solutions 2018 for integrating photovoltaicpanels …...of the xx power systemscomputation...
TRANSCRIPT
Solutions for IntegratingPhotovoltaic Panels into Low-voltage Distribution NetworksFrédéric OlivierAdvisor: Prof. Damien Ernst
2018
2
LV distribution network
• Distribution transformer• Electrical lines• Nodes (buses)• Houses• Smart meters
2
3
PV panels and LV distribution networks
Voltage too high
Disconnection of PV
Loss of renewable energy production
Loss of earnings
Slowing down the energy transition
4
How to increase the hosting capacity?
• Reinforcing the network• Balancing the phases of the network• Identifying the phases of the smart meters• Balancing the production and consumption on the three phases
• Actively managing the network• Controlling the active and reactive power flows inside the network
Part 1Phase Identification of Smart Meters
6
The phase identification problem
?
?
𝑎𝑏𝑐𝑛
Network
𝑟&
𝑟'
𝑠' 𝑡'
𝑠&
𝑡&
𝑟*
Cluster 𝒞&
Cluster 𝒞'
Cluster 𝒞*
6PSCC 2018
𝑎𝑏𝑐𝑛
Smar
t met
erSm
art m
eter
𝑟𝑠𝑡
𝑟
𝑛
𝑛
𝑀-&𝑀.&𝑀/&
𝑀-*
Measurements
7
Why is the phase information important?What are the existing solutions?What is the algorithm we propose?What is its performance ?
8
On the importance of phase identification
If you are a DSO
If you are a researcher
8
Change connection
phase of the customer
Reduce the imbalance in the network
Increase the hosting capacity
Phase-to-neutralactive and reactive
power measurements
Three-phasepower flow
Comparison between
measured and simulated voltages
9
Manual
Existing solutions
Automatic
9
Phase identifiers
Smart meters with PLC
k-means clustering
Graph theory
10
Contributions
1. Novel algorithm 1. Using the underlying structure of the network2. Using the advantages of both graph theory and correlation 3. Identifying the measurements that should be linked together and cluster
them. 2. Performance compared to those of a constrained k-means
clustering3. Tested on real measurements from a distribution network in
Belgium, in a variety of settings.
10
11
• Distance between two voltage measurements:• Pearson’s correlation
𝑑 𝑀1,𝑀3 = 1 − 𝑃𝐶 𝑀1,𝑀3 , ∀𝑙, 𝑖 ∈ ℐ
• Distance between a voltage measurement and a cluster
Δ 𝒞@,𝑀3 = min1∈𝒞D
𝑑(𝑀1,𝑀3)
Distances
11
12
Reference algorithmConstrained k-means Clustering
12
13
Proposed algorithmConstrained Multi-tree Clustering
13
Olivier,F.,Sutera,A.,Geurts,P.,Fonteneau,R.,&Ernst,D.(2018).Phaseidentificationofsmartmeters byclustering voltagemeasurements.In ProceedingsoftheXXPowerSystems ComputationConference (PSCC2018).
14
Test system
• Belgium LV distribution network• 5 feeders, star configuration 400 V• 79 three-phase smart meters• 2 single phase smart meters• Average phase-to-neutral voltage
measurements every minute
14
15
Results for the test setsDiscussions on the selection of the root
0
0,2
0,4
0,6
0,8
1
CKM CMT
Transformer House
Performance measureThe ratio between the measurements correctly identified and the total number of measurements
16 16
17
Influence of the voltage-averaging period
CMT
CKM
18
Influence of ratio single-phase – three-phase smart meters
CKM CMT
19
Influence of the number of smart meters
CKM CMT
Part 2 Active network management
21
Active network management
DesignstrategiestocontroltheactiveandreactivepowerinjectedbythePVpanelsintothedistributiongridsoasto
ensurethatthevoltagesstaywithintheirlimits
Objective
22
Active network management
Centralized controller• Model required• Communication infrastructure
Distributed controllers• Model-free• Little communication• Simpler logic
23
What are the key principles of the control scheme?What is the structure of the algorithm?How can we adapt the control scheme to three phases?What is its performance compared to other algorithms?
24
Key principles
• Distributed algorithm• No central entity • Each inverter applies locally the algorithm• No model required• First use reactive power• Curtail active power as a last resort• Limited communication in the form of a distress signal to pool the
resources
Olivier,F.,Aristidou, P.,Ernst,D.,&VanCutsem,T.(2016).Activemanagementoflow-voltagenetworksformitigating overvoltages duetophotovoltaic units. IEEETransactionsonSmartGrid, 7(2),926-936.
25
Assuming that the three phases are balanced
M M M M M M M M M M
N1 N2 N3 N4 N5 N6 N7 N8 N9 N10
Feeder bus
26
27
Mode𝑄1HI↓↑ (𝑉)
Mod
es o
f the
alg
orith
m
28
Mode 𝑄1HI↓↑ (𝑉)
Qf
Vtm
V2
V3
�Q(i)
max
⇣P (i)
set,t
⌘
Q(i)
max
⇣P (i)
set,t
⌘
V1
V4
Mode Q#"loc
(V )
29
Capability curve of PV inverters
Q
P
Inverter rated forreactive support at
full output(Curve A)
Inverter rated forreactive support atpartial output only
(Curve B)
30
Mode𝑄 ↑
Mode𝑄 →
Mode𝑄 ↓
Mode𝑄1HI↓↑ (𝑉)
Mode𝑃 →
Mode𝑃 ↓
Mode𝑃 ↑
31
Single feeder test system
M M M M M M M M M M
N4AB1
N4AB2
N4AB3
N4AB4
N4AB5
N4AB6
N4AB7
N4AB8
N4AB9
Node 4B
0.4 kV15 kV
Node 4ANode 4
MV LV
32
0
5
10
15
20
25
30
0 400 800 1200 1600 2000 2400 2800 3200
t (s)
(kW)
First try to restore active power
Second try to restore active power
PV N4AB1 PV N4AB2 PV N4AB3 PV N4AB4 PV N4AB5
PV N4AB6 PV N4AB7 PV N4AB8 PV N4AB9 PV NODE4B
-10
-8
-6
-4
-2
0
0 400 800 1200 1600 2000 2400 2800 3200
t (s)
(kVAr)
PV N4AB1 PV N4AB2 PV N4AB3 PV N4AB4 PV N4AB5
PV N4AB6 PV N4AB7 PV N4AB8 PV N4AB9 PV NODE4B
0
5
10
15
20
25
30
35
0 400 800 1200 1600 2000 2400 2800 3200
t (s)
(kW)
PV N4AB1 PV N4AB2 PV N4AB3 PV N4AB4 PV N4AB5
PV N4AB6 PV N4AB7 PV N4AB8 PV N4AB9 PV NODE4B 0.98
1
1.02
1.04
1.06
1.08
1.1
0 400 800 1200 1600 2000 2400 2800 3200
t (s)
(pu)
First try to restore active powerSecond try to restore active power
N4AB1 N4AB2 N4AB3 N4AB4 N4AB5
N4AB6 N4AB7 N4AB8 N4AB9 NODE4B
33
When the three phases are unbalanced
• Four groups of inverters• One for each phase and one for three-phase inverters
• Three distress signals• One per phase
• Each group applies the control scheme independently
34
Unbalanced test system
PV0
L0
PV1
L1
PV2
L2
PV3
L3
PV4
L4
PV5
L5
PV6
L6
PV7
L7
PV8
L8
PV9
L9
PV10
L10
PV11
L11
PV12
L12
PV13
L13
PV14
L14
PV15
L15
PV16
L16
PV17
L17
PV18
L18
External network
35
PV connection scenarios
PV0
L0
PV1
L1
PV2
L2
PV3
L3
PV4
L4
PV5
L5
PV6
L6
PV7
L7
PV8
L8
PV9
L9
PV10
L10
PV11
L11
PV12
L12
PV13
L13
PV14
L14
PV15
L15
PV16
L16
PV17
L17
PV18
L18
External network
Thre
e-ph
ase
Thre
e-ph
ase
Scenarios 3P
Thre
e-ph
ase
Thre
e-ph
ase
Thre
e-ph
ase
Thre
e-ph
ase
Thre
e-ph
ase
Thre
e-ph
ase
Thre
e-ph
ase
Thre
e-ph
ase
Thre
e-ph
ase
Thre
e-ph
ase
Thre
e-ph
ase
Thre
e-ph
ase
Thre
e-ph
ase
Thre
e-ph
ase
Thre
e-ph
ase
Thre
e-ph
ase
Thre
e-ph
ase
36
PV connection scenarios
PV0
L0
PV1
L1
PV2
L2
PV3
L3
PV4
L4
PV5
L5
PV6
L6
PV7
L7
PV8
L8
PV9
L9
PV10
L10
PV11
L11
PV12
L12
PV13
L13
PV14
L14
PV15
L15
PV16
L16
PV17
L17
PV18
L18
External network
Thre
e-ph
ase
Thre
e-ph
aseA B C A B C A B C A B C A B C A B
Scenarios ABC
37
PV connection scenarios
PV0
L0
PV1
L1
PV2
L2
PV3
L3
PV4
L4
PV5
L5
PV6
L6
PV7
L7
PV8
L8
PV9
L9
PV10
L10
PV11
L11
PV12
L12
PV13
L13
PV14
L14
PV15
L15
PV16
L16
PV17
L17
PV18
L18
External network
Thre
e-ph
ase
Thre
e-ph
aseA A B C A A B C A A B C A A B C A
Scenarios AABC
38
PV connection scenarios
PV0
L0
PV1
L1
PV2
L2
PV3
L3
PV4
L4
PV5
L5
PV6
L6
PV7
L7
PV8
L8
PV9
L9
PV10
L10
PV11
L11
PV12
L12
PV13
L13
PV14
L14
PV15
L15
PV16
L16
PV17
L17
PV18
L18
External network
Thre
e-ph
ase
Thre
e-ph
aseA A A B C A A A B C A A A B C A B
Scenarios AAABC
39
PV connection scenarios
PV0
L0
PV1
L1
PV2
L2
PV3
L3
PV4
L4
PV5
L5
PV6
L6
PV7
L7
PV8
L8
PV9
L9
PV10
L10
PV11
L11
PV12
L12
PV13
L13
PV14
L14
PV15
L15
PV16
L16
PV17
L17
PV18
L18
External network
Thre
e-ph
ase
Thre
e-ph
aseA A A A A A A A A A A A A A A A A
Scenarios AAA
40
Simulations for a sunny day
06:00 08:00 10:00 12:00 14:00 16:00 18:00 20:00
0
0.2
0.4
0.6
0.8
Time
[kW
/kW
p]
41
Comparison between algorithms
On-off• The one already implemented in commercial
inverters
Distributed• The one we propose
Optimal power flow• The one centrally optimizing the power
produced by the PV
42
Energy curtailed
3P ABC AABC AAABC AAA0
100
200
300
400
500
50
95
155
205
450
41 31
113
185
447
14 1.2 38
98
324
Ene
rgy
curt
aile
d[k
Wh]
On-off Distributed OPF
43
Reactive energy usage
3P ABC AABC AAABC AAA0
10
20
30
40
50
60
0 0 0 0 0
31
22
28
25
22
30
23
51 53
35
Rea
ctiv
een
ergy
[kVA
rh]
On-off Distributed OPF
44
Losses
3P ABC AABC AAABC AAA5
10
15
20
25
30
19.3
18.8
17.5
18
16.5
20.1 22
.3
17.9
14.8
10.8
21.8
24.7
23.6
20.7
16.2
Los
ses
[kW
h]
On-off Distributed OPF
45
Batteries
46
Local energy communities
• Definition
Electricitydistributionsystemcontainingloadsanddistributedenergyresources(suchasdistributedgenerators,storagedevices,orcontrollableloads),thatcanbeoperatedina
controlled,coordinatedway
47
Local energy communities
• Drivers• With a shared infrastructure between the members• Without a shared infrastructure
• Network operation• Energy market
Communities extend theperimeter ofself-consumption fromoneprosumer toseveral topool
productionandflexibility means
47
48
Electric vehicles
49
Conclusion – From research to industry
Question1Identifying thenetworktopology andthephasesofthesmartmeters
Question2Computing thehostingcapacity networks
Question3Preventive actions
Question4Correctiveactions
Solutions for IntegratingPhotovoltaic Panels into Low-voltage Distribution NetworksFrédéric OlivierAdvisor: Prof. Damien Ernst
2018