A new approach has been developed for establishing cost-optimal targets for a heat exchanger network (HEN). We formulate a mixed-integer-linear-programming (MILP) transportation problem that simultaneously optimizes for heat exchanger units, heat exchange area and loads on each utility. Heat exchange matches are placed between each hot stream in each temperature interval and all cold streams in all the subsequent lower temperature intervals. The objective function is linear and minimizes the total annual cost of the HEN subject to heat balance constraints. The temperature intervals are generated with a temperature shift just greater than zero. Each temperature interval is small enough that we can linearize the log-mean temperature difference on each match while maintaining the accuracy. Flowrate continuity constraints are written on utility streams so that we could account for non-point utilities. Furthermore, constraints on heat exchange area that occur in retrofit scenarios are incorporated into the model. The solution of the optimization model provides the heat loads on each utility, in case of multiple hot and cold utilities, the heat load on every match pair, the heat exchange area target and the number of units target such that the total annual cost is minimum. Using these targets, it is straightforward to construct heat exchanger networks.