Reading a load profile before you read the nameplate
Interval data tells you how a facility actually runs. Nameplate data tells you how somebody once expected it to run. Start with the first one.
- Energy audit
- Interval data
- M&V
Every audit starts the same way: a facility manager hands over a stack of nameplate data and a floor plan, and the temptation is to start sizing things immediately. That is backwards. The nameplate says what a machine can draw. The meter says what the building did draw, at 15-minute resolution, for the last twelve months.
Overnight load is the cheapest thing you will ever find
Plot demand against hour of day and look at the floor of the curve — the level the facility never drops below. In a single-shift plant that floor should be small: security lighting, refrigeration, a couple of servers. When it sits at 40% of peak, something is running that nobody asked to run.
The gap between those two curves is not a projection. It is what the meter recorded after the compressed air leaks were fixed and the air handlers were put on a schedule.
Three things the curve gives you for free
- Operating hours that are real. Not the hours in the production schedule — the hours the equipment was actually energized.
- A weather-dependent fraction. Regress daily consumption against degree days and the slope tells you how much of the load is HVAC and how much is process.
- A baseline you can defend. Measurement and Verification is much easier when the pre-period was characterized properly the first time.
Where it stops helping
Whole-building interval data cannot disaggregate. If two measures affect the same end use, you will not separate them from the utility meter alone, and you should not pretend otherwise in the savings calculation. That is what submetering and short-term logging are for.