The professional staff has the following proprietary Harbourfront modeling systems to support our relevant work efforts:
Smart Meter/Load Research Sample Design & Selection System – this software package develops stratified sample designs for any class of customers, at a confidence level and error specification selected by the user. An additional module allows the analyst to select multiple samples that follow the distributional shape of the original design. This facilitates the “refreshing” of in-field samples without the need for wholesale replacement of sample points.
Non-Linear Weather Normalization System – this package develops weather normalized hourly, daily, or monthly sales for any class of business specified. A unique multivariate non-linear weather normalization model structure is customized for each class of customer for each client. Heating to cooling “balance points “ are automatically estimated from client data. The end result is a weather normalization system that corrects the inherent estimation biases found in both linear and non-linear heating and cooling degree day-based weather normalization models.
Generating Asset Cost & Performance Modeling System – these models and databases have been developed for nuclear, fossil , wind and geothermal generating assets . They have been used on behalf of both utility senior management in benchmarking their own assets as well as in regulatory proceedings explaining generating asset performance.
Electric Utility Econometric Load Forecasting System – our load and energy forecasting assignments are facilitated by our standard multi-model econometric forecasting process that has been successfully employed in all of our client forecasting engagements. This system utilizes ForecastPro statistical modeling options coupled with our customized model specification and variable selection procedures. Weather normalization of historical data is accomplished in a data preparation stage using our Non-Linear Weather Normalization System.