Transcript Introdução

CIRED’2003
Beta session 4a: Distributed Generation
Controllability of DG helps managing
Distribution Grids
J. A. Peças Lopes
([email protected])
Barcelona– May 2003
CIRED’2003
Beta session 4a: Distributed Generation
Exploiting DG to improve system operation
•
DG has been considered as non controllable and non
dispatchable, since all the energy production has priority to be
absorbed by the network;
•
The increase in DG foreseen for the next years will require a
different approach regarding the way how DG units will be
operated:
– Concepts of controllability should be developed and exploited:
• Participation in reactive power control;
• Interruptability;
• Delivery of ancillary services (primary and secondary reserves,
according to the conversion technology and primary energy
sources);
• Participation in system restoration strategies;
– Development of concepts related with control of clusters of DG and
virtual power stations;
Barcelona– May 2003
CIRED’2003
Beta session 4a: Distributed Generation
Main characteristics of DG units and
controlability concepts
• Three main types of energy conversion systems can be
found among DG units:
– Conventional synchronous machines (cogeneration, CHP,
mini-hydro);
– Asynchronous generators (wind power, mini-hydro);
– AC/DC/AC electronic conversion systems used together with
synchronous or induction machines (micro-turbines, fuel cells,
wind generators).
• Classification (according to primary energy source and
conversion system used):
– Non- controllable (Ex: Wind park with asynchronous stall
generators);
– Partially controllable (Ex: Wind park with synchronous
variable speed gen. and AC/DC/AC converters);
– Controllable (Ex: Mini-hydro or Cogeneration plant with
synchronous units).
Barcelona– May 2003
CIRED’2003
Beta session 4a: Distributed Generation
DG units used to optimise the distribution system operation
•
DG can be used to optimise the operation strategy of distribution
networks.
The Problem can be formulated an optimisation problem:
Min (active power losses)
Subj. to:
Vmax < Vi < Vmin
Sij max < Sij
Qgmaxi< Qgi < Qgmini
taking into account the type of generator
Qimpor max < Qimpor
Transformer tap limits are kept
Control variables: Qg, capacitor banks and transformer taps
The need to use a motor of optimisation
(Evolutionary Particle Swarm Optimisation – EPSO)
Barcelona– May 2003
CIRED’2003
Beta session 4a: Distributed Generation
Some results of the participation of DG in Voltage
VAR control
•
Test System: 60 kV distribution network with a large penetration of
DG (mini-hydro and wind generation).
Activate control
on reactive power
generated in the DG
Units.
Barcelona– May 2003
CIRED’2003
Beta session 4a: Distributed Generation
Some results of the participation of DG in
Voltage VAR control
Changes in active Losses
– Peak load scenario
– A clear reduction on actives
losses was obtained
Initial Losses
2500
Final Losses
2000
1950
Difference
1641
Losses (kW)
•
1500
1000
500
309
0
Barcelona– May 2003
CIRED’2003
Beta session 4a: Distributed Generation
Some results of the participation of DG in
Voltage VAR control
Results concerning voltage in network busses
Initial Scenario
Final Scenario
1.14
1.12
1.10
1.08
1.06
1.04
1.02
1.00
44
43
42
41
40
39
38
37
36
35
34
28
27
26
25
24
17
16
15
14
13
12
11
9
10
8
7
6
5
4
3
2
0.98
1
Voltage (p.u.)
•
Barcelona– May 2003
CIRED’2003
Beta session 4a: Distributed Generation
Dynamic Impacts
• Dynamic behaviour impacts need to be addressed using
adequate DG modelling and DG equivalent representation:
– Considering disturbances resulting from DG operation;
– Considering disturbances in distribution networks;
– Considering disturbances in the transmission system
Barcelona– May 2003
CIRED’2003
Beta session 4a: Distributed Generation
Dynamic behavior analysis
•
Scenario: Week peak with maximum dispersed generation
•
Disturbance: Outage of Power Plant H: 7,346MVA, production of
6,692+j3,03 MVA, injection of 2,678+j1,081 MVA (tg j = 0,404)
•
Voltage profile:
•
60kV bus at the substation
1 .0 4 8
P e r fi l d e te n sã o n ó 2 (p . u . )
1 .0 4 7 5
1 .0 4 7
1 .0 4 6 5
1 .0 4 6
1 .0 4 5 5
sec
1 .0 4 5
1 .0 4 4 5
0
5
10
15
20
25
Barcelona– May 2003
CIRED’2003
Beta session 4a: Distributed Generation
Dynamic behavior analysis
P e r fi l d e te n sã o n ó 3 (p . u . )
1 .0 5 1
1 .0 5
1 .0 4 9
1 .0 4 8
1 .0 4 7
1 .0 4 6
1 .0 4 5
1 .0 4 4
sec
1 .0 4 3
15 kV bus of the
feeder where the
power plant was
connected
1 .0 4 2
1 .0 4 1
0
5
10
15
20
25
P e r fi l d e te n sã o n ó 4 (p . u . )
1 .0 5 0 2
15 kV bus of the
feeder where the
power plant was
not connected
1 .0 5
1 .0 4 9 8
1 .0 4 9 6
1 .0 4 9 4
1 .0 4 9 2
sec
1 .0 4 9
1 .0 4 8 8
0
5
10
15
20
25
Barcelona– May 2003
CIRED’2003
Beta session 4a: Distributed Generation
Dynamic behavior analysis
(Impact in the other generators)
C a n iç o s
f e m C a n iç o s ( p .u .)
1 .2 3
1 .2 2
1 .2 1
1 .2
1 .1 9
sec
1 .1 8
1 .1 7
1 .1 6
0
5
10
15
20
25
C a n iç o s
Pe C a n iç o s ( M W )
4 .3
Pm e c C a n iç o s ( M W )
4 .2
4 .1
4
3 .9
3 .8
sec
3 .7
3 .6
3 .5
3 .4
0
5
10
15
20
25
Barcelona– May 2003
CIRED’2003
Beta session 4a: Distributed Generation
Relay coordination
•
•
Under voltage relay coordination is needed;
Energy conversion systems need to able to withstand low voltages
during short-circuits up-stream.
Frequency changes
Changes in contractual inter-area power flows
A)
B)
A)
B)
Barcelona– May 2003
CIRED’2003
Beta session 4a: Distributed Generation
Impacts on Operation
•
•
•
•
•
•
•
•
•
•
Load flows become bi-directional;
Voltage profiles have different patterns;
Losses change as a function of the production and load levels;
Congestion in system branches is a function of the production
and load levels;
Short-circuit levels increase;
Power quality may be affected;
Voltage transients will appear as a result of connection
disconnection of generators;
Risk of islanding operation;
Reliability may be reduced;
System dynamic behavior may be largely affected;
Protections coordination is needed;
Barcelona– May 2003
CIRED’2003
Beta session 4a: Distributed Generation
Conclusions
• The future:
– DG units should be more actively used to help in the
management of the distribution grid;
– New DMS tools need to be developed:
•
•
•
•
Topology processor with capabilities of identification of energised areas;
Voltage and reactive power control;
Load and current forecasting;
Load flow including new generator models and load allocation
algorithms to allow load flow to run;
• Optimum network reconfiguration;
• State estimation (considering that some DG units will not be monitored
and new pseudo-measures need to be defined);
– Cluster control strategies should be implemented, involving
the development of local dispatch centres;
– Development of DMS training simulators for distribution grids
with large amounts of DG (steady state and dynamic
behaviour).
Barcelona– May 2003