Probabilistic Impact Assessment of Residential Charging of Electric Motorcycles on LV Feeders Courtney Rhoda 1 , Bernard Bekker 2 , Justice Chihota 3 1,2,3 Department of Electrical and Electronic Engineering 1,2,3 Stellenbosch University Cape Town, South Africa 1 crhoda@sun.ac.za Abstract—Motorcycles form a popular mode of transport in East African countries, and policies in countries like Rwanda are encouraging a transition to electric motorcycles (EMs). This paper aims to identify the impacts of EM charging on a low voltage residential distribution network in future high uptake scenarios. A stochastic-probabilistic analysis is conducted on a residential network, looking at the effect of EM charging on voltage level, voltage unbalance as well as cable and transformer loading. The Monte Carlo Simulation method is used to account for the randomness in the placement of EMs along the network while the extended Herman Beta transform is used to account for the variability in the residential consumer loads. This paper found transformer overloading to be the limiting factor with regard to EM uptake for the sample network modelled. A sensitivity analysis then highlighted the effects that the feeder properties, transformer size as well as EM and residential load model had on the simulation outcome. The sensitivity analysis found the results most sensitive to the residential load modelling as this affected the transformer loading prior to any EM charging. Keywords— Rwanda, electric motorcycles, impact assessment, residential charging, stochastic-probabilistic analysis I. INTRODUCTION With a global transition towards a cleaner and greener environment many countries have set national targets with regard to electric vehicles (EVs), with the implementation of EV policies worldwide [1] and campaigns such as EV30@30 launched by the Eighth Clean Energy Ministerial in 2017 [2]. As the world moves towards electric mobility, it is anticipated that the movement from carbon-based fuel motorcycles towards electric motorcycles (EMs) will follow suite. There are millions of motorcycles in East Africa, with between 20 000 and 30 000 in Kigali, Rwanda [3]–[5]. This makes countries like Rwanda a good basis for information to use in case studies regarding the impacts of EMs. Ampersand, an EM company, with a mission to “build affordable electric vehicles and charging systems for the three million motorcycle taxi drivers in East Africa, starting with Rwanda.” plans to extend to Uganda and Kenya in the near future [6], [7]. In May 2019, Ampersand launched their pilot programme with 20 EMs to test its battery swap out system [8]. The system makes use of three battery swap out stations where users exchange their fully or partially depleted battery for a full one and only pay for the battery capacity consumed [4], [8]. In August 2019, Paul Kagame, president of Rwanda, announced the movement of the entire country towards EMs stating “We will find a way to replace the ones (motorcycles) you have now” and implored current motorcycle operators to help with the “phase-out process” [3], [4]. Since then, the waiting list of users for the Ampersand EM grew from 1 300 to 7 000 [7]. Ampersand is planning to build 500 more EMs in 2020, however the government wants them to build 5 000 more [7], [9]. Following this, Safi Motors - a local EM company in Rwanda - launched in late October 2019 [10]. Safi Ltd does not make use of a battery swap out system, and were the first company to install EM charging stations in Rwanda [11]. To reduce the downtime due to charging and to accommodate different financial positions, charging works similarly to filling up with fuel, where one can charge depending on how much time or money one has available [10]. The first phase of the launch introduced 60 EMs and three charging stations located next to fuel stations, allowing a total of six EMs to charge at a given moment [12]. A variety of impact assessment studies have been done focusing on the impacts that EV charging and discharging has on the grid [13]–[16]. A study has also been done looking at the economic and environmental effects of each section of the EM life cycle, from manufacture, operation, to end of life [17]. However, the technical impacts of EMs on the distribution network have not been explored. This is likely due to the fact that EM loads may be considered smaller than EV loads and the possible negative effects deemed insignificant. With little research done to assess the technical impacts of EMs, it may not be sufficient to simply assume that the effects of EM charging are negligible. This paper explores whether, with high uptakes of EMs, the effects of these loads may be significant when superimposed onto residential consumption loads, especially during periods of mass simultaneous charging. The aim of the paper is to investigate the technical impacts of EM charging on LV residential feeders. The paper proposes a stochastic-probabilistic approach that is implemented to address the diversity in customer loads and EM loads, and the uncertainty in EM allocation. The approach makes it possible to analyse an extensive set of EM penetration scenarios: varied scenarios of EM location and size, and varied penetration limits per household. The performance of the studied networks, and the respective hosting capacity are determined based on the conditions of four technical variables: voltage- deviation, unbalance, thermal loading of conductors, and transformer loading. The next section describes the simulation methodology while the case study simulation inputs, considerations, method and parameters of interest are discussed in section III. The simulation results are reviewed in section IV. This is followed by a sensitivity analysis of the simulation inputs in section V. The paper then concludes.