Clean Energy: A Truth Too Painful to Dissect
title: "Clean Energy: A Truth Too Painful to Dissect" date: "2026-09-20" author: "Zhigeng" channel: "ecology" excerpt: "The universe's temperature spans from near absolute zero to billions of degrees, yet the window habitable to humans is only a dozen degrees wide—hypothermia at a 35°C core, heat stroke at a 35°C wet-bulb, two life-and-death lines bracketing our entire living space. So-called 'clean energy' can only wash away 'carbon': wind, water, tides, and geothermal heat are the very engines of the ecosystem; coal, oil, and gas are irreplaceable material feedstocks; hydrogen carries two hidden debts of leakage and escape; and both nuclear and fusion keep feeding the 'big quilt.' Heat is the final pollution. What we lack was never cleaner energy, but less energy." tags: [] readTime: 24
In the essay Global Warming: A Truth Too Painful to Dissect, published on this account on September 13, I promised several follow-up pieces. Today I make good on "Is clean energy really clean?"
I. What Is Happening: A Fragile "Leaf Puddle"
Let me begin with a picture.
In summer, people who spend time in nature often notice that some larger leaves, for one reason or another, curl into a shallow bowl, and a small puddle of rainwater collects in the bottom. Look closely and there are tiny floating creatures inside. Their entire world is this little puddle of water; their life and death hang entirely on whether that puddle dries up—the span of the puddle's existence is the entire life cycle of these floating organisms.
In truth, humanity's situation in the vast universe is comparable to those floating creatures in the leaf puddle.
How vast is the universe? Vast enough that its range of temperatures is despairingly wide.
At the cold end, the temperature of the cosmic microwave background is about 2.7 kelvin (minus 270.45 degrees Celsius)—this is the universe's "room temperature." But for humans, it is an extreme cold only 3 degrees above absolute zero (−273.15°C). At the hot end, the temperature of the Sun's core is about 15 million degrees; the cores of massive stars can reach hundreds of millions of degrees; and at the instant of a supernova, the core temperature soars to tens of billions of degrees—these are the "body temperatures" of stars. Pushing further into theory, the "Planck temperature" at the very first instant of the Big Bang reaches 10 to the 32nd power degrees.
From near absolute zero to tens or hundreds of billions of degrees—this is the actual temperature span of the universe. And how narrow is the temperature range suitable for humans, and for every carbon-based life form we know of?
The average surface temperature of the Earth is about 15°C. Liquid water can stably exist, at normal pressure, between 0 and 100°C. But the global average temperature window that can actually sustain an ecosystem woven from complex life—like the Earth's—is pitifully narrow: somewhere between a few degrees and a little over twenty. Nudge it a bit to either end, and for water it either freezes or boils; for the human body, it means either hypothermia or heat stroke.
Hypothermia and heat stroke are life facts everyone should understand:
Hypothermia: when a person's core body temperature (the temperature of the internal organs and brain) drops below 35°C, they enter a hypothermic state—consciousness blurs, the heart races, judgment declines; without timely rewarming, life is at risk. Humans normally maintain body temperature through energy converted from food and the insulation of clothing. In a prolonged cold environment, the ambient temperature need not reach freezing; so long as "cold + wind + wetness" stack together (wind strips away surface heat, wet clothing accelerates heat loss), a healthy person can fall below the hypothermia line within a few tens of minutes.
Heat stroke: at the other end, when the environment is hot, the body dissipates heat by sweating. When the ambient "wet-bulb temperature" (which accounts for both temperature and humidity) reaches 35°C, sweat can no longer evaporate, and the body's heat dissipation fails completely—even a healthy person sitting still can suffer heat stroke within hours and die of it. This is precisely the physical root of the repeated "sitting still can kill you in the heat" reports during recent global heatwaves.
In other words, the ruler of the universe's temperature stretches from near absolute zero all the way to tens or even hundreds of billions of degrees (theoretically up to 10 to the 32nd power)—endless beyond sight; yet the segment that belongs to our survival is only a thin dozen degrees wide—35°C below the body is the life-and-death line of hypothermia, 35°C wet-bulb above the environment is the life-and-death line of heat stroke. Like the puddle in a leaf bowl, one touch tips it over.
And even this dozen degrees is a small accident in the billion-year evolution of the universe, the solar system, and the Earth—the geomagnetic field blocks the solar wind and cosmic rays, the atmosphere retains heat and oxygen, the oceans regulate temperature, the water cycle transports energy... these boundary conditions mesh together like gears to maintain the stability of this dozen degrees: an "ultra-complex giant system" that could come apart at any moment.
Yet now, the intensity of human activity is hammering many of these critical boundary conditions. Scientists call them "planetary boundaries"—climate, biodiversity, the nitrogen and phosphorus cycles, land use, fresh water, ocean acidification... Of the nine major "planetary boundaries," six have already been crossed beyond the safe line.
Among them, the warming of the water-air system is the most severe, and the most irreversible.
II. What Solutions Has Humanity Come Up With?
Faced with warming, humanity has come up with, in all, just two moves.
The first: reduce greenhouse gas emissions. Burn less coal and oil, cut carbon emissions, and thin out the "big quilt" of greenhouse gases. This is the logic I explained in Global Warming: A Truth Too Painful to Dissect—and it is that "truth" I could not bear to dissect, one that is "physically incomplete and game-theoretically unworkable."
The second: switch to "clean energy." Use solar, wind, hydro, nuclear, tidal, geothermal, hydrogen, biomass, and other energy sources that "don't burn fossil fuels" to replace coal and oil. This has almost become the loudest slogan of our time—"green transition," "energy revolution," "carbon neutrality."
The question is: are these energy sources really "clean"?
III. What Does "Clean" Energy Even Mean?
The term "clean energy" is so familiar to everyone that it has been accepted by default. But if you press on what "clean" actually means, the answer is not so simple.
First, look at the Earth's heat circulation system—this is the key to understanding everything.
Energy on Earth can be roughly divided into a few categories:
Primary energy includes:
Solar energy: the most important source. The heat the Sun shines onto the Earth every day is the "main engine" of the Earth's climate system—driving atmospheric circulation, the water cycle, and photosynthesis. This heat is the "heat that should be there" in the biosphere.
Geothermal energy + heat from radioactive decay in the crust: coming from the Earth's interior, sustaining the geomagnetic field, driving plate tectonics, earthquakes, and volcanoes. These too are heat that "should be there" in the Earth system.
Wind, tidal, and hydro energy: the kinetic energy of the Earth's ecosystem—wind is the kinetic energy of atmospheric circulation, tides are the ocean kinetic energy driven by lunar gravity, river flow is the driving force of river ecology. In nature they are always "doing work": wind transports water vapor and heat, tides drive ocean currents and deliver nutrients to nearshore ecosystems; river water energy is potential energy—water flowing from high to low is part of the Earth's water cycle and ecosystem.
Coal, oil, and natural gas: chemical energy, the "ancient solar energy" sealed underground hundreds of millions of years ago, locked away, existing as static matter that can be converted to heat through combustion.
Secondary energy includes electricity and hydrogen.
Here is the question:
Which are the heat that "should be there" in the biosphere, and which is the extra heat humans release into the biosphere—either by burning something into heat, or by converting it first to electricity and then discharging it as waste heat?
True "cleanliness" should be broad and oriented toward sustainable development. In an ideal state, it should not pollute, not warm, and not destroy ecology.
No pollution—not just atmospheric pollution, but also the wastewater, waste material, heavy metals, and toxic substances produced when manufacturing these devices;
No warming—this is the most easily overlooked. The heat the Sun shines onto the Earth is heat that should be there; but apart from the solar energy already reaching the Earth, the vast majority of energy humans extract and convert will ultimately become excess waste heat discharged into the water-air sphere. And heat is the final pollution.
No ecological destruction—not flooding forests, not clearing habitats, not poisoning soil and water sources, not creating waste that cannot degrade for hundreds of years, or nuclear waste that cannot degrade for tens of thousands of years. This should also include: not altering ocean currents, not disturbing the air currents of the troposphere, not weakening the Earth's magnetic field.
IV. What "Clean Energy" Is There—and Is It Really Clean?
Whether something is "clean" cannot be judged only by the day it generates electricity; it must be judged across its entire life cycle—from being dug out of the ore, to being made in the factory, to being installed on the mountain, to being buried after decommissioning.
1. Solar Energy
Photovoltaics: it is true that PV panels emit no carbon while generating electricity, but PV panels are not naturally formed—they are "made."
The core material for making PV panels is polysilicon—purifying the silicon in sand to "electronic grade" is an extremely energy- and electricity-intensive process requiring temperatures of over a thousand degrees. That is to say, before a PV panel produces its first kilowatt-hour, it must first "eat" a large amount of electricity—and today, most of that electricity still comes from burning coal.
In other words, the first half of a PV panel's life is spent "using coal to build a device that doesn't burn coal." The industry calls this "energy payback time"—a panel must generate electricity for several years to "pay back" the energy consumed in making it. This period has been greatly shortened by technological progress (from several years down to a little over a year), but it genuinely exists and cannot be erased.
More troublesome are the byproducts of manufacturing: polysilicon production emits toxic waste such as silicon tetrachloride, which pollutes if mishandled; and the lead, cadmium, and other heavy metals in PV panels become new e-waste after decommissioning. Meanwhile, the first global wave of PV panel "retirement" is arriving—most of these panels currently have no mature recycling pathway.
Concentrated Solar Power: CSP takes another route: using vast mirror fields (parabolic troughs, tower heliostats, dishes) to focus sunlight onto a point, heating a working fluid (thermal oil or high-temperature molten salt) to produce steam, which then drives a turbine to generate electricity.
It has an ability PV cannot match: thermal storage. Molten salt can be heated to five or six hundred degrees and held for over a dozen hours—storing heat by day and releasing it to generate power at night. This gives CSP plants the ability to "store heat by day and supply power at night," partially solving solar energy's "at the mercy of the weather" intermittency problem. This is its most brilliant advantage.
But CSP also carries its own debts.
First, it occupies enormous land. To concentrate enough light requires spreading out vast mirror fields, sited where sunlight is strong and the land is flat desert or gobi. Once the mirror field is laid, that whole expanse of land, and the desert ecosystem upon it, is entirely "covered"—altering surface albedo and crowding out the habitats of local plants and animals.
Second, and most ironic: it consumes water, yet is built precisely where water is scarce. The turbines of a CSP plant need cooling, and the mirrors need regular washing to remove dust—all of which requires water—while its best sites are exactly the sunniest and most arid deserts. Building a power plant that needs large amounts of water in a water-scarce place is itself a twisted contradiction.
Third, high construction cost and long cycles. The mirror field, molten-salt system, and thermal storage tanks are all heavy investments, far costlier than PV—one reason it still accounts for a far smaller share of global new energy than PV.
Within the solar family, PV and CSP each have strengths and weaknesses: PV wins on being cheap and fast to deploy, but is at the mercy of the weather and strikes at night; CSP wins on thermal storage and peak-shaving, but occupies land, consumes water, and is expensive and slow. Their common undertone is that sentence—they all use sunlight that already shines on the Earth, which in one sense is the "cleanest"; but to turn sunlight into electricity available on demand still exacts the price of manufacturing, land, water, and waste disposal.
2. Wind Energy
It is true that the spinning blades of a wind turbine emit no carbon, but wind turbines do not come from nothing.
Many modern wind turbines use "direct-drive" technology, which requires a key material—rare-earth permanent magnets, in which neodymium and dysprosium are among the most precious and the "dirtiest" of the rare earths. Refining rare earths uses strong acids and alkalis and produces radioactive slag, which once exacted a heavy environmental toll on China's Bayan Obo mine and on rare-earth mines around the world.
Another unspoken difficulty of wind power is the blades. Wind turbine blades are made of fiberglass and resin—large and rigid—and there is almost no mature recycling technology for them worldwide. Europe has already begun piling up hundreds of thousands of decommissioned blades that can neither degrade nor be economically recycled; they can only be cut up and buried—and these blades will sit in landfills for hundreds of years.
And then there is the ecological hurdle: the impact of wind farms on birds and bats, and on local airflow and even surface temperature, is still under study, and the conclusions are not all "harmless."
The wind energy intercepted by turbines also originally had ecological functions. If large-scale wind farms cause regional atmospheric stagnation, there would presumably be ecological consequences as well.
3. Hydroelectric Power (Dams)
Hydropower has always been regarded as the "cleanest" energy—water falls from a height, drives a turbine, and not a drop of carbon is burned.
But this "cleanliness" has a fatal blind spot: reservoirs behind dams flood vast areas of land and vegetation. The submerged forests, grasslands, and farmland rot slowly in oxygen-deprived water, releasing large amounts of methane—and methane's warming potential is dozens of times that of carbon dioxide (over 80 times on a 20-year scale).
That is to say, a hydropower station, even as it generates electricity "cleanly," may be quietly emitting a greenhouse gas stronger than CO₂ into the atmosphere through the rotting matter at the bottom of its reservoir. This is another illustration of the "big quilt" logic: a reservoir does not emit carbon dioxide, but it emits methane.
Moreover, dams sever the connectivity of rivers, altering downstream sediment, water temperature, and fish migration paths—its changes to river ecosystems are permanent.
4. Nuclear Energy (Fission Reactors)
Nuclear energy is an even more twisted existence.
From the "carbon dioxide" angle, nuclear power plants are nearly zero-carbon in their generation process—this is the biggest reason they have been welcomed back into the "clean energy" camp.
But nuclear energy has three problems that cannot be avoided:
First, nuclear waste. The spent fuel from a nuclear plant remains radioactive for tens of thousands of years, and humanity still has no truly safe, widely accepted final disposal solution. The so-called "deep geological disposal" has been attempted for decades, yet no country has truly implemented it.
Second, accident risk. Chernobyl and Fukushima have already told the world twice: nuclear energy is 99.99% "safe," but when that 0.01% happens, the cost is generations of people, a swath of land, and a stretch of ocean.
Third, and most easily overlooked—waste heat. The cooling towers of nuclear plants are among the most efficient and concentrated "waste heat emitters" on Earth. Nuclear power does not emit carbon dioxide, but it still discharges enormous heat into the oceans, rivers, lakes, and atmosphere.
5. Fusion Energy
Fusion energy, dubbed the "artificial sun," has been pinned with "savior"-level hopes. The principle of fusion is to fuse hydrogen isotopes into helium under ultra-high temperature and pressure, releasing enormous energy. Unlike fission, it produces no long-lived radioactive waste, burns no fossil fuel, and its feedstock (deuterium and tritium in seawater) is nearly inexhaustible. At first glance, this does seem the closest to a "perfect" energy.
But is it truly "perfect"?
First, it emits waste heat just the same. This is the most direct—and most stinging—extension of the greenhouse "big quilt" logic: fusion, fission, or burning coal—any method that converts matter that should not enter the surface heat cycle into energy and releases it as waste heat is ultimately "lighting a fire under the quilt." Fusion energy will not "bring down the Earth's fever"; it will only let humanity "add fuel" to that quilt more fiercely, faster, and with less restraint.
Second, it has carried a nickname for decades: "always 30 years away." Since the 1950s, scientists have said "controlled nuclear fusion is 30 years away." As of 2026, it is still "30 years away." The International Thermonuclear Experimental Reactor (ITER) budget has grown again and again, its schedule slipped again and again, and although various private fusion companies periodically announce "ignition achieved" and "energy gain," those are laboratory instants measured in seconds, still separated from "stable, sustained, commercially grid-connected power" by several mountains.
Third, and most counterintuitive: even if fusion succeeded tomorrow and went commercial tomorrow, it might not be the cure but an accelerant of the poison.
Think about it: if humanity truly possessed an energy source that was "inexhaustible, everlasting, and seemingly carbon-free," what would happen?—humanity's last shred of awe toward energy would vanish completely.
By then, the root disease of "unrestrained growth in total demand" would lose its final brake. Today we still restrain ourselves a little because "electricity is too expensive" and "oil is running short"; once fusion makes energy "free," that last bit of restraint disappears too. The "prisoner's dilemma" described in the Global Warming essay—whoever first imposes limits on themselves loses—would only be amplified to the extreme.
6. Importing Solar Power from Space, and Helium-3 from the Moon
There are two even grander visions, farther and more "divine" than fusion.
The first is Space-Based Solar Power (SBSP): building giant PV power stations in orbit, converting sunlight directly into electricity, and "beaming" the power back to the ground via microwave or laser. The second is lunar helium-3: an ideal fusion fuel (producing almost no neutrons, cleaner than deuterium-tritium fusion) that barely exists on Earth but is relatively enriched in the lunar soil. So some envision mining helium-3 on the Moon and shipping it back to Earth as fusion fuel.
Both visions point to the same impulse: humanity finds Earth's energy insufficient, and finds it "dirty," so it wants to reach into space to "take" energy.
This very act is wrapping the quilt that can no longer be held down with yet another layer, and adding another fire.
Take space solar first. It sounds beautiful: in space the sun never sets, there are no clouds, and the radiation intensity is several times that of the ground. But the accounting goes like this: to put a power station capable of supplying a city into orbit requires launching tens of thousands, even hundreds of millions of tons of material—and launches rely on chemical rockets that burn fossil fuels, with every launch first dumping a huge amount of carbon into the atmosphere; assembling and maintaining these materials in orbit is another ocean of energy consumption. Worse still is how it transmits power back—microwave or laser transmission itself suffers significant losses, and the energy dissipated en route does not vanish; it still becomes heat, remaining in the atmosphere.
An even more counterintuitive point: a space solar power station artificially "increases" the solar radiation the Earth receives. As discussed in the September 13 essay, one plan to cool the Earth is "shading"—blocking 1.7% of solar radiation at the L1 point; but the direction of space solar is exactly the opposite: it is taking that portion of sunlight that should not reach the Earth, additionally "moving" it onto the Earth, then converting it into electricity, and finally into heat. One is shading, the other is moving—their directions are diametrically opposed.
Now lunar helium-3, which has two inescapable knots:
First, it depends on fusion technology that is still "always 30 years away"—deuterium-tritium fusion is not yet commercial, and pinning hopes on "cleaner" helium-3 fusion (harder, with more extreme temperature requirements) is staking an answer that has not been born on another even more distant unborn answer.
Second, and most ironic: going to the Moon to mine helium-3 is itself "consuming vast amounts of energy for the sake of energy." Mining, purifying, and shipping it back across vast distances—every step is fossil-fuel rockets and mega-projects burning money and energy. And even if, after untold hardship, it were mined and brought back, the electricity generated from it would still ultimately become waste heat that does not belong to the Earth, tucked into the greenhouse gas "quilt."
So whether it is space solar or lunar helium-3, neither escapes that most fundamental conclusion: so long as the root disease of "humanity wanting ever more energy" is not cured, taking energy from space is merely moving the hand that "adds fuel" from beside the stove to outside the stove—the fire will only burn hotter.
7. Tidal Energy
Tidal energy is ocean kinetic energy driven by lunar gravity. It "burns nothing and emits no smoke," sounding cleaner than anything. But tidal energy hides two special debts.
First, engineering destroys ecology. Tidal power stations either build a barrage at a bay mouth (tidal barrage) or deploy large numbers of underwater turbines (tidal stream). A barrage blocks the exchange between the bay and the open sea, altering the rhythm of tides, causing silt accumulation, blocking fish and shrimp migration, and degrading tidal-flat ecology—its transformation of marine ecology is the same thing as a reservoir dam's transformation of a river.
Second, and more fundamental, rarely mentioned: the tide is originally the "engine" of marine ecology. The rise and fall of tides drives ocean currents, stirs the seawater, and transports nutrients and oxygen to nearshore waters, sustaining the flourishing of the intertidal and nearshore ecosystems. When humans intercept it and turn it into electricity, then into heat, it is equivalent to hijacking a set of ecological energy that was already "doing work" for the ocean, and converting it into waste heat released into the atmosphere.
This is not to say tidal energy cannot be used at all; rather, the price of its "cleanliness" is not only engineering and ecology, but a deep paradox of "unhooking the ecosystem's engine and burning it as firewood."
8. Geothermal Energy
Geothermal energy comes from the Earth's interior—heat from radioactive decay, plus the primordial heat left over from the Earth's formation. It is true that it "burns no carbon," but geothermal energy likewise has two problems.
First, the pollution of extraction itself. Deep geothermal extraction (especially enhanced geothermal systems) requires injecting water underground to fracture rock, which may induce microseisms or even earthquakes; the extracted geothermal water often contains dissolved hydrogen sulfide, heavy metals, and small amounts of radioactive material, which pollute if mishandled.
Second, and most critical, geothermal heat is originally the "nourishment" of the Earth's magnetic field. The thermal convection of the core and mantle is exactly the energy source that sustains the Earth's magnetic field "dynamo"—and the geomagnetic field is the "invisible shield" that protects the surface from solar wind and cosmic rays, one of the key boundary conditions sustaining the stability of the Earth's biosphere. When humans channel underground heat out to generate electricity, small-scale use may be negligible, but if mined on a large scale, the long-term impact on the geomagnetic field and deep geological stability remains unclear to this day. Some might say that compared with the enormous thermal energy of the Earth's interior, what humans can extract is negligible and the consequences can be ignored. Friends who think this way should consider how the vast ocean has been destroyed by humanity's repeated "tiny" pollution discharges. The impact of 8 billion people on the Earth can no longer be ignored.
Tidal energy, geothermal energy, along with the preceding hydro (potential energy) and wind (kinetic energy), share one common feature—they are the "ecological energy" of the Earth's ecosystem. Wind must blow, water must flow, tides must rise, the Earth's core must stay hot—these energies sustain the stability of climate, ocean currents, and geomagnetism in nature. When humans intercept them and turn them into electricity, ultimately into waste heat released into the atmosphere, it is essentially "turning the ecosystem's engine into a waste-heat boiler."
9. Biomass Energy
Biomass (burning straw, wood chips, and charcoal for power) is touted as "carbon neutral" because plants absorbed carbon as they grew, and it is released again when burned—on paper, "income and expenditure balance." The problem with biomass energy, on the one hand, is that it ignores the deforestation, occupation of farmland, and the fact that the "carbon debt" repayment period may stretch for decades—in the most urgent decades of the climate crisis, biomass may actually be "accelerating carbon release"; on the other hand, biomass is an organic product formed by biological systems absorbing carbon dioxide, water, and other nutrients through photosynthesis—it is the essence of the ecosystem, a precious ecological resource. In the material cycle of the ecosystem, it should come from dust and return to dust, "turning into spring mud to better nourish the flowers." If biomass is used as energy, it amounts to severing the material cycle of the ecosystem and causing ecological imbalance.
10. Hydrogen Energy
Hydrogen has been hyped by many media as the "ultimate clean energy"—because it produces only water when burned, which sounds perfect.
But hydrogen has a fundamental problem: hydrogen almost never exists in free form in nature; it is secondary energy—it must be "made."
So whether hydrogen is "clean" depends on the entire process of making, transporting, and storing it—what electricity is used, what fuel is burned. This leads to the famous "color" trap.
Hydrogen used as energy today is graded by how it is made:
Grey hydrogen: produced by reforming natural gas or coal at high temperature. This is the absolute mainstay of today's hydrogen supply, over 95%. Making one kilogram of hydrogen this way emits more than a dozen kilograms of carbon dioxide—essentially no different from burning natural gas, except that the carbon emission is shifted from the "user end" to the "factory end."
Blue hydrogen: grey hydrogen with an added "carbon capture" step to grab and bury part of the emitted CO₂. The name turns blue, but the capture is not thorough—at best seventy or eighty percent, with the rest still emitted—and the "capture-storage" process itself consumes extra energy. It is grey hydrogen wearing a light blue coat.
Green hydrogen: splitting water into hydrogen and oxygen via electrolysis, touted as "zero-carbon." However, electrolysis is extremely electricity-hungry, requiring forty to fifty kilowatt-hours to make one kilogram of green hydrogen. "Where the electricity comes from" determines whether it is truly green—if the electricity comes from solar, the hydrogen is indeed fairly green; if it comes from coal power, it is merely "grey hydrogen in a green skin."
Apart from the manufacturing process, there is another difficulty: hydrogen is terribly hard to handle.
Hydrogen is the smallest, lightest, most reactive molecule, highly corrosive to nearly all materials, penetrating everywhere and extremely prone to leakage. Compressing it into high-pressure gas, or liquefying it into an ultra-low-temperature liquid at −253°C, or first converting it to ammonia and back again... every step leaks and consumes energy. Some estimates put the total energy loss of hydrogen from production, through storage and transport, to final use as high as thirty to forty percent.
The above is the "explicit account"—efficiency, cost, storage, safety—that every article on hydrogen covers. But hydrogen energy also has a "hidden account" that the vast majority of articles omit—and it is this account that truly determines whether it deserves the four words "ultimate energy." This is the indirect greenhouse effect of hydrogen leakage.
Hydrogen itself is not a direct greenhouse gas, but once it leaks into the atmosphere, it does two bad things:
First, it consumes the atmosphere's "detergent." The atmosphere contains a substance called OH (hydroxyl radical), which acts like a detergent, responsible for breaking down methane and other greenhouse gases. Hydrogen binds with hydroxyl to form water—water vapor is itself a greenhouse gas—and the more hydroxyl is consumed, the longer methane lives—and methane's greenhouse effect is 28 times that of carbon dioxide.
Second, it promotes tropospheric ozone formation, and ozone is itself a greenhouse gas. Taken together, hydrogen's global warming potential over a 20-year horizon (GWP₂₀) is roughly 33–37 times that of carbon dioxide. This mechanism is not a new discovery—it was quantified by the atmospheric chemistry community in the early 2000s; but only in recent years has it moved from academic papers to being recognized as an issue that could affect the direction of an entire industry. At the end of 2025, a Stanford-led team published the first comprehensive accounting of global hydrogen emissions in Nature, confirming that between 2010 and 2020, hydrogen emissions had already warmed the Earth by about 0.02°C (other studies using a 1990–2020 timeframe estimate the same order of magnitude). 0.02°C sounds small, but remember: this is the contribution of today's "tiny amount of leakage" alone. Imagine, if hydrogen really scales up as the optimistic narrative expects, and leakage rises by orders of magnitude, how high would this number climb?
Hydrogen energy has two very different modes of use. Centralized industrial hydrogen—steel mills, chemical plants, ports—has short chains, few interfaces, and professional operation, with an average full-chain leakage rate of about 4.5%, still within manageable range. But decentralized hydrogen—passenger cars, building heating, residential refueling stations—doubles the links and interfaces, and shifts maintenance from professional engineers to ordinary car owners, with leakage risk amplified systemically.
A 2025 assessment by the Polytechnic University of Turin shows that by 2050, the overall leakage rate of the hydrogen supply chain could fall below 2% in the optimistic scenario, or approach 20% in the worst case. What does 20% mean? Roughly equivalent to each kilogram of hydrogen produced, the climate cost of the leaked portion equals 7.4 kilograms of carbon dioxide; while the carbon saved by replacing gasoline with one kilogram of hydrogen is only about 8–10 kilograms—meaning, in the worst case, leakage can eat away seventy to ninety percent of hydrogen's carbon-reduction benefit.
This is the crux of the crux: hydrogen is not "naturally clean energy," but "clean energy with strict preconditions." That precondition is that the full-chain leakage rate must be held below 1–2%. Below 1%, the carbon-reduction benefit is essentially fully realized; at 5%, the benefit clearly shrinks; at 10%, green hydrogen's 20-year advantage is greatly eroded, and blue hydrogen may even become worse than directly using natural gas.
Moreover, leaked hydrogen has yet another, more hidden consequence: escape.
Hydrogen is the lightest molecule in the universe. The Earth's gravity cannot hold it. Every year, about 95,000 tons of hydrogen escape from the top of the Earth's atmosphere into outer space—this is the most important pathway of the Earth's continuous mass loss.
This raises a question most people do not know about: escaped hydrogen is a permanent, irreversible loss—once hydrogen escapes the atmosphere, it will never return to the Earth. The hydrogen reserves the Earth accumulated over billions of years through geological and biological processes (the vast majority locked in water) diminish a little with every escape.
Of course, to be truthful: the vast majority of currently leaked hydrogen (over 99%) is oxidized by OH in the atmosphere within 50 kilometers of the ground, or absorbed by soil and turned back into water, remaining in the Earth's water cycle—only a tiny fraction truly escapes into outer space. The claim that "hydrogen will immediately all escape" is physically unsound, and should not be used as a sensationalist pretext.
But the direction—that "escape is a permanent loss"—is still worth remembering. It reminds us that hydrogen's "hidden account," besides "leakage indirectly warms the Earth," also includes a longer-term account concerning a "non-renewable resource": letting hydrogen escape bit by bit is also quietly "leaking away" a resource that could have been recycled.
This reasoning was raised four years ago. In April 2022, when "green hydrogen" was still the hottest capital darling, a researcher writing under the pen name "Mr. Wuhao" issued a warning on Zhihu: could large-scale hydrogen use cause irreversible environmental consequences through leakage? How much hydrogen escapes into outer space each year, and as ground-level hydrogen leakage increases, would it accelerate hydrogen escape? That answer was tagged by the platform as "controversial content with no reliable source," restricted in reach, and ignored.
Four years later, the scientific community has progressively confirmed the direction of this judgment through Nature, Harvard, and RMI research reports. The industry, meanwhile, has paid an expensive tuition in real money—by incomplete counts, in the past 18 months at least 52 commercial-scale clean hydrogen projects have been publicly cancelled worldwide, totaling millions of tons of capacity: in July 2026 Air Products abandoned its Louisiana blue hydrogen project, Germany's HyScale100 green hydrogen project was halted, Plug Power wrote off 13.6 million euros; further back, BP cut four large hydrogen projects within half a year, and Shell, ExxonMobil, and Equinor all retreated. The IEA's Global Hydrogen Review 2025 lowered its 2030 low-emission hydrogen capacity forecast for the first time, by about 25%. The reasons are highly consistent: cost too high, demand insufficient, over-reliance on subsidies.
Hydrogen is indeed irreplaceable in some scenarios—as a reducing agent in steel metallurgy, a feedstock for chemical ammonia synthesis, ocean shipping, and long-duration energy storage—but these scenarios share one feature: centralized, short-chain, professionally operated.
(Note: the arguments in this section concerning the indirect greenhouse effect of hydrogen leakage, decentralized leakage risk, and the tracing of the "Mr. Wuhao" warning are provided by the Zhihu and Toutiao author "Mr. Wuhao," referencing his pieces A Systematic Calculation, Clean Hydrogen Projects Densely Abandoned Globally, Hydrogen Does Not Deserve the Title "Ultimate Energy"—the Missing Half of the Cards, and Minutes of the Hydrogen Environmental Risk Assessment Technical Exchange Meeting, among others. Readers interested in "Mr. Wuhao's" articles may search his account on Zhihu and Toutiao.)
V. Waste Heat and Primary Energy
Every "clean energy" above, whether or not it emits carbon, cannot escape the same iron law of physics—conservation of energy. Energy is neither created nor destroyed; it only changes from one form to another; and its final destination is waste heat.
From the perspective of the Earth's heat cycle, all energy can be divided into three categories.
The first: "ecological energy" that already exists on Earth. Wind is the kinetic energy of atmospheric circulation, hydro is the kinetic energy of the water cycle, tides are the ocean kinetic energy driven by lunar gravity, geothermal is the deep energy sustaining geomagnetism. These energies have been "doing work" in nature for hundreds of millions of years, sustaining the stability of climate, ocean currents, and geomagnetism. They are the "ecological energy" and "geological energy" that "should be there" in the water-air sphere—the Earth's "primary energy."
The second: "fossil stored energy" locked underground. Coal, oil, and natural gas are solar energy sealed underground hundreds of millions of years ago in the form of chemical energy. They were originally lying there, not participating in the surface heat cycle. It is humans who dig them up and burn them, merging this "dormant energy" into the atmosphere's heat budget in a short time.
And the squandering of this category is not merely "injecting extra heat"—coal, oil, and natural gas are also "irreplaceable" material feedstocks for human material production. Oil is the source of plastics, chemical fibers, medicine, fertilizer, synthetic materials, and thousands of chemical products; coal is the parent of metallurgical coke and much fine chemistry; natural gas is the feedstock for synthetic ammonia, methanol, and other basic chemicals. To burn such a "material wealth"—accumulated over hundreds of millions of years, and non-renewable once used—into heat and carbon dioxide is a double waste: it adds waste heat to the atmosphere, and it personally burns away the "family assets" that future generations could have used as feedstock.
The third: nuclear energy that humans "create" themselves. Fission and fusion release energy originally sealed in atomic nuclei. This too is extra injection that "was not there" in the biosphere's heat cycle.
Thus the question becomes clear: to judge whether an energy source is "clean," ask whether it increases the waste heat that "should not be there" in the biosphere.
Wind, hydro, tides, geothermal: these are the ecological and geological energy already "doing work" in the Earth system. When humans intercept them and turn them into electricity, then heat, it is like unhooking the "ecosystem's engine" to burn as firewood—energy that should be circulating within the system is converted into excess waste heat. Coal, oil, natural gas: digging up and burning fossil stored energy buried deep underground, which did not participate in surface circulation, is pure "extra injection"—and what is burned is humanity's irreplaceable material feedstock. Nuclear energy: releasing energy that should not be released from atomic nuclei is likewise "extra injection." Solar energy: it is the only heat that "already shines on the Earth." But even solar energy, once "captured" by PV panels and concentrated into electricity and heat, changes its natural distribution of "being reflected back to space versus absorbed."
This is the truth about "clean energy" that I most cannot bear to dissect: "clean" can only wash away "carbon emissions"; but "warming"—apart from the solar energy already shining on the Earth—no energy can wash clean. Because their common essence is to turn "energy that should not enter the surface heat cycle" (deep-buried fossil energy, nuclear energy in atomic nuclei) or "ecological energy that should circulate within the system" (wind, water, tides, geothermal) ultimately into waste heat released into the atmosphere. And heat is the final pollution—it has nowhere left to go.
VI. What Is to Be Done?
At this point, some readers may feel discouraged: by your account, no energy on Earth is truly "clean," so why should we transition at all? Should we just give up?
What I want to say are three things that are more honest, and more urgent:
First, that "clean energy" is not clean does not mean it should not be used. Compared with coal and oil, on the three rulers of "pollution, warming, and ecological destruction," they remain, at least for now, the overall better choice—the per-kWh emissions of solar and wind are an order of magnitude lower than coal power. What we should do is "shift from dirtier to less dirty," rather than retreat to "dirtier" because "the less dirty is also imperfect."
Second, stop being fooled by the adjective "clean." Its real danger is not that it is inaccurate, but that it stops people from questioning further. One sentence—"we are developing clean energy"—is enough to reassure the public and policymakers that the problem is being solved—the same psychological mechanism as the "1.8°C is a placebo" and "carbon neutrality is a painkiller" in the September 13 essay: using a new "good word" to build the illusion that "the problem is already on track."
Third, and most important—true "cleanliness" requires a full accounting. Not whether "smoke comes out at the moment of generation," but the full life cycle "from cradle to grave": how much ore was mined, how much energy consumed, how much waste emitted, how much ecology destroyed, and whether it can ultimately be recycled. Only by making this full accounting clear can we distinguish which "clean energies" are truly better, and which merely pollute in a different way.
What is truly "clean" is energy that adds as little "extra fuel" to the surface heat cycle as possible. Solar energy, being heat that already shines on the Earth, does not count as "extra fuel" when used; wind, water, and tides are ecological energy already in circulation, and borrowing a small part of them "appropriately and on a small scale" may, for now, disturb the Earth's heat cycle less than digging up and burning deep-buried fossils or releasing nuclear energy (long-term consequences need detailed assessment).
In other words, energy is not a black-and-white "dirty/clean," but a spectrum from "should be there" to "extra injection." What we should do is move from the "extra injection" end of the spectrum (fossil, nuclear) as much as possible toward the "should be there" end (solar). In between, wind, water, and tides may serve a transitional role.
And more fundamental than the full accounting is this: so long as humanity's total demand for energy keeps growing without restraint, no "clean energy" can fill this bottomless pit. However green PV is, making PV requires electricity; however green wind power is, making turbines requires rare earths; they will all ultimately become waste heat, tucked under the "big quilt."
The truth about "clean energy" and the truth about "global warming" are actually the same truth: so long as humanity's total energy use on Earth keeps growing wildly, neither "carbon reduction" nor "clean energy" can change the reality of global warming, global pollution, and ecological collapse. What humanity needs to find is an energy-saving, energy-reducing economy—a sustainable economy in harmonious coexistence with the Earth's ecosystem.
Those floating creatures in the summer "leaf puddle" will not ponder whether their home will dry up—they lack the ability. But we have this ability. May we not give up thinking.