Four levers actually move this line on the bill, and cutting your consumption is not one of them. Which lever applies to your site is a question your interval data answers, not a brochure.
Full detail below.
It is the most common assumption and it is close to useless here. The demand charge is calculated from the single highest interval in the billing period. Switch equipment off for three weeks and you will reduce the energy charge substantially and the demand charge hardly at all, because the peak that set it happened in one quarter hour you did not change.
Everything that works operates on the height of the peak rather than the volume underneath it.
On a grain operation in the south of the province, demand charges made up 43.7% of a $75,389 annual electricity bill. A 15 kW / 30 kWh battery, tiny next to the site's 150 kW array, cut the annual demand line by roughly 20% and paid back its $3,770 net cost in under a year on demand savings alone. Read the full case study.
It is held by your distributor and available on request at no cost. Fifteen minute resolution over twelve months tells you which intervals set each month's billing demand and what plant was running during them. Without it, every recommendation that follows is a guess dressed up as engineering.
Check whether your bill is being driven by highest metered demand, by contracted demand, by a rate minimum, or by a ratchet carried forward from an earlier peak. These respond to completely different interventions, and it is entirely possible to shave the metered peak beautifully and see no change on the invoice because a contracted level was setting the charge all along.
The cheapest fix by a wide margin. If two large processes start within the same quarter hour and either one could move, moving it costs nothing and removes the peak permanently. A surprising share of commercial peaks turn out to be habit rather than necessity.
If your demand charge is denominated in kilovolt-amperes rather than kilowatts, a poor power factor is inflating it directly, and capacitor correction is far cheaper than storage. Diagnose the unit on your bill before anyone quotes you a battery.
Whatever peak remains after scheduling and correction is the target for a battery, and only that. Storage sized against the remaining unschedulable peak is a small, cheap system. Storage sized against a solar array or against a vague sense of resilience is a large, expensive one that may not move the demand line at all.
If capacity charges are billed against a contracted level set years ago for load you no longer run, sustained lower peaks are the evidence base for renegotiating it downward with the distributor. This is a paperwork saving rather than an equipment one, and it is routinely overlooked.
If your interval data shows a peak created by two schedulable processes overlapping, we will say so, and the fix is a conversation with your operations lead rather than a purchase order. If you are billed on kVA with a poor power factor, correction hardware costs a fraction of storage and does more. If your account sits with a distributor that does not levy demand charges on your customer class, there is nothing here to shave at all, and that is worth knowing before anyone models a system for you.
Storage earns its place when the remaining peak is short, sharp, and cannot be moved. That is a real and common situation, and it is also a narrower one than most sales material implies.
There are four practical levers: reschedule the loads that collide to flatten the peak, correct power factor if you are billed on kVA, install storage that discharges during your highest intervals, and renegotiate contracted demand if your tariff bills capacity charges against a level you no longer need. Which one applies depends on what your interval data shows is creating the peak.
Barely. The demand charge is set by the single highest interval in the billing period, so reducing consumption across the rest of the month leaves it largely untouched. Conservation lowers the energy charge. Only flattening the peak lowers the demand charge.
On a completed project in the province, storage cut the annual demand line by roughly 20%, from $32,958 to $26,319. Single summer months fell much further, in one case from 134 kW down to 64 kW. Winter months barely moved because the load was a sustained plateau rather than a spike. The realistic answer for any given site depends on the shape of its peaks, not on a percentage anyone can quote in advance.
From interval data, against the height and duration of the peaks that are setting billing demand, not against the size of any solar array. The completed project referenced on this site pairs a 150 kW array with only a 15 kW / 30 kWh battery, because the job was covering specific quarter hours rather than shifting bulk energy.
Yes, and the value is usually higher, because preventing a spike prevents the floor that spike would have set under subsequent bills. The Utilities Consumer Advocate illustrates a ratchet at 85% of an earlier peak. Under that structure one unshaved interval raises many months of invoices, not one.
Get your interval data from your distributor and find out which intervals set your peak and what was running during them. Everything else is guesswork until that is on the table, and it costs nothing to request.
