LIT_047: Kim, Jang & Choi (2023) — Estimated vs Actual Power Self-Sufficiency in a Solar Energy-Sharing Community

LIT_047: Kim, Jang & Choi (2023) — Estimated vs Actual Power Self-Sufficiency in a Solar Energy-Sharing Community

Source

doi:10.3390/en16247941 — original publication (opens in a new tab; the file is not redistributed)

Summary

Kim, Jang & Choi (2023) did something rare and directly useful: they took a solar energy-sharing community, recorded the self-sufficiency rate estimated at the design stage, and compared it against the actual self-sufficiency rate measured over the following year (April 2022 – March 2023). The estimated rate was 171%; the realised rate was 133% — a ~38-point shortfall, with actual performance only ~78% of the design estimate. They trace the gap mainly to the difference between the standard insolation assumed in design and the actual insolation experienced, with strong seasonal variation. For Neobiome Intelligence this is a calibration anchor: it quantifies how much a design-stage self-sufficiency figure can overstate reality, and confirms that the dominant error source is the solar resource — exactly what the model’s P50/P90 bands are meant to bracket.

Key claims

- claim: "Kim, Jang & Choi (2023, Energies 16:7941) compared the design-stage ESTIMATED energy self-sufficiency rate of a solar energy-sharing community against the ACTUAL rate calculated from one year of operational data (April 2022 – March 2023); the study area is a Korean national smart-city pilot (Busan Eco Delta City)."
  source_location: "Abstract; §1 Introduction; §2 Methods"
- claim: "The estimated energy self-sufficiency rate was 171%, whereas the realised (actual) rate was 133% — a roughly 38-percentage-point shortfall (actual ≈ 78% of the design estimate). The authors' hypothesis that the design estimate would differ from the operational value by more than 20% was confirmed."
  source_location: "Abstract; §1 (hypothesis); §3 Results"
- claim: "The discrepancy was attributed mainly to the difference between the standard insolation assumed at the design stage and the actual insolation experienced, with the differences analysed by total, district and season — seasonal variation in the solar resource was the principal driver."
  source_location: "Abstract; §3 Results (seasonal/insolation analysis)"
- claim: "The study community comprises 56 households on a site of ~3,620 m² (18 of them two-to-three-storey detached houses, the remainder multi-family), in an area with ~1,358 mm average annual precipitation; both the estimated and actual self-sufficiency rates exceed 100% (net over-generation), but operational performance is materially below the design estimate."
  source_location: "§2 Study area / community description"
- claim: "The authors note a gap in prior research: self-sufficiency rates are usually estimated OR calculated, but design-phase estimates are rarely compared against actual post-occupancy rates — yet accurate estimation is critical for planning energy-efficient communities."
  source_location: "§1 Introduction (research gap)"

Neobiome Intelligence relevance

• Self-sufficiency calculation — the calibration anchor. This is direct evidence that a design-stage self-sufficiency estimate can substantially overstate the measured outcome (here 171% → 133%, ~78% realisation), and that the dominant error is the solar resource (actual vs standard insolation). The NI model should treat modeled self-sufficiency as an estimate to be discounted and bracketed, which is exactly the role of its P50/P90 bands (D16) and the pv_yield_p90 bad-year companion — this paper is the empirical justification for running the frontier at P90, not just P50 LIT_047.
• D01 / reinforces “modeled, not metered.” It quantifies the modeled-vs-measured gap for community solar self-sufficiency, alongside the same caveat already attached to the Totarabank LCOE (LIT_032) and HEEP energy figures LIT_047.
• ⚠ Scope: Korea-context (insolation, household mix and the >100% over-generation differ from a NZ pilot). The magnitude (~78% realisation) is illustrative, not a NZ calibration constant — but the direction (design overstates actual) and the cause (solar-resource variance) transfer. The “self-sufficiency rate” here is a generation-to-demand ratio (so it can exceed 100%), not a capped self-consumption fraction.

Research targets

Research gaps

• None proposed for retrieval — the finding reinforces the model’s existing P50/P90 solar-yield bands (D16 / pv_yield_p90, RT_102); a NZ-specific estimated-vs-actual community-SS study is unlikely to exist, so it is not worth chasing.

Connections

Links to

Sources (1): LIT_032

Referenced by

Concepts (1): Self-Sufficiency Calculation Framework

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