Numerous degrees of freedom are often available for the heat integration of batch processes. Usually, only few of these are actually considered and optimized. Time slice-dependant Tmin have been investigated and optimized using a supertargeting procedure originally developed for multiple base case design. The trade-off direct <=> indirect heat exchanges has also been studied and tools for the preliminary assessment of the heat storage system are proposed. Further improvements are being considered for both targeting and design of batch heat recovery schemes.