Hydrogen Production Analysis Using the H2A v3 Model (Text Version)

Below is the text version of the webinar titled "Hydrogen Production Analysis Using the H2A v3 Model," originally presented on July 9, 2013. In addition to this text version of the audio, you can access the presentation slides and a recording of the webinar (WMV 158 MB).

Alli Aman:      
Thanks so much for joining today's webinar. Just to go through a few housekeeping items before we get started. Today's webinar is being recorded so a recording, along with slides, will be posted to our website in about 10 days. I will send out an email once those post to our website. And everyone is on mute, so we will have a Q&A at the end of the presentation so please submit your questions as our presenter's presenting and we will cover those the last 10 or 15 minutes of the call today.

And also, I encourage you to check back to our website for future webinars. We do host these monthly, sometimes bi-monthly, so I encourage you to check back to our website often for future webinars and I also encourage you to sign up for our monthly newsletter. That will keep you up to date on webinars as well as other things going on within the Fuel Cell Technologies Office. And on that note, I'm going to turn it over to Eric Miller who's the hydrogen production technology development manager at the DOE Fuel Cell Technologies Office. Eric?

Eric Miller:      
Thanks, Alli. And welcome to Brian and all of our webinar attendees. Our presenter today is Brian James. He is the director of energy programs at Strategic Analysis Inc. and has 25 years of professional experience working on high technology projects and alternative energy analysis. Particular specialties include fuel cell power systems, hydrogen reformer systems, and systems performance and cost effectiveness analysis.

He has conducted numerous techno-economic analyses and manufacturing cost analyses for the Department of Energy, NREL, and private industry. Brian was a 2005 and 2007 recipient of the DOE Hydrogen Program R&D award and holds five U.S. patents. And I know Brian has a lot to cover for you today so I'll keep this intro relatively short. The webinar is titled "Guidance for Filling out a Detailed H2A Production Case Study."Basically, H2A is a discounted cash flow tool using a consistent set of technical and economic inputs and assumptions allowing for relative cost analyses of different hydrogen production technologies at central, semi-central, and distributed production scales.

As such, it has been an extremely valuable tool for us in the identification and analysis of promising pathways for meeting the DOE hydrogen production cost targets. The H2A tool itself is publicly available and we encourage all interested stakeholders in hydrogen production to learn its use and to develop detailed case studies of their favorite production technologies. And to help you with this, Brian will be describing the use of H2A version 3, which is the most recent version of this techno-economics tool. And with that, I will hand it over to Brian.

Brian James:   
Thank you, Eric. Once again, Brian James from Strategic Analysis and I'll be talking about filling out a detailed hydrogen production case study.

[Next slide]

There we go. So, overall, I want to explain the H2A model capabilities, including how it can be used to compare various hydrogen generation technologies and chart their progress. As part of current or future DOE contracts or even proposals, you in the audience might be requested to prepare an H2A case study, so I want to give you some ideas as to exactly how to do it from a user perspective.

So, this presentation is going to review elements of the H2A Excel model, give examples of a fully detailed case, identify key numbers and common pitfalls and errors that people often make as well as clarifying the level of depth and accuracy and transparency needed for the analysis. And I'll conclude by discussing some metrics and common issues associated with a variety of cases.

[Next slide]

Once again, H2A is a discounted cash flow analysis that computes the required price of hydrogen for a desired after-tax internal rate of return. It's written in a variety—well, it's written in Microsoft Excel. It uses custom macros and we're up to version 3 of the model.

There are two main types of H2 analyses, production and delivery, and I'll be primarily focusing on the production elements today—or actually, exclusively on the production elements. And the objective for the overall H2A analysis is to establish a standard format for reporting production costs to allow comparison of technologies as well as to provide needed transparency and consistency of approach.

[Next slide]

So, there's a process flow diagram that basically shows the inputs and the outputs of the system. We begin with standard price and property data and these are largely populated from data supplied by the Annual Energy Outlook—AEO—from Energy Information Agency and some of the physical property data comes from the GREET model developed at Argonne National Lab.

So, we combine the standard price and property data with specific information about the case study to be examined—the description and the title—and then very specific technical analysis aspects, input parameters—performance, and a process flow diagram and stream summary—defining exactly what the particular case study is going to be physically doing. So, we combine that technical along with the cost elements. So there's a variety of financial inputs and costs—capital cost assumptions, replacement schedules—and I'll talk about all these inputs both on the technical and the cost side in the remainder of the talk. So, they all combine to come up with the results of the model, which is the production cost of hydrogen including a specified rate of return, a cost breakdown of the main contributors to that production cost, as well as doing a brief sensitivity analysis. And all this combines to give information on the key cost drivers for the overall case.

[Next slide]

So, in terms of hydrogen production case studies, there are two main types. There's distributed or forecourt. And forecourt refers to the service station—sort of the corner gasoline service station where hydrogen, in this case, would be produced and then dispensed directly into the vehicle. And those typically are between 1 and 5 metric tons delivered—of hydrogen—per day in the vehicles, but the nominal value is 1,500 kilograms per day of hydrogen. And that's contrasted with the central case, which is further outside the city and is much larger in scope. It's a true plant—100 to 500 metric tons of hydrogen per day.

We also differentiate between the cases in terms of the time frame of the technology, breaking them down generally into current, future, and ultimate cases. For the current cases, it's as if you were fabricating today at some level of production volume. So it's not "onesies" and "twosies", but it should be a short-term projection of technology that's already been demonstrated in the lab. So, it doesn't have to be what is fully demonstrated today but it shouldn't be far in the future—should be pretty much grounded in what you're highly confident can be done. And it can include potential price reductions from both production volume scale as well as implementing design changes that you've already identified.

The future cases, in general, use advanced—more advanced materials. They often have increased efficiencies and longer plant lifetimes and often improved assumptions regarding replacement schedules and lifetimes of constituent parts. And it goes along with sort of a—in general—a reduction in the capital cost for the future cases compared to the current cases. And then finally, the third case is what we call the ultimate target case, and these are meant to be representations of, based on assumptions, for the limits of the thermodynamic, physical, or economic limits. So, they're supposed to be the limit of the particular approach and in general, we would like it to approach the DOE production target—once again, it's production, not all of pump price, but just production price of $2.00 per kilogram of hydrogen. But it doesn't necessarily have to meet that. We'd like it to, obviously, but we want you to be honest as to what the ultimate target in the advanced or the limit of the technology is.

[Next slide]

So, what exactly does the model do? Well, there are some general governing equations. Once again, the objective is to solve for the required price of hydrogen that returns a desired after tax internal rate of return after adjusting for all expenses. And the method that it does this with is the discounted cash flow analysis.

So, I won't go into the actual formulas that are used here but it's fairly standard. For each year of the lifetime of the analysis, it adds up all the yearly revenues and subtracts off all the yearly expenses, and thereby determines what the overall production price of the hydrogen needs to be to return that internal rate of return. And there's more of a description of the governing equations and the approach in the user guide shown here at the DOE website.

[Next slide]

So, lots of different technologies have been demonstrated within H2A in the past. There are a variety of existing technologies—natural gas steam methane reforming, electrolysis, ethanol reforming, biomass, coal, nuclear water splitting.

And then there's a variety of general emerging technologies—photoelectrochemical PEC, photo-bio, just called bio in many cases, and the STCH cases, solar thermochemical. All of these production cases, both the existing and emerging listed here, can be found at the DOE website shown here. And I would encourage you, when you go to fill out your own case, to go to the website, find the particular technology case that most closely matches yours—it doesn't have to be exact, but it should be generally the same—and use that somewhat as a template in the system to filling out your particular case.

As I mentioned before, there are two broad categories—the forecourt and the central—but in general, as we move forward in hydrogen production cases, or H2A production cases, we want to focus on an increased level of details in the analysis. We want to focus on emerging technologies. We want to increasingly apply uniform, primary metrics from one technology to the other. In many cases, we'll have a primary metric that is a comparison—a term used for multiple technologies—then they'll have individual sub-metrics that are appropriate specifically to that technology. We want to enhance our sensitivities—both tornadoes and waterfalls that I'll discuss in a minute—and we also want to involve multiple versions—this is the current, future and ultimate target time frames—to chart technology progress. So, in today's presentation, I'll be using an electrolysis forecourt case study as an illustration of how to fill out the model.

[Next slide]

And this is specifically what I mean by the electrolysis case. It's a forecourt case delivering 1,500 kilograms of hydrogen a day and it's a stand-alone, grid powered electrolysis unit that's based on some performance and price quotes from Norsk Hydro a number of years ago. These are the system components. I won't go into the details here. There's a website with the case study and also, there was a 2009 independent review of this particular electrolysis case study that was done and is available at the website shown here at the bottom. But at its heart, it's a forecourt station where you put in water and electricity and you get out hydrogen and oxygen.

[Next slide]

So, let me dive in and talk to you specifically about the H2A model itself.

[Next slide]

The model was organized as an Excel workbook with multiple tabs—they're shown across the bottom. Each of these bubbles here—buttons—is the name of a tab that appears at the bottom. There's a variety of them that are informational—title, description, process flow, and I'll go over each of these in a minute. There's a set of them that are system inputs, and these are the key ones which you need to particularly pay attention to.

There are four of them—four tabs—that are largely results oriented and then there's a half dozen or so additional tabs that you really don't need to enter anything to or even reference unless you're curious. For purposes of transparency, they're all listed here, but they contain standard price and property data as well as financial schedules such as depreciation schedules and constants and conversions that are important to have in the model for transparency and important to have in the model for consistency for comparison among technologies. So, now what I'll do is I'll through each of these three top rows, each of the tabs associated with each of those buttons there.

[Next slide]

The first one is the Title tab. It is, generally, exclusively informational only. It doesn't draw any numbers from this for use in the actual computations, but it's very important to fill out the information here because it is the quick snapshot as to what the model actually represents. And we also strongly encourage you to fill out the dates and a brief description of what you've done to the model because these things are iterative by nature and there will be multiple versions of your very own model so it helps to keep track of the changes.

[Next slide]

Next tab is the Description tab, which once again is informational only. The main block here is for a system description. We generally want a paragraph or two of information. We want you to be as descriptive and detailed as conveniently possible. So we don't want one sentence and we don't necessarily need multiple pages. But we want a fairly specific description, in words, of what the system actually is doing and what it's—how it works. And any references that you have, this is the place to put them.

[Next slide]

We also want a process flow diagram on the Process Flow tab—and system schematics are a very important means to convey information concisely to the user. On the left, we have a process flow diagram from a forecourt steam methane reformer system that we analyzed. It was separately analyzed in Aspen HYSYS, and HYSYS prints out these PDFs, so it was convenient to do a screen capture and paste it in here. And HYSYS also has very detailed reports listing the temperature, pressure, and flow rates at every node in the process flow diagram. So, because it was so convenient, we were able to screen capture a very lengthy report and paste it in this tab. It was convenient, and fast, and it's information that's useful to people, so go ahead and put it in.

In contrast to that is what we did for the electrolysis forecourt case, which was not separately performance analyzed but rather was based on the performance numbers of the material vendor, or system vendor. So, for that, we put in their actual system diagram and we created our own simple table—shown down here—with just the temperatures and pressures and flow rates at key points throughout the electrolysis system, which in general is much simpler than a steam methane reformer.So, we encourage you to put something in here—it's required to put something in here—a process flow diagram as well as a table of supporting node information.

[Next slide]

The most important, perhaps, tab in the H2A workbook is the Input Sheet Template. This is just the top level of it. It goes down quite a few rows below and I'll talk about those in a minute. But a key aspect of it is this Calculate Cost button over here. After you finish putting in your input information and you want to view the results, you need to come back to the Input tab and press this button, Calculate Cost. It will initiate a macro that starts the discounted cash flow and then spits out the results in the results tab. If you forget to push this button, it will prompt you to do so, so don't forget this.

Also, importantly, the input cells within the Excel workbook are color-coded. Blue means it's a calculated cell and you shouldn't try to change the values in the blue cells because those are there for a purpose—they're pasted in by the macro or they're really a formula that is invariant with the technology and therefore, you shouldn't be messing with it. This peach color here are the ones—are the cells where you want to input your data so always watch out for the peach colors. Hopefully you won't see any red, because it represents an error, and the green and the yellow are largely informational or explanatory cells.

[Next slide]

So, on the Input sheet, there's also several rows for basis years. And what I mean by that is there's a couple of different years that are defined in the H2A spreadsheet that are worth distinguishing between. There's the reference year, which is the dollar year for which the results are reported, and by DOE specification we want that to be 2007, so this takes out inflation when comparing one technology to another. And when projecting on what the costs of this specific technology will be, they always report it in the same 2007 year dollars. And the way you tell it that is by going up to this peach cell under Reference year and typing in—I think there's actually a pull down menu—for what year you want. Second type of year is the basis year, and this is the dollar year for the costs—the capital costs—that you enter. It can be any value you want, you just have to tell the spreadsheet what to do and it will do the computation for you to convert the dollar years, your capital costs, into the reference year dollars for which the cost results will be reported.

And then finally, there's the assumed startup year, and this is the year of the plant startup. It's used primarily with looking up the projected costs of feedstocks and utilities that are drawn, once again, from the Annual Energy Outlook annual projection or year-by-year projection in the future of what feedstock costs will be. And this could be any year you want, but it has to be appropriate to your specific case study. So obviously, it would be different for a future case than for a current case or an ultimate case. And then there are some more words here to further explain the things I've gone over.

[Next slide]

There's also a variety of financial parameters. These are input on the Input Sheet tab; however, they don't appear exactly like that. I break them down into two different categories. There are some of the financial parameters that you may want to change but maybe not. Plant life is typically—and by H2A default—20 years for forecourt and 40 years for the central case. These are really good rules of thumb to live by but if there was some really, really strong reason that your technology would have a longer life in a forecourt than 20 years, then you could go ahead and change that. Chances are it doesn't. And that doesn't mean that it has to last 20 years. You could have a reactor replacement or some type of component replacement in say 5 years or 10 years or something shorter than 20. So, we generally keep to this unless you have a very compelling reason to deviate. These are contrasted with other values such as the after-tax rate of return, depreciation schedule, taxes that we really want to be constant across the various technologies. So, these values really should not be altered in the analysis.

[Next slide]

There is also a Capital Cost tab, which is a very important tab, as you might imagine. Once again, the peach colored is input and the blue are for outputs. What we want to do here is we're looking for roughly 5 to 20 lines of a capital cost breakdown. You should use your judgment as to which—how many lines to do. We don't want one line, because it doesn't inform us, and we don't want 50 lines.

We want you to definitely add comments in the green zones to describe the basis of your cost estimates. If possible, you can use formulas rather than pasted-in values to better explain your logic. The example I give there is if you have a concentrating mirror and the cost basis is based on so many dollars per square meter of concentrator area, we'd like to see that in that formula and also described in the comments rather than just a numeric value of dollars for the entire system. It just better explains exactly what the basis of your numbers is.

A lot of times, people don't know what the capital cost of their system is. We need you to do your best to estimate that, and you can estimate it based on approximate material cost with the mark-up for manufacturing. In general, it's best to break it down into as many sub-systems as you can and then estimate each of those individually and summarize them—and sum them. Use your best judgments to come up with a very clearly explained rational basis for your capital cost estimates.

[Next slide]

There's also a tab called Plant Scaling tab, which is new to version 3. It is only for the very experienced or so-called H2A power user because it represents an ability to scale or to have a different capital cost—a different size plant for capital cost input than for the hydrogen production analysis that you were doing. I would strongly recommend that you don't use this feature if you're a novice user, or even if you're an experienced user you probably shouldn't use this. And just ensure that the baseline design capacity shown here, our 1,500 kilograms a day of hydrogen, is the same as entered in the peach colored tab on the Input sheet—and that's explained in words down here which two values need to equal each other. When they do that, then the scale ratio will be one and there won't be any scaling in here and therefore, it won't lead you astray. It's possible to get erroneous numbers that are hard to detect the error if you—by using this Scaling tab and I encourage you not to use it if you're a novice.

[Next slide]

There's also a Replacement Cost tab, and this is the location where you type in as much information as you want as to the annual replacement schedule for your system. In this particular example of electrolysis, it has a 25 percent system capital cost charge every 7 years and this represents restacking the electrolyzer system. If you know—or have a shot at knowing—what the replacement schedule for your system is, then you should put in as much detail here as you can, both numerically—in this column here, as the costs that correspond to this percentage, it's just listed as a percentage for your convenience—and also in the comments column which is over here on the right, not shown on the screen, but in which you would explain each of these rows as to what they are.

There are always going to be some unplanned replacement costs. And although listed as zero here in red from the Input sheet, on the Input tab there is a percentage in there—typically 0.5 percent—that is incurred in systems every year as sort of an unanticipated, unplanned replacement expense. And were there a percentage on the Input tab, it would show up repeated here and then this column over here for unplanned replacement costs would be automatically populated as an annual expense for each year of the analysis.

[Next slide]

Let's hop back to the Input sheet. After one enters the capital costs, those capital costs are transferred over to this blue cell here representing the total direct capital costs and then there's a series of other input factors that must be added covering site preparation, engineering design, process contingency, et cetera. Most of these parameters can be computed as percentages of capital costs and they're listed as H2A default values. So, what I would encourage you to do is when you've located a case study that is closest to your technology, use that as a template. Look to see what they did. And if you don't have any better value and data or a strong feeling that the cost in this category is different than the H2A default value, then use the H2A default value. Likewise, some of the H2A default values are discussed in the model user guide, which is also available at the DOE website.

Also, always look in the Notes column, or add in the Notes column exactly what your assumption is over here. It helps both to spot errors because it's sometimes easy to overlook a formula in a cell where all you can see is the result, where over here you can quickly see that it's based on a percentage, and you can spot check your results and your assumptions much more easily.

[Next slide]

Also on the Input sheet, further down the page, is a spot to enter feedstock, utility, and byproducts costs. There are other sections that I'll go over in a minute for other materials and byproducts and fixed costs, but this particular section of the Input sheet is for feedstocks, utilities, and byproducts. What you do is you go over to this peach-colored cell and you're picking out the Annual Energy Outlook data table from which to draw the prices. And there are three choices in the pull-down menu. You should pick the top one, which is the 2009 reference case which is the H2A default value—assumption. And then you selected whether you want a feedstock, a utility, or a byproduct.

When you select one of those, then you'll have a further pull-down list of these particular—production feedstocks. So, when you click on one of those—the value's over here in this pull-down—it will appear in this cell here, and it's gigajoules per kilowatt hour, whatever parameters will be automatically populated in these blue cells down here, and you'll be asked to put in the appropriate usage of that particular product. In this case, it's industrial electricity and it has a usage of 53 kilowatt hours per kilogram of hydrogen. So, what you want to do there—and it's going to use the look up tables for the year that it's doing—once you've done that, you go over and you hit the Add button here and it will go through a macro that takes this information that you entered and it enters it into a row down below here in blue, as well as knowing internally into the system that it needs to do that feedstock at this conversion, or this usage at the specified price in the look up tables in each year of the discounted cash flow.

So, you do this for each of the utilities, feedstock, or byproducts that you have in the analysis—going through, selecting them, and adding. If you make a mistake, don't change the numbers down here in blue. Remember, don't change things in blue. You have to delete it and then go ahead and re-add it in, so using the Add and the Delete button. Some people make the mistake of trying to change the numbers down here. You can do that. It'll physically let you do it. But it will not be properly enumerated within the discounted cash flow.

[Next slide]

Second block here is the Other Materials and Byproducts sheet further down the Input sheet. It has a similar pull-down menu for Other Materials. These are your options—a variety of cooling waters, oxygen, acids, et cetera. You go through the same selection in Add as well as Usage—it's blank right now, Usage, type in there and then it appears down in the list. So, this is particularly one that has process water, cooling water, and compressed inert gases all used with a specific usage per kilogram of hydrogen, and then it will then be used in the model.

[Next slide]

And the third category is where you enter fixed operating costs, which to a large degree is, in the items shown here, led by manpower listed in full time equivalents. So, you carefully consider how many people you need at the plant, explaining your basis for the assumption here and a lot of these other parameters could be H2A default values. Once again, look at the User's Guide as well as the existing case studies for guidance as to what to use there. But always explain what your thinking is over in the Comments column. And finally we get to the results.

[Next slide]

This is just a summary of the results. It does not appear in this form in the H2A model. But down here, we have what appears on the H2A sheet, is a breakdown of levelized costs. The model also does compression storage and dispensing for forecourts, but we're primarily interested in production here so these are the breakdowns. These are the categories that it breaks them down into as standard pieces and then sums them up into this projected cost of $4.17 per kilogram of hydrogen.

[Next slide]

Also as outputs are the tornado charts, which it does automatically. Data is entered on the Sensitivity Analysis tab. And the way tornado charts work is there is a base cost—this is the levelized cost of hydrogen here as the center line—and then independently, for each critical parameter, you postulate a low value to that parameter and a high value to that parameter. And then you compute the cost per kilogram and type it into the chart down below. And then the macro within the H2A model automatically creates a tornado chart for you. So, it automatically draws the widths based on the data in these peach-colored cells that are entered there. So, you don't have to actually figure out how to do the offsets for the tornado charts—it does it for you down here for your convenience. You may have to reorder the parameters here to get the tornado shape with the widest, most sensitive parameters on top and the least sensitive parameters on the bottom. That's easily accomplished.

[Next slide]

Although not automatically in the H2A model, the DOE is particularly interested in examining waterfall charts going forward into the future. For those not familiar with waterfall charts, there's this one here for electrolysis. It's a generic one. I've removed the dollars per kilograms just to show the main features. It begins with a sort of a status set of assumptions and a cost breakdown of the key components of the system. It's representing—the top bar here—it's representing the cost of hydrogen. And then we go through each specific design change and cumulatively chart how they drop costs. So, starting here, and then efficiency improvements drop the price to here, and capital cost reduction drops the costs down to here. The order of these matter, because unlike the tornado charts, which are independently varying the parameters, these cumulatively vary parameters. So this cost to achieve this is the cumulative effect of each of these, and then of the future costs here, we should further show the breakdown. Once again, these are going to be used in the future, you should start thinking about these, but they're somewhat of a work in progress as to the exact guidance the DOE is to give in their construction.

[Next slide]

There are also a variety of other examples of H2A cases that I'll briefly outline.

[Next slide]

Of particular interest are the solar-based approaches. Those are typically biological approaches either using algae or bacteria; solar thermal chemical approaches, STCH; and then PEC—both colloidal slurry suspension systems, which are baths, a particle suspended in a liquid, and then electrode PEC systems, which are effectively a combination of photovoltaic and electrolysis where a plate electrode is suspended in a water bath and the sunlight causes oxygen to evolve on one electrode and hydrogen gas on the other. So, it is water splitting but it's PEC water splitting.

[Next slide]

Let me talk briefly about some STCH concepts. This is a solar dish concept that was done in a case study. It was based on a design by Sandia that involves individual dishes with a STCH converter system at the focal point. But as prep work into the H2A analysis, one had to consider the overall size of the plant being modeled, how many dishes it would take, what the spacing of the dishes is, how much land it encompasses, the size of each dish because it affects the capital cost of the dish and also of the converter. And one also has to consider the line and the piping connections to connect the dishes to each other and to collect and manifold the product and the feedstocks.

[Next slide]

So, following on with the solar theme, the STH efficiency stands for solar-to-hydrogen, STH. Energy conversion efficiency is a critical prime parameter for all of the solar cases. Simplistically, it is the lower heating value of net hydrogen—not gross, but the net that you get out of the system hydrogen—divided by the total solar input into the collector system. However, each of these terms has a bit of a nuance associated with it. The solar energy must be the full spectrum energy into the system. You cannot choose the wavelength that you want to include in the efficiency calculation; it has to be full solar spectrum. I know many systems only have specific wavelengths that they're reacting to.

And secondly, we're not interested in the—for the purposes of the efficiency calculation—the photons that hit the ground in between the collectors or in dead parts of the collector. We're really capturing the full active area of the collector system, not the space in between. Additionally, you have to make sure that your major terms are consistent with each other. It's common sense, but it bears repeating. There are solar energy and intensity assumptions that need to go into your system.

You should consider and be mindful of the direct versus indirect insolation effects. If you have a concentrator system, you pretty much only can use the direct energy. The total—direct and indirect—is included in the efficiency calculation, but in terms of what is useful to you, it's really only the direct component. You have to consider tracking, because that affects insolation, as well as potential blockage, which affects potential spacing of the systems next to each other. The collection area is obviously tied into solar intensity as well as the conversion efficiency.

Capital cost has to be scaled appropriately for your system and you have to size, in general, the systems for peak hourly production. So that would be the longest day of the year, the brightest day of the year, at noon, not necessarily the average, which in general increases your capital costs, to size for the peak. With that in mind, your hydrogen production rate is mostly referred to, but not necessarily exclusively, as a yearly average. That's the shorthand we use for hydrogen production, although the capital costs need to be scaled for peak production. So, you have to reconcile the hourly peak as well as daily and seasonal variations, both for the hydrogen production as well as the capital costs associated with the system.

[Next slide]

While STH efficiency is the key metric, we want to know more. We want to understand how you derived that STH efficiency and so we suggest that you break down STH into the appropriate component efficiencies and define exactly what you mean by them and explain what the basis is. This STCH example breaks it down to the optical, receiver thermal, and reactor conversion, and then has a product going at this 6.2 percent. One also has to break it down by year for the case studies and so we want to be able to compare how the individual component technologies advance in the future and to generally increase the STH efficiency shown here in the example. So, by all means, break it down into the appropriate sub-parameters you have, but also report the macro metric.

[Next slide]

You're encouraged to include supporting calculations as an added tab at the end of the workbook. This is an example that the Sandia guys did for a STCH model. They calculated the actual spacing between the solar dish arrays out in the field to avoid blockage. They did this in a separate spreadsheet and really, the only results that go into this, into the H2A model per se numerically, is the amount of land that is required for their array. Because the model doesn't really care what the spacing is but it does care how many acres of land it has to buy to support the plant.

If you don't want to do it in the Excel tab, you can alternately use a Word document, as was done here, to further explain and show the basis of your assumptions.

[Next slide]

And finally, the DOE in their MYRD&D plan—that's hard to say—has a series of technical target tables. These are, in general, broken down for each of the technologies, each of the main technology types, and they're broken down into two types of tables. One is a table of actual targets and metrics. These are meant to be measurement metrics that are applicable to multiple different approaches, and they tend to be normalized per ton per day of hydrogen or per year or even for both.

Supporting those are a table of supporting assumptions, and these are listed for illustrative and informational purposes only. They hang together to support the metrics in the column up above, but there are many ways to achieve these top table metrics. So these are not really targets per se, they're guidelines or suggested metrics that you can compare against that show one possible method of achieving the top metric. But there are other ways that you're more than willing to pursue; you just have to explain what you're doing. But they are included here just to give you guidance into what DOE is thinking and a potential pathway to achieving the upper metrics.

Each of the tables, as you see here, has lots of footnotes attached to it…

[Next slide]

…and these copious words appear in the MYRD&D plan to describe the details that went into each of the metrics and the assumptions, and they can be quite lengthy.

[Next slide]

This is the bio table. I won't go into the details here but it is likewise configured with a table of metric targets as well as some supporting assumptions that feed into the metrics along here.
 
[Next slide]

The footnotes for the bio tables.

[Next slide]

Two pages of them.

[Next slide]

Then there is a table for the PEC photo-electrode…

[Next slide]

[Next slide]

…as well as a table for the PEC colloidal suspension system, which has a further specification of being a dual bed system as opposed to a single bed system. A single bed system would co-evolve hydrogen and oxygen and the water splitting reaction. A dual bed system has one bed for hydrogen evolution and one bed for oxygen evolution. So it's showing the specificity of it that would not necessarily affect the overall target table for the system, but it would certainly have an impact on these supporting metrics down below.

[Next slide]

And once again, this presentation is going to be available after the WebEx. Alli will have further details on that.

[Next slide]
           
And then this is a summary of the various websites that contain the existing case studies that you should use for comparisons, as well as the User Guide that has more specifications as to the default values, as well as the look-up tables for the Annual Energy Outlook price projections, as well as property data projections contained within the GREET model. With that, I'd be happy to take any questions.

Eric Miller:      
All right. Thanks, Brian. We have a number of questions that were typed in during your presentation. I think we've got several minutes to go through most of them so let's take them one at a time. The first question is as follows. "I have a project idea for an offshore platform for hydrogen production with energy from wind, sun, tide, and hydro gravity. Where can I find help?"

Brian James:   
Where can you find help? Hm. That is a broad set of technologies that you have there. I would direct you to the DOE websites, derivatives of the various websites shown there at the last slide that have case studies for each element of those. I think there are also DOE documents for various constituent technologies—say, from the wind office—that would have perhaps other parameters.

But holistically, you're gonna have to combine the H2A analysis with side calculations to make sure that all of the pieces fit together. There's going to be a lot more specification required for a complex technology system like that that combines elements, than the need to go into the capital costs and the manpower aggregate numbers for the H2A page. Eric, did you want to add anything to that?

Eric Miller:      
Yeah, well, maybe Brian, let's kind of postulate that maybe this is an electrolysis process that's utilizing these other renewable resources to power it. Is H2A model modular enough to be able to construct an analysis that would be useful to the questioner?

Brian James:   
Yes. So, in that particular instance, if it was an electrolysis project with electricity generated from various renewable sources, then the way the H2A model would interpret that is just a standard electrolysis case—or however standard you want to make it on an oil platform—with a cost of electricity entered not from the EIA look-up tables but with a typed-in cost of electricity that would be calculated separately from your estimation of what the wind cost would be or the solar cost or whatever you want. So, you would use the H2A with a custom schedule for electricity cost. That's generally how you would approach that.

Eric Miller:      
Well, that leads right into the second question which is, "For our feedstock, we plan to use exclusively wind power at night, which has a cost near zero. Is that cost input accepted in this model?"

Brian James:   
Yes. So, you would, once again—and I don't know if I can clearly show it in the table here—when you're adding in the feedstocks, you can add in a utility feedstock—or industrial, it doesn't really matter which one you add in—but you're going to be changing this Look Up Prices from a yes to a no. And there's a way to do it, when you click on it you can say no. When you do that, then it will prompt you for the price of electricity. This cell, instead of being a blue value from the data tables in 2007 dollars per kilowatt-hour of electricity, it will change to a peach color where you type in a constant price of electricity.

And then it will use that. Obviously, if it's zero, it doesn't mean anything. But to support whatever value of electricity you put in there—and I think it has to be a constant value rather than a schedule—you need to do a side analysis showing your assumptions and justifying the use of that electricity price. That will be a key component to it because obviously, with electrolysis, the price of electricity is the key determining factor of the cost of hydrogen and so you have to have a solid basis for that price of electricity.

Eric Miller:      
All right. Thanks, Brian. Let me go through a couple of hopefully quick questions then I'll come back to the more lengthy ones, or the answers may be more lengthy. Let's do the quick ones first. One question is, "What does forecourt mean or stand for?"

Brian James:   
So, the forecourt is the term of art for current gasoline stations. I think it specifically refers to the tarmac where the cars pull in to be refueled. So obviously, you're just dispensing gasoline in the gasoline station from an underground tank. In our forecourt stations, you're actually going to be—for these production cases for forecourt—you're actually producing the hydrogen at the station itself and then storing it and then dispensing it into the vehicles. So, forecourt is meant to be the roughly 1,500 kilogram per day dispensing station where hydrogen is also produced at that location.

Eric Miller:
All right. Let's go to the hopefully very quick one. "What is the base case required of pressure for produced hydrogen?"

Brian James:   
300 psi.

Eric Miller:      
I thought you'd get that one pretty fast. I'm just trying to see if there's any other quick ones. Okay, so here, I'm gonna give you a little more challenging question. "Slide 15—why does the spreadsheet use reference years from 2007 and 2005? Shouldn't the spreadsheet model be updated to recognize 2013 as reference year for its input data?" You knew that was gonna come up, right Brian?

Brian James:   
Yeah. It very easily could be. It was somewhat of a legacy because it was an update of a model that had been done earlier and therefore, I think the years of the analysis were more maintained to allow comparisons to other cases. But it would be a—because of the automation within H2A, it would be a trivial update to change the date of the date years.

Eric Miller:      
That's the reference year, right? So, that's not the base year though, right?

Brian James:   
Excuse me, the reference year. Yes. I misspoke. Not the—specifically not the base year. The reference year is what I meant to say.

Eric Miller:      
Okay. We'll see if more questions come up on that one. Let me give you one I think you can answer with a little bit of finesse. "Does the DOE $2.00 per kilogram target reflect approximately gasoline parody for current generation of fuel cell electric vehicles? If so, where is SMR and what is next?"

Brian James:   
SMR is—I was going to say it's north of that, but I'm not exactly sure what the actual value is for SMR right now. Remember that these are production prices of hydrogen, not necessarily the delivered pump price which would have to have delivery and dispensing added to it. So, when you add those costs in—and even from SMR—it tends to be more expensive than gasoline, at least on a per Btu basis. Beyond that, I don't know that I would add much.

Eric Miller:
Okay. "Can H2A be utilized in residential units?"

Brian James:   
Yes. I have—it's not a standard use of the model but at its heart, as I've said multiple times, it is merely a discounted cash flow analysis with input parameters that can easily be changed to correspond to a residential arrangement. You would just have to go through each of the assumptions and modify the input parameters to reflect the residential scenario. So, you could change the installation cost. You could change the lifetime. You could change the maintenance schedule—all those parameters could be changed to reflect a residential arrangement.

Eric Miller:
And as a follow on, I think, to this is, "What is the cost of such a system, for example, at a capacity generation of five kilowatt…"—of power I think maybe you're alluding to—"How do you find the cost of such a system or how do you input that cost?"

Brian James:
I'm not sure I understood the question. Is that the capital cost of the small system?

Eric Miller:
I'm guessing what is needed to do the modeling, right, for the H2A case?

Brian James:
Okay. Well, frankly, what I would consider the trickiest part of filling out an H2A case study is coming up with a credible basis for the capital cost of the equipment. And that is—although it's entered into the model as a series of rows with some level of detail, the model typically really only cares about the bottom line installed capital costs. So, all one needs is the cost for H2A purposes, but for credibility and confidence that you're going to get the right results in terms of hydrogen price or hydrogen costs, you have to do your homework on the capital costs. And that's typically an external analysis to further break down the elements within your system and to realistically assess what those capital costs would be. So, it's not really an H2A analysis, it's a side analysis.

Eric Miller:
Okay. Just a couple more questions. We're running—I think we're on good time. You may be able to get to all of them. Here's a question. "How do you go about obtaining and vetting detailed process information for a given technology and do you have trouble with commercial vendors and sensitive and/or proprietary information?"

Brian James:
When we have done case studies in the past, we speak to a lot of people that are involved in that technology. Sometimes we deal with proprietary information and sign NDAs, and most of the time we don't. I personally find that what is important in the H2A cost analysis is to get the big numbers correct or not to be too far off. So, you don't necessarily need to know the intricate details of the process. You just need to distill it down to a parameter or a value that captures the true cost aspect.

So, as such, you can talk to a vendor or a merchant and you can have a general discussion with him that usually can be done at a non-proprietary level. You might start off with him saying, "I can't tell you something," but after a conversation, you can bring it up a level in less detail to where you can reach agreements on a number that's representative.

Eric Miller:
Okay. And this may be a follow up, but a little more difficult, I think. "Process inputs for cases such as SMR and electrolysis seem straightforward, but how do you go about doing case studies for the longer-term production technology such as the solar cases, which haven't been demonstrated out on the process level?"

Brian James:
Yeah. They definitely are more challenging and they require often more side computations in a variety of approaches. A lot of times, we will puzzle over what the capital costs should be of a system and I have found that really, the only way to have moderate confidence in the result is to start making specific design assumptions. For instance, when we looked at the biological systems, one had to define what the beds will look like and how deep they are, what the pumps looked like, what the layer over the top is, what the material is. When you start specifying those kinds of conceptual design details, it moderately quickly shakes out as to where the costs are and what's important and what isn't.

So, it's definitely harder on the embryonic cases and the only thing I could say is that you gotta dive in. And instead of talking in generalities, you have to start picking specific approaches. And by picking specific approaches, then you can stop waving your arms about different things and you can be more confident in the results. An example is—another example is—the PEC system where we originally tried to do one case study for PEC and we quickly saw that it could be a colloidal suspension or an electrode, and the electrodes could either be with concentration or without concentration, and the colloidal suspensions could be a single bed or a dual bed. It was a fool's errand to try to come up with a single case for PEC that would encompass all of those. Only by specifying—going down that road and doing the specific design in enough specificity to narrow the field, were we able to come up with numbers that didn't have a huge range of variability.

Eric Miller:
Right. And yeah, there's a lot of flurry of last minute questions. We'll get to a few more. Maybe following right on with the PEC example, there was a question, "Is there any specific way to estimate the pricing with respect to the conditions in India and currency, especially for the PEC hydrogen generation?"

Brian James:
In regards to India? Is that what the question is?

Eric Miller:
That's what I have. I suspect you would need to look at maybe the sun conditions for one thing on one side and also look at maybe currency translation type.

Brian James:
Yeah. H2A is able to encompass all of that because you just have to ask yourself, "How does its location in India impact things?" If there are reduced labor costs, then there's a row in there for the dollars per hour so you put in the appropriate labor rate. If there is a different cost of land there, then it's that. If there's a specific location for insolation—solar insolation—then you can look that up and put that value in there. If it's raw material costs, then those go into the capital costs. So, it should be fine. You just perhaps would find yourself using non-H2A default values in a lot of parameters, which is okay, as long as there is a strong rationale for why you're deviating from the default values. You have to explain that.

Eric Miller:
Great. And you know, I think we're running out of our hour here. Let me give you a real quick yes or no question. Is there a method to include California carbon credit values in the economic analysis?

Brian James:
Yes, there is. There is an assumption row in the Input tab basically for byproduct credits and you can basically finagle that cell to basically be any carbon credit you want or any other type of credit. So, there's a way to do that.

Eric Miller:
All right. Thanks, Brian. Now, I'm gonna hand it back over to Alli since we've run out of time. Sorry I didn't get to all of the last minute questions but she'll tell you how to get answers to those questions.

Alli Aman:
Yes, thank you so much Brian for taking the time to give the presentation today. Just a few housekeeping items to wrap up. Today's webinar was recorded so we will have that, along with slides, posted to our website in about 10 days, and I encourage you guys to all check back to our website for future webinars. We are having another one at the end of this month so definitely check back to our website for details on that. And that concludes today's webinar. Thanks again, Brian.

Brian James:
You're welcome.