Extension of Aspiration Level Model for Optimal Planning of Fast Charging Stations

Introduction

Internal combustion engine vehicles, though comfortable damages the environment by the release of harmful greenhouse gases (0.3 kg/mile) during the exhaust stroke. Electric vehicles (EVs) are found to be a sustainable solution that mitigates the harm caused by the conventional vehicles. The release of greenhouse gases by the power sector for the fueling of EVs is significantly less (0.16 kg/mile). Many countries have started adopting EVs in their transportation sector as a substitute for internal combustion engine vehicles. Apart from the reduction of greenhouse gases, usage of EVs has other advantages such as regenerative braking, mitigation in noise pollution, smart charging, good running performance, etc(Ban (2019), Xu (2017), Zeng (2019), Shabarish (2020), Zhenhe (2019)).

The power delivered to the wheels of the EV is obtained using a battery that is embedded inside the vehicle. The battery can be charged at the EV battery refilling stations. Battery charging and Battery Swapping are the two modes of refilling EV batteries. In the former, the EV charges in a Battery Charging Station (BCS), and the depleted battery is charged using the standard charging piles. Range anxiety and long charging time(7-15h) cease the development of EVs when batteries are refueled by this technique. This can be overcome by developing a Fast-Charging Station (FCS) where it takes less than 1 hour to charge the battery. In a Battery Swapping Station (BSS), the depleted battery is swapped with a high SOC battery. The swapping is done using swapping robots. The swapping time is within 4-12 minutes. The depleted batteries are then refueled using the inbuilt BCS available inside the BSS. When the EV user is far away from BSS, Battery Swapping Van (BSV) is used to refill the batteries. BSV is the active mode of replenishing energy, where the van contains a stock of fully charged batteries which reaches the EV user and swaps the depleted battery. The battery swapping process has many advantages compared to battery charging method in terms of service time, mileage, easier integration with renewables, grid balance, etc. State of Health (SOH) of the battery can be maintained by appropriately scheduling the charging of the depleted batteries in the inbuilt BCS. BSS also has a stock of Fully Charged Batteries (FCB) which serves the demand during the shortage of FCBs in the inbuilt BCS. However, the growth of BSS is less than FCS owing to the heavy infrastructure requirement of the former. In the future, BSS can replace the FCS due to its outstanding advantages(Neha (2020), Sasikumar (2018), Saxena (2020), Shao (2017)).

There is a requirement for the proper infrastructure of an FCS, to ensure the quality of service for the EV users. The queuing models study the various types of queues, servers, and system capacity concerning to a service station (BCS). These models are derived from the pure birth model and pure death model where there is only the occurrence of arrivals and departures respectively. A system may have a single server or multiple servers based on the incoming arrival demand. Systems with single servers are studied using single sever queuing models, whereas the performance of the system having multiple servers is analyzed through multi server queuing models. If the system can hold infinite arrivals, infinite system capacity queuing models can be used to examine the behavior of the system. However, due to area constraints most of the real time systems cannot accommodate infinite arrivals, In such cases, finite system capacity queuing models are used to evaluate the performance of the system. In this present work, performance of FCS is analyzed using finite system capacity queuing models to account for the practicality. The arrivals in queuing theory assume Poisson probability distribution thus, the service time follow an exponential probability distribution. The reason being the memoryless property associated with these probability distributions. The memory-less property is closely related to the Markov process where for the given present value, the future values do not depend on the past values. The average arrival rate and the average service rate must be known before applying these models. However, some queuing models follow Gaussian probability distribution for both arrival and service times but their practical significance is very low. Various queue measures such as average waiting time of user in system and queue, average queue length in the queue and the system, percentage idleness of servers helps to evaluate the optimal service level of the system. These measures depend upon the queuing model chosen to serve the incoming users (Manikanta (2020), Thiruvonasundari (2020), Mohd (2019), Till(23)).