From the article:

"High-efficiency electrolysis of iron oxide can store as much as 80 percent of your input energy in the iron fuel"

"Using this kind of cyclical process to generate electricity could approach a theoretical efficiency around 40 percent"

Moving fuel around is also a problem with coal and oil. We're basically burning fossil fuels to move fossil fuels around. In the case of oil, we than also have to run it through energy intensive processes to turn it into more palatable fuels than bunker oil, which is what is typically used for shipping things around.

A typical oil tanker consumes a couple of hundreds of tons of bunker fuel per day to move a couple of thousands of tons of fuel around. So, that's a sizable percentage of its load that is lost just to get the raw product from A to B. I'm sure the math is similar for producing and moving coal around.

It only makes sense if you get to pollute at will without financial consequences; which happens to be the case with shipping, which mostly takes place in international waters without any constraints or regulations on the amount of crap put in our atmosphere.

Ships actually switch to cleaner fuels when they get close to populated areas because otherwise they'd end up killing some of the locals with their fumes. Bunker fuel is nasty. Even more so than coal.

So, if we are debating efficiencies, we'd do well to apply the same scrutiny to existing solutions as well. The efficiency of coal should also include digging it up using diesel burning heavy duty equipment, moving it around, etc. I'd bet that knocks some percentage points of even the most efficient plants or ICE cars.

Efficiency matters when you run out of space or resources. But you got vast areas like Sahara where solar power production could be combined with metal production. Wasting space in Sahara is a non-issue if it means you get the overall cost of system down.

This article suggests at top speed Panamax tankers carry 1.5-2mn gallons of fuel while burning 63,000 gallons per day (i.e. around 3-4% of capacity) at top speed.

That is considerably less than the 10% (200 tons to move 2,000 tons) that you suggest.

Also the article notes that reducing speed by 10% can reduce fuel use by a third. The tankers travel at 19mph, significantly less than their top speed of 23-28mph so fuel consumption is likely closer to 1%.

While it certainly makes sense to calculate for the entire journey, if you were talking about the "whole journey" why did you specify "per day"? It's hard to blame 'ximeng' for relying on your written comment rather than what you intended to say. :)

But second law of thermo dynamics applies. If this works at all it is because you are capturing a little more of the energy that was put in at the top and would otherwise be lost. You stil need to put in external energy to keep this system running.

If the process was 100% reversible, and friction free (probably a couple other loss factors I forgot about) it would be energy positive as the oxygen is making its way back up in the air is an external energy input. Of course in the real world both assumptions are wrong by more than enough that you can never get this to work.

You could build the infrastructure to do the electrolysis immediately next to the coal plant to solve this problem. If this technology takes off that's most likely what will happen anyway.

We are not trying to transfer green energy to less windy/less sunny places, the grid can take care of that.

It also seems like it can make use of existing infrastructure and save money there.

At first glance it seems reasonably transportable, but I'm worried about how you prevent it rusting in the air. Maybe it would be useful to blend it with biodiesel for that purpose.

But secondly, wind power makes a lot more sence in northern latitudes.

I imagine methane synthesis will beat some thinglike this out. Pipeline infrastructure exists, is convenient, etc.

Methane seems nice for the re-use of gas fired plants. I do think its gaseous nature has more downsides than upsides on the other two factors. There's also a gas-leak hazard with methane that doesn't exist here.

If the capital requirements are low enough, I could see a 40% efficient energy battery for smoothing out peaks being nicer than more complicated storage schemes.

For fuel that can be transported, methane is a winner. In addition to the distribution infrastructure, there's an awful lot of residential use, with lots of appliances that don't need replacement.

Yes for short term batteries are better but for say seasonal storage or when power is generated far away from where it is used then metal is better.

But it is a great idea to burn any fuel whose oxidized product is not a gas. The only reason why burning hydrocarbons is problematic is because carbon dioxide is a gas, otherwise it could be captured and recycled like this.

If you wanted to convert that heat back to electricity, there would be losses, but nowhere near as much as some people are saying. 1800 degrees minus ambient is a very big delta T. You are not bound to a closed cycle, so Carnot does not apply. So it is likely 80+% would be achievable, maybe over 65% round trip.

Burning hydrocarbons is problematic for a host of reasons besides CO2 release: fracking, oil spills, release of other pollutants, political issues with the supplier countries, pipelines, ...

Emphasis mine. Don't be a pedant.

70% vs 40% is a big difference, especially when you consider it from the losses column instead of the wins column. If I gave you a system that was losing 30% of a resource to replace a system that lost 60%, you've cut your losses in half.

If I told you a mechanical system could get that to 15%, you'd replace that again. Anyone who tells you they can get a heat engine down to 15% losses is a charlatan and should be reported for fraud.

What I haven't seen is many people doing both at the same time. I might have seen a desal unit powered by green energy, which tops up the water towers when power is cheap, but not much beyond that.