Three Applications of Operations Research in a Large Electric Utility
Abstract
The studies described are (1) the development of a “self-correcting” system for estimating, up to twelve months in advance, the future monthly mean levels of Lake Erie, which control the flow available for power generation at Niagara Falls, (2) the development of a system for forecasting daily peak loads up to twenty-four hours in advance, based on a new model of load behavior, whose input is a normal weather forecast and whose output has a median error of less than 1 per cent, (3) the development of a method for determining the optimum replacement point for trucks in a large fleet under changing economic conditions.

