Low-voltage (LV) distribution networks are bound by operational limits for voltage, conductor loading and transformer loading. Networks are described as being constrained when these limits are violated. In the future, LV distribution networks may become constrained because of the additional load due to electric vehicle (EV) charging. The EV hosting capacity is the maximum EV penetration level a network can accommodate before becoming constrained. Distribution network operators (DNOs) need to understand EV hosting capacity in order to allocate planning and investment.

EV hosting capacity can be assessed using load-flow modelling. When modelling LV distribution networks, individual customer loads are explicitly modelled. At the level of individual customers, loads are highly uncertain, and therefore, a stochastic approach is appropriate. Unlike deterministic load-flow simulations which produce a single EV hosting capacity, stochastic load-flow simulations produce a distribution of EV hosting capacities. A high EV hosting capacity is preferable because it allows network reinforcement to be deferred; reducing the costs which are ultimately passed on to customers. On the other hand, a compromise is necessary between deferring network reinforcement and the possibility of violating operational limits.

In the literature, constraints are assessed using hard criteria. When using hard criteria, all voltages must be within the statutory limits. Similarly, all conductor and transformer loadings must be within the continuous rating. Using hard criteria can restrict the EV hosting capacity, when, in reality, customer equipment can handle a small number of minor voltage violations; this is typically referred to as device immunity. Similarly, conductors and transformers can handle a small number of minor overloads; in practice, DNOs consider emergency and cyclic ratings. Together, the magnitude and duration of violations provide a clear indication of severity. When using hard criteria, the severity of violations is not considered, which makes understanding the necessary compromise difficult.

The contribution of this thesis is a novel framework for assessing EV hosting capacity using flexible constraint criteria. When using flexible criteria, a small number of violations are permitted. The number of permitted violations depends on the violation magnitude.

Using flexible criteria allows the EV hosting capacity to be increased, while still ensuring that severe violations are not permitted. Severe violations are explicitly defined in terms of magnitude and duration; the definition can be adjusted as deemed appropriate by DNOs. The proposed framework allows EVs to be integrated in a safe and cost-effective manner.