As a step forward in closing this gap, the purpose of this thesis is to discuss heat integration across plants in the "total site". This integration can be accomplished either directly using process streams or indirectly using intermediate fluids. By applying pinch analysis to a system of two plants, it is first shown that the heat transfer that effectively leads to energy savings occurs at temperature levels between pinch temperatures. In some cases, however, heat transfer in other regions is required to attain maximum energy savings. Therefore, a systematic procedure is presented to identify energy-saving targets, and it is followed by a strategy to determine the minimum number of intermediate-fluid circuits needed to achieve maximum energy savings. Next, an MILP problem is proposed to determine the optimum location of these intermediate-fluid circuits. Subsequently, the targets identified are employed in the synthesis of multipurpose heat exchanger networks that are capable of operating each plant stand-alone as well as both plants integrated. An economic analysis shows that the use of a single intermediate-fluid circuit sometimes can be economically more beneficial than direct integration.

The quest for energy-saving opportunities has driven academia to develop methods for energy-efficient design of individual plants. Several practical applications of these now-available tools have been proved useful to the industry. On the other hand, the scarce knowledge about the potential energy savings that can be obtained by integration of many plants in a complex (i.e. the "total site") usually has been attributed to difficulties in implementing these savings. Hence, a gap exists between the absence of information about integration in the "total site" and the actual practical instances in which this integration is implemented.

Then, the models previously developed for the two-plant case are extended to many plants. These models lead to the identification of the maximum energy saving targets, establish the minimum number of connections required between the two-plant combinations, and determine the location of independent intermediate-fluid circuits. Alternative solutions exist that allow flexibility for the subsequent design of a multipurpose heat exchanger network. For cases in which the "total site" is partially shut down, the optimal location of multi-operation circuits that allow flexibility of operation is presented. The use of steam as an intermediate fluid is briefly discussed within the heat integration framework. Finally, the new concept of a "heat belt, " which is a single pipe circuit used to extract heat from and release it to the plants, is introduced.