Research on improving power supply capacity of urban distribution network based on load reorganization

Research on improving power supply capacity of urban distribution network based on load reorganization

Jiang Jian1, Lou Jian1, Wu Shunyu1, Liu Haiqiong2

(1. Hangzhou Power Supply Company, State Grid Zhejiang Electric Power Co., Ltd., Hangzhou 310009; 2. Hangzhou Worui Electric Power Technology Co., Ltd., Hangzhou 310012)

Citation information for this article: Jiang Jian, Lou Jian, Wu Shunyu, Liu Haiqiong. Research on improving the power supply capacity of urban distribution network based on load reorganization[J]. Zhejiang Electric Power, 2020, 39(01): 9-15.

0 Introduction

As the end of the entire power network, the distribution network is most closely connected with users and is a very important part of the power network. With the rapid development of social economy and urban scale, the growth rate of urban electricity consumption has significantly accelerated, the demand for access to various business expansion projects has continued to rise, and the problem of insufficient available power supply capacity of the distribution network has become increasingly serious. In order to meet the needs of urban development, the transformation and expansion of the power grid has been fully launched, but the land use in urban built-up areas is tight, the land cost of substations and feeders is high, and the resources of sites and line channels in some central areas are already very tight. It has become very difficult to obtain the site of a new substation and the underground channel for a new feeder [1-2].

At present, domestic and foreign scholars have conducted various studies on improving the power supply capacity of urban distribution networks, but they usually improve the distribution network structure from the perspective of distribution network planning, and the measures taken are usually to build or change the interconnection switch, build new lines, etc. Reference [3] proposed a multi-objective optimization model for the interconnection structure between substation main transformers based on power supply capacity, and used genetic algorithm to solve it; Reference [4] established a distribution network line optimization planning model based on maximum power supply capacity, and explored the potential of distribution network equipment by means of line switching, installation of switches and new interconnection lines; Reference [5] proposed to change the position and state of the interconnection switch in the interconnection line to switch the existing load line to improve the power supply capacity when the substation and network structure are determined. Although the above methods meet the requirements of the distribution network for line load rate margin, they aggravate the problem of low utilization rate of existing power grid equipment and large amount of network management resources being occupied during the low power consumption period, and at the same time cause the phenomenon of low return rate and waste of power grid construction investment.

To this end, under the grid conditions of the existing distribution network, without building new substations and feeders, this paper adopts a technical route of global optimization and reorganization of the load distribution points of the busbars and intervals on the line, and proposes a big data analysis and calculation strategy for load peak shifting, complementarity, and balanced optimization, to tap the power supply capacity of the existing power grid, reduce the peak and volatility of line loads, and improve the available power supply capacity of the distribution network to absorb more loads.

1. Potential tapping ideas

Urban power loads can usually be divided into the following typical loads: residential loads, office building loads, commercial loads, industrial loads, school loads, etc. Each type of load has different changing patterns, and their typical daily load curves and load forms vary greatly. For example, commercial loads are affected by people's consumption and living habits, with double peaks at noon and evening, and a load trough around 4:00 in the morning; while office building loads are affected by working hours, with a full-day load peak at 9:00 and a significant load reduction around 17:00 [6-7].

In addition to the inconsistent peak and valley values ​​in the daily load curves of different types of users, the daily peak and valley differences and volatility are also different. In actual distribution network lines, many different types of users are connected, and the proportions of each type of user are also different, which makes the daily load characteristic curves of the lines different. If the load composition on the line is changed, the daily load curve of the line will also change accordingly.

As shown in Figure 1, a ring network includes line 1 and line 2. The sub-loads of line 1 include load 1-1, load 1-2, and load 1-3, and the sub-loads of line 2 include load 2-1, load 2-2, and load 2-3.

Figure 1 Schematic diagram of tapping the potential of line power supply capacity

Under the initial connection mode, the historical daily load curves of lines 1 and 2 fluctuate greatly, and the peak-to-valley values ​​differ greatly. Considering the differences in the types of user loads connected to each sub-load, the load peaks, the time of peak power consumption, and the volatility of the loads of each sub-load are different. For example, the peak times of loads 1-2 and 2-2 are staggered, and when load 1-2 is in the peak period, load 2-2 is in the valley period. Therefore, the loads on lines 1 and 2 are reorganized, and loads 1-2 on line 1 are interchanged with loads 2-2 on line 2, that is, the connection mode after reorganization and optimization becomes: line 1 is connected to loads 1-1, 2-2, and 1-3, and line 2 is connected to loads 2-1, 1-2, and 2-3. As shown in Figure 1, after the reorganization and optimization, the fluctuation amplitude of the load curves on lines 1 and 2 becomes smaller, and at the same time, the load peak of line 1 is reduced by P1, and the load peak of line 2 is reduced by P2, so more loads can be connected, and the power supply capacity is improved.

2 Potential tapping model

In order to comprehensively judge the rationality of the optimization scheme, the comprehensive evaluation index V is established by considering factors such as line load peak, load balance and optimization cost in the optimization scheme:

in:

In the formula: Pmax is the peak drop of the optimized line; pagenumber_ebook=15,pagenumber_book=11 is the normalized value of Pmax; Pvar is the line load balance; pagenumber_ebook=15,pagenumber_book=11 is the normalized value of Pvar; C is the implementation cost of the optimized plan; C* is the normalized value of C; H is the number of data points; μi is the average load value at the i-th data point; pi is the load value at the i-th data point; α, β, γ are the weight coefficients of each indicator respectively.

The normalization function used is:

In the formula: x is the quantity to be normalized; x* is the quantity after normalization; xmax and xmin are the maximum and minimum values ​​respectively.

3 Ways to tap potential

In order to ensure the reliability of power supply in the distribution network, this potential tapping method only optimizes the distribution network ring network lines, selects the lines that need to be tapped by using the line screening method, and proposes two load reorganization methods: busbar optimization and interval optimization, combined with the actual engineering transformation volume.

3.1 Line Filtering

Line screening is based on the following two conditions, and potential tapping calculations are performed only if the conditions are met.

3.1.1 Equipment conditions

Equipment condition determination is to screen the power supply path transformation conditions. It is determined that different power supply lines must have more than 60% of the same switch stations at the same time, and the lines are all cables or overhead lines on the same pole to meet the optimization conditions, that is:

Where: Ωa_s, Ωb_s are the sets of switch stations passed by path a and path b respectively; N is the number of switch stations passed by the power supply path set; Ωa_l, Ωb_l are the line sets of path a and path b respectively; Ωl is the set of all lines; Ωcab is the set of cable lines; Ωove is the set of double-circuit overhead lines on the same pole.

3.1.2 Characteristic conditions

The load characteristic condition refers to whether the average load rate, peak shifting degree and peak difference of the optimized path objects a and b meet the conditions for opening capacity potential. When the optimized object meets the equipment judgment condition, it is further judged whether it meets one of the following three load characteristic conditions.

(1) During the peak period of electricity consumption, the average line load rate pagenumber_ebook=15,pagenumber_book=11 is greater than 65% (the line overload standard is 80%).

Where: pt is the line power value at time t; T is the time length; pmax is the rated capacity of the line.

(2) For lines with obvious peak shifting between different feeders, if the peak time difference Δtab between different lines is greater than 2 h, it is judged that they have peak shifting.

Where: ta_max is the peak time point of path a; tb_max is the peak time point of path b.

(3) The load peak difference Δpab between the optimization objects is greater than 30%.

Where: pa_max, pb_max are the maximum power values ​​of path a and path b respectively; pa_rat, pb_rat are the rated capacity of the line.

3.2 Load reorganization method

To reorganize the line load, it is necessary to consider factors such as the implementation workload and power outage cost of the reorganization plan in the actual engineering transformation. Therefore, in order to minimize the implementation workload and avoid power outages, this paper selects two reorganization methods, busbar optimization and interval optimization, for optimization.