Every figure here was published to the client and projected from the model, the load analysis and quoted equipment. When the project completes its Verify cycle, the measured result replaces the projection, and the gap between them stays on the record.
Every project reports the same five numbers — its all-in cost, monthly cost, cost per kWh delivered, savings against diesel, and CO₂ avoided.
Same five every time, so the last project tells you something useful about the next one.
They come out of the same modelling engine every time — the system that produces the client’s decision package, not a spreadsheet rebuilt per job.
12-month project · Wolff 180 B tower crane and STROS NOV 3238 hoist · 189,370 kWh total site energy
The conventional spec puts a generator on site big enough for the heaviest lift, then idles it at low load for the rest of the shift, burning fuel to stay ready. The lift peak is real. The generator sized around it doesn’t have to be. Four architectures were modelled against the same equipment schedule before one was recommended — and the recommendation was not a smaller generator.
The recommendation took diesel off the site entirely: a grid connection with battery buffering to carry the lift peaks. 76,023 litres of fuel never arrived on site. Savings come from fuel and fuel delivery, partly given back by higher equipment rental and utility charges. Across 189,370 kWh of site energy, the project keeps $103,088.
36-month project · four architectures modelled against the same equipment schedule
The rental premium on a battery is fixed. The fuel saving is not — it compounds every month the job runs. What looks marginal in month three is decided by month twelve.
A diesel-and-battery hybrid took 83% of generator hours off the schedule and cut fuel cost by 60%. Where a grid connection of the right size is available, the same model prices lower again — $0.64/kWh with battery, $0.18/kWh on grid alone, against the same $1.36 diesel-only baseline.
But these are the conditions where the maths usually favours one — most are even obvious when you see it modelled.
Ask for the model behind either project. You get the same decision package the client got — the assumptions, the architectures compared, and the one that was recommended.
