ESTIMATION
Type of Fuzzy Information considered on the MicroGrids
Membership Functions for the Measurements and for the
Results of the Voltage Magnitude in the Buses 2 and 20
Membership Functions for the Measurements and for the
Results of the Active and Reactive Power injected on Bus 42
© 2009 11
Local State Estimation
DISTRIBUTED STATE ESTIMATION
Each area is governed by a
Local State Estimator that is
responsible for estimating its
own state
The areas exchange
information through
communication links to a
Coordination State Estimation
© 2009 12
Coordinated Voltage Support and Control
• Two types of tools for the Coordination of Volt/VAR Control:
– Tools based on a local control approaches (using conventional
techniques)
– Tools based on a global coordinated approach (using meta-heuristics –
EPSO)
• The developed approaches make full use of the control
capabilities provided by MicroGrids, DG Units, OLTC
Transformers and Capacitor Banks
© 2009
Coordinated Voltage Support and Control
• High DG penetration may cause voltage rise problems, especially in the case
of weak distribution networks
• The effects of voltage rise may propagate to the LV side
• From the power flow equations (in LV networks where R<<X does not apply):
• In conclusion, high DG and microgeneration penetration will require the
development of an effective voltage control scheme based on active and
reactive power control
cos12
2
2
R
VV
R
VPinj
R
V1
V2
MV Network
MicroGrid
Pinj
13
© 2009
Coordinated Voltage Support and Control
Voltage/Control in Distribution Systems integrating
MicroGrids becomes a hierarchical optimization problem
that must be analyzed in a coordinated way between LV
and MV levels
Given the characteristics of the networks, both Active and
Reactive Power Control is needed
14
© 2009
Coordinated Voltage Support and Control ControlGeneral
• Forms of Control
– Generation curtailment
– Reactive power
(Droop control)
– OLTC
V1 V2
R+jX
P+jQ Local
Generation
Load
Transformer
Network
LineShunt
Compensator
Q generation
Vmax
Vmin
V0
Q absorbtion
DVmax
iM SQ
iM S maxQ
iM S max
Q
iM SV
15
© 2009 16
Coordinated Voltage Support and Flow Control
0,9
0,95
1
1,05
1,1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Vo
ltag
e (
pu
)
Hours (h)
Initial
Final
0
0,02
0,04
0,06
0,08
0,1
0,12
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Act
ive
Po
we
r (M
W)
Hours (h)
Initial
Final
In order to maintain voltage profiles within admissible
limits, Microgeneration Shedding was required
A Neural Network was used to emulate
the behavior of the LV MicroGridGlobal Approach
Optimizing distribution
network operation in
interconnected mode, when
dealing simultaneously with
DG connected directly to
the MV grid and
microgeneration installed
at the LV side
Controls:
• OLTC Transformer taps
• Reactive power provided
by DG Sources and
Capacitor Banks
• Active power control at the
MicroGrid level in extreme
scenarios (using
Microgeneration Shedding)
© 2009 17
Flow Control and Peak Shaving – Optimization at the MV Level
iLcP
iLncP
iLP
iGc
Pi
GncP
iG
P
Optimization problem within the MMG:
Controlled from the CAMC
© 2009 18
Coordinated Frequency Support
• Effective frequency support can be obtained exploiting decentralized
control strategies together with a Hierarchical Approach housed at
CAMC and MGCC
• Coordination with Load Curtailment is required
• Dynamic Equivalents of MicroGrids may be needed
• The feasibility and profitability of Ancillary Services provision (namely
regarding frequency control) from Multi-MicroGrid should be
addressed at different levels:
– By contributing to general system frequency control (primary and
secondary):
– By allowing islanding operation and black start
© 2009 19
Development of Equivalents of MicroGrids
Boundary
bus PV
PV
Microturbine
Fuel cell
Wind
generator
Microturbine
Storage device
Boundary
bus
MG slow dynamics
reduced model
VSI
IdR; IqR
Vd; Vq
Study
subsystem
External
subsystem
MicroGrid
Dynamic equivalent
© 2009
Coordinated Frequency Support (Microgrids)
• MicroSources providing no
frequency response
• MicroSources providing primary
response using special VSI control
solutions
• MicroSources providing primary
and secondary responses,
exploiting P,Q controlled inverters
together with central control
f0
fMG
Primary
response fp
2MSP
2ref
MS
P
2p
MS
P
2m
ax
MS
P
Primary
response
Secondary
response f0
fMG
fp
3MSP3ref
MS
P
3p
MS
P
3m
ax
MS
P
20
© 2009
Coordinated Frequency Support (Microgrids)
• Flywheel
– Inertia
– Short circuit current
– Initial frequency support
• Frequency target of the flywheel
0
f0
fMG
Primary
response
Secondary
response
FWP
pF
WP
max
FW
P
21
© 2009 22
Coordinated Frequency Support – Multimicrogrid
CHP
MicroGrid
Hydro
DFIMVSI
Diesel
TEST SYSEM
© 2009 23
Coordinated Frequency Support
0 10 20 30 40 50 60 70 80 9048.5
49
49.5
50
50.5
Time (s)
Fre
qu
ency
(H
z)
With Hierarchical Control
Without Hierarchical Control
0 10 20 30 40 50 60 70 80 90
0
0.2
0.4
0.6
0.8
1
Time (s)
Ou
tpu
t P
ow
er (
pu
)
MGCC 3
Hydro
CHP
Frequency Deviation following Islanding of
the Multi-MicroGrid System
Active Power Set-Points sent by the CAMC to
the DG Units and MicroGrids
Local secondary frequency control should be designed such that load shedding is also managed
© 2009 24
Emergency Functions – Islanding with several MicroGrids
0 10 20 30 40 50 60 70 80 9047
47.5
48
48.5
49
49.5
50
50.5
Time (s)
Fre
qu
ency
(H
z)
With Load-Shedding
Without Load-Shedding
8 9 10 11 12 13 14 150
1
2
3
4
5
6
7
8
Time (s)
Po
wer
(M
W)
In some extreme situations, Load
Shedding may be needed to prevent
long frequency excursions and even
System Collapse
Load Shedding Steps
ISLANDINGSystem Collapse
© 2009
Centralized Load-Shedding (acting as secondary reserve)
• Controllable loads help to reach rated frequency faster and
economically (integrated in an optimal allocation algorithm)
• But are not able to avoid large initial frequency variations.
0 10 20 30 40 50 60 70 80 9048.5
49
49.5
50
50.5
Time (s)
Fre
qu
ency
(H
z)
With Load Shedding
Without Load Shedding
0 10 20 30 40 50 60 70 80 900
0.2
0.4
0.6
0.8
1
Time (s)
Lo
ad (
MW
)
Loads NMVCHP
Loads NMVCHP2
25
© 2009
Using MicroGrids for Service Restoration
• Black-Start is a sequence of events controlled by a set of rules
– A set of rules and conditions are identified in advance and
embedded in MGCC software
– These rules and conditions define a sequence of control actions
to be carried out during the restoration stages
– The electrical problems to be dealt with include:
• Building LV network
• Connecting microgenerators
• Connecting controllable loads
• Controlling frequency and voltage
• Synchronization with the MV network (when available)
26
© 2009
MicroGrid Black Start – Test System
20 kV
0.4 kV
Appartment
building
Appartment
building
15 kW wind
generator
30 kW SSMT
30 kW SSMT
10 kW PV
Group of 4
Residences
10 kW PV
MG Main
Storage
Industrial
Load
2x30 kW SSMT
The fast transients associated to the
initial stages of the MG restoration
process were analysed using an
EMTP-RV tool, being the long term
dynamics evaluated using the MatLab
Simulink simulation platform
27
© 2009
Results from Simulations – Long Term Dynamics
• Development of the Service Restoration Procedure
90 100 110 120 130 140 150 160 170 180 190 200 210 22049.6
49.8
50
50.2
50.4
Fre
qu
en
cy (
Hz)
90 100 110 120 130 140 150 160 170 180 190 200 210 220-20
0
20
40
Ac
tiv
e P
ow
er
(kW
)
90 100 110 120 130 140 150 160 170 180 190 200 210 220
0
20
40
60
Time (s)
Ac
tiv
e P
ow
er
(kw
)
MG main storage
SSMT 1
SSMT 2
SSMT 3
load connection
PVs connectionWG connection
Motor load start up
28
© 2009 29
Emergency Functions –Black Start
10 15 20 25 30 35 40 45 500.9
0.95
1
1.05
Time (s)
Vol
tage
(p.
u.)
Diesel
CHP1
CHP2
0 100 200 300 400 500 600 700 800 900 1000
49.8
50
50.2
50.4
Time (s)
Fre
quen
cy (
Hz)
A Bottom-Up Strategy is
followed for rebuilding the
Multi-MicroGrid network
comprising two main stages:
1. MV network energization
and synchronization of
small islands
2. Load supply and
integration of generation
BLACKSTART
© 2009 30
Ancillary Services Markets
• An Ancillary Services Markets can developed for Voltage Support and Reserves for Normal and Emergency Operation (separated from the main Energy Market)
• An Optimization Algorithm was developed for setting controllers response characteristics in both Energy and Ancillary Services Markets simultaneously
• Functions (existing and new) incorporated in the MGCC and information exchange between CAMC and MGCC regarding Ancillary Services were studied and MicroGrid contribution to voltage violation management has been investigated
© 2009 31
Ancillary Services Markets
0
0,5
1
1,5
2
2,5
Good Citizen Ideal Citizen
VA
R D
em
and
[M
VA
R]
Microgrid Policy
HV BUS
CHP
DIESEL
DFIM
µG5
µG4
µG3
µG2
µG1
Selected VAR Bids (“Good Citizen” Policy vs. “Ideal Citizen” Policy)
Normal Interconnected Operation
VAR MARKET
Normal and Emergency Mode
Import required No Import required
© 2009
SmartMetering infrastructure fostering Microgrids
•
ICTs
32
© 2009
Microgrids and Plugged in Electric Vehicles
• PEV are: controllable charges and mobile storage devices that
need to be controlled and managed.
33
© 2009 34
Final Remarks
• The main issues to be dealt with in the future are the deployment of Smart Metering as a mean of pushing forward the development of MicroGrids (MG) as an integrated part of the general Smart Grid concept
• Massive integration of Distributed Storage, based on mobile storage (Electrical Vehicles) or on stationary storage (fuel cells, regenerative fuel cells, ion-lithium batteries, etc.) needs to be further studied
• A full assessment of active demand side management strategies (managed by the DSO) should also be carried out.
• Integrated management of MicroGrids allows the integration of flexible DG, flexible consumption (including EV battery charging – smart charging) and flexible storage. Flexibility
• Regulatory issues need to further addressed (quantity of needed flexibility and value of flexibility) for system operation in normal and emergency modes.