A chapter by chapter summary of Bill Gates’s brilliant new book “How to Avoid a Climate Disaster”
It’s rare to find a book so topical, concise and pressing that is also a great delight to read. If you’re keen to learn the important takeaway details from this book, but don’t have time to read it, this article is for you.
We emit 51 billion tons of CO2 equivalents each year. We need to get to zero.
The idea of climate change first entered the national debate in 1970. At this time, it was widely accepted that the best way to raise the quality of life and spread economic development was to expand the use of fossil fuels. Now we have a new problem, but are struggling with the remnants of this old mindset. But we did a large scale change in energy then, and we can do it again.
During the pandemic, greenhouse gas emissions dropped just 5%. This is proof that we can’t get to zero by just flying and driving less. We need serious innovation.
Chapter 1: Why zero?
Once greenhouse gasses are in the atmosphere, they stay there for a long time. 1/5 of the CO2 we emit now will still be there in 10,000 years. In a zero carbon future, we’ll probably be emitting some CO2 still, but we’ll remove the carbon we emit.
A small increase in global temperature is a big deal. During the last ice age, the average temp was just 6 °C lower than today. During the age of the dinosaurs, it was just 4 °C hotter than today on average, and there were crocodiles living above the arctic circle.
Molecules with 2 of the same gas don’t trap heat (like O2). When they’re different, like CO2, they trap radiation and contribute to warming of the planet. We have already raised the temp by 1 °C since preindustrial times, and if we keep emitting at this rate, temp will go up 1.5–3 °C by mid century, 4–8 °C by end.
Due to rising temps, storms are getting more powerful (more water rises into the air when it’s hot, and hotter air can hold more moisture). Droughts are happening more often for same reason, worse wildfires too. In California, wildfires are now 5 times are frequent as they were in the 1970’s. Sea levels are also going up because polar ice is melting, and seawater expands when it’s warmer. A rise of 2 °C will cut geographic range of vertebrates by 8%, plants by 16%, and insects by 18%. Extra heat will make agricultural animals less productive and more prone to dying young, so animal based food will get more expensive. Coral reefs might disappear completely, as will the seafood population they support, which feeds more than a billion people.
Political instability will also follow as the world gets hotter. 2007–2010 was Syria’s worst drought on record. 1.5 million people left farming areas for cities, which set the stage for the armed conflict in 2011. That drought was made 3x more likely by climate change. By 2080, lower crop yields might make 2–10% of Mexican adults try to cross the border into the US.
By mid century, we’re on track for climate change to be just as deadly as COVID, unless we do something.
Chapter 2: This will be hard
By 2060, the amount of buildings (accounting for number and size) will double. That’s like putting up a new NYC every month for 40 years. This is mainly happening in developing countries like China, Nigeria, and India. Emissions from the US and EU have been pretty stable or even dropped since 1990, but developing countries are skyrocketing. Global energy demand will go up 50% by 2050. In sub-Saharan Africa, only half of people have reliable electricity at home now.
In terms of just emissions, China is the biggest emitter by far. Note that the US’s emissions would be about 8% higher if we factored in the goods that we use, but are manufactured elsewhere. Britain’s would be 40% higher!
The energy transition we need now is being driven by something that never mattered before, the environment. Before it was just about cost and efficiency of fuel. We’ve gotten really good at extracting fossil fuels and getting energy out of them, and their cost doesn’t reflect the damage to the environment. So their monetary cost is extremely low.
The US needs to invest more in nuclear. More people die from coal pollution each year than have died in all nuclear accidents combined.
Chapter 3: Five questions to ask in every climate conversation
- How much of the 51 billion tons per year are we talking about?
At Breakthrough energy, they only fund technologies that can get up to getting rid of 1% of global emissions each year, that’s 500 million tons a year.
2. What’s the plan for all 5 emissions categories?
Making things, plugging in, growing things, getting around, keeping warm and cool. All these things matter, so if it’s a technology that cuts down on one, but requires tons of cement and steel for example, that needs to be considered.
The percent of global emissions produced by each category:
Making things (cement, steel, plastic) 31%
Plugging in (electricity): 27%
Growing things (plants, animals): 19%
Getting around (planes, trucks, cargo ships): 16%
Keeping warm and cool (heating, cooling, refrigeration): 7%
3. How much power are we talking about?
A watt is a bit of energy per second. Gigawatt is billion watts, megawatt is million. Kilowatt is thousand.
The world: 5,000 gigawatts
The US: 1,000 gigawatts
Mid-sized city: 1 gigawatt
Small town: 1 mega watt
Average American home: 1 kilowatt
4. How much space will it need?
There’s limited space, so we need to think about this.
Think about power density, in watts/sq meter
Fossil fuels: 500–10,000
Hydropower (dams): 5–50
5. How much will this cost?
Consider the green premium. Remember, we need green options to be cheaper than fossil fuel options if we want them to be adopted globally.
Direct air capture can help us capture some of the CO2 back, but note that if we relied on it exclusively to offset our emissions, it would cost $5.1 trillion/yr, about 6% of the global economy. It’s not feasible. Point capture right at the source is another way of capturing it. While these won’t help us get to zero, we’ll need to be removing 10 billion tons of CO2 per year by 2050 even if we get emissions under control now. But note this capture tech only exists for carbon, not other, more potent greenhouse gasses.
Chapter 4: How we plug in
Electricity in the US is really cheap. It costs about half a penny to leave a 40W lightbulb on for an hour. Electricity is 27% of emissions, but if we get a source of clean energy, we could use it to decarbonize some of the other categories. Getting clean reliable electricity is the single most important thing we can do to avoid a climate disaster.
One of the US’s early forms of energy was dams. Hydropower can be good, but there’s hidden costs. When you cover land with water, carbon in the soil turns into methane and escapes into the air. A dam can actually be a worse emitter than coal for 50–100 yrs before it becomes greener.
Today, fossil fuels account for 2/3 of electricity generated worldwide. Hydropower and nuclear are next.
The Government does a lot to keep fossil fuels cheap, this encourages their use and production. In 2018 alone, subsidies were about $400 billion.
Between 2000–2018, China tripled the amount of coal power it uses. That’s more capacity than US, Mexico, and Canada combined.
Changing America’s electricity to zero carbon would add up to about $18 per month extra for the average home, but most countries aren’t so lucky. We have a lot of natural resources. Small scale solar power works for individuals in rural homes, but it won’t be enough to grow the economy and attract industries. Poor countries will turn to fossil fuels just like China did, unless they have a cheaper option.
Transmission and distribution of electricity actually account for over 1/3 of the final cost of electricity. Storing electricity during the night, and moreover during seasonal intermittency is a huge problem because batteries are expensive, and actually close to as efficient as possible. Gates estimates they can get 3x as good as now, but not 50x. We need to figure out how to store electricity on a large scale.
In the US, we have many power grids, so we can share electricity locally, but not over multiple states. We could make lots of high voltage current wires, but there’s a lot of political hurdles to work across state lines with such a big project. Also, they’ll be in the air, burying them is really expensive because you have to add a cooling mechanism so they don’t melt. In the air, the heat just dissipates. But if we don’t upgrade our grid and left it to individual states, the green premium will be much higher.
Between now and 2050, America needs to build renewables for electricity 5–10 times faster than our current rate.
Making carbon-free electricity
The US gets about 20% of its energy from nuclear plants. France has highest rate, at 70%. We didn’t stop using cars because they were dangerous, instead we innovated to make them safer. We need to do the same for nuclear energy. TerraPower is working on ways to use less dangerous fuel, including waste from other nuclear plants, and it would produce less waste.
Nuclear fusion is a cool idea, that’s putting atoms together, it powers the sun. You get hydrogen super hot (50 million degrees C) to turn it into plasma. Particles move so fast that they collide and fuse, release energy. This could run on hydrogen extracted from sea water. Less dangerous waste too, but it’s hard to do. ITER is the biggest project working on this. By 2030’s it should be able to generate 10x more power than it uses, if it does, we can start to scale it up, but it’s a few years away still.
Offshore wind is a good option, US has lots of windy coast to put it on. But currently, the US uses very little of it. We should use more. The wind out there is less intermittent, and it could be generated near lots of places that will use it, so you won’t need to transport it as much.
Geothermal is cool, where you can get it. But you need a volcanic area, and 40% of wells dug for geothermal turn out to be duds. Innovation is happening, but this will likely only make a modest contribution overall.
Batteries will likely only improve 3x, not 50x. Some innovations are happening that might change this, like the use of liquid metals instead of solid ones. We’ll see.
Pumped hydro is cool, use electricity to pump water when it’s cheap (or small rocks) up a hill (or underground under pressure), then release later to generate electricity when electricity demand goes up. People are working on this, but it hasn’t taken off yet.
Thermal storage: use electricity to heat something up when it’s cheap, when demand goes up, use the heat to generate power via a heat engine. Molten salt is a promising material to store heat.
So storing electricity is tricky on a grid-sized scale, but if we figure out how to get cheap hydrogen, that problem would suddenly become obsolete. Hydrogen is a key ingredient in fuel cell batteries. These work through a chemical reaction between hydrogen and oxygen, and the byproduct is water. If we could use clean electricity to generate hydrogen, that would solve the storage problem, but right now, it’s expensive to make hydrogen without emitting carbon. Also, you have to use electricity to make the hydrogen in the first place, and it tends to leak out of containers because it’s so lightweight, so it’s inefficient. Also, the electrolyzers involved in the process of making hydrogen are very expensive.
We can help out by using less, and also implementing load shifting to even out when we use electricity.
Chapter 5: How we make things
31% of global emissions
Each year, America produces 600 pounds of cement for every person in the country. Same again for steel. In raw amount, that’s a small fraction of what China has been doing since 2001. China makes more cement than the rest of the world combined.
To make steel, you need iron. You need to separate the iron and oxygen in iron ore. Then you add a little carbon. To do this, you melt iron ore at really high temps (3000F) in the presence of oxygen and coal coke. The coke will release carbon, some of which bonds with the iron, the rest binds to the oxygen, creating CO2 as a byproduct. Making 1 ton of steel produces 1.8 tons of CO2! One approach is trying to use electricity to break apart the iron ore, so no CO2 is produced at all, but it hasn’t worked at scale yet.
To make cement, you need the calcium out of limestone. Limestone has calcium, carbon, and oxygen. So you melt it at super high temps, and CO2 is made as byproduct. It’s an inevitable 1:1, make a ton of cement, produce a ton of CO2.
To make steel and cement, you need such high temps that it’s only realistically possible from fossil fuels or nuclear. Cement is the toughest case. Some people are trying to inject some carbon into the cement after manufacturing it keep it sequestered, but this would only reduce emissions by 10%. There are other more theoretical approaches, but it’s early.
Plastics have only been around since 1950 or so. They all contain carbon, usually from refining fossil fuels. When we make plastic, half of the carbon stays in the plastic, unlike cement and steel. Plastics are a huge environmental concern, but at least they’re not contributing as much to CO2 emissions. If we can find a way to use carbon captured from the air and generate the plastics with clean electricity processes, plastics could actually become a carbon sink, with net negative emissions.
So basically we use fossil fuels to make things:
- To generate the electricity that factories need to run their operations
- When we use them to generate heat need for manufacturing (like melting iron ore)
- When we actually make the materials, like how cement and steel manufacturing inevitably produces CO2
Right now we don’t have a lot of options other than carbon capture to deal with the emissions created in stage 3, and that adds substantially to the green premium. For plastics 9–15%, steel is 16–29% and cement is 75–140%.
We need a good source of clean electricity to replace fossil fuels wherever we can in manufacturing.
Chapter 6: How we grow things
19% of global emissions
In agriculture, the main culprit is actually methane, which causes 28x more warming than CO2, and nitrous oxide (265x). Food is getting cheaper. Average household spends less of a percent of income on food than it did 30 years ago, general global pattern. Global population is headed towards 10 billion people by 2100. And as people are getting richer, they’ll eat more, and eat more meat.
2 key innovations: Norman Borlaug’s semi-dwarf high-yield wheat and synthetic fertilizer via the Haber-Bosch process. It’s allowed us to grow the global population 40–50% larger than it could have otherwise.
Cow burps and farts account for about 4% of global emissions each year. They produce it via enteric fermentation of cellulose. The second biggest emitter in agriculture is poop, from pigs, cattle, etc. Poop releases nitrous oxide, methane, sulfur, and ammonia. We can cut emissions by spreading the efficient Western breeds to poor areas of the world. This will also make poor farmers richer.
Plant-based meat imitations are getting better, but it comes with an 86% green premium right now. There’s also cultivated meat, won’t be available for a few more years, and not sure how much it will cost. We can also waste less food. US wastes 40% of food, that’s double many industrialized nations. One approach is using invisible coating to make fruits and veggies last longer.
Fertilizer: in Africa, many farmers can’t afford it, so they get 1/5 of the yield per acre as a Western farmer. Plants get the nitrogen they need from ammonia in the soil, which is made by microbes. To grow crops, you need way more nitrogen than is naturally present in soil. To make fertilizer, we make ammonia, which requires heat usually produced by burning fossil fuels. Then we transport that fertilizer using gas powered vehicles. Then once it’s in the soil, half the nitrogen runs off to contaminate water sources or evaporates as nitrous oxide, which is a greenhouse gas. Fertilizer accounts for about 2% of global emissions. We can use clean energy to generate the fertilizer, but it comes with a 20% green premium right now. One approach genetically modifies plants to hold more nitrogen so less runs off, but it’s not very efficient. Another recruits microbes to the plants that fix nitrogen, but it’s early.
So all the stuff above, broadly agriculture, accounts for 70% of emissions from farming. The other 30% is deforestation. We’ve lost 3% of forest since 1990. When you burn down a tree, the CO2 stored in the tree is quickly released, and the soil is loosened, so more carbon is freed up. Soil actually contains more carbon than the atmosphere and all plant life combined. Planting trees is often talked about, but realistically can’t keep up with the pace we’re emitting.
In short, we’ll soon need to produce 70% more food than we are now, while cutting down substantially on our current emissions. We’ll need new ways to fertilize plants, raise livestock and waste less food. We should also eat less meat.
Chapter 7: How we get around
16% of global emissions
Gasoline is a dense energy source, and it’s really cheap. Transportation contributes 16% of global emissions. However, in the US, it’s actually the biggest emissions category because we drive and fly a lot.
Globally, passenger vehicles account for half the emissions, medium vehicles (garbage trucks, 18 wheelers etc) are about 30%, planes are 10%, cargo and cruise ships are about 10%.
Cars: There are about a billion cars on the road today. Electric cars are getting cheaper because of battery innovation, and Gates thinks the green premium will be zero by 2030. But they take an hour to fully charge, and the range is smaller than for a tank of gas. Also, remember that this is only a good thing if the electricity is clean. Also, it’ll take a while to switch because people keep cars for several years.
Some fuels are using carbon that was already in the air, that’s a good approach as it’s not adding more carbon.
Biofuel: In the US, most fuel is already 10% ethanol from corn, but growing and refining corn might make just as many emissions, so it might not be a good choice. Existing biofuels just aren’t efficient and cheap enough. We need a lot more innovation. Key: we need to use biofuels that aren’t using land used for food, and don’t require fertilizer. The green premium for the best biofuels now is 106%!
Hydrocarbon fuels: also called electrofuels. we can use clean electricity to combine the hydrogen in water with carbon from CO2 that you capture from the air. These are a great option. But right now they’re super expensive (237% green premium) because it costs a lot to make hydrogen without emitting carbon.
Medium size vehicles: For busses and garbage trucks, batteries can work because they’re not super heavy and they don’t have to travel long continuous distances. But for freight trucks, batteries are not a practical option. China is using a lot of electric busses. Green premium for busses will probably reach zero in the next decade.
The best lithium ion batteries now pack 35x less power per weight than gasoline. If we made freight trucks run on batteries, to go the same distance on one charge as a gas-powered one, the whole weight would need to be batteries. So for freight trucks, our only options right now are biofuels and electrofuels.
Ships and planes: 20–40% of a plane’s weight is fuel, so no way can we use batteries, way too heavy. So same as fright trucks, our best bet is biofuels and electrofuels. Green premiums for these are 141%-296%. Same goes for cargo ships (3% of global emissions).
So to cut emissions: bike and walk more, carpool etc. Use less carbon-intensive materials in manufacturing vehicles. Enforce standards for fuel efficiency to force innovation. Where possible, switch to electric vehicles and alternative fuels.
How to lower the green premium: make policies to buy electric vehicles, making more charging stations. Need to innovate in clean electricity. Nuclear powered container ships could help. Military submarines and aircraft carriers already use this tech. Need to innovate in alternative fuels.
Chapter 8: How we keep cool and stay warm
7% of global emissions.
AC is not just a luxury, the modern economy depends on it. It keeps server farms cool for computing for example. Domed stadiums wouldn’t be possible without it. In your home, AC uses more electricity than your lights, fridge and computer combined. But more energy is taken up by furnaces and water heaters actually, but this is often gas powered. Right now in poor countries, only 10% of households have AC, so this number is going to go way up. AC will triple by 2050. Also, we’ll need it more as the world gets hotter.
Most people buy to most energy conscious AC machines. Sometimes a unit is cheaper to buy but more expensive to maintain long term, so we need to label these more clearly. Many countries also don’t have efficiency standards for AC units. We should do that. But it’s not just the demand for electricity. AC units emit refrigerants, called f-gasses, very potent greenhouse gasses). F-gasses are responsible for 3% of emissions. Less harmful refrigerants are being developed, but not available yet.
Heating: we need to heat water, and keep our homes warm in cool seasons. Furnaces and water heaters are responsible for 1/3 of emissions from the world’s buildings. And unlike AC, it mostly runs on fossil fuels not electricity. So like transportation, we need to electrify where we can, and use clean fuels where we can’t. Electric heat pumps are a great option, and can actually save you money. We don’t replace this stuff often, but when we do, it’ll be adopted easily. There are some outdated regulations that won’t let you put in an electric one because it’s less efficient than fossil fuels (though it emits less), so we should update that.
We can make buildings more efficient with tight sealing and double glazed windows. We need to introduce policy to set standards for building efficiency to encourage innovation to drive cost down.
Chapter 9: Adapting to a warmer climate
In poor countries, climate change will shorten growing seasons, leading to malnutrition and death. We need to raise the odds that malnourished children will survive. We need to improve health care there. We need to help poor farmers grow more food. CGAIR is the world’s largest agricultural research group, and they’re doing the most of any organization to help poor farmers adapt to climate change. They’re developing new crop and livestock varieties. They educate farmers on best practices. They developed an app to allow farmers to identify pests with their smartphones. Organizations like CGAIR are chronically underfunded. We should double their funding.
We need to change the ways cities grow. As urban areas rapidly expand, they build over floodplains, forests and wetlands that absorb waters during a storm or reservoirs that can be used in drought. City planners need to work with projections from climate models. Today, many city leaders in the developing world don’t have maps that indicate the most flood-prone areas of a city for example. Forests store and regulate water. Wetlands prevent floods and provide water reservoirs. We need to protect these natural resources. Mangroves for example prevent some $80 billion per year in losses from floods, and these events will get more frequent with climate change.
As lakes and aquafers shrink or get polluted, by mid century, 5 billion people will not have enough access to clean water. Desalinating is an option, but it’s really energy intensive. Taking water out of the air is a cool approach that might work. It’s a solar power dehumidifier which also filters the water to take out the air pollution. It’s available, but expensive now.
The financial problem: we need to pay the cost of adaptation up front, but it might not pay off for years. We need to make adaptation an attractive investment. Governments can set goals with climate change in mind and remove some of the risk for private investors to attract more money. The benefits will come in terms of bad things that don’t happen. We don’t often think that way in terms of finding good investments.
Extreme poverty has plummeted from 36% in 1990 to 10% in 2015, but climate change might regress a lot of this positive change.
Geoengineering: this is like an emergency case. We could make temporary changes in the ocean or atmosphere to lower the temperature and buy us some time. It doesn’t absolve us of having to deal with it entirely though. Basically, to compensate for all the warming caused by greenhouse gasses, we need to reduce the amount of sunlight hitting the earth by about 1%. Like scattering tiny particles in upper atmosphere to reflect light away. Can use a salt spray to make clouds more reflective. This would last about a week, and we can keep doing it as long as we need to stop with no long term impacts.
Chapter 10: Why government policies matter
In the 1950’s and 1960’s, big cities like LA and London had horrible incidents of smog that killed thousands of people. Policy makers responded by funding research into the problem and passing laws like the Clean Air Act, which created zones in cities where only cleaner fuels could be used. In 1970, Nixon created the EPA to continue environmental protection. China passed several laws starting in 2014 that caused 35% reduction in some types of pollution in Beijing.
Luckily, we have a lot of experience regulating energy. It’s one of the most heavily regulated sectors of the economy.
Seven high level goals governments should aim for
- Mind the investment gap
Right now, the private sector and energy companies themselves, spend very little money innovating in energy. It’s a risky and long term investment, with no consumer demand. We need government incentives like a price on carbon or mandates for percent of energy that has to come from clean sources etc.
2. Level the playing field
We should make the price of fossil fuels reflect the price on society, like with a carbon tax or cap and trade program. Government can also add incentives for using clean energy.
3. Overcome nonmarket barriers
There’s a lack of education about some available green technologies, a lack of dealers to install and maintain, and in some cases regulations to prevent their use, like in the case of electrical heat pumps. Governments can get rid of these barriers.
4. Stay up to date
In the case of concrete, there are performance standards and chemical composition standards. New green cement might not be the same chemical composition, but it might pass performance standards. Standards need to include the latest advances in technology.
5. Plan for a just transition
Some communities will lose lots of jobs in the fossil fuel industry, as well as the associated tax revenues to pay for roads and schools etc. Local leaders need to plan for solutions. Federal government can also help by providing funding, technical advice, etc.
6. Do the hard stuff too
Electric cars and more solar power are easier to demonstrate progress in, and that’s really important, we need to be doing the easy stuff way more quickly, but we can’t ignore the hard stuff like storing electricity, clean fuels etc. Governments need to invest more R&D in the hard stuff.
7. Work on technology, policy, and markets at the same time
These three things have to work in a coordinated way. Policies can drive innovation, and conversely, technology should change policy.
Chapter 11: A plan for getting to zero
We need to plan for getting to zero emissions by 2050. We need to expand the supply of innovation, as well as the demand for it.
We need to quintuple energy and climate related R&D over the next decade. Right now, it’s $22 billion globally, that’s 0.02% of the global economy. US is the largest investor by far, and spends $7 billion. NIH budget is $37 billion/yr, we need something on that scale. Right now, government investments in clean energy are safe and short sighted. These projects should be funded by the private sector. The role of the gov is to invest in high-risk high-reward projects that don’t need to pay off right away, like the human genome project. For that, every $1 invested returned $141 to the US economy. Governments and industry should work together from the earliest stages of a project and focus on areas of greatest need.
To drive demand: gov can prioritize buying green energy to ensure a market. Gov can also give private sectors incentives to go green. These policies need to be technology neutral (favor any tech that reduces emissions, not just funding a few favored ones), predictable (rather than the cycle of extending and expiring that happens now), and flexible enough for many different companies to use. Gov needs to infest in the infrastructure necessary, like transmission lines for wind and solar, charging stations for electric vehicles etc. Update regulations to match existing tech and prioritize getting to zero emissions.
We need to put a price on carbon. This could be a price on carbon, or a cap and trade where companies can buy and sell the right to emit carbon. This extra revenue can go to consumers to help cover the extra energy costs or go into R&D.
We need clean electricity standards. 29 US states and the EU have adopted a renewable portfolio standard. Electrical utilities must get a certain percentage of their electricity from renewable resources. But these currently exclude nuclear and carbon capture, which drives prices up. We should instead emphasize clean energy, not necessarily renewable. Nuclear isn’t renewable. We can do the same to enforce clean fuel and product (including imported) standards. We nee to create incentives to shut down fossil fuel burning plants earlier than we would have otherwise. Out with the old.
Chapter 12: What each of us can do
The most important thing you can do as an individual is engage with the political system. Make calls, write letters, attend town halls. Run for office.
As a consumer, you can send a signal to the market that consumers are willing to pay a green premium for greener options.
Sign up for a green pricing program with your electric utility, will cost the average American home $18/month. Not all states offer this though. Reduce your home’s emissions by using a smart thermostat, efficient lights, use recycled steel for renovations, use good insulation etc. You can buy an electric vehicle. Eat less meat.
As a company, you can do things like impose an internal carbon tax.