CalTRACK Technical Documentation

Contents:

• CalTRACK Methods
  • Section 1: Overview
  • Section 2: Data Management
  • Section 3(a): Modeling - Billing and Daily Methods
  • Section 3(b): Modeling - Hourly Methods
  • Section 4: Aggregation
• Technical Appendix
  • 2.2.1.2. For fitting baseline models using the hourly methods, no minimum baseline period length is required.
  • 2.2.4.1. Temperature data may not be missing for more than six consecutive hours.
  • 2.4.1. Weather station to be used is closest within climate zone that meets CalTrack data sufficiency requirements.
  • 3.1.3. Models are fit to baseline data in the 365 days immediately prior to the intervention start date.
  • 3.1.4.1. Select and qualify balance points for candidate models for each period for each meter.
  • 3.2.1. A grid search of models is performed using a wide range of candidate balance points.
  • 3.2.3. Maximum gap between candidate balance points in the grid search is 3 degrees F or the equivalent in degrees C.
  • 3.3.1.2. Independent variables
  • 3.4.1. Models using daily data are fit using ordinary least squares.
  • 3.4.3.2. Candidate model qualification.
  • 3.4.3.3. The model with highest adjusted R-squared will be selected as the final model.
  • 3.6.5. Baseline periods.