The method of optimal, automated technological design of heat exchange systems is considered, which allows to reduce the initial problem of mixed nonlinear programming to a matrix form convenient for organization of decomposition synthesis. The initial problem is proposed to be divided into a sequence of tasks of a non-linear programming of search of private economic evaluations on the organization of heat exchange of two flows of different energy potential at a fixed set of a group of control variables, linear programming of search of optimum topology as a combination of optimum structural elements, and non-linear programming of search of a new approximation for the vector of fixed variables. The solution of the problem is based on the construction of a rigorous redundant model of the heat exchange system including all possible variants of the technological scheme construction. The algorithm for solving the problem is built in the form of program conjugation of technological modeling package Aspen Plus/ Aspen Hysys, universal mathematical package Matlab using COM/ActiveX. Output of intermediate and final results is organized in the form of adjacency and reachability matrices in MS Excel and in the form of PFD diagrams through the graphical interface Unity. The proposed method, algorithm and program complex are tested on a number of model examples, including process plants of chemistry, petrochemistry, oil and gas refining. The reached results approach the ultimate estimates of the energy saving potential obtained by the pinch-analysis method. The received solutions are superior to semi-empirical pinch design methods and integral synthesis methods, but the quality of the solution depends on the initial estimate in the range of extremums of the optimality criterion - the amount of capital and operating costs normalized to the same value. To reduce the number of local minimums further, it is proposed to determine the initial estimates by means of a one-time realization of a transportation problem with intermediate points, in which the total costs of hot and cold utilities are used as the target function.