The aim of this work is to propose a tool for the design assistance of District Heating Network (DHN) at the beginning of an urban project, as well as for an extension of an existing network. Two goals of DHN optimization are handling at the same time: the optimization of the configuration (network layout, choice between the different production technologies, existence or not of such utilities) and the optimization of the design (mass flow rate, temperature, thermal generating capacity to install, area of the heat exchanger -HX- in sub-station...). The optimization objective is to minimize the global cost of the DHN over 30 years. It includes both operating cost (heating and pumping cost) and investment cost (line, trench, heating plant, HX).

The formulation of the optimization problem leads to a mixed integer non-linear programming (MINLP) problem. That means the optimization problem has a single nonlinear objective function (the global cost) subject to numerous linear and nonlinear constraints (mass flow rate, energy conservation...) with both continuous variables (mass flow rate, temperature distribution, area of the HX, thermal generating capacity to install) and discrete (logical existence of the pipe between sites).

The originalities of this work are the potential offered by the formulation and the method of resolution. The optimization of an academic example is discussed. One of the results is the layout of the DHN, supplied in star or in cascade: a first consumer with high temperature (HT) requirement can supply another consumer with lower temperature (LT) requirement. Furthermore, the localization of the heating plant(s) and which technologies of production are analysed in the case of an isolated consumer.

Concerning the optimization issue, the problem is described globally within GAMS. The model is solved with DICOPT, a MINLP solver which uses the deterministic Outer Approximation (OA) method.