It charges for speed, not distance
Energy charges (c/kWh) price the total you use. The demand charge prices the fastest you ever draw it — the highest 30-minute average, measured in kW or kVA, inside the tariff’s demand window.
NSW commercial tariffs · explainer
Why one bad hour can set the whole month.
A demand charge is a fee for the fastest your site ever draws power, not the total it uses. On most NSW commercial tariffs, the highest 30-minute interval inside a set window becomes the charge for every day of the billing month — which is why the bill can stay high in a month you barely used anything.
How it actually works
Energy charges (c/kWh) price the total you use. The demand charge prices the fastest you ever draw it — the highest 30-minute average, measured in kW or kVA, inside the tariff’s demand window.
On most NSW demand tariffs there is no averaging. The single highest qualifying interval becomes the billed demand for every day of the month — a short, sharp peak costs the same as running at that level all month.
Each tariff defines when demand is measured — commonly weekday afternoons and evenings. A peak outside the window may not count at all; a peak inside it sets the bill. The window matters more than the total.
kW or kVA? Some tariffs bill real power (kW); most large-business tariffs bill apparent power (kVA), which also counts reactive load from motors and compressors. If your bill is in kVA and your power factor is poor, part of the “peak” is electrical inefficiency — a different, cheaper problem to fix. The bill shows which basis applies.
The arithmetic
N19 bills the highest 30-minute kVA reading between 16:00 and 20:00 on business days. High-season network demand rate (November–March, GST-inclusive): 58.421 c/kVA/day.
50 kVA × 58.421 c/kVA/day × 31 days ÷ 100 = $905.53A site whose January peak is 50 kVA higher than it needs to be — one compressor bank starting during the evening window — pays about $905 for that interval alone. Held every month, roughly $10,100 a year across the seasonal rates.
Network rates are not retail quotes: the retailer’s pass-through decides what your bill does. And on a kVA tariff, confirm how much of the peak is reactive before counting the full reduction.
See the current N19 rates and windowsReducing it
The cheapest fix is operational: stagger compressors, ovens, pumps or chargers so they do not coincide. Free to test, and interval data shows whether the peak is a scheduling accident.
kVA includes reactive power that does no work. Where power factor is poor, correction equipment can cut billed kVA for a fraction of any battery’s cost. Check the bill basis before anything else.
Some sites sit on a demand tariff with a load shape that would cost less on time-of-use energy pricing — or the reverse. The assigned network tariff is printed on the bill and can sometimes be changed.
A battery that reliably covers the qualifying interval lowers the measured maximum. This works when peaks are short and predictable — and fails quietly when they are long or random.
What the battery physically does, in the worked example above: it sits between the meter and the site, and discharges through the 16:00–20:00 window so the grid never sees more than a set limit. Erasing that 50 kVA January peak means sustaining roughly 50 kW of discharge for as long as the peak lasts — and that duration is the kWh you are really buying. A unit that covers the first hour and then runs flat leaves the peak, and the $905, intact.
A battery is the fourth lever, not the first. The dispatch reliability question — can it cover the qualifying interval every time — is what separates a real saving from a brochure number. How battery peak shaving works.
Worth ruling out
Quick answers
The next step
The assigned network tariff, the demand window and twelve months of interval data are enough to say which lever applies — and whether a battery belongs in the conversation at all.
Rate sources
General information only. Network and retailer rates change; the retailer’s contract decides what a site actually pays.