Coolroom running costs, and what peak and demand charges do to them.

Three separate charges, not one

A small business electricity bill is usually built from three parts, and they behave completely differently. Reading your bill as a single number is what makes refrigeration costs feel unmanageable.

• Usage. Cents per kilowatt hour for the energy you actually consumed. On a time of use tariff this rate changes depending on the hour of the day.
• Supply. A fixed daily charge for being connected. Nothing you do to the coolroom changes this one.
• Demand. On a demand tariff, a charge based on your single hardest half hour inside a defined window, billed in kilowatts. This is the one most operators have never had explained, and it is the one refrigeration is worst at.

Usage rewards using less overall. Demand rewards never spiking, even once. They are different problems, and a coolroom is stuck in the middle of both.

The peak windows, by network

Which window applies to you depends on the distributor for your address, not on your retailer. Sydney, the Central Coast, the Hunter and Newcastle are Ausgrid. Western Sydney, the Blue Mountains, the Southern Highlands and the Illawarra are Endeavour. The rest of regional New South Wales is Essential.

Last checked: 14 August 2026. Ausgrid windows from Ausgrid's time of use pricing page. Endeavour and Essential business tariff codes and windows compiled by WATTever. Network windows change, typically on 1 July. Your retailer can also apply its own windows on top of the network's, so the tariff code printed on your bill is the only authority for your site.

Two things in that table are worth sitting with. Ausgrid dropped its separate shoulder rate in 2024 and folded it into off peak, so on that network there are now only two rates and four months of the year with no peak at all. And on Endeavour's newer tariffs the cheapest window is the middle of the day, not the middle of the night, because there is more solar on the network at noon than the network knows what to do with. If your instinct is "run it late", that instinct is now wrong on some tariffs.

Why demand charges punish refrigeration specifically

A demand charge does not care how much electricity you used. It cares about the worst 30 minutes you had inside the window, measured as an average draw in kilowatts, and it bills you for that one half hour as though it described your whole month.

Ausgrid makes the point neatly in its own guidance: a high wattage appliance like a hair dryer is generally not a demand problem, because it only runs for moments inside any given half hour block, and the average over the block stays low. That is exactly the property a compressor does not have. When a coolroom compressor is working hard it is working hard for the entire block, and the average is the peak.

So the loads that set your demand charge are the ones that make the compressor run flat out for a sustained period:

• Defrost cycles and the recovery afterwards. The unit has just added heat to the space on purpose, then has to pull it all back out.
• A big delivery. Doors held open, warm stock loaded in, and the compressor running continuously for as long as it takes to get back to setpoint.
• Restart after a power interruption. Every unit on the site comes back at once, all of them warm, all of them pulling at the same moment. This is the single biggest demand spike most food businesses ever record.
• A door left open, or a failing seal. The compressor never gets to cycle off, so it sits at full draw for the whole window.

Notice that three of those four are not really energy problems. They are operational problems that happen to show up on the bill.

How long one bad half hour follows you

This is the detail that decides how much any of it matters, and it is set by the tariff code on your bill rather than by anything about your site.

• On Ausgrid EA251 and EA256, the demand figure resets monthly. A bad Thursday costs you that month and then clears.
• On Ausgrid EA302, the charge is based on your highest half hour across the previous 12 months. One delivery mishandled on a February afternoon is still on your bill the following January.

If your bill shows EA302, the arithmetic of prevention changes completely, because the saving from avoiding a single spike is annual rather than monthly. That is worth knowing before you decide none of this is worth the effort.

Working out your own number

We do not publish cents per kilowatt hour on this page, for the same reason we do not publish a single price for an alarm system: your retailer sets your rates, they differ between plans, and a number copied off someone else's bill would only mislead you. Ausgrid says the same thing plainly on its own site, that the demand charges you pay are set by your retailer and may not match the network's. What we can give you is the method, which does not go stale.

Take the last bill and find four things:

1. The tariff code. Usually a short alphanumeric like EA251 or BLNT2AL. This tells you your windows and whether demand resets monthly or annually.
2. The peak usage rate in cents per kilowatt hour, and the off peak rate next to it.
3. The demand charge, if there is one, usually shown as dollars per kilowatt for the month, with the kilowatt figure that triggered it.
4. The compressor nameplate from the plant itself, in kilowatts.

Then the rough sum for one coolroom is: nameplate kilowatts, times the hours it actually runs in a day, times the rate. A compressor cycles rather than running flat out, so on a well sealed room in reasonable ambient conditions somewhere around half the day is a common working assumption, and your own runtime data beats that guess every time. Do that once at the peak rate and once at the off peak rate, and the gap between the two answers is the entire prize available from load shifting. For many single room sites that gap is smaller than people expect, and the demand line is where the real money is.

If you want the same treatment applied to the risk side rather than the running side, the cool room failure cost calculator turns your stock value and likely time to notice into what a single breakdown would cost.

What actually reduces the bill

In rough order of how much they return for the effort:

• Move defrost cycles out of the peak window. Most controllers let you set defrost times. If yours are running mid afternoon in February, that is a free change to make.
• Stagger the restart after any outage. Bringing units back one at a time over 20 minutes rather than all at once flattens the largest spike you are likely to record. This matters most on an annual reset tariff.
• Book deliveries outside 3pm to 9pm on weekdays where the supplier allows it, so the recovery load lands in an off peak hour.
• Fix the seals and the door discipline. A door that does not close properly raises both your usage and your demand, all day, every day. It is the least glamorous item here and usually the biggest.
• Pre-cool slightly before the window where the stock tolerates it, so the compressor is doing less work during the expensive hours. Check this against your food safety obligations first, because temperature compliance outranks the electricity bill every time.

What you should not do is raise the setpoint to save money. The margin between a safe coolroom and a food safety incident is not where you go looking for savings, and the temperature records you are required to keep exist precisely because that margin matters.

Where monitoring fits

Everything above depends on knowing what your refrigeration is actually doing, and most sites do not. Temperature logging tells you when a room drifted, how long recovery took after a delivery, whether a defrost cycle is running when you think it is, and whether a door is being left open on a particular shift. Those are the same facts that drive both the food safety answer and the running cost answer, which is why we treat them as one problem rather than two.

What we install and monitor is temperature, door state and power interruptions, with the alerting and the logs described on our what we do page. If your question is specifically about the tariff mechanics rather than the refrigeration, our sister site has the full explanation of how time of use tariffs work, with the peak and off peak windows set out network by network.

Questions people ask