A Rocket Exploded. We Need to do Math

Source: A Rocket Exploded. We Need to do Math., Hank Green, 23:28, uploaded 2026-06-01, playlist index 28.

On the night of 28 May, a Blue Origin New Glenn rocket exploded during a hot-fire test at Cape Canaveral. The engines were fired whilst the vehicle stayed bolted to the pad, so the test could check the rocket, its fuel systems, and the launch site before a flight. The vehicle’s first stage uses seven BE-4 engines burning methane and liquid oxygen. Its upper stage uses hydrogen and oxygen.

New Glenn is Blue Origin’s heavy-lift, partly reusable orbital rocket and a major part of the company’s attempt to enter the commercial launch market. Blue Origin also has contracts connected to NASA’s Artemis programme, future lunar landers, and cargo delivery. The failure therefore reaches beyond one destroyed rocket. It can delay other missions, damage infrastructure, and leave engineers with a long investigation before anyone knows what went wrong.

The footage is hard to ignore. The engines begin to fire, the rocket disappears into a fireball, and a large cloud rises above the pad. Blue Origin calls the event an anomaly. Green accepts the word with dry disbelief while keeping the cause open. The satellites meant for the next mission were elsewhere, so they survived. The rocket and much of the launch pad did not.

The comparison that starts the calculation

Social media supplied a ready-made environmental verdict. Green reads a Threads post from someone who had rinsed a yoghurt pot for recycling the day before, then compared Jeff Bezos’s failed rocket launch with the annual CO₂ output of several thousand diesel cars. The post had about 22,000 likes. Green sees the appeal and follows the claim into the arithmetic.

Space flight deserves environmental scrutiny. Carbon emissions affect the climate, and satellites that re-enter the atmosphere add another set of questions around the ozone layer. Green points to Anton Petrov’s video on satellites and ozone as a related discussion. The rocket calculation still needs a boundary. The fuel would have burned during a successful test or launch, although the failure burned it in one uncontrolled event. Green chooses to count the emissions from the explosion and focuses on the first-stage methane.

That choice leaves some uncertainty. The hydrogen in the upper stage carries a carbon cost if it comes from methane rather than water electrolysis, although that upstream carbon was not part of the explosion itself. Green leaves it outside the calculation because the question concerns the fireball. The result is an estimate of the event’s direct impact, not a full life-cycle account of the rocket.

Methane, carbon dioxide, and the missing fuel

Green says that New Glenn’s methane tank holds almost 25,000 cubic feet of liquid methane. He converts that volume to roughly 300 tonnes of methane. The conversion from methane to carbon dioxide depends on molecular mass. Methane is CH₄, with a molecular mass of 16. Carbon dioxide is CO₂, with a molecular mass of 44. When methane burns, its carbon combines with oxygen from the air, so the resulting carbon dioxide weighs more than the methane that supplied its carbon.

The calculation is:

Green compares that with a typical car’s reported annual emissions of about 4.6 tonnes of CO₂. If all of the methane burned, the explosion would therefore equal roughly 180 car-years of emissions released at once. The number is large enough to deserve concern. It also falls far short of several thousand diesel cars.

The missing variable is methane that escaped without burning. Methane acts as a stronger greenhouse gas than CO₂ over the time periods Green uses. He gives a factor of about 28 over 100 years and about 80 over 20 years, since methane breaks down in the atmosphere and its relative effect changes with the time horizon. If all 300 tonnes escaped unburned, the result would be about 8,400 tonnes of CO₂ equivalent on the 100-year measure and 24,000 tonnes on the 20-year measure. That equals roughly 1,800 or 5,000 car-years when divided by 4.6 tonnes per car per year.

The fireball shows that much of the methane burned. It cannot show how much. Green says that a thousand cars’ worth of CO₂ equivalent seems like a reasonable upper estimate, although he expects it to overstate the result. The source therefore supports a range shaped by combustion, leakage, and the chosen time scale. It cannot support a single precise comparison with several thousand cars.

The explosion also released material that the CO₂ arithmetic leaves out. Soot, particulates, nitrogen oxides, and combustion products from the rocket and pad can affect local air quality. Soot high in the atmosphere has climate effects of its own. Green declines to build those pollutants into the headline calculation because the available footage cannot establish their quantities.

He then compares the event with air travel. If most of the methane burned, the emissions sit in the range of a few complete transatlantic flights, meaning the aircraft and its passengers together. Airports produce comparable emissions constantly, whilst a rocket explosion is an unusual one-time event. The comparison changes the scale without making the failure harmless.

The part about Jeff Bezos

The arithmetic does not settle Green’s broader anger. He says the deeper frustration concerns Jeff Bezos and his view that long-term billionaire wealth can insulate a person from ordinary correction. Those remarks are Green’s personal judgement, and they carry a different kind of claim from the methane calculation. The environmental effect of one explosion can be estimated. A judgement about the person who owns the company requires a moral and political view.

Green’s son, who is nine, helps him separate these questions. The boy can estimate the explosion’s carbon impact against cars and aeroplanes. He struggles with the recycling comparison because a single person’s contribution feels too small to register. Green treats that failure of intuition as the reason to keep doing the maths.

Plastic, aluminium, and what recycling can do

Green says plastic recycling has limited value in its present form. Plastic often contains several materials that are difficult to separate. A bottle can combine different plastics in the body, cap, label, and seal. Contamination makes collection harder, and the material can degrade after repeated processing. Recycled plastic also costs more than new plastic in many cases.

He still wants people to recycle plastic because the material stream gives researchers and industry something to improve. He draws a sharper line around aluminium. Aluminium remains valuable, can be recycled repeatedly without the same degradation, and takes far less energy to recycle than to make from bauxite. Green says recycling an aluminium can uses about five percent of the energy required to make a new one.

His figure for one recycled can is about 98.7 grams of CO₂ equivalent avoided. Against the 825-tonne estimate for the rocket, the comparison looks absurdly uneven:

One person will not recycle 8.4 million cans. That fact can feed the social-media conclusion that individual action is pointless. Green turns the same number around by asking how many cans the United States uses. He gives roughly 100 billion cans in one year and about 46 billion recycled in the previous year. At 98.7 grams per can, those recycled cans avoid about 4.5 million tonnes of CO₂ equivalent. Divided by 825 tonnes, the total equals about 5,500 rocket explosions each year, or roughly 15 every day.

The discarded cans matter at the same scale. Green says the United States threw away around 61 billion aluminium cans in 2023. If those cans had been recycled, they would represent about six million tonnes of avoided CO₂ equivalent, which he converts to roughly 7,300 rocket explosions. Full recycling of the country’s cans would avoid around 12,800 explosions’ worth of emissions each year, in his calculation.

The single can remains insignificant on its own. The system of cans, households, collection schemes, and industrial processing has a large effect because it contains billions of repetitions. Green’s point depends on keeping both facts in view. Individual effort has a small marginal effect. The practice becomes substantial when a large population follows the same rule through a deposit law, a city programme, a public campaign, or a piece of media.

The scale of primary aluminium

Green widens the comparison again. He says the global aluminium industry produces about one billion tonnes of CO₂ each year when the electricity used to make aluminium is included. A year contains 8,760 hours, so the industry emits about 114,000 tonnes per hour. The 825-tonne rocket estimate is therefore:

The explosion represents about 0.7 percent of one hour of global aluminium production. That comparison keeps the New Glenn failure visible while placing it inside a much larger continuous process. The event is spectacular and wasteful. Primary aluminium production creates more than a hundred similar quantities of emissions every hour.

The source’s description adds that New Glenn itself is made from aluminium, perhaps the equivalent of five to ten million cans, although it says the exact amount is hard to know. Green wonders whether the wreckage will be recycled after the investigation, as happens with the material from some aeroplane investigations. The description supplies the observation and the question without a documented mass calculation.

Care under conditions of bad scale

The modern information system keeps presenting isolated disasters beside private acts of care. A rocket becomes a single visible event with a time, a fireball, and a replayable image. A recycled can disappears into a stream of repeated actions. The first is easy to feel. The second requires a person to imagine a number that exceeds ordinary experience.

Green says the internet can turn that mismatch into nihilism. The original post may have meant to express disgust with Bezos, while its numerical comparison can leave readers asking what use recycling has. The objection then lands on the wrong target. The rocket’s emissions deserve scrutiny, and the aluminium industry’s emissions deserve much more scrutiny. The existence of those problems does not erase the measurable effect of a practice carried out by millions of people.

Green’s conclusion stays modest. Care can become so diffuse and demanding that people lose contact with the actions they can share. Information should help people direct care towards the parts of a system where it has an effect. The source’s case for recycling comes from scale, with a material difference between one person’s can and billions of cans processed through the same system.

Limits of the calculation

The video gives the methane volume, molecular conversion, car comparison, aluminium savings, US can counts, and global industry figures as Green’s working numbers. It provides no underlying studies or links for those figures. The note keeps them attached to his account rather than presenting them as independently verified results.

The rocket estimate also depends on how much methane burned, how much escaped, which methane time horizon is used, and which emissions are counted. The direct CO₂ calculation leaves out upstream fuel production, the upper stage’s hydrogen supply, damage to the pad, pollutants beyond CO₂, and the later fate of the aluminium. The comparisons with cars, flights, and aluminium production make scale legible. They do not form a full environmental assessment of New Glenn or of space flight.

Further reading / references

• NASASpaceFlight’s Space Coast livestream, the source of some of the launch footage.
• NASASpaceFlight’s original response and footage compilation, both linked in the video’s description.
• SpaceFlightNow on YouTube, another footage source named in the description.
• Anton Petrov on satellites and ozone, the related video Green mentions when introducing the environmental effects of space flight.