This report presents a new methodological approach for the optimal design of energy integrated batch processes. The main emphasis is put on indirect and, to some extend, on direct heat exchange networks with the possibility of introducing closed or open storage systems. The study demonstrates the feasibility of optimizing with genetic algorithms while highlighting the pros and cons of this type of approach. The study shows that the resolution of such problems should preferably be done in several steps to better target the expected solutions. Demonstration is made that in spite of relatively large computer times (on PCs) the use of genetic algorithm allows the consideration of both continuous descision variables (size, operational rating of equipement, etc...) and integer variables (related to the structure at design and during operation). Comparison of two optimization strategies is shown with a preference for a two steps optimization scheme. One of the strengths of genetic algorithms is the capacity to accommodate heuristic rules, which can be introduced in the model. However a rigorous modelling strategy is advocated to improve robustness and adequate coding of the decision variables. The pratical aspects of the research work are converted into software developed with MATLAB to solve the energy integration of batch processes with a reasonable number of close or open storages. This software includes the model of superstructures, including the heat exchangers and the storage alternatives, as well as the link to the Struggle algorithm developed at MIT via a dedicated new interface. The package also includes friendly preprocessing using Excel, which is to facilitate application to other similar industrial problems. These software developments have been validated both on an academic and on an industrial type of problems.