Solar carparks for electric vehicle charging in a grid with limited capacity

The electricity grid in the Netherlands is currently unable to provide sufficient capacity for both the integration of new renewable electricity powerplants as well as for the integration of new electricity demands like electric vehicle charging. Symptoms of this scarcity of capacity, also seen in other countries undergoing an energy transition, are observed in various forms.

On the generation side, newly planned solar photovoltaic projects at both commercial and residential scales are increasingly being denied permission to connect to the grid or face long delays for grid reinforcement before they are connected. Since 2020, new utility-scale solar Photovoltaic (PV) installations were provided a maximum of 70% grid connection capacity relative to the solar installed capacity. In 2022, this permitted grid connection capacity has been further lowered to 50% for new projects larger than 1 MWp.

On the demand side, recent mapping studies by the Dutch grid operators show that a majority of the country faces structural congestion in the distribution and transmission grids. The Dutch ambition, as stated in the Regional Energy Strategy (RES), is to integrate 12 GWp of additional solar installed capacity to the existing 14 GWp by 2030. Also by 2030, the total number of Electric Vehicles (EVs) in the Netherlands is expected to increase from about 390,000 (4.4% of the total Dutch passenger vehicle fleet) today to about 1 million (10%), increasing the peak electricity demand.

The scarcity of capacity in the electricity grid to integrate both low carbon solar generation and electric vehicle charging presents an obstacle to the realisation of both short and long term emissions targets. Even though significant grid expansion is already planned and commissioned, this scarcity of capacity is expected to be a characteristic feature of the electricity grid over the coming decades. This thesis aims to investigate how the coupling of solar carparks and EV charging can enable their integration in a grid with scarce capacity while lowering operational carbon emissions.

Two configurations of solar carparks for EV charging are analysed, with the aim of reducing the grid capacity needed:

Chapter 3 analyses the first configuration: a solar carpark for charging EVs at a workplace in the Netherlands where demand peaks are caused by the simultaneous charging of EVs. The inclusion of EV demand forecasting within the scheduled charging reduces annual peak EV charging power by 36-39% relative to immediate charging. These reductions in peak demand enable a more effective use of the available power capacity (now mandated to 50% of solar installed capacity) as well as increase the utilisation of generated solar energy by reducing the need for solar curtailment.

Chapter 4 investigates the second configuration: an off-grid solar carpark for EV charging at a long term (>24 hours) parking lot in a Dutch airport. Offgrid solar charging would enable rapid planning of charging facilities for EVs, removing the uncertainty, delays and costs associated with a grid connection. However, these benefits come with a trade-off: not all vehicles are fully charged at the time of departure. With immediate charging, 20% of EVs over the year leave with a state-of-charge lower than 60% and 3% of EVs leave with a state-of-charge lower than 40%. The adequacy of fleet-level charging is lowest during the low irradiance month of December, during which 63% of vehicles leave with a state-of-charge lower than 60% and 11% of vehicles leave with a state-of-charge lower than 40%. Prioritising the charging of plugged-in vehicles with the lowest state-of-charge ensures that no vehicles leave with a state-of-charge lower than 40% over the entire year, even in the low irradiance winter months. Increasing the minimum duration of parking reduces the fraction of vehicles leaving with state-of-charge below 60% by about 2% per day.

The consequences of scheduled charging on greenhouse gas emissions are investigated in Chapters 5 and 6: Chapter 5 analyses a recently constructed solar carpark located in Dronten, the Netherlands, which includes a solar array, a nickel metal hydride battery and charge points for electric vehicle charging. The aim of the study is to quantify the magnitude of offset carbon emissions per year by the solar carport, and the contribution of battery storage to this offset. The prevalent practice of using the annual average carbon intensity is found to be unsuitable for estimating the annual offset carbon emissions since it does not account for the intra-day patterns of solar production, EV charging and battery cycling. To overcome this, we propose a novel method to calculate the annual offset carbon emissions, making use of the hourly average and hourly marginal carbon intensity. The choice of approach is found to make a difference to the calculated values of the annual offset emissions of the solar carpark. The use of hourly average carbon intensity, which takes into account variation in production, generation and storage, leads to a higher calculated value of annual offset carbon emissions by about 7% relative to a method using the annual average carbon intensity. The use of the hourly marginal carbon intensity to calculate the annual offset carbon emissions suggests that solar carparks have about a 55% higher incremental effect on the carbon intensity associated with the new load of EV charging than what is conventionally calculated. When comparing the annual offset carbon emissions from the solar carpark with and without a battery, we find that the use of the battery has a negligible effect on the annual carbon offset by the system. This result is found to be robust across all the methods of calculation. We therefore conclude that the use of batteries in solar carparks have a low contribution to the total carbon offset by the solar carport.

Chapter 6 investigates the effect of price-based scheduling of EV charging on the carbon intensity of the electricity used by a scheduled fleet of EVs. Real data of over 55,000 home charging sessions collected from 1031 charge points in the Netherlands is analysed. A simulation is made with a commercial smart charging algorithm to create a scheduled charging profile ex post from the EV charging data set. The profile results in an average price reduction of 25% for the overall fleet relative to the costs for unscheduled charging of the fleet over the same period. The time dependent hourly carbon intensity of electricity consumed in the Dutch low voltage grid in 2018 is used to find the impact of price-based scheduling on the mean carbon intensity of electricity used by the fleet. A small decrease of 1.2% in carbon intensity used by the entire EV fleet is observed over the year. Although price optimisation has large effects on the carbon intensity in individual sessions, the effect is found to balance out over the large number of sessions in the year.

Chapter 7 investigates the factors affecting the consumer acceptance of Vehicle-to-Grid (V2G) charging, which remains an insufficiently investigated barrier for the use of the full potential of EVs in demand response and storage. The research work comprises two stages of semi-structured interviews: the first with EV drivers who have never experienced V2G charging, and the second with EV drivers who experienced V2G charging. The participants in the second stage are given access to a V2G-compatible Nissan LEAF and the V2G charging facilities set up in a living lab on the University campus for at least a week each, after which they are interviewed. Clear communication of the battery impacts, financial compensation and operational control are all found to foster acceptance and were, in many cases, necessary conditions for acceptance. The main barriers for acceptance found are range anxiety in various forms, concerns about the effects of V2G charging on the EV battery and the perceived loss of freedom associated with private vehicles. A majority of participants interviewed from both groups are found to accept or conditionally accept V2G charging. This suggests that the use of EVs for demand side storage in addition to demand response is already acceptable to a subset of current EV drivers. The study also clarifies the conditions under which V2G charging would be more acceptable to a broader group of EV users.

The results obtained in this thesis show that the coupling of solar photovoltaics and EV charging enables the integration of both in a grid with scarce capacity. We therefore recommend that solar carparks for EV charging be more widely implemented at workplaces and at longterm (>24 hour) parking lots, though without stationary batteries.