Doc #40 — Refrigeration Transition

---

RECOMMENDED ACTIONS (BY ACTUAL URGENCY)

First months (Phase 1)

1. Inventory all refrigerant stocks nationally — include HVAC wholesale distributors, mechanical contractors, supermarket chains’ maintenance stores, dairy factory maintenance stores, hospital facilities management. Classify by type (R-134a, R-410A, R-404A, R-32, R-22, R-290, R-717, others). Include this in the national asset census (Doc #8) and consumables requisition (Doc #7).

2. Classify refrigerant as a controlled substance under emergency rationing — no servicing of non-essential refrigeration with scarce HFC stocks. Priority allocation to: (a) pharmaceutical cold chain, (b) dairy and meat processing, (c) essential food cold storage, (d) hospital and laboratory cooling.

3. Issue leak minimisation guidance to all refrigeration operators — proper maintenance, regular leak checks, tightening connections, replacing degraded seals and valves where possible. Every gram of refrigerant conserved extends the cold chain. [Urgency: moderate — this matters over months and years, not days]

4. Inventory existing ammonia refrigeration systems in NZ — identify every operating ammonia plant and its capacity. These become strategic cold chain assets. Identify ammonia stocks nationally (agricultural, industrial, cleaning products).

5. Identify and secure all qualified refrigeration technicians — these are among the most strategically important tradespeople. Their skills are needed both for maintaining existing systems and for building new ones. Register them through the skills census (Doc #8).

First year (Phase 1–2)

6. Begin ammonia system construction at priority sites — large cold stores, dairy factories, and meat processing plants that currently use HFC systems and will lose cooling as refrigerant depletes. Target the highest-volume, most critical cold chain nodes first.

7. Begin hydrocarbon refrigerant conversion program for domestic and small commercial refrigerators. R-600a (isobutane) conversions for domestic refrigerators require trained technicians with recovery equipment, vacuum pumps, precision charging scales, and model-specific knowledge (see Section 3.2) — the process is well-understood but not trivial, and flammability risk requires careful handling. R-290 (propane) for small commercial units.

8. Commission absorption refrigeration pilot projects at geothermal sites — Taupō Volcanic Zone facilities with access to geothermal heat can operate ammonia-water absorption chillers for cold storage without any compressor or synthetic refrigerant.

9. Establish refrigeration technician training program as part of accelerated trade training (Doc #157). Existing HVAC technicians retrain for ammonia and hydrocarbon systems. New trainees enter the pipeline.

Ongoing (Phase 2+)

10. Progressively decommission non-essential refrigeration — recover refrigerant from abandoned domestic and commercial units for reuse in priority systems. This extends the HFC bridge period.

11. Expand ammonia and absorption systems as construction capacity allows. Target: every major food processing facility and regional cold store on alternative refrigerant by end of Phase 2.

12. Develop ice-making capability using ammonia or absorption systems for distributed cold chain — ice transport was the standard cold chain method before mechanical refrigeration and remains viable for short-distance distribution.

---

ECONOMIC JUSTIFICATION

The value of the cold chain

Refrigeration underpins three critical recovery functions:

Food safety and preservation. NZ’s dairy industry — approximately 11,000–12,000 farms producing roughly 22 billion litres of milk per year pre-event⁴ — depends on refrigerated vat storage at the farm and cold processing at factories. Meat processing (approximately 50–60 export-licensed plants)⁵ requires refrigeration for carcass chilling, aging, and storage. Without cold chain, these industries must either process everything immediately into shelf-stable forms (cheese, butter, salted/smoked meat) or accept massive spoilage losses.

Pharmaceutical preservation. Insulin, biologics, vaccines, and other cold-chain pharmaceuticals are irreplaceable (Doc #116, Doc #4). Loss of pharmaceutical cold chain means loss of these medications — a direct threat to the lives of the approximately 250,000 New Zealanders with diabetes and others dependent on temperature-sensitive drugs.⁶

Food storage for population. NZ has approximately 1.8–1.9 million households, virtually all with at least one refrigerator.⁷ Domestic refrigeration extends food shelf life from days to weeks, reducing waste and reducing the frequency of food distribution — both critical under rationing conditions.

Cost of alternative systems

Ammonia system construction for a medium-sized cold store (500–1,000 m3 storage volume, approximately 100–200 kW refrigeration capacity) requires:

• Materials: Steel pipe and fittings (from NZ Steel, Doc #89), welding (Doc #94), insulation (wool or wood fibre — Doc #163 techniques are adaptable), compressor (salvaged or fabricated — see Section 7), electrical controls, ammonia charge (50–200 kg depending on system size)
• Labour: Estimated 4–8 person-months for design, fabrication, installation, and commissioning, assuming trained personnel
• Ongoing: 0.5–1 person per facility for operation and maintenance

Hydrocarbon conversion of existing domestic refrigerators requires approximately 2–4 person-hours per unit, plus a small charge of isobutane (typically 50–150 grams per domestic refrigerator).⁸ At scale — converting, say, 500,000 priority domestic units — this represents approximately 1,000,000–2,000,000 person-hours, or roughly 500–1,000 person-years of technician time spread over several years. This is skilled refrigeration technician time — a scarce pool — not general labour that can be staffed from the wider unemployed workforce, making the real cost higher than the raw person-year figure suggests.

Comparison against alternatives

The alternative to maintaining mechanical refrigeration is a full transition to non-refrigerated food preservation (Doc #78) — salting, smoking, drying, fermentation. These methods work and should be expanded regardless, but they cannot fully replace refrigeration for:

• Dairy processing (fresh milk, yoghurt, and soft cheese require cold chain)
• Pharmaceutical storage (no substitute for cold chain)
• Meat quality (refrigerated aging produces better outcomes than immediate salt-curing of all product)

Mātauranga Māori (Doc #160) includes a substantial body of food preservation knowledge developed over centuries without mechanical refrigeration. These are tested, NZ-specific techniques for managing food safety and extending shelf life using local materials:

• Smoking and drying of fish and meat: Actively practised in many coastal Māori communities. Applies to surplus protein from pastoral destocking (Doc #78).
• Tītī (muttonbird) preservation: Ngāi Tahu communities harvest and preserve tītī in their own fat in pōhā (bull kelp containers), producing a product that keeps for over a year without refrigeration. A demonstrated large-scale protein preservation system using entirely NZ materials.
• Rua kūmara (storage pits): Underground pit storage for kūmara and root vegetables, using earth’s thermal mass to maintain stable cool temperatures. Directly applicable to household and community-scale food storage as domestic refrigeration declines.
• Pātaka (elevated food stores): Raised storage structures designed to protect dried and preserved foods from pests and moisture while allowing airflow. A practical design principle for non-refrigerated bulk food storage.
• Natural cold sources: Spring water, earth cellars, and high-altitude storage provide passive cooling without any mechanical system. These were understood and used in pre-European NZ and remain viable.
• Traditional food spoilage indicators: Knowledge of when preserved foods are safe to consume — colour, smell, texture changes — is partly encoded in traditional knowledge systems and is practically important as households shift away from refrigeration.

These methods reduce the total demand that mechanical refrigeration must serve, and they remain functional across all phases of refrigerant depletion. Integration with Doc #160 recommendations is appropriate: engaging Māori knowledge holders through iwi and marae networks to document and teach these preservation methods is a direct refrigeration transition measure.

Assessment: Investing in alternative refrigeration systems is justified for food processing, pharmaceutical storage, and critical cold chain nodes. Maintaining universal domestic refrigeration at current levels is neither feasible nor necessary — households will increasingly rely on non-electric preservation methods, traditional knowledge, and the cooler ambient temperatures of nuclear winter. The priority is industrial and medical cold chain preservation.

---

1. NZ’S REFRIGERATION INFRASTRUCTURE

1.1 Scale and composition

NZ’s refrigeration infrastructure includes: an estimated 200–400 commercial cold storage and food processing facilities (meat, dairy, seafood, produce — concentrated in Waikato, Canterbury, Southland, Hawke’s Bay)⁹; approximately 1,000–1,200 supermarkets with extensive refrigerated display and storage¹⁰ (a single large supermarket may hold 200–500 kg of refrigerant)¹¹; approximately 1.8–1.9 million domestic refrigerators¹²; an estimated 400,000–500,000 heat pumps (primarily for heating — Doc #163)¹³; a fleet of refrigerated trucks numbering in the low thousands; and approximately 11,000–12,000 dairy farm milk cooling vats.¹⁴

The equipment uses a mix of refrigerants: R-134a and R-404A dominate commercial and industrial systems, R-410A and R-32 in heat pumps, R-134a or R-600a in domestic units, and ammonia in some large industrial systems. Total refrigerant charge across all NZ equipment is likely several thousand tonnes.

1.2 Refrigerant stock and depletion timeline

In-country service stocks. NZ’s HVAC and refrigeration wholesale distributors hold service refrigerant in cylinders. The total volume at any given time is uncertain but is estimated at 300–600 tonnes across all types, based on the relationship between annual import volume (~1,000–1,500 tonnes) and typical distributor inventory levels (a few months of supply).¹⁵ This is an assumption that should be verified through the national asset census.

Refrigerant charged in equipment. The total refrigerant charge across all NZ equipment is much larger than the service stock — likely several thousand tonnes. However, this refrigerant is not available for reuse while the equipment is operating. It becomes available only when equipment is decommissioned and the refrigerant is recovered.

Depletion mechanism. Refrigerant is lost through:

• Leaks: Joints, seals, valve stems, and corrosion points allow slow refrigerant escape. Annual leak rates vary widely: hermetically sealed domestic units may lose less than 1% per year; large commercial systems with many joints and service valves may lose 5–15% per year; poorly maintained or older systems may lose more.¹⁶
• Service losses: Each time a system is opened for repair, some refrigerant is lost despite best practices.
• Equipment failure: Compressor burnout, pipe rupture, or corrosion failure can release the entire charge at once.

Estimated depletion timeline. This is necessarily rough:

• Domestic refrigerators and freezers: Hermetically sealed units with small charges. Many will continue operating for 10–20 years without refrigerant service, failing eventually from compressor wear, electrical component failure, or corrosion rather than refrigerant loss. These are the longest-surviving segment.
• Commercial and industrial systems: Higher leak rates, larger charges, more service points. Without recharge capability, systems begin failing within 2–5 years as cumulative leakage degrades performance below useful levels. Well-maintained systems in good condition last longer; older, poorly maintained systems fail sooner.
• Heat pumps: Split systems (indoor and outdoor units connected by refrigerant lines) have field-made connections that are prone to slow leaks. An estimated 5–15% annual failure rate from refrigerant-related causes is plausible based on the higher leak rates of field-connected split systems, though this figure is uncertain and should be verified through monitoring data.
• Transport refrigeration: Vibration and thermal cycling accelerate leaks. These may be the first systems to lose charge — perhaps 2–4 years average without recharge.

The bridge period: With aggressive refrigerant conservation (leak repair, recovery from decommissioned units, priority allocation), NZ could maintain critical cold chain functions for perhaps 5–10 years on existing HFC stocks. This is the window for transitioning to alternative systems. If alternative systems are not ready when HFC stocks are exhausted, cold chain functions served by those systems are lost.

---

2. AMMONIA REFRIGERATION

2.1 Why ammonia is the primary alternative

Ammonia (NH3, refrigerant designation R-717) was the dominant industrial refrigerant from the 1850s through the mid-20th century and remains widely used in large food processing and cold storage worldwide.¹⁷ Its high latent heat of vaporisation (approximately 1,370 kJ/kg at -33°C) means less refrigerant mass is needed per unit of cooling compared to HFCs.¹⁸ Its pungent odour is detectable at concentrations well below dangerous levels (5–50 ppm detection vs. 300 ppm IDLH), providing a natural leak alarm.¹⁹ Most importantly, ammonia is produced domestically at the Ballance Kapuni plant in Taranaki as an intermediate in urea production (Doc #7).

2.2 What ammonia requires

Ammonia refrigeration is not a drop-in replacement for HFC systems. Converting or building ammonia systems involves:

Steel piping and components. Ammonia is incompatible with copper and copper alloys — it corrodes brass, bronze, and copper pipe, which are standard in HFC systems.²⁰ Ammonia systems use carbon steel pipe, steel valves, and steel or stainless steel fittings. NZ Steel (Doc #89) produces structural steel, and NZ has welding capability (Doc #94) for steel pipe fabrication. Steel piping is heavier and more labour-intensive to install than copper, and requires welded joints rather than the brazed or flared connections used in HFC systems.

Compressors. Ammonia compressors differ from HFC compressors — they typically use reciprocating or screw designs with steel internals and mineral oil lubrication (ammonia is miscible with some synthetic oils but mineral oil is standard). NZ’s existing ammonia compressors are a finite strategic asset. Fabricating new ammonia compressors is within NZ’s machining capability (Doc #91) but represents significant precision engineering — cylinder bores, valve plates, shaft seals, and bearings must be machined to close tolerances. A realistic timeline for fabricating a medium-capacity ammonia compressor from NZ-produced steel in a well-equipped machine shop is 2–6 months per unit, assuming trained machinists and suitable materials. This is an estimate; the actual difficulty depends heavily on the specific design and the capabilities of the shop.

Safety systems. Ammonia is toxic. At concentrations above 300 ppm it is immediately dangerous to life; at 25 ppm it causes irritation; liquid ammonia causes severe chemical burns.²¹ Ammonia refrigeration plants require:

• Ventilated machinery rooms (natural or forced ventilation to outside)
• Ammonia gas detectors and alarms
• Emergency ventilation that activates on leak detection
• Self-contained breathing apparatus (SCBA) accessible to operators
• Emergency shower and eyewash stations
• Written emergency response procedures
• Trained operators who understand the system and its hazards

These requirements are well-established — NZ’s existing ammonia plants comply with them, and the relevant NZ standards (AS/NZS 1677, NZS 5442) provide design guidance.²² The challenge is scaling this to new installations under constrained conditions with potentially limited instrumentation and safety equipment.

Ammonia supply. A medium cold store ammonia system requires perhaps 50–200 kg of ammonia charge, depending on system size and design.²³ Total NZ demand for new ammonia refrigeration systems over the transition period might be 50–200 tonnes — a fraction of the Kapuni plant’s production capacity (which produces ammonia as an intermediate in urea manufacturing at roughly 200,000 tonnes per year).²⁴ Ammonia supply is not a constraint if the Kapuni plant continues operating. If the Kapuni plant fails, ammonia becomes much harder to source — small-scale ammonia production from other nitrogen fixation methods is theoretically possible but difficult (Doc #114).

2.3 NZ’s existing ammonia refrigeration and applicability

NZ already operates ammonia refrigeration at some large cold stores and food processing facilities — including some meat processing plants (Silver Fern Farms, ANZCO, Alliance), ice cream manufacturing (Tip Top/Fonterra), and large commercial cold storage operators. The exact number is not publicly documented. These systems are strategic assets — proven cold chain capacity independent of imported refrigerant. The national asset census (Doc #8) should identify every ammonia system, its capacity, condition, and operating personnel.

Ammonia is suitable for large cold stores, food processing plants, ice production, and institutional facilities with trained operators. It is not suitable for domestic refrigerators (toxicity risk in enclosed kitchens), small commercial units (disproportionate safety infrastructure), road vehicles (crash and proximity risk), or any location without trained operators.

---

3. HYDROCARBON REFRIGERANTS

3.1 Propane (R-290) and isobutane (R-600a)

Hydrocarbon refrigerants fill the gap that ammonia cannot — small-scale and domestic refrigeration. They have been used in domestic refrigerators since the 1990s (particularly in Europe, where R-600a is the dominant domestic refrigerant in new appliances).²⁵

Properties:

Property	R-600a (Isobutane)	R-290 (Propane)	R-134a (HFC, for comparison)
Boiling point (°C)	-11.7	-42.1	-26.1
ODP	0	0	0
GWP	~3	~3	1,430
Flammability	Yes (A3)	Yes (A3)	No (A1)
Typical domestic charge	50–100 g	100–200 g	100–200 g

Key advantage: Hydrocarbons are excellent refrigerants — high energy efficiency, good thermodynamic properties, compatible with existing mineral oil lubricants, and widely available. NZ has abundant propane and isobutane from its LPG supply (Taranaki natural gas processing, and LPG import stocks).²⁶

Key risk: Flammability. Propane and isobutane are flammable gases (lower explosive limit approximately 2.1% and 1.8% by volume in air, respectively).²⁷ A refrigerant leak from a domestic refrigerator releases a small quantity (50–200 g), which in a typical kitchen volume (30 m3) produces a concentration well below the lower explosive limit. This is why hydrocarbon domestic refrigerators are considered safe by international standards (IEC 60335-2-24) for charge limits up to 150 g.²⁸ However:

• Larger charges in commercial equipment increase explosion risk
• Leak into a confined space (under a counter, in a closed cabinet) can create locally flammable concentrations
• Ignition sources near refrigerant connections must be managed
• Technicians must be trained to handle flammable refrigerants — different procedures from HFC work

3.2 Conversion of existing equipment

Many existing R-134a domestic refrigerators can be converted to R-600a by recovering the existing charge, flushing with dry nitrogen, adjusting or replacing the capillary tube (R-600a has different flow characteristics), and recharging at approximately 40–60% of the original R-134a charge weight. This requires a trained technician with recovery equipment, vacuum pump, precision charging scale, and model-specific knowledge. Poorly executed conversions cause system damage or safety hazards. A realistic conversion rate is perhaps 4–8 units per technician per day.²⁹ Converted units may show a modest reduction in cooling capacity (estimated 5–15% depending on the original system design and capillary tube match) compared to the original R-134a charge, and compressor discharge temperatures may differ, potentially affecting compressor longevity — these are acceptable trade-offs for continued operation, but the performance gap should be acknowledged.

Commercial system conversion to R-290 is more complex — larger charge quantities increase flammability risk, and existing multi-kilogram-charge rack systems may require redesign to use multiple small, self-contained hydrocarbon circuits.

3.3 Hydrocarbon supply

NZ has domestic LPG production from Taranaki gas processing and imported LPG stocks. LPG is a mixture of propane and butane in varying proportions. Separating propane from butane requires fractional distillation, which is within NZ’s industrial chemistry capability but not currently set up for producing refrigerant-grade product.

Refrigerant-grade propane and isobutane require higher purity than fuel-grade LPG — specifically, moisture content must be below approximately 10 ppm and non-condensable gas content must be minimal.³⁰ Producing refrigerant-grade hydrocarbon from fuel-grade LPG requires: fractional distillation equipment capable of separating propane from butane (a distillation column with sufficient theoretical plates, condenser, and reboiler); drying using molecular sieve beds or desiccant (molecular sieve material is imported and has finite capacity before requiring regeneration); filtration to remove particulates; and quality testing equipment to verify purity meets AHRI 700 specifications. The distillation column and associated equipment could be fabricated from NZ steel (Doc #89), but molecular sieve adsorbent is an imported consumable with limited NZ stocks. This is an industrial process requiring purpose-built equipment and trained chemical process operators — not something that can be done with a propane bottle from a barbecue.

Isobutane availability. Isobutane (2-methylpropane) is present as a minor component in some LPG streams. It can also be produced by isomerisation of n-butane, which requires a catalyst and elevated temperature. NZ’s ability to produce refrigerant-grade isobutane at scale is uncertain — this needs assessment. In the near term, existing isobutane stocks (from aerosol propellant supplies and industrial gas distributors) may be the primary source.

---

4. ABSORPTION REFRIGERATION

4.1 How it works

Absorption refrigeration produces cooling using heat rather than mechanical compression. The most common system uses an ammonia-water solution:

1. A generator (heated by electricity, geothermal water, combustion, or any heat source above ~120°C) boils ammonia out of an ammonia-water solution
2. The ammonia vapour condenses in a condenser (releasing heat to ambient)
3. The liquid ammonia evaporates in an evaporator (absorbing heat — this is where cooling occurs)
4. The ammonia vapour is absorbed back into a weak ammonia-water solution in an absorber
5. The cycle repeats

The key advantage is that the only moving part is a small solution pump (and in some designs, even this is eliminated using bubble-pump or thermosiphon arrangements). There is no compressor — the device that is hardest to manufacture and most prone to failure in vapour-compression systems.

4.2 Performance and limitations

Absorption systems have a COP of approximately 0.5–0.7 — meaning they use 3–8 times more energy per unit of cooling than vapour-compression systems (COP 2–4).³¹ In practical terms, a 50 kW absorption chiller requires 70–100 kW of continuous heat input, compared to 12–25 kW of electricity for an equivalent vapour-compression system. If the heat source is geothermal or waste heat that would otherwise be discarded, this penalty is irrelevant — the energy is free. If electricity must be used to generate the heat, absorption is a poor choice compared to vapour-compression. Ammonia-water systems can achieve evaporator temperatures as low as -30°C to -40°C, suitable for freezing applications. They can be built at virtually any scale. The main limitations are lower efficiency (making them unsuitable where the heat source is electricity or scarce fuel), slower response to load changes (minutes rather than seconds to adjust cooling output), the requirement for a continuous heat source above ~120°C, larger physical size (roughly 2–3 times the footprint of an equivalent vapour-compression unit), and the remaining ammonia toxicity hazard (though contained within a sealed system with fewer leak points than a vapour-compression system).

4.3 NZ-specific opportunities

Geothermal-powered absorption chillers. NZ’s Taupō Volcanic Zone provides geothermal heat at temperatures well above 120°C. A geothermal-powered absorption chiller at a Taupō-area cold store could provide refrigeration with no imported fuel, no imported refrigerant, and no compressor — the entire system could be fabricated from NZ steel and charged with domestically produced ammonia. This is one of the most self-sufficient refrigeration options available to NZ.

Potential geothermal cold storage sites include:

• Taupō — central to the North Island, road and rail access
• Rotorua — existing geothermal infrastructure, large population
• Kawerau — existing industrial geothermal use (Norske Skog paper mill, geothermal power station)
• Mokai/Ohaaki — existing geothermal development

Waste heat from industrial processes. Any factory producing waste heat above 120°C can drive an absorption chiller. Candidates include:

• The Glenbrook steel mill (Doc #106) — significant waste heat from the electric arc process
• Geothermal power stations (waste heat from rejected steam or binary cycle outlets)
• Charcoal kilns (Doc #98) — waste heat during burns
• Glass production (Doc #98) — furnace waste heat

4.4 Construction feasibility

An ammonia-water absorption chiller can be constructed from:

• Steel pressure vessels (generator, absorber, condenser, evaporator) — fabricable from NZ steel by qualified pressure vessel welders
• Steel pipe and fittings — standard materials available from NZ sources
• Heat exchangers — shell-and-tube designs fabricable in NZ machine shops
• Insulation — wool or wood fibre (see Doc #163)
• Ammonia charge — from Kapuni production
• Solution pump — the only moving part; a small, low-pressure pump that can be fabricated in a machine shop

The dependency chain is short compared to vapour-compression systems: NZ steel, NZ welding capability, NZ ammonia, and NZ engineering knowledge. No imported compressor, no imported refrigerant, no imported electronic controls. This makes absorption refrigeration particularly attractive for long-term self-sufficiency.

Estimated construction effort: A 50 kW absorption chiller suitable for a community cold store would require approximately 6–12 person-months of fabrication and installation by a team including a mechanical engineer, pressure vessel welder, and refrigeration technician. This is a rough estimate; the actual effort depends heavily on the specific design and the fabrication capability available.

---

5. TRANSITION TIMELINE AND END STATE

5.1 Nuclear winter advantage

Lower ambient temperatures during nuclear winter (estimated -5°C to -15°C reduction from normal, Doc #74) significantly reduce cooling loads. Refrigeration systems cycle less frequently, reducing compressor wear and refrigerant leakage from vibration and thermal cycling. Outdoor ambient temperatures in many NZ regions during nuclear winter may be near or below refrigeration target temperatures (4°C) for much of the year, meaning that well-insulated cool stores may require little or no mechanical cooling during cooler months. This is a meaningful benefit — it extends the HFC bridge period.

5.2 What the end state looks like

By Phase 3–4 (years 7–15), NZ’s cold chain will look fundamentally different from today:

• Large-scale cold storage and food processing: Served by ammonia vapour-compression and/or absorption systems, fabricated from NZ materials. These facilities continue to operate, providing critical cold chain for dairy, meat, pharmaceutical, and food distribution.
• Regional and community cold storage: A mix of ammonia systems (where scale justifies them), absorption chillers (where geothermal or waste heat is available), and ice-based cooling (transported from production facilities).
• Domestic refrigeration: Significantly reduced. Some households retain operating refrigerators (either original units still functioning or hydrocarbon-converted units). Many households rely on cold cellars, outdoor cooling in cooler months, community cold stores, and non-electric food preservation. This is a genuine quality-of-life regression that should be acknowledged.
• Pharmaceutical cold chain: Preserved at hospital and pharmacy level through priority ammonia or hydrocarbon systems. Pharmaceutical storage consolidated into fewer, better-protected locations (consistent with Doc #4 recommendations).
• Transport refrigeration: Largely lost. Food distribution relies on insulated containers with ice, rapid distribution of perishables, and a shift toward shelf-stable products.

---

6. ICE AS A COLD CHAIN MEDIUM

Before mechanical refrigeration became widespread in the early 20th century, manufactured ice was the standard cold chain medium.³² An ammonia or absorption refrigeration plant can produce ice in block moulds (25–50 kg blocks), which is then stored in insulated warehouses and distributed by truck or cart. Ice production is energy-intensive (approximately 25–30 kWh of electricity per tonne via vapour-compression, or 130–190 kWh of heat via absorption)³³ but extends cold chain to locations without mechanical refrigeration.

Under nuclear winter, natural ice may be available in inland South Island during winter — a free cooling resource. Ice-based cold chain is most suitable for milk collection from farms, transport of perishables, community-level food storage, and fish preservation at harbours.

---

7. COMPRESSOR SOURCING AND FABRICATION

NZ does not manufacture refrigeration compressors. The primary near-term strategy is salvage and reuse from the approximately 2.2–2.5 million compressors in existing domestic refrigerators, heat pumps, and commercial systems (see Section 1.1). As non-essential systems are decommissioned, their compressors can be recovered, tested, and redeployed. Matching is important — compressors are designed for specific refrigerants, pressures, and temperature ranges, and an R-134a compressor cannot be used with ammonia due to copper incompatibility.³⁴

Fabricating reciprocating ammonia compressors in NZ machine shops (Doc #91) is feasible but demanding. The dependency chain includes: NZ-produced steel (Doc #89) for cylinders, valve plates, and housings; spring steel for reed valves (availability depends on existing spring steel stock or the ability to heat-treat carbon steel to appropriate hardness); bearing materials (bronze bearings are not suitable for ammonia-wetted surfaces — steel or Babbitt metal bearings are needed); gasket materials for shaft seals; and precision machine tools capable of boring and honing cylinders to 0.02–0.05 mm clearance.³⁵ A machine shop focused on compressor production might produce 4–10 medium-capacity units per year with a team of 3–5 machinists, though this estimate is uncertain and depends heavily on the specific machine shop’s capabilities and tooling. Early NZ-fabricated compressors will likely have shorter service lives (perhaps 3,000–8,000 operating hours vs. 20,000–40,000 for factory-built units) than imported units, with iterative improvement over successive generations.

---

8. CRITICAL UNCERTAINTIES / KEY RISKS

Uncertainty	Why it matters	How to resolve
Total NZ refrigerant stock (service cylinders + charged in equipment)	Determines the HFC bridge period — how long existing cold chain functions	National asset census (Doc #8)
Leak rates across NZ’s installed equipment base	Determines how fast the stock depletes	Monitoring and data collection from serviced systems
Number and capacity of existing ammonia refrigeration systems in NZ	These are the strategic cold chain assets that do not depend on imports	Census of industrial refrigeration facilities
Kapuni ammonia plant operational continuity	If the plant fails, ammonia supply for new refrigeration systems becomes very difficult	Designated national strategic asset (Doc #7); secure workforce and consumables
NZ’s ability to produce refrigerant-grade hydrocarbons from LPG	Determines whether domestic refrigerator conversion is feasible at scale	Technical assessment and pilot production
Compressor fabrication capability and quality	Determines whether NZ can build new vapour-compression systems or must rely on salvaged compressors	Pilot fabrication program in partnership with machine shops (Doc #91)
Training pipeline for ammonia and hydrocarbon refrigeration technicians	These skills do not currently exist at scale in NZ — most HVAC technicians work with HFC systems	Accelerated training (Doc #157)
Nuclear winter duration and severity	Cooler temperatures extend the bridge period but also increase space heating demand (potentially competing for electricity)	Monitoring; plan for range of scenarios

---

CROSS-REFERENCES

Document	Relationship
Doc #1 — National Emergency Stockpile Strategy	Refrigerant stocks included in consumables requisition
Doc #3 — Food Rationing and Distribution	Cold chain management for food distribution
Doc #4 — Pharmaceutical and Medical Supply Management	Cold chain for insulin, biologics, vaccines
Doc #7 — Agricultural and Industrial Consumables	Refrigerant included; Kapuni ammonia plant as strategic asset
Doc #156 — Skills Census	Inventory of refrigerant stocks, refrigeration equipment, and skilled technicians
Doc #21 — Chemical Safety Data	Ammonia safety data and handling procedures
Doc #34 — Lubricant Production	Compressor lubrication requirements
Doc #74 — Pastoral Farming Under Nuclear Winter	Nuclear winter temperature reduction extends cold chain
Doc #74 — Dairy Adaptation	Milk cooling requirements; dairy cold chain dependence
Doc #78 — Food Preservation Without Imports	Non-refrigerated preservation as the long-term transition
Doc #160 — Heritage Skills Preservation	Traditional food preservation (smoking, drying, fat-packing, pit storage, pātaka design) as refrigeration transition support
Doc #88 — Spare Parts Triage	Refrigeration compressors and components in spare parts inventory
Doc #89 — NZ Steel Glenbrook	Steel for ammonia system piping and pressure vessels
Doc #91 — Machine Shop Operations	Compressor fabrication capability
Doc #94 — Welding Consumable Fabrication	Welding for ammonia system construction
Doc #102 — Charcoal Production	Waste heat for absorption chiller operation
Doc #114 — Ammonia Synthesis	Long-term ammonia supply if Kapuni fails
Doc #116 — Pharmaceutical Rationing	Cold chain critical for insulin and biologics
Doc #157 — Accelerated Trade Training	Refrigeration technician training
Doc #163 — Housing Insulation	Heat pump fleet decline; refrigerant depletion for heat pumps

---

---

1. New Zealand’s Ozone Layer Protection Act 1996 implemented the Montreal Protocol, banning import of CFC refrigerants and phasing down HCFCs. NZ subsequently ratified the Kigali Amendment to the Montreal Protocol, committing to HFC phase-down. The Environmental Protection Authority (EPA) administers import permits for ozone-depleting substances and HFCs. See: https://www.epa.govt.nz/industry-areas/hazardous-substanc...↩︎

2. NZ refrigerant import volumes are tracked by the EPA under HFC import levy and permit systems. The 1,000–1,500 tonne estimate is based on NZ’s market size relative to Australia (which imports approximately 8,000–10,000 tonnes of HFCs per year for a population roughly 5 times NZ’s). This is an estimate that should be verified against EPA import data. The actual figure may be higher or lower depending on NZ’s specific equipment mix and service demand patterns.↩︎

3. Refrigerant leak rates vary widely by system type and condition. The 2–15% range is based on industry surveys and regulatory impact assessments. The European Commission’s review of the F-Gas Regulation cited average annual leak rates of approximately 3–6% for commercial refrigeration systems and 1–3% for smaller sealed systems. Older systems with more joints and service valves tend toward the higher end. Source: European Commission F-Gas Regulation impact assessment; various HVAC industry publications.↩︎

4. DairyNZ statistics. NZ has approximately 11,000–12,000 dairy farms. Total milk production approximately 22 billion litres per year (2023/24 season). See: https://www.dairynz.co.nz/↩︎

5. NZ meat processing capacity is distributed across approximately 50–60 export-licensed plants. Source: MPI (Ministry for Primary Industries) registered premises data.↩︎

6. Diabetes NZ and Ministry of Health data. NZ has approximately 250,000 people diagnosed with diabetes, of whom approximately 50,000–60,000 have Type 1 or insulin-dependent Type 2 diabetes. Source: https://www.diabetes.org.nz/↩︎

7. Stats NZ household data. NZ has approximately 1.8–1.9 million households. Refrigerator ownership is near-universal. Source: https://www.stats.govt.nz/↩︎

8. Hydrocarbon refrigerant charge quantities for domestic refrigerators: IEC 60335-2-24 sets a maximum flammable refrigerant charge of 150 g for domestic appliances. Typical R-600a charges in European domestic refrigerators are 40–80 g. Conversion from R-134a to R-600a typically uses approximately 40–60% of the original R-134a charge weight. Source: IEC standards and manufacturer technical data.↩︎

9. The number of commercial cold storage facilities in NZ is not precisely documented in public sources. The estimate of 200–400 is based on the combined meat processing, dairy processing, seafood processing, produce cool store, and general cold storage sector. The actual number should be established through the national asset census.↩︎

10. NZ retail grocery store count is based on data from Woolworths NZ (Countdown), Foodstuffs (New World, PAK’nSAVE, Four Square), and independent grocers. Total is approximately 1,000–1,200 stores.↩︎

11. Supermarket refrigerant charges: A large supermarket with extensive refrigerated display cases, walk-in cool rooms, and freezer cabinets may contain 200–500 kg of refrigerant across multiple circuits. This is based on typical commercial refrigeration system specifications. Source: HVAC industry technical literature; Environmental Investigation Agency supermarket refrigerant surveys.↩︎

12. Stats NZ household data. NZ has approximately 1.8–1.9 million households. Refrigerator ownership is near-universal. Source: https://www.stats.govt.nz/↩︎

13. NZ heat pump installations are estimated at 400,000–500,000 units based on industry sales data and import volumes accumulated over approximately 20 years of strong market growth. EECA (Energy Efficiency and Conservation Authority) and the NZ Heat Pump Association provide relevant data. The exact installed base is uncertain.↩︎

14. DairyNZ statistics. NZ has approximately 11,000–12,000 dairy farms. Total milk production approximately 22 billion litres per year (2023/24 season). See: https://www.dairynz.co.nz/↩︎

15. NZ refrigerant import volumes are tracked by the EPA under HFC import levy and permit systems. The 1,000–1,500 tonne estimate is based on NZ’s market size relative to Australia (which imports approximately 8,000–10,000 tonnes of HFCs per year for a population roughly 5 times NZ’s). This is an estimate that should be verified against EPA import data. The actual figure may be higher or lower depending on NZ’s specific equipment mix and service demand patterns.↩︎

16. Refrigerant leak rates vary widely by system type and condition. The 2–15% range is based on industry surveys and regulatory impact assessments. The European Commission’s review of the F-Gas Regulation cited average annual leak rates of approximately 3–6% for commercial refrigeration systems and 1–3% for smaller sealed systems. Older systems with more joints and service valves tend toward the higher end. Source: European Commission F-Gas Regulation impact assessment; various HVAC industry publications.↩︎

17. Ammonia was first used as a refrigerant by Ferdinand Carré in 1859 and has been in continuous industrial use since. It remains the dominant refrigerant in large industrial refrigeration worldwide, particularly in food processing and cold storage. Source: ASHRAE Fundamentals Handbook; IIR (International Institute of Refrigeration) publications.↩︎

18. Ammonia thermodynamic properties: Latent heat of vaporisation approximately 1,370 kJ/kg at -33°C (atmospheric boiling point). Charge quantities for ammonia systems are typically lower than for equivalent HFC systems due to ammonia’s higher latent heat. Source: ASHRAE Fundamentals Handbook; standard refrigeration engineering references.↩︎

19. Ammonia toxicity: IDLH (Immediately Dangerous to Life and Health) concentration is 300 ppm. Odour threshold is approximately 5–50 ppm. OSHA PEL (permissible exposure limit) is 50 ppm TWA. Ammonia causes severe irritation at 50–100 ppm and is potentially fatal at concentrations above 300 ppm with prolonged exposure. Source: NIOSH Pocket Guide to Chemical Hazards; Doc #21.↩︎

20. Ammonia’s incompatibility with copper is well-established in refrigeration engineering. Ammonia reacts with copper alloys in the presence of moisture, causing stress corrosion cracking. All ammonia system components must be steel, stainless steel, or aluminium. Source: ASHRAE Refrigeration Handbook; IIAR (International Institute of Ammonia Refrigeration) guidelines.↩︎

21. Ammonia toxicity: IDLH (Immediately Dangerous to Life and Health) concentration is 300 ppm. Odour threshold is approximately 5–50 ppm. OSHA PEL (permissible exposure limit) is 50 ppm TWA. Ammonia causes severe irritation at 50–100 ppm and is potentially fatal at concentrations above 300 ppm with prolonged exposure. Source: NIOSH Pocket Guide to Chemical Hazards; Doc #21.↩︎

22. NZ standards for ammonia refrigeration: AS/NZS 1677 “Refrigerating systems” covers design, construction, and safety requirements for refrigeration systems including ammonia. NZS 5442 relates to the safe handling and use of refrigerants. WorkSafe NZ provides guidance on ammonia refrigeration safety.↩︎

23. Ammonia thermodynamic properties: Latent heat of vaporisation approximately 1,370 kJ/kg at -33°C (atmospheric boiling point). Charge quantities for ammonia systems are typically lower than for equivalent HFC systems due to ammonia’s higher latent heat. Source: ASHRAE Fundamentals Handbook; standard refrigeration engineering references.↩︎

24. Ballance Agri-Nutrients Kapuni ammonia-urea plant. Capacity approximately 260,000 tonnes of urea per year, requiring corresponding ammonia production (approximately 200,000+ tonnes/year of ammonia as intermediate). See: https://www.ballance.co.nz/ ; Doc #7.↩︎

25. Hydrocarbon refrigerants in domestic appliances: R-600a (isobutane) has been the standard domestic refrigerant in most European-manufactured appliances since the 1990s, driven initially by German environmental policy (the “Greenfreeze” campaign by Greenpeace in 1992) and subsequently by EU F-Gas Regulation. Source: UNEP Technology and Economic Assessment Panel reports; various HVAC industry publications.↩︎

26. NZ LPG supply: LPG (propane/butane mix) is produced domestically from Taranaki natural gas processing and supplemented by imports. NZ’s total LPG consumption is approximately 150,000–200,000 tonnes per year. Source: MBIE Energy in New Zealand data.↩︎

27. Lower explosive limits: Propane 2.1% by volume in air; isobutane 1.8% by volume in air. Upper explosive limits: propane 9.5%; isobutane 8.4%. Source: Standard chemical safety references; NFPA data.↩︎

28. IEC 60335-2-24: International standard for safety of household refrigerating appliances. Sets maximum flammable refrigerant charge limits (150 g for isobutane/propane in sealed systems) to ensure that a complete release of refrigerant into a typical kitchen does not create a flammable concentration. This standard is adopted in NZ.↩︎

29. Conversion rate estimate based on practical experience with hydrocarbon refrigerant conversions. Each conversion requires system recovery, evacuation, possible capillary tube modification, recharging, and performance verification — approximately 2–4 hours for a straightforward domestic unit. More complex units or those requiring component replacement take longer.↩︎

30. Refrigerant-grade purity requirements: AHRI Standard 700 specifies purity requirements for refrigerants including moisture content (maximum 10 ppm for most refrigerants), non-condensable gas content, and other contaminant limits. Fuel-grade LPG does not meet these requirements without additional purification.↩︎

31. Absorption refrigeration COP and temperature ranges: Single-effect ammonia-water absorption systems achieve COP of approximately 0.5–0.7 using heat sources at 120–180°C. Double-effect systems achieve higher COP but require higher source temperatures and more complex construction. Evaporator temperatures as low as -40°C are achievable with ammonia-water systems. Source: ASHRAE Fundamentals and Refrigeration Handbooks; standard absorption refrigeration engineering references.↩︎

32. The natural ice trade was a major global industry from the early 19th century through the early 20th century. In the US alone, the ice harvesting industry employed tens of thousands of workers and shipped millions of tonnes of ice annually. Mechanical ice production using ammonia refrigeration largely replaced harvested ice by the 1920s–1930s. Source: Weightman, Gavin, “The Frozen Water Trade,” HarperCollins, 2003.↩︎

33. Ice production energy requirements: Producing one tonne of ice from water at 15°C requires approximately 93 kWh of thermal energy (to cool from 15°C to 0°C and then freeze). With a vapour-compression system COP of 3–4, electrical energy input is approximately 23–31 kWh per tonne. With an absorption system COP of 0.5–0.7, heat energy input is approximately 130–190 kWh per tonne. These are theoretical minimums — practical systems require 20–50% more due to losses.↩︎

34. Ammonia’s incompatibility with copper is well-established in refrigeration engineering. Ammonia reacts with copper alloys in the presence of moisture, causing stress corrosion cracking. All ammonia system components must be steel, stainless steel, or aluminium. Source: ASHRAE Refrigeration Handbook; IIAR (International Institute of Ammonia Refrigeration) guidelines.↩︎

35. Compressor fabrication tolerances and materials: Cylinder bore tolerances of 0.02–0.05 mm are based on standard reciprocating compressor engineering practice for ammonia service. Service life estimates for locally fabricated vs. factory-built compressors are rough projections based on the likely lower precision of field-machined components and less controlled material quality compared to purpose-built compressor factories. The 4–10 units per year production estimate assumes a well-equipped machine shop with lathe, milling machine, and honing capability, dedicated to compressor production. Source: IIAR (International Institute of Ammonia Refrigeration) guidelines; general reciprocating compressor engineering references.↩︎