OT_159: Amitrano et al. (2014) — Building Energy End-use Study (BEES) Part 1: Final Report (BRANZ SR297/1)

OT_159: Amitrano et al. (2014) — Building Energy End-use Study (BEES) Part 1: Final Report (BRANZ SR297/1)

Source

https://www.branz.co.nz/pubs/research-reports/sr297/ — original source (opens in a new tab; the file is not redistributed)

Amitrano et al. (2014) — Building Energy End-use Study (BEES) Part 1: Final Report (BRANZ SR297/1)

Scope — commercial office & retail, NOT community buildings; an UPPER anchor for the #12 intensities, not a validation

BEES is the non-residential companion to HEEP, but it sampled commercial office and retail buildings (plus a mixed “Other BEES” residual) — not the halls, workshops, barns, glasshouses or schools the model’s community-building electricity intensities (*_elec_kwh_m2_yr, engine params.py #12) describe. Its measured intensities are high because these are intensively-serviced, fully-occupied commercial premises: national EnPIelec 173 ±28 kWh/m²·yr, smallest stratum S1 (≤650 m²) 143 ±57, “Other BEES” 158 ±36. The model’s assumed community values (barn 5, glasshouse 10, hall 30, workshop 40, school 50 kWh/m²·yr) are 3–30× lower — which is correct for intermittently-used, lightly-serviced buildings. So BEES bounds the assumptions from above and confirms they sit well below any commercial floor (a directional sanity check), but it is not a per-type benchmark for those buildings and does not move them off assumed. Use BEES as the authoritative NZ non-residential electricity-intensity primary; keep RT_328 open for the low-intensity community-building types (EECA end-use database route).

Summary

BRANZ Study Report SR297/1 (Amitrano et al., 2014) is the final report of the Building Energy End-use Study (BEES) — the largest and most detailed measurement of how New Zealand’s non-residential buildings use energy and water, and the direct non-residential counterpart to the residential HEEP study (OT_037). Running for six years from 2007, BEES combined a high-level survey of buildings and a telephone survey of businesses (848 premises in 412 buildings) with intensive targeted monitoring of ~101 individual premises, plus web-search building characterisation of ~3,000 buildings and extensive computer modelling (a 48-model, seven-climate-zone national dashboard). It established, for the first time, that there are an estimated 41,154 ±1,286 non-residential (“BEES”) buildings in NZ with 39.93 ±2.14 million m² of floor area (average ~970 m²/building), using 6,370 ±1,100 GWh/yr of electricity at a national electricity performance indicator of EnPIelec = 173 ±28 kWh/m²·yr and 1,130 ±840 GWh/yr of gas (EnPIgas = 31 ±23). EnPIelec rises with building size (S1 143 → S5 223 kWh/m²·yr) and differs by use (Commercial Office 186, Commercial Retail 176, Other BEES 158). For Neobiome Intelligence it is the primary NZ source for non-residential building electricity intensity — but because BEES samples only commercial office and retail buildings, it functions as an intensively-serviced upper anchor for the model’s community-building intensities rather than a direct per-type benchmark.

Key claims

- claim: "BEES is the non-residential companion to HEEP: for BEES, non-residential buildings have been defined using categories in the New Zealand Building Code, but in general terms the study looked at commercial office and retail buildings — these vary from small corner-store dairies to large multi-storey office buildings. The research started in 2007 and ran for 6 years, with two main methods of data collection: a high-level survey of buildings and businesses, and intensive detailed monitoring of individual premises."
  source_location: "Preface, p.iii ('For BEES, non-residential buildings have been defined using categories in the New Zealand Building Code…'; 'The BEES research started in 2007 and ran for 6 years'); §ii text 'The study had two main methods of data collection'"
- claim: "It is estimated there are 41,154 ±1,286 (95% CI) BEES buildings in New Zealand with a total floor area of 39.93 ±2.14 million m² (36.86 ±2.60 million m² of which were BEES areas), giving an average of approximately 970 m² per building. For comparison there are ~1.5 million occupied dwellings with ~222 million m² total floor area, ~160 m² per dwelling."
  source_location: "Executive Summary, p.iv ('Non-residential Buildings Energy Use')"
- claim: "The building-size distribution is extremely skewed. Divided into five equal-floor-area strata: S1 (5–649 m²) = 27,609 buildings (67% of buildings, avg 298 m²); S2 (650–1,499 m²) = 8,007 (19%); S3 (1,500–3,499 m²) = 3,544 (9%); S4 (3,500–8,999 m²) = 1,496 (4%); S5 (≥9,000 m²) = 499 buildings (1%, avg 17,014 m²); total 41,154. The very large buildings are numerically small and hold only ~20% of floor area — the other ~80% of floor area is in buildings under 9,000 m²."
  source_location: "Executive Summary, Table A (p.iv) and text p.iv–v"
- claim: "Across all buildings, total electricity use was 6,370 ±1,100 GWh/yr — an electricity performance indicator EnPIelec of 173 ±28 kWh/m²·yr. Total gas use was 1,130 ±840 GWh/yr — a gas performance indicator EnPIgas of 31 ±23 kWh/m²·yr."
  source_location: "Executive Summary, p.v ('Across all buildings, total electricity use was 6,370 ±1,100 GWh/yr…')"
- claim: "EnPIelec increases with building size (Table C): S1 (0–649 m²) 143 ±57; S2 (650–1,499 m²) 153 ±53; S3 (1,500–3,499 m²) 154 ±65; S4 (3,500–8,999 m²) 201 ±73; S5 (≥9,000 m²) 223 ±66 kWh/m²·yr — attributed to the increased level of services provided as building size increases."
  source_location: "Executive Summary, Table C (p.v)"
- claim: "EnPIelec by building-use stratum (Table C): Commercial Office (CO) 186 ±61; Commercial Retail (CR) 176 ±45; Other BEES 158 ±36; Total 173 ±28 kWh/m²·yr. When separated into individual use strata, average EnPIelec ranged from ~150 kWh/m²·yr (Other BEES, lowest) to ~190 kWh/m²·yr (Commercial Office, highest)."
  source_location: "Executive Summary, Table C (p.v); §2 p.7 ('the average EnPIelec ranging from 150 kWh/m².yr to 190 kWh/m².yr')"
- claim: "Office and Multiple-Use premises display the 'one-third rule': approximately one-third of electricity goes to lighting, one-third to plug load, and one-third to space conditioning and other — consistent across both premise and building-size groupings. Lighting is very important across most categories (especially non-food retail); commercial refrigeration dominates electricity end-use in Food Storage premises."
  source_location: "Executive Summary, p.vi ('The Office and Multiple Use premises display the one-third rule…')"
- claim: "Targeted monitoring was undertaken in 101 premises, with end-use electricity data available for 84. Plug loads and lighting were found in 100% of premises; identified circuit-wired space conditioning in 74%, circuit-wired water heating in 64%, process energy 24%, non-domestic cooking 21%, non-domestic refrigeration 10%, and a catch-all Miscellaneous category in 61%."
  source_location: "Executive Summary, p.vi ('Targeted monitoring was undertaken in 101 premises…')"
- claim: "As building size reduced, the prevalent source of heating (and often cooling) was electric heat pumps; only in the two smaller building-size strata did simple electric resistance heaters as the primary heating source exceed 30% of the sample. Independent of building size, about half of monitored premises contained some electric resistance heaters (fixed or portable) and about half contained some portable electric fans."
  source_location: "Executive Summary, p.vii ('As building size reduced, the prevalent source of heating…')"

Neobiome Intelligence relevance

context: ni — BEES is the authoritative NZ primary for non-residential building electricity intensity. It feeds D01 (the demand-side benchmark behind the community-building electricity intensities, engine params.py #12) and D04 (non-residential building energy-performance structure). Its load-bearing caveat is scope: it samples commercial office/retail, not community buildings, so it anchors rather than validates.

• D01 — non-residential electricity-intensity primary; the UPPER anchor for the #12 community-building intensities. BEES gives the only measured NZ non-residential electricity-intensity dataset: national EnPIelec 173 ±28 kWh/m²·yr, smallest stratum S1 (≤650 m²) 143 ±57, and lowest use stratum Other BEES 158 ±36 OT_159. These are 3–30× above the model’s assumed community-building values (barn 5, glasshouse 10, hall 30, workshop 40, school 50 kWh/m²·yr) — as they should be, because BEES premises are fully-occupied, intensively-serviced commercial offices and shops, whereas the model’s buildings are intermittently used and lightly serviced. So BEES confirms the assumed values sit correctly well below any NZ commercial-building floor (a directional sanity check) and bounds them from above, but it is not a per-type benchmark and does not move barn/glasshouse/hall/workshop/school off assumed.
• D04 — how non-residential building electricity splits (structure, not magnitude). The office “one-third rule” (≈⅓ lighting / ⅓ plug load / ⅓ space conditioning + other), the strong lighting–floor-area relationship, and the end-use presence rates (space conditioning 74%, water heating 64%, cooking 21%, refrigeration 10% of monitored premises) give the structure of non-residential electricity demand that complements HEEP’s residential split (OT_037) OT_159. For a community build with mixed residential + shared non-residential spaces (hall, workshop, common house), this is the reference for how the non-residential slice divides — with the same vintage/scope caveats as the intensities.

Research targets

Documents to retrieve

• BEES Part 2 (BRANZ Study Report SR297/2, appendices) — the methodology + detailed-results companion to this Part 1 overview, holding the granular per-premise-type EnPIelec breakdowns (by Revised QV category, CPA and DAC) and the end-use audit detail that get closer to disaggregated building-type intensities than the Part 1 headline strata. No new RT raised — folded into the RT_328 residual below as an alternative route to a finer NZ non-residential split, alongside the EECA EEUD route. Retrieve only if RT_328 is actively re-worked.

Research gaps

• RT_328 (existing — ADVANCED, keep open). BEES delivers the “BRANZ BEES small non-res benchmarks” this RT sought and is now the authoritative NZ non-residential electricity-intensity primary — but it samples commercial office/retail only, so it bounds the model’s community-building intensities (*_elec_kwh_m2_yr, #12) from above rather than validating them. The per-type values (barn 5 / glasshouse 10 / hall 30 / workshop 40 / school 50 kWh/m²·yr) remain assumed. Two routes to the low-intensity, intermittently-used community-building types BEES does not sample: (1) the EECA energy-end-use database (partly in corpus as RD_007 / RD_024); (2) BEES Part 2 (SR297/2) for the finer per-premise-type (QV/CPA/DAC) split. Do not close RT_328 on this ingest.

Notes

Authoritative BRANZ publication (BRANZ Study Report SR297/1, © BRANZ 2014, ISSN 1179-6197; jointly funded by the Building Research Levy, MBIE Public Good Science Fund, and EECA; 194 pp), read verbatim via pdftotext -layout — not AI-prepared, so no retrieval-provenance block is required. Every figure quoted traces to the raw. data_quality: verified (every figure traces to a directly-read primary): BEES is BRANZ’s own original measurement study, the strongest verified case in the corpus, and this matches the recent corpus convention for verbatim-read BRANZ primaries (OT_131 = verified). Note the residential twin HEEP (OT_037) carries the older high value — a corpus inconsistency, not a difference in evidentiary strength.

⚠ Scope caveat (load-bearing — repeated from frontmatter/callout). BEES sampled commercial office and retail buildings only (plus a mixed “Other BEES” residual). It does not sample community halls, workshops, barns, glasshouses or schools. Its intensities are an intensively-serviced-commercial upper anchor, not a benchmark for the model’s intermittently-used community buildings. Do not apply BEES EnPIelec figures (143–223 kWh/m²·yr) directly to a hall, barn or glasshouse cell.

⚠ Data vintage. Fieldwork spanned ~2007–2012 (survey + monitoring over the 6-year programme; final report 2014). LED penetration was only 2% of monitored premises and lighting was a full third of office electricity — both have shifted materially since, so lighting-share and absolute EnPIelec figures are dated; treat the structure (size/use gradients, one-third rule) as more durable than the absolute kWh, exactly as for HEEP. Christchurch was excluded after the 2010/2011 earthquakes removed 11% of the sample frame.

Cross-references (no edits required): the residential twin OT_037 (HEEP — same BRANZ end-use-study lineage, several shared authors; BEES is the non-residential complement) and the EECA end-use database pages RD_007 / RD_024 (the modern cross-check and the route to the low-intensity community-building types BEES omits).

Connections

Links to

Sources (4): OT_037 · OT_131 · RD_007 · RD_024

Referenced by

EDT domains (2): D01: Renewable Energy & Storage Systems · D04: Sustainable Habitat & Building Technology