Powerwall on APS’s demand-charge TOU: a different game
In plain English: APS’s time-of-use plan with a demand charge — part of the bill is set by your single worst hour of grid draw.
Arizona’s APS R-3 pairs modest energy rates with a demand charge measured in your worst 4–7 PM hour. The battery’s job here isn’t buying low and selling high — it’s making your peak hour look boring.
The plan at a glance
| Period | Hours | Rate |
|---|---|---|
| Winter energy | Layered off/mid-peak | $0.035–0.099/kWh |
| Demand charge | Billed on your peak 4–7 PM weekday draw | Per-kW, the real bill driver |
Energy rates are modest; the demand charge is where APS bills are won and lost. Rates come from the published tariff as reflected in SolDial's rate database in mid-2026 — always verify current numbers with your utility.
What this plan means for a Powerwall
With energy spreads of a few cents, classic arbitrage barely moves an APS bill. The demand charge does: it’s billed on your single highest draw during the 4–7 PM weekday window, so one hour of A/C + oven + EV charger overlapping can set your whole month’s charge. The battery’s highest-value behavior is peak shaving — discharging during 4–7 PM not to save on energy, but to cap the demand meter.
Settings: reserve low enough that the pack always has multiple kWh of working room entering the window; never let EV charging overlap 4–7 PM (a 10 kW charge inside the window can cost more in demand than a month of energy shifting saves); and keep the battery covering the house through the window every weekday.
The automations worth having
A hard EV block for the demand window; a pre-window top-up so the pack enters 4 PM full; an alert if household draw spikes during the window anyway; and storm-watch override. Demand-charge strategy is unforgiving of one bad hour — which is exactly why it belongs to software, not memory.
Does this apply to my exact situation?
Plans vary by zone, season, and enrollment vintage, and utilities revise rates. SolDial matches your account to the actual published schedule from ~50,000 tariffs and recomputes the strategy — and the honest savings math — from your windows, not a blog’s snapshot. If your plan differs from this page, the app follows your plan.