The asset is in trouble: something overheats, a component fails, safety limits are breached.
Response: Stop now, fix now.
Priority: safety and asset protection, not optimisation.
Reality: this is where you have to react, not choose.
Regular service based on hours, days, or OEM recommendations (e.g. every 6 months / 5,000 operating hours).
The what is fixed (scope of work).
The when is often flexible – but in practice, usually set by technician availability and OEM calendars, not by market conditions.
You’ve spotted something early:
Anomaly in SCADA
Sensor drift
Performance degradation trend
Response: “It’s not broken yet, but it will be – let’s fix it before it fails.”
Scheduled maintenance windows
Preventive interventions once an issue is detected
And that’s where most portfolios quietly lose money.
When the maintenance team can come
When there’s a generic “low season”
When the OEM has a slot
We simply help you choose the window for type 2 and type 3 downtime based on:
Production forecasts (how much will this asset produce per hour?)
Price forecasts (what is each MWh worth in each hour?)
Contract exposure (PPA, merchant, imbalance, penalties)
The optimal window where lost production × price is lowest
Alternative windows if weather or market conditions shift
The true cost of doing it “next Tuesday at 09:00 because the team is free”