In recent years, the research in sustainable and energy efficient mobility increased substantially with the rising concerns about climate change. One of the major contributors to this change are green house gas emissions created by the industry, during energy production, agriculture, and transportation. Electric powered mobility (e-mobility) is one of the directions currently taken in order to address those concerns in the transportation sector. Nowadays there exist several classes of electric vehicles to replace classical internal combustion engine vehicles (ICEV). Two prominent examples are the pure battery electric vehicle (BEV) and the plug-in hybrid electric vehicle (PHEV). BEVs use an electric power-train using the energy stored in a rechargeable battery. Its limited capacity and the relative long recharging times imposes significant operational challenges to the planner. However, the price for electricity is far lower than fossil fuel, which can lead to more cost-efficient tours compared to ICEVs. PHEVs do have both engines embedded and thus avoiding the range limitations while benefiting from the cheap electric engine for short distances. This comes at the expense of smaller batteries and higher consumption rates for both energy and fuel due to the additional weight of an extra engine. In addition, the initial costs are quite large for both BEVs and PHEVs, which requires requires cost-efficient tour plans to render them an economically sound choice. In the literature, several definitions for heterogeneous fleets of ICEVs have been investigated [1]. The operational routing problem for BEVs has been studied with increased interest in recent years, starting with the work of [2]. However, a formulation and solution method to tackle the combined problem considering different ICEV, PHEV and BEV types is still missing in the literature. 2 Hybrid heterogeneous electric fleet planning The hybrid heterogeneous electric fleet routing problem with time windows and recharging stations (H 2 EFTW) is a rich vehicle routing problem considering demand and time windows at customers, an energy resource for PHEVs and BEVs and different cost metrics based on the engine used to travel between pairs of nodes. Time-window bounds are modelled as hard constraints and waiting times prior to serving customers are not penalized. The energy resource can be replenished using optional recharging stations. As part of the problem, the amount of