Planning Method of Low-Voltage Distribution Network Applicable to the Access of Multiple Flexible Entities

Since the establishment of the goal of “peak carbon and carbon neutrality”, various types of emerging loads have been developing rapidly. However, the application scenarios and working principles of various loads are different, and there is no clear standard for distribution network requirements. Thus, the existing low-voltage distribution network planning methods do not take emerging loads into account. Therefore, this paper proposes a low-voltage distribution network planning method that is applicable to the access of multiple flexible subjects. Firstly, the operational characteristics of flexible resources on the power side, user side, and energy storage side are analyzed and studied. Secondly, the low-voltage distribution network planning method applicable to the access of multiple flexible subjects is studied. Finally, the low-voltage distribution network of Song Yang County, Lishui City, is selected for the empirical application.


Introduction
With the in-depth development of electric energy substitution, emerging clean energy supply stations and customer-side demand response loads have put forward new demands for the reliability of power supply in distribution networks [1] .However, low-voltage distribution network planning is less involved in the access of flexible resources such as distributed photovoltaic and energy storage under the new power system [2] .In the context of the new power system, considering that the original planning methods are no longer adapted to the system planning requirements in the new environment, it is necessary to construct a planning model study for the new distribution network applicable to the access of multiple flexible subjects.
The distribution network planning method focuses on achieving optimal siting and equipment models for various system components (substations, feeders, transformers, circuit breakers, and capacitors) and ensuring the safety and reliability of power supply operation [3] .In the power planning model, Liu et al. and Li et al. [4][5] considered the substation siting and capacity determination from an economic point of view under the influence of substation load factor and supply radius, intending to minimize the initial investment and operate costs of low-voltage lines and substations.In the grid planning model, the traditional grid model is mainly aimed at the economy [6][7][8] .When all load nodes are considered at once under the constraints of power balance constraint, node voltage constraint, and network connectivity, the grid planning model suffers from too many decision variables that lead to the high complexity of the solution problem; hence, many papers carried out power supply partitioning for distribution networks to reduce the solution difficulty [9][10] .
In this paper, we deeply analyze the operational characteristics of multiple flexible resources and analysis of the current situation of traditional distribution network planning, flexible resource prediction, and full-time sequence power balancing.We propose a low-voltage distribution network planning method applicable to the access of multiple flexible subjects to achieve standardized, differentiated, and lean planning of distribution networks.It provides a reference basis for the planning preparation considering the integration of source-grid-load-storage during the construction of new distribution systems.

Analysis of current problems of low-voltage distribution network planning under diversified flexible subjects
Facing the access of multiple flexible entities, the current problems of traditional distribution system planning mainly focus on load forecasting and power balance analysis.
First, the access to multiple flexible entities makes the load forecasting of the distribution network more difficult.Since distributed power sources with random nature will provide most of the electricity for users, it greatly increases the difficulty for distribution network planners to make load predictions and reactive power optimization.In the process of power balancing, not only traditional loads need to be considered, but also the impact of new power sources, such as new energy sources and storage, mainly scenery.Flexible and adjustable resources also need to be considered, such as new energy vehicle charging, 5G loads, virtual power plants, demand-side response, etc., on the final grid power consumption.
Secondly, distribution network planning usually assumes that the grid load grows year by year and does not consider the impact of multiple flexible entities.When many flexible entities are connected, the previous planning may not be able to meet the requirements.
The analysis of the current problems shows that traditional low-voltage distribution network planning can no longer meet the needs of planning under multiple emerging loads.Therefore, this paper proposes a low-voltage distribution network planning method applicable to the access of multiple flexible subjects.Based on the analysis of multiple flexible resource characteristic curves to perform power balancing calculations, it solves the real problem of missing power balancing methods for multiple load characteristic curves at the low-voltage distribution network level.

Power-side flexible resource operation characteristics
Low-voltage power side balancing resources mainly include small hydropower and photovoltaic power sources.Among them, renewable energy equipment such as hydropower, photovoltaic power, and subwind power are more volatile due to their generation output being affected by objective weather, wind/light intensity, water flow, etc.
Affected by the peak-to-valley price difference in power generation, hydropower tends to concentrate its power generation during the day and cause backward delivery peaks.From the analysis of daily output characteristics, the daily output of hydropower plants is smaller in the morning before 7:00 a.m. and larger from 7:00 to 9:00 p.m., which correspond to the lower and higher periods of system load, respectively.This is shown in Figure 1(a).
PV output varies with natural factors such as light intensity, weather, season, and temperature, with random fluctuations.PV output is mostly concentrated in the daytime, especially during the midday hours, while the night output is zero, showing significant temporality.This is shown in Figure 1  Electric vehicles participate in the dispatch of the grid in the form of charging and switching stations and are transferable loads.The coefficient is about 0.2 to 0.6.Industrial high energy load is divided into two types, interruptible load, and shiftable load.The flexible resource coefficients are 0.3 to 0.5 and 0.6 to 0.8.For large commercial users with large capacity and concentrated power consumption time, the flexible resource factor of commercial users is about 0.6 to 0.8.About 60% of loads of residential users can be used as flexible loads, and the flexible resource factor for residential customers is about 0.5 to 0.7.

Energy storage side balancing resources out of operating characteristics
The main feature of energy storage-type devices is that they can be used for both supply-side generation and demand-side consumption during operation.Pumped storage and battery storage are two forms commonly used for grid demand response.
Pumped storage efficiency is generally about 75%, with daily regulation capability, and can be applied to peak regulation and standby.Its peak-shaving flexibility is 100%, and its valley-filling flexibility is 75%.The flexibility of battery storage is affected by the performance of the battery, which should be combined with the type of battery and battery performance parameters for an in-depth analysis of battery flexibility.

Research on the planning method of low-voltage distribution network adapted to the access of multiple flexible entities
Distribution network planning has the contents of current situation analysis, flexible resource prediction, full-time sequence power balance, etc.They follow a certain sequence and have certain connections.Combining the steps of distribution planning and low-voltage line planning, the process of low-voltage distribution network planning is obtained, as shown in Figure 2. The flexible resource characteristic curve-based power balancing method is based on load forecasting.The calculated load is obtained after considering all aspects of flexible resources, such as source network load and storage, to participate in power balancing.Grid planning is carried out on this basis.The key to the method is the emerging load forecast and flexible resource overlay.The specific equation is as follows: where P is the distribution capacity, P 1 is the net load after balancing, P 2 is the conventional utility distribution load, P 3 is the electric vehicle load, P 4 is the 5G base station load, P 5 is the wind power output load, P 6 is the PV output load, and P 7 is the hydropower output load.

Empirical applications
The grid of Yuying Road, Xiping Sub-district, Songyang County, Lishui City is selected for empirical application.

Analysis of the current situation
The current situation of Yuying Grid has a PV power plant with a total installed capacity of 1.66 MW at the farmers' market, which is connected to line 186.There are 38 distribution substations in the grid with a total distribution capacity of 22.26 MW.The average maximum load ratio of public distribution substations is 25.58%, the overall power supply capacity is sufficient, and there are no heavy overload distribution substations in the grid.

Load forecasting and power balancing (1) Load forecast
According to the natural growth load data of Songyang from 2016 to 2020, the natural growth rate is predicted to be 10.67% during the "14th Five-Year" Plan period.In 2025, the natural growth load of Yuying Grid will be 14.88 MW, and the maximum load of public distribution is 7.12 MW.By using the load density index method to carry out the vision load distribution of Yuying Grid, the forecasted load distribution of Yuying Grid is 21.82 MW, and the utility distribution load is expected to be 10.5 MW.By predicting and superimposing the multiple loads in Yuying Grid, the utility distribution load of the Yuying Grid will reach 7.3 MW by 2025.The utility distribution load of Yuying Grid will reach 10.78 MW by the year of saturation, and the load forecast results are shown in Table 1.(3) Power balance analysis Based on the original grid planning results and load forecasting results considering flexible resources for power balance calculation, the calculation result is shown in Table 2.As the maximum load of urban public distribution occurs at night, the distributed PV output has no reduction effect on the maximum load.Still, the overall construction of urban low-voltage distribution networks is more mature.The distribution load ratio is relatively low, from the overall capacity-load ratio to the vision of the Yuying grid without new distribution.

Low-voltage distribution network planning scheme
Combining the current situation of the low-voltage distribution network station area of Yuying Grid and the development demand for flexible resources, it is expected that by 2025 the urban area will be transformed with 4 distribution transformers, with a capacity of 1900 kVA; It is expected that by the saturation year, 4 new and renovated distribution transformers with a capacity of 3260 kVA will be built.The details are shown in Table 3.

Conclusion
The low-voltage distribution network planning method applicable to the access of multiple flexible subjects proposed in this paper solves the real problem of the lack of power balancing method of the multi-load characteristic curve of low-voltage distribution network layers in the context of the new power system.Breakthroughs in the calculation and the low-voltage distribution network frame structure and planning methods under different typical scenarios provide technical support for the power grid companies to realize an accurate investment in distribution networks so as to avoid blind investment.This paper is based on the development requirements of the new power system.This paper provides scientific guidance for low-voltage distribution network planning based on the typical planning model of a low-voltage distribution network required by the development of a new power system. (b).

Figure 1 .
Typical daily output curve of hydropower and photovoltaic3.2User-side flexible resource operation featuresAs an important resource for demand response, user flexible resources are divided into four main categories: electric vehicle flexible resources, industrial user flexible resources, commercial user flexible resources, and residential user flexible resources.

Figure 2 .
Figure 2. Low-voltage distribution network planning flow chart

Table 1 .
Results of multivariate load prediction for the Yuying grid Flexible resource forecast According to the distributed PV planning of Songyang County, there is no low-voltage distributed PV connected to the grid in Yuying Grid in the current year.The installed low-voltage distributed PV is expected to be 0.4 MW by 2025 and 2.1 MW by the vision year.

Table 2 .
Yuying grid utility distribution power balance sheet

Table 3 .
Yuying grid low voltage distribution network project inventory