Source
https://www.branz.co.nz/pubs/research-reports/sr237/ — original source (opens in a new tab; the file is not redistributed)
Pollard (2010) — The Energy Performance of Heat Pump Water Heaters (BRANZ SR237)
Vintage & technology — 2010 R-134a-era field units, NOT modern CO₂ HPWHs
SR237 is the only dedicated NZ field-monitoring dataset of heat-pump water heaters, and its value is exactly that: real installed-system COPs, not lab or marketing numbers. But the 11 monitored units are 2010-vintage R-134a integral and split systems — the generation before the CO₂-refrigerant HPWHs that dominate today (SR488’s modern CO₂ unit at ~COP 2.5, OT_038). So SR237’s measured COP 0.53–2.00 (most integral 1.1–1.6) is a real-world field floor and a low-draw-off warning, not the modern design COP the model should carry. Use it to (a) reality-check that installed HPWH performance sits well below the “up to 3 units of heat per unit of electricity” premise, and (b) enforce the ≥140 L/day demand threshold below which HPWH savings collapse — not to reset the DHW heat-pump COP cell (that stays SR488/CR_014).
Summary
BRANZ Study Report SR237 (Pollard, 2010) is a New Zealand field-monitoring study of the energy performance of 11 heat pump water heating (HPWH) systems. Three of the systems carried over from an earlier solar-water-heating project (BRANZ SR188, monitored a full year); the other eight were an audit subsample of a much larger (>160-system) EECA HPWH grant programme, to which BRANZ added detailed data-logging. Systems were located in Auckland or Wellington and monitored for at least 151 days, spanning integral, split (once-through) and split (recirculated) types. The headline result is that measured field coefficients of performance (COP) ranged from 0.53 to 2.00 — integral systems were reasonably consistent (mostly 1.1–1.6), the three once-through split systems performed best, and the single recirculating split system performed poorly (COP 0.53, never exceeding 1). Performance is strongly driven by how much hot water the household draws: to achieve a reasonable COP of 1.5 a household needs at least 140 L/day of hot water, yet 7 of the 11 households used less than that, and cutting the draw-off from a “small” (7.1 kWh/day) to an “extra-small” (3.4 kWh/day) profile dropped the average COP from 1.61 to 1.14 (the report labels this “41% lower”, but that figure is measured off the extra-small 1.14 base — on the 1.61 baseline it is ~29% lower; equivalently the small-draw-off COP is 41% higher than the extra-small). Colder ambient conditions add a further penalty (energy use ~27% higher at 6 °C than 15 °C; Invercargill ~14% less efficient year-round than Auckland). For Neobiome Intelligence, SR237 is the NZ field reality-check for the heat-pump water-heating duty — but as a 2010 R-134a-era dataset it bounds and cautions the DHW heat-pump cells rather than setting them (the modern CO₂ design value stays with SR488).
Key claims
- claim: "The energy performance of 11 heat pump water heating (HPWH) systems was examined by installing data-logging equipment. Three of the systems were carried over from an earlier solar-water-heating project (previously identified as H36, H37, H38, now with a full year's data); the remaining eight were an audit subsample of a much larger (more than 160 systems) but less detailed EECA study of HPWH performance. The systems were located in either Auckland or Wellington, were monitored for at least 151 days, and were of integral, split (once-through) and split (recirculated) types."
source_location: "Abstract (p.i); §3 Data Collection (p.6); §4 Results & Table 2 (pp.7–8)"
- claim: "Measured field COPs of the 11 systems ranged from 0.53 (System 11, a recirculating split system — the poorest) to 2.00 (System 10 — the best). Table 6 measured COPs by system: 1) 1.50, 2) 0.70, 3) 1.16, 4) 1.80, 5) 1.73, 6) 1.63, 7) 1.88, 8) 1.50, 9) 1.14, 10) 2.00, 11) 0.53. The integral systems were reasonably consistent; the three once-through split systems performed better than the integral systems; the one recirculating split system performed poorly, not exceeding a COP of 1 over the extra-low and low draw-off ranges it operated over."
source_location: "§4.3 Table 6 (p.13); §5 Discussion & Conclusions (p.17)"
- claim: "COP reference anchors: an instantaneous electric water heater has a COP of 1 (100% of electrical input converted to heated water for the users). An average electric storage hot-water cylinder has standing losses of around 33% of total water-heating energy (Isaacs et al. 2006), giving a COP of 0.67. A system with a COP of 1.34 would therefore require half as much water-heating energy as a standard electric storage cylinder for a given amount of hot water."
source_location: "§2.3 HPWH performance measures (p.4)"
- claim: "The performance of a HPWH system is reduced when daily water draw-off is low. To ensure a reasonable level of performance (a COP of 1.5) is achieved by a typical HPWH system, the household hot-water demand should be at least 140 L per day. Seven of the 11 households examined had average daily hot-water usage of less than 140 L per day."
source_location: "§5 Discussion & Conclusions (p.17)"
- claim: "Excluding the two systems with little extra-small use (Systems 7 and 10), the remaining eight systems had an average COP of 1.61 for a small draw-off (7.1 kWh) at 15 °C, falling to an average COP of 1.14 (which the report describes as '41% lower' — but that is relative to the extra-small 1.14 base; the drop from the 1.61 baseline is ~29%) when an extra-small (3.4 kWh) draw-off was used in the regression model. As the water draw-off decreases the standing losses become a far more important component of energy use, and many cylinders were located outside (higher heat loss)."
source_location: "§4.4 Linear regression of daily energy use (pp.16–17); Figure 10"
- claim: "At an outdoor temperature of 15 °C the average daily electrical energy input to the 10 modelled systems was 5.0 kWh; when the temperature was lowered to 6 °C this increased 27% to 6.3 kWh."
source_location: "§4.4 (p.15)"
- claim: "The ratio of winter (July) to summer (January) modelled energy use was around 1.26–1.27 for most NZ centres, rising to 1.35 for Christchurch and 1.32 for Hamilton (which have cooler winters). Modelled energy use in Invercargill was 18% higher than in Kaitaia, and the year-round performance of a HPWH system in Invercargill would be around 14% less efficient than an equivalent system in Auckland. Some models of HPWH are only suited to warmer areas such as the North Island (excluding the volcanic plateau)."
source_location: "§4.4 Tables 7–8 (p.16); §5 (p.18)"
- claim: "Hot-water usage levels used in the analysis (assuming 60 °C outlet and 15 °C inlet water): Extra Small 3.4 kWh/day (12.2 MJ/day ≈ 65 L/day); Small 7.1 kWh/day (25.6 MJ ≈ 136 L); Medium 10.8 kWh/day (39.0 MJ ≈ 207 L); Large 14.4 kWh/day (52.0 MJ ≈ 276 L). These small/medium/large draw-offs approximately align with AS/NZS4234:2008."
source_location: "§4.1 Table 3 (p.10)"
- claim: "Water heating accounts for about 29% of total residential energy in New Zealand (Isaacs et al. 2006, HEEP). Prior NZ experimental HPWH studies reported COP ranges of 2.4–3.0 (Carrington et al. 1984), 1.6 (Lloyd & Kerr 2007) and 1.2–3.0 (Whitley 2009); Lloyd & Kerr (2008) estimated the equivalent in-use COP for Carrington's actual installed systems at 1.1–1.7 — somewhat below the experimental performance. The pooled average hot-minus-cold water temperature difference across the 11 systems was 36.3 ± 5.2 °C."
source_location: "§1 Introduction (p.1); §2.4 Table 1 & text (p.6); §4.2 (p.12)"
- claim: "High-efficiency HPWH systems using CO₂ as a refrigerant, popular in Japan (over half a million units sold per year), are around 50% more efficient than models based on R-134a refrigerant and provide water temperatures up to 80 °C, but carried a high price of around NZ$9,000 to NZ$14,000 (2008 figures)."
source_location: "§2.2 HPWH systems (p.3)"Neobiome Intelligence relevance
context: ni — SR237 is the NZ field-performance reality-check for the heat-pump water-heating (DHW) duty. It feeds D01, sitting alongside SR488 (OT_038, modern CO₂ lab-instrumented unit) and CR_014 (CR_014, space-heating SCOP). Its load-bearing caveat is vintage: 2010 R-134a-era units, so it bounds and cautions the DHW heat-pump cells rather than setting the modern design COP.
- D01 — the NZ HPWH field floor and the ≥140 L/day demand threshold. SR237 gives measured installed-system COPs of 0.53–2.00 (most integral units 1.1–1.6) OT_162 — the empirical answer to the “up to 3 units of heat per unit of electricity” marketing premise the report itself flags as incomplete. The decisive NI signal is draw-off dependence: a HPWH only reaches a reasonable COP 1.5 when the household draws ≥140 L/day hot water (7 of 11 monitored households used less), and dropping from a small (7.1 kWh) to extra-small (3.4 kWh) profile cut average COP from 1.61 to 1.14 (the report’s “41% lower” is taken off the extra-small 1.14 base — on the 1.61 baseline that is ~29% lower; equivalently 1.61 is 41% above 1.14) OT_162. For a small remote household or a low-occupancy dwelling this is a real risk that HPWH savings largely evaporate — a feasibility/sizing caution the model should carry on the DHW heat-pump option, not just a headline COP. A cold-climate penalty stacks on top (~27% more energy at 6 °C vs 15 °C; Invercargill ~14% less efficient year-round than Auckland), reinforcing the South-Island / cold-site derating already applied to heat pumps.
- D01 — corroboration of the DHW COP anchors (independent of OT_105/OT_038). SR237 independently restates the reference COPs the corpus already uses — electric storage cylinder COP 0.67 (~33% standing losses), a system at COP 1.34 using half a standard cylinder’s energy OT_162 — the same anchors SR239 (OT_105) gives for SWH, so three BRANZ reports now agree on the baseline. It does not move the DHW heat-pump COP cell: SR488’s modern CO₂ unit (~2.5, OT_038) remains the design value; SR237 is the older-technology field floor beneath it.
Research targets
Documents to retrieve
- None raised. SR237 is the document RT_357 sought; its own references (SR188/OT_084, HEEP/OT_037, AS/NZS4234:2008, the draft AS/NZS5125 HPWH test standard) are already in the corpus or already targeted (RT_356 for the EECA AS/NZS4234:2008 modelled listings). Per the batch “minimise new RTs” discipline, no new documents are raised.
Research gaps
- RT_357 — RESOLVED → OT_162. The named document (BRANZ SR237, Pollard 2010, the HPWH companion to SR239) has been retrieved and ingested. ⚠ Residual (folded into existing targets, not a new RT): SR237’s field data are 2010 R-134a-era units, so while it fully satisfies RT_357’s ask for “the SR237 HPWH field study”, the broader goal of a modern (CO₂-generation) NZ HPWH field dataset is not met — that gap sits with the modern-DHW residuals already tracked on RT_282 (modern SWH/DHW field data) and RT_356 (EECA AS/NZS4234:2008 modelled per-model listings). No new RT raised.
Notes
Authoritative BRANZ publication (BRANZ Study Report SR237, © BRANZ 2010, ISSN 1179-6197; funded by the Building Research Levy; 23 pp), read verbatim via pdftotext -layout — not AI-prepared, so no retrieval-provenance block is required. Every figure quoted traces to a stated table, figure or page. data_quality: verified per the batch rule (every figure traces to a directly-read primary): SR237 is BRANZ’s own original field-measurement study, matching the corpus convention for verbatim-read BRANZ primaries (its sibling OT_105 and OT_038).
⚠ Vintage/technology caveat (load-bearing — repeated from frontmatter/callout). The 11 monitored units are 2010-vintage R-134a integral and split systems, not the modern CO₂ HPWHs that dominate the current market. SR237 is therefore a real-world field floor + low-draw-off warning, not the modern design COP. Do not overwrite the DHW heat-pump COP cell (SR488/CR_014 ~2.5) with SR237’s lower field figures — use SR237 for the draw-off/climate sensitivity and the “installed performance sits well below marketing” reality-check.
⚠ Scope. Energy-performance only — SR237 contains no cost figures (as with SR488). The HPWH capital/installed-cost cells stay with CR_038 / OT_029. Monitoring covered Auckland and Wellington only (no South Island field sites); the cold-climate degradation is modelled (regression extrapolated to NIWA 1971–2000 temperatures for seven centres), not directly measured in the south.
Cross-references (no edits required): the SWH companion OT_105 (SR239 — same author, same year, same BRANZ programme; SR239 explicitly names SR237 as its HPWH companion, and RT_357 was raised from OT_105’s ingest); the modern CO₂ HPWH lab study OT_038 (SR488); the SWH field primary OT_084 (SR188, the source of Systems 1–3); the space-heating heat-pump COP/cost synthesis CR_014; and the HPWH cost/appliance benchmarks CR_038 / OT_029.
Connections
Links to
Referenced by
Sources (1): OT_105
EDT domains (1): D01: Renewable Energy & Storage Systems