NZ regional residential space-heat demand + P90 cold-year band (compiled synthesis)
Compiled research — upstream-referenced; gross delivered energy, actual behaviour
A Perplexity Deep Research synthesis. Figures are delivered energy (all fuels, gross input) reflecting NZ’s partial/spot-heating culture (households heat the living room a few hours each evening, not the whole house) — so they sit below health-standard (WHO 18°C) need yet above useful-heat in solid-fuel regions (≈65% burner efficiency). Confidence varies: HEEP-direct/EECA regions HIGH; HDD-interpolated regions ±25–30%. Tag the model’s
heat_space_kwh_hhaccordingly.
Summary
Annual residential space-heating demand (kWh/household/yr) for all 16 NZ regions, with a cold-year (P90) uplift, for the NI heat-demand parameter. National average ~3,220 kWh/hh/yr — low by international standards (UK ~12,000) because of NZ’s temperate climate, partial-heating culture, and near-absence of central heating (~5% of homes vs 87% UK). Demand runs ~2,300 kWh (Northland) to ~7,460 kWh (Southland), driven by heating degree days but amplified by a South Island shift from spot-heating to whole-home solid-fuel heating. The Tasman pilot sits at ~2,900 kWh/hh/yr (P90 ~3,200). Anchored on BRANZ HEEP (SR155/SR221), EECA Electric Homes 2024, EECA EEUD 2021, and NIWA HDD normals. Fills RT_179 (② heat_space_kwh_hh + heat_space_cold_year); feeds D01. (The HEEP SR221 primary is now ingested as OT_037 — it reports 34% space-heat of 11,410 kWh/dwelling ≈ 3,880 kWh/hh, consistent with the 3,820 kWh/hh used here before the 0.843 modernisation factor; ⚠ 2005-vintage, which is exactly why the modernisation discount is applied.)
Key claims
- claim: "NZ national average residential space-heating demand is ~3,220 kWh/household/yr (EECA EEUD 2021: 21,332 TJ across ~1.84M dwellings); BRANZ HEEP measured 3,820 kWh/hh/yr in 2000–05, reduced to modern values by a 0.843 factor (heat-pump uptake, insulation retrofits)."
source_location: "§4 National Baseline Derivation"
- claim: "Per-region typical (and P90 cold-year) space-heat demand in kWh/hh/yr: Northland 2,300 (2,540); Auckland 2,690 (2,970); Waikato 2,385 (2,630); Wellington 3,200 (3,530); Tasman 2,900 (3,200); Nelson 2,850 (3,140); Canterbury 3,990 (4,400); Otago 6,380 (7,030); Southland 7,460 (8,230)."
source_location: "§7 Per-Region Results / §10 Quick-Reference"
- claim: "The P90 cold-year uplift factor is 1.10 (uniform across regions): a 1-in-10 cold year needs ~10% more space-heating energy, derived from ~10% coefficient of variation in annual HDD × an ~80% climate-sensitive demand fraction."
source_location: "§8 P90 Cold-Year Uplift Factor"
- claim: "NZ households practise partial/spot heating: ~85% heat the main living area daily in winter but only 24–34% heat bedrooms regularly (41% never); HEEP found 28% of houses failed to reach 16°C in living rooms on winter evenings — so the measured kWh reflect under-heated actual behaviour, below the WHO-recommended 18°C."
source_location: "§1 Executive Summary / §9.1"
- claim: "Figures are delivered energy, all fuels, gross input: solid fuel is 56% of gross space-heating energy nationally but only 45% of delivered heat (~65% burner efficiency). Deploying whole-house heat pumps could increase per-dwelling demand short-term as households heat previously un-heated rooms."
source_location: "§9.2 Fuel Mix and Efficiency"
- claim: "NZ space heating is ~32% of ~10,000 kWh/yr household energy — low internationally (UK ~12,000 kWh / 54%; Canada ~18,000 / 45%); only ~5% of NZ homes have central heating vs 87% in the UK."
source_location: "§9.3 International Benchmarks"
- claim: "Dwelling basis for all figures: detached single-family house (~84% of stock), ~120 m² floor area, mixed stock (~1977 median build year), ~2.7 persons; HDD-interpolated regions carry ±25–30% uncertainty."
source_location: "§2 Dwelling Basis / §11 Confidence"Regional space-heat demand — model data (② heat_space_kwh_hh)
Per-region typical and P90 cold-year space-heat demand (kWh/hh/yr), with confidence, all from CR_015:
| Region | Typical | P90 (×1.10) | Confidence / basis |
|---|---|---|---|
| Northland | 2,300 | 2,540 | moderate — HDD-derived |
| Auckland | 2,690 | 2,970 | high — HEEP direct |
| Waikato | 2,385 | 2,630 | high — HEEP direct |
| Bay of Plenty | 2,550 | 2,810 | moderate — HDD-derived |
| Gisborne | 2,500 | 2,760 | low–moderate |
| Hawke’s Bay | 2,510 | 2,770 | moderate |
| Taranaki | 2,450 | 2,700 | moderate |
| Manawatū-Whanganui | 2,350 | 2,590 | moderate |
| Wellington | 3,200 | 3,530 | moderate — HEEP+EECA blended |
| Tasman (pilot) | 2,900 | 3,200 | low–moderate — Nelson Aero proxy |
| Nelson | 2,850 | 3,140 | low–moderate |
| Marlborough | 2,800 | 3,090 | moderate |
| West Coast | 2,400 | 2,650 | low–moderate |
| Canterbury | 3,990 | 4,400 | moderate — EECA index |
| Otago | 6,380 | 7,030 | high — EECA+HEEP |
| Southland | 7,460 | 8,230 | moderate — HDD-scaled |
| National avg | ~3,220 | ~3,550 | EECA EEUD 2021 |
Neobiome Intelligence relevance
Fills heat_space_kwh_hh and heat_space_cold_year (②). This resolves the §9 open data decision in the spec (HDD method vs published figure) — the report does both: HEEP/EECA published anchors where they exist, HDD-interpolation elsewhere. Pilot (Tasman): 2,900 kWh/hh/yr, P90 3,200, tagged assumed/derived (Nelson Aero proxy, HDD-interpolated, ±25–30%). HEEP-direct regions (Auckland, Waikato, Otago) can be tagged sourced; the rest assumed.
Closes the heat loop with CR_014. Heat demand × (1 ÷ COP) = the electricity load the heat pump adds. Pilot: 2,900 kWh space heat ÷ SCOP ~3.0 ≈ ~970 kWh electricity/hh for space heating if fully heat-pumped (before the building_standard multiplier and water heat). This is the cross-domain term the generation sizing must cover.
Two opposing biases to carry (don’t double-count). The figures (a) under-state health-standard demand because NZ under-heats (28% below 16°C) — a community heating to WHO 18°C whole-house would use more; and (b) over-state useful heat in solid-fuel regions because they are gross fuel input at ~65% burner efficiency. For an MVP “standard build, actual-behaviour” baseline they are the right anchor; the D17 building-standard multiplier (RT_185, pending) and any “heat to health standard” scenario are the explicit levers on top.
Research targets
Documents to retrieve
- BRANZ HEEP (SR155 Year 10 / SR221 Final) — primary measured regional space-heat breakdowns behind the HEEP-direct figures. Already tracked as RT_095. No new target.
- EECA Electric Homes Technical Report (2024) — source of the regional space-heating index (Auckland 74% … Dunedin 198%). Already RT_096. No new target.
- BRANZ HEEP2 (SR495 + full energy results, ~2025–26) — will supersede HEEP1 regional estimates, especially Canterbury (post-EQ), Wellington, South Island. Already tracked as RT_098. No new target.
Research gaps
- Gross-input vs useful-heat — the model’s heat-demand cell wants useful thermal energy; these figures are gross fuel input. A burner-efficiency discount for solid-fuel-heavy regions (Waikato, Canterbury, Otago, Southland) would refine them. Method note, not a retrieval target.
- Sub-regional cold pockets — representative-station HDD understates inland/alpine areas (Tasman inland valleys colder than Nelson Aero; Central Otago colder than Dunedin). Parcel-level refinement deferred to the GIS layer.
Connections
Links to
Referenced by