How Gaia benchmarks welfare energy
The benchmark diagnoses a possible gap in a welfare compound. It is not proof of waste, and it is not a measurement of your site. DataMate measures your site. This page sets out what the benchmark compares against, and where it stops.
Measure it · Fix it · Scale it · Check how your site compares
The two modes
Waste Risk Score. For a reader who does not know their welfare electricity use. It scores how the compound is configured and returns low, medium or high risk. It can say a setup looks exposed, never what it is drawing.
Energy Benchmark. For a reader with a reliable figure: welfare kWh for a week or a month, or fuel used by a generator feeding welfare and nothing else. That figure is compared with Gaia’s measured reference sites.
What a result shows
Every result prints the reported usage, the comparison basis, the reference sample size, the period, and the caveats. Where the sample behind a comparison is thin, it says so. The result appears before any email is asked for.
What the configuration score looks at, and why
- Drying rooms and other high-load welfare spaces
- A drying room draws more than the office next to it, and it gets switched on in October and left. Uncontrolled, it runs 168 hours a week for a few wet coats.
- How the cabins are heated and cooled
- Convectors hold a temperature whether or not anyone is in the cabin, so they heat empty rooms. Air conditioning that also heats can end up fighting a convector in the same cabin, and both appear on the bill.
- Working hours, active days and how many people use the cabins
- These give the unoccupied hours. A compound worked five days from seven until five stands empty most of the week, and the loads do not know it.
- What controls the heating today
- Where the answer is nothing, or a thermostat on the appliance, no control on site knows whether a room is occupied. It is the largest thing the score reacts to.
- The power source
- A generator with no battery is sized for the peaks, so it burns fuel overnight for a load close to nothing.
The weights behind the score, and the control rules AutoMate would apply, are not published. What is reasoned about, and why, is what makes a result checkable.
A generator shared with plant cannot be ranked
Where a generator also feeds cranes, hoists or chargers, its fuel use is not welfare energy. The benchmark says so, and does not rank the result.
The measured band behind the figures
Where a result puts a number on removable waste, it uses the band Gaia has measured on grid sites: 58 to 106 kWh saved per cabin per week. The low end is Maendy, the grid portion of 7 May 2024 to 31 October 2025. The high end is Swynnerton, 3 March to 31 May 2026, a spring window. Both endpoints are measured, not modelled. Spring carries the most removable waste, which is why the output is a range and why both periods are printed beside it.
Rates and carbon
Rates are supplied by each customer, and the ones we hold report 24p to 30p per kWh. The assumed rate is printed beside every £ figure, so a reader can apply their own. Carbon is shown on grid supply only, at 0.13096 kg CO₂e/kWh, DESNZ 2026, generation only, with the kWh beside it. Generator and hybrid supply carry no carbon figure, because the diesel factor is unresolved.
The reference set grows
Every DataMate install adds to the record the comparison is drawn from, so the reference set is larger this month than last. That is why the sample size is printed on every result instead of stated once here.
Who this is for
Project managers wanting a read on a compound before committing to a visit, quantity surveyors checking whether a claim holds, and anyone doing diligence on how Gaia reaches a number.
or get DataMate free and have it measured on your own cabins