• Hamiltonian;
  • calculus of variations;
  • Pontryagin;
  • design parameters;
  • refrigeration;
  • economics


The design of controlled processes is a combined optimal control and design problem (OCDP). Literature on solving large OCDPs is rare. This paper presents an algorithm for solving large OCDPs. For this algorithm system dynamics, objective function and their first-order derivatives must be continuous in the state, control and design parameters. The algorithm is successfully applied to the combined control and design problem of a cold store with three possible refrigeration technologies: mechanical refrigeration, ventilation and evaporative cooling. As a result, insight into cost effectiveness of the refrigeration technologies is generated. It is concluded that for this cold store in the Netherlands evaporative cooling is too expensive. Ventilation is economically viable if the cold store is to be used in January only. In case the cold store is to be operated all year then it is most economical to rely on mechanical refrigeration only and use the overcapacity during most part of the year to shift refrigeration to low-tariff hours. Copyright © 2008 John Wiley & Sons, Ltd.