Álvaro Gomes, A. Gomes Martins

Faculdade de Ciências e Tecnologia da Universidade de Coimbra, INESC

Rua Antero de Quental, 199

3000 Coimbra, Portugal

Fax: 351 - 39 - 24692 e-mail: agomes@inescc.pt or amartins@inescc.pt



Power systems demand simulation is a useful technique in studies dealing with load flow and load forecasting and also to predict the effects of Load Management (LM) programs both on the electric utilities load curve (EELC) shape and the customers load curve (CLC) shape. Regressive models, frequently used simulation studies, are not so suitable to predict the efects of LM programs as the so called physically based models (PBM). These models are based on physical phenomena related to loads operation, usually the transfer of thermal energy, insted of using great amounts of data collected in power systems. Recently, studies have been carried out in order to obtain models capable of anticipating the effects of LM actions (Malhamé, 1990; Laurent and Malhamé, 1993). The availability of suitable software packages capable of simulating both the LC and LM actions have several advantages: the risks of implementation of LM actions are reduced once one can simulate several LM programs and antecipate the effects, contributing for the design of the LM program itself and allowing the identification of suppluy-side resources; the necessity of collecting data from LC is reduced; real time support to decision dealing with the implementation of LM is available; contacts with consumers in the context of LM programs become easier; reduction of LM programs implementation time is possible. The paper deals with the implementation of this kind of models in SIMCAE2 software package and the results of some simulations with both individual load models and aggregate load models are presented.