Recently, vehicular energy network (VEN) has been proposed to transport energy over a large geographical area by means of electric vehicles (EVs) over a road network. In the near future, an abundance of EVs, plentiful generation of renewables, and mature wireless energy transfer and vehicular communication technologies will expedite the realization of VEN. VEN routing refers to the establishment of energy paths, which are composed of segments of vehicular routes, to transmit energy from a source to a destination. In this paper, we study the fundamental energy loss minimization problem, which studies routing energy for a particular source-destination pair along possible energy paths such that the total energy loss is minimized while guaranteeing a transferable amount of energy. We first discuss how to construct energy paths for a particular source-destination pair, followed by some analytical results. Then we formulate the problem based on the established energy paths constituting linear-programming-based solution methods. We test the performance with a real-world traffic network. This paper lays the foundations of VEN routing.