1、 Index Termsflexible distribution network,SNOP,DG,allowable penetration capacity,MPGA AbstractThis paper adopts a new type of power electronics-soft normally open point(SNOP)to solve the bottleneck problem of bus/branch in distribution network,and improve the DG penetration capacity in flexible dist
2、ribution network(FDN).Firstly this paper establishes the optimization model of the allowable capacity of DG in the FDN.A multi-population genetic algorithm(MPGA)is used to get the global optimal solution,in both traditional distribution network and FDN.Then this paper builds the system model of IEEE
3、 33-bus distribution network to verify the conclusion,which finally proves that the application of SNOP in FDN will allow more DG accessing to the network.What is more,this method has certain reference value to determine the optimal capacity and access location of DG in the future FDN.I.INTRODUCTION
4、 As distributed generation(DG)is more widely used in recent years,the high DG penetration level of all kinds become new challenges to the stability of power systems.When the DG penetration in both medium voltage and low voltage distribution networks reaches a high proportion(i.e.high penetration),it
5、 is difficult to maintain balance and stability of the distribution network,also the reliability and quality of the power cant be ensured.How distribution network can accommodate the maximum DG penetration level becomes a hot research topic.In the past research,the maximum DG penetration of the dist
6、ribution network is obtained by considering the constraints of the power flow and the topology of the Manuscript received July 14,2018.This work was supported in part by the technological project of State Grid Tianjin Electric Power Company Research on Key Technologies of Source-Network-Load Coordin
7、ated Planning and Design for Energy Internet(SGTYHT/16-JS-198).network2.In paper 3,by considering the impact of DG location and capacity in the distribution network,the concept of DG operating domain is proposed,and the range of allowed DG capacity of a given node can be obtained.The above research
8、is mainly aimed at the traditional distribution network.In paper 4,it proposes the concept of the flexible distribution network(FDN),which is based on the application of a new power electronic device-soft normally open point(SNOP).At present,the research on SNOP mainly focuses on the operation optim
9、ization of distribution network 7-10.The SNOP optimizes the power flow by accurately controlling the active power exchange on both sides and provides reactive power compensation as needed.It provides the real-time and refined power flow adjustment and optimization,which is not available in the prese
10、nt distribution automation system.It can quickly track the dynamic changes of DG and load,reduce power loss,and ensure that the distribution network is in an optimal state all the time.Paper 11 uses a hybrid algorithm based on simulated annealing and cone optimization to improve the economy of the d
11、istribution network with SNOP,but the core technical target is still to reduce network losses.In fact,in addition to optimize power flow,SNOP can also be used to improve DG penetration.In FDN,relying on SNOPs capability of redirecting power adjustment,the operation state of bottleneck bus/branch be
12、improved,thereby enlarging the operating domain in paper 3 and achieving the high DG penetration.Some research results have proved that it can increase DG penetration with SNOP.In 12,it explores the effect of SNOP in improving PV capacity in distribution networks.In paper 13,it conducts a research o
13、n the network of urban and rural in UK respectively,pointing out that the reactive power supporting capability of SNOP is the key to improving the DG penetration.The paper 14 mainly focuses on the improvement of the Allowable DG Penetration Capacity Calculation of SNOP-based Flexible Distribution Ne
14、twork LIANG Haishen1,ZHANG Kai1,LI Shengwei2,GE Leijiao3,WANG Qingbiao1,HAN Tao4 1.State Grid Tianjin Baodi Power Supply Company 2.Economic and Technical Research Institute,State Grid Tianjin Electric Power Company 3.Tianjin University 4.Tianjin Electric Power Overhauling Company 2018 China Internat
15、ional Conference on Electricity Distribution Tianjin,17-19 Sep.2018CICED2018 Paper No.201805090000012 Page1/52148DG penetration using energy storage coordinated with SNOP,aiming at smoothing PV output fluctuation.However,none of the previous studies could obtain the optimal combination of DG and SNO
16、P from the perspective of planning.This paper establishes the model of the FDN with highest DG penetration.As the model is high-dimensional and nonlinear,a multi-population genetic algorithm is used to solve the problem,which can avoid the premature convergence and obtain the global real optimal sol
17、ution.A model of 33-bus distribution network is used for case study.On the basis of analysis on the capability of SNOP to increase the DG penetration of the given node,the optimal location and capacity of DG in FDN with maximum DG penetration level can be obtained.II.MODEL OF ALLOWABLE DG PENETRATIO
18、N CAPACITY IN FDN A.Outline of SNOP The most widely used SNOP device is back-to-back voltage-source converters(B2B VSC).The SNOP is usually applied to replace the traditional contact switch,as shown in Figure 1.Different from the traditional contact switch,who only has two states:0 and 1,SNOP has th
19、e ability of flexible four-quadrant power controlling,meaning that it can control the real and reactive power flows of both-side continuously.This feature will not only reduce the network loss,but also regulate node voltage and branch current.Traditional interconnection switchSNOPConverterBranch imp
20、edanceBranch impedanceNode loadTransformerSourceFeederFeederDG DGDG DGNode loadNode load Node loadBranch impedanceBranch impedanceConverterFig 1 SNOP in FDN B.Model of Maximum Capacity of DG in the FDN Since the influence of DG and load on the network voltage is reversed,the worst case of DG integra
21、tion is the scenario that the distribution network runs at its minimum load with DG at maximum output.At this time,the DG capacity is equal to the installed capacity.The DG penetration discussed in this paper is based on this scenario.Establish the DG accessing capacity model and choose the maximum
22、DG penetration in the FDN as the optimization goal:,1maxNDG iiE=(1)Where,1NDG iiE=is the total capacity of all the DG in the FDN,and N is the number of nodes in the network.The constraints are as follows:,1(cos sin)0ND G i S N O P i L i i j ij ij ij ijjP P P U UG B=+=(2),1(cos sin)0ND G i S N O P i
23、L i i j ij ij ij ijjQ Q Q U UG B=+=(3)Where,DG iP,SNOP iP and,LiP denote active power injection from the DG,SNOP and load at node i,While,DG iQ,SNOP iQ and,LiQ denote reactive power injection.iU and jU are the voltage amplitude of nodes i,j.,ij is the phase between the connected nodes i,j.,ijG and,i
24、jB are mutual conductance and mutual susceptance.This paper considers DG at maximum output,which is equal to the installed capacity.Therefore,the,DG iP in equation(2)is the same as,DG iE in equation(1).maxCC(4)min max iU UU(5)In the formula,maxC is the maximum capacity of the feeder.minU andmaxU are
25、 the maximum and minimum voltage limits.DG,DG,max iEE(6)WhereDG,iE is the DG capacity of node i in the network,and DG,maxE is the maximum DG penetration capacity of one node at a specific voltage level.,0SNOP i SNOP jPP+=(7)22,SNOP i SNOP i SNOP iPQ S+(8)22,SNOP j SNOP j SNOP jPQ S+(9)Where,SNOP iP,
26、SNOP jP,SNOP iQ,SNOP jQ show the power of SNOP connected to the two nodes i and j respectively,they denote the active and reactive power of the converters at both ends.And,SNOP iS and,SNOP jS are the apparent capacity of the two converters respectively.III.SOLUTION METHOD OF THE MODEL A multi-popula
27、tion genetic algorithm(MPGA)is used in this paper to solve the model.In MPGA,2018 China International Conference on Electricity Distribution Tianjin,17-19 Sep.2018CICED2018 Paper No.201805090000012 Page2/5214925 26 maxCC=5 6 maxCC=22 23 maxCC=multiple populations are adopted for searching.Linked by
28、genetic operators,it can achieve co-evolution of multiple populations.Compared with the standard GA,MPGA uses multiple populations to simultaneously search the set of solutions,taking into account the global and local searching ability of the algorithm.The sensitivity of the calculation related to t
29、he genetic control parameters is greatly reduced,and immature convergence is overcome obviously.So it is suitable for high-dimensional nonlinear optimization problem this paper put forward.When using MPGA to solve the model,the fitness of MPGA is the sum of the DG in the whole network in Equation(1)
30、.For different situations,the solution can be solved by setting different variables of MPGA.IV.CASE STUDY A.Case Data This paper takes a IEEE 33-bus distribution network for analysis.0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1718 19 20 2122 23 2425 26 27 28 29 30 31 32TS1TS3TS4TS5TS2Fig.2 IEEE 33-bus
31、 distribution system In the figure,TS1TS5 are five contact switches,all of which are disconnected.The networks reference voltage is 12.66kV,and the feeders rated load is 5.28MW.The basic load is 3715+j2290 kVA.The maximum and minimum voltage limits are set to 1.07 p.u.and 0.93 p.u.According to a Chi
32、nese general planning and design technical guideline of distribution network,the accessing DG capacity of one node to the 10kV network should not exceed 7 8MW,thus this paper sets 7MW.DG is set to PQ type with a power factor of 0.95.B.The calculation of the allowed capacity of single-point DG in dis
33、tribution network without SNOP In order to obtain the maximum DG penetration of each node,position of DG is constant and the capacity is variable,while parameters of SNOP are set 0.Table 1 Allowable DG capacity of different node Node Max capacity/kW Bottleneck factor Node Max capacity/kW Bottleneck
34、factor 1 Invalid 17 1854 17 maxUU=2 7000 2 maxEE=18 Invalid-3 7000 3 maxEE=19 Invalid-4 7000 4 maxEE=20 Invalid-5 7000 5 maxEE=21 Invalid-6 6187 22 6083 2 22 maxCC=7 5618 7 maxUU=23 6069 8 4419 8 maxUU=24 4901 24 maxUU=9 3673 9 maxUU=25 6155 5 25 maxCC=10 3574 10 maxUU=26 6105 11 3399 11 maxUU=27 49
35、49 27 maxUU=12 2798 12 maxUU=28 4261 28 maxUU=13 2606 13 maxUU=29 3946 29 maxUU=14 2439 14 maxUU=30 3359 30 maxUU=15 2263 15 maxUU=31 3198 31 maxUU=16 1971 16 maxUU=32 3011 32 maxUU=The DG penetration capacity of different nodes can be divided into four cases:1.DG access location is 1,1821.With DG a
36、ccess,the voltage of nodes 817 and 2832 is still below the minimum voltage 0.93 p.u.,so it is invalid.2.DG access location is 717,2732.These nodes are located at the end of the feeder and the allowable DG penetration level is low.3.DG access location is 6,22,23,25,26.These nodes are located at the f
37、ront of each feeder branch,and the allowable DG penetration level is high.4.DG access location is 25.These nodes are located at the front of the entire 33-bus distribution network,and the allowable DG penetration level is higher.Among the above four cases,nodes 18,30,22,and 5 are respectively select
38、ed for DG access with their highest DG penetration level(node 18 takes 7000 kW),and the corresponding network voltage is as follows:0 5 10 15 20 25 300.920.940.960.9811.021.041.06NodeVoltage Without DGDG in node 5DG in node 18DG in node 22DG in node 30 Fig.3 Node voltage of distribution system with
39、DG According to Table 1 and Figure 3,the conclusions are as follows:2018 China International Conference on Electricity Distribution Tianjin,17-19 Sep.2018CICED2018 Paper No.201805090000012 Page3/52150(1)DG access will change the voltage of the distribution network,causing the voltage of the access n
40、ode to increase,so are the nodes close to it and this effect will gradually decrease as the distance from the access node increases.(2)The closer the access node is to the end of the long feeder,the lower the DG penetration level is.This is because the node voltage in the end of long feeder is more
41、sensitive to DG than other nodes,and reactive power of DG mainly affects the access node,making it easier to exceed the voltage limit.(3)As the access location moves to the head of the feeder,the DG penetration capacity increases.This is because the reactive power from DG can spread throughout the n
42、etwork,resulting in reducing voltage sensitivity of the access node.And feeder power limitation becomes the main restriction.(4)When the access location is at the head of the network,the DG penetration level gets further promotion,and the main factor that hinder the DG penetration level becomes the
43、allowed capacity of DG in single point at a certain voltage.C.Promotion impact of SNOP to the DG penetration level of the specified node Among the above four cases,DG penetration level in case 1 and 2 has large promotion space.The SNOP is introduced to distribution network to form the FDN,and the SN
44、OP1SNOP5 are located in the position of TS1TS5 respectively.Examine the enhancement to DG penetration level by SNOP in different position in three cases.These are DG in the single node(18/30),multi-node of single-feeder(14,17)and multi-node of multi-feeder(17,32).The SNOP is set to 1 MVA.The results
45、 are shown in Table 2 to Table 4.1P is the active power of SNOP in the small-number node,1Q is the reactive power of SNOP in the small-number node.When SNOP sends active power to the head and absorb from the end node,the P value is positive.When SNOP absorbs reactive power,the Q value is positive.Ta
46、ble 2 Promotion of SNOP to the allowable DG penetration level in single bus case Scheme DG location DG penetration without SNOP/kW SNOP DG penetration with SNOP/kW SNOP output 1P/kW 1Q/kVar 2Q/kVar 1 30 3359 SNOP1 4367 302 792 953 SNOP2 3733-213 757 762 SNOP3 3743-770 534-347 SNOP4 3700-994 105 74 S
47、NOP5 4072 475 586 849 2 18-SNOP1-SNOP2-SNOP3 6378 963 267 192 SNOP4 6250 894 249 348 SNOP5-Table 3 Promotion of SNOP to the allowable DG penetration level in single-bus/multi-branch case DG penetration without SNOP/kW SNOP DG penetration with SNOP/kW SNOP output Sum 14 17 sum 14 17 1P/kW 1Q/kVar 2Q/
48、kVar 2439 2439 0 SNOP1 3440 2172 1268-461 879 887 SNOP2 3512 3455 57 0 989 1000 SNOP3 3056 2983 74-375 912 924 SNOP4 2782 2761 21-289 957 943 SNOP5 2681 2670 11 290 956 886 Table 4 Promotion of SNOP to the allowable DG penetration level in multi-bus/multi-branch case the DG penetration without SNOP/
49、kW SNOP the DG penetration with SNOP/kW SNOP output Sum 17 32 Sum 17 32 1P/kW 1Q/kVar 2Q/kVar 3906 1342 2564 SNOP1 5364 2083 3281 0 1000 1000 SNOP2 4853 2091 2762 0 1000 1000 SNOP3 4357 1706 2651-135 991 956 SNOP4 4234 1412 2822-367 930 468 SNOP5 4490 1403 3087 101 983 994 2018 China International C
50、onference on Electricity Distribution Tianjin,17-19 Sep.2018CICED2018 Paper No.201805090000012 Page4/52151 It can be seen from the results that SNOPs capability to control power flow can improve the DG penetration level.Different DG and SNOP access modes reflect the following different characteristi
