Energy you pay while nothing is supposed to run
When power is still drawing and no production order is active, that idle energy is recoverable OPEX, not only a facilities chart. You need draw beside place and order state on the same hall to prove it.
Idle draw is an operations leak, not only a facilities chart
Idle energy is what you pay while a cell is not supposed to be making anything: heaters, hydraulics, air, and auxiliaries that stay up between orders, through a break, or through a starve nobody treated as a stop. The downtime board can look calm. The meter does not. Restart and idle load already sit inside the wider cost of a stop, next to idle labor and the overtime used to catch up. When the cell was never coded down because there was no order to begin with, that draw does not even get the dignity of a downtime line. It arrives later as a utility invoice.
Industry energy practice already separates non-operating base load (nights, weekends, planned shutdowns) from operating idle load during paid time when lines are stopped or empty. The two have different owners and different fixes. This page is the money around the operations join: meter beside order state. It is not an energy-management design and it is not a sustainability report. The join that defines the mismatch sits under energy draw with no active production order. Same-shift recovery of other leaks that also become Saturday sits under same-shift recovery from joins, not overtime.
The bill is hours times draw times your tariff
Fill three numbers you already have, or can measure on one cell for two weeks. Count how many hours the asset drew above its true idle floor while MES showed no active order, or while the station was starved and the shutdown rule said it should have been off. Measure the kilowatt-hours (kWh), or the utility unit you actually buy, in those hours. Use the tariff for that window, because a peak afternoon and a quiet night are not the same euro or dollar per unit. Multiply those three numbers. That product is the recoverable slice, and only for assets where a shutdown was allowed. A cell that must stay warm for a fast restart is not waste. It is a rule you should write down so the alert does not fight the process.
Do not borrow a percentage reduction from a smart-building case and apply it to this hall. There is no twinzo energy-saving percent to paste here. If you cannot yet meter the cell, you cannot yet claim the saving. You can still time the gaps between orders and ask facilities what that base load costs per hour. The gap without a rate is an incomplete invoice. It is still a better start than a site total that hides one high-draw cell inside the building.
A meter alone never makes the call
Site kilowatt-hour totals tell you the month was expensive. They do not tell an area lead which cell to shut before the next tariff step. A machine meter without order state has the opposite blindness: a warm idle and a good run can look alike on the trend. MES without the meter tells you the job is closed and says nothing about draw. The two screens can both be true and still leave the shutdown unowned. That is the same daily gap as any other join that stays in its own app, described under the daily gap between data and a call.
An operational digital twin does not replace the historian or the meter. Facilities keeps consumption truth. MES keeps the order. The twin places both on a hall people recognize so the lead can act in the shift, which is a practical smart manufacturing move only when the shutdown rule already exists. Fifteen-minute averages hide the lunch gap. Minute-scale samples on the few high-draw cells are enough to start. Production monitoring is the surface for that call, next to the other exceptions the shift already owes a decision.
The saving is a cancelled hour, not a new energy program
One cancelled idle hour on a high-draw cell is the unit of benefit. If the lead sees draw with no order and the rule says shut down, the kilowatt-hours that do not happen are the operating line you can defend. If the same picture shows the cell is starved rather than finished, the call is logistics, and any energy drop is a side effect of clearing the wait. Those are different actions. Do not add them into one percentage and call it a utility project. Facilities still owns the tariff and the meter. Operations owns whether someone acted before the window closed.
Keep the pair of draw and order state once it is trusted. Later you can ask which mixes keep leaving the same cells powered through the gap. That question waits until the shift can already stop one idle hour. It is not a reason to delay the shutdown rule until a plant-wide model exists. Start where the invoice already hurts, on the assets a weekend base load has made obvious.
Cancelled idle hours are only one of three equal twin jobs. Draw-and-order events still feed the spatially oriented dataset under the live view. Correlation and causality analytics use that record when the same cells keep drawing through gaps with no order. Same-shift shutdown, the stored spatial record, and hall-wide pattern work stay equal pillars. That is a practical OPEX cut, not a facilities rebrand of the twin.
What to sample before a plant-wide meter map
1. Pick the cells whose weekend or between-order load already annoys the energy team - A handful, not the campus.
2. Write who may power down and what must stay warm - An alert without that rule becomes noise, and a wrong shutdown becomes a worse restart.
3. Log idle-draw windows for two weeks beside order state - Hours, draw, tariff. The product is your pilot case.
4. Compare one cancelled idle hour with the cost of seeing the mismatch in place - Keep the meter system. Add the shared place. How a rollout is priced sits on pricing. What the hall can show sits on the features overview.
You are already paying for the hours nothing was supposed to run. Get in touch if the utility bill keeps rising on cells that had no order, and nobody on the shift can see both facts together.