Can You Move Your Peak Outside ComEd's 9-to-6 Window, and What Would It Save?

ComEd does not bill you on your highest half hour. It bills you on your highest half hour inside a window, and the window is narrow: nine in the morning to six in the evening, Monday through Friday, holidays excluded. Everything outside it is invisible to the demand charge.

That is a strange and exploitable fact, and the first instinct it produces is the right one. If the meter only looks between nine and six, move the load to seven. The instinct is sound. The size of the prize is where people go wrong, and it is worth getting that right before anyone reorganises a production schedule around it.

The window is real, and it is unforgiving at the edges

The tariff defines billed demand as the highest thirty minute demand established during nine to six on weekdays. There is no separate off-peak demand charge anywhere in the commercial delivery structure. A kilowatt drawn at two in the morning contributes nothing to the demand charge, no matter how many of them there are.

The edges are literal. On one reconciled ComEd account, the highest reading of the day was 41.94 kW at 8:30 in the morning, and it set nothing. The interval that set the charge was the next one, at 9:00, at 40.91 kW, and it produced a Distribution Facilities Charge of $596.88 for the month. Thirty minutes, and about a kilowatt of difference, decided which reading counted.

So the mechanism is exactly as advertised. Shifting works.

What shifting is actually worth, and this is the part that is never said

Here is the cap. When you remove your highest on-peak interval, you do not stop paying a demand charge. Your second-highest on-peak interval becomes your billed demand.

So the saving from moving a load is not the size of the load. It is the gap between your highest qualifying interval and the next one down. Move a 40 kW chiller out of the window, and if some other combination of equipment was already reaching within 3 kW of your peak on another afternoon, you have saved three kilowatts, not forty.

This single fact separates the sites where load shifting is transformative from the sites where it is theatre, and it explains why so many demand management projects underdeliver against a proposal. The proposal multiplied the shifted load by the rate. The bill pays the gap.

Two shapes, and most sites are recognisably one of them:

A spiky site has one or two events per month standing well above everything else. A single startup sequence, a test, a batch process. The gap is large, shifting one event moves the whole bill, and the payback is immediate. This is the case worth acting on.

A flat site runs near its own ceiling through the whole working day. The peak is not an event, it is the ordinary shape of the operation. The gap is a rounding error, and shifting anything short of the entire operation does very little. On a site like this the honest answer is that demand management is not the lever, and knowing that early is worth more than a project that fails slowly.

How to tell which you are, from what you already have

The bill alone will not tell you, because it reports the peak and never the runner-up. But two cheap steps get most of the way.

Ask what was running. You know your own operation. If you can name the event that sets your peak, and it is a scheduled or deferrable thing, you are probably spiky and worth measuring properly.

Look at the month-to-month billed demand across a year of bills. If billed demand jumps around, the peak is being set by events. If it is nearly the same number every month, the peak is your baseline shape, and no schedule change will move it much.

The definitive version reads the interval data behind the bill and sorts the on-peak intervals, so the gap between first and second is a measured number rather than an impression. That is the number a shifting project should be justified against, and it is the number a proposal almost never contains.

The trap on the other side of the window

One caution, because it is the failure mode of a successful shift. Moving a load to 6:30 in the evening removes it from the window. Moving it to 8:45 in the morning also removes it, until the day the process runs fifteen minutes late and lands at 9:05, at which point it sets your demand for the whole month.

Shifting to the edge is fragile. Shifting well clear of it is not. The tariff will not give you partial credit for a near miss, and the bill remembers the one time it happened.