Research on the impact of electric vehicle charging on residential area distribution network

With the continuous development and progress of electric vehicle related technologies, electric vehicles will surely achieve large-scale development  . However, a large number of electric vehicle charging will have an impact on distribution network equipment, thereby affecting the safety and economic operation of the distribution network  . The impact of electric vehicles on the distribution network is mainly achieved through charging load. However, the cluster load characteristics of a large number of electric vehicles have a large randomness in time and space, resulting in complex charging load characteristics. Therefore, conducting an analysis of the impact of electric vehicle charging on the distribution network and studying demand-side management strategies aimed at reducing negative impacts is of great significance to maintaining the safe and stable operation of the power grid.

This paper takes urban residential communities as the research object, establishes a Monte Carlo calculation model for the charging power of electric vehicles in residential communities based on the charging behavior of users on weekdays and holidays, predicts the charging load of electric vehicles with different penetration rates, and combines the original load characteristics of residents in typical communities to analyze the impact of electric vehicle charging on community load characteristics and distribution transformer load rate, as well as the community’s ability to accept electric vehicle charging. An effective demand-side management strategy is proposed to guide the benign interaction between user charging behavior and the safety and reliability of the system, as well as the need for peak shaving and valley filling, so as to improve its impact on the friendliness of the power grid and weaken its adverse effects.

1 Electric vehicle charging load prediction

1.1 Factors affecting electric vehicle charging load

1.1.1 Electric vehicle charging mode

Electric vehicle charging modes are mainly divided into fast charging and slow charging  .

1) Fast charging can provide short-term charging services for electric vehicles. The charging time is about 20 minutes to 2 hours, and the charging current is about 150 to 400 A. It is generally carried out at electric vehicle charging stations or large parking lots with conditions.

2) Slow charging uses a small amount of AC power to charge, generally using a single-phase 20 V/32 A AC power supply (7 kW). For private passenger cars, the charging time is usually 5 to 8 hours. Slow charging of electric vehicles is generally carried out through charging piles, which is mainly suitable for residential areas or some working areas with conditions.

1.1.2 User Charging Behavior

The driving routes of private cars are relatively flexible. The main use of private cars on weekdays is to travel to and from get off work, and the main use of private cars on holidays is leisure and entertainment, visiting relatives and friends, etc.

SOC (State of Charge) is used to represent the state of charge of a battery, also known as the remaining capacity. It represents the ratio of the remaining capacity of a battery to its capacity when it is fully charged. SOC is determined by the charge state and mileage of the electric vehicle at the time of departure, and its value range is 0-100%  . According to a survey, the annual mileage of Chinese cars is (1.5×104-2 × 104 ) km, and the average daily mileage is about 40 km. Based on the actual endurance of electric vehicle batteries in the market, users need to charge once a day on average, and the charging start SOC satisfies the normal distribution N (0.4, 0.12 ) .

The charging mode of users under natural conditions is defined as disordered charging. According to user travel habits, on weekdays, about 80% of private electric vehicles start charging in community parking lots between 18:00 and 24:00, which follows the normal distribution of N(19, 2.5 2 ), and the remaining 20% of electric vehicles follow the uniform distribution of U(8, 17); on holidays, 70% of private electric vehicles need to be charged, and about 40% of electric vehicles start charging in community parking lots between 9:00 and 16:00, which follows the uniform distribution of U(9, 16), and 60% start charging between 17:00 and 22:00, which follows the uniform distribution of U(17, 22)  . According to the corresponding time periods of private electric vehicle charging behavior and charging type and the above assumptions, the parameter table required for establishing the electric vehicle charging load model can be obtained, as shown in Table 1 .

 

Tab. 1 Charging characteristics of private cars

scene	Charging pile utilization rate/%	Initial charging time	Charging probability	Initial SOC distribution	Start time distribution
Working Days	100	8:00～17:00	0.2	N (0.4, 0.22 2 )	U (8, 17)
	100	18:00～24:00	0.8	N (0.4, 0.22 2 )	N (19, 2.52 2 )
Holidays	70	9:00～16:00	0.4	N (0.4, 0.22 2 )	U (9, 16)
	70	17:00～22:00	0.6	N (0.4, 0.22 2 )	U (17, 22)

1.2 Electric Vehicle Charging Load Modeling

In order to ensure the accuracy of the calculation results, a day is divided into 1440 minutes, and the charging power of a single electric vehicle is calculated every minute to obtain the charging load curve of a single vehicle. The charging load curve of each electric vehicle is superimposed to obtain the charging load curve of all electric vehicles for 24 hours a day. The program is run in the MTALAB environment, and 10,000 Monte Carlo simulations are performed to generate the charging load curve of electric vehicles. Due to the different charging behaviors of users on weekdays and holidays, there are obvious differences in the charging curves of electric vehicles in residential areas on weekdays and holidays.

Assuming that there are 1,000 households in a residential area, each of which owns a private car, and considering that the penetration rate of electric vehicles is 10% to 60%, the charging load of private electric vehicles on weekdays and holidays is predicted.

The charging load curves of electric vehicles in residential areas on weekdays and holidays are shown in Figures 1 and 2 , and the maximum charging load is shown in Table 2. It can be seen that the penetration rate of electric vehicles and residents’ work and rest time have a greater impact on the charging load of electric vehicles. In general, the charging load in residential areas on weekdays is greater than that on holidays. The peak period of electric vehicle charging load on weekdays is mainly concentrated between 19:00 and 22:00 after work. The main reason is that most car owners choose to charge their electric vehicles immediately after getting home from work, causing a sharp increase in load; there are two peaks in the charging load of electric vehicles on holidays, among which

Fig. 1 EVs charging load curve on weekday

Fig. 2 EVs charging load curve on weekend

 

Tab. 2 EVs charging load kW

Permeability/%	10	20	30	40	50
Residential area work	350	690	1 100	1 360	1 700
Residential District Holidays	210	410	610	820	1 020

The charging load is higher during the evening peak, mainly concentrated between 19:00 and 21:00. There are fewer vehicles charging on holidays than on weekdays, and users’ charging time is more dispersed, making the peak charging load on holidays lower than that on weekdays. Whether it is weekdays or holidays, the charging load of electric vehicles has obvious peak-to-valley differences, and it coincides with the peak period of electricity consumption of community residents.

2 Impact of electric vehicle charging on community distribution networks

2.1 Selection of Typical Residential Areas

According to the different single-family area and income level of the residents in the community, it is divided into three typical communities: ordinary community, mid-range community and high-end community [ 9 ] . Assuming that the number of households in ordinary, mid-range and high-end communities is 1,000, the maximum load on weekdays and holidays is calculated by combining the load density index corresponding to different types of communities [ 10 ] , and then the total capacity of the community distribution transformer is obtained. The load characteristic curve is obtained by investigating communities of the same type, as shown in Table 3 .

 

Tab. 3 Daily maximum load and distribution transformer capacity selection

Cell Type	Ordinary community	Mid-range residential area	High-end residential area
Number of households	1 000	1 000	1 000
Calculated load/kW	4.600	6.000	10.00
Demand factor	0.600	0.600	0.600
Load planning margin	0.700	0.650	0.600
Power Factor	0.850	0.850	0.850
Simultaneity rate	0.400	0.400	0.400
Maximum load on working days/kW	1 104	1 440	2 400
Maximum load on holidays/kW	1 187	1 636	2 243
Reference capacity of distribution transformer/kVA	1 855	2 606	4 706
Actual capacity of distribution transformer/kVA	2 000	3 000	4 800

2.2 Impact of EV Charging on Load Characteristics

The charging load when the electric vehicle penetration rate is 10% to 60% is superimposed on the original load curves of the community on weekdays and holidays, and the load curves of three typical communities on weekdays and holidays that include electric vehicle charging loads are obtained as shown in Figures 3 , 4 , 5 , 6 , 7 , and 8 .

Fig. 3 Ordinary-level community load curve on weekday

Fig. 4 Mid-level community load curve on weekday

Fig. 5 High-level community load curve on weekday

Fig. 6 Ordinary-level community load curve on weekend

Fig. 7 Mid-level community load curve on weekend

Fig. 8 High-level community load curve on weekend

It can be seen that a lower penetration rate will affect the load characteristics of ordinary communities; in medium and high-end communities, due to the large original load base, the proportion of electric vehicle charging load to the original load is relatively small, and the impact on load characteristics is relatively small. The charging time of electric vehicles is relatively concentrated on weekdays. When the electric vehicle penetration rate is 10% to 60%, the maximum load of ordinary, medium-range, and high-end communities increases by 21% to 167%, 16% to 127%, and 10% to 67% respectively compared with the original residential electricity load; the charging time of electric vehicles on holidays is relatively scattered. Compared with the original residential electricity load, the maximum load of ordinary, medium-range, and high-end communities increases by 9% to 94%, 5% to 59%, and 4% to 29% respectively.

On weekdays, the peak-to-valley difference rates of ordinary, mid-range, and high-end communities increased from 0.46, 0.46, and 0.57 to 0.55-0.8, 0.51-0.73, and 0.6-0.71, respectively; on holidays, the peak-to-valley difference rates of ordinary, mid-range, and high-end communities increased from 0.38, 0.43, and 0.37 to 0.42-0.68, 0.45-0.64, and 0.49-0.58, respectively. It can be seen that the peak of electric vehicle charging coincides with the peak of residents’ original electricity consumption, which makes the peak-to-valley difference rates of the three types of communities on weekdays and holidays increase to varying degrees with the increase of electric vehicle penetration rate. The increase in the peak-to-valley difference rate increases the peak-shaving pressure of the system, and the system needs to provide a larger spare capacity, resulting in a waste of resources.

2.3 Impact on distribution transformer load factor

According to the calculation results in Table 3 , the total capacity of distribution transformers in ordinary, mid-range and high-end communities is 2 MVA, 3 MVA and 4.8 MVA respectively. When the penetration rate of electric vehicles is 10% to 60% respectively, in order to prevent the distribution transformers in the community from being overloaded on weekdays and holidays, the maximum penetration rate of electric vehicles accepted by ordinary, mid-range and high-end communities is shown in Table 4. It can be seen that the load rate of distribution transformers increases with the increase of electric vehicle penetration rate. The charging time of electric vehicles on weekdays is relatively concentrated, and it overlaps with the original load time of residents during the evening peak, which increases the power supply pressure of distribution transformers; the charging time of electric vehicles on holidays is relatively dispersed. Therefore, compared with weekdays, the charging load of electric vehicles on holidays has less impact on the power supply pressure of community distribution transformers. In addition, after electric vehicles are connected to the distribution network, the overload of distribution transformers is closely related to the economic positioning of the community [ 11 ] . When the penetration rate of electric vehicles is the same, the load rate of distribution transformers in high-end and mid-range communities is significantly lower than that in ordinary communities. This is mainly because high-end and mid-range communities have reserved higher power load margins for distribution transformers during planning. The capacity and capacity margin of distribution transformers in ordinary communities are smaller than those in mid-range and high-end communities. As the penetration rate of electric vehicles increases, the power supply pressure increases during peak load periods.

Table 4 Maximum penetration rate of electric vehicles accepted by community distribution transformers

Tab. 4 Acceptance of electric vehicle penetration maximum in district distribution transformer %

scene	Ordinary community	Mid-range residential area	High-end residential area
Working Days	≤20	≤50	≤60
Holidays	≤40	≤60	≤60

3 Demand-side management strategy

A large number of electric vehicles are charged at the same time during the peak period of the original load of the power grid (especially the evening peak), resulting in a “peak on peak” load curve, which exceeds the power supply capacity of the existing distribution equipment. If the existing distribution equipment is increased in capacity and modified for this reason, more distribution equipment will run at low load during valley hours, greatly reducing the utilization efficiency of the distribution equipment.

Therefore, it is necessary to effectively guide or control the charging behavior of users. On the basis of meeting the charging needs of electric vehicles, it is necessary to consider using the peak-valley electricity price strategy to guide electric vehicles to charge in an orderly manner, avoid the peak period of the original load of the power grid, and “shift the peak to fill the valley” of the power grid load, thereby delaying the capacity expansion and transformation of distribution network equipment, and achieving coordinated development of electric vehicles and the power grid [ 12 ] .

3.1 Electric vehicle orderly charging load curve

Electricity is the main driving cost of electric vehicles. Users can effectively reduce electricity expenses by adjusting the charging time, so users will charge as much as possible during off-peak hours. Assuming the valley price range is 1:00-8:00, after adopting the peak-valley electricity price strategy, 80% of users charge during off-peak hours, and 20% of users are not affected by electricity prices. When the penetration rate of electric vehicles is 10%-60%, the load forecast for electric vehicle charging on weekdays and holidays is carried out.

The charging load curves of electric vehicles in residential areas on weekdays and holidays are shown in Figures 9 and 10 , and the maximum charging load is shown in Table 5. It can be seen that the peak periods of electric vehicle charging load on weekdays and holidays are mainly concentrated in the valley price period of 4:00-8:00 and 19:00-21:00 in the evening. The main reason for these two peaks is that most users choose to charge at the valley price and a small number of users who are not affected by the peak and valley price choose to charge their electric vehicles directly after going out and returning home, resulting in an increase in charging load. There are fewer charging vehicles on holidays than on weekdays, and the charging time of users is more dispersed, making the maximum charging load on holidays lower than that on weekdays. Compared with disordered charging, the peak and valley electricity price strategy reduces the maximum charging load on weekdays and holidays by about 25% and 17% respectively.

Fig. 9 EVs charging load curve on weekday

Fig. 10 EVs charging load curve on weekend

 

Tab. 5 Orderly charging load of EVs kW

Permeability/%	10	20	30	40	50	60
Residential area working day	246	484	732	973	1 211	1 456
Residential District Holidays	176	341	515	680	844	1 023

3.2 Effect of Orderly Charging on Load Characteristics

The charging load when the electric vehicle penetration rate is 10% to 60% is superimposed on the original load curves of the community on weekdays and holidays, and the load curves of three typical communities on weekdays and holidays that include electric vehicle charging loads are obtained as shown in Figures 11 , 12 , 13 , 14 , 15 , and 16 .

Fig. 11 Ordinary-level community load curve on weekday

Fig. 12 Mid-level community load curve on weekday

Fig. 13 High-level community load curve on weekday

Fig. 14 Ordinary-level community load curve on weekend

Fig. 15 Mid-level community load curve on weekend

Fig. 16 High-level community load curve on weekend

It can be seen that electric vehicles are mainly charged during off-peak hours on weekdays and holidays, which coincides with the off-peak hours of residential electricity consumption. When the penetration rate of electric vehicles is 10% to 60%, compared with the original residential electricity load, the maximum load of ordinary, mid-range, and high-end residential communities on weekdays increased by 5% to 114%, 3% to 67%, and 2% to 13%, respectively; the maximum load of ordinary, mid-range, and high-end residential communities on holidays increased by 2% to 70%, 1% to 24%, and 1% to 6%, respectively.

On weekdays, the peak-to-valley difference rates of ordinary, mid-range and high-end communities changed from 0.46, 0.46 and 0.57 to 0.39-0.67, 0.45-0.62 and 0.54-0.52 respectively; on holidays, the peak-to-valley difference rates of ordinary, mid-range and high-end communities changed from 0.38, 0.43 and 0.37 to 0.29-0.52, 0.33-0.39 and 0.39-0.32 respectively.

Under the guidance of the peak-valley electricity price strategy, users actively adjust the charging time of electric vehicles, making full use of the low-load period at night to achieve the purpose of orderly charging. Compared with the disorderly charging state, orderly charging can effectively “shift the peak to fill the valley” of the power grid load, greatly reducing the peak-valley difference rate of the system, thereby avoiding the negative effect of “peak on peak” during disorderly charging, improving equipment utilization efficiency, and benefiting the economic operation of the power grid.

It is recommended that charging facility companies regard bookable charging time as an essential function of charging facilities and promote it nationwide, which can lay a technical foundation for peak shifting and valley filling.

3.3 Effect of Orderly Charging on Distribution Transformer Load Factor

When the penetration rate of electric vehicles is 10% to 60%, the maximum penetration rate of electric vehicles accepted by ordinary, mid-range and high-end communities is shown in Table 6. It can be seen that under the guidance of the peak-valley electricity price strategy, car owners will try to choose to charge their electric vehicles during the valley hours when the electricity price is lower, thereby transferring the charging load of electric vehicles during the evening peak hours to the valley hours of the power grid load to a certain extent. A large number of electric vehicles are concentrated in the valley hours for charging, which effectively reduces the power supply pressure of the distribution transformers in the residential areas during the evening peak hours. The distribution transformers in the residential areas that adopt the peak-valley electricity price guidance strategy can accept more electric vehicles to charge at the same time.

 

Tab. 6 Acceptance of electric vehicle penetration maximum in district distribution transformer %

scene	Ordinary community	Mid-range residential area	High-end residential area
Working Days	≤50	≤60	≤60
Holidays	≤60	≤60	≤60

4 Conclusion

This paper takes three typical communities, ordinary, mid-range and high-end, as the research objects, explores the impact of electric vehicle charging on the load characteristics of community distribution networks at different penetration rates on weekdays and holidays, and analyzes the ability of different types of communities to accept electric vehicle charging. The main conclusions are as follows:

1) A large number of electric vehicles are concentrated in charging during the evening peak, causing the grid load to “peak on peak”. If the charging behavior of users is not guided, the low penetration rate will affect the load characteristics of ordinary communities; in mid- and high-end communities, since the original load base is large, the proportion of electric vehicle charging load to the original load is relatively small, and the impact on the load characteristics is relatively small.

2) For the same electric vehicle penetration rate, the load rate of distribution transformers in high-end and mid-range communities is significantly lower than that in ordinary communities. This is mainly because high-end and mid-range communities have reserved higher power load margins for distribution transformers during planning. The capacity and capacity margin of distribution transformers in ordinary communities are smaller than those in mid-range and high-end communities. As the penetration rate of electric vehicles increases, the power supply pressure gradually increases during peak load periods.

3) The peak-valley electricity price guidance strategy is adopted to enable electric vehicles to charge in an orderly manner. Compared with the disorderly charging state, the community load is “shifted to fill the valley”, which effectively reduces the peak-valley difference rate. The peak-valley electricity price strategy greatly reduces the negative effect of “peak on peak” during disorderly charging, improves the utilization efficiency of equipment, and is conducive to the economic operation of the power grid.

 

--- END ---