Most facilities teams read the energy line on their bill. The line that hurts is the one underneath it — a charge set by the single busiest 15 minutes of the month, whether it happened once or every day.
The demand line on your bill
On a commercial tariff you pay twice: once for every kWh you use, and once for the highest rate at which you used it. The utility meters your load in 15-minute blocks, takes the worst block of the month and multiplies it by a demand rate.
At Harbor Logistics that rate is $15.50 per kW. In August 2024 the cold docks, chargers and HVAC overlapped on a hot Tuesday afternoon and set a 1,840 kW peak. That one quarter-hour cost $28,600 — more than 30% of the month’s bill.
$15.50 / kW
Demand rate on the Ashford commercial tariff, per month
1,840 kW
Highest 15-minute average in August 2024
$28,600
Demand charge for that one month
Why solar alone barely touches it
A roof array is excellent at the energy line. It is unreliable at the demand line, because the demand charge only needs one cloudy quarter-hour — or one late-afternoon spike after the panels fade — to reset the peak.

In our first model, solar alone cut Harbor’s August peak by just 9%. The busiest block simply moved from 14:00 to 17:15. The energy saving was real; the demand saving was almost nothing.
“Solar lowers how much you buy. Storage lowers how fast you buy it. The demand charge only cares about the second.”
Owen Reyes · Head of Engineering
Sizing the battery to the shoulder
The battery does not need to run the building. It needs to cover the gap between what the grid should supply and what the site wants — only during the shoulder, only on the worst days.
Three numbers we pull from interval data
- The shoulder peak. Your highest 15-minute reading after solar output starts to fall — usually 16:00 to 20:00.
- Its duration. How many hours that shoulder lasts on the worst weekday of the month. That sets the kWh.
- The ratchet. Some tariffs bill 80% of your annual peak for eleven months. One bad day then costs you a year.
| Battery | Peak shaved | Saved / month | Payback |
|---|---|---|---|
| 250 kWh / 125 kW | 125 kW | $1,940 | 8.9 yrs |
| 500 kWh / 250 kW | 230 kW | $3,570 | 6.4 yrs |
| 1 MWh / 500 kW | 290 kW | $4,500 | 9.6 yrs |
The payback maths
The 500 kWh battery shaves about 230 kW from the shoulder peak, worth $3,570 a month at $15.50 per kW. It also shifts midday solar into the evening, which adds roughly $9,000 a year on the energy line. Against an installed cost of $340,000, simple payback lands at 6.4 years.
Inside a PPA the client never sees that payback — it is priced into the rate. Owning the battery outright makes sense when the tariff has a ratchet; leasing makes sense when it does not.
Run it on your own bill. We will send the worked model with Harbor’s data swapped for yours — send us 12 months of interval data.






Our tariff has a 12-month ratchet at 75%. Does that change the sizing logic, or only the payback?
Both. With a ratchet we size for the worst day of the year, not the month, so the kWh usually goes up by 20–30%. Payback improves because every avoided kW is saved twelve times. Send us the interval data and we’ll run it.
Useful table. Would be good to see the same numbers for a lease, where the battery stays off our balance sheet.