Where the $720 / yr leak in a typical NZ bill goes.

MBIE publishes the Quarterly Survey of Domestic Electricity Prices (QSDEP) ^[1], the reference figure for what the typical New Zealand household pays for electricity. As of the last published edition, an MBIE modelled consumer using around 22 kWh per day (about 8,000 kWh per year) faces an all-in retail rate in the high 30s of cents per kWh. The all-in annual bill lands close to $2,400 a year, varying by region.

The build-up below works from this baseline. The three leaks are subtracted in sequence so the numbers do not interact, and so a reader can audit any line individually.

Consumer NZ ^[2] publishes a rolling median saving for households that use the Powerswitch tool^[3] to find the cheapest plan available at their address. The published median has run between roughly $350 and $524 a year over the past two years, with $400 a fair midpoint.

• What it captures.The gap between the household's current retail plan and the cheapest comparable plan in their region.
• What it does not.It does not capture savings from changing plan structure (e.g. flat to TOU). That sits in section 04.
• Population.Powerswitch users are self-selected, so the published median is conditioned on households that already suspect they can save. A randomly chosen NZ home likely captures less. The build-up therefore treats the $400 as an addressable maximum, not a guaranteed minimum.

The figure refreshes monthly. The number cited on this page will track the latest published Powerswitch median, with the reporting window noted in the source list.

BRANZ Household Energy End-Use Project (HEEP)^[4] and the companion ACEEE 2006 paper^[5] measured the continuous baseload of New Zealand homes. Typical residential baseload sits in the 100 to 300 watt range continuously, which is 876 to 2,628 kWh a year before any appliance is deliberately turned on.

• Why baseload is not all ghost load.Fridges, freezers, modems, alarms and standby clocks make up the non-negotiable share. Only the discretionary slice (TVs and consoles in standby, sound systems, chargers without devices, second fridges, set-top boxes, idle inverters) is addressable.
• The assumed discretionary fraction.Literature ranges from 40% to 60% of measured standby load. This page uses a conservative 40% for the headline number.
• The calculation.Take a 200 W median baseload, multiply by 8,760 hours and current retail rate. That gives roughly $700 a year of baseload spend. Apply the 40% discretionary fraction. Result: about $150 / yr addressable.

Discovering ghost loads is the lever that most directly requires per-circuit visibility. A whole-house meter shows a household uses 200 W overnight; only a circuit-level view tells you which 200 W to turn off.

Several NZ retailers offer time-of-use plans with materially cheaper off-peak rates: Electric Kiwi "hour of power", Contact "Good Nights" overnight free, Mercury "free hour"^[6]. The spread between peak and off-peak on these plans is large enough that running flexible appliances inside the off-peak window produces a meaningful saving.

The flexible-load population in a typical NZ home:

• Hot water cylinder.About 3,000 kWh a year, ~30% of the average bill. Assume 70% of that is shiftable into off-peak windows without affecting hot-water availability.
• Heat pump heating.About 1,500 kWh a year in heating-dominant months. Assume 30% shiftable (some heating must happen during peak hours).
• Dishwasher, laundry, dryer.About 600 kWh combined. Assume 80% shiftable.
• EV charging (where present).About 2,500 kWh a year per EV. Assume 95% shiftable to overnight. Excluded from the headline so the number applies to the median home without an EV.

Shiftable kWh totals roughly 2,500 kWh a year for a non-EV household. On a TOU plan where the off-peak window discounts or zeros the rate for the shifted kWh, the annual saving lands near $170. EV households see materially more (the EV alone contributes another ~$250 a year at current spreads).

Like ghost loads, this lever depends on knowing which circuits are flexible. A whole-house meter cannot tell you which 2,500 kWh to shift.

Each lever is applied to the post-previous-lever residual, not to the gross bill. Step 1 switches the household to the cheapest flat or flat-equivalent plan. Step 2 removes discretionary baseload from the resulting consumption. Step 3 moves the remaining flexible loads into the cheapest available TOU windows.

• Why no double-count.Step 1's plan switch is to a flat-equivalent plan, so step 3's TOU-shift saving is purely incremental. Step 2's ghost-load removal is energy that is no longer drawn, so it does not show up in step 3's shiftable kWh.
• Median vs. sum-of-medians.The build-up sums median values across three independent levers. A real median household typically captures the median in one lever and less in the other two. The $720 is best read as the addressable maximum for a household that takes all three actions, not as a guaranteed median outcome.
• Status.Modelled from primary NZ sources (MBIE^[1], Consumer NZ Powerswitch^[2][3], BRANZ HEEP ^[4], NZ retailer rate cards ^[6]), not measured against Basis customer bills. As Basis fleet data matures, the page will publish a measured replacement with confidence intervals.

• Plan.The Basis app compares your home's actual usage profile against every retail plan in the market, refreshed continuously. One tap to switch, or set it to alert when a cheaper plan becomes available.
• Ghost loads.Per-circuit measurement at 10,000+ samples per second. Each ghost-load circuit shows up as a named line item in the app with a dollar value, so the homeowner can act on the biggest first.
• Timing.Dispatch automation shifts hot-water heating, EV charging, and other flexible loads into the cheapest windows on the household's tariff. No timers, no manual scheduling.