Toyota says that the Mirai fuel tanks hold 122.6 litres. But hydrogen fuel is sold in $/kg. Typically pressure is 70Mpa or 10kpsi. So how many kilograms of H2 fuel does the Mirai hold?

I don't think that is right. How did you arrive at that figure? I read in one of the articles that it was 5kg based on converting from the EPA MPGe figure, but MPGe is grossly exaggerated to promote EVs and FCVs it is not real. I believe the answer is about 7kg but I have no proof.

I think I got it. One mole of H2 is 2.01 grams. 1 mole of gas at STP is 22.4 litres. A Mirai tank is 122.4 litres. 10kpsi = 680.48 atmospheres so tank is equivalent to 680.48 * 122.4 litres = 83,290 litres at STP. Divide that by 22.4 litres per mole = 3,718 moles at 2.01 g/mol = 7.47 kg. Now for some fun facts. What is MPGe? How do you define equivalency? I define it on $ that is equivalent cost. EPA says the Mirai has a MPGe of 66 and if you take gas as $3.25/gal and Mirai range as 312 miles then you get $15.36 to fill up. At 7.47 kg that works out to $2.06 per kg H2. If you do this backwards, that is assume H2 is $15/kg then the MPGe is 9.05! Not EPA's 66.

If my understanding is correct, the answer is 10 kg. Quoting from Wikipedia's article on the Mirai: The Mirai has two hydrogen tanks with a three-layer structure made of carbon fiber-reinforced plastic consisting of nylon 6 from Ube Industries and other materials. The tanks store hydrogen at 70 MPa (10,000 psi). The tanks have a combined weight 87.5 kg (193 lb). and 5 kg capacity. Two tanks at 5 kg capacity apiece = 10 kg.

Thank you for looking this up. But two things, one the line states that the combined weight and 5kg. So why would you assume the word "combined" applies only to the weight? Second although there is a citation for the weight, there is no citation for the 5kg. I think it is just a common mis-conception. Simply put I believe in science and according to my calculations it comes out to 7.5kg any other answer would say scientists don't know what they are doing, i.e. the definition of a mole is wrong. I will add that in practice it may turn out that you cannot fill the tanks to their theoretical capacity, because temperature rises as the gas is compressed.

I did spend some time on Wikipedia trying to understand EPA's formula for MPGe. It is not clear, they give what appears to be conversion factors that come out of nowhere and go to some length to explain that the MPGe is based on equivalent energy, not equivalent money. In round figures 1 gallon of gasoline has the same energy as 1kg of H2 according to them. This makes good sense when you realize that the Mira gets somewhere between 46 and 66 miles/kg (depending on 7.5kg capacity or 5kg), and it is easy to relate this to a gas car getting 46mpg. However what is ignored is the price difference between a gallon of gasoline and a kg of H2.

I will concede that the practical limit of the Mirai H2 capacity is likely 5kg based on the fact that I am sure you cannot drive the car until the tank is completely empty. There is probably some minimum pressure below which the fuel cell will not work and as stated above temperature rise will also limit amount of H2 it is possible to load.

Well okay, I admit I was trying to fit what it said in the Wikipedia article into what I thought I had read elsewhere, which is that the Mirai holds 10 kg of compressed H2. But exercising my Google-fu, it looks like it really is just 5 kg. Here's a report by someone actually filling up a Mirai. The fuel capacity is 5 kg. And dang! that means the Mirai's range is only half of what I thought it was. This fool cell car is even more foolish than I realized!

Thanks. Although I still stand by my calculations it appears that the practical (useable) limit is 5kg. I did spend a little time trying to find out the minimum pressure, but only found out that there is a pressure regulator (reducer) between tank and fuel cell as expected but no value on what it reduces to. Also the incoming air required to get oxygen has a compressor which I assume brings the oxygen supply up to the same pressure as the regulated H2. In any case it is difficult to imagine these pressures are very high, that is it would require a minimum pressure of 10k/(7.5/(7.5-5))= 3,333 psi to support my theory. An admittedly unrealistic minimum pressure so I still don't have a reasonable explanation for the discrepancy.

While the Mirai's specs list ~122.4 liters ( ~ 7.47 kg per OP's calculations ), real world fill up is actually 5 kg which is about 67% (5/7.47). That's in line with my 2015 Toyota Sienna having a 20 gallon capacity but real world fill up from empty (with light on) is about 15.5 gallons. I've only did that fillup from empty once to test. I usually fillup when it gets down close 1/4 tank.