Decisions made in the earliest stages of design can have a major impact on the final design. Therefore, these decisions should take into account as many of the important aspects of the problem as possible. One aspect that is critical to the economic and environmental performance of a process design is the energy utilization. However, it is common practice to choose a process design without considering any process integration until late in the design process. Such a sequential procedure is known to miss optimal heat-integrated solutions. This paper describes the combination of discrete programming and data analysis techniques for considering heat integration at the earliest stages of the design process. Through an iterative procedure, the engineer is able to identify process designs that are amenable to heat integration.