Heat from solar for a hydrogen production system working like a conveyor belt
In solar reactors that use solar-heated thermo-chemistry to produce hydrogen from water, heat recovery is needed to increase efficiency and, thus, lower costs to replace today’s fossil-fueled thermo-chemical process to make hydrogen.
Ceria is state of the art for this redox process and has two main steps: Reduction and Oxidation. In the first (Reduction) step, ceria (CeO2) is heated, causing it to release its oxygen. Then, at a lower temperature, water is added, and the ceria reabsorbs the oxygen from the water, leaving hydrogen, which can be removed. The ceria is heated again to restart the cycle. Each cycle lasts 40-60 minutes, and by running it repeatedly, hydrogen is extracted each time.
Researchers have been working on ways to recover this waste heat at ETH in Switzerland, at DLR in Germany, and other international research centers to improve the efficiency of the process.
Now, MIT researchers have come up with another heat recovery idea—passing heat between a revolving train of reactors that showed in modeling that it was able to recover 70% of the waste heat, while increasing the heat-to-hydrogen efficiency (what fraction of heat that you put into the system comes out as hydrogen) to 40%, much higher than today’s 10%.
To generate the heat at over 1,000°C to run solar thermo-chemistry, a solar field of mirrors focuses sunlight up to a solar receiver on a tower (either surrounding the tower as here, or in just front)
In this new research field, Solar Thermo Chemical Hydrogen (STCH), where concentrated solar heat is used to make hydrogen, a solar receiver atop the tower is heated by a solar field of heliostats, concentrating the sunlight onto a heat transfer fluid in the receiver. The heat is typically stored to provide steady heat for the reactor where the thermochemistry occurs.
“The most promising system is where we have a solar receiver and store that energy thermally so that it could be either solid particles or molten metal because of the very high temperatures. Then, we use that stored heat to heat the reactors,” said solar researcher Aniket Patankar in a call from MIT in Massachusetts.
“Doing this with nuclear would not be ideal because of the temperatures we need, and with fossil-fuel-based systems, we have the whole emissions issue. So, concentrated solar thermal energy is probably the best option when looking at renewable sources. As a backup, we have PV or wind, we can store that electricity, but that is less preferred because you’re taking electricity and converting it to heat, which is less efficient. So our primary focus at MIT is concentrated solar.”
Typically, solar researchers in hydrogen place one reactor adjacent to the solar receiver on the tower. Patankar’s MIT team has devised a design that deviates from this more typical single solar reactor on the tower, which they described in A Reactor Train System for Efficient Solar Thermochemical Fuel Production. Instead, they propose a train of reactors circling the tower base at ground level, like an airport baggage handling roundabout.
“You can think of it like a belt that these reactors sit on,” lead author Patankar explained.
“The conveyor belt transfers them around that loop. It’s curved, a circular conveyor belt. Each reactor has a gas exchange port at the back. These reactors are not moving continuously. Every minute, the conveyor belt moves the reactors one step ahead. Then the conveyor belt moves the next one into position and repeats the exchange.”
During each of these rest stops, on the hot side, the oxidation zone, the oxygen is pulled out – while on the cool side, the reduction zone -the water is injected in and the hydrogen pulled out.
“In some technologies like photocatalytic water-splitting, the oxygen and hydrogen come out in the same gas stream, and there is a risk of them recombining to form water. We do not have that risk here,” he added.
Heat solar hydrogen: This single pipe seen on the cold side on the right has a dual function. It is actually two concentric pipes, one inside the other. The outer part supplies the H₂ O (as steam) into the reactor, while the inner pipe is the outlet where the hydrogen gas comes out. On the hot side, there is no inlet gas, just oxygen coming out of the reactor.
The thermochemical reaction itself also generates some heat, but the main source of heat is the hot reactors themselves as they exit the reduction zone. They must be cooled down before they reach the oxidation zone. The heat they lose as they cool down gets transferred radiatively to the other set of reactors on the opposite side.
“It is radiative transfer, a no-contact method of transferring heat,” Patankar noted.
“The heat transfers from the porous material in the first reactor window, then through the window of the second reactor and into its porous material, into the ceria. At one end, where they curve, we’ll have the hot zone that puts in the high-temperature heat at 1500°C. The reactors on the hot side will be releasing oxygen from the ceria. As they move along the flat section, they cool down till at the other end. Once the reactors are cool enough (800°C), they release the hydrogen.”
“The hot reactors are moving left to right, the cold reactors are moving right to left, so they’re moving in opposite directions and exchanging heat from one side to the other,” he added.
The team had previously presented this train idea to improve the efficiency of this hydrogen production at a Solar PACES conference.
Now, to estimate the efficiency of the heat recovery idea, they have devised a new software method which sped up the modeling to make this analysis possible; GREENER. Their new tool overcomes obstacles with traditional methods of capturing this data that are either too slow or fast enough but less accurate. The team applied the GREENER method to study the performance with varying ceria cavity thickness and cycle times.
Researchers in this field of solar hydrogen work on improving the radiative heat transfer in the ceria cavities while reducing temperature differences and maximizing hydrogen production. The two traditional methods of analyzing the efficiency are Monte Carlo Ray Tracing (MCRT), which is accurate but slow, and RDA (Rosseland Diffusion Approximation), which is fast but less accurate. They say GREENER combines the accuracy of the slower MCRT method with the speed of RDA.
“If you make the ceria thicker, then more of the active material can fit in each reactor. But there’s a trade-off with the ceria thickness,” he explained.
So, one of the calculations they did was to determine the optimal thickness of the ceria.
“If you make the ceria material thicker,” he said, “then the positive side is that you can theoretically get more hydrogen from a single reactor. So that’s a good thing, but the challenge is that you start developing temperature gradients within that ceria. It absorbs heat through the window; only that portion of the ceria facing the window gets hot, and the back of it remains cold. So that reduces the amount of hydrogen you can make because the back portion of the ceria doesn’t quite get as hot as the front. So we wanted to find where the optimal thickness is.”
The team could calculate the efficiency and the temperature gradients for multiple thicknesses because the computation was fast.
The geometry of each reactor is the same —like the thickness of the porous ceria structure—and doesn’t change much over time, so they can be calculated once for all of them, no matter how many there are.
“The optimal system will have tens of these reactors, and so that’s a big problem to solve,” he explained.
“So if you can calculate that for a single reactor at a single time, so there’s no time dependence, and apply that for all reactors at all times during the cycle, that can reduce computation very significantly, by several orders of magnitude, while maintaining the same accuracy as the Monte Carlo ray tracing method.”
So far, the work has been done in simulations. The team is now building a demo. If they can show that the demo validates the heat-to-fuel efficiency predicted by GREENER, it will mark a step forward for solar thermochemical hydrogen.
“Twenty percent is where it starts becoming commercially viable, at least compared to other sources of green hydrogen,” he pointed out.
“That is just with the heat recovery aspect. If you do better oxygen removal and better steam hydration separation, it could be pushed to 30%, but just as with the heat recovery concept, we feel it can go to 20 to 25%. Existing systems that don’t include heat recovery – their efficiency is less than 10%, so the heat recovery was the biggest piece of the puzzle, or like the biggest energy loss, and we are solving for that, counter flow, heat exchange, and then get up to above 20% then, yeah, there are a few more knobs we can turn to get to even higher efficiencies.”
