Number 1/38 (March 2019)

Acta Energetica 01/2019
no 1/2019

Analysis of the Impact of Charging Electric Cars on the Power System Load

Publication date: 2019-08-30
DOI: 10.12736/issn.2330-3022.2019108
About authors | How to cite | Publication history (expand all)

1. Use of electric cars

In recent years, there has been a growing interest in electric cars around the world. This is evidenced by numerous investments in the infrastructure supporting the operation of electric cars. Many countries are interested in the development of an appropriate infrastructure, which includes chargers, communication protocols (for data exchange between vehicles and charging stations), and billing systems. This is connected with growing requirements for the reduction of harmful emissions to the environment, and the spread of electric car use may contribute to the reduction of exhaust gas and transport pollution emissions. Governments of Germany, the Netherlands, the USA and China in recent years have made a lot of effort to achieve the goal of a well-developed infrastructure supporting electric cars. The Chinese government has adopted a plan that assumes that the production of electric cars will account for 20% of the total number of vehicles produced in 2025. Therefore, it aims to quickly expand the network of chargers, so that one charging station will be available to every eight electric cars. In addition, the construction of 120,000 fast charging stations and 500,000 public charging stations by 2020 has been planned. The goal of the German government is to achieve the number of 5 million electric cars sold by 2030. In addition, it intends to allocate EUR 200 million for the construction of 5,000 fast charging stations and an additional EUR 100 million for 10,000 so-called level 2 charging stations. The German government has also specified communication and payment requirements: all charging stations must comply with the Open Charge Point Protocol, so that – using any charging station – users can pay for the electricity with a single identifier [1].

Many states have focused primarily on the appropriate funding of charger network expansion projects, at both local and national levels. Cooperation was also undertaken with power system operators, research centres and automotive sector enterprises. They also seek to increase public awareness and encourage citizens to use electric cars. Currently, many countries are also focusing on the standardization and unification of communication and billing systems. Electromobility measures, objectives and achievements so far of various countries are described in more detail in [1].

Appropriate initiatives for the development of electromobility have also been undertaken by the Polish government. These include Package for Clean Transport consisting of three documents setting out the principles of electromobility development until 2025, as well as Electromobility Development Plan in Poland. The primary goal set out in these documents is to increase the number of electric cars. To this end, since these cars are expensive, financial incentives for their purchases are contemplated. At the same time, a fee depending on their emissivity would be imposed on the other types of vehicles. It is planned to increase the number of electric cars to 1 million in 2025. Moreover, electromobility development measures have been divided into three stages. The first is the preparation stage, which assumed the adoption of an act on electromobility and the completion by 2018 of preparatory works for the construction of charging stations. This would also be the stage of promoting electromobility. The second stage was to be the construction of the first charging stations in 2018–2020. The last stage involved increasing their number and gradual replacement of combustion cars with electric vehicles [24].

Now, based on data collected by the Polish Alternative Fuels Association (PSPA) and the Polish Automotive Industry Association (PZPM), an Electromobility Counter of the actual number of electric vehicles and charging stations available in Poland has been developed. The Counter is available on these organisations’ websites [4, 6]. According to the Counter, ca. 4,100 EVs were registered in Poland at the end of 2018, and ca. 5,600 at the beginning of June 2019. This increase in the EV number resulted mainly from company purchases and development of car-sharing, i.e. car rental for minutes. The EVs registered in Poland constitute about 0.4% of all electric vehicles in Europe.

With the increasing number of electric cars, the development of a charging infrastructure can be observed. At the beginning of June 2019 there were ca. 670 charging stations in Poland, most of them in Warsaw. DC fast charging stations in Poland account for 33% of all charging stations. This share of DC fast charging stations is among the largest in Europe, larger than 30% in the Czech Republic, but smaller than 50% in Sweden. Many companies are involved in the development of the charging infrastructure in Poland. For example, PKN Orlen has invested in its own charging network and has announced the installation of ca. 50 chargers by the end of March 2020. Tauron has opened 23 charging stations in Katowice. PGE Polska Grupa Energetyczna plans to launch ca. 1,500 charging stations by 2022. Also, Energa plans investment in the construction of 54 charging stations in 2019 and another 100 by 2022 [6].

2. G2V and V2G concept

With the growing interest in electric cars and investment in electromobility, the impact of charging a large number of electric cars on the power system has been noticed. Electric vehicles are equipped with batteries that can potentially be used as mobile energy storage. When connected to a charging station, an EV can intake electricity for charging, but with the appropriate equipment of the vehicle and the station alike, it can also output electricity to the power system. These opportunities are offered by G2V Grid-to-Vehicle and V2G Vehicle-to-Grid solutions, respectively.

The G2V charging process is the unidirectional flow of energy, from power system to electric car battery. It applies to charging electric cars and allows only the option to purchase electricity. This is the simplest way to integrate an electric car with the power system. Due to widely available and well-developed charging technologies, it is anticipated that the G2V solution will be implemented first. Moreover, this solution does not require such advanced communication and security systems as the V2G solution. This implies lower capex and opex costs [711].

V2G is a network service that offers energy exchange between a power system and electric car battery. Car parking hours coincide with the hours of the system’s highest loads in the 24 hours of the day. Therefore, these cars can be used as a source of reserve power. The V2G solution supporting electric cars would output some of the energy stored in their batteries to the power system. Since these cars supply energy to the system during the day, their batteries are discharged more than in the G2V solution. Therefore, they would have a higher demand for power for charging at night. Such management of a large number of charged vehicles by the grid operator through messages and instructions issued to vehicle owners allows increasing the power system operation’s stability, flexibility and security.

The V2G solution and technologies are still being tested and examined. It requires a more advanced infrastructure than that for G2V, and a communication system that allows the exchange of data between cars, grid operator, and vehicle owner is also necessary, as well as appropriate channels for bi-directional energy exchange. Also important is a system of incentives, so that vehicle owners agree to supply energy to the system. The energy available to the system depends on how many vehicles are connected to chargers at a given moment and on the decision of the users, i.e. the vehicle owners who would decide how much energy to output to the system. The entire battery capacity can not be available for transfer to the system, since some of it must cover the vehicle demand [1215].

Despite many requirements, the V2G solution enables filling the valleys and also cutting the peaks, while the G2V solution only allows filling the valleys. The valley-filling strategy consists in increasing the load in times of low power demand, i.e. during the night valley. This demand can be realized by charging cars at night. Cutting peaks, on the other hand, reduces the system load during its peaks. It would be enabled by EVs’ returning electricity to the system at this time [7, 9]. Both strategies aim at equalisation of the system load during the day and of the daily load curve [7, 13, 15]. An example of a daily load curve shape change due to the V2G strategy is shown in Fig. 1.


Fig. 1. Smoothing the daily load curve by electric cars supporting V2G technology. The red line represents the daily load curve with the participation of electric cars [7]

Another available option is smart charging. This service consists in controlling the EV charging process. Implementation of this service may reduce the load during the day, i.e. it will allow cutting peaks. At the same time, this service does not require the charger to be equipped with a connector enabling bi-directional energy flow. After exceeding a set level of electricity consumption for charging purposes, the energy intake can be automatically limited or interrupted by the operator [16]. This paper discusses the implementation of the G2V and V2G services only.

3. Daily load curves with electric cars impact

As part of the study of the impact of charging a large number of electric cars on the power system operation, daily load curves were developed for the average working day in January 2016, which consider the possible impact of electric cars under the G2V and V2G scenarios. In order to examine the possible impact of the use of electric cars it was assumed that the number of electric cars is ca. 1 million, and 40% of them support the V2G two-way energy exchange. It was assumed that the average car battery capacity would be 30 kWh. The consumption of energy contained in the battery for the vehicle’s own needs, which include the user’s commuting to the place of work and to the place of residence, was also considered. This consumption amounted to approx. 4 kWh per day and was calculated on the basis of the formula from [17]. The efficiency of energy conversion by the charging (and outputting) system is 95%. The energy conversion efficiency and energy consumption for own needs were included in the calculation of PG2V and PV2G. The other assumptions are presented in [18]. Under these assumptions, possible system loads were charted, assuming the current manner of using electricity by consumers and additional loads resulting from the impact of electric cars. The curves were based on formula (1) for the G2V variant and on formula (2) for the V2G solution [18]:

wzor1.jpg                                                                                                         (1)

where: PG2V – power system load including the impact of G2V electric cars [GW], PKSE – power system load without the impact of electric cars at a given time [GW], PG2Vp – power intake from the system by G2V cars for charging [GW].

wzor2.jpg                                                                                     (2)

where: PV2G – power system load including the impact of V2G electric cars [GW], PV2Gp – power intake from the system by V2G cars for charging [GW], Ppraca (work) – power output from electric cars at the place of work [GW], Pdom (home) – power output from electric cars at the place of residence [GW].

Assumptions regarding the EV operation mode and the resulting changes in the system load depending on the variant are described in [18]. In the G2V variant, the energy demand would increase by ca. 4.6 GWh per day. The threat of increased load during peak loads is also visible here. This increase is undesirable, and with more vehicles it can become a threat to the power system. Therefore, the grid operator’s appropriate management of the number of vehicles charged during the day is important. The valley-filling effect would also be accomplished, because cars intake energy for charging at night [18].


Fig. 2. Daily load curve in the average working day in January 2016 G2V variant [18]


Fig. 3. Daily load curve in the average working day in January 2016 V2G variant [18]

In the V2G variant, the energy demand would increase by ca. 13.2 GWh per day, while ca. 8.6 GWh would be returned to the system. The more efficient balancing of the system load is not the only advantage of the V2G scenario. With this solution, cars charged during the day would not increase the peak loads. It would also allow peak-cutting, because some peak loads could be covered by vehicles’ energy output to the system. The V2G solution supporting electric cars also demand more energy for charging at night, which contributes to more efficient valley-filling [18].

4. Load rise and fall dynamics during the day

In each variant in the daily load curve diagrams two areas can be identified with the largest change in the load value in a relatively short time. The first is the load increase from the night valley to the morning peak. The other is the load decrease from the evening peak to the night valley. To examine the changes that would occur in the load rise and fall dynamics, the load gradients (for both V2G and G2V variants) were calculated. The gradients are indicators of the magnitude of change in the power system load at the time the change occurred. They were calculated by two methods – as a derivative of the trend line equation, and as the ratio of the load increase to the time over which it occurred. For the purpose of analysing the gradients, the gradients were calculated for:

  • curve of the actual system load (reference chart)
  • curve including changes in the power system load in the G2V variant
  • curve including changes in the power system load in the V2G variant.

The load change dynamics were examined for the period from the lowest load that occurred in the night valley to the first highest peak load, and for a smaller range of data – in the periods of the highest load jump at the transition between time zones. The gradients are presented in Tab. 1.


Tab. 1. Load gradients for the load rise period

Gradients were also determined for the time interval of a significant load decrease – from the afternoon peak to the night valley. The gradients ​​are listed in Tab. 2.


Tab. 2. Load gradients for the load fall period

For a larger time interval of 10 hours (3.00 a.m. – 1.00 p.m.) the gradients vary depending on the chosen calculation method. The actual system loads would change from ca. 16.4 GW to ca. 23.6 GW. In the G2V variant the load increases in the night valley, which is also associated with the increase of the lowest load – it would amount to ca. 16.9 GW. Since some electric cars intake energy during the day, the highest load in the morning peak would also change. Considering the small number of vehicles charging during the day (ca. 5%), this value at 1.00 p.m. would increase by ca. 0.1 GW in relation to the actual load at this time and would amount to ca. 23.7 GW. Despite the increased demand for power during the day, the gradients indicated that the dynamics of load changes would improve and would be smaller. The smoothest load changes can be observed for the V2G solution. This is due to the decrease in peak load and the larger load increase in the night valley than in the G2V variant. The effect of managing the EV battery charging and discharging process in V2G is the more even shape of the load curve. The load at 3.00 a.m. would amount to ca. 17.8 GW i.e. it would be higher by ca. 1.4 GW than the actual load and ca. 0.9 GW higher than the load in the G2V variant at the same time. Whereas the load at 1.00 p.m. would amount to ca. 23 GW – i.e. would be lower by ca. 0.6 GW than the actual load and lower by ca. 0.7 GW than in the G2V variant. Due to these changes, the power demand jumps in the smoothest manner among the three variants presented. This is the most beneficial for the power system operation. Sections of the discussed curves along with the trend line equations are shown in Fig. 4.


Fig. 4. Gradient for load rise period 3.00 a.m. – 1.00 p.m.: 1) variant G2V and 2) variant V2G, vs. the actual load curve

Another time interval considered is the direct load rise period. This time during the largest load jump was 3 hours, lasting from 6.00 a.m. to 9.00 a.m. For this narrowed period of time, the load change dynamics can be determined more accurately. These gradients are larger compared to previous calculations. These calculation results confirm the conclusions from the analysis of the 3.00 a.m. – 1.00 p.m. time interval. The actual system load would increase at ca. 1.8 GW/h rate. The load would change over this time from ca. 17.5 GW to ca. 22.9 GW. In the G2V variant, the load would jump from 17.8 GW to ca. 23 GW, i.e. by ca. 1.7 GW/h. Due to the initial value increase, the change dynamics is lower. This is reflected in the smaller (by about 0.1 GW/h) gradients. The smallest load jump occurs in the V2G solution. The system load would increase from ca. 18.6 GW to ca. 22.5 GW. The results are the smallest gradients (ca. 1.3 GW/h) compared to the previous variants. This is a change that would positively affect the power system operation, as less rapid load changes are easier to accommodate. Curves of the load rise in 6.00 a.m. – 9.00 a.m. are shown in Fig. 5.


Fig. 5. Gradient for load rise period 6.00 a.m. – 9.00 p.m.: 1) variant G2V and 2) variant V2G, vs. the actual load curve

The challenge for the power system is not only to cover the rapidly growing demand in the morning, but also to properly adjust the energy output during a high decline in demand. The biggest drop in the system load occurs in 8.00 p.m. – midnight. In this period, as in the case of load rise, the highest gradient values ​​were obtained also for the actual load curve. Negative gradients indicate a decrease in power demand. The actual load would change over these four hours from ca. 23.9 GW to ca. 18.8 GW. Whereas the load in the G2V variant would decrease from ca. 24 GW to ca. 19.3 GW – hence load change dynamics lower by ca. 0.1 GW/h. The lowest load drop dynamics was that of the load curve including the V2G EV impact. The load would decrease from ca. 23.6 GW to ca. 20.1 GW, i.e. the load change rate would be ca. 0.9 GW/h – i.e. would be lower by ca. 0.4 GW/h compared to the actual curve, and lower by ca. 0.3 GW/h than in the G2V variant. This is the smallest system load change.

The smallest gradients were obtained for the V2G load curve. This indicates low dynamics of changes in the V2G solution. The V2G and, albeit to a lesser extent, G2V variants also have a positive effect on the load curve in this time interval. Fragments of the system load drop curves are shown in Fig. 6.


Fig. 6. Gradient for load fall period 8.00 p.m. – midnight: 1) variant G2V and 2) variant V2G, vs. the actual load curve

Based on these results it can be concluded that the G2V solution slightly affected the gradients’ reduction. The smallest gradients were obtained for V2G technology. This means that after applying the V2G variant, compared to G2V and the values obtained without the EV participation, the smallest load changes would be observed at the same time. This is beneficial for the power system operation. In the G2V solution or without the EV participation, the system had to cover a large increase (or decrease) in electricity demand in a relatively short time. The V2G solution, even if it is not supported by every electric car (in this case 40% of the vehicles return energy to the system), significantly reduces the load jump and the gradients. The V2G solution reduces the demand increase or decrease in the same time interval. To accommodate a smooth electricity demand increase or decrease is less burdensome for electricity generators and for the system.

5. Analysis of electricity load indicators

The daily load curves can be described by indicators – with them their shapes can be compared [19, 20]. Daily load variability can be characterized by three types of indicators. The reference quantities can be, for example, the peak and average installed capacities. By reference to them, several types of indicators (degrees) can be identified. Coefficients referred to the daily peak power are called load factors and designated with the symbol md. Coefficients referred to the average power are called daily balancing factors and designated with the symbol ld. Load variability coefficients referred to the installed capacity are called utilisation factors and designated with symbol nd.

In this paper the daily base (md0) and average (mdśr) load factors, base (ld0) and peak (lds) balancing factors, and base (nd0), average (ndśr) and peak (nds) utilisation factors are calculated. Their values are listed in Tab. 3.


Tab. 3. Load variability factors for various daily load curve variants

The base load factors in the G2V and V2G variants are higher than for the actual system load. This indicates an increase in the base load (as a result of EV charging at night when there is the lowest load). The average load factor is the most frequently used indicator. It is interpreted as a measure of the daily chart’s filling. A marked change of this factor is visible in the V2G variant – the chart is there much better filled as a result of concurrent valley-filling and peak-cutting in relation to the base variant.

The base and peak balancing factors in the V2G variant are closer to one than in the actual load. This means better equalisation of the daily load curve as a result of the lowering of peak loads and increasing of base loads.

The basic utilisation factors also indicate an increase in the base system loads, both in the G2V and V2G variants. The average utilisation factors also indicate an increase in the average loads, both in the G2V and V2G variants. Similarly, in the case of the peak utilisation factor – the biggest changes occurred in the V2G variant. The peak factor for this example is the smallest. Differences between the utilisation factors are the smallest.

6. Conclusions

Charging a large number of electric cars can have a significant impact on the power system operation. In order to avoid negative effects, e.g. uncontrolled increase in power demand, the EV charging process should be properly managed. This not only allows avoiding risks but can also bring many benefits.

In the G2V solution most electric cars would be charged at night, and the process would be spread over time. As a result, vehicles would not create an additional load jump, and would not cause a significant increase in power demand during the day. As follows from the calculations presented, it can also cause changes in the way the load increases or decreases. The load gradients ​indicate that the load, depending on the time of day, would increase or decrease less rapidly. Implementation of the G2V scenario would also increase the load equalisation, as indicated by the daily load variation indicators. Moreover, the G2V solution can be implemented in already developed and well-known technical solutions. Unfortunately, the G2V option also carries with it a certain risk – some electric cars will be charged during the day. The more such vehicles there are, the more power demand increases during peak loads, which is an undesirable effect. Therefore, the G2V solution is perceived as a good way to integrate electric cars with the power system in the initial phase of electromobility development.

Another mode of EV interoperability with the power system is the V2G solution. According to the idea of ​​this scenario, electric cars would not only intake energy during charging at night, but also would give back some of it during the day. Such management of the EV battery charging (and discharging) process by the grid operator through issuing instructions and messages to vehicle owners can bring many positive effects on the power system operation. It would increase the load in the night valley more than the G2V solution. Moreover, it would allow reducing the load in the peak period. In addition, this would increase the system load equalisation, which is reflected in the calculated indicator values. The V2G solution significantly reduces the power demand increase and decrease rates. The low gradients also indicate increased load equalisation and smoothed daily load curve. Unfortunately, the V2G solution is still being tested, and its supporting technologies still require proper development.

In technical and economic terms, the V2G solution is more demanding than G2V. Its implementation entails higher opex and capex costs. Nevertheless, it has many advantages and greater potential as regards the impact on the power system load development.

  1. Hall D., Lutsey N., Emerging Best Practices for Electric Vehicle Charging Infrastructure, White Paper, International Council on Clean Transportation, 2017.
  2. Plan rozwoju elektromobilności w Polsce [Electromobility Development Plan in Poland], Ministry of Energy, Warsaw 2016.
  3. Sipiński D., Bolesta K., Cicha rewolucja w energetyce. Elektromobilność w Polsce [Silent Revolution in the Energy Sector. Electromobility in Poland], Polityka Insight, 2017.
  4. Flasza J., Elektromobilność w Polsce – wyzwania i możliwości z uwzględnieniem inteligentnych instalacji OZE [Electromobility in Poland Challenges and Opportunities Including Smart RES Plants], Autobusy: Technika, Eksploatacja, Systemy Transportowe, No. 6, 2017.
  5. Polish Automotive Industry Association (PZPM) [online], [access: 03.06.2019].
  6. Polish Alternative Fuels Association (PSPA) [online], [access: 03.06.2019].
  7. Alonso M. et al., Optimal Charging Scheduling of Electric Vehicles in Smart Grids by Heuristic Algorithms, Energies, Vol. 7, 2014, pp. 2449–2475.
  8. Raustad R.A., The role of V2G in the smart grid of the future, Interface, Vol. 24, No. 1, 2015, pp. 53–56.
  9. Gracia-Valle R., Joao A., Pecas L., Electric Vehicle Integration into Modern Power Networks, Springer, New York 2013.
  10. Damiano A. et al., Vehicle-to-Grid Technology: State of the Art and Future Scenarios, Journal of Energy and Power Engineering, No. 8, 2014.
  11. Monteiro V. et al., Batteries Charging Systems for Electric and Plug-In Hybrid Electric Vehicles [w:] New Advances in Vehicular Technology and Automotive Engineering, ed. J. Carmo, J. Ribeiro, Rijeka 2012.
  12. Mullan J. et al., The Technical, Economic and Commercial Viability of the Vehicle-to-Grid Concept, Energy Policy, Vol. 48, 2012.
  13. Wang Z., Wang S., Grid Power Peak Shaving and Valley Filling Using Vehicle-to-Grid Systems, IEEE Transactions on Power Delivery, Vol. 28, No. 3, 2013.
  14. Yu R. at al., Balancing Power Demand through EV Mobility in Vehicle-to-Grid Mobile Energy Networks, IEEE Transactions on Industrial Informatics, Vol. 12, 2016.
  15. Deivanayagam R., Vehicle-to-grid Technology: Concept, Status and Challenges, Journal of Undergraduate Research, No. 9, 2016.
  16. EVBox [EV chargin station manufacturer] [online], [access: 02.06.2019]
  17. Pistoia G., Electric and Hybrid Vehicles – Power Sources, Models, Sustainability, Infrastructure and the Market, Elsevier, Amsterdam 2010.
  18. Stahl W., Rozwiązania V2G I G2V jako sposoby wykorzystania samochodów elektrycznych do zmiany kształtu krzywej obciążenia dobowego systemu elektroenergetycznego [V2G and G2V solutions as the ways of using electric cars to change the shape of the daily load curve of the power system], Zeszyty Naukowe Wydziału Elektrotechniki i Automatyki Politechniki Gdańskiej, No. 61, 2018, pp. 69–72.
  19. Góra S., Gospodarka elektroenergetyczna w przemyśle [Power management in industry], Państwowe Wydawnictwo Naukowe, Warsaw 1975.
  20. Marecki J., Gospodarka elektroenergetyczna [Power management] [in:] Poradnik inżyniera elektryka [Electrical Engineer’s Guidebook], Vol. 3, edited by L. Bożentowicz, Wydawnictwa Naukowo-Techniczne, Warsaw 2011.
Share with LinkedIn Tweet