Calculator
Peak Shaving and Demand Charge Calculator
For a commercial site, the electricity bill has two parts: the energy consumed and the highest half-hour of demand recorded in the period. The second is often a quarter of the bill and responds to a battery far more efficiently than the first.
This calculator sizes the storage needed to hold demand below a target and values the reduction against the tariff's capacity charge.
Enter the demand profile
Power the battery must supply
25 kW
After 15 kW of solar support
Energy per peak event
50 kWh
Nominal battery capacity required
55.6 kWh
At 90% usable depth
Annual demand-charge saving
€3,840
Estimated storage capex
€22,222
Simple payback on demand charges alone
5.8 years
Energy arbitrage and resilience add further value
Annual round-trip losses
864 kWh
How the calculation works
Demand charges bill the single highest measured interval, so the battery only needs to cover the events that set that maximum — typically a handful of half-hours a month, not continuous operation.
Two ratings matter. The power rating must cover the difference between the site peak and the target, less whatever the solar array reliably contributes at that hour. The energy rating must cover that power for the full duration of the event, with margin for consecutive peaks.
The saving is the reduction in billed demand multiplied by the charge and by twelve months. Unlike energy arbitrage, it does not depend on how often the battery cycles, only on never missing a peak.
The formula
P_batt = P_peak − P_target − P_solar E_event= P_batt × t_event C_nom = E_event / DoD Saving = (P_peak − P_target) × charge × 12 Payback= C_nom × cost_per_kWh / Saving
Assumptions behind the numbers
- The peak duration and frequency are known from interval data. Estimating them from monthly bills is unreliable.
- Solar contribution during peaks is firm. If the peak occurs on a cloudy afternoon or after sunset, set it to zero.
- The control system predicts and responds to peaks correctly. A single missed event resets the month's billed maximum and destroys that month's saving.
- No ratchet clause is modelled. Some tariffs bill a percentage of the highest peak in the previous eleven months, which makes a single miss far more expensive.
- The battery is available at full state of charge before each event, which requires charging outside peak windows.
How to read the result
Demand-charge savings alone rarely justify storage below about €6/kW/month. Above €10/kW/month, and especially with a ratchet, the case becomes strong.
Stack the value streams. The same battery can shave demand, arbitrage a time-of-use tariff, absorb solar that would otherwise be curtailed and provide backup — payback assessed on any one stream in isolation is misleadingly long.
The most common failure is not technical but contractual: sites shave demand successfully, then a process change moves the peak to a new time the controller does not anticipate.
Worked example
A factory bills 180 kW peak against a €8/kW/month capacity charge, wants to hold 140 kW, and its peaks last two hours about twelve times a month. Solar reliably contributes 15 kW during those hours.
The battery must supply 180 − 140 − 15 = 25 kW for two hours, or 50 kWh per event, which needs about 56 kWh nominal at 90% depth of discharge. At €400/kWh installed that is roughly €22,200.
The saving is 40 kW × €8 × 12 = €3,840 a year, giving a 5.8-year payback on demand charges alone. Adding evening energy arbitrage and avoided solar curtailment typically brings that under four years, which is why these systems are usually specified for stacked value.
Mistakes that skew the answer
- Sizing from monthly bill totals instead of half-hourly interval data, which hides both the height and the duration of real peaks.
- Ignoring consecutive peak events that leave no time to recharge between them.
- Assuming solar output during the peak when the peak is set on the worst weather day of the month.
- Overlooking a ratchet clause, which makes one missed event expensive for a year.
- Setting the target so aggressively that the battery is exhausted before the event ends, achieving nothing.
Questions readers ask
- What is a demand charge?
- A charge based on the highest rate of consumption in a billing period, usually measured over a half-hour or fifteen-minute interval, and billed in euros per kilowatt per month. It exists because networks must be sized for peak flow, not average flow.
- How much peak reduction is realistic?
- Typically 15 to 30% of the billed peak, limited by the battery power rating and by how predictable the peaks are. Beyond that the storage becomes large enough that the capital cost outruns the saving.
- Can solar alone shave the peak?
- Only if the peak coincides with strong sun and the array is large. Because the charge is set by the single worst interval, an intermittent resource cannot guarantee the reduction — one cloudy peak restores the full charge for that month.
- What control strategy is needed?
- A predictive controller that watches the rolling interval average and discharges before the threshold is crossed, not after. Reactive control consistently misses the first event, which is often the one that sets the bill.
- Does peak shaving shorten battery life?
- Less than daily arbitrage does. Peak shaving typically uses 100 to 200 shallow cycles a year at moderate C-rate, well within the warranted throughput of a commercial lithium iron phosphate system.