Case Study of Alternatives to Grid Expansion: Reducing Grid Connection Requirements for Large Scale EV Charging Stations through Photovoltaic and Energy Storage Integration

Abstract

The increasing share of electric vehicles requires more charging stations, which in turn de- mands more available grid capacity. In many situations, this necessitates an expansion in the regional and transmission grid, which is both expensive and time consuming. This thesis explores the possibility that an integrated energy system consisting of PV, battery energy storage systems, and hydrogen energy storage systems can reduce the investment costs for grid connection at new large-scale charging stations (1-10 MW). Additionally, the load pat- tern and the correlation between load and traffic patterns were examined. The thesis looks at four different locations near the main highway in Agder County, where the proportion of seasonal tourists is high. An optimization model developed by S. Myhre was used as a comparison tool to identify the most cost-effective solution for grid connection. This model aims to find the most optimal solution for grid connection based on minimizing investment costs, the model considers both grid reinforcement and alternatives technologies such as solar power and energy storage.

The following cases were simulated: • The lowest investment cost for grid connection, considering today’s (2024) cost of grid connection and alternative technologies. • The lowest investment cost for grid connection in a future cost situation, evaluated from a long-term perspective that takes into account technological developments and cost reductions for alternative technologies. • How much investment costs need to fall for the various elements in Integrated Energy System (IES) for the technologies to have a lower investment cost than traditional grid expansion. • What is the lowest investment cost for the combination of grid connection when the model is forced to limit grid expansion. • What is the best investment for grid connection considering the lifetime of the grid and alternative technologies, when the model is forced to limit grid expansion.

The load at the charging stations near the main thoroughfare in Agder County has high activity during the holiday seasons of Christmas, Easter, and the summer holiday. Addi- tionally, the charging stations typically have peak load demand in the middle of the day during the high season, while during the rest of the year the peak load demand is usually after working hours. It was also observed that the load and traffic patterns correlated better at locations where the proportion of commuters is lower.

The results show that in the current cost scenario, the alternative technologies are not more profitable than grid expansion. However, by 2030, the cost of PV could fall enough to have a lower investment cost than traditional expansion, but will lead to weaker supply security. By 2050, the cost of batteries with a C-rate of 0.5 (indicating they can fully discharge in 2 hours) could potentially decrease significantly, thereby helping to reduce some of the reliance on grid connections. Batteries with lower C-rates and hydrogen energy storage systems, will not be profitable by 2050. If one wishes to replace grid capacity with alternative technologies the total investment cost will grow exponentially with grid capacity reduction. A small reduc- tion in grid capacity will have a small increase in investment cost, while a large reduction in grid capacity will lead to a significantly higher investment cost. Over a 30-year perspective, it will still be profitable to invest in solar cell production, when one can sell the excess energy.

The authors recommend replacing the current optimization solver with a commercial prod- uct that is faster and can handle more complex models. This provides an opportunity to expand the model to include a more complete grid tariff, include more technologies, rebuild the model to consider the total investment, a stochastic analysis of load and demand analysis and/or use traffic patterns as the basis for the estimation