Long-term load forecasting is a critical function that is ever more valuable as regional markets evolve and capacity becomes more constrained. As generation assets become increasingly expensive to construct, forecasting has become a critical method to avoid building excess capacity, which can dramatically increase costs to ratepayers. Conversely, generating capacity shortages can entail excessive purchased power costs and adversely affect system reliability. Long-term load forecasting requires skills and experience to efficiently develop, analyze, and apply data from multiple sources to accurately forecast demand and energy needs over the planning horizon.

Exeter has conducted numerous long-term forecasts of electricity consumption and peak demands for state and federal government clients. Exeter has consistently relied on an econometric approach to forecast electricity consumption and peak demands because this approach lends itself to the analysis of alternative scenarios and provides a clear method to assess the reasonableness of the forecast. The series of steps involved in performing an econometric forecast typically include:

  • Collection and preparation of data
  • Development of econometric models
  • Econometric estimation of equations
  • Development of base case forecasting assumptions
  • Development of alternative (e.g., high case and low case) scenarios
  • Running the models to calculate the forecast
  • Report preparation