Hacker News Reader

Comment by eloff | Hacker News Reader

How can you burn fossil fuels to produce electricity, to use to turn the combustion byproducts back into fuel at any reasonable level of efficiency? Some people here are saying the electricity doesn't matter, because it is from renewable sources - but then you could just shut down that dirty coal plant and use the renewable energy directly. Something doesn't add up here.

View on HN

This is for soaking up renewable energy that either isn't connected to the grid yet, or is being generated at the wrong time. Since there's no fuel cost for wind or solar, you might as well do something with it. Anyone willing to buy electricity at inconvenient places or times will help make renewable energy generation more profitable, so more will be built.

There are lots of potential competitors, though: other kinds of energy storage (such as charging a battery), industrial users that can run in the off hours, or maybe generating Bitcoin. It's probably too soon to say whether this technology will win.

Buying electricity at a low price and selling it at a different time or place at a higher price is basically a form of arbitrage. Hopefully energy storage will become more efficient at getting electricity to customers when they need it, so they can outbid things like Bitcoin mining.

That makes sense, I was struggling to grasp how it could be economical.

Demand elasticity. We are still really far from reasonable deployment.

Talking to a friend who worked for a power utility here, it's a very foreign concept. Some room for disruption there.

For this you would need to store the CO2. For each ton of coal burned you generate about 3 tons of CO2. At atmospheric pressure thats 1500 cubic meters. A small 200 MW power plant could burn 80 tonnes per hour. Thats 120 000 cubic meters. That is a building 12 m high and 100 m square.

I guess everything is possible given the right subsidies...

You wouldn't store the CO2 gaseous at atmospheric pressure, you'd use the energy to compress it, presumably liquefying it. Liquid CO2 has a density of around 1100kg/m3, according to a quick Google search, so your example power plant would need only about 200m^3 to store the CO2 produced in one hour, or a cube of about 6m wide. Or 120m wide per year. Still seems like a lot, and I don't know how much energy is needed to compress and cool all that CO2 but it's dubious the surplus renewable energy would suffice to compress all the CO2 produced by fossile-burning power plants.

That indeed seems like a better idea.

Bitcoin mining unfortunately is typically useless for sucking up spare electricity. The hardware is too expensive, unless it's running 85%+ uptime, you will not make back your initial hardware purchase.

We can thank Moore's law for that. Bitcoin miners go obsolete every time a new generation of silicon hits maturity. 85% uptime requirement will drop substantially after Moore's law tapers off.

> We can thank Moore's law for that.

Except that Moore's law has tapered off substantially. The latest 10nm process node is 3 years late[0], and the next one will probably be substantially more difficult than that.

[0] https://www.anandtech.com/show/12436/intel-10nm-dualcore-can...

Yeah, I had to write a proposal for a carbon capture project. It was really hard to make it make any sense from a systems perspective.

I conclude that talking about using it to reduce emissions from fossil plants is idiocy only useful for PR. You are always better off not burning the carbon in the first place and just getting the electricity from another source.

The only even remotely compelling case is that it lets the plant respond faster to load by dumping excess electricity into chemicals. And even there, it's not the best solution, it's probably better handled by the way we do it now with load-following plants.

I think the researchers know that, they talked about using renewable power for it. But I'm sure their estimates are also based on consuming coal flue gas because they have to goose the numbers to make the CO2 cheap and concentrated enough to make any sense. So my assumption with basically any of these is that they're setting up a paradoxical situation -- assume that we're using fossil power plants for the CO2, but getting the electricity/hydrogen from renewables. That way the chemicals look close to free.

But if we're still generating electricity with the fossil plant, it will always be better to just not use the plant from a CO2 and systems efficiency perspective and to instead build more renewable generation to make up the difference.

Basically this is a reasonable idea for transport fuels. Sunlight gives you 1000 W/m², or 160 W/m² at 16% photovoltaic efficiency. A Boeing 777 has a surface area on top of something under 1000 m², and so if you cover it with solar panels, you get 160 kW when the sun is shining directly on it. But its engines produce somewhere between 700 and 1100 kN at a cruising speed of 900 kph, or 250 m/s, which if you work it out means it would need 175–275 MW if the engines are 100% efficient, a bit over 1000 times what it can gather from the sun. (A bit less since it's not running the engines at full power when it's cruising, but a lot more when you take into account that the engines are quite inefficient.)

There are ways you can design the plane to use less power, and Solar Impulse 2 demonstrated that it is in fact possible to fly on solar power alone, but Solar Impulse 2 also took 23 days to go around the world, has a maximum speed of 140 km/h, and has a maximum passenger capacity of 1, rather than a few hundred. Design compromises like these are necessary to lower the power intensity of flight to a solar-feasible level. So, it's likely that liquid-fueled aircraft (the 777 burns kerosene) will continue to have substantial advantages over solar aircraft for the foreseeable future.

The obvious solution is to use some 10⁸ m² of area on the ground — something like a 10 km by 10 km square for each 777 in flight — to generate electricity to produce the liquid fuels for the liquid-fueled aircraft. That sounds like a ridiculous thing to do, building a city-sized solar plant for a single airplane, and today it would be — the photovoltaic modules would cost US$5 billion and installing them would cost another US$5 billion, for a total of US$10 billion. The entire Boeing 777 development cost was only US$5 billion, and each plane sells for US$300M or so. So the cost of the power plant would be something like 30× as high as the cost of the airplanes themselves.

From an exponential point of view, though, that's only 5 doublings. If the price-performance of photovoltaic modules improves by a factor of 2 five times, the power plant falls to the cost of the plane. If PV modules continue dropping in price by 30% per year, as they have been for the last four years or so, that crossover happens in 2028; if they continue dropping in price by 20% per year, as they have been over a somewhat longer period, it happens in 2033. It's only in a scenario where PV modules effectively stop dropping in price where this remains a ridiculous idea.

The same logic applies, but in a weaker form and a nearer crossover, for other challenging transport fuel uses such as ships and trucking.

This is an excellent point, however even so you'd still have another problem. Let's say this all works great, and we get to the prices that make it feasible. You're still just adding one go around to the carbon cycle. The fuel you burn at the power plant once gets back converted to fuel once, then turns into CO2 that you can't recover when coming out the back of a 777.

The only way this technology gets to the point where it's net carbon neutral or negative is if it can process the highly dilute CO2 coming out of the free air efficiently rather than the highly concentrated CO2 coming out of a power plant. I have a sense that the dilute case is quite a bit more challenging to make work at scale and efficiently.

I suppose you could use the power generated to desalinate water, grow forests where there otherwise isn't enough water, then bury the trees before they have an opportunity to decay into the atmosphere.

Atmospheric carbon dioxide removal via direct air capture is feasible already. The cost is small compared to the price of the fuel; estimates range from US$30–1000 per tonne. More details are at https://pubs.acs.org/doi/10.1021/acs.chemrev.6b00173

That is a big range. One ton of Diesel is $788, but produces about 3 tons of CO2.

Wait, burning diesel produces energy and mass?

Diesel isn't the only ingredient in the reaction. Most of the weight in co2 comes from the oxygen that is taken from the atmosphere.

Indeed. Atomic weights are: carbon 12; oxygen 16. So the oxygen in CO2 is 32/12 = 8/3 as heavy as the carbon inside it.

So, if 100% of diesel were carbon that gets burned, you would get 3.67 tons of CO2. That leaves room for incomplete combustion and for other things than carbon (hydrogen, nitrogen, sulphur, etc).

Thanks for a reply to a sincere question rather than just downvoting.

> You're still just adding one go around to the carbon cycle.

You mean like a tree does, or corn for ethanol production, or...?

The tree would take what you burned and bring it back down because it's working in the dilute context already. Ethanol, or other biofuels, are nominally a closed cycle because you burn what you captured then capture it again. Now it's true corn ethanol is problematic for a variety of reasons (not least of which is use of fertilizer), but the concept of carbon neutrality in biofuels is mostly sound. The tricky part is to do it at scale, and reliably.

This might be an off-tangent, but comments like yours are the main reason why I so much love HN, and why it has become more than 1/3rd of the time I spend on the Internet.

Thanks for investing the time to write this. You gave me so much.

If you enjoyed this sort of first-order quantitative reasoning about energy use and production, can I recommend the book 'Sustainable Energy: Without the Hot Air' by the late Prof David Mackay? It's essentially 300 pages just like the grandparent post, about making an energy policy that adds up, given various bounds of allowable CO2 emissions etc.

Free online, and a the best seller physical book in the field: https://www.withouthotair.com/

Dammit. I didn't know he had died. He was twelve years younger than me.

The book is brilliant, we need more people like him.

Hugh Hunt started a climate lecture series in his honour: https://www.climate-series.eng.cam.ac.uk/

Also the Do The Math blog -- https://dothemath.ucsd.edu -- which does the math of many environmental ideas by a UC San Diego physicist who got his degree at Caltech.

Yes! I've read it a long time ago. Brilliant book.

His comments about the viability of PV are always to be understood to be about the insolation in the UK. Some people like to generalise from them to say PV can't work in Australia or North Africa.

His comments about scale are true, but also aren't entirely complete. You probably can't replace the entire economy we have with lower intensity sources of energy without also having to reduce energy consumption, is his key argument.

But, striking out the worst ones with the best ones we can invent is still interesting.

It's a good read, but it's not holy canon. Parts are contestable and parts are highly contextual and parts are conjecture. Actually.. that makes it quite like holy canon.

It's very specifically not a holy book because it does not hide any reasoning or ask you to take anything on faith. Quite the opposite - all of his working is shown, every assumption is stated, and you are very explicitly given the tools to calculate and update his work as new advances and considerations emerge. There is a whole chapter on not-the-UK, and how the math changes. The reason the book was such a global success is not because it was a statement of opinion about UK energy policy, but a proper treatment in how to thing about energy production and use, the tools for which can be applied anywhere.

...and then the other side of the deal, which is the widespread use of the text as a giant cudgel to belabour people with in an argument: in that situation it's exactly like hitting people with the Bible, wooden cover, locks and all.

Thank you! I'm very happy it pleased you so much :) I second the recommendation of Without Hot Air. It's very unfortunate MacKay won't be able to update it to reflect the developments of the last decade, and the next one.

I also recommend this book heartily - it's amazing. There have been mutterings about crowdsourcing an update, and I think I saw a github repo once but I can no longer find it. Here's an example:

https://www.carboncommentary.com/blog/2017/3/30/l6qcqgoedse1...

Here's the github repo for "Sustainable Energy Without the Hot Air: https://github.com/titojankowski/sustainable_energy_without_...

Would love a new commit of some sort :)

For example he says: It’s possible that floating wind turbines may change the economics of deep offshore wind.

They are indeed now being deployed, as are significantly better generators per turbine. His methodology is sound. His specific numbers are going to have to be adjusted. His overall polemic will probably stand (that high energy density sources like nuclear make more sense but are political)

I'm not an expert in the area, so I could be wrong about this, but I think the steam turbine generators needed to get electricity from heat from coal, nuclear, or CSP sources now cost more per nameplate watt than PV—and, in many places, more per delivered watt. This is a result of PV prices falling by more than a factor of ten since MacKay published his book. If this is correct, then high-density energy sources like nuclear power no longer make sense economically, political factors aside. (Wind and hydro excepted, because they don't need steam turbines.)

From my math I got something around 1 square kilometer of solar panels to power a jet.

Basically one square kilometer is 1 million square meters times 160w/sq meter is 160mw.

Yeah, but there are inefficiencies in the production process and the engine. I conservatively assumed each of these is about an order of magnitude.

10km * 10km = 38 square miles, BTW

In nevada, you have about 3600 sun hours per year, so about 40% of all hours.

So ignoring transmission losses, "reverse fuel cell efficiency", wouldn't we expect an average of 64 MW from each square kilometer of solar panels?

As the French say: Avec des "si" on mettrait Paris en bouteille. (With some "if" one could put Paris inside a bottle.)

All in the name of being opposed to CO2, an essential nutrient of plants and a by-product of industrial/modern productive activity. Billions of humans are and remain alive today thanks to fossil fuel.

You must have intended to respond to a different comment, or perhaps you are a bot triggered by the occurrence of particular keywords.

The only problem with this analysis is the 777 (and all modern aircraft) is designed around a fuel cost of a certain price. Moving to higher cost renewable energy would change the whole design of passenger aircraft.

You can put the solar panel in space and beam concentraded light to the plane in laser form and get much higher watts per sq ft.

No fuel weight should be an advantage. Is there any disadvantage? Would transmission loss be too high? You’d still need batteries for takeoff/landing. But you could charge them at cruising altitude.

1) Prohibitively expensive.

2) Too many easily broken parts and too little fall-back mechanisms. (What do you do if it fails? You have to be able to reach the nearest airport, which, over the big oceans, will be near-impossible without carrying at least half-a-journey full of energy anyway.) (Good luck getting something like this through certification..)

3) NIMBYs will try to stop you from darkening their skies. If you want to take the energy from next to earth, there will be additional energy influx (and you can put a global-warming equivalent on that).

There are many more if I think for more than 10s.

All energy transfer mechanisms double as weapons. Orbiting solar power stations are also known as death rays. This has military implications, to put it mildly.

Beaming to a plane would be difficult, but space solar power for the grid is actually starting to look feasible. I just read a book [1] about it.

Lasers are an option but they're not very efficient, and blocked by clouds. Instead we can use microwaves. The watts per square foot would be just a little higher than sunlight, which is good for safety anyway. Total efficiency from satellite to grid would be about 50%, which isn't bad since it keeps running at full strength at night and on cloudy days, and sunlight is 30% stronger in space to start with.

NASA's monolithic design from the 1970s would have been absurdly expensive even if launch were free. Some recent designs are completely different. The book focuses on SPS-Alpha, a modular design made of many thousands of identical components of eight different types, each only a meter or two wide, which self-assemble in orbit. There's no single point of failure, and the components are cheap since you mass-produce them.

The book estimates a retail electricity price of $0.15/kWh. I plugged in the extremely low launch cost SpaceX is claiming for the BFR, and got a price of $0.045/kWh. That would decrease further over time. U.S. national average is about ten cents.

[1] https://www.amazon.com/Case-Space-Solar-Power-ebook/dp/B00HN...

How long does it take for the panel to produce the amount of energy it took to launch it into space?

An ESA study in 2006 [1] estimated total energy payback including launch and manufacturing at a year or less, depending on various assumptions. It's similar to ground solar; launch energy is offset by the higher and constant solar flux in space.

[1] Peter Ongaro, Leopold Summerer; "Peter Glaser Lecture, Space and a Sustainable 21st Century Energy System"; (57th International Austronautical Congress, 2006)

What about a 50 to 100 passenger, large triangle glider shape PV coated. Medious speed, light motors, no duel and take off energy provided from a ground propeller ?

Taking it further, if need be could something like that be pulled to altitude like a glider, the idea being using liquid fuel only for the the portions of the flight where it has massive advantages.

Specializing into liftoff and cruise is interesting, since requirements are so different. Perhaps one could even lift 5-10 gliders in one go. These can then go off to potentially different destinations.

I could see the feasibility having a strong seasonal and geographic dependency, but the idea seems interesting. The SpaceX vertically landing rockets seem possibly relevant as well.

Surely this whole idea has been discussed before but I don't recall ever coming across it before.

You numbers are off. 45,220 US gal gives 14,260 km range at 905 km/h + emergency reserve. ~= 16 hours of flight time 2830 gal / hour of fuel used. 37 kWh per gallon * ~.40 effecency * 2830 ~42,000kWh.

40% solar panels are available but 23% are low cost. So 230 kWh or 400kWh which means 100x to 180x not 1000x Before considering fuel vs panel weight.

PS: At the other extreme solar can hit 2.4kW/kg which would be worth installing on aircraft to reduce APU fuel consumption.

I was recently thinking about this problem. Why not beam energy to planes (as also mentioned in comments). We could create a network of zeppelins that use/store wind and solar energy and beam them towards planes. As an extra bonus these zeppelins can be used for a multitude of things like internet and weather stations.

The renewable power source for the electricity isn't close enough to where it is needed (hydro-electric), or not stable enough (wind, solar). And the fuel produced can be used for applications other than power generation. And used as a storage medium to even out the bumps in solar/wind.

For that last category, this process needs to be more effective than battery storage.

And if there is a price on carbon pollution that is high enough, that would help make the case.

It wouldn’t help because at best this is carbon neutral. But in reality it’s unlikely to be anywhere close to carbon neutral, in which case a tax on carbon makes this much more costly than battery storage.

And if it was carbon-neutral, it would also be energy neutral, at best (realistically: much worse). But there is a time-shift opportunity in there, capture the CO2 when renewable output is lower than demand, recycle the CO2 into fuel when renewable energy is abundant. Retrofitting existing gas peaker plants into giant flow batteries would be the holy grail of grid scale energy storage.

Except we know that existing fossil fuel burning plants are about 30% efficient. No battery is that bad. This is a really weak idea that only sounds good if you don’t think about the details at all.

(And remember that 30% is only on the recovery side, you’d have to factor in any inefficiencies on the storage side before getting a final efficiency number to compare against other means such as batteries.)

> Except we know that existing fossil fuel burning plants are about 30% efficient

High efficiency gas plants are reaching 60%, and it gets better when you have consumers for waste heat who would otherwise be burning fossils just for heating. Maybe you are be confusing the numbers with well-to-wheels efficiency of electric cars?

This could be useful for developing countries, if the following supply chain became feasible:

1. Rich country builds nuclear reactors to generate power 2. Rich country uses nuclear power to synthesize hydrocarbons for export 3. Developing country imports hydrocarbons for domestic energy needs

Developing countries don’t have the resources to build nuclear reactors, and we might not want to export the nuclear technology anyway, but we can export the actual energy this way.

Because we have convenient gas station network and a lot of combustion engines already. That also have the benefit of a huge range and instant top up.

A tech to transition would be a good thing.

The advantage of fossil fuels in this case is energy storage density, and potentially longer engine life than that afforded by modern batteries.

Rechargeable batteries aren’t currently as mass or volume efficient as these chemical forms.

https://en.m.wikipedia.org/wiki/Energy_density#/media/File%3...

If this actually works and global warming gets bad enough, we could build a bunch of nuclear plants for this, if nothing else.

If you apply this idea, fossil fuels are no longer a source of energy, they are now a way of storing and transporting energy. You could have a bunch of solar panels in Arizona where there is a lot of sun and cheap land and then ship the fuel somewhere else where solar is less efficient.

They are also no longer fossil fuels, just chemically equivalent or at least similar hydrocarbons.

I wonder if language will eventually change to reflect that.

Time... It is all about when you want to use the electricity.

Your suspicion is correct. This is a film flam. At best this is an extremely inefficient way to store energy.

The CO2 is already there, now. Might as well use it.

I think the worry is that it might cost more energy to process the CO2 than it actually produces. In which case the energy would be more of a byproduct of removing the CO2 from the atmosphere, I guess?

No, the worry is that without negative emissions, there's no realistic way to head off worst case climate change. The IPCC scenarios that have us averting disaster already assume that we're going to develop carbon sequestration technology that is beyond what we can currently do. There will inevitably be some inefficiency in the conversion, but even a fairly inefficient process that is still carbon negative is desirable.

And what will you do when the current interglacial ends? the last one --the Eemian-- was much warmer, and it did end.

Or do you believe that this particular interglacial will, somehow, never end?

It seems like we have the technology to warm up the planet within reach already

You believe that you can control the cycles of the changes in orbit and axis of our planet, as well as the cycles of solar activity?

Or do you have evidence for other primary cyclical causes of ice ages, interglacials, and glaciations?

Might?

I’m pretty sure that’s not a question. If they were claiming it would generate more energy than it took, this article would have been about a perpetual motion machine.

Reply on news.ycombinator.com