OPTIMAL EXPANSION OF ENERGY EFFICIENCY PROGRAMS

Electric utilities in the US are increasing their commitment to energy efficiency, spending over $1 billion on such programs in 1991. The qu estion we address is: how quickly should energy efficiency programs be expanded? This is called the 'ramp up' problem. Although such program s have short lead times, significant inefficiencies can occur if they grow too quickly. This is because the capacity to install efficiency m easures cannot be changed instantaneously. Staff must be hired and tra ined, while non-utility contractors must increase their capability. Ma nagement capabilities, such as database systems, must also be expanded . Adjustment costs also occur later as opportunities for inexpensive c onservation are used up. Downsizing programs too quickly results in ha rdship for contractors and high employee severance costs. Optimal prog ram expansion might spread the effort over more years in order to avoi d the inefficiencies of a quick run-up and run-down. We present an opt imal control framework for scheduling program expansion to maximize ne t benefits. Among the costs considered is the expense of altering prog ram capacity. The variables that describe program status include staff ing level and the number of potential installations that have already been made. The decisions that the utility makes are the number of inst allations, which is influenced by incentive levels and promotional eff orts, and the change in capacity. This control problem is solved using dynamic programming. A simpler version of the problem can also be sol ved with a spreadsheet, permitting extensive sensitivity analysis. As examples, we present applications to the design of residential weather ization programs.