Skip to main content
Back to blog
Engineering8 min read0 of 8 min read

Understanding Weather-Normalized Baselines: A Practical Guide

Share:Share

Weather-normalized baselines are the foundation of credible energy performance analysis. Without them, you're comparing apples to oranges — a mild winter will always look like "savings" even if nothing changed operationally.

Pro tip: Weather normalization is required by ASHRAE Guideline 14 and IPMVP for any M&V report that will be used in an energy performance contract or utility incentive claim.

Why Weather Normalization Matters

Buildings consume energy primarily for heating, cooling, and baseload operations. Weather directly impacts the first two, which means raw year-over-year comparisons are inherently misleading.

ASHRAE Guideline 14 provides the statistical framework for creating regression models that separate weather-dependent consumption from operational changes. EdiMono implements these models automatically, using heating and cooling degree days (HDD/CDD) as independent variables.

R² > 0.70ASHRAE minimum threshold for regression model fit

The Regression Approach

The most common model is a changepoint regression — sometimes called a "3-parameter" or "5-parameter" model depending on whether the building has heating only, cooling only, or both.

3-Parameter Heating Model: Energy = Baseload + Heating_Slope × max(0, Balance_Point - OAT)

5-Parameter Model (Heating + Cooling): Energy = Baseload + HS × max(0, HBP - OAT) + CS × max(0, OAT - CBP)

EdiMono fits these models using least-squares regression and validates them against ASHRAE thresholds for R² (>0.7) and CV(RMSE) (<20%).

SavingsAreaChangepoint020040060080010003005007009001100Heating Degree Days (HDD)Energy (GJ)Monthly readingsRegression line

Important: The balance point temperature (the OAT at which heating or cooling kicks in) varies by building and must be found by optimization, not assumed. EdiMono sweeps candidate values from 10°C to 22°C to find the best fit.

Setting Up Baselines in EdiMono

  1. Import 12+ months of utility data — monthly bills or interval data
  2. Select the baseline period — typically the year before an efficiency project
  3. EdiMono auto-selects the best model — testing 3P, 4P, and 5P fits
  4. Review the statistical diagnostics — R², CV(RMSE), t-statistics
  5. Generate normalized comparisons — see true performance changes

Key Takeaway

Always use at least 12 months of baseline data. Fewer data points dramatically increase uncertainty and may not satisfy ASHRAE or IPMVP requirements for M&V reports.

Common Pitfalls

  • Too few data points: You need at least 12 monthly readings for statistical validity
  • Occupancy changes during baseline: If the building's use changed mid-baseline, split the period
  • Multiple fuels: Model electricity and gas separately — they respond differently to weather

Key Takeaway

Weather normalization isn't optional — it's the difference between guessing and knowing. With EdiMono, the math is handled automatically so you can focus on making decisions, not building spreadsheets.

Share:Share

Ready to put these insights into practice?

Request a Demo

More in Engineering

Understanding Weather-Normalized Baselines: A Practical Guide | EdiMono Blog