Utility Demand Charge

Residential customers are billed for energy and nothing else, which is why almost nobody arrives at a commercial bill prepared for the second line. A commercial or industrial tariff charges twice: once for the kilowatt-HOURS consumed, and again for the highest RATE at which they were consumed. The second charge is the demand charge, and on many sites it is a third to a half of the total bill.

The mechanism is worth stating precisely, because the arithmetic is trivial and the misunderstanding is not. The meter does not record instantaneous power. It records the AVERAGE power over each demand interval — fifteen minutes on most commercial tariffs, thirty on some — and it keeps the largest interval it saw all month. That single quarter-hour sets the billed demand. Running a load for one interval costs exactly as much as running it every interval of the month, and the kilowatt-hours involved in that one interval are trivial. A 400 kW site on a $14.50/kW tariff pays $5,800 in demand charges whether the peak lasted fifteen minutes or thirty days.

Which is why demand is managed by SEQUENCING rather than by efficiency. A more efficient motor cuts both bills a little. Making sure the chiller and the air compressor and the electric reheat never start in the same quarter-hour cuts the demand bill a lot, and costs nothing but a control strategy. Staggered starts, soft starters on the big motors, interlocks between electric heat and mechanical cooling, and shifting a batch process off the site peak are all standard, and a demand controller that sheds a non-critical load for four minutes when the interval is trending high can pay for itself in one season.

Three clauses on the rate schedule will move this answer, and every one of them is worth reading before trusting an estimate. A RATCHET bills the greater of this month's peak and some percentage — 75 to 90 per cent is common — of the highest peak in the preceding eleven months, so one bad quarter-hour follows the site for a year and a single commissioning test can be expensive. A KVA TARIFF bills apparent power instead of real power, which means a poor power factor inflates the billed figure even though the work done is identical; on such a tariff, power-factor correction is a direct and immediate saving rather than a penalty avoidance. And SEASONAL or TIME-OF-DAY rates make $r_d$ different in summer and winter, or on-peak and off-peak, so the peak that counts is the largest one inside the charged window and not necessarily the largest one of the month.

One last thing worth doing with this page: divide the demand line on an actual bill by the demand register reading and see whether you get the headline rate. Very often you do not, and the gap is a ratchet, a minimum billing demand, a two-block demand structure or a power-factor adjustment — each of which is a separate conversation to have with the supplier.