Video Url

Below is the text version of the webinar titled "Highly Efficient Solar Thermochemical Reaction Systems," originally presented on January 13, 2015. In addition to this text version of the audio, you can access the presentation slides.

Amit Talapatra:
Hello, everyone, and thanks for joining today's webinar. Today's webinar is being recorded, so a recording, along with slides, will be posted to our website in about ten days. We will send out an email once these are posted to our website.

[Slide 2]

Everyone in this webinar is on mute, so please submit your questions via the question function, and we'll cover those during the Q&A at the end of the presentation. Since we have multiple speakers today, please indicate who your question is for when submitting your questions, and please check back to our website for future webinars, as we host them monthly.

Also, I encourage everyone to sign up for our newsletter that we send out monthly. You can sign up for that on our website. And with that, I'd like to hand it off to Eric Miller. He is the program manager for the hydrogen production and delivery team in the Fuel Cell Technologies Office. Eric?

[Slide 3]

Eric Miller:
Thanks, Amit. Got it. We are fortunate to be joined by Robert Wegeng and Jamie Holladay of the Pacific Northwest National Laboratory to tell us about the STARS project, which received the prestigious R&D 100 Award in 2014. Developed jointly by Barr Engineering, Diver Solar LLC, Oregon State University, and PNNL, the Solar Thermochemical Advanced Reactor System, or STARS, converts natural gas and sunlight into electricity—actually, into syngas, which then can be burned to make electricity in power plants.

STARS has actually set a world record with 69% solar-to-chemical energy conversion, which has earned it the 2014 R&D 100 Award thanks to its technical significance. So with that, I'd like to hand it over to Bob to hear the details of the project. Do we have Bob? Check the mute button.

Robert Wegeng:
OK, this is Bob Wegeng. Can you hear me now?

Eric Miller:
Now, we can. Great. Take it away, Bob.

Robert Wegeng:
OK. So this is Bob Wegeng. As Eric said, I work at the Pacific Northwest National Laboratory, which is up in Richland, Washington. I'm happy to be here and thank you for the kind introduction. I think what I'll do is kind of jump right in to what the presentation and this webinar is about.

The system I'll be describing efficiently converts solar energy to chemical energy. It employs advanced micro-channel and meso-channel process technology. Those are heat exchangers and reactors that are very process intensive and quite compact that allow us to take concentrated solar energy from a solar concentrator, like a parabolic dish concentrator, to drive a high-temperature endothermic reaction, therefore running the reaction and adding to the chemical energy content of the stream so that the resulting product stream has, both some original chemical energy in it, plus some solar augment to it. So let's go on to the next slide.

[Slide 4]

So before I get into the real technical details, I want to first acknowledge some things and to give thanks, first, to the DOE Fuel Cell Technologies Office, FCTO. Their support in the 1990s and also in the early 2000s for the initial development at PNNL of micro- and meso-channel process technology allowed us to build our first individual components and our first prototype systems that included the micro-channel reactors and heat exchangers.

They were, at the time, being pushed in a direction for possible use within automobiles, for fuel-cell powered automobiles, as a way to convert gasoline to hydrogen for the fuel cell. Obviously, that's an application where you need the system to be very efficient and very compact, and more so, it needs to be something that can be mass produced.

I want to thank the DOE Solar Energy Technologies Office. They saw a value in this for concentrated solar power production, and so they have funded this, now, over the last three and a half years to work to adapt what we had done for FCTO, to apply it to the concentrated solar power case. And I also want to thank FCTO for the fact that they've organized this webinar. It gives us a great opportunity to do some outreach and communicate what we've done and what we intend to do. So let's go on to the next slide.

[Slide 5]

So then, I want to also acknowledge the project team. Eric already mentioned the majority of them. Here at PNNL, the researchers here; Diver Solar, which some of you may know is Rich Diver, a retired employee from Sandia, who's a consultant to us on the project; Barr Engineering; Infinia Technology Corporation; and Oregon State. Plus, recently, we've added Southern California Gas Company to our team of collaborators. So in the next slide, I want to get at sort of the gist of it.

[Slide 6]

If you think about solar energy being converted to chemical energy, a quick way to think about it is to be thinking about what photosynthesis does. And in photosynthesis with algae or with corn, you end up with a few percent gross conversion of solar energy to chemical energy. That's one route through photosynthesis.

What we have been working is a thermochemical route that includes catalytic reactions—no photocatalysts though, all just thermochemistry—where we operate at high temperatures, like with heat engines. High efficiencies can be obtained when you get to high enough temperatures, and the result is, as the title says, our system has accomplished a conversion of solar energy to chemical energy with an efficiency of 69 percent. That was the system that we operated in the year 2013, and it's pictured here on this slide.

I like to talk about different technology readiness levels as a ladder that you climb as you go from components to low-fidelity systems to high-fidelity systems and then to systems that are designed for manufacturing, etc. And so we were, at that time, at what's called technology readiness level 4. That's a fairly low-fidelity system but with individual components that are designed for the task.

And I want to also comment that when I talk about solar-to-chemical energy conversion efficiency, the metric I'm using is the increase in the higher heating value of the stream divided by the direct solar energy that hit the parabolic dish concentrator. So that's the metric that I'm using. So I'm including the dish concentrator and everything else in the system as part of the calculation.

Now in the upper left of this, you see the chemical reactor that we built for that system. It's about a foot in diameter and a couple of inches thick. In that, we have meso-channels, channels that are in the neighborhood of five to ten millimeters thick, in which we have loaded catalysts. That's the unit that sits near the front of this item on the right. It sits near the front and receives the concentrated solar energy from the parabolic dish concentrator.

So the image on the right is the solar nacelle. That's where we load the reactor into. On the left side of the image, you can just sort of see the opening, about a three-inch-diameter opening, that receives the sunlight from the solar concentrator. The reactor is just behind that, and then past that are the individual heat exchangers, one of which is shown in the left, bottom image here.

That's a micro-channel counter flow recuperative heat exchanger made out of Inconel. It allows us to preheat the reactants before they go into the reactor so that they're already hot. If we get them hot enough, then there's very little for the solar energy to do other than run chemical reaction, and that is one of the tricks—or the secrets or whatever you want to call it, the special sauce that makes this system so efficient—to do a great job at preheating first.

We then take the products from the reactor and run them counter flow through the same heat exchanger network. So we're cooling the stream down using the heat from that to do the preheating. And then that's the other part —great recuperation in very compact devices is what you have to do to make the system work well.

So that system on the right, that overall nacelle, which was supplied by Infinia, that nacelle is about four feet long. So that gives you an idea of what we have to mount at the focal point of the dish concentrator. So let's go to the next slide.

[Slide 7]

On this next slide, we show in the center, within that box, the level zero process diagram. It just basically says that thermochemical reaction system receiving radiant energy, concentrated energy, from solar concentrators.

And I should mention, at that top portion in the circle, in the center, we show an example parabolic dish concentrator. There are some other types of concentrators that could also be employed to do this kind of operation. The left circle at the top is a central receiver. We'll show more of those in a little bit. The right is a beam-down tower. Those can give you high enough concentration ratios to reach the temperatures we need for our chemical reaction.

And the chemical reaction is shown just under the thermochemical reaction system box, methane steam reforming. So we bring in methane, either from natural gas or from biogas, from biomass. We bring that in and react it with water, as the reaction shows. And this is the idealized version of the reactor, showing the production of carbon monoxide and hydrogen as the energy carriers in the product stream.

Chemical engineers will know that to push the reaction far you want to come in with a little extra water, and that will help you get more conversion of methane. It will also result in a little bit of carbon dioxide production, a little less carbon monoxide and a little more hydrogen too. So the realistic reaction ends up with some CO2, a little bit of unreactive methane, and some unreactive water on the right-hand side.

But the main point is, by using this radiant energy, we increase the energy content of the product gas, what we call synthesis gas, by about 25 to 28 percent. That means, in the end, that somewhere around one-fifth of the energy in the stream is solar—came from the solar. The other four-fifths came from the original methane, which, again, if it came from natural gas, then it was a fossil fuel. If it came from biogas, then it was a renewable energy content completely.

The syngas we produce could be burned, for example in a power plant, in a conventional power plant, to make electricity, and that is one of the desires of the DOE solar program—to make electrical power from solar energy at a cost no more than six cents per kilowatt-hour levelized cost of electricity over the lifetime of the plant.

It could be consumed in a fuel cell as well, for power generation, or the syngas could be used to make other chemicals. And in fact, industry uses syngas for that purpose for many different commodity chemicals, or it can be used to make hydrogen. And of course, the interest of the hydrogen program is hydrogen as a transportation fuel at no more than about $2.00 per GGE. So let's go on to the next slide.

[Slide 8]

So, so far, I have told you sort of bottom line, the efficiency we gained and what some of the applications may be. And the rest of the presentation, I'll talk some—about some background on concentrating solar power. And I'll talk a little bit about what's been done in the past, what other workers have done to demonstrate and to develop solar thermochemical process technology.

Then we'll talk a little bit about the application areas for the system that we are developing. I'll talk a little bit about the core technology that allows us to do this, the micro- and meso-channel process technology. Then, I'll talk a little bit more about the performance we gained in 2013 from the technology readiness level 4 system.

And then I'll talk about the work we've done in the last year to improve the system, advancing it another level on the TRL ladder, to TRL 5. And I'll show some preliminary performance data from that. And then we'll talk a little bit about the plans for what would be the next reactor improvement and some conclusions. So let's go on to the next slide.

[Slide 9]

So on this slide, we're going to talk a little bit about concentrating solar power, and we're going to start off with what is, more or less, a conventional application. In the left-hand image, upper-left image, in Nevada, that is a solar concentrator power plant that was commissioned in the year 2007. My recollection is that it's about 80 megawatts of power output from that system.

They use parabolic trough concentrators, and you can see those in an array, line concentrators in the image, and toward the back of it is the steam power plant. So they basically use the parabolic troughs to heat the heat transfer oil, which could be stored or used right away to make steam. At any rate, it's used to make steam to run a conventional power plant.

The upper-right image is another architecture. This is the central receiver architecture. That's a unit that's in Spain that produces about ten megawatts of power. It employs a number of heliostat mirrors. You can see all those little rectangles throughout it, which track the sun and reflect the sunlight to a location near the top of the central tower.

There, a molten salt is heated, and then the salts can be stored down below in an insulated tank or used at any point to make steam. So this allows you to make steam when the sun is shining for the steam turbine or make steam also when the sun isn't shining for the steam turbine because it's not very expensive to store that molten salt. So this is another power system.

The image across the bottom of the screen is yet another architecture that was pioneered by a couple of companies in the United States, one of them being SES Corporation—Stirling Energy Systems—and the other one being Infinia Corporation. Each of them—and this image, by the way, is from Spain, as well. And it is an array of parabolic dish concentrators of the Infinia type.

Those units intercept about 13 kilowatts of solar energy each and apply it to a nacelle that has a heat engine, a Stirling cycle heat engine, within it which converts the solar energy to power, then, at about a net 25 percent efficiency. SES Corporation and Infinia Corporation both had the same kinds of problems that new companies often have moving forward to get to a commercial point.

And basically, things like the low-cost photovoltaics that came out and made it very difficult for them. Neither of those companies have survived, though Infinia Technology Corporation, a spinoff from Infinia, does still exist, and they are a partner on our team. So let's go to the next slide.

[Slide 10]

So now we're seeing a recent CSP power plant that was commissioned just in the last—just a little over the last year in Southern California in the Mojave Desert. This system has three central receivers with the heliostat mirrors. As we list here, over 170,000 individual heliostats, each with a motor and with two mirrors, go into the system of three central receivers.

The system, altogether, puts out nearly 400 megawatts of power. And so this is a fairly recent operation down in California, and it's part of the wave to come that shows that CSP power really has a strong future for power generation, particularly in areas where there's a good strong solar resource. So let's move on to the next slide, where we'll talk a little bit about work that's been done by others to apply solar concentrators toward chemical reactions.

[Slide 11]

So just as you can heat up fluid to make steam or to drive a Stirling cycle heat engine, you can also heat it for endothermic reactions, and this slide lists a little bit of work that was done, started in the 1980s and in the 1990s and the 2000s, to start to do that kind of work.

Sandia National Lab has done quite a bit of research in the area. There's been research in Israel, and in German and Spain as well, and the bullet that I've got at the bottom left of this relates to a collaboration in the early '90s between Sandia and the German Aerospace national lab—I'm not going to try to pronounce it but the acronym is DLR—where they applied a solar reformer that used CO2 rather than water as the oxidizer to a system and obtained a pretty good chemical efficiency, 54 percent, in that application. Let's go to the next slide, where we can see a picture of that system.

[Slide 12]

So this slide shows the 150-kilowatt solar dish system in Germany, which the Germans supplied, and it shows the catalytic reactor at the focal point that Sandia developed and supplied. And again, this system obtained a very reasonable solar-to-chemical energy conversion efficiency in the early '90s doing what we call "dry reforming" CO2 with methane.

The system, again, we've been developing is wet reforming. We use steam as opposed to CO2, and that will give you more hydrogen. And it'll also give you some other things that are different. So let's go on to the next slide.

[Slide 13]

And that one will talk more about what we're doing, but I'm going to specifically go in the direction of how it could be used to make power using a conventional heat engine. So in the image on the right, in the upper portion of that, we show solar energy coming into a parabolic dish concentrator, which reflects it to a focal point—and the focal point in this is just a cartoon.

The focal point shows the thermochemical reactor, and it also shows one of the recuperative heat exchangers. It shows that we bring in methane and water to the nacelle, preheat it in recuperative heat exchange mode, then run the chemical reaction using the solar energy, then take the products, run it backwards through the heat exchange network, taking that syngas, then, to a combustion system. This could be a very conventional power plant. It could be a combustion turbine, gas turbine-type system, or a combined cycle power plant, which is part of what we are targeting in our application to make electricity.

Because we've increased the chemical energy content, the fuel value of that stream by about 25 percent, we end up with more power generation. We did not increase the amount of carbon in the stream. So since we need less fuel, now, for power generation, we have lower CO2 emissions.

In the image, lower portion of that, we show the application when the sun isn't shining. When the sun isn't shining, just bringing methane into the combustion turbine allows you to make power. So this system allows you to make power around the clock. The power plant is always available for operation, so you can end up with high-capacity factors.

By operating with a very efficient concentrator, such as we are, and a very efficient reaction system combined with a very efficient power system, you have the opportunity for efficient solar electric conversion and the opportunity for low costs. And again, we're targeting six cents per kilowatt-hour for power generation.

Now, I should mention that high-efficiency power generation doesn't have to be done through a heat engine. Fuel cells are another option, and in fact we're very interested in using the syngas in either a high temperature or a low temperature fuel cell for power generation. And we have some work where we're examining that too. Now, again, the net thermochemical reaction from our system is shown at the bottom left. Let's go on to the next slide and look at the option where we make hydrogen for either a PEM fuel cell or some other application.

[Slide 14]

In that case, you would like to really reduce the amount of carbon monoxide in the stream. So we here, we show a case where we bring the solar energy into the solar methane reforming, the SMR reactor in the top. That operates at about 750 to 800 degrees C or a little bit hotter.

And after it leaves that solar methane reformer, we've again cooled the stream and then would run it to a water-gas-shift reactor, which would be operating in the 410 to 275 degrees C range. Now, I've listed it that way, by the way, because the most efficient way to run water-gas-shift would be to start it at a high temperature, 410 degrees C, and as the slightly exothermic reaction proceeds, to be cooling it as you do it.

Starting at the high temperature means fast kinetics, and that allows you then to have a smaller reactor at the high temperature end. As you cool the stream, you would be always keeping the reaction near chemical equilibrium. And finishing at about 275 C therefore allows you to operate that reaction.

The net reaction—again, this is the idealized version—at 275 C, there would still be a little bit of CO in the stream. But the net reaction is shown down below, and that is a suitable place, then, to use something like pressure swing separations to pull out the hydrogen.

To make the system even more efficient, you would use something like the heat from the water-gas-shift reactor to support the water vaporization. There's not quite enough energy in that to really vaporize all the water, but to get the most efficient system, you'd want to do what you could.

Now, conventional methods that use natural gas to produce hydrogen produce about 11 kilowatts of CO2 that has to be emitted per kilogram of hydrogen produced. But they do that because they have to burn hydrocarbon to drive the solar methane or drive the methane-reforming reaction.

Since we're using solar energy, we don't have to do that. It allows us to optimize the system in another way, and the result is we get pretty much the theoretical minimum amount of CO2 in the process stream. Mainly, from 11 kilowatts we intend to drive it down to about 5.5 kilowatts of CO2 per kilogram of hydrogen. So that's a significant reduction in CO2 associated with hydrogen from methane. Let's go on to the next slide.

[Slide 15]

So on this slide, I'll bring up one more item. I had mentioned before that syngas can be used to make lots of different chemicals. One that is of particular interest because it's easy to make and because it's very easy to store is methanol. The production of solar methanol would enable us in a power plant setting to take some of this solar energy and store it as methanol as renewable energy storage—thermochemical energy storage.

And as I said, methanol is very easy to store and easy to make. And so in the system I show on the left, I show it's coming out of the solar reformer, going through that high-temperature recuperative heat exchanger, and then going to a methanol synthesis subsystem, which would convert a portion of the syngas stream—a portion of that, then, to methanol. And a portion of it would remain syngas for a power plant operation.

Again, the methanol would be easy to store, and the heat from methanol reactor can be used to make the steam. And in that case, there is enough heat to make all the steam that we need in the system.

And methanol is a commodity chemical. It's sold worldwide. The process we're talking about, as I said, makes it very easy to make methanol and then store it while the sun's shining. It could then be used when the sun isn't shining to drive the heat engine or the fuel cell-type power plants. So this enables very inexpensive thermochemical energy storage of the solar energy.

And as we've talked about here, there's also a reduction in CO2 emissions compared to classical methods of making methanol from natural gas. So I wanted to comment that if you're making methanol from natural gas today in a conventional approach, again, you have to burn some hydrocarbons to drive the reformer. And a typical methanol production plant emits about 25 to 40 percent of the incoming carbon as CO2. In our system, you could reduce that to very close to zero percent.

And then finally, that methanol we're talking about, because we didn't emit carbon while we were producing it, that allows us to have the lowest life-cycle carbon-intensive methanol available. And that could be valuable for applications like in states that apply a low-carbon fuel standard that need to have an additive to put in with gasoline so that the gasoline has a net lower carbon life-cycle emission. Let's go on to the next slide, then.

[Slide 16]

So in this slide, I talk a little bit about the core technology that we're developing at PNNL that, again, received support initially from FCTO and now it's being applied to the solar applications. It's been developed at PNNL since really the early 1990s. What I'm showing in the right side, in the large picture on the right on the upper side, is a micro-channel reactor from the late '90s—I'm sorry, a micro-channel heat exchanger from the late '90s.

The individual channels that you see are about 250 microns wide, so about the thickness of human hair. That allows heat transfer to occur really quickly between the metal walls and the fluid extremely rapidly, so we end up with rapid heat transfer in short channels as opposed to longer channels or longer piping within conventional heat exchangers. That makes the system very process intensive.

It's both heat and mass transport that are rapid, and so we can apply this to catalytic chemical reactors as well. And because the systems were designed with mass production in mind, we get the opportunity for exploiting that, economies of mass production, as opposed to classical chemical process technologies that would exploit economies of scale.

The image on the bottom right was our first counter flow micro-channel recuperative heat exchanger, where we interwove the channels for the hot gas and the cold gas, again going counter flow in a really compact device that provided a kilowatt's worth of heat transfer, high heat transfer effectiveness. So let's go to the next slide.

[Slide 17]

The next slide illustrates some of the work that we did for FCTO. The image on the left was a combustor evaporator for the onboard vehicle fuel cell application. It would take the anode off gas from the fuel cell that still had about six to eight percent hydrogen in it, would combust it so that we could vaporize gasoline for the fuel processor. And it won an R&D 100 Award in the year 1999.

The system in the center picture is a system that we built in the year 2000 for the hydrogen program. This one employed four micro-channel steam-reforming reactors and over 20 micro-channel heat exchangers, again for preheating and cooling down the streams, and it had an exergetic efficiency. For those of you that are, I'll say, that are thermodynamics aficionados, you know that exergetic efficiency relates to the potential work that could be created. So the exergetic efficiency of this network of heat exchangers and reactors was 85 percent, which is quite high and was very pleasing to us.

The image on the right shows the flow sheet in terms of an exergy balance. So the width of the channels—I mean of the individual items within this—shows the chemical and the physical exergy in the stream. I'm not going to try to talk through this. I'm sure it's too small on your screen, but I will say that we use the second law analysis to help us plan out the flow sheets.

It's very convenient to identify exergy destruction on a component-by-component basis, and then ask yourself, "Why is it this amount and this component, etc.?” in designing a flow sheet. It allows you to see, quantitatively, where efficiency penalties will occur and then allows you to say, "What can I do about it?" And that's how we ended up with a high exergetic efficiency here, and it's how we ended up with a high exergetic efficiency within our solar system. So let's go on to the solar system, and in the next slide, let's look at the test setup that we have.

[Slide 18]

So we have a solar concentrator test stand at PNNL. We'll see it in a minute. This shows the basic P&ID-type diagram—simplified version—for that test stand with the contents of the solar nacelle inside the dotted-line box. So that's the reactor and the heat exchangers.

And what's outside of it—let's look to the upper right, you see the parabolic dish concentrator. It delivers about ten kilowatts of solar energy into the reacting stream within our reactors. So you see the reactor. It says "reactor/receiver" because it also is the receiver for the solar energy. That's where the chemical reaction occurs, but let's go to the left and look to see where the methane comes in first.

So at extreme left, we show where we bring methane coming in. This shows it at 120 SLM. We bring the methane in, and then it's preheated first within what is called the "low-temperature recuperator." That's LTR-M for methane preheating. You see those words and that item just a little bit under the words solar nacelle within this diagram.

There, it's mixed with steam. We get the steam that's water that's brought in at the pump at the bottom of the system. We preheat it, going through a low-temperature recuperative heat exchanger, water preheater. Then, for this test setup, we often employ electrical resistance heat for vaporizing the steam, though on a real application, we'll use another source of heat for that.

We mix our steam, as I said, with methane. Together, they flow into the high-temperature recuperative heat exchanger that brings the temperature up in the neighborhood of 700 degrees or hotter. We then get into the reactor, flowing back down through the network again.

And at the bottom right, you see the point where we separate the condensed water from the syngas. The syngas currently is flared off, but this is the stream that you would take for power generation or hydrogen production, and that basically shows what the test setup is. So the way to think about the things I'm going to show you next are all in the context of test reactors and test systems. So let's move on to the next slide, and let's jump right into the results from that set of tests in year 2013 for the TRL 4 reactor system.

[Slide 19]

So the left image on this shows the—basically information about the efficiency of that reaction system. On the x-axis it says, "receiver energy input." This is the energy that came into the nacelle through the aperture, and it also includes any energy that got lost by the nacelle. So this could be by re-radiation off of the reactor surface. It could be by convection to the outside air. It could be by conduction through the structure, but the real point is this is the energy that comes into the nacelle from the dish.

The y-axis is the energy that goes into the reacting fluid. So that's both for the heated reaction, and it's also for any temperature change or any physical enthalpy change in the stream.

So we're graphing this across testing, again during 2013, where we varied the solar energy into the nacelle through the use of a screen. So a convenient way to do this is to point at the sun when the sun's very intense. Use a meter to measure what the direct normal solar is, but put a screen in the way. And so with different screens, we're able to limit—with different calibrated screens, we limit the solar energy into the system.

This allows us, then, to plot the information. As you see, we plotted on a line. The x-intercept that says 2 kilowatts are the fixed losses of our nacelle system, and the slope of the line relates to variable losses. At higher solar flux, the losses will change again because of things like convection and re-radiation.

And so what we're looking at at two kilowatts is a system that clearly has a fixed loss of two kilowatts energy. So of the 11 kilowatts brought in, 2 kilowatts are being lost by the system, not going into the reacting stream. Now that ends up affecting the overall solar-to-chemical energy conversion efficiency.

So on the right graph, we've got those points plotted, so taking into account the 2-kilowatt loss and also taking into account variable loss. But actually, these measurements are made by the GC measuring what the change and the chemical energy cuts into the stream is and then relating that to the higher heating value.

So on the x-axis, we've got the solar energy incident upon the dish now, and the y-axis, we've got the efficiency value, which again is the ratio of higher heating value increase to the direct solar that hit the dish. And you can see from this something that is a fairly happy observation, that at fairly low solar concentrations, we are still able to get fairly high efficiencies. The highest concentrations occur both when we have the highest solar energy—I'm sorry—the highest efficiencies occur when we have the highest solar energy but also when we're at the highest temperatures.

So we mentioned this before, the efficiency with which we can convert heat to chemical energy is very dependent upon the temperature at which you do it. There is a constraint in thermodynamics—it's much like the Carnot cycle constraint for heat engines—that says, basically, your theoretical efficiency increases as you go higher.

So again, at the highest solar concentrations and at the highest temperatures, we get the highest efficiencies, and that's where we obtained our 69 percent value. So this shows what we did in the year 2013, but it also implies for us that there was some low-hanging fruit that we could go after right away in reducing the losses, the two-kilowatt losses that are fixed from the system. So let's go on to the next slide where we show a little bit about the evolution of the technology as we've worked to reduce the losses.

[Slide 20]

So the two left-hand images, one above the other, showed the TRL 3 reactor that we did for the SunShot Program three years ago. That one obtained a 63 percent value solar-to-chemical energy conversion efficiency, and that was at the technology readiness level 3. So that shows you the reactor we used and the high-temperature recuperative heat exchanger.

In the year 2013, we used the same reactor, but we improved the heat exchanger. Again, the key is doing a great job of preheating the stream. So with an improved heat exchanger, that's located in the top center image, we improved the efficiency of the system to 69 percent. And again, that's where we were observing that we have a two-kilowatt fixed loss.

So to improve the efficiency further yet, the right-hand image shows a new reactor and a new heat-exchange network. So we improved the heat exchangers, and we also improved heat transfer within the reactor. So this system was designed to get us above 70 percent with the initial testing occurring in October and then additional testing to occur later this year.

So let's go a little bit further and look at a little bit about that reactor, as we fabricated it. I want to give you a sense of what the design of the reactor is like. So on the next slide, please, you will see the reactor manufacturing and assembly.

[Slide 21]

So the three images on the top are for the three plates that make up the reactor: a front plate, the bottom of which receives the solar energy; a center plate; and then a back plate. We bring the methane and steam into the center of that front plate, through the center, and then it flows radially outwards, then drops or comes up through orifices in the center plate to reach the back plate. And then, we flow it counter flow, radially again, toward the center.

This means we're able to take the hot reaction gases and actually do some recuperation into the reaction as it's operating, and not just use solar energy to drive the chemical reaction. So this recuperation makes it a little bit more efficient yet.

The bottom images show the system a little bit further along. So I mentioned that just right of center at the bottom, you see the front plate with the catalysts put into the system. The two images on the left are the system once it's been diffusion bonded together, and the image on the right is from testing the system, where we got to confirm that we had a good bond, and therefore, that the system would be ready for testing with the network of heat exchangers and on-sun. So let's go on to the next slide.

[Slide 22]

This slide shows, on the left, the heat-exchange network in a little bit larger form. There are two micro-channel heat exchangers in it. And the upper portion of that one is the high-temperature recuperator, and then practically just below it, in a similar shape, is one of the low-temperature recuperative heat exchangers.

And on the right—I threw this in here, probably should have shown it earlier. This shows the exergy flow for the system. It was for the TRL 4 system, but it shows the exergy flow with chemical exergy dominating the stream—that's in red—and the physical exergy related to enthalpy and a little more than just that as gray. As the system gets hotter, the physical enthalpy increased, but also in the steam-reforming reactor, the chemical exergy increased.

So in this set of calculations, we showed 40 kilowatts of methane chemical exergy coming in, and we see exiting in the syngas about 47 kilowatts. That's all due to the solar upgrade. Going on to the next slide, we show just a little bit more about the assembly.

[Slide 23]

I think what I'll mention mostly here is that on the right side, you can see the reactor. The other side of it is the side that—the side you can't see—is the side that receives the solar energy. The side you can see is where we did a lot of instrumentation. So we added over 30 thermocouples through the reactor into the system so that we could get a good job of measuring temperature profiles through the reactor and therefore do a better job, too, of determining where the losses are in the system. And this was extremely helpful in chasing down the heat leaks.

At the test site, which is in the next photograph or the next slide—

[Slide 24]

—at the test site in Richland, which is the upper-left image, that's where we set up the test then in October. The parabolic dish concentrator I mentioned before is one of the ones that had been developed by Infinia. And as I mentioned before, also, the nacelle is an Infinia nacelle without their heat engine but with our chemical system.

Now, I want to mention that next year we're going to start doing some testing in Southern California as well. So with help from Southern California Gas Company, we're going to have access to a test site that's at the San Diego State University branch campus in Brawley, California. They get considerably more solar energy resource than we get in Richland.

So—and you might not have imagined this if you haven't been here, but we get pretty good solar as well, particularly during the summer months. We are located—you can see on the image on the right where we are in Washington State, on the east side of the state. We're located east of the Cascade Mountains, which means we're in the rain shadow. We only get about eight inches of rainfall a year, and much of that comes in the March/April timeframe.

So when the winds are from the west, the water tends to drop out of the air in the Cascades. We get lots of days that are barely cloudy, or they're without clouds altogether. And in summer, they're very long days, so it's a great time to be getting test data points, where you go to steady state and wait and then do it again with a new condition.

And again, it's a great place to do that during the summer months in Richland, but it's not a very good place to try to test during the winter, as we can attest to by the fact that we were still testing our system in November this year. So we're really looking forward to going to Brawley for testing next year, where we can test around the clock, and where also we can have a little bit higher-intensity solar on the best days. Let's go on to the next slide.

[Slide 25]

That slide shows the setup of our concentrator's test stand in the year 2013. The upper-left image is the dish concentrator being assembled, and the upper, center image is the concentrator assembled with our test lab trailer.

On the right side—upper right—is a cold-water calorimeter that we built. We put that in the nacelle and put it up at the focal point so that we could measure how much of the solar energy that had impacted the dish, which we get separately from the DNI meter, how much of that actually makes it through the aperture. What I should mention is that a little bit of the solar energy that's reflected spills outside of the aperture on the nacelle and is lost. So we wanted to know how much that is. That's part of how we then knew what the real heat leak inside the reaction system is.

And the lower three images were kind of a fun activity that we took in June 2013. We decided to point it during the supermoon at night—at the moon. The supermoon is the largest moon that you get of the year, and it is also almost the same size as the sun, just slightly smaller.

So by putting a target near the focal point and moving it back and front, past the focal point, etc., we were able to confirm that our dish that would give us a reasonable amount of solar intercept into the device. And we also got a sense that the reactor in the nacelle would be properly positioned. So let's go on to the next slide, then.

[Slide 26]

This is the TRL 5 system, the system this year—or sorry—just last year that we put up on sun in October and November. There it is, fully instrumented and fully insulated with everything but the shroud that goes around the nacelle and connected to our dish. In the very next slide, we get to see results, and I should mention this is the first time we have shown results from this testing out to the public.

[Slide 27]

The left-side image is the one that corresponds to what we talked about before that tells us the heat leak in the system. The right-side image is what corresponds to the solar-to-chemical energy conversion efficiency.

And on the left, we have graphed both the data from the TRL 4 system and the TRL 5. TRL 5 is in blue. Now, the preliminary tests we did in October and November, we operated with a screen. That took out about 50 percent of the solar. So these are very preliminary with a lower solar concentration than we'll get to when we do some testing in the spring.

At that point, we were able to confirm—and this is after quite a bit of work, by the way, while we tested, to chase down those heat leaks and to reduce them. We were able to confirm that we reduced the fixed loss from two kilowatts to one kilowatt, and that's a pretty big deal for us. Because we're already paying the penalty of preheating the stream completely to the reaction temperature, that extra kilowatt is an extra kilowatt for the chemical reaction. And so that means we should expect, and we did see, higher efficiencies out of the system on the right.

Again, we tested at the lower power levels, but you can see that with the blue dots that we got higher efficiencies out of it, and just as the TRL 4 system showed higher efficiencies yet as we moved to higher powers, this system should show that too. So we are confident that we will get into the mid-70s at least when we test down at Brawley later this year. So let's talk to the next slide.

[Slide 28]

Let's talk—and we're getting near the end. We'll talk now briefly to the plan for our project later this year, as we move from a TRL 5 system to a more improved system at TRL 6.

Now our target at TRL 6 is still going to be the mid-70s for solar-to-chemical energy conversion efficiency, but we're redesigning the system based on manufacturing studies so that there will be considerably less expensive high-temperature alloy mass in the reactor. Our studies show that is a major part of the cost of our system. So by reducing that mass, we reduce the cost a lot.

And we're also redesigning the system so the thermal stresses are less because we want to have a long lifetime system too. So our TRL 6 system is designed to take us in that direction. We're also going to add power generation with the syngas product.

Now, our options are a microturbine or a fuel cell, and we're currently leaning toward operating with a fuel cell in that demonstration, and also, over the next year, we'll do more evaluation of manufacturing, and we'll do more techno-economics. It's working the problem backwards with economics that helps us move the system toward a 6-cents-per-kilowatt-hour power generation case or less than $2.00 per GGE hydrogen production case.

So let's go to the next slide, where I sort of wrap up and conclude.

[Slide 29]

So what we've talked about is what we've accomplished in the last couple of years. Clearly, we're getting very high-efficient operation. We're getting high solar-to-chemical energy conversion efficiencies, a lot more than you can get with other types of systems, especially a lot more that you can get with photosynthesis and things like that.

Because we are operating differently, we're using metals that can go to high temperatures, and we're exploiting the thermodynamics of these reactions, we expect to get into the mid-70s this year. We believe values exceeding 80 percent are feasible.

So it also requires improvements in the parabolic dish concentrators so there's less spillage outside of the nacelle. The core technology for this that enables it are the micro- and meso-channel reactors and heat exchangers, in part because they're very, very efficient, and partly because they can be mass produced, but also in part because they are so process intensive that they easily fit in that nacelle at the focal point.

The heat exchanger I showed you before—the high-temperature recuperative heat exchanger—has channels that are about 500 microns wide. If we went to a conventional heat exchanger, the unit would have to be almost 100 times bigger in volume; that would make it hard to fit up there and would make it very heavy, requiring a lot more structural mass to hold everything up.

As we said, we expect reasonable cost based upon our studies, and the near-term applications, we're working the problem backwards to make sure we hit a 6-cents-per-kilowatt-hour goal for power generation and a less than $2.00 per gallon of gasoline equivalent or $2.00 per kilogram goal for hydrogen production. Our preliminary calculations suggest we may be able to get closer to $1.00 per GGE, but there's still a lot to be proven to do that and the devil is always in the details. And finally, of course, the other applications are production of other chemicals, including synthetic transportation fuels. So on the next slide, just acknowledgments again.

[Slide 30]

We want to thank FCTO and SETO. I also want to be grateful and appreciative of the project team, the organizations and the individuals of the team, and some of them are on the next slide. So let's go one more slide over, and you can see some of the faces.

[Slide 31]

I just realized, for this one, they used a very old picture of me in the upper left. I no longer have a mustache. That's about 20 years old, so my hair looks a little bit different now, too. And finally, the next slide, we want to open this for questions. I guess we've got about ten minutes left.

[Slide 32]

Eric Miller:
All right.

Robert Wegeng:
Oh, and Amit, you might move it to the next slide so people can see the references we put up. Those are all papers that you can acquire to read on the progress of our system.

[Slide 33]

Eric Miller:
Great. Thanks, Bob. This is Eric again. I've got a couple of questions that were submitted online. Let me get to maybe the simpler clarification questions, first.

Robert Wegeng:
OK.

Eric Miller:
The first was—and I think this refers to a statement you made with regards to steam-methane reforming: "Does—Bob—does he mean 11 kilograms of CO2 per kilogram of hydrogen—H2? I think you might've said kilowatts."

Robert Wegeng:
Oh, if I said kilowatts, I didn't mean that, no. The conventional system is 11 kilograms of CO2 per kilogram of hydrogen.

Eric Miller:
I think you've answered the question. I think the answer is "yes." Terrific.

[Crosstalk]

Robert Wegeng:
Yeah, yeah.

[Crosstalk]

Eric Miller:
And another clarification, someone has asked you Bob, "Where might one obtain a parabolic dish?"

Robert Wegeng:
Oh, and that's a tough question. At the moment, the two companies that were working to commercialize them have gone out of business—Infinia and SES Corporation—and so that's a challenge I have also, is to get this to the point of mass production, I need to also encourage companies to develop parabolic dish concentrators.

If you go online, I think you can find that there are companies elsewhere in the world. There's one in China. There's one in Nova Scotia. There are a couple other places. I don't know what their prices are, and I don't know what their quality is either. So I'll just say that you have to search for that.

Eric Miller:
OK. Thanks. I've got—a bunch came in at the last minute here on clarifications. So let me get through this, and Jamie, interrupt me if you've got ones that you want to throw in—priority questions. But let me get through these—as many as possible. We have a question that—to Bob: "How was steam generated in the on-sun tests?"

Robert Wegeng:
OK, yes, for this test stand in the current time, both in 2013 and 2014, we used an electrical resistance-heated vaporizer that is inside the nacelle for generating the steam. Now, in an actual application, we'd be using heat from elsewhere for that. We wouldn't use electrical resistance, but this is what we do for the test stand.

So other options would be methanol synthesis—the heat from that. A microturbine has a lot of heat in the exhaust from that—in fact, a lot more than we need. So there are good ways to recuperate energy for steam generation outside of the nacelle. But all we did was an electrical-resistance vaporizer, is a cheap way to do that.

Eric Miller:
OK. Great. A couple of questions related to the feedstock—there's one question: "Is biomass the primary source of the methane charged to the reactor?" You mentioned the option up front. Maybe you can address that again—natural gas vs. bio-derived gas.

Robert Wegeng:
Well, I wouldn't say one is the primary over the other. Natural gas is one targeted application. Biogas would be another targeted application. Our LCOE calculations have been assuming natural gas, and part of the reason for that is natural gas is very available and very cheap in the U.S. today. But of course, if we worked with biogas methane, then we have an option to be able to say, "Wow, it's all renewable energy with no net CO2 emissions," which is very attractive, as well. But I wouldn't say one's the primary over the other. The technology can work with either.

Eric Miller:
Great. And I think I'll jump to on that, just a springboard from what you just said about the LCOE calculations. One question came in: "Have you analyzed the LCOE with vs. without the solar-reforming step and including sensitivities to natural gas spot prices and also to solar-to-chemical conversion efficiencies?"

Robert Wegeng:
OK. So yeah, I think the general answer to that is almost always yes to all parts of that. In our LCOE calculations, we include a base case that's a natural gas combined cycle power plant that's only working from natural gas so that we have that as a comparison. We also work the case that I showed on this, where we are using our solar concentrator to derive the syngas when the sun is shining, but it operates off natural gas when the sun isn't.

We have worked with various—in our sensitivity calculations, we have worked with various values for the natural gas cost. And we've also worked sensitivities with various values for the efficiency of the system. Working those things helps us know what our targets should be to stay under six cents per kilowatt-hour, so we include all of those.

Eric Miller:
Great.

[Crosstalk]

Jamie Holladay:
Bob, one of the questions was—

[Crosstalk]

Eric Miller:
Good to hear, huh?

Robert Wegeng:
Yeah.

Eric Miller:
Go ahead, Jamie.

Jamie Holladay:
Oh, yeah, one of the questions: "What percent methane conversion have you obtained?"

Robert Wegeng:
OK, so this is all a function of pressure and temperature because thermodynamics limits conversion. In our cases, we have been as high as 98 to 99 percent methane conversion, but if we go to higher pressures or lower steam-to-carbon ratios than we usually have tested at, then the values can't be quite as high. So it's been a range of values, but 98 to 99 percent are values we have obtained during some of the tests, and we know the conditions for that.

I'm not convinced that that's the most economic case. Sometimes getting the most out of a system is you pay so much more for the system, so we're not sure yet what the—from a sensitivity perspective—what the target methane conversion should be.

Eric Miller:
Great. Jamie, do you have another?

Jamie Holladay:
Yeah, sure. One of the other ones was, "You mentioned that you're looking to commercialize this system. Do you have a timeline to get this to a commercial application?"

Robert Wegeng:
Yeah, we do have a timeline for it. Our goal is to get to a commercial point in about the year 2020. That means that after this next year of testing, we've got to start building multiple systems and get them out in the field to test so that we can start confirming failure modes and mean time between failures and things like that, refine the design, you know, go to a larger number for testing, etc. But we've also get in place mass production facilities for the systems, so we're looking at that.

And we are also looking at that other problem I talked about, which I now consider one of my biggest problems, and that is a mass-produced parabolic dish concentrator for the system. So 2020 is our target, but the devil is definitely in the details.

Jamie Holladay:
So another interesting question that I think, Bob, you can point to is some references for this one. Was any modeling used to optimize the system?

Robert Wegeng:
Oh, yes. Definitely. We've used COMSOL Multiphysics in addition to CHEMCAD and Aspen, but we've used COMSOL Multiphysics to model what happens inside the chemical reactor itself. So it's a 3-D reacting flow model, and we calibrated the model with results from the tests so that that helps us improve to the next stage. So it's a major effort, but very worthwhile.

Oh, and I should point out, in the references that are up on you—up on the screen, you can see one paper. The one at the bottom listing is one of the papers from 2010, where we first started doing that kind of modeling for the reactor, but we've done much more since then.

Eric Miller:
Hey, Bob, I just want to let you know, at the last minute, we got a flurry of online questions. We'll get through as many as we can, but we'll try to deal with the rest offline. We have about four or five more minutes. Let's try to get through a couple more.

Robert Wegeng:
OK.

Eric Miller:
And I'll try to combine a few. There's one question related to what kind of improvements are needed in the parabolic concentrators, and combining that with, "Is dust a problem, a significant problem, on these concentrators?"

Robert Wegeng:
Yeah, I don't think we have a significant problem in the technology for the concentrators, but we would like to see less energy spilled outside of the aperture on the nacelle. So at the moment, parabolic dish concentrators, the good mirrors give you about a 93 to 94 percent reflectivity. You can't do much better than that.

But the system we're testing with at our test stand only brings about 88 to 90 percent of the solar that's reflected in through the aperture. So we're losing 10 to 12 percent solar that shows up as brightness on the outside of the nacelle. We're losing that at the moment, and parabolic dish concentrators have been tested in past that were up around 98 to 99 percent—what we called the "solar intercept."

And so I would really like to see the parabolic dish concentrator get a little bit better—at least 95 percent. Every bit of that gain is gain to the chemical reaction, and so it makes sense to start engineering the parabolic concentrator now, too, and not just the reaction system.

Eric Miller:
Great. This must be from a fuel cell person. This is in regards to one of your slides that mentioned the $2.00 per GGE, the gallon of gasoline equivalent. The question is, "By $2.00 per GGE, is that equivalent in chemical energy or electrical output?" And they note that "since fuel cells are a lot more efficient than turbine or internal combustion engines."

Robert Wegeng:
Yeah, no, that's in terms of chemical energy. It's not electrical output, the $2.00 per GGE.

Jamie Holladay:
Yeah, it's the standard Fuel Cell Technologies Office definition of GGE, so it's the equivalent of one gallon of gasoline has about the equivalent energy as—almost about the same amount as one kilogram of hydrogen on a lower heating value.

Eric Miller:
And another fuel cell-related question is, "What, on fuel cell technology, are you envisioning for your new power generation options?"

Robert Wegeng:
Well, for the overall application, I'm willing to let the fuel cell manufacturers compete and deliver the best value for the dollar. This system could be worked with PEM fuel cells. It could be worked with solid oxide fuel cells. It could be worked with molten carbonate fuel cells.

The main thing I'll say, though, in addition to that is that a different fuel cell—one fuel cell as opposed to another imposes different requirements on the processing of the stream. And it also provides different opportunities. So the high-temperature fuel cells provide sufficient heat that we could easily make steam from that and have a high-temperature operation. In fact, we may be able to make the overall system more efficient by using some of that heat to support some of the reforming.

The PEM fuel cell looks like it could be very low cost, and so it's a very nice opportunity to bring this in for low-cost electrical power generation. So the PEM fuel cell doesn't produce as high temperature heat. It's actually moderately cold, so we can't really use it for steam generation. But on the other hand, as I said, it looks like it has a very good cost and a nice efficiency. So I'm willing to end up adapting this to whatever makes the best economic value proposition.

Eric Miller:
All right. Good. And let me finish this up with a couple of questions about the micro-channel technology. I'll combine two questions: "What percentage of the total system cost is the micro-channel and meso-channel devices? And are there manufacturing innovations needed to bring the cost down?"

Robert Wegeng:
OK, so I don't have the numbers on top of my head. I believe that the manufacturing aspect of this is not a high percentage cost. A higher percentage cost in the reaction system is the raw metal, the high-temperature alloys in the system. So the manufacturing, the fabrication part, isn't really a high part of the cost.

But are there additional developments that could be done, etc.? Well, definitely, yes. At the moment, to my knowledge, no one mass-produces micro-channel reactors and heat exchangers. We've been examining different methods, and we've been using our friends at Oregon State at the Microproducts Breakthrough Institute to do testing with different methods so that we could have what we hope are good projections then on costs. But there is quite a field to be worked here to—and it could become quite competitive for micro-channel reactors and heat exchangers mass-produced, and various methods are becoming quite exciting.

I think the 3-D rapid prototyping approach or direct laser metal sintering looks really promising right now. I'm very interested in whether that's going to lead to cheaper devices than when we go through a more conventional method that we've used the last few years for the fabrication.

Eric Miller:
Great. And it looks like we are actually out of time. We've gotten through as many questions as possible. I'd like to hand it back over to Amit for final housekeeping.

Jamie Holladay:
Hey, Eric, one last point for the cost information. One of those references did go through that cost information so the—

Eric Miller:
Great. Thank you.

[Crosstalk]

Jamie Holladay:
—can look there. Sorry to interrupt.

Eric Miller:
No, thank you. Thanks for clarification, Jamie. OK, Amit, take it away.

[Slide 34]

Amit Talapatra:
OK. Thanks, Eric. I'd just like to thank everyone who attended today's webinar, and I'd especially like to thank Bob for presenting today. If you have any questions that we weren't able to get to, you can send them to our email addresses shown here, and I also wanted to remind everyone that a recording of today's webinar and a copy of the slides will be available on our website in roughly ten business days. Thank you.