Optimal multiperiod planning methodology for utility systems is proposed considering internal energy demand, which is formulated as a mixed integer nonlinear programming (MINLP) problem, The problem is decomposed into two levels: the energy distribution network level as a mixed integer linear programming (MILP) problem, and the steam generation unit as a nonlinear programming (NLP) one. The internal steam demand is calculated from the first principle model, and the internal electricity demand is obtained from design data. For multiperiod operation with varying demand as a series of piecewise constant for utilities, the optimal configurations of utility pumps (UP) are determined. The objective function is comprised of the operational costs for each period, transition costs and switch costs between periods. Total cost obtained from the proposed planning methodology has been reduced by 2.34% compared with the planning result without considering switch costs between periods and 3.15-9.14% compared with the result by Lee et al, (1998).