Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Impact of Distributed Generation On Unbalanced Distribution Networks

Download as pdf or txt
Download as pdf or txt
You are on page 1of 4

CIRED 22nd International Conference on Electricity Distribution Stockholm, 10-13 June 2013

Paper 1020

IMPACT OF DISTRIBUTED GENERATION ON UNBALANCED DISTRIBUTION


NETWORKS
Seyed Hossein Hashemi Mohammad hosein Ashouian Hamidreza Pirpiran
Esfahan Province Electricity Dis.Co – Iran Behrad Consulting Engineers Co. – Iran Esfahan Province Electricity Dis.Co – Iran
Azad University of Khomeini Shahr -Iran h.ashourian@fkegraduate.utm.my pirpiran@epedc.ir
h.hashemi1363@gmail.com
Reza Karami
Esfahan Province Electricity Dis.Co – Iran
mohandesi@epedc.ir

Unbalanced Voltage that is one of power quality index of


ABSTRACT distribution networks. An unbalanced voltage can cause
increased losses in motor loads and abnormal operation of
Unbalance active distribution network (UADN) is the sensitive equipment„s. unbalance voltage problem will be
combination of unbalance distribution network (UDN) and improved when DG operate as PV mode. In [5] the authors
distributed generations (DGs) such as gas generator, wind proposed a definition on unbalance voltage and some effects
turbines, fuel cell, photovoltaic system, etc. of it on utilities. Juanuwattanakul et. al [6] investigate the
Regarding to unbalance structure, different kinds of DGs effect of a voltage regulator and DG unit location in
models and bidirectional power flow, some network improving voltage profile, reducing power loss and
parameters such as amplitude voltage, voltage unbalance increasing voltage stability margin.
Factor (VUF), and power losses of network will alter As discussed above, DGs in UADN have significant effect
accordingly. A suitable power flow analysis will provide a on power flow analysis and may improve parameters of
platform to analyse these behaviours in steady state network such as unbalanced voltage, system losses and so
condition. This paper offers appropriate method for on. Although, various researches have been done in this
handling DGs in a Power flow program which is developed field, there are less examination to study the interaction of
using MATLAB® for practical unbalanced distribution power between DGs and unbalanced distribution network
network at Iran and investigate the impact of DG location that is named UDN.
and DG control mode (as PV and PQ mode) on voltage This Study is structured in several sections. In sections I and
profile, reducing power loss and VUF. II provide prepared models and an unbalance radial load
flow based of forward-backward sweep method. In section
I. INTRODUCTION III, a definition of percentage unbalance voltage is
Active Distribution Networks (ADNs) are defined as a explained. In section IV, numerical results of power flow
distribution network combined with distributed generations analysis with handling DGs as PQ and PV nodes on one
(DGs). DGs are included from green energy resources such practical unbalanced radial distribution systems and the
as wind turbines, PV, etc and non-green energy resources behavior of these models on UADN are presented and
like Diesel/Gas generators, micro turbine, etc. Energy discussed. Finally in the last section is the conclusion.
efficiency, reduction of the power loss, reduced and deferral
distribution investments, network (voltage) support, quality
of supply improvement are some of the expected benefits of II. UNBALANCED ACTIVE DISTRIBUTION
using DGs. Though, there are lot of advantageous of using NETWORK MODELING
DGs, but some concerned issues can be put to using DGs
such as network protecting, load shedding, DG discarding, structure of UDN
voltage supporting and bidirectional power flowing[1,5]. The important step in modeling unbalance distribution
In order to development of ADN, appropriate operation and network is line modeling. A precise model of line (both
control, It should be concerned both DG operation and overhead and underground) in three-phase unbalance
distribution network structures in power flow analysis which distribution network, which is shown in Fig.1, is developed
is an unavoidable tool for power system analysis. Radial by Kersing [7].
network structure, an extremely large number of branches/ In distribution system, loads can be modeled as: constant
nodes, an unbalanced distribution load, unbalanced real and reactive power (constant PQ), constant current,
operation and high penetration of DGs. a suitable method constant impedance, and any combination of these which
three- phase unbalanced radial power flow solution has to each of them maybe wye or delta connections with one,
be addressed. Some works have been proposed to satisfy two or three-phase structure. A comprehensive description
some of these features [2]. Handling DGs as PV and PQ of all types of loads with mathematical modeling is taken
models in power flow study are discussed in [4]. into account in [7].

CIRED2013 Session 4 Paper No 1020


CIRED 22nd International Conference on Electricity Distribution Stockholm, 10-13 June 2013

Paper 1020

Delta-grounded wye, ungrounded Wye-delta, Grounded interface that improve power quality in ADNs. Fig.2 shows
wye-grounded wye, delta-delta and open wye-open delta a combination of DGs in distribution network.
transformers used in unbalance distribution network are
modeled in [7]. Capacitors in unbalance distribution
network can be modeled as constant impedance loads.

Fig.2: Combination of micro sources in distribution


network [5]
UADN Load-Flow Procedure
Fig 1: Three-phase line segment model. [7]
The procedure to study load flow for UADN needs to
follow several steps. Before starting the main power-flow
Structures of DGs steps, the networks should be recognized. Reading network
Structures of DGs are introduced by PQ and PV models in data, identifying all nodes beyond all the branches,
power flow study. Designating real and reactive power constructing PV node sensitivity matrix ZV are some of
constrains PQ model at DG nodes which will prepare a these cases which were introduced by Shirmohammadi and
Constant power factor and specifying real power- voltage Gosh in [4,5]. DGs as PQ operation are modeled constant
(PV) will make fixed voltage at DG nodes within limited PQ-load. After this, load-flow study without considering PV
reactive power. In the same way, constant PQ-load with node will be done. These steps within kth interaction for mth
current injecting into node is substituted by PQ model and node are shown as follow:
since in radial power flow handling PV model cannot be in Step 1: Nodal Current Calculation
the direct manner and need some supplementary procedures, Step2: Backward Sweep
PV model is represented by a compensation current injected Step3: Forward Sweep
by the DG that is a function of the terminal voltages. It Step4: Handling DGs as PV Node
should be noted that identifying the appropriate PQ and PV At PV node, it is considered to hold voltage node at specific
models for each DG in power flow study need to be value .For this purpose, appropriate reactive current will be
recognized its connection to network (direct or indirect) injected to node by DGs and it means DGs should prepare a
beside of its operation. DGs can be connected to network by proper reactive power according to this reactive current
means of induction generators or synchronous generator or injection. A compensation-based method can be used for
static power converter interfaces. this purpose, as described in literatures [2,7]. Voltage
Induction generators which are connected to network convergence criterion, after updating reactive power which
without power electronic interfaces (directly) can be has been described in step 4, return to step 1, 2 & 3. This
modeled as PQ nodes. Some induction engines and wind algorithm will continue till the convergence of all PV nodes
turbines are in this group. Synchronous machines which reach. Fig. 3 shows the flowchart of UADN load flow
depend on either regulating excitation voltage type or fixed procedure with considering DGs as PV and PQ operation.
excitation voltage type may be modeled by PV and PQ node
respectively. Gas turbine and some internal combustion III. UNBALANCE VOLTAGE MODELING
engines, which are connected directly to network, are The uneven distribution of single-phase loads, asymmetrical
located in this category. Fuel cells, photovoltaic systems, transformer winding impedances, open wye and open delta
micro-turbines and more wind turbines use a mixture of transformer banks, asymmetrical transmission impedances
both power electronic utility and electrical machine or only many other causes may make a voltage unbalance in ADN.
by means of power electronic utility to inject power into the The voltage unbalance Factor (VUF) in percent is defined
network. Power electronic interface, which are inverters, by the National Electrical Manufacturers Association
rectifiers or AC/AC converters, use both independent P, Q (NEMA) in Standards Publication no. MG 1-1993:
control and independent P, V control interface. Based on
the control strategy which is used in converter it can be Maximum Deviation From Average
VUF  100 (1)
defined as PQ and PV nodes. Flexible and high speed Average of Three Phase to  Phase Voltage
control active and reactive interaction power between DGs NEMA standard defines that induction machine and utilities
and network is significant feature of power electronic may be derated when the voltage unbalance Factor goes

CIRED2013 Session 4 Paper No 1020


CIRED 22nd International Conference on Electricity Distribution Stockholm, 10-13 June 2013

Paper 1020

beyond 1.0% As can be seen in Fig. 5, with missing phase b at bus 62,
Read Network Data VUF at some buses increase more than 1 percent (more than
Form Breakpoint NEMA standard margin). The maximum VUF is about
Impedance Matrix&
PV sensitivity Matrix 1.048% at Bus 62. Total real losses leads to 256 KW.
Calculate P & Q at DG
Node
Network analysis with DG as PQ and PV
Solve 3-Phase Radial

Nodal Current
Load-Flow Operation
Calculation and Then
Calculate ΔQ regarding
Backward & Forward
Sweep again to sensitivity Because of limited land, 4 locations are chosen to install
Reactance Xv
no Breakpoint V Mismatch ≤ ϵ1
DG (bus number 16, 38, 56 and 74).The result of load-flow
yes
analysis for unexpected condition after connecting 2MW
no DG (at buses 16, 38, 56 and 74) as PV and PQ control
PV Node |V|Mismatch ≤ ϵ2
mode are shown in Table. 1.
yes

Print Results Connect DG at Bus 16:


Fig.3 Load- flow solution algorithm for UADN With connecting DG at bus 16, VUF Reduced to less than
1% in PV mode operation but the maximum voltage
IV.CASE-SYUDYDESCRIPTION, SIMULATION AND deviation for PV and PQ modes exceed from standard
ANALYSIS margin (5%). The minimum power losses for this
To verify the unbalance radial load flow program, an 89- installation is 119 KW for PV operation mode.
bus, 20KV practical unbalanced distribution network at Iran Connect DG at bus 38:
are used and developed in MATLAB®. The load and line With connecting DG at bus 38, voltage deviation at all
data and single line diagram for case-study networks are buses for PV and PQ mode operation put to standard
given in APENDIX. The result of power flow analysis margin, while VUF at PQ mode exceed from 1%. Total
without DGs is show in Fig.4. power loss for PV mode reduced to 105 KW.

Connect DG at bus 56:


When connecting DG to bus 56, voltage deviation and VUF
at all Bus for PV and PQ mode reach to standard margin.
Voltage deviation at all buses is less than 4.9%.
As can be seen in Table. 1, Total power loss for PQ and PV
mode reduced to 106 KW and 101KW, respectively.

Connect DG at bus 74:


When connecting DG to bus 74, maximum voltage
Fig.4 : voltage profile and VUF of case-study
deviation for PV and PQ mode exceed than standard margin
Since the favorable voltage range is between 0.95 and 1.05 (5%) and total power loss is 117 KW for PQ operation
p.u, there is violation at many buses. The minimum voltage mode and 112 KW for PV operation mode.
occurs at bus 62 with about 0.93 p.u. Total real losses for
this case is 270 KW. In general, with connecting DG at Bus 56, maximum
voltage deviation, VUF and Total power loss reach to the
Unexpected condition lowest possible value and this place is best candidate for
To study unbalance condition and show significant installing. The result of power flow analysis with connect
behaviour of DGs as PV&PQ models in UADN, phase (b) DGs at bus 56 as PV mode are shown in Fig6.
of load at bus 62 in case study is missed. The result of
power flow analysis in this situation show in Fig.5.

Fig.6 power flow analysis with connect DG at bus 56 as


PV mode operation
Fig.5 voltage profile and VUF after missing phase b at
Bus 62

CIRED2013 Session 4 Paper No 1020


CIRED 22nd International Conference on Electricity Distribution Stockholm, 10-13 June 2013

Paper 1020

Table.1: result of load-flow analysis and Analysis, CRC Press, New York.
DG Total Maximum Maximum Voltage Feeder Input Feeder Input
Location
Opera tion
DG Status Loss Voltage Deviation @ Bus @ Active Power Reactive APENDIX:
Mode
(Bus No.) (KW) Unbalancing Phase (MW) Power(MWAR)
no DG Normal Condition 270 0.77 % at 62 7.61% at 62 at phase B 6.28 3.63
Table.2: Load and line data
sendi ng recei vi ng conductor l ength Recei vi ng end l oad i n KW
no DG Unexpected Condition 256 1.048 % at 62 7.33% at 62 at phase B 6.13 3.54 branch bus bus type (m) Pa Qa Pb Qb Pc Qc
P(kw) 2000 1 1 2 1 126 0 0 0 0 0 0
P-Q 128 1.01 % at 62 5.5% at 62 at phase B 4 2.38 2 2 3 2 385 10 6 11 7 13 8
Q(kwar) 970 3 2 4 1 828 0 0 0 0 0 0
16 P(kw) 2000 4 4 5 2 95 17 11 16 10 15 9
5 4 6 1 889 0 0 0 0 0 0
P-V Q(kwar) 1400 119 0.93 % at 62 5.1% at 62 at phase B 3.99 1.94 6 6 7 2 243 170 82 110 53 100 48
V (p.u) 0.97 7 6 8 1 839 24 15 22 13 20 12
8 8 9 1 126 0 0 0 0 0 0
P(kw) 2000 9 9 10 2 170 7 5 6 4 7 4
P-Q 115 1.00 % at 62 4.9% at 62 at phase B 3.99 2.36
Q(kwar) 970 10 9 11 1 175 0 0 0 0 0 0
11 11 12 2 105 25 12 25 12 25 12
38 P(kw) 2000 12 11 13 1 1156 0 0 0 0 0 0
P-V Q(kwar) 1420 105 0.88 % at 62 4.5% at 62 at phase B 3.98 1.89 13 13 14 2 75 67 42 85 53 57 35
14 14 15 2 164 43 27 65 40 30 19
V (p.u) 0.97 15 13 16 1 150 0 0 0 0 0 0
P(kw) 2000 16 16 17 1 140 6 3 6 3 5 3
P-Q 106 0.98 % at 62 4.9% at 82 at phase B 3.98 2.35 17 17 18 1 114 0 0 0 0 0 0
Q(kwar) 970 18 18 19 2 39 7 4 6 4 8 5
56 P(kw) 2000 19 19 20 2 122 0 0 0 0 0 0
20 20 21 2 139 13 8 6 3 11 7
P-V Q(kwar) 1130 101 0.80 % at 62 4.7% at 82 at phase B 3.97 2.18 21 20 22 3 120 0 0 0 0 0 0
V (p.u) 0.97 22 22 23 2 109 0 0 0 0 0 0
23 23 24 2 180 20 10 15 8 19 9
P(kw) 2000
P-Q 117 1.01 % at 62 5.4% at 62 at phase B 3.99 2.36 24 23 25 2 182 0 0 0 0 0 0
Q(kwar) 970 25 25 26 2 104 2 1 2 1 2 1
26 25 27 2 147 63 39 20 12 43 21
74 P(kw) 2000 27 18 28 1 105 63 31 40 19 40 19
P-V Q(kwar) 1170 112 0.92 % at 62 5.2% at 62 at phase B 3.98 2.15 28 28 29 1 39 0 0 0 0 0 0
29 29 30 2 18 21 13 19 12 24 15
V (p.u) 0.97 30 29 31 1 110 21 13 18 12 25 16
31 31 32 1 207 0 0 0 0 0 0
32 32 33 2 477 31 19 41 25 29 18
33 32 34 1 80 0 0 0 0 0 0

V. CONCLUSION 34
35
36
34
35
34
35
36
37
2
2
1
398
452
85
24
30
20
15
19
13
17
15
10
11
7
5
30
25
16
19
15
8
37 37 38 1 443 0 0 0 0 0 0
38 38 39 1 192 0 0 0 0 0 0
Analysis of simulation Results and utilizes voltage 39
40
39
39
40
41
1
1
41
312
11
14
7
9
25
8
13
5
20
14
12
9

Deviation, Unbalanced Voltage Factor and power loss can 41


42
43
38
42
42
42
43
44
1
2
1
225
214
54
0
17
80
0
11
50
0
10
30
0
6
19
0
15
40
0
9
19

be used to validate the placement of DG units. For this 44


45
44
45
45
46
1
2
62
94
0
42
0
20
0
30
0
15
0
25
0
12
46 45 47 1 320 0 0 0 0 0 0
study operation of DG as PV mode not only decrease VUF 47
48
47
47
48
49
2
1
492
520
25
30
12
19
14
15
7
7
24
27
15
17

and voltage deviation but also decrease total power losses. 49


50
51
49
50
51
50
51
52
1
2
2
713
95
452
0
165
30
0
107
15
0
135
15
0
84
7
145
0

17
0
70
8
52 50 53 1 227 0 0 0 0 0 0
53 53 54 2 28 165 77 141 68 120 60
54 54 55 2 24 49 29 40 25 67 34

REFERENCES 55
56
53
56
56
57
1
2
394
240
0
49
0
31
0
43
0
28
0
55
0
34

[1] R. Lasseter, A. Akhil, C. Marnay, J. Stephens, J. 57


58
59
56
58
59
58
59
60
1
1
1
165
726
77
25
38
0
12
24
0
15
15
0
9
8
0
10
32
0
5
20
0

Dagle, R. Guttromson, A. Melliopoulos, R. Yinger, 60


61
60
60
61
62
1
2
482
238
63
180
31
112
40
0
19
0 130
30 15
63
62 16 63 1 160 30 15 20 10 60 29

and J. Eto, 2002, White Paper on Integration of 63


64
63
64
64
65
1
1
68
210
9
0
5
0
4
0
3
0
6
0
4
0
65 65 66 2 42 135 65 120 58 80 39
Distributed Energy Resources, The CERTS Microgrid 66
67
65
67
67
68
1
1
67
56
5
13
2
6
5
6
2
3
5
14
2
7

Concept, Office of Power Technologies of the U.S. 68


69
68
69
69
70
1
1
840
263
13
0
8
0
15
0
7
0
11
0
8
0
70 70 71 2 67 2 1 2 1 2 1

Department of Energy, Contract DE-AC03-76SF0009. 71


72
70
72
72
73
1
1
85
107
9
2
4
1
11
2
6
1
9
2
4
1
73 73 74 1 271 0 0 0 0 0 0
[2] D. Shirmohammadi, H.W. Hong, A. Semlyen, and 74
75
74
75
75
76
1
1
136
175
0
5
0
2
0
5
0
2
0
5
0
2

G.X. Luo, 1988, “A Compensation-based power flow 76


77
78
76
75
78
77
78
79
1
1
2
302
529
103
45
0
160
28
0
78
40
0
125
25
0
55 112
30
0
18
0
52

method for weekly meshed distribution and 79


80
78
80
80
81
2
2
1157
424
0
11
0
5
0
11
0
5
0
11
0
5
81 81 82 2 259 125 67 50 30 88 43
transmission networks”, IEEE Transaction on Power 82
83
80
74
83
84
2
1
79
162
140
0
68
0
120
0
71
0
163
0
88
0

Systems, vol. 3, no. 2, 753-762. 84


85
86
84
84
86
85
86
87
1
1
2
627
188
115
7
0
170
4
0
82
4
0
120
3
0
58 145
5
0
2
0
70

[3] C.S. CHENG, D. SHIRMOHAMMADI, 1995, “A 87


88
86
88
88
89
1
1
88
349
18
20
9
10
14
21
7
10
18
35
9
20

three-phase power flow method for real -time Table.3: Conductor's Parameter
distribution system analysis”, IEEE Trans. Power Type Impedance in ohm/km
0.3359+0.8975i 0.0593+0.4826i 0.0593+0.4826i
Systems, 10, (2), 671–679. 1 0.0593+0.4826i 0.3359+0.8975i 0.0593+0.4826i
0.0593+0.4826i 0.0593+0.4826i 0.3359+0.8975i
[4] S. Khushalani, J.M. Solanki and N. N.Schulz, 2007, 0.5225+0.952i 0.0593+0.4826i 0.0593+0.4826i

“Development of Three-Phase Unbalanced Power


2 0.0593+0.4826i 0.5225+0.952i 0.0593+0.4826i
0.0593+0.4826i 0.0593+0.4826i 0.5225+0.952i
0.5112+0.8605i 0.0593+0.6917i 0.0593+0.6917i
Flow Using PV and PQ Models for Distributed 3 0.0593+0.6917i 0.5112+0.8605i 0.0593+0.6917i
0.0593+0.6917i 0.0593+0.6917i 0.5112+0.8605i
Generation and Study of the Impact of DG Models”,
IEEE Transactions on Power Systems, Vol. 22, NO. 3,
1019 – 1025.
[5] M.H. Ashourian, A.A.M Zin, A.S.B Mokhtar and Z.
Muda, 2010, “Steady-State Behavior of Distributed
Energy Resources in Unbalance Radial Active
Distribution Networks”, IEEE International Power
and Energy Conference, 282 – 287.
[6] P. Juanuwattanakul, M.A.S. Masoum, P.S Moses,
2011, “Voltage analysis for placement of DG in
unbalanced distribution networks”, 11th International
Electrical Power Quality and Utilisation
Conference,1–4.
[7] W.H. Kersing, 2002, Distribution System Modeling Fig.7 : single line diagram

CIRED2013 Session 4 Paper No 1020

You might also like