To evaluate the effectiveness of the cooling system, a constant engine speed of 3750 revolutions per minute (RPM) was maintained while the vehicle remained in neutral gear, to heat the engine and ensure additional load would require more cooling than could be provided by the unassisted radiator. Coolant temperature was monitored by the vehicle’s internal coolant temperature gauge. Temperatures on the gauge were unmarked, so the 1/4–3/4 span was assumed to be between 90°C, and 105°C, as is standard for similar vehicles [53,54].

A Hanmey WCX5 PTO woodchipper, rated for chipping wood up to 5 inches (125 mm) in diameter, was connected by the PTO shaft to the driving wheel’s hub, as shown in Fig. 4. The moment-restricting frame was bolted to the vehicle’s towbar and mounted to the woodchipper by the standard pins.

The woodchipper shaft’s stated operating speed was 430 RPM, corresponding to a vehicle engine speed of approximately 2000 RPM in first gear. This operating RPM was maintained by a driver, who adjusted the throttle according to load. Before chipping, wood volume and mass were measured. During operation, the woodchipper was spun to 430 RPM, then the wood was fed in with the thickest end first. The following wood types were tested: (i) wet wood, diameter 1–3.5 inches (25–90 mm); (ii) wet wood, diameter 4 inches (100 mm); (iii) dry wood, diameter 5 inches (125 mm). The volume of wood, chipping time, instantaneous fuel consumption, and engine rotational velocity were recorded during operation.

An Agriquip GP33 PTO generator with a 30 kVA synchronous alternator was connected by the PTO shaft to the driving wheel’s hub, using the same mechanism shown in Fig. 4. The generator was equipped with an automatic voltage regulator (AVR) to maintain constant output voltage. The generator was connected by a New Zealand-standard 60-Amp three-phase plug to a load bank with variable resistance. To generate three-phase 400 V 50 Hz electricity, the generator required a rotational velocity of 1500 RPM, corresponding to 430 RPM at the PTO, and an engine speed of ~2000 RPM with the vehicle in first gear. During operation, the electrical load of the load bank was increased in increments of 2.5 kW, to a maximum of 20 kW. Power, current, frequency, and voltage were measured at the load bank and generator, and engine rotational velocity and instantaneous fuel consumption were measured at the vehicle.

All three wood types were chipped successfully. Although the test for dry wood (type iii) was interrupted by a knot in the wood causing the woodchipper to jam, the vehicle remained running, and chipping was restarted, allowing successful completion of the test. The dry wood produced 0.00152 m³ of wood chips, with a total mass of 700 g, and required 0.0299 L of petrol. Assuming a petrol energy content of 32 MJ/L [55] and wood chip energy content of 10 MJ/kg [56], the wood chipping process used 0.96 MJ of energy from petrol to produce 7 MJ of energy from wood chips, of which 3.5 MJ (conservatively) would likely be used to produce syngas [57,58]. Thus, the modified vehicle’s energy returns on energy investment (EROI, the amount of usable energy output for a given amount of energy input [55]) is 3.7. An example of engine speed and fuel consumption during the chipping of dry wood, before restarting, is shown in Fig. 6.

Electricity generation of up to 20 kW was sustained, with voltage fluctuations within the limits of the New Zealand Electricity Regulations [59]. Minimal vibrations were observed in the vehicle, and the electrical power output variation was under 1 kW. Electrical frequency varied when the load changed, as the vehicle RPM was managed by manual control of the engine throttle, which typically overcompensated for load increases. Engine speed and power at the load and generator during electricity generation are shown in Fig. 7. Engine speed and power frequency at the load and generator are shown in Fig. 8.

The radiator coolant temperature reached 105°C after 10 min of engine operation at 3750 RPM in neutral gear. Auxiliary cooling for two minutes reduced coolant temperature to 90°C.

With light-duty vehicles generating 20 kW of electricity for an average of 120 h per week (accounting for 30% downtime [71]), equivalent to 124,800 kWh per year, 9.6 million vehicles would be required to produce electricity equivalent to the demand of the water sanitation sector. With an EROI of 3.7 from wood chipping for gasification, 3.5 million vehicles would be required to chip wood for gasification to power electricity generation, for a total requirement of 13.2 million vehicles. Older vehicles are expected to be better suited to modification than newer vehicles, given their limited internal electronics and thus lower likelihood of damage in an HEMP event, but newer vehicles are still typically unplugged and would likely survive. Approximately 830 million light-duty vehicles were produced between 1990–2005 [72], of which approximately 5% are operational [73], leaving roughly 41 million suitable vehicles expected to be operational. Thus, the supply of light-duty vehicles is expected to exceed requirements.

When not connected to an electrical grid to induce a magnetic field in their rotor, induction motors require self-exciting capacitors, which use the reactive power of the spinning motor [74]. The annual global capacitor market is valued at USD 25 billion [75], with aluminum electrolytic capacitors accounting for 22% (USD 5.5 billion) [75]. Of these, approximately 6% (34 million capacitors per year, assuming an average price of 10 USD [75]) are within the required 50–500 μF range and rated about 400 V [75]. Assuming an average lifespan of 10 years, 330 million suitable capacitors are expected to be currently operational. To fulfill the capacitor requirement for electricity generation, 9.6 million would be required, indicating even a very low survival rate of capacitors to HEMP (anything greater than 3%) would be sufficient.

The PTO used in this work included an integrated gearbox with a 3.49:1 ratio, to meet the required 1500 RPM of the generator. Most continuously variable transmissions (CVTs) could meet this gearing requirement and be used instead in a catastrophe [76]. With a market size of 22.1 billion USD [77] and an average price of around 5000–10,000 USD [78], approximately 2.2–4.4 million CVTs are produced annually. Assuming each lasts for 15 years, approximately 33–66 million would be operational at the time of a GCIL event. Some gearboxes with the appropriate gear ratio could be manufactured or salvaged from other applications, so this supply is likely to be sufficient for the 13.1 million required.

Recreational vehicles and residential boats both have water pumps of sufficient sizes for auxiliary cooling systems. With an annual market of 55 billion USD [79] and an average cost of 50,000 USD [80], approximately 1.1 million recreational vehicles are produced annually. Assuming all have usable pumps and an average lifetime of 15 years, 16.5 million pumps could be sourced from recreational vehicles. Residential boats have an annual market of 34 billion USD [81] and average cost of 100,000 USD [82]. Assuming an average lifetime of 15 years and 75% with sufficiently large pumps, 3.8 million pumps could be sourced from residential boats. Neither recreational vehicles nor residential boats are typically connected to electricity grids. Thus, both are expected to survive a HEMP or electrical storm, so could provide pumps for auxiliary cooling systems. Combined, recreational vehicles and residential boats are expected to provide up to 20.3 million pumps, more than the 9.6 million required to produce 1200 TWh per year. However, in a catastrophe affecting the availability of transport, boat pumps may only be available near coastal areas where the boats are located. As only 14% of the global population lives within 10 km of coastal areas [83], transportation limitations could necessitate alternative solutions for cooling system pumps. Additionally, future work could consider developing open-source water pump designs for such emergencies, using distributed manufacturing methods, such as those used in the COVID-19 pandemic [84].

Other components for auxiliary cooling systems are expected to be sourced without difficulty in a catastrophe. A single nozzle supplier can produce up to 2.5 million nozzles per year [85]; assuming an average lifetime of 10 years, demand could thus be met even by the 25 million nozzles available from a single supplier. Other components, such as tubing and reservoirs, could be adapted from common polymer components and are expected to be sufficiently ubiquitous in a GCIL scenario.

While vehicles with larger engines have higher-capacity cooling systems and are unlikely to require auxiliary cooling systems to produce 20 kW of electrical power, these vehicles could be used to generate higher loads and could thus still require auxiliary cooling. Alternatively, the production of electrical power from each vehicle could be scaled according to their cooling capacity, removing the need for auxiliary cooling systems altogether. Furthermore, salvaged auxiliary radiators could obviate water cooling.

Synchronous generators, particularly those with AVRs, can provide stable electricity generation with high efficiency [86]. While large-scale commercial synchronous generators (2–50 MW) have an annual market of ~6 billion USD [87], data on smaller systems (1 kW–2 MW) are not readily available. Diesel gensets pair a synchronous motor with AVR to a diesel engine, and the market for small (<100 kVA) gensets is approximately 7 billion USD [88]. Assuming an average cost, power, and power factor of 10,000 USD, 50 kVA, and 0.8 [89], respectively, approximately 24 GW of genset capacity is produced annually. Assuming an average lifetime of 10–15 years and a 30% survival rate from an EMP, 72–108 GW of installed capacity is expected to be operational, which is less than the required 192 GW.

Induction motors, while providing less stable electricity generation than synchronous generators, are widely available and have long lifetimes [86]. Induction motors have an annual market of 21 billion USD [90], and more than half are estimated to be of appropriate sizes (in the range of 1–75 kW) [91]. With an average price of approximately 1000 USD [92], an average lifetime of around 15 years, 158 million suitably sized induction motors. Assuming an average size of 3 kW and an HEMP survival rate of 30%, 142 GW of induction motor capacity is expected to be operational.

Hybrid electric vehicles (EVs) contain integrated motor generators (IMGs), which generate electricity to store in the vehicle’s battery or power the drivetrain [93]. If the availability of gensets and induction motors is insufficient, these IMGs could be used in a catastrophe for electricity supply. The total number of EVs (including hybrids) is estimated to be 40 million, of which hybrids constitute approximately 30% [94], so approximately 12 million hybrid EVs are expected to be operational. The average electrical output of an IMG is between 20–90 kW [93], so the total generation capacity from hybrid EVs is expected to be approximately 240–1080 GW. Assuming a HEMP survival rate of 30% to account for some hybrid EVs being plugged in, 72–324 GW of generation capacity is expected to be available from hybrid EVs. Additionally, the number of hybrid EVs is increasing globally [95], so the total generation capacity from IMGs is likely to increase.

Combined available generation capacity from gensets, induction motors, and hybrid EVs is expected to be 285–573 GW, which exceeds the 192 GW demand for water sanitation. Alternatively, future work could assess the feasibility of directly powering hybrid EVs with wood gas to directly generate electricity from their IMGs, with no modifications.

The annual woodchipper market size is approximately 430 million USD [96], and average costs are approximately 2000 USD per unit [97]. Assuming an average lifetime of 15 years, approximately 3.2 million woodchippers are expected to be available in a catastrophe, which is just under the required 3.5 million. Furthermore, the size of the chips is smaller than the chunks ideal for gasifiers.

With 120 h of engine operation per week and an average wheel speed equivalent to 30 km/h, each vehicle would cover an annual equivalent distance of 187,200 km. With oil changes for light duty vehicles (~6.5 L per vehicle) recommended every 7000–10,000 km [98], each vehicle would require 18–26 oil changes per year. With 33 million vehicles required, total annual requirements are expected to be 4.0–5.7 billion L. While other consumables, such as transmission fluid, would also be required, these consumables typically have much longer replacement intervals than engine oil, so are not expected to limit scalability while vehicles use the consumables already in their systems. However, while the stockpiles of these consumables are unknown, they could likely be salvaged from non-operational vehicles post-catastrophe.

Annual engine oil consumption in the USA is approximately 2.1 billion L [99]. Assuming engine oil consumption is proportional to contribution to gross world product (GWP), of which the USA contributes 25% [100], global annual engine oil consumption is estimated to be 8.4 billion L. Assuming one month’s worth of consumption is available at a given time in warehouses and in transit, approximately 700 million L of engine oil would be available at the time of a GCIL event, equivalent to 3–5 months of required consumption. Although the availability of new engine oil would be low after production ceased in a GCIL event, engine oil could likely be salvaged from other vehicles. The total number of light-duty vehicles in the world is approximately 1.4 billion [101]. Assuming engine oil tanks are an average of 40% full, approximately 3.8 billion L of engine oil could be salvaged from existing vehicles, equivalent to 20–28 months of required consumption. Together with oil in the supply chain, the total available engine oil is expected to be between 23–33 months of required consumption. Additionally, some research has shown oil change intervals can be extended up to 2–3 times those typically recommended by vehicle manufacturers, without incurring mechanical damage or impairing engine operation [102,103]. Thus, available engine oil could likely be extended up to 47–100 months of required consumption.

Decentralized wood gasification could mitigate the effects of electricity loss on the critical water sanitation sector following a GCIL event. Supplies of vehicles, capacitors, gearboxes, and electrical generators could be sufficient to provide equivalent electricity to the demand for water sanitation facilities. While pump supply could be a limiting factor for auxiliary cooling systems, the required number of these could be reduced by pairing external machinery loads with vehicles with sufficient cooling capacity.

The supply of woodchippers could be a limiting factor for scalability. Production of woodchippers could potentially be increased following a GCIL event, as many factories are expected to be capable of producing woodchippers, which are simple machines, consisting essentially of a single cutting wheel in a housing to direct logs towards, and wood chips away from, the wheel. Fabrication capacity, however, may be limited following a catastrophe and could be constrained by other requirements, so further research is required to assess the scalability of woodchipper production in a GCIL scenario. Although plans have also been developed for bicycle-based wood chipping [104], this method is best suited to chipping small-diameter wood, so the supply of wood would be limited. Additional work is needed to investigate alternatives, such as the industrial woodchippers used for paper production (which also produce chips, not chunks), wood splitters, chainsaws, band saws, table saws, circular saws, and splitting by hand.

Engine oil could also be a limiting factor for scalability. Current engine oil stores are expected to last 3–5 months of continuous wood chipping and electricity generation but could be extended up to 47–100 months with longer oil change intervals and/or salvaged engine oil, which may be sufficient to allow repair of power plants, although the time required for this work is unknown. Alternative methods of supplying conventional engine lubricants, or alternative lubricants, such as bio-lubricants [105,106], could potentially replace conventional engine oil if production was increased. For example, an open-source controlled pyrolysis system can convert waste plastics to oil-based lubricants [107,108], so could potentially be produced quickly in a catastrophe. While the production of bio-lubricants could be decentralized, the potential for global production to increase is unknown and could be the subject of future research.

Current supplies of wood gasifiers are negligible, so the scalability of wood gasification to produce electricity for critical needs would depend on post-GCIL production. Wood gasifiers can be built with simple tools and supplies from open-source plans [109,110] and thus could potentially be produced quickly in a catastrophe, particularly with simple pre-planning [49]. Thus, the supply of wood gasifiers may not be a limiting factor for scalability.