Commercial energy / Buyer guidance
Gas use per square foot: separate weather, occupancy and process loads
Calculate a clearly bounded gas intensity metric and avoid treating a weather-related change as verified efficiency.
Gas intensity needs an operating explanation
Therms per square foot can help track a building, but space heating, hot water, cooking and industrial processes do not respond to floor area in the same way. Record the gas end uses and the area actually served. A campus boiler serving several buildings needs a different boundary from a single tenant kitchen meter.
Use consistent energy units. If invoices show volume, retain the actual heat-content conversion before combining periods. Do not add Ccf from one account to therms from another. EIA's units reference supports the energy-versus-volume distinction; the actual utility bill supplies the conversion for the account.
Account for weather without assuming a perfect relationship
Heating degree days are a weather measure, not a direct reading of a boiler's load. ENERGY STAR's degree-day tool explains the heating and cooling context. Non-heating baseload, occupancy, process output and schedule changes can also affect gas use. A simple therms-per-degree-day ratio can be misleading if a large hot-water or process load stays constant.
Keep weather-sensitive and non-weather-sensitive uses separate where data permits. A regression or engineering baseline may be appropriate for a real project, but its data quality and assumptions must be checked. Do not claim a verified improvement from dividing two annual totals alone.
Worked example: a colder year can obscure a change in intensity
Illustrative property: area is 40,000 ft². Baseline gas use is 80,000 therms, or 2.0 therms/ft²-year, with 4,000 heating degree days. The reporting year uses 84,000 therms, or 2.1 therms/ft²-year, with 4,800 heating degree days.
Raw use rose by 5%. A simple whole-load ratio falls from 20 to 17.5 therms per heating degree day. It would be incorrect to call that a verified efficiency gain because it assumes all consumption varies with heating weather and ignores occupancy or process changes. If a separately supported 20,000-therm annual baseload is assumed in both periods, the weather-sensitive ratios would instead be 15 and approximately 13.33 therms per degree day. That still requires validation; the example shows how the assumption changes the measure.
Questions before assigning a target
- Which meters and buildings are included, and did that boundary change?
- Are gas units and heat-content factors consistent across periods?
- Which end uses respond to weather and which respond to activity?
- Are hours, occupancy, output and equipment records available?
- Is the comparison a screening ratio or a formal project baseline?
Use the ratio to find questions for the operating team. Keep supplier-price changes separate from consumption and weather effects. A favorable therms/ft² result can coexist with higher total cost if rates or fixed charges rise.
Sources and limits
ENERGY STAR EUI, its degree-day calculator, and EIA units support the definitions. The property, weather figures and baseload are invented teaching assumptions, not observed performance or a benchmark for a sector.