Results Analysis and Interpretation
Choosing the right metric
Different questions need different metrics. EUI (energy use intensity, kBtu/ft²·yr or kWh/m²·yr) supports benchmarking but is weather- and schedule-sensitive. End-use breakdowns explain where energy goes. Peak demand and load profiles inform equipment and utility decisions. Unmet load hours reveal simulation or sizing problems and, under 90.1 performance rating, excessive unmet hours invalidate a comparison. Energy cost captures rate structure that EUI hides.
Calibration
For calibrated models, ASHRAE Guideline 14 gives statistical criteria: normalized mean bias error (NMBE) within ±5% and CV(RMSE) within 15% for monthly data, with tighter limits for hourly. Always state the baseline, the time interval, and the weather normalization used.
Interpreting comparisons
When comparing alternatives, isolate one variable at a time and control for weather. Report uncertainty and sensitivity so a decision-maker knows whether a 2% difference is meaningful. Watch the site-versus-source energy distinction: a measure can reduce site energy while increasing source energy.
Traps
Common errors include comparing EUIs from two different climate zones as if they were equivalent, treating unmet hours as an energy result rather than a diagnostic flag, and reporting a single annual number without the end-use breakdown that explains it. Another trap is presenting savings that come from schedule or setpoint changes rather than the measure under study.
Worked example
A 100,000 ft² office uses 8,000 MMBtu/yr. EUI = 8,000,000 kBtu / 100,000 ft² = 80 kBtu/ft²·yr. If the benchmark is 65, the building is about 23% above benchmark - investigate plug loads and after-hours operation before assuming envelope failure.