Saving to Trane Reference makes this config visible to every user.
🏗️
Building Energy Model
Analyse your building's energy fingerprint — climate exposure, load regression, hourly load shapes, and a full-year 3D simulation of predicted demand.
Complete Building Onboarding and Utility Tariff steps first.
⚠️Inputs changed — results below may be out of date.
📊 Model Summary
—
Climate Zone
—
Annual HDD (°F·days)
—
Annual CDD (°F·days)
—
Base Load α (kWh/mo)
—
Heat Sensitivity β_h
—
Cool Sensitivity β_c
—
Bill Match
🌡️ Climate Fingerprint — Monthly Degree-Days
📐 Fitted Load Model
① Monthly OLS Regression
Emo = α + βh·HDD + βc·CDD
α —βh—βc—R² —
② Daily Disaggregation
Eday = α/Nm + βh·HDDd + βc·CDDd
③ Hourly Shape (DOE Profile)
Ph = Eday · [focc(h) + fhvac·|Th−65°F|]
④ Bill-Anchoring Correction
Run model to compute
📐 How the Load Model Works
① Monthly OLS Regression
Emo = α + βh·HDD + βc·CDD
Non-negative least squares fitted to your 12 monthly kWh bills. HDD/CDD are climate normals for your ZIP. Gives α (base load), βh (heating sensitivity), βc (cooling sensitivity).
② Daily Disaggregation (365 days)
Eday = α/Nm + βh·HDDd + βc·CDDd
Monthly kWh → daily using each day's actual temperature. HDDd/CDDd computed from hourly temps vs 65 °F base. Weekday vs weekend occupancy scaling applied separately per building type.
③ Hourly Shape — DOE Profile Blend
Ph = Eday · [ focc(h) + fhvac·|Th − 65°F| ]
Daily kWh distributed across 24 hours via a DOE reference building occupancy shape focc(h) blended with a weather-following HVAC component fhvac·|Th−65°F|. The blend is normalised so the 24-hour sum equals Eday exactly. This hourly kW matrix is what populates the 3-D load surface.
🌐 Full-Year Simulated Load Surface
✅ Model Validation — Actual Bill vs Regression vs Corrected
