Powerwall 3 vs 2 for Demand Charges: What Actually Changed

Most Powerwall comparisons focus on backup time and solar self-consumption. Almost nobody talks about demand charge performance. That's a problem, because demand charges are where homeowners on SRP E-27, APS Time-of-Use with Demand Charge, or California TOU plans actually get hurt the most.

The difference between a PW2 and a PW3 for demand management comes down to one number. Not the battery capacity (both units are 13.5 kWh), and not the warranty. It's the continuous output.

The Number That Actually Matters

Powerwall 2 delivers 5 kW continuous. It can spike to 7 kW for a few seconds, but sustained output is 5 kW.

Powerwall 3 delivers 11.5 kW continuous. That's more than double.

Here's why that matters for demand charges: your Powerwall can only reduce grid draw by as much as it can output. If you're pulling 9 kW from the grid and your battery maxes out at 5 kW, the best you can do is bring that down to 4 kW. The Powerwall isn't failing — it's just physically limited. Demand charge math doesn't care about your intentions.

On SRP's E-27 plan, summer on-peak demand charges typically run in the $6–8 per kW range (SRP adjusts rates annually — check your current bill for the exact figure). APS's demand-charge plan — Time-of-Use 4pm–7pm with Demand Charge — runs steeper: $19.585/kW in summer at current published tariffs. That 4 kW output gap can cost you $24–32/month on SRP, or close to $78/month on APS at current rates — and no amount of automation fixes it, because it's a hardware ceiling.

Three Real-World Scenarios

Scenario 1: Summer Evening AC + Cooking

This is the most common summer spike. AC unit running at peak load (around 4–5 kW), oven or range adding 2–3 kW. Total draw: 7–8 kW.

• With PW2: Battery outputs 5 kW, grid sees 2–3 kW. Peak demand stays manageable.
• With PW3: Battery outputs 7–8 kW, grid sees under 1 kW. You're essentially invisible to the meter during this window.

PW2 handles this scenario. Not perfectly, but well enough that demand charges stay reasonable.

Scenario 2: AC + Dryer + EV Charging

This is where PW2 hits its ceiling. AC at 4 kW, dryer at 5 kW, Level 2 EV charger at 7.2 kW. That's a 16 kW draw before accounting for anything else in the house.

• With PW2: Battery outputs 5 kW, grid still sees 11 kW. At SRP E-27 summer on-peak rates (roughly $7/kW — verify your current rate schedule), that's a demand charge around $77 from one appliance combination.
• With PW3: Battery outputs 11.5 kW, grid sees 4.5 kW. Demand charge drops to roughly $32 at the same rate. Same household, same behavior, a $45+ difference per month.

The PW3 doesn't eliminate demand charges in this scenario either (nothing short of a 16 kW battery could), but it cuts them roughly in half.

Scenario 3: Two PW2s vs One PW3

PW2 is no longer sold new by Tesla, but a lot of homes have existing installs and the output comparison still comes up. If you're stacking two PW2s, your combined continuous output is nominally 10 kW (2 × 5 kW), though actual usable output may vary depending on your Backup Gateway configuration and wiring topology. The PW3 edges that out at 11.5 kW — not a dramatic gap.

The cost comparison is murky because PW2 is discontinued: secondhand pricing is inconsistent and there's no reliable installed cost to anchor against. For demand management alone, if your load profile tops out around 9–10 kW during peak, the two configurations perform similarly on paper. But chasing a discontinued product complicates the decision in ways that go beyond the output numbers.

Where two PW2s still make sense: homes that already have them and want to add capacity, or situations where a longer outage buffer matters more than raw output. For pure demand management on a new install, one PW3 is the cleaner path.

Battery Capacity: Same, and It's Not the Constraint

Both units are 13.5 kWh. A lot of buyers fixate on this number because it's the easy one to compare. For backup purposes, capacity matters a lot. For demand charges, it's secondary.

What matters is whether your battery can discharge fast enough during the peak window to offset your demand. SRP's on-peak window is 3–8 PM on weekdays — 5 hours. APS's Time-of-Use 4pm–7pm with Demand Charge plan runs 4–7 PM weekdays, which is 3 hours. The math below shows why neither window is a capacity problem.

At 5 kW continuous, a PW2 can deliver its full 13.5 kWh over about 2.7 hours. At 11.5 kW, a PW3 could theoretically drain in 70 minutes, but demand management doesn't ask it to do that. Grid Getter's DemandGuard spreads discharge across the full peak window, so you're not burning through the battery in the first hour.

13.5 kWh clears both peak windows with room to spare. The question is always how fast you can put it out when the load spikes.

What This Means for DemandGuard

DemandGuard (Grid Getter's automation that monitors your real-time demand and dispatches the Powerwall to hold below a target threshold) works with both PW2 and PW3. The mechanics are identical. What changes is how aggressive you can set the target.

With a PW2, you need to account for the 5 kW output ceiling. Setting a demand limit of 2 kW sounds great, but if your baseline load regularly hits 6–7 kW, the battery physically can't hold that limit. A realistic target on PW2 is 4–5 kW, depending on your load profile.

With a PW3, a 2–3 kW demand limit is achievable for most homes. The battery can respond to spikes fast enough and sustain enough output to keep grid draw genuinely low across the full peak window.

This isn't a knock on PW2. A well-configured DemandGuard automation on a PW2 can still deliver meaningful monthly savings on SRP E-27, APS Time-of-Use with Demand Charge, or similar demand-based plans — run the rate math above against your own recorded demand peaks to get a number you can actually trust. It's just that PW3 gives you more room to work with.

Should You Upgrade?

Run this specific analysis on your own bills before deciding anything:

1. Pull your last 12 months of utility bills. Find your recorded demand for each month.
2. Estimate what your demand would have been if it were capped at 5 kW. Sum the difference.
3. Multiply that by your utility's current demand rate.

If that number is under $100/month, a PW2 with good automation gets you most of the way there. If your demand regularly spikes above 8–9 kW during peak — especially in Arizona summers when the AC is running hard and you've got an EV — the PW3's output ceiling starts paying for itself.

The upgrade math for existing PW2 owners is harder. You're not comparing $0 to a new unit. You're comparing marginal demand savings against the cost of swapping hardware. For most PW2 owners, optimizing automation first is the right move.

For new buyers in Arizona or California: if you have a large home with central AC and an EV charger, the demand math typically pencils out for the PW3. You won't hit the output ceiling on hot summer afternoons.

Check Your Own Numbers First

Before making any hardware decision, you need your actual recorded demand peaks — not guesses, not utility averages.

Grid Getter's free tier at gridgetter.com/free/ pulls in your Powerwall usage data and surfaces your historical usage patterns, including when your demand tends to spike relative to your utility's peak window. You'll see exactly when those spikes happen and how often they'd push past the PW2's 5 kW ceiling.

Spend five minutes on that before spending five figures on hardware. Your July and August bills from last year are the most important data points in this decision. Pull those first and look for the recorded demand line.