Source
https://personales.upv.es/thinkmind/ENERGY/ENERGY_2023/energy_2023_1_60_30045.html — original source (opens in a new tab; the file is not redistributed)
Apperley & Toki (2023) — An Islanded Community Solar Microgrid with Capability of Future Fractal Growth (LIT_085)
The primary conference paper behind LIT_059's Motairehe case — solar+battery islanded-microgrid DESIGN for a Māori community
Apperley (Univ. of Waikato) & Toki (Aotea Energy), IARIA ENERGY 2023, pp.29–35. Develops from first principles a non-grid-connected (islanded) smart community microgrid for Motairehe marae + 10 households on Aotea/Great Barrier Island — needs analysis, sizing, and a fractal-grid configuration that lets intra-community sharing cut supply shortfalls dramatically on the same total PV + battery, and that can grow to neighbouring communities without reconfiguration. This is the source LIT_059 (JRSNZ 2024) and CR_012 draw their Motairehe figures from; it adds the design-level detail those summaries omit.
Summary
The paper designs an islanded (off-grid) smart community microgrid for the Motairehe community on Aotea/Great Barrier Island — the majority-Māori part of an island whose entire ~1,000 population lives off-grid on under-sized solar/battery systems backed by fossil gensets. Working from real 8,760-hour household load profiles (scaled to a 7.5 kWh/day average base load) plus a synthesised marae load with 14 randomly-timed hui/year, it sizes each of 10 houses and the marae at a standard 4 kW solar + 22.5 kWh battery using NZ rules-of-thumb (solar ≈ average daily load ÷ 4; battery ≈ average daily base load × 3). An hourly annual simulation shows that operated as standalone systems the marae fails to meet demand for 2,184 hours/year and the houses for 355 hours combined; reconfigured as a fractal microgrid sharing the identical total generation and storage, whole-community shortfall falls to just 64 hours/year (only 16 coinciding with hui). Optimising the resource split (swept 18–30%) shows placing 20% of the PV and 20% of the battery at the marae minimises both peak and average grid flow, giving a recommended configuration of Marae 8,800 W + 49.5 kWh and Houses 3,520 W + 19.8 kWh, with an LV interconnect cable sized for the ~10 kW peak flow. For Neobiome this is a NZ-grounded design worked-example of the community-pooling principle (D9), a sizing cross-check, and a reticulation-cost signal.
Key claims
- claim: "SITE & COMMUNITY. Motairehe is on Aotea/Great Barrier Island, ~100 km NE of Auckland; there is NO reticulated power system on Aotea — the entire island population (~1,000) lives off-grid on their own solar/battery systems supplemented by petrol/diesel generators, natural gas and wood, and 'in almost all cases, the solar/battery systems do not provide nearly enough of the households' energy needs, so there is a heavy reliance on the fossil-fuel powered back-up generators.' Motairehe contains the majority indigenous Māori population of Aotea — approximately forty households and 90–100 people. The initial design is for the central part of the community: ten households together with a marae, intended to extend by phases to the rest of the households and then neighbouring communities."
source_location: "§I Introduction, p.29"
- claim: "HOUSEHOLD LOAD PROFILES. Annual load profiles were established for ten individual households from real NZ-mainland households at similar latitude (selected for NO electric hot-water and NOT using electricity as primary heating), using hourly data over a whole year — 8,760 data points per house. Profiles were scaled (retaining inter-household variation) to an average daily base load across the ten houses of 7.5 kWh."
source_location: "§II Load Profile Estimation, p.29–30 (Table 1)"
- claim: "MARAE LOAD PROFILE. Built from household profile A4 doubled to represent 4 adults (estimated permanent marae occupancy) → base-load daily average 19.171 kWh. Fourteen hui were added over the year with randomly generated start dates and durations of 1–3 days (community-suggested frequency); the hui supplementary load adds cooking, lighting and heating. With hui, the marae's annual average daily consumption rises to 24.275 kWh and the maximum consumption in any one hour over the year rises from 6.051 to 9.465 kW."
source_location: "§II Load Profile Estimation, p.30"
- claim: "SIZING RULES + STANDARD CONFIG + DISCRETIONARY LOAD. NZ design guidelines used: solar panel capacity ≈ average daily load ÷ 4; battery capacity ≈ average daily base load × 3. Standard per-site configuration (each household and the marae, operating independently): 4 kW solar panels + 22.5 kWh battery. A discretionary/divertible load (e.g. hot-water or EV charging) of up to 5 kWh/household/day was invoked only when the battery was at ≥95% charge and there was surplus solar — producing a community-total average discretionary load of 30.4 kWh/day, with just 9 days in the year when no discretionary load was possible at all."
source_location: "§II–III, p.30 (Table 2)"
- claim: "STANDALONE (NON-NETWORKED) RESULTS. Operated as independent units over the 8,760-hour year: for FOUR of the ten households there were zero base-load supply failures; House 10 was the worst case at 317 hours (3.6%) of failed base-load supply; the marae, because of its high hui peaks, had 2,184 hours in which the system could not meet demand (≈ nearly 9 kWh/day averaged over the year). All the failed hours across the ten households were spread across 355 hours of the year. House 5 achieves perfect generation/load balance 75% of the time (zero failed hours); House 10 achieves perfect balance 83% of the time but with 317 hours (3.6%) of unmet demand."
source_location: "§III Individual Household Design, p.30–31 (Table 2, Figures 1–2)"
- claim: "FRACTAL MICROGRID RESULT (same total resources). Sharing the SAME total solar generation and battery storage as the standalone case via a fractal microgrid, the overall number of shortfall hours over the year for the whole community falls to 64 — 'significantly less than the 2,184 previously experienced by the marae, and the 355 by all of the houses together' — with only 16 of the shortfall hours coinciding with hui, and 1,870 hours of overall surplus. The shared total resource was: solar 110 × 400 W panels (each of the 10 houses + marae fitted with ten 400 W panels); battery 247.5 kWh total (22.5 kWh × 11 sites)."
source_location: "§IV Fractal Microgrid Design, p.31–32 (Figure 5)"
- claim: "OPTIMAL RESOURCE DISTRIBUTION (marae allocation). To minimise grid flow (peak flow dictates cable size; average flow dictates cable energy losses), the marae's share of the total resources was swept from 18% to 30%; a 20% allocation of BOTH panels and battery to the marae produced a minimum of both peak grid flow AND daily-average grid flow. With 20% at the marae the largest peak flows (~9.5 kW) occur TO the marae and almost all coincide with hui, although overall more energy flows FROM the marae; the marae flow is used as the basis for cable sizing."
source_location: "§V Resource Distribution for Optimal Microgrid Operation, p.33 (Figures 9–11)"
- claim: "RECOMMENDED FRACTAL-OPTIMISED CONFIGURATION. Based on the analysis (20% marae allocation), the suggested configurations are — Marae: 8,800 W solar panels + 49.5 kWh battery; Houses: 3,520 W solar panels + 19.8 kWh battery ('these numbers, of course, need to be rounded up for sensible use of currently available technology'). The houses and marae must be connected by a cable capable of carrying at least the calculated 10 kW load with minimal losses."
source_location: "§VI Conclusion and Future Work, p.34"
- claim: "RETICULATION-COST SIGNAL. 'A preliminary calculation, taking into account the relatively close proximity of the households and marae at Motairehe, suggests that the cost of the interconnecting microgrid cables to handle this 10 kW load could have provided ~15% increased solar panel capacity across the site, if no microgrid was included. However the advantages of the microgrid interconnection shown in this analysis, far outweigh the benefits of increased, but isolated, generation.' — i.e. the internal LV interconnect cable costs on the order of ~15% of the site PV capex, and buys a far larger shortfall reduction than the equivalent extra isolated PV would."
source_location: "§VI Conclusion and Future Work, p.34"
- claim: "FRACTAL ARCHITECTURE + FUTURE GROWTH. In the fractal-grid model any terminal node can combine load, generation and/or storage (a household, a community solar array, a community service such as street lighting or EV charging, or a community battery); non-terminal nodes present as a single intelligent entity to the higher-level grid ('fractal microgrid'). The initial Motairehe grid is a single level-0 grid with the marae as the level-1 node. Smart per-node dispatch priority: (i) local base load → (ii) local battery charging → (iii) microgrid (community) needs → (iv) household discretionary load. The structure readily enables future expansion (further community facilities, other close-by communities as extensions or new level-0 grids, or expansion to level 2+) without significant reconfiguration. Developed with support from the MBIE Māori Housing Renewable Energy Fund; the research was supported by the MBIE Endeavour Programme (Public Housing and Urban Regeneration: Maximising Well-being, Grant ID 20476 UOOX2003) and the MBIE SSIF Ahuora grant (CONT-69342-SSIFAETP-UOW)."
source_location: "§IV, §VI, Acknowledgements, p.31–35"Neobiome Intelligence relevance
A NZ-grounded, off-grid, Māori-community DESIGN worked-example — the primary behind the Motairehe figures NI already leans on via LIT_059 and CR_012. Its value beyond those summaries is design-level detail: (1) The pooling result in raw numbers. Whole-community shortfall drops from marae 2,184 hrs + houses 355 hrs to 64 hrs on identical total PV+battery (110 × 400 W = 44 kW PV, 247.5 kWh) — the empirical underpinning of LIT_059’s “~97%” headline and direct support for NI’s core assumption that the community pools generation/storage (D9) rather than islanding per household. (2) A fractal-optimised resource split. Sweeping the marae’s share 18–30% and finding 20% minimises both peak and average grid flow is a concrete community-vs-per-household resource-distribution worked example; the recommended asymmetric sizing (Marae 8.8 kW / 49.5 kWh vs Houses 3.52 kW / 19.8 kWh, from a uniform 4 kW / 22.5 kWh standalone base) is a NZ cross-check on the model’s PV×battery sizing (D15) and on the solar ≈ load÷4 / battery ≈ base-load×3 heuristic. (3) A reticulation-cost cross-check. The paper’s finding that the LV interconnect cable for the ~10 kW peak load “could have provided ~15% increased solar panel capacity” is an independent order-of-magnitude sanity check on CR_027’s ~$8,500/household internal-reticulation line — internal wires cost a meaningful but minority fraction of PV capex, and the sharing benefit outweighs the equivalent extra isolated PV. (4) Dispatch priority (local load → local battery → community → discretionary) matches NI’s slice-dispatch logic (D19). Thesis value: an off-grid Māori community (marae, hui, papakāinga-adjacent), community “ownership” of the microgrid, and the MBIE Māori Housing Renewable Energy Fund / Public-Housing-Endeavour framing — governance, equity and mātauranga-Māori material.
What it is NOT. A design/simulation study, not a built/measured deployment — every figure is a model output for one specific site (10 houses + one marae, Aotea). The sizing and shortfall numbers are Motairehe-specific and should be transferred as method and order-of-magnitude, not as calibration constants; no figure here is wired into an NI calculation cell (the model already sources its sizing heuristics and reticulation cost from LIT_059 / CR_027). Resolves RT_244 and cross-references the existing Motairehe evidence rather than replacing it.
Research targets
Documents to retrieve
- None. This resolves RT_244 (Apperley & Toki 2023 — the Motairehe conference primary), the doc raised by LIT_059.
Research gaps
- No new RT. The paper’s method references are the authors’ own prior IARIA conference papers on the fractal-grid model and energy-balance modelling (Apperley 2019, 2017) and a solar-simulation tool (Suppers & Apperley 2015) — methodology, not NZ data worth a target. The reticulation-cost and community-microgrid-cost questions are already tracked (CR_027 / RT_228-lineage); no NZ recalibration target is opened here.
Notes
Primary peer-reviewed conference paper (IARIA ENERGY 2023, ISBN 978-1-68558-054-4, pp.29–35), read verbatim via pdftotext -layout; every figure in key_claims traces to a stated section/table/figure. data_quality: high, not verified — it is a single authoritative NZ-relevant primary, but LIT_059 and CR_012 report the same Apperley analysis, so they corroborate without being independent (the verified ↔ high line is independent corroboration per model_design.md §2). Design/simulation outputs, not field measurements. context: both — a D01 design cross-check (NI) that also supplies off-grid Māori-community governance/equity material (thesis). Resolves RT_244; cross-references LIT_059, CR_012, CR_027; no new RTs.
Connections
Links to
Referenced by
Sources (1): LIT_059
EDT domains (1): D01: Renewable Energy & Storage Systems