In this paper the first results of a new method for the configuration of district energy systems are presented. District energy systems are believed to help decreasing the CO2-emissions due to energy services (heating, cooling, electricity and hot water), by implementing polygeneration energy conversion technologies. However, because polygeneration technologies are complex, it is meaningful to use them not just for one single building, but for several buildings connected together with a network. The configuration of the network is an important but not trivial task, mainly because the problem involves a large number integer variables and results in an MILP that needs to be optimised.