Optimization problems of heat source operation are solved by linear (LP) or nonlinear (NLP) programming. The optimization methods can be compared on the basis of the complexity of the computational program setup, the time required to input and update data, and the computational time for optimal shifting and loading of installed power machinery. LP methods are preferred (e.g. simplex method, method characteristics), also the Lagrange multipliers NPL method is rarely applied. In this paper, the method of proportional increments (LP, NLP) is applied to compare the optimal loading of the power machinery installed in the thermal power plant using linear and quadratic economic characteristics.