Making a Dent (The China Edition)
TLDR
The IPCC tells us that we need to make a real start in reducing emissions. The US showed us a way to do this, dramatically reducing the amount of CO2 the nation produced through a switch from burning coal to burning natural gas to make electricity. China generates the most electricity in the world and relies on coal to do it. Cutting their emissions by reducing or eliminating coal use there could drop global CO2 emissions by as much as 9 Gt per year, or 16%.
But it's highly unlikely, physically or geopolitically, that we could do anything like this by changing coal use to natural gas use in China. A more palatable solution would be a combination of nuclear, solar and wind replacing coal. Even this would demand a level of cooperation among nations that is highly unlikely to happen, today.
So, could we drop global emissions by 16% if we concentrated on replacing China's coal use with renewables? Yes, we probably could. Will we do it, as involved nations? Likely no, because we are too busy squabbling over much less important issues. What we are doing now is likely a preview of the bun-fights we will be having over the necessity to conduct geoengineering operations to reduce heat levels in the future.
The Introduction
Let's see if it's actually possible to start making real cuts in global CO2 emissions, over a reasonable timeframe, that would meaningfully move us toward a 50% drop in global emissions.
The United States accidentally showed us the way, by following Rule #1 in (ahem) Hykawy's Rules for Climate Change Mitigation (I will be changing the name to reach ever more grandiose heights, over time):
Through to the 1990s, the majority of electricity generated in the US was made by burning thermal coal. There really was no good alternative. But in the late 1990s, US companies, seeking a way to unlock shale oil deposits, started widespread deployment of hydraulic fracturing, or 'fracking'. And in the 2000's, fracking dramatically increased the volume of US oil production, at least while oil prices were high, and also produced a large amount of natural gas. So utilities in the USA decided that if natural gas were available and inexpensive, they might as well burn natural gas instead of coal.
The use of coal to generate electricity in the United States peaked in 2007 and it's been downhill, since. Once generating stations were refitted to burn cheaper natural gas, there is no quick or cheap path back. Donald Trump promised, during his 2016 election campaign, to bring back 'clean' coal, but economics is a much more powerful force than dreams and empty promises, and coal continues to decline.
An unintended result of this change was that CO2 emissions from the United States dropped, significantly. Unfortunately (for me; fortunately for most of the rest of you), LinkedIn doesn't have an easy way to incorporate equations into these articles. But coal is mostly carbon, chemical formula C, and burning it with the oxygen gas, O2, in air makes carbon dioxide, CO2. CO2 is the major greenhouse gas driving climate change and is produced by a wide variety of human activities and industries.
But burning natural gas, much of which is methane, CH4, produces energy, some CO2 and also water, H2O. In fact, burning methane instead of carbon cuts carbon emissions significantly; most sources tell us the amount of CO2 is cut in half (which matches up reasonably well with calculation). There are other advantages, including a very large reduction in fine particulate emissions and some nasty co-pollutants from coal like sulfur compounds.
The United States went from about 48% of its electricity being generated from coal in 2007 (peak coal use) to only 16% in 2023; the gap has been filled mostly by natural gas and some from the increased use of renewables. But the effect on CO2 emissions from the US power sector has been impressive. In only 16 years, the US reduced effective CO2 emissions from the power sector by almost 1 Gt while saving money and reducing local environmental pollution. Remember, this was not done to reduce CO2 output, this was strictly to save operating costs.
But what if we throw the doors open and have some fun? Let's see what could be done if we aimed to reduce greenhouse gas emissions: how much and whether (or not) the proposed change is even technically feasible.
China (Hearts) Coal (Not)
Let me state something as a fact, for which there is little supporting data: the powers-that-be in China would very much like to reduce coal use and clean up their environment. Now, I know that coal use in China is on the rise, because I see a news story every couple of days about how China just doesn't care, is building more coal plants and it's just more than the authors of these stories can even bear, deep and dramatic sigh.
But China is also building more wind and solar photovoltaic capacity than any other nation, by a very long way (some sources opine that, in 2023, China generated 37% of all the wind and solar photovoltaic electricity generated on Earth; enough electricity to power Japan). They also built, against massive pressure from environmentalists outside China, by the way, the Three Gorges Dam Project (I have visited and it is impressive). Electricity demand has been going up every year in China and, unfortunately for all of us and in spite of pretty drastic measures to try to curtail demand and build any and all alternatives, coal use is still rising:
The obvious major differences in this graph, compared to the US profile in Figure 1, are that the US graph for total amount generated is roughly flat and has been for some time where the Chinese graph is continuing up and to the right. In other words, China's modernization and demand for electricity has not yet stabilized. But the bigger difference, from a climate change perspective, is that there was no dramatic shift to natural gas use in and around 2008.
The reason for this is simple enough. China is not blessed with abundant shale oil deposits like those in the United States. They would love to be. China imports a lot of oil and gas and considers this dependence on foreign feed a strategic weakness. But the lack of domestic petroleum deposits means that there is no easy way for China to produce huge amounts of domestic natural gas. And, unlike the healthy relationship between the United States and Canada (at least up until recently), China does not consider Russia reliable enough and friendly enough to rely on with respect to supplying the fuel necessary for its domestic electricity production.
China does import a lot of coal, as well. But this is because, when feasible, they import cleaner-burning coal from other nations. Were those supplies to be cut off by a hostile force, however, China also has large domestic reserves of dirtier coal that could be relied on. It wouldn't be ideal but it would suffice in an emergency.
Chinese entities are also planning or building a large number of new nuclear reactors. There are now 57 operating nuclear reactors in China (2nd largest fleet in the world, behind only the US with 94) but is building another 28 out of the 62 that are under construction, worldwide (second in terms of reactors under construction is India with a whopping 7; there are no new nuclear reactors under construction in the United States). It would be a valid question to ask why China isn't building 128 nuclear reactors instead of 28 and planning to shut down coal plants across China. The answer is much the same as for coal and natural gas, above; China does not have large reserves of uranium and is unwilling, as I was once told, to hand control over the fuel necessary for domestic electricity production to the United States and Russia (I 'corrected' the person telling me this by saying "You mean Canada and Kazakhstan", and was properly, politically re-oriented by a wry smile and "I mean the United States and Russia!"). Never mind the issues stemming from the fact that we are in a period of sustained uranium shortages, globally (but that's another story).
Just for fun, let's make a wild leap. Let's say the world recognized the value of helping China reduce CO2 emissions because, as Hykawy's (ahem) Rule #1 so eloquently and correctly points out, replacing the burning of coal with the burning of gas or oil is preferable. Let's ignore both cost and China's concerns regarding national security for a moment and see whether this would make a reasonable impact.
China Burns Gas Instead of Coal
In 2023, about 61% of China's electricity generation came directly from coal. Other sources (US EIA and CEIC) tell us that China's total electricity generation from coal was 93.809 quadrillion BTU. The US EIA tells us that burning coal to produce a million BTU makes about 98 kg of CO2 (and a lot of other nasty stuff like fine particulates, sulfur compounds, mercury and aerosolized radioactive elements like uranium and thorium, which is why Chinese authorities would really like to clean up their power generation; CO2 is not at the top of their list of concerns). A quadrillion of anything is a billion million, so burning enough coal to make a quadrillion BTU would make 98 billion kg of CO2, or 98 million tonnes. China made 93.809 quadrillion BTU from coal, thus also making 9.2 billion tonnes of CO2 (in metric units, 9.2 Gt).
From Figure 2, above, the total manmade CO2 equivalent emissions, globally, were 57.1 Gt. So Chinese coal burning, alone, is 16% of worldwide carbon emissions. Now, before we go suggesting China should just turn off their electricity (to which China would likely reply "you first, dummy", but in a more diplomatic way), let's also note that the US has not completely removed coal from its own generating mix. People have to eat and move and live, no matter what nation they call home. The hard work needed is to find a way to do that while also reducing or eliminating carbon emissions and not destroying the economy in the process.
According to the US EIA, burning natural gas to make a million BTU would generate 53 kg of CO2 (burning other hydrocarbons to make the same amount of energy, like liquid fuel oil, generates more CO2 because the liquids generally contain longer chains, with relatively more carbon and less hydrogen than simple methane). So, if China could, somehow, convert all those coal-burning operations to natural gas and then find enough natural gas, they would only generate 5.0 Gt of CO2 equivalent.
That would yield, on it's own, a 4.2 Gt drop in CO2 emissions at 2023 levels, or 7.4%. Considering, according to Figure 2, that all of road-going transportation (light-duty, like driving to the train station to get downtown; medium-duty, like an Amazon delivery van; and heavy-duty, like a semitrailer rig that runs all day for several days with a pair of drivers) is only good for 11% of all global emissions, ridding ourselves of 7.4% by converting over a countable number of thermal plants seems easier.
But is it possible?
Where Would the Gas Come From?
Now, let's figure out what the volume requirement is for that gas.
If China needed 93.8 quadrillion BTU in 2023, we can work out the mass of coal that is needed, just for laughs. According to the US EIA, a million BTU worth of energy is contained by just under 36 kg of coal (depending on type of coal, quality of coal and amount of moisture). So 93.809 billion million BTU comes from 3.38 Gt (billion tonnes) of coal. So, a lot.
Again, according to the US EIA, a million BTU from natural gas requires 26.8 cubic meters of it. So, making 93.809 billion million BTU means we need 2,514 billion cubic meters of gas. Statista estimates that global natural gas production in 2023 was 4,080 billion cubic meters. Enerdata says slightly higher production, about 4,397 billion cubic meters. Still, adding slightly more than 60% of total global production, plus finding a way to get it all to China in a reliable enough manner (commercially as well as politically) that China would consider making the switch is, I think we would all agree, not something that can happen overnight.
领英推荐
How Do You Get It All There?
To get all that natural gas to China, we would need to liquefy all of it and put it on ships to move it from North America, South America and the Middle East. Other possible solutions don't improve the energy density of the transport very much, if at all (for example, taking natural gas and steam reforming it, locally, to produce hydrogen gas that is then transported does nothing useful; too much energy loss along the way and hydrogen is a pain to ship).
The International Institute of Refrigeration suggests that making liquefied natural gas (LNG) requires 30-35% of the energy content of the gas being liquefied. So, in 2023, we would have needed 3,394 billion cubic meters of natural gas to make a total of 2,514 billion cubic meters of natural gas in liquefied form. Now we are looking at increasing global natural gas production by more like 75% from present levels.
Worse, we were previously suggested that natural gas could save as much as 4.2 Gt of CO2-equivalent emissions by replacing coal in China. But now, we are lowering that to as little as a saving of 2.5 Gt of CO2-equivalent (really depending on where the energy comes from for making LNG; if we are in the Middle East and can derive the energy from sun and wind, the saving would be higher). A 2.5 Gt reduction in global CO2-equivalent emissions in 2023 would still have been 4.3%, but this isn't making me feel like 50% reductions in emissions by 2030, as the IPCC is imploring us to do, are achievable.
OK, What About Existing Renewables?
I'm inclined to look most carefully at nuclear, but I'm a physicist so you'd probably expect that. I am certainly not set against using solar photovoltaic (PV) systems or wind, but there are issues. Those issues are so serious that refusing to consider using nuclear because wind and solar are somehow "perfect" is making perfection the enemy of saving ourselves from the negative impact of climate change.
To see what might work, we need to determine what our target is. We know that about 61% of Chinese electricity comes from coal. And the CEIC tells us that, in 2023, Chinese electricity production was 8,972 TWh. So we need to replace 5,473 TWh of coal-fired electricity with something else that doesn't produce (as much) CO2.
OK, what about solar PV, the use of solar panels? Well, there are a couple of issues, here. The first is that solar photovoltaics don't produce electricity when it's dark or even as much when it's partially cloudy. The second is something we can describe as "power density".
So, the first issue, first. We want to generate 5,473 TWh of green electricity to replace coal in China. OK. There are something like 8,766 hours in the (average) year. So, if we had some perfect renewable electricity generator that is able to work for 24 hours every day, then we could generate 5,473 TWh of green electricity, in a year, with only 624 GW of generating capacity (only!). But the big problem with solar is that day/night cycles mess things up. No electricity at night. Less electricity being generated in the winter. Less electricity generated during daylight hours when it's cloudy. We need to build more solar generating capacity to compensate for the times when electricity is not being generated, as much or at all.
The US helpfully collects data on this. The data we are looking for are called 'capacity factors'. Statisticians helpfully tally up all the nominal generating capacity for a specific source, like solar photovoltaic, and then tally up all the electricity generated by these assets over a year. Then they calculate a ratio for how much was generated versus what could have been generated had peak power output been produced all the time. The US EIA tells us that the capacity factor for solar photovoltaic systems across the USA was only 23.3%. Widespread use in China will probably be worse because a lot of the open and best land for solar PV is in the north, where winter sunlight is scarce. But if we use 23.3%, then we see that we actually need to install not 624 GW of solar PV capacity but 2,680 GW. China installed 253 GW of solar PV in 2023, out of a worldwide total of 447 GW, according to Solar Power Europe.
Now, this doesn't seem that bad, right? China could replace coal in less than 11 years. Unfortunately, electricity demand in China is growing. China's National Energy Administration reported that total generating capacity grew by about 360 GW in 2023. So there was quite a bit more thermal capacity, meaning coal, in use by the end of 2023 than at the beginning. Even with large-scale construction of renewables, China is getting further and further behind. But like us, the people in China want to go to work and make enough money to make things better for their families. And that requires cheap energy, so China is not going to voluntarily shut down the coal plants without replacement, nor should we expect that.
Now, the second problem. When you put up a solar farm, you can't put another solar farm on top of it. Our current-generation solar cells are pretty much as efficient as they are going to get, bar small improvements, so the output power for a given land area is the output power. And that output power isn't great. The US DoE has some papers on solar PV power density that could put anyone to sleep, but the upshot is that you aren't going to do much better than 0.35 MW/acre, no matter where you put your solar farm, in 2023 (and, in fact, you could do a lot worse depending on location and latitude). 0.35 MW/acre is the equivalent of 86.5 MW/square kilometer.
So, to generate 2,680 GW of solar power, we need to cover about 31,000 square kilometers with solar panels at current best specifications, mounting it all up and also wiring up all those solar panels, spread out over a 176 km by 176 km square, to get the power out. This is about the same area as the entire State of Maryland in the USA. All this bearing in mind that the best possible solar PV farm locations are likely already used, so we are progressing into poorer and poorer areas that might be extremely remote, as we go.
We can do the same calculations for wind. It's the same 624 GW of perfect generating capacity that we need to get the job done. Unfortunately, the wind doesn't always blow at the perfect speed, just as the sun doesn't always shine when we need it. The capacity factor of wind power in the USA, on land, is 33.5%, better than for solar PV. From that, we would only (!) need to install 1,863 GW of wind turbines. In 2023, China did install about 77 GW of wind. This isn't exactly sprinting toward what would be needed.
But if the problem with areal power density is big in solar PV, then it is positively massive with wind. Again, you can't plop another wind farm down on top of an existing wind farm. Unfortunately, if you put wind turbines too close to one another, the wake and turbulence from one will damage downwind machines. The TNO of the Netherlands has done studies on this topic (windmills, the Dutch; go figure) and you really can't do much better than about 5 MW per square kilometer, onland, with the current biggest turbines.
So, to install a generating capacity of 1,863 GW, we would need to cover 372,600 square kilometers with wind turbines. And wire them up. And transmit the power. This means covering an area the approximate size of the US States of Montana or New Mexico with wind turbines. And while you could do some things with the land in-between the turbines, this isn't an undisturbed natural environment, anymore.
OK, how about nuclear, then? Well, the story isn't perfect here, either; far from it. If we need 624 GW of some perfect electricity generating technology (and we actually need more because of peak versus average demand, but that's the least of our issues, right now) then we only need 670 GW of nuclear because the capacity factor of nuclear sites in the USA in 2023 was 93.1% (yes, nuclear power plants do go out-of-service for maintenance and repair, but they operate at the highest level of any electricity generating asset because they are, and definitely should be, overengineered). And based on the 54 nuclear plants in use in the United States, the US-based Nuclear Energy Institute says that nuclear plants produce about 1,000 MW per 1.3 square miles, so a high 294 MW per square kilometer.
This means we might only need a land area of 2,279 square kilometers to get the nuclear plants installed. This is a little more than half the area of the State of Rhode Island in the USA. Given that we are actually thinking of replacing coal plants, which also require sources of cooling water for their boilers and the like, it might be possible to build a number of these nuclear plants on existing coal plant sites and piggyback on the available power transmission infrastructure, too (we get the connections to the grid for free, maybe). In 2024, there were 1,161 coal-fired power stations in mainland China, so we might even get our pick.
But there's a catch. There's always a catch. A large nuclear reactor, in rough terms, can generate about a GW of energy. Some do less. In rough terms, we need about 700 new, large nuclear power plants to be built. The entire world production of uranium serves a fleet of only 440 nuclear reactors, as of May 2024. In other words, we would need to roughly triple current uranium production levels to get close to replacing coal in China (and convince China that supplying nations will continue to supply them with fuel, barring whatever future arguments arise). But at least we can conceivably handle demands for land to do this with.
Other Possibilities?
These do exist, and range from the possible (nuclear reactors using a fuel cycle involving thorium to reduce the strain on global uranium mines) to the more fanciful (a much more efficient but still cheap solar cell) to the very hopeful (FUSION!!!).
How about thorium as a nuclear fuel? Well, thorium is more common than uranium, by something like a factor of 3-4x in nature. Even better, thorium only has one isotope, the one we want to use in a nuclear reactor, while uranium has a few natural isotopes but the one we are most interested in, 235U, is only about 0.72% of natural uranium. The actual effect is that thorium is about 400-550x more common, from the point of view of nuclear fuel.
Unfortunately, I need to say 'but' here. 235U is fissile, which means it occasionally breaks apart on its own and produces energy plus additional neutrons. It is also very good at capturing a slow-moving neutron, becoming an unstable isotope of 236U and then splitting apart into smaller nuclei, some energy and between 2-3 fast moving neutrons that can be slowed down and used to split even more 235U nuclei, what we know as a chain-reaction. 232Th, the one and only natural isotope of thorium, is only fertile. If it absorbs a neutron, it quickly becomes a 233U nucleus, which is now able to absorb another slow-moving neutron and fission. But this means that we can't simply throw thorium into a reactor as a direct replacement for uranium. We need to design the reactor carefully so we don't waste our precious neutrons and we still need some uranium as the 'spark', if you will, to allow our thorium to 'burn'. Some existing designs for a thorium fuel cycle suggest that lifetime uranium demand could be as low as 1/3 of what was required to make uranium-fueled reactors, but building enough reactors to do coal replacement in China would still demand a lot more uranium than we produce today.
How about a more efficient solar cell? What we are used to, single crystal silicon solar cells, mostly top out at around 20-22% efficiency. We can fiddle with esoteric factors like coatings and layering and other similar things, but we are very near the limit. However, a newer idea is to use a class of chemicals called perovskites and layer these with silicon. Perovskites are chemical compounds that look like CaTiO3, although the most common perovskite currently used in solar cells is actually CH3NH3PbCl3. By combining perovskite with silicon, with the two materials absorbing different parts of the solar spectrum, efficiencies can get to be near 30%. This would improve power generation by nearly 50% from current levels, which is not insubstantial.
By now, you are probably expecting a 'but'. But. Perovskites are somewhat unstable in the presence of water and heat, both of which are kind of a given for a solar panel installed outside. Though more robust perovskites are being investigated, they tend to have poorer efficiencies. So that leaves more complete environmental exclusion as the best option, which generally means higher costs. And we don't need or want higher costs while we try to replace coal.
And as for nuclear fusion, I came across an old volume of Scientific American that I purchased when I was young (yes, I am a geek). It was published in 1986, iirc. It proudly proclaimed that recent advances in something known as fusion by inertial confinement (using lasers to flash heat the outside of a pellet of fuel so that the inside collapses and generates high densities and temperatures leading to a burst of fusion reactions and energy production) were so encouraging that commercial use was not far off. Well, not only did that not happen, 39 years later there are very few government-supported inertial confinement projects happening.
The core problem with fusion stems from the fact that stars like our sun use fusion to make energy by cheating. The stars pile so much hydrogen into one place that gravity provides the force needed to keep the hot hydrogen nuclei close enough to one another that they crash into each other and fuse. On Earth, we can't put that amount of mass together, for obvious reasons. So we try to keep a much smaller mass together using stronger forces, like electromagnetic ones, and end up needing to increase both the speed that the hydrogen nuclei are flying at each other with (the temperature) and the number of chances that the collision will cause hydrogen nuclei to fuse together (the confinement time).
To a large degree, these necessary factors are opposed to one another. The hotter the nuclei, the less able we are to keep them in one small area. Now, we can go down a rabbit hole of fusion and start discussing how the very hot hydrogen gas, usually referred to as hydrogen plasma, in a device like a tokomak is inherently unstable, causing significant loss of energy. There are designs, like stellerators, that can create a much more stable plasma, but stellerators have significant design issues that some companies are working on ways to overcome. The net outcome of all of this is that nuclear fusion is probably not going to pop up and save us from our own carbon emissions in the next couple of years.
We need to act on bringing down carbon emissions as quickly as possible. Waiting for technological miracles before we begin is not a viable strategy. We need to start acting on what we can do, today, and if better solutions come along then we can incorporate them as we go.
(Is There) A Possible Solution (?)
So, the answer to whether we could actually displace coal in China and cut as much as 16% of worldwide CO2-equivalent emissions in one swoop is a qualified 'probably not'.
On the one hand, yes, we are capable, as a planet of concerned individuals, to do things to make it happen. We need China to streamline the use of nuclear power and develop something that would fit that framework. Something like a modular nuclear reactor but one that relies on thorium or mixed nuclear fuels so as to minimize dependence on scarce uranium could make a serious dent. Naturally, to be acceptable to China the manufacturing of that modular reactor would need to happen in China and would require politically and commercially reliable supplies of uranium to facilitate the thorium fuel cycle.
Continued development and deployment of both wind and solar power in China would have to go hand-in-hand with the building of nuclear reactors. Use of solar and wind, especially with their inherent unreliability, demands the use of large amounts of cheap grid storage, so it is good that China dominates world production of both lithium iron phosphate batteries, as well as their potential successor when it comes to grid storage, sodium batteries. And, where the space is available, let's not discount mechanical electricity storage such as compressed air or gravity energy storage such as pumped hydro.
But the largest impediment to all of the above, at this point in time, is the lack of international cooperation around the supply of critical materials to the markets that need them. In this case, putting an emphasis on mining more uranium and getting fuel-grade enriched uranium to China as rapidly and reliably as possible is falling by the wayside as China and the US engage in trade squabbles.
The Chinese situation around coal is one of the best examples out there that solving the global climate change problem requires a global strategy. It is not enough to simply have every nation promise to cut emissions without a clear plan as to how or when. In most cases, as with China, drastically cutting emissions requires materials that China does not, naturally, possess. If a nation cannot depend on a long-term, reliable supply of something to reduce emissions, especially something as critical as a material necessary to generating electricity, then that nation cannot adopt a plan that incorporates such a material. I can guarantee that China is not avoiding building a lot more nuclear reactors because they prefer polluting their air with coal smoke; they are avoiding the construction of too many nuclear reactors so that the United States and the European Union can't suddenly turn to China and inform the central government that no more uranium will be delivered due to 'national security concerns'.
So. Could we, as a sensible people, cut 16% of global carbon emissions out of the air by helping China convert their use of coal to much higher levels of nuclear, solar and wind use? Absolutely yes. Will we? Likely not. And so our climate change problems will continue to worsen, year on year, until the point where it's too late to do anything about them. And, at that point, we'll likely be arguing about geoengineering efforts to try and reduce temperatures, and not coming to any serious conclusions, there, either.