Could a petrol engine run on Mars, and what would it take?

Mars has air. It is just 1% as dense as ours and almost none of it is oxygen. Work it through and a 200 cc bike engine breathing Martian air makes about one watt.

Rocky reddish ground of Jezero crater on Mars photographed by the Perseverance rover on its fourth day
Photo: NASA/JPL-Caltech/ASU, Public domain, via Wikimedia Commons

A petrol engine cannot run on Martian air. The atmosphere is about 1.3% as dense as Earth's at sea level and only 0.16% oxygen by volume, so a 200 cc bike engine that makes about 14 kW on Earth would make roughly 1 watt breathing Mars air. It can run on Mars only if it carries its own oxygen, about 3.4 kg for every kg of petrol, or switches to a fuel such as magnesium that burns in carbon dioxide. That is why NASA's Mars rovers run on a plutonium generator and solar power, not combustion.

Calculated in this post from the NASA Mars and Earth fact sheets (air density and oxygen share) and the engine assumptions in the table below.

Key takeaways
  • 0.016 kg/m³ is Mars surface air density (NASA Mars fact sheet), about 1.3% of Earth's sea-level 1.217 kg/m³ (NASA Earth fact sheet).
  • 0.16% oxygen by volume: each litre of Martian air holds about 15,000 times less oxygen than a litre of ours.
  • About 1 W from a 200 cc single that makes 14 kW on Earth, if it breathes Mars air (calculated below from the NASA Mars fact sheet).
  • 3.4 kg of oxygen per kg of petrol must be carried, or made by a MOXIE-style machine.
  • 16.7 MJ per kg of magnesium burned in Martian CO2, the only fuel that uses the planet's own air.

This is a tier-2 what-if, so it gets the full treatment: assumptions, the combustion maths, a sensitivity check and a verdict. It starts somewhere much closer to home, at Gurudongmar in North Sikkim.

The story

On 3 April 1933, two Westland biplanes took off from Lalbalu airfield near Purnea, in Bihar, to become the first aircraft to fly over Mount Everest. Each carried a highly supercharged Bristol Pegasus engine, built to force enough thin air into its cylinders to keep burning fuel high up, plus oxygen and heating for the crews. Lord Clydesdale flew one with Stewart Blacker as observer; David McIntyre flew the other with the cinematographer Sidney Bonnett.

The engines held. The people nearly did not. Bonnett damaged his oxygen line, passed out briefly from hypoxia and patched it with a handkerchief; McIntyre's mask strap broke and he flew holding the mask to his face. At about 10:05 am both planes crossed the summit with roughly 100 ft to spare. Mars asks the same question of an engine, with almost no oxygen left to compress.

Wikipedia: Houston–Mount Everest flight expedition

What are the assumptions?

One engine, three places: sea level, Gurudongmar and the Martian surface. Values marked as estimates are mine.

SymbolMeaningValue used
VdDisplacement, a 200 cc single200 cm³
NEngine speed9,000 rpm (estimate)
ηvVolumetric efficiency0.9 (estimate)
ρEEarth surface air density1.217 kg/m³
ρMMars surface air density0.016 kg/m³ (±0.006)
YO2Oxygen mass fraction: Earth / Mars0.231 / 0.00118
LHVPetrol energyabout 43 MJ/kg
ηEngine efficiency30% (estimate)

The Mars oxygen fraction comes from NASA's composition: 0.16% O2 by volume, times 32 (oxygen's molar mass) over 43.49 (Mars air's mean molar mass), gives 0.118% by mass. The petrol figure of 43 MJ/kg is a round working value for pump petrol and sources differ: Wikipedia's Gasoline article gives 46.7 MJ/kg, and the US DOE lists 112,114 to 116,090 Btu per gallon. The results scale in proportion to it.

How much oxygen does a petrol engine need?

About 3.4 kg of oxygen for every kg of petrol. Petrol burns completely at an air-fuel ratio of about 14.7 to 1 by mass, and 23.1% of Earth's air by mass is oxygen.

\[ \frac{m_{O_2}}{m_{fuel}} = 14.7 \times 0.231 = 3.40 \]

An engine is really an oxygen pump. Its power is set by how much oxygen it can swallow per second, and the fuel just follows. That gives one compact equation for shaft power:

\[ P = \eta \cdot LHV \cdot \frac{\rho \, Y_{O_2} \, \eta_v \, V_d \, N/120}{3.40} \]

The N/120 turns rpm into intake strokes per second for a four-stroke, which fills its cylinder once every two turns.

What does that give on Earth, in Sikkim and on Mars?

  1. Air swallowed. \( \eta_v V_d N/120 = 0.9 \times 0.0002 \times 75 = 0.0135 \) m³/s at every location (calculated here from the estimates in the assumptions table above).
  2. Earth, sea level. Oxygen: \( 0.0135 \times 1.217 \times 0.231 = 3.80 \) g/s. Fuel: 1.12 g/s. Shaft power: \( 0.30 \times 43 \times 10^6 \times 0.00112 = 14.4 \) kW, about 19.6 PS. Air density is from the NASA Earth fact sheet.
  3. Gurudongmar, about 5,300 m. Terrain data puts the lake at about 5,150 m and the Sikkim government at 5,430 m, so I use the midpoint. NASA Glenn's standard-atmosphere fit gives density about 58% of sea level, so roughly 8.4 kW, if the fuel is metered correctly. A carburettor does not do that, so real output drops further.
  4. Mars. Oxygen: \( 0.0135 \times 0.016 \times 0.00118 = 2.5 \times 10^{-7} \) kg/s. Fuel: \( 7.5 \times 10^{-8} \) kg/s. Shaft power: about 1 W. Density and oxygen share are from the NASA Mars fact sheet.

On Martian air, a 200 cc engine at full throttle makes about as much power as a bicycle dynamo lamp. In practice it would not even do that, because a mixture this lean cannot be lit at all.

Power from a 200 cc single breathing the local air
See the data as a table
Shaft power (W)
Earth, sea level14400
Gurudongmar (5,150-5,430 m)8400
Mars surface1

Source: Calculated in this post from NASA Earth and Mars fact sheets and the NASA Glenn standard atmosphere · Assumes 9,000 rpm, 90% volumetric efficiency, 30% efficiency and correct fuelling. Mars density varies seasonally from about 0.010 to 0.022 kg/m³.

Gurudongmar Lake in North Sikkim, a high-altitude lake surrounded by barren brown mountains under a blue sky
Gurudongmar Lake, North Sikkim: about 5,150 m on terrain data, 5,430 m on the Sikkim government figure. At either height the air holds under 60% of sea-level oxygen per litre, enough to halve a carburetted bike's pace.Photo: Yoghya, CC BY-SA 4.0, via Wikimedia Commons

Could a turbocharger fix it?

No. It would need a pressure ratio of about 15,000. A turbo works by squeezing more air into each stroke. To match Earth, the intake would have to deliver as much oxygen per litre as Earth's air does.

\[ \frac{\rho_E Y_{O_2,E}}{\rho_M Y_{O_2,M}} = \frac{1.217 \times 0.231}{0.016 \times 0.00118} = \frac{0.281}{1.89\times10^{-5}} \approx 14{,}900 \]

Road-car turbochargers run pressure ratios of roughly 2 to 3 (my estimate). Fifteen thousand is not an engineering stretch; it is a different machine. Squeezing the gas that hard also heats it, and the compressor would eat far more power than the engine could give back, by the first law of thermodynamics. The fuel-metering side of this problem, even on Earth, is covered in fuel injection vs carburettor.

Oxygen in 1 m³ of intake airEarth, sea level: 281 gGurudongmar: about 163 gMars: 0.019 g (a sliver you cannot see at this scale)Mars air is 95.1% CO2; oxygen is just 0.16% by volumeRatio Earth to Mars: about 15,000 to 1
The whole problem in one picture. An engine's power follows the oxygen it can swallow, and Martian air has almost none.

What if the bike carries its own oxygen?

Then it runs, and on Mars it would not need much. A petrol engine fed from an oxygen tank is closed-cycle: it no longer cares what the outside air is. The question becomes how far a kilogram of propellant takes you, and Mars helps here because gravity is 3.73 m/s² and the thin air gives almost no drag.

Worked solution: range per kilogram of petrol

  1. Road load at 30 km/h. Rolling resistance with a loose-ground coefficient of 0.05 (my estimate): \( F = 0.05 \times 250 \times 3.73 = 46.6 \) N. Power: \( 46.6 \times 8.33 = 389 \) W.
  2. Drag. \( \tfrac12 \times 0.016 \times 0.5 \times 8.33^2 = 0.28 \) N, about 2 W. Negligible.
  3. Engine at light load. Say 20% efficient (my estimate). One kg of petrol gives \( 0.20 \times 43 = 8.6 \) MJ at the wheel.
  4. Range. \( 8.6 \times 10^6 / 389 = 22{,}100 \) s, about 6.1 hours, or about 184 km at 30 km/h (calculated here from the estimates in the steps above).
  5. Oxygen bill. 3.4 kg of oxygen per kg of petrol, so 4.4 kg of propellant for those 184 km (the ratio is derived above from the 14.7:1 air-fuel ratio).

The oxygen has to come from somewhere. Shipping it from Earth is expensive, which is exactly the problem NASA's MOXIE experiment was built to test.

Mars Oxygen ISRU Experiment (MOXIE): Preparing for human Mars explorationJeffrey Hoffman, Michael Hecht et al. · Science Advances 8(35), eabp8636 · 2022

MOXIE rides inside the Perseverance rover. It pulls in Martian air through a HEPA filter, compresses it with a scroll pump, heats it to 800 °C and splits CO2 in a solid oxide electrolysis stack, passing oxygen ions through a zirconia ceramic. It made oxygen seven times between landing in February 2021 and the end of 2021, at 6 to 8 g per hour, and the team targets over 99.6% purity.

A six-person ascent rocket would need about 31 tonnes of oxygen, so a scaled-up unit would have to make 2 to 3 kg per hour. The limit: it had not yet been run at dawn or dusk, when temperatures swing.

Read the paper (free)

Put the bike next to MOXIE. The 184 km tank above needs 3.4 kg of oxygen. At 6 to 8 g per hour, MOXIE would take 425 to 570 hours to make it. A scaled-up 2 to 3 kg-per-hour plant would do it in under two hours.

Calculated in this post from 3.4 kg of oxygen (above) and MOXIE's 6 to 8 g per hour and 2 to 3 kg per hour scale-up, both from Hoffman et al., 2022.

Engineers installing the gold-coloured MOXIE oxygen experiment box into the body of the Perseverance rover
MOXIE going into Perseverance at JPL. It made oxygen from Martian CO2 at 6 to 8 grams an hour.Photo: NASA, Public domain, via Wikimedia Commons

Is there a fuel that burns in Martian air?

Yes: magnesium. Magnesium burns in carbon dioxide, pulling the oxygen out of the CO2 itself, so the 95% of Mars air that is useless to petrol becomes the oxidiser.

\[ 2\,\text{Mg} + \text{CO}_2 \rightarrow 2\,\text{MgO} + \text{C} \]

The heat released follows from Hess's law: subtract the enthalpy of formation of what you start with from what you end with. NIST lists MgO at −601.6 kJ/mol and CO2 at −393.5 kJ/mol.

  1. Products. \( 2 \times (-601.6) = -1203.2 \) kJ.
  2. Reactants. \( -393.5 \) kJ (magnesium and carbon are zero by definition).
  3. Heat released. \( 1203.2 - 393.5 = 809.7 \) kJ per 2 mol of Mg, which is 48.6 g.
  4. Per kilogram. \( 809.7 / 0.0486 = 16{,}660 \) kJ, so 16.7 MJ per kg of magnesium. Calculated here from the NIST MgO and NIST CO2 values above.

Petrol holds 43 MJ/kg, but on Mars it drags 3.4 kg of oxygen along, so a kilogram of carried propellant gives only \(43/4.4 = 9.8\) MJ. Magnesium gives 16.7 MJ per kilogram carried, about 1.7 times better, because the oxidiser is free.

Calculated in this post from the NIST enthalpies above and the 43 MJ/kg petrol value discussed under the assumptions table.

Nickel-coated aluminum particles: a promising fuel for Mars missionsEvgeny Shafirovich and Arvind Varma, Purdue University · NASA/TM-2004-213114 · 2004

A short research summary on burning metals in CO2 for Mars propulsion. It notes that magnesium is the leading candidate because it ignites easily and burns fast in CO2, and cites calculations that a 200 kg ballistic hopper with CO2/metal rockets and a CO2 collector could make 10 to 15 flights covering 10 to 15 km in total. The authors then test nickel-coated aluminium particles, using laser ignition and levitation, to lower aluminium's ignition temperature. The limit: it is a brief abstract of work in progress, not a flight-tested engine.

Read the paper (free)
Option on MarsUses local air?Energy per kg carriedMain problem
Petrol, breathing Mars airYesAbout 1 W of outputAlmost no oxygen
Petrol plus carried oxygenNo9.8 MJCryogenic oxygen tanks
Magnesium burning CO2Yes16.7 MJPowder handling, MgO ash
Battery, solar chargedNoUnder 1 MJ (my estimate)Dust on panels, cold
Plutonium generator (MMRTG)NoAbout 110 W, for yearsCost, regulation, low power

How would you cool it in thin air?

Air cooling would barely work. Forced-convection heat transfer in turbulent flow scales roughly with \( (\rho v)^{0.8} \), the familiar Reynolds-number scaling from heat transfer textbooks.

\[ \frac{h_{Mars}}{h_{Earth}} \approx \left( \frac{0.016}{1.217} \right)^{0.8} = 0.031 \]

So fins would shed about 3% of the heat they do on Earth, an estimate that ignores CO2's different conductivity. The bike would need a liquid loop and a radiator, part convective and part glowing infrared, as worked through in how to build a motorcycle that runs in space. The −59 °C average helps a little, but thin air is a poor coolant whatever its temperature.

Why did NASA's rovers go nuclear and solar?

Because they need steady low power for years, not bursts of high power for hours. Combustion is the wrong tool for that job.

  • Perseverance (2021): 1,025 kg, powered by a multi-mission radioisotope thermoelectric generator holding 4.8 kg of plutonium dioxide, giving about 110 W at launch and falling a few percent a year. Two lithium-ion batteries cover peaks. Top speed: 152 metres per hour.
  • Ingenuity (2021–2024): a 1.8 kg solar-charged helicopter with rotors 1.2 m across, flying in air less than 1% as dense as Earth's. It made 72 flights and covered about 17 km.
  • The trade: 110 W from plutonium sounds tiny, but it never needs refuelling, works at night and in dust storms, and its waste heat keeps the electronics warm.
Technicians handling the finned black cylinder of a multi-mission radioisotope thermoelectric generator at Kennedy Space Center
A multi-mission radioisotope thermoelectric generator, the type that powers Perseverance. About 110 W, with no air and no fuel stops.Photo: Kim Shiflett, Public domain, via Wikimedia Commons

Try it on your own engine

Use this to check what you just read. Put in your bike's displacement and see what it would make on Earth (density 1.217, oxygen 23.1%) or Mars (0.016 and 0.118%).

Engine power from local air

Liquid O23.4 kg per kg fuelPetrol43 MJ/kgEnginesealed intakeExhaustCO2 + H2ORadiatorWheel: about 390 W at 30 km/hOutside: 0.6 kPa of CO2, unused
The only petrol engine that works on Mars ignores the Martian air completely and breathes from its own oxygen tank.

So, could it be done?

  • Breathing Mars air: about 1 W from 200 cc. Impossible in any useful sense.
  • Turbocharged: needs a pressure ratio near 15,000. Impossible.
  • Carrying oxygen: works, about 184 km per kg of petrol at 30 km/h, with 3.4 kg of oxygen per kg (the worked solution above rests on my estimates).
  • Burning magnesium in CO2: works in principle, 16.7 MJ per kg carried, and NASA-funded research has studied it for hoppers.
Verdict: possible with changes

A petrol engine cannot run on Martian air, which holds about 15,000 times less oxygen per litre than ours, so a 200 cc engine would make about 1 W. It can run on Mars as a closed-cycle engine fed from its own tank of liquid oxygen, at 3.4 kg of oxygen per kg of petrol, for roughly 184 km per kg of fuel at 30 km/h. The better Martian combustion engine burns magnesium in the planet's own CO2, releasing 16.7 MJ per kg.

Calculated in this post from the NASA Mars and Earth fact sheets, the NIST enthalpies and the estimates in the assumptions table.

For anything that must run for years, NASA's answer stands: a 110 W plutonium generator or solar panels.

The high-altitude version of this problem is the one Indian riders actually meet, and the engine basics behind it are in how engine power and torque work. The other ways an engine dies when starved are in riding without engine oil and holding a bike at the redline for an hour. More machines live on Yantra.

What if next?

Questions people ask

Can fire burn on Mars?

Ordinary fuels cannot burn in Martian air, which is 95.1% CO2 and only 0.16% oxygen at about 0.6 kPa. Reactive metals such as magnesium can burn in CO2 itself, which is why they have been studied for Mars engines.

Why does a bike lose power at high altitude?

Each litre of air holds less oxygen, so the engine burns less fuel per stroke. At about 5,300 m air density is roughly 58% of sea level. Fuel injection adjusts the fuel to match; a carburettor runs rich, so power falls even more.

How are Mars rovers powered?

Perseverance and Curiosity use a radioisotope generator of about 110 W, fed by decaying plutonium dioxide, plus batteries for peaks. Earlier rovers and the Ingenuity helicopter used solar panels.

Could you make oxygen for an engine on Mars?

Yes. NASA's MOXIE split CO2 into oxygen at 6 to 8 g per hour. A scaled-up plant making 2 to 3 kg per hour could fill an oxygen tank for a bike in under two hours.

Sources

  1. NASA NSSDC: Mars fact sheet (source for 0.016 kg/m³, 6.36 mb, 95.1% CO2, 0.16% O2, molar mass 43.49, 3.73 m/s², −59 °C)
  2. NASA NSSDC: Earth fact sheet (source for 1.217 kg/m³ surface air density)
  3. NASA Glenn Research Center: Earth atmosphere model (source for the troposphere fit used to get about 58% density at 5,300 m)
  4. Hoffman, J. A., Hecht, M. H. et al.: Mars Oxygen ISRU Experiment (MOXIE), Science Advances 8(35) (2022) (source for 6–8 g/h, 800 °C, seven runs, 31 t of oxygen, 2–3 kg/h scale-up)
  5. Shafirovich, E. and Varma, A.: Nickel-coated aluminum particles, a promising fuel for Mars missions, NASA/TM-2004-213114 (2004) (source for the 200 kg hopper, 10–15 flights, magnesium as candidate fuel)
  6. NIST Chemistry WebBook: Magnesium oxide (source for −601.6 kJ/mol)
  7. NIST Chemistry WebBook: Carbon dioxide (source for −393.5 kJ/mol)
  8. US DOE Alternative Fuels Data Center: Fuel properties comparison (source for petrol energy content)
  9. Wikipedia: Air–fuel ratio (source for 14.7:1)
  10. NASA Science: Perseverance rover components (source for 110 W, 4.8 kg PuO2, 1,025 kg, 152 m/h)
  11. NASA Science: Ingenuity Mars Helicopter (source for 1.8 kg, 1.2 m rotors, 72 flights, 17 km)
  12. Wikipedia: Gurudongmar Lake (source for the 5,150–5,430 m elevation range)
  13. Wikipedia: Houston–Mount Everest flight expedition (source for the 1933 Everest flight)
  14. Wikipedia: Westland Wallace (source for the supercharged Bristol Pegasus engines of the PV-3 and PV-6)
Marscombustionair-fuel ratioMOXIEmagnesium CO2 engineMMRTGhigh altitudewhat if
Sapta

Sapta (Saptarshi Sarkar) is a mechanical engineer turned marketer in Bengaluru and a solo touring rider. He writes the seven Saptaverse blogs. Work with him or read more of RoadHook Machines.

Keep reading in RoadHook Machines

From the other blogs