LIT_069: Kerr (2024) — Empowering Energy Innovation in the Communities of Aotearoa (Masters thesis, Univ. of Waikato)

LIT_069: Kerr (2024) — Empowering Energy Innovation in the Communities of Aotearoa (Masters thesis, Univ. of Waikato)

Source

https://researchcommons.waikato.ac.nz/ — original source (opens in a new tab; the file is not redistributed)

Kerr (2024) — Empowering Energy Innovation in the Communities of Aotearoa

Masters of Engineering thesis (Univ. of Waikato) pairing an NZ industrial-solar case study (Chia Sisters juicery, Nelson — 9 months of custom 5-min monitoring: 36.8% solar self-supply, 24% panel conversion, ~$5.3k saved in 9 months) with an Excel simulation of school-bus Heavy-Vehicle-to-Grid (HV2G) peak-shaving in Wellington (21 buses → 4.62 MW evening "virtual battery"; scaling to 150 buses → ~33 MW), then theorises a Community Energy Management System (CEMS) for remote/rural communities.

⚠ RT_024 called this a “Doctoral dissertation” — it is a MASTERS thesis (Michaela Jane Kerr, 2024). Primary read, figures traceable → but a non-peer-reviewed thesis with loose units (power stated as “kW per hour”, “60 kW/h” chargers), self-admitted rough capacity estimates, and simulation/borrowed figures → data_quality: medium, not verified. Case data is Nelson/Wellington-specific, not national. Use the Chia solar-performance numbers and the Rowallan microgrid spec as concrete NZ anchors; treat the HV2G MW figures as an indicative proof-of-concept, not a validated dispatch model.

Summary

Kerr’s thesis argues that Aotearoa’s rural, remote and isolated communities were “yet to be considered when energy policy was written” and asks how existing and already-planned energy infrastructure can be repurposed to give those communities more security of supply while advancing decarbonisation. It builds the argument through two NZ case studies and a synthesising concept. Case Study One (Chapter 3) instruments the Chia Sisters solar-powered juicery in Nelson: because the industrial site had no smart meter, the author built a low-cost open-source monitoring stack (Hioki clamp-on data logger → Linux box → GitHub → a Streamlit web app) and logged ~9 months of 5-minute load and solar data, quantifying how much of the factory’s electricity its rooftop array actually offsets and recommending operational changes (reschedule the analogue-timer boiler, add boiler insulation, consider a battery). Case Study Two (Chapter 4) is an Excel proof-of-concept simulating Heavy-Vehicle-to-Grid (HV2G) use of Wellington’s electric school-bus fleet as a distributed “virtual battery” for evening peak-shaving, run against a real high-load Transpower day at three fleet-electrification levels (existing ~14%, 50%, 100%). Chapter 5 fuses the two into a theorised Community Energy Management System (CEMS) — a distributor-owned layer that monitors community generation (factory/residential solar), EV/bus storage and demand to shift load locally, akin to a community-scale Virtual Power Plant. The document sits alongside the existing NZ community-microgrid corpus (LIT_002, the RT’s own reference tag) and complements the design-led NZ autonomy thesis OT_018 and the Waitati community-energy case URL_003. It resolves the RT_024 retrieval target.

Key claims

- claim: "NZ national context (thesis framing): New Zealand commits to reducing emissions 30% below 2005 levels before 2030 and to being 100% carbon neutral by 2050; contributes ~0.3% of global GHG emissions yet has the 12th-highest emissions per capita in the developed world; water is 78.6% drinkable across all categories (bacteriological, protozoal, chemical, water safety plan) per MoH 2020; land is being lost to sea-level rise at ~1.8 mm/year. On any given day ~80% of the NZ electricity grid is powered by renewable sources (Transpower, 2023)."
  source_location: "Section 1.2 Problem Statement, p.2-3 (emissions, 0.3%, 12th, 78.6%, 1.8mm); Section 2.1.1, p.8 (~80% renewable)"
- claim: "Aotearoa's overall ENERGY (not electricity) usage is a 60:40 non-renewable:renewable split — 60% from coal, oil and gas, 40% from renewables (Energy Resources Aotearoa, 2020). The non-renewable 60% breaks down as transport 36%, agriculture 4.7%, industrial 34%, domestic 11%, commercial 9%. The Climate Change Commission recommends 50% of all energy consumed by 2035 be from renewable sources (CCC, 2022)."
  source_location: "Section 2.1.2 New Zealand's Energy System, p.10-11 (Figure 2.1)"
- claim: "NZ electricity demand is predicted to increase 68% over the next thirty years (Transpower, 2022b). Peak electricity consumption is predicted to rise ~66% over 30 years — a jump from 8,600 MW to around 13,700 MW. Transpower estimates over 40 new power stations will connect to the grid by 2035 (more generation development than in the last 40 years), needing ~30 new grid connections. Wind and solar were identified (MBIE, 2019) as the best way to increase NZ generation by 66% over 27 years."
  source_location: "Section 1.1, p.2 (68%); Section 4.3.2.1, p.46-47 (8,600→13,700 MW, ~66%); Section 2.1.1, p.8-9 (40 stations, 30 connections, MBIE 66%/27yr)"
- claim: "NZ distributed generation (Electricity Authority, 2023): industrial participants have 1,255 ICP (Installation Connection Point) connections and commercial participants 1,649, at an average installed capacity of 165 kW — much higher than residential systems, which average about 4.6 kW. EECA predicts industrial and commercial solar DS will eventually provide approximately 6% of NZ's overall electricity supply. A residential distributed-solar install inclusive of a battery ranges from $15,000 to $30,000."
  source_location: "Section 3.1.3 Commercial or Industrial Distributed Systems, p.20-21 (ICP counts, 165 kW, 4.6 kW, 6%); Section 3.1.2, p.20 ($15,000–$30,000)"
- claim: "NZ receives an average of over 2,300 sunlight hours per year, with up to 15 hours 10 minutes of daylight per day (NIWA, 2007). NIWA monitoring stations further north average closer to 25 MJ/m² per day of radiation (1 MJ = 0.277778 kWh), i.e. ~6.94 kWh/m² per day of solar potential; many NZ regions are relatively similar in daily production potential (except Antarctica)."
  source_location: "Section 3.1.1 Solar in New Zealand, p.18-19 (2,300 hours, 15h10m); Section 5.1.3, p.65 (25 MJ/m²/day, 6.94 kWh/m²/day, 1 MJ = 0.277778 kWh)"
- claim: "Chia Sisters juicery, Nelson: installed a series of 32 Sumec solar panels in 2018 with a nominal ~16 kW capacity ('16kW per hour'), stated as usually twice the factory's ~8 kW maximum requirement — the author explicitly flags this stated capacity claim as 'not correct' once monitored. The system cost nearly $30,000; on current bills the owners estimated a ~10-year payback, and the monitored data implies close to a 6-year payback (installed 2018, still running). The 2000 L boiler heats to 65–70°C on an analogue timer at 4 am and 12 pm for sterilisation."
  source_location: "Section 3.2.1 Introduction, p.22 (32 Sumec panels, 16 kW, 8 kW, 2018, $30,000, ~10yr); Section 3.4.3.2, p.36 (~6yr); Section 3.2.2, p.23-24 (2000 L boiler, 65–70°C, 4 am/12 pm)"
- claim: "Chia Sisters monitored results over ~9 months (5 Jul 2022 – 11 May 2023): total factory usage 45,587.89 kWh, total rooftop solar generation 18,332.67 kWh → 36.814% of usage supplied by solar (monthly %-solar ranged 8.56% in early May to 60.38% in December). Costed at Chia's retailer rates (anytime 28.99 c/kWh; solar buyback 12 c/kWh): cost with solar $7,901.13 vs $13,215.79 without → savings of $5,314.66 in nine months (Table 3.2)."
  source_location: "Table 3.2 (Total load and generation data), p.36; Section 3.4.3.2, p.35-36"
- claim: "Chia panel performance: comparing generation to NIWA recorded radiation (Nelson Airport station, ~1 km away), the Sumec panels' overall conversion efficiency of radiation to generation is calculated at 24%, giving about 1.735 kWh per m² of solar panel. NIWA's Solarview prediction model (which uses ~10 years of historical radiation and accounts for orientation/obstructions) was found to be ~61% accurate (calculated value 0.612171) at predicting Chia's March generation."
  source_location: "Section 5.1.2, p.63 (24% conversion, MJ/h ÷ 3.6); Section 5.1.3, p.65 (1.735 kWh/m²); Section 5.1.1.1, p.61-62 (Solarview ~61%, 0.612171, 10 yr data)"
- claim: "PowerNet Rowallan microgrid (Southland) — an existing NZ off-grid community microgrid, built because a large electricity user was frequently cut off from the PowerNet network by poles washed away from erosion: PowerNet installed a microgrid at the Rowallan lodge powered by two 6 kW solar arrays, a 27 kWh battery storage system, and a diesel generator as backup (PowerNet, 2020)."
  source_location: "Section 2.4 Microgrids, p.13-14"
- claim: "Wellington electric school-bus specifications (Metlink, 2020): single-decker eT12-max buses carry a 350 kWh battery, run all day and take ~5.5 hours to fully charge on 60 kW slow chargers (arriving at depot with ~50–80 kWh remaining); double-decker UT200RHDF buses carry a 508 kWh battery and use fast charging (two fast chargers at 450 kW and 1,080 kW). A bus uses ~0.56 kWh per km; the model derives ~4.86% of battery capacity used per hour of travel (17 kWh/hour over a 19-hour day from a 330 kWh usable window, Eq. 4.1–4.2). Buses never drop below 20% state of charge (bus-company requirement); the model counts a maximum of 75% of electric-bus capacity."
  source_location: "Section 4.3.2.3–4.3.2.4, p.47-48 (350/508 kWh, 5.5 h, 60 kW, 450/1,080 kW, 0.56 kWh/km); Section 4.6, p.51-52 (Eq. 4.1–4.2, 4.86%, 17 kWh); Section 4.3.2.5, p.48-49 (20% SoC, 75%)"
- claim: "HV2G school-bus peak-shaving SIMULATION (Wellington, Metlink 150 school-bus routes, run against a real high-load Transpower day). Simulation One — existing fleet, ~14% electric = 21 buses → a modelled 4.62 MW distributed 'virtual battery' available in the evening, a 5–10% reduction in localised evening load, estimated to power ~1,000–1,500 homes (using Mercury's Auckland-battery estimate). Simulation Two — 50% uptake (75 buses), load increased 25% → 16.5 MW. Simulation Three — 100% uptake (~150 buses) by 2035 → a combined ~33 MW battery, providing over 50% of the capacity required in localised evening peaks."
  source_location: "Section 4.6.1, p.52-53 (21 buses, 4.62 MW, 5–10%, 1,000–1,500 homes); Section 4.6.2, p.53 (75 buses, 16.5 MW, +25%); Section 4.6.3 & 4.7, p.53-55 (150 buses, ~33 MW, >50%)"
- claim: "Comparators and definitions: Hornsdale Power Reserve (South Australia) — the world's first big battery, a 100 MW Li-ion Tesla system commissioned Dec 2017 with a 50 MW expansion in 2020, estimated to have saved over AU$14 million by providing islanding/inertia support (Aurecon, 2022). NZ 'rural communities' are defined (Stats NZ, 2023) as a population between 200–999 people with at least 40 dwellings; the 2011 country-wide fibre rollout 'forgot and left behind 13% [of] New Zealanders'. There are ~4.4 million registered vehicles in NZ (2021), most expected to be electric by 2050."
  source_location: "Section 4.2, p.42-43 (Hornsdale 100 MW + 50 MW, AU$14M); Section 2.3, p.12-13 (rural 200–999 / ≥40 dwellings, 13% fibre); Section 5.1.4, p.66 (4.4 million vehicles)"

Neobiome Intelligence relevance

This is the document behind RT_024. Routed context: both — it feeds D01 (renewable energy & storage) and supports the thesis argument about technology-enabled resilience for remote/rural NZ communities. Its NI value is concrete-but-caveated:

• NZ industrial/distributed-solar performance data (D01). The Chia Sisters case is a rare piece of measured (not modelled) NZ rooftop-solar performance: a ~16 kW array on a Nelson juicery self-supplied 36.8% of factory electricity over 9 months (18,332.67 of 45,587.89 kWh), at a monitored panel conversion efficiency of 24% (~1.735 kWh/m² of panel), saving ~$5,315 in nine months at a 28.99 c/kWh anytime tariff / 12 c/kWh buyback. These are usable real-world calibration anchors for D01’s solar-yield and self-supply-fraction logic, and they corroborate the “solar covers roughly a third of load without storage” pattern. The NIWA-Solarview finding — the model over-predicts, actual generation ≈ 61% of predicted — is a useful haircut factor when NI uses modelled solar yields.
• A concrete NZ off-grid microgrid spec (D01). The PowerNet Rowallan microgrid (Southland) — 2 × 6 kW solar + 27 kWh battery + diesel backup, built specifically because grid poles kept washing out — is a small, real, erosion-driven off-grid NZ community microgrid that sits at the very small end of the community-microgrid size ladder already in the corpus (LIT_031 / LIT_032 / LIT_033 / LIT_068). Its full cost/spec beyond the component sizes given here is a potential follow-on retrieval (noted, not opened).
• Vehicle-to-grid as a community storage lever (D01 storage). The HV2G school-bus concept quantifies an existing-infrastructure storage option: a Wellington fleet at 100% electrification (~150 buses) modelled as a ~33 MW distributed virtual battery covering >50% of localised evening peak, with per-bus specs (350–508 kWh, 0.56 kWh/km, 20% SoC floor) that are reusable if NI ever models EV/bus storage. Caveat: these are Excel proof-of-concept outputs with loose units and borrowed rules-of-thumb (the “~1,000–1,500 homes” figure is Mercury’s estimate for a different battery), so treat the MW numbers as indicative, not validated dispatch.
• NZ national baseline figures (D01 context). Handy corroborating anchors: 68% predicted 30-year demand growth (peak 8,600 → ~13,700 MW), ~80% daily grid renewables, the 60:40 non-renewable:renewable total-energy split, the CCC 50%-renewable-by-2035 target, and distributed-generation fleet stats (industrial DS avg 165 kW, residential 4.6 kW; residential solar+battery 15k–30k). Most are cited secondhand from Transpower/MBIE/EECA — use as cross-checks, not primaries.

I01 (financial/economic sufficiency) touch, not a feed. The Chia payback economics (~30k system, ~6-year monitored payback, ~5.3k/9-month saving) is genuine cost/savings data but is a single-firm result, not a community financial-sufficiency benchmark — recorded here and available if I01 later wants a distributed-solar payback illustration, but not strong enough to add as a feeds: bullet.

Scope caveat. Everything measured is Nelson (solar) or Wellington (buses) specific; the CEMS is theoretical (“still need[s] to be developed… not on a community level”); and the thesis is engineering-narrative, not a policy dataset. It resolves the retrieval target but does not, on its own, deliver a national community-energy policy compendium.

Research targets

Resolved

• RT_024 (RESOLVED → this page): Kerr M.J. (2024) “Empowering Energy Innovation in the Communities of Aotearoa” retrieved and read in full. ⚠ It is a Masters of Engineering thesis (Univ. of Waikato), not a doctoral dissertation. Delivers NZ community-energy deployment/performance data (Chia industrial solar, Rowallan microgrid, Wellington HV2G) and policy context; the “policy data” anticipated is background-level, not a primary policy analysis.

Potential future work (not opened as targets): a PowerNet Rowallan full-cost/spec retrieval and an NZ school-bus HV2G benefit-cost/dispatch validation (the thesis’s own flagged Future Work). The Rowallan component sizes and the indicative HV2G figures already captured above are sufficient for D01’s present purposes.

Notes

Masters of Engineering thesis, University of Waikato (2024), author Michaela Jane Kerr (supervisor Mark Apperley). Single-file PDF (94 pp) read in full via pdftotext -layout — every cited figure is traceable to the raw, but the source is not peer-reviewed and carries the usual thesis roughness, so data_quality: medium, not verified:

• Unit sloppiness throughout — power is repeatedly written as energy-per-time (“16kW per hour”, “8kW per hour”, “6.94kWh/m²” mixed with radiation, “60 kW/h” chargers, “MJ/J by 3.6”). The intended physical quantities are clear from context and preserved above, but figures should be sanity-checked on units before entering any NI cell.
• Self-flagged rough estimates — the author explicitly notes her own Chia 16 kW / 8 kW capacity statement is “not correct” once monitored; the HV2G “~1,000–1,500 homes” is imported from a Mercury estimate for a different Auckland battery; the school-bus MW figures are Excel proof-of-concept outputs, not a validated grid model.
• Secondary national figures — the 68% demand growth, 8,600→13,700 MW, 60:40 energy split, ~80% grid renewables, ICP/DS statistics etc. are cited from Transpower / MBIE / EECA / Energy Resources Aotearoa and were NOT verified against those primaries here; use as corroboration, not as primaries.

⚠ Citation-type correction: this work is a Masters thesis, not a doctoral dissertation (verified against the raw’s cover page: “submitted in fulfilment of the requirements for the Degree of Masters of Engineering at The University of Waikato by Michaela Jane Kerr, 2024”). The source_type: thesis / display_name here record the correct level.

⚠ URL not fully resolved: the raw’s cover page cites the Waikato Research Commons (researchcommons.waikato.ac.nz) but the exact item handle (/handle/10289/…) is not printed in the PDF. Frontmatter records the repository base.

Prefix/store decision: filed as lit_ / source_type: thesis (academic research thesis with case studies + simulation), matching the LIT_012 thesis precedent, rather than ot_ (used for the design-portfolio thesis OT_018). Store: source-files/01_literature/.

Connections

Links to

Sources (8): LIT_002 · LIT_012 · LIT_031 · LIT_032 · LIT_033 · LIT_068 · OT_018 · URL_003

Referenced by

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