Total site heat integration provides more energy saving opportunities than conventional heat integration within single process. Some total site design methods including direct integration using process streams and indirect integration using intermediate-fluid circuit have been proposed during these years. In these works, it is found that direct heat integration can recover more heat, require less heat transfer area, but incur higher cost in pipeline compared with indirect heat integration. However, most research focused on total site heat integration with considering only indirect heat integration or direct heat integration, no methodology simultaneously considered both direct and indirect heat integration. Therefore, some optimal designs will be missed by using current total site integration methodologies for they do not fully utilize the features of indirect and direct heat integration. In this work, the situation that only applying direct heat integration, only applying indirect heat integration and applying both heat integration in total site heat integration are analyzed. The new idea allows the direct and indirect heat integration to be considered simultaneously. The application of the new methodology can bring a significant energy saving in total site. Heat integration across plants can bring large energy savings and has been studied for many years. The concept of total site heat integration, first introduced by Dhole and Linnhoff (1993), describes heat integration of multiple plants with a central utility system. The site sink-source profiles proposed by them can be used to determine the different levels of steam that can be generated in order to indirectly integrate heat through multiple processes. Hu and Ahmad (1994) developed a total site heat integration methodology that considers the utility system. In their work, different levels of steam were used to transfer heat between processes, and such integration using intermediaries was defined as indirect integration. Klemes et al. (1997) further developed the total site profile and the site utility grand composite curve to evaluate total site potential heat recovery. Rodera and Bagajewicz (2001) introduced a mathematical methodology to compare the difference between indirect integration and direct integration. Moreover, they analyzed factors such as number of intermediate-fluid circuits and type of intermediate-fluid in terms of cost. Because total site contains a number of processes and each process may require different minimum temperature difference, Varbanov et al.(2012) developed a modified total site targeting procedure that can obtain more realistic heat recovery targets for total sites by specifying the minimum temperature difference. More recently, Perry et al. (2008) extended the total site concept to a broader spectrum of processes in addition to the well-studied industrial processes in terms of carbon footprint. Varbanov and Klemes (2011) further developed the total site methodology of Perry et al. (2008) by incorporating the aspects of renewable energy sources and CO2 emissions. Kapil et al. (2012) proposed a methodology for estimating the cogeneration potential for a site utility system via bottom-up and top-down procedures. In their methodology, the low-grade heat is used through heat pumping, organic Rankine cycles, energy recovery from exhaust gases, absorption refrigeration and boiler feed water heating. Although many aspects of total site heat integration have been studied by various researchers, there are still some unexplored problems. In the work of Wang (2013),they found that because of long distance between plants, pipeline cost becomes a major part of capital cost, and the number of heat exchange