OT_061: EECA Appendix One — Assessing residential solar at different time scales (Miller 2024)

Source

EECA Appendix One — Assessing residential solar at different time scales (Miller 2024)

Small-sample resolution study — not a demand shape

Appendix One (Dr Allan Miller / AMCL, 31 May 2024) to the EECA report “Understanding the value of residential solar PV and storage in New Zealand”. It quantifies how DATA RESOLUTION (1-min vs 30-min vs 60-min) affects modelled solar self-consumption/export, using GREEN Grid 1-minute data for 7 households. It does not publish a diurnal demand curve — the demand-shape firming (RT_165/RT_224) still needs the EPECentre dataset processed.

Summary

A methods appendix assessing the error introduced when residential-solar economics are modelled at coarser time resolution than the underlying 1-minute metering. Using the MBIE/Transpower/EEA-funded GREEN Grid 1-minute dataset (Centre for Sustainability, Otago; Ford/Anderson), Miller selected 7 households (Taranaki + Hawke’s Bay, ~5 kW PV, AC capacity factor 0.151–0.167) with ≥1 full year and <0.5% missing samples, and compared self-consumption and export computed at 1-minute, 30-minute and 60-minute resolution. Headline: coarser resolution understates solar export (and overstates self-consumption) because NZ net-metering records import and export separately within each half-hour trading period, and sub-half-hour alternation is lost when the data is averaged up.

Key claims

- claim: "Modelling residential solar at 30-minute resolution understates annual solar export by ~1.0–2.0% of total household load relative to 1-minute analysis; at 60-minute resolution the understatement is ~2.0–2.6% — because coarser averaging hides sub-half-hour import/export alternation under NZ's separate-register net-metering regime."
- claim: "NZ residential solar self-consumption (GREEN Grid, 7 Taranaki/Hawke's Bay households): at a 3 kWp-ac system, 21–44% of generation is self-consumed per household; at 6 kWp-ac, 12–30% (self-consumption falls as array size rises relative to load)."
- claim: "Basis: GREEN Grid 1-minute circuit/PV/incomer data from 44 households (Hawke's Bay + New Plymouth, ~2014–2018, Centre for Sustainability, University of Otago), of which 7 had a full year of coincident load+PV samples; Taranaki PV ≈5 kW, AC capacity factor 0.151 (2018)–0.167 (2015)."

Neobiome Intelligence relevance

  • Corroborates the engine’s E5 “export is a lower bound” caveat. The NI electricity slice runs a 3-slice representative day — coarser than the 60-minute resolution studied here — so it understates solar export (economic value) by at least the ~2–3% Miller finds, and likely more. A directional validation that the model’s known coarse-resolution simplification errs conservatively on export, not optimistically.
  • The 21–44% self-consumption at 3 kWp range is a NZ yardstick to sanity-check the engine’s self-consumption output (locally-used ÷ generated).
  • Confirms GREEN Grid as the NZ residential-metering backbone (same data behind CR_019 / CR_020).
  • It does not provide a diurnal demand shape, so it does not firm the demand-shape options (RT_165/RT_224 remain the path — a dataset task).

Research targets

Advances RT_225 (the EECA 2024 Miller value-of-solar study): this Appendix One is now ingested; the parent report (“Understanding the value of residential solar PV and storage in NZ”) is still to be retrieved as a working source (the public PDF URL returned a 2021 graphic asset, not the data report).

Connections

Links to

Sources (2): CR_019 · CR_020

Referenced by

Sources (3): OT_071 · OT_076 · OT_104

EDT domains (1): D01: Renewable Energy & Storage Systems