Below is the text version of the webinar titled "DOE Analysis Related to H2USA," originally presented on July 24, 2013. In addition to this text version of the audio, you can access the presentation slides.

Sunita Satyapal:
[Audio starts mid-sentence] …companies typically have internal models that cannot be shared publically while the focus of the DOE model is on transparency and accessibility of the analysis as well as the assumption.

[Next slide]

So if we go to the next slide as a quick overview of H2USA, which many know is being formed as a public-private partnership among DOE as well as other federal agencies and state agencies, as well as industry including auto makers, hydrogen suppliers, academic institutions, and other stakeholders.

And the main objective is to promote the widespread adoption of fuel cell electric vehicles. There are a number of activities that are planned. And the main focus is on enabling the hydrogen infrastructure, which will allow the widespread deployment of fuel cell electric vehicles. So as you can see on this slide—and once again all of these slides will be provided on our website—there are a number of initial proposed activities within H2USA, such as forming a strategy for say fuel cell vehicle and hydrogen infrastructure rollout, looking at synergies between hydrogen and other fuels such as natural gas, and evaluating business cases that are required for commercialization of fuel cell electric vehicles.

Now before H2USA was announced we had a preliminary analysis meeting among technical experts from the DOE national labs as well as academia. And I will now turn it over to Todd West from Sandia National Lab on slide four to cover the results of the initial analysis meeting.

[Next slide]

And then we'll hear from Fred Joseck, the systems analysis lead from the Department of Energy Fuel Cell Technologies Office. I'll turn it over to Todd.

Moderator:
Todd are you there?

Sunita Satyapal:
So while we are waiting for Todd if we go to slide four, as a summary there was an analysis meeting that was held at the National Renewable Energy Lab in March to assess the information and modeling requirements for a coordinated roll-out plan for fuel cell vehicles and hydrogen infrastructure across the nation. The purpose was to assess the models and tools that are available that would support the market launch. And we also identified and collected feedback on a number of specific areas such as what is the information and what is the data required to conduct the studies, and assessed the specific models that are needed for an infrastructure rollout market launch, for example, what type of information and what types of tools are necessary for in-depth business case assessments. We also looked at where the gaps are in terms of the models and tools and analyses, and identified next steps in organizing the team as well as potential critical workshops with a broader group of stakeholders to complete the analysis. So we will go onto the next slide and see if Todd West is available.

[Next slide]

Fred Joseck:
Since Todd is not available this is Fred Joseck. I'll provide the overview of our modeling. We look at the hydrogen and fuel cell vehicles as part of a portfolio of technologies to increase the energy security and reduce greenhouse gas emissions and criteria emissions. And we look at the hurdles for fuel cell vehicle sales, which are relative to expanse of market, as well as the hurdles for hydrogen infrastructure. We have a wide array of analysis capabilities to provide market research on location and number of potential buyers, optimizing refueling infrastructure and locations to match the potential customers, and determine the financial scenarios necessary for successful hydrogen infrastructure and fuel cell deployment. And we can also assess the risk and uncertainty around mitigation.

[Next slide]

I'm going to move to the next slide, which is slide seven, and talk about our analysis framework.

[Next slide]

Slide seven provides an overview of the modeling framework that is used by DOE analysts for their projects. The analysis inputs consist of consistent and transparent input from multiple sources such as Energy Information Administration. And the inputs are utilized by a portfolio of analytical tools, which are exhibited in this figure.

The tools range in applicability and complexity, which we'll be explaining in our following slides. The tools and models are used to conduct a wide array of studies and analysis to support DOE Fuel Cell Technologies Office. This same set of tools could be used for analysis outside of DOE could support the modeling requirements for entities such as H2USA. These models as well as assumptions have been peer reviewed and vetted. And the models also benefit from multiple peer reviews and input from the stakeholders.

Before the results are published the preliminary results are vetted back through stakeholders and subject matter experts. The final results and deliverables can be in the form of recommendations; input for policy decisions, such as impact on investment tax credits; and inputs for plans such as the FCTO Program Plan and guidance on key areas of R&D. The details of the modeling structure will be presented in the next slide.

[Next slide]

This slide provides a hierarchy of our models and analysis. The framework consists of the base models, which are component models that most of the other models and tools are built on or based on. And most of these other upper-level models use results from these component models. Some examples are the H2A model from NREL and the HDSAM delivery model from Argonne, which you see on the left-hand side. There are other component models as well, but for a matter of time we're going to move on to the next level—which is around the vehicle penetration modeling, which utilizes component models or their outputs to examine multiple vehicle platforms such as PHEVs and fuel cell electric vehicles, internal combustion engines, battery electric vehicles, and compete each other with their attributes to be able to understand the penetration of these vehicle platforms as they compete.

Examples of these vehicle penetration models are the Oak Ridge MA3T model and HYTRANS model, which you'll hear more about shortly. The benefits of the various vehicles can be evaluated by the environmental life cycle assessment and are shown on the left of the triangle. These benefits include the reduction of greenhouse gas emissions, petroleum use, and criteria emissions. And the total benefits can be rolled up based on the vehicle penetration for a given year. And examples of these environmental models and life cycle assessment models are the GREET and VISION models from Argonne.

Based on the understanding of vehicle penetration and fueling infrastructure requirements, market assessment modeling can be completed. These models enable assessment of infrastructure build-out, and fueling station placement relative to fuel demand and traffic flow. Examples of these models are to the left of the triangle again and they consist of the Oak Ridge MA3T model, NREL SERA model, University of California-Irvine STREET model, and UC-Davis model, which will be discussed in the coming slides.

Following infrastructure demands and build out, we look at the financial and employment impacts. And at this level the financing requirements, cash flow, investment needs, and impact on capital or debt financing can be evaluated. This information is needed by investors as well as government and companies and other entities planning or making key investment decisions. Examples of this modeling capability of conducting this analysis are through the SERA model of NREL, the STREET model, the Energy Independence Now, or referenced as EIN, and the University of California-Davis and Argonne JOBS models.

Finally we look at the pinnacle where we're looking at complex analysis of policy impacts and scenario modeling when conducting or examining policy impacts and understanding multiple variables through scenario analysis to understand the limits of potential ranges and bounds of analysis. And the impact of potential policies can be evaluated to understand effectiveness of vehicle penetration and infrastructure development. And examples are the SERA model, the UC-Davis, HYTRANS, and STREET. In some cases certain models can be used to perform multiple modeling tasks, which we're seeing with the cross cut. And each can focus on key aspects, and overall they complement one another.

Now we'll give you some more details about each of the models. You see that we have a wide portfolio of models and some of the other modeling outputs. We'll be discussing the applicability to infrastructure and we will provide some examples. The first model to be discussed will be the H2A model. This will be a component model which Marc Melaina will be discussing. Marc?

[Next slide]

I'll fill in for Marc and talk about the H2A model. The Hydrogen Analysis model, or H2A model, is the main cost model used to estimate cost of producing hydrogen. And the H2A calculates the cost of hydrogen in dollars per kilogram using a standard and transparent discounted cash flow framework. The framework allows for consistent comparisons across multiple production and delivery technologies used to inform technology and R&D priorities as well as to access market viability.

The H2A was first developed in 2002 with input from the hydrogen stakeholders in industry and government and academia and has since expanded to include multiple models including production models, delivery components, HDSAM, which Amgad Elgowainy from Argonne will discuss shortly. The hydrogen production costs from H2A production model cases are shown in the figure on the right. The results are indicated from a central production and provide you with—and also production onsite for hydrogen fueling stations labeled as forecourt costs.

Current costs are for near term technology. Future results indicate estimates for longer term production technologies. The costs are shown in dollars per kilogram with one kilogram having approximate energy content as one gallon of gasoline. For comparison fuel cells are roughly twice as efficient as today's gasoline vehicles, so dividing by two of an estimate of comparable fuel costs per mile driven. Essential production technology costs range from $1.50 to $2.00 per kilogram but will include the cost of delivery for delivering hydrogen to the vehicle.

Forecourt on the onsite hydrogen costs tend to be higher but do not require hydrogen to the station. At the bottom of the page it gives you an idea of the process flow for the H2A model. The model is user friendly and has the ability to take input variables. It's a discounted cash flow model and it is publicly available as indicated on the right and gives you the URL. There was also a webinar that explained the application and details of the H2A model, which you can get more details on.

[Next slide]

The application of the hydrogen cost analysis is shown for distributed reforming. And this gives you an idea of how we apply the H2A model. This slide—this was for distributed reformer and provides the total cost of production, station compression, storage, and dispensing. And this provides you the total cost of hydrogen production and then on the right-hand side is just the hydrogen production without compression, storage, and delivery. But we examined multiple natural gas cases for years, and that was the inputs to being able to examine what the impact would be on the cost of hydrogen produced. And it's on a levelized cost basis. So this is an application of the H2A model.

Next we will go to Marc Melaina. Are you on the phone? Next we'll go to a cost calculator that was developed to be able to examine the cost of a fueling station. And it's on the same platform as a…

Marc Melaina:
Can you hear me now Fred?

Fred Joseck:
Yes.

Moderator:
Yes.

[Next slide]

Fred Joseck:
I'm sorry Marc. Now we'll go to the HDSAM model and Amgad.

Amgad Elgowainy:
Thank you Fred.

Fred Joseck:
Amgad Elgowainy of Argonne National Laboratory will talk.

Amgad Elgowainy:
Thank you Fred. HDSAM stands for Hydrogen Delivery Scenario Analysis Model. It has been developed with collaboration from NREL, PNNL, as well as Argonne—the Argonne team. It was meant first to be developed at the component level, mainly looking at delivery components, such as trucks, the liquefier, the terminals, the refueling decisions. And then based on a market scenario that covers all urban areas in the U.S. with population of 50,000 or more, and based on the market scenario you can size the components and tie them together to estimate the delivery costs and split it into capital, operation and maintenance, and also provide cash flow by component and the total.

The model attributes are that it is capable of sizing the components, delivery components, in particular the refueling decision is a key component based on engineering fundamentals, and also estimate the performance and with the cost data and the sizing of the component we can estimate the cost of components and then project that into levelized cost with the capital and the cash flow. The model in the past year has been evolved into the SAE J2601 protocol by tracking pressure temperature and mass between the delivery components although [inaudible].

[Next slide]

In this slide you see a depiction of how the model works. On the left you will see—the top left—the definition of the market, the fuel cell vehicle penetration. And then at the bottom left you will see the delivery mode, the trucks or pipeline, gaseous or liquid. And as you see in the middle section at the bottom, once you specify the dispensing option as far as pressure of the refueling system configuration, all of that tie together into sizing the component as you see in the top middle of the slide, and also evaluating the cost.

And in the far right you will see the type of cost estimates by capital, operation, energy. You will see in the middle this refueling decision by top component, what contributes to that. And you see also at the bottom there the cash flow by component and the total. As Sunita mentioned the cost data are mainly from vendors and from input from industrial experts. And the modeling and analysis are also vetted by tech teams and experts around the industry. Next please.

[Next slide]

So this slide shows a sample analysis. At the very bottom you will see a depiction of a refueling station that is capable of 900 bar. There you will see the cascade levels to dispense into the vehicles that we pull in. Next please. So here you will see that the sizing of the components will depend mainly on the capacity of the station as well as the peak demand as you see in the right figure. During the peak hours how many back-to-back vehicles could be filled in each hour is key for sizing the component. And on the middle left you will see the tracking of the mass pressure and the temperature following the SAE J2601 protocol. Next please.

So after sizing the components to satisfy the refueling profile we will be able to predict or estimate the capital cost, as you will see on the vertical left axis there. Each bar represents a capacity and daily dispensing capacity. And you will see sections of the model which show the contribution by component. On the right vertical axis you will see the contribution of the refueling decision to the dispensed hydrogen cost in dollars per kilogram. And now I turn it back to Marc Melaina from NREL.

[Next slide]

Marc Melaina:
Great. Thank you Amgad. Can you hear me okay now Fred?

Fred Joseck:
Yes.

Marc Melaina:
Okay, thanks. So I'm going to talk about the hydrogen station cost calculator. The cost calculator is a simplified version of the H2A Model that Fred Joseck described earlier. And we used this model as part of a workshop that we held in 2011, the Market Readiness Workshop. We had about 60 people attend the workshop and provide input on opportunities to reduce costs in hydrogen stations. And then we used the hydrogen station cost calculator as a follow up with a subset of experts who had experience installing stations to get quantitative feedback. And then we combined that with the workshop proceedings.

So a key part of the hydrogen station cost calculator is the four station types listed here in these bullets. The first one is State of the Art. This is a category of station that's operational in the 2011/2012 timeframe, with the most recent generation of major components for those stations. So we expect those costs to be higher and in each of the next categories we expect improvements in the station types.

So the second one is the Early Commercial stations—EC. This was probably the most important one to identify clearly. So to the respondents we described this as a station installed within the next 5 to 20 years. The station type would be financially viable with little government support. It would support growing demands in a promising market region, adequate return on investment, and the design of the station would be replicable. So that's sort of an anchor point for early commercial deployment. And then we have two more categories that go beyond that.

The first one—More Stations—is the same type of station but deployed in larger numbers, or greater number of units. And then the fourth and final category is Large Stations deployed with higher capacities—same design but higher capacity. As I mentioned this was distributed to a select number of experts with direct experience. And if we go to the image here on the screen there's a screen shot of the calculator itself, what it looks like in Excel. And you can see that there are four columns for each of these four station types. And then each row has different cost attributes so that the respondents can compare their responses for each element.

Then at the bottom once they've filled out these values you can click calculate and it gets you the resulting cost of hydrogen. You can see how your responses vary between station types. We had a third independent party, IDC Energy Insights, actually administer this cost calculator to those respondents. And they weighted and aggregated the responses to protect the anonymity of the people who responded and then presented the aggregate results. Those are included in a report. If we click one more time it shows the cover of the report and the URL where you can access that report. I have a couple more slides. We can go to the next one to show some of the results from this calculator.

[Next slide]

So as I mentioned it gives us a simple cost of hydrogen, dollars per kilogram, in aggregate from all the different respondents. So in this figure the SOTA station or State of the Art station costs are around $20.00 per kilogram, which is not too surprising, it goes off the chart here. And what we really wanted to know was how much the reduction would be from that current generation of technology down to the early commercial.

You can see here that the results of the calculator suggest that the cost for early commercial stations would be about $6.00 per kilogram and the size of the stations would be about 450 kilograms per day, that's the horizontal axis along the bottom. This cost per kilogram represents capital costs and fixed operating costs for the stations but it does not include some of the variable or upstream costs that might be associated with that type of station. So then we can move on to our other two categories shown on this figure.

The next one is the More Station or MS station. We see a 19 percent reduction in the cost of hydrogen and again this is when those stations are mass produced, a greater number of units, number of stations. And we actually got feedback that the size of the station would increase as well as we reach that technology development level. And then our furthest out, sort of longest term technology category was LS, Larger Stations. We see another 27 percent reduction down to about $3.50.

Those are some of the simple results that we got from the cost calculator and I have a few more results on the next slide.

[Next slide]

Some of this is a little bit more elaborate because we were able to tease out of the results from the calculator a learning function. Here we see three different curves that again capture economies of scale for the size of the station. The horizontal axis is the size of the station and you see costs—this is capital cost for capacity declining as you get to larger station sizes.

The red line is for the Early Commercial station type. And you can see the large red dots show that capital cost for that station size. And then if we move to the next one, the green triangle, you can see the reduction as we move from the red line down to the green. That reflects both economies of scale for a larger station and the learning and experience that happens from putting out more units, higher volumes, and economies of scale to shift from the red line down to the green line.

Then to go out even further the line shifts down again and we move out on the horizontal access to the bottom right to the larger stations. This is an aggregate learning function experience curve for this technology category of hydrogen stations. And we use this as a quantitative result that has a lot of different things rolled into it—not just technological advances but a lot of other cost reduction opportunities that were identified by the participants at our workshop.

We have a few of those listed on the right—the bullets. That includes compressors as a technological improvement, but also facilitating codes and standards for different equipment, reducing installation costs. And then the last bullet here on the slide emphasizes that efforts to increase the utilization of individual stations is really key. That's mostly for the dollar per kilogram results. That's important for the early station rollout when we expect some stations to be underutilized as the vehicle market expands.

That was another opportunity reduction that was identified during the workshop. Now I want to turn to the—those are the only slides I have on the calculator so I want to turn now to Tyson Eckerle from Energy Independence Now who's going to talk about his cash flow model.

[Next slide]

Fred Joseck:
Okay, I'll fill in for Tyson. This is Fred Joseck. The cash flow model was developed in collaboration with California Fuel Cell Partnership. It was to be able to identify some of the financial risks that were associated with infrastructure development. We'll talk about some of the station cash flows in a moment and give you some examples. But it's a spreadsheet model and it's to be able to understand some of the solutions for incentives and also delays in the station construction, also any financing opportunities.

And also—talk about the capacity utilization of the station as well so you can be able to examine when your cash flow is going to become positive for the station. Also some of the other characteristics—you could look at the CapEx, capital expenditure, and operating expenditures, impacts of grants, and also loans and tax incentives. I'll go to the next slide. Tyson?

Tyson Eckerle:
Yeah can you hear me?

Fred Joseck:
Yeah, okay. So we'll move to the next slide.

[Next slide]

Tyson Eckerle:
Great, thank you. This is the conceptual diagram of our cash flow model. On the top we have inputs which have the market scenarios that can be varied: vehicle sales, fueling patterns, and prices, and the infrastructure build out and cost, which a lot of it leans on the hydrogen station cost calculator data that Marc Melaina has put forward before me. And also investment package scenarios. We can explore a variety of combos of grants and CapEx grants, operation support, debt, and tax incentives.

And what that does is it allows us to look at key outputs on the left side, the station, return on investment and internal rates of return at present value and the impact of incentives on the station's cash flow. And on the right side, kind of network wide funding needs for support and build out. We have two kinds of snapshots from the models. One, on the left, is that single station cash flow, just showing what it looks like on a quarterly and annual basis from an individual station owner perspective.

And that will vary based on the market scenarios and investment package scenarios that we feed through the model. We can look for optimum solutions to make it more attractive for private investment. And then on the right side is the snapshot of what would be government investment in our model or large private investment for how much it would take to actually put that incentive scenario forward and develop stations at the level that we decide in the model makes sense.

It allows us really to balance both the private side and the public side and try to find the optimum solutions. That's our cash flow and incentive model in a nutshell. And with that I'd love to hand it over to David Greene, corporate fellow at the Oak Ridge National Laboratory.

[Next slide]

David Greene:
Thank you Tyson. Good afternoon. I'm going to introduce two models. The first is the Market Acceptance of Advanced Automotive Technologies Model which was developed by the Department of Energy's Vehicle Technologies Office to analyze the potential impacts of 20 different advanced alternative passenger car and light truck drivetrain technologies, which includes fuel cells and plug-in fuel cell vehicles, and evaluate them in terms of cost, fuel savings, and reductions in petroleum use and greenhouse gas emissions.

The key features of the model are its consumer heterogeneity. It has 1,458 market segments which are, for example, differentiated by region—rural, suburban, urban location; types of daily driving distribution (how many miles each day for each vehicle); and by housing type, which is obviously important for plug-in vehicles. It uses rigorous technology projections for all the technology types. They're typically based on runs of Argonne National Laboratory's Autonomy Model.

The Vehicle Choice Model, which is a nested multinomial logit Vehicle Choice Model. The parameters are derived from fundamental behavioral assumptions to increase the transparency of the model. That is to say, for example, the value of range or the cost of limited range is dependent on the value of people's time and how long it takes to recharge the vehicle, what the energy storage is on board, what the efficiency is. That is to say range, per se, not limited range and long recharging time.

That's handled a different way, again, derived from basic behavioral assumptions. The MA3T Model is continuously updated and enhanced. It's been calibrated using historical sales of hybrid vehicles. It does a pretty good job reproducing the past ten years of sales—not perfect, but pretty good. And it's publically available. You see the URL on the slide.

[Next slide]

The next slide shows one particular example of the use of the MA3T Model. It's an analysis of the impacts of meeting DOE's technology R&D targets on the market's success of hydrogen fuel cell vehicles.

So here we are testing the impact of meeting these R&D targets and the MA3T Model is predicting what the market shares will be as a result. In this particular graph you see here the PG stands for Program Goals are met. The base represents a much slower rate of improvement in cost and technology performance. In the base case the technology targets are not even met by 2050. And all the technologies, all 20, are included in the analysis so that the fuel cell technologies are competing against advanced versions of the other technologies, except in the cases labeled base and FC, where in the base and FC case only the fuel cell vehicles meet their technical targets.

Hydrogen refueling infrastructure is assumed to be provided in all cases but the deployment of infrastructure is slower in the RefInfra cases, illustrating the importance of infrastructure deployment on market success. This study, which is published in the International Journal of Hydrogen Energy, has a couple of key conclusions I think. One is that fuel cells compete successfully against other advanced vehicle technologies, even if all of the technologies meet their technical R&D targets. And that delay in meeting the goals does indeed delay market success. It reduces the petroleum reduction benefits and the greenhouse gas reduction benefits. But still the fuel cell vehicles achieve the same market shares at the end of the time period in 2050. Next slide please.

[Next slide]

The next slide deals with the HyTrans Model. The HyTrans Model was developed to integrate in a competitive market framework the comprehensive analyses of hydrogen supply, delivery, vehicle manufacturing, and consumer choice carried out for DOE's 2008 transition study.

Many of the same team members presenting today, colleagues from NREL, Argonne, and the University of California-Davis, were key members of that analytical team. And many of the inputs to the High-Trans Model come, as Fred pointed out early, from the models you're hearing about today such as H2A, HDSAM, GREET, and so on. And by integrating these studies in a dynamic market model, HyTrans produced a comprehensive market-based scenario, creating a vision of the transitions of hydrogen vehicles in the U.S.

HyTrans is not a geographically detailed model like the SERA Model you'll hear about later. The inputs in modeling analysis were extensively vetted with the stakeholder community during the study and before the final report. It is a large, non-linear optimization model. And because of the many positive feedback loops in this process of transition there are multiple optima. There is path dependency, there are tipping points. All of these things make it a challenge to solve the model.

It's programmed in the GAMS language, which is a commercial language, but costs a few thousand dollars. As a consequence of all of that the model is available to researchers who are interested, but we've actually had very few requests to date for people to obtain the model. The lessons learned from the HyTrans Model building, as well as from earlier studies of transition to hydrogen such as the NRC study A Focus on Hydrogen inform the development of the LAVE-Trans Model, which was used in the recent NRC study Transitions to Alternative Fuels and Vehicles.

LAVE-Trans is a lighter Excel model that's only partially optimizable and more useful for simulation analysis and scenario construction. Next slide.

[Next slide]

This is one example of outputs from the 2008 DOE transition study. HyTrans was used to demonstrate how an economically efficient transition to a self-sustaining hydrogen and fuel cell vehicle system could be accomplished. All of the economic factors were costed out, including the barriers: things like risk aversion of the majority consumers, lack of scale economies, learning by doing, lack of diversity of make and model choices, and lack of fuel availability early on.

Numerous policy options—you see them here on the right-hand side in the table, it's not really possible to explain all of them at this time but they include various vehicle and fuel subsidy strategies—and just a variety of policy packages which are possible were evaluated using the model. And in the end the excess cost of the transition subsidies and so on were on the order of $40 billion to $50 billion, very similar to those found by other studies such as the NRC's Transition to Alternative Transportation Technologies—a Focus on Hydrogen.

Much of that modeling work was done by Professor Joan Ogden whom you'll hear from shortly. That concludes my presentation. Thanks for your attention.

[Next slide]

Now it's my pleasures to introduce Tim Brown who is a senior scientist in the Advanced Power and Energy Program at the University of California at Irvine.

Tim Brown:
Thank you David. I'm glad we were able to hear you and I hope everyone can hear me all right. I want to talk about the Spatially and Temporally Resolved Energy and Environment Tool called STREET. This is a suite of software packages that was developed here at the Advanced Power and Energy Program at UC-Irvine through extensive collaboration with automakers, energy companies, and other stakeholders. STREET was originally funded in part through the DOE's California Hydrogen Infrastructure Project and continues to get support in development from both automakers and the California Energy Commission.

STREET is a broad holistic energy supply chain model which can evaluate greenhouse gas emissions, criteria pollutant emissions, energy resource consumption, in a very high resolution spatial and temporal fashion. It can therefore evaluate a number of supply chains for various fuels and regions and can go a step further actually and because of the spatial and temporal resolution actually evaluate not just emissions but air quality throughout its—air chemistry transport in the atmosphere.

But today I want to talk about the portion of the STREET Model that can do infrastructure siting and specifically for hydrogen refueling siting. And as shown here in this first step—the first step in that process is to determine cluster regions within an area. And shown here is a map of Southern California showing three broad cluster regions: Santa Monica, Torrance, and Orange County. And these are regions that we've identified through the use of demographic data, OEM fuel cell vehicle market projections, and vehicle sales data as areas of initial fuel cell vehicle high intensity sales in these areas. So these are the regions where we're about to have very robust infrastructure.

So step two is to break each of these regions (shown here in this map is the Santa Monica and West L.A. regions) down into a series of links and nodes. And this is sort of a mathematical description of the STREET network. The links represent roadway sections and nodes represent intersections. And by doing that we can then perform an optimization within this area to understand how many hydrogen stations are required to meet some minimum requirement of service coverage.

And in this case what we're trying to do is represent a six-minute coverage within each of these regions. And this is based on travel time analysis, station land use, vehicle travel density, and service coverage. The third step is to identify secondary markets. These are the areas where we expect to have some initial fuel cell vehicle customers but they don't need to have the robust infrastructure as the primary clusters. And these are selected using demographic data and vehicle sales data.

And the fourth step is to determine the connector and destination locations that can bring these clusters and secondary markets together to form a cohesive network throughout the area, in this case California is shown in the map. And these are selected based on travel times and distances, automaker input, as well as demographic and sales data.

[Next slide]

On the next slide I show that the STREET Model has been used in California successfully to develop a plan that calls for 68 hydrogen stations in the state. And this has been adopted by the California Fuel Cell Partnership and forms the foundation of two documents that they've created both called California Road Map.

It's also been used by the California Energy Commission in the recent grants solicitation for $30 million for hydrogen infrastructure to help allocate where the infrastructure should be placed within the state. The STREET plan has also been adopted by the 2013 ZEV Action Plan which is Governor Jerry Brown's initiative to put 1.5 million zero emission vehicles on the roads in California by 2025. And we performed similar analysis on the island of Oahu and analysis is underway in three additional states as well.

Shown in the top left is a map of Northern California showing the main cluster regions in the area and the stations that we designate need to be in the area according to the STREET Model. The next map shows Southern California and the stations required down there. We'd need about 40 stations in Southern California, 22 in Northern California, and an additional six as connections to destinations. And last but not least there's a map of the island of Oahu and the analysis there showing that there are 15 stations needed to provide infrastructure for Oahu.

And with that I'd like to turn it over to Professor Joan Ogden from UC-Davis.

[Next slide]

She's a professor of environmental and science policy and also co-director of Hydrogen Pathways Program and Institute of Transportation Studies. Thank you.

Joan Ogden:
Thanks very much. Can you guys hear me?

Tim Brown:
Yes.

Joan Ogden: 
Okay great. Thanks Tim. I want to talk today about several models that we've developed at UC-Davis over a period of time to look at the introduction of hydrogen fuel cell vehicles, infrastructure build out—sort of two sets of models. One is focused really on the next decade, on the early introduction. And then there are some other models that focus on the longer term that take out to 2050 and look at transition issues.

First I'll just mention some tools that we've developed which are very spatially specific to look at station network design. And these rely on GIS data and optimization routines to take traffic data, information about population density, about where early adopters might live, and that sort of demographic data, and to look at where optimal placement of stations would be and how many stations you would need in order to provide a certain level of consumer convenience which we measure by travel time to stations. We would use these models to then develop scenarios for station placement over time and look at alternative rollout strategies.

So far we've focused on California but these models could be extended to any other region in the U.S. The second set of models I'll talk about are infrastructure rollout economic analysis where we have talked with the industrial gas companies, oil companies, and others who have built stations, to get good data that we could on near term hydrogen station capital costs and operating costs. And we've coordinated with the team at NREL and also the California Fuel Cell Partnership as well as vetting with industry.

We consider different build out scenarios over the next decade based on this cluster idea which Tim introduced. The idea is basically you co-locate early stations and early fuel cell vehicles in geographically distinct clusters. And that allows you to have a station near where the early adopters live. And then we analyzed the economics from several perspectives. From the perspective of the whole station network, if you were thinking about building this network or supporting it, how many stations would you need? How much would that cost?

We also looked at it from the point of view of the single station owner—many of the stations throughout the U.S. and certainly in California are owned by individuals. So if you were considering putting in hydrogen pumps what would the economics look like for you? And then finally we estimated fuel costs for hydrogen over time as the rollout proceeds and stations scale up in size. This is a rather similar concept to what Marc Melaina presented earlier about learning and sizing up as you go.

We looked at cash flow, breakeven year—that being the year when the station could produce hydrogen competitively with gasoline on a cent per mile basis. And we also tried to estimate subsidies that might be needed to support early infrastructure to bring it to this launch point where it could produce hydrogen competitively. And then we've done some sensitivity studies. So that's an Excel-based model that we've developed, I'll describe that a little more.

And then finally we've looked at longer-term hydrogen infrastructure build out, transition costs, and benefits. These were Excel-based tools, something called the SSCHISM Model that I'll mention that was used in the NRC's 2008 study on transition to hydrogen that David Greene mentioned a little earlier. I guess I'm ready—that's kind of an overview of the models relevant in my talk about this.

[Next slide]

In the infrastructure rollout analysis—this is looking at the early, maybe the next ten years—we did some spatial network analysis to find hydrogen stations and locations. We input GIS data for consumer travel patterns and also possible station sites focusing on early adopter areas that were identified through surveys by the California Fuel Cell Partnership.

So we looked at how you would lay out a network to meet growing fuel demands. The key result really is that a cluster strategy of co-locating enables good fuel accessibility with an initial sparse network. And in the case shown above here we have a total network of 16 stations which is less than one percent of the gas stations in Los Angeles. And the average travel time is on the order of four minutes. The trick is really to put the early adopters together.

We also have done economic analysis for early rollout, developed an Excel-based spreadsheet to model the economics of different station types, varying between 100 and 1,000 kilogram a day, and different types of truck delivered stations, onsite production by either onsite steam reforming of natural gas or electrolysis.

And I show the sample results in the upper right-hand corner there. This shows hydrogen cost over time. And we got some estimates from the industrial gas companies about how they expected the cost to progress both over time with technological improvement and with scaling up size. So we showed different sizes of stations and different years. And we find that initially we're over $10.00 a kilogram, but as we go out in time, both through technological improvements and scaling up from 100 to then 250 and 500 kilograms a day you get down into the $7.00 to $9.00 per kilogram range which would be equivalent or comparable on cent per mile basis to gasoline at about $4.00 to $6.00 a gallon. Maybe click again for me Fred and see if one more picture comes up.

This is another example of what we can do with this model. This is a sensitivity study and the point here is that we started with a base case for onsite SMR and then varied lots of things like the capacity factor, like the assumed prices of natural gas and electricity, and station size. And we were able to look at how the costs depend on a lot of inputs that can vary. Next slide please.

[Next slide]

This is just a picture of a schedule. This again is an input to this rollout economic model for looking at the early years. And here we've set up a schedule, a scenario, where we imagined putting a certain number of stations in over time and building these up, both starting with smaller stations, going to larger stations over time. And the capacity stays ahead of the demand here because we're building stations a little in advance of the vehicles getting there. And over time the demand builds up. By the time you get to 2017 we have a total of 78 stations. This particular scenario is based on compressed gas truck delivery but you could also look at a variety of supply options like liquid truck delivery or onsite reformers. Next slide please.

[Next slide]

This is a result of a cash flow analysis looking at the whole network. And what's shown here in this slide are various costs, first of all the capital costs over time for that 78 station scenario that we looked at. We're building out here—out to 2017. Then we're building stations right along as we go along. We also have O&M costs. Notice that the O&M costs and the hydrogen sales costs (shown in yellow) go up over time. And that's because we assume these stations are not perfectly utilized right from the beginning and that as more vehicles are coming the station becomes more and more utilized.

And this network as we look at—this sort of light blue picture there is the cash flow. That goes positive after about 2017 or 2018. So what this says is with this network that we envisioned on the previous slide, a schedule of growing sizes of stations, eventually you get to a point where you're selling enough hydrogen to make up for the cost. And then the lowest curve there is an integral of that cash flow. So that shows how much money you'd have to be putting out. Eventually that turns around and bottoms out when the cash flow goes positive. And eventually that breaks even.

So what this says is that these 78 stations, which would cost maybe about $110 million to build, will eventually be able to break even, assuming that you can sell the hydrogen at about $4.00 a kilogram more than you pay for it. These models on the cash flow are available. If anybody is interested feel free to contact me about those. We've also collaborated quite a bit with Tyson and his team and certainly talked with the other folks on the call, David Greene, Marc Melaina, and so on, and have used H2A and others as a basis for some of these numbers. Next slide please.

[Next slide]

I'm going to now go back to the national scale. This is a model called the SSCHISM Model, which stands for Steady-State City Hydrogen Infrastructure Systems Model, that we initially developed (Chris Yang and myself) as a part of the 2008 NRC study on hydrogen transitions. And this is a national level model. We consider a rollout where we have a series of lighthouse cities, one after another, where you adopt hydrogen. And then we used some simplified models of infrastructure, drawing on the city population density and other aspects to develop infrastructure estimates.

And what we see is in city after city as we introduce hydrogen we have the cost coming down over time. And after maybe five to ten years it sort of plateaus out to a learned out cost, which is on the order of between $3.00 and $4.00 a kilogram in this model. This model is now—we're now revising this with some better models for the early infrastructure rollout based on the detailed geographic modeling that I presented earlier. And this will be available online and allows one to look across the country with a lot of detail on 73 different cities across the country, looking at the hydrogen rollout.

So I think that's everything. One of the plans—I might mention that at UC-Davis we have a program called the Sustainable Transportation Energy Pathways Project. The hydrogen infrastructure modeling is really part of that. And we're also continuing to collaborate with pretty much all the other analysts on the call. With that I'd like to pass this along—back to Marc Melaina from NREL.

[Next slide]

Marc Melaina:
All right, thank you Joan. I'm going to talk about the SERA Model. SERA stands for Scenario Evaluation and Regionalization Analysis Model. We have a list here of the goals and key analysis questions for the model. I should say that is really NREL's integrating model where we integrated our different sub-models and attempt to optimize on the cost of hydrogen to look at different scenario questions. So we're trying to build on a lot of the model results and lessons and results that we've already presented in earlier slides.

In general what the SERA Model is able to do is generate self-consistent vehicle adoption rates as they roll out across the country at a high level of geographic detail, and then match supply—optimal supply—across any number of different supply pathways and hydrogen production and delivery technologies to meet that demand. So a lot of detail, a lot of temporal resolution. We can do any time step, we can optimize on different metrics. It's an open, flexible model.

And the last bullet here says the goal is really to look at early niche markets and different synergies between different technologies. For example we've looked at combined heat and hydrogen stationary fuel cells and how those might be leveraged for vehicle markets—so it's an exploratory optimization tool. Some of the key analysis questions are which different pathways, meaning production, delivery, storage, and dispensing, are going to be least cost under different conditions out into the future, and especially as the whole infrastructure evolves over time?

What are the different network economies that can be achieved? For example if you have a large production facility serving multiple cities it's going to operate a little bit differently than one near just one large city. The same kind of economies for pipeline networks, and a little bit on the station scale as well. And then once we set up those scenarios we can vary particular technologies to see what makes them more or less competitive.

So if we go to the figure on the right we can quickly walk through some of the steps in a typical SERA study. So the figure at the top labeled A just shows how we can take demand from a fleet of vehicles rolling out over time. The different colors here show urban areas of the U.S. with different numbers of fuel cell vehicles being introduced so we can generate demand based upon that fleet model. And then B, Figure B, shows the resulting distribution of demand and how it would expand over time given different input assumptions about how the market evolves across the country and different regions.

And in Figure C we've provided a lot of detail on the supply side including detailed station representation. So this is Chicago after a high level market share of fuel cell vehicles is being supplied by stations. We want a lot of detail there because of the initial market barriers to do with hydrogen stations; I can talk about that in the next couple slides. In Figure D we do have demand at neighborhood level, zip code level. So here we're trying to build upon the work of Joan and Tim Brown and others to look at early cluster markets. And especially California Fuel Cell Partnership has looked a lot at this so that we can try and have our supply dynamics be responsive to those early market clusters.

Figure E just shows a quick map of how we do optimize delivery pathways with different types to link production centers to demand centers. And then finally Figure F shows how we can do a lot of analysis on the different financing cash flows and metrics useful for understanding the business case.

[Next slide]

So if we go to the next slide I have a couple of results here from different case studies we've done. This first one is basically the scenario we've looked at most. It's taking high level nationwide results, especially from some of the National Academies studies that have been mentioned earlier. If we take that total demand on national level we can break that down to any level of geographic detail we want, allocating that demand, and then determining what the optimal supply pathways would be to supply that demand.

So the map on the left shows the production types. And the green represents natural gas which is typically the winner in the different production types. A little bit of blue is coal gasification without carbon capture and storage. So for this scenario we don't have a carbon signal and we're not necessarily trying to meet any particular low carbon mix of production sources. But because we're able to disaggregate it to a lot of detail the figure in the middle shows that we can roll up demand and supply at a state level.

Here you can kind of see Texas, New York, and California as being some of the bigger states. And as the market evolves over time we're going to see different types of production serving those different states and different regions. And the figure on the right shows the different pathways. This is the truck, pipeline delivery pathways that end up being optimal in any given scenario. So all of that is integrated together for our optimization ranking.

If we click to the next set of figures on the bottom we can see some more results here, especially at the station retail level. We've developed the SERA Model over about eight years and the last couple of years we've really focused more on the retail distribution of stations. So we have various different algorithms and mostly we've been trying to learn from the results from UC-Davis and UC-Irvine for their traffic flow models, where they've determine the optimal location of stations or best location of stations to serve early adopters, and then we've tried to generalize those results to other cities in the country using different correlations.

We've also introduced constraints on the size of stations that would be introduced so that they mimic the existing network of gasoline stations and their size distributions. And that ends up being just the result of basically following the distributional population density in different urban areas. It's a pretty consistent pattern.

On the right we show an example of Chicago and how we identified some of the early neighborhoods. Our top down algorithm places stations to serve those neighborhoods and then starts filling in with connector stations and moving on to the next most promising locations. This top down algorithm does not have the kind of traffic detail as the UC-Davis, UC-Irvine models, but it is able to use demographics and vehicle purchase decision results from historical purchases to look at early adopter neighborhoods. And I should say we are working with a company, KSS Fuels, to try and improve on this, to learn how existing retailers building new gasoline stations sort of tackle similar problems. So I have one more slide on SERA.

[Next slide]

These are a couple more case studies we've done. As I mentioned SERA is an open programming framework. It doesn't just solve in the same way each time. You can set it up to solve a lot of different scenarios if you have different initial conditions and constraints. So this is a study we did just looking at the Northeast corridor states. We developed a demand scenario based upon meeting one method of complying with the ZEV mandate in these states

You can see in these bars demand increases over time. It's color coded by state. You can see New York and Massachusetts are the major sources of demand. And then the SERA Model optimizes the hydrogen supply system to serve those different cities in each of the states. And the figures on the right show some of the results, which is the delivered cost of hydrogen declining over time—the figure on the left. And that's broken out again by state. And then on the far right you can see the cash flow. This is again trying to mimic the same framework that Joan Ogden just mentioned that's been used in the National Academies study reports and elsewhere to look at how quickly the infrastructure can pay for itself as the different markets ramp up over time.

One thing that's interesting here is that SERA is able to see sort of the succession of rollout across different geographies and different regions. I have one more case study on the bottom of this slide. This is a case study for rolling out hydrogen vehicles and stations in Hawaii. We collaborated with Tim Brown on this. He showed a map earlier of some of the initial hydrogen station placements. And then we used a financial algorithm within SERA to look at some of the financing associated with that build out. On the bottom left, to show that we can set different targets for how to balance debt and equity and how a typical investor might behave to finance this effort.

And on the right-hand side we break out a lot of the cost and detail especially at the station level to understand how they contribute to the overall business case. And this is over a long timeframe out to 2050. But you can see the inset there, there's a red cross-hatch which is where in this particular scenario we had assumed that production incentives would be required to some level in order to shore up the revenue stream. And this depends upon your assumptions about the clearing market price of hydrogen, what people would pay for it, and the availability of those incentives over time.

Those are all my SERA slides and I think now we'll go back to Fred Joseck.

[Next slide]

Fred Joseck:
Thank you Marc. You've seen that we have quite a portfolio of analytical models and pools. Many of the results that you've just seen are just snippets of many reports and their modeling results so they're—you weren't presented with full granularity or full vivid views of those models or outflows, outputs, but that can be obtained from some of the reports that were identified. You can go to those. These tools have been developed with a number of technical experts that have been funded by the Department of Energy and vetted through industry and our other stakeholders to ensure strong mathematical, scientific, and economic basis.

We've undertaken the complex analysis and subject analysis, such as infrastructure analysis, for H2USA, and model activities will require specific and well-defined inputs from the stakeholders. The models themselves stand alone and are able to do almost anything we want, and we have the capability of a wide range. However it's really incumbent on consumer information, credible data from stakeholders on vehicle rollout and infrastructure cost, to be able to identify what may happen in a case such as to support H2USA.

Since stakeholder data is very sensitive and confidential we do have the capability to be able to aggregate that and to be able to sanitize it so that we don't give attribution to a single entity. And we've been able to do that through the NREL Secure Data Center. This Secure Data Center enables us to be able to take this data, combine it, and then bring it out in a washed form to be able to use in the models, to be able to do our analysis that would be able to support an effort such as H2USA.

This has also been done for our vehicle demonstration project with the technology validation program. So our next steps would be to continue to coordinate our analysis for the demands and needs of H2USA as they move forward with their analysis and needs to be able to satisfy their goals and their requirements.

[Next slide]

Some of the references are given in the back of the presentation where you can find some of these models. There are the URLs for being able to acquire the model if you so desire to work with it. And as many of the presenters mentioned some of these are publically available and you can acquire these through these URLs.

[Next slide]

Now I'd like to acknowledge the presenters. Dr. Sunita Satyapal, Director of the Fuel Cell Technologies Office. I'm Fred Joseck from DOE. Marc Melaina from National Renewable Energy Laboratory, Dr. David Greene of Oak Ridge National Laboratory, Todd West of Sandia National Laboratories, Amgad Elgowainy of Argonne National Laboratory, Tyson Eckerle of Energy Independence Now (EIN), Tim Brown of the University of California-Irvine, and Joan Ogden of the University of California-Davis.

So I'd like to thank you for your participation. Also thanks to Kristen Nawoj for helping put the webinar together. She did extensive work. As well as Kathleen O'Malley and also the presenters who I've just mentioned. And the presentation will be going up on the website. Kristen is going to be talking to that in a couple of minutes.

Now we'll start with the questions that have been requested during the course of the webinar.

We'll start with the first one, are the webinar slides online? They should've been and we will be making them public.

The next question was, what are the states that are currently undergoing STREET analysis for hydrogen stations. Tim, do you want to address that? Tim Brown?

Male:
Hi my name is Chris [inaudible], I am here with Pete Willette. We're graduate students at UC-Irvine. We're speaking on behalf of Dr. Tim Brown who just stepped out of the room a few minutes ago. And yes we can answer that question. In addition to California we are applying the STREET methodology to such states as New York, New Jersey, Connecticut, and some of the other surrounding states. Were you guys able to hear us okay?

Fred Joseck:
Yes.

Sunita Satyapal:
It looks like the next question is: is this based on a single program or multiple programs to generate the detail? Now we assume that question refers to the various models and tools that were discussed. So the response is that they were funded through DOE activities over a number of years. So what was discussed today was primarily relevant to the Fuel Cell Technologies Office within the Office of Energy Efficiency and Renewable Energy. But these were all DOE related analysis and activities but with of course significant industry and stakeholder input. But they are different models and tools. We'll go to the next question.

Fred Joseck:
One other question was it would make a lot of sense starting with the existing fueling infrastructure where alternative renewable fuels can gradually replace traditional petroleum-based fuels. You're very much on point. As mentioned with some of the modeling such as what Marc Melaina had exhibited when we do our modeling the lowest cost options right now are natural gas reforming and would evolve—but our long-term goal is to be able to drive the cost down through R&D to be able to have the renewable production of hydrogen from renewable resources.

Sunita Satyapal:
Now we have one more question on what is DOE's projection for the current cost of hydrogen per kilogram. The question discusses how the multiple models provide various results. So if we go to slide 10 I think this would be for Marc Melaina. It indicates that $4.50 to $3.70 per kilogram is the cost of hydrogen. That was a high-volume projected cost. And the questioner also points out that slide 15 indicates a higher $21.00 per kilogram. And then another slide relates to $17.00 to $11.00 per kilogram. The question is, is there an effort to commonize the assumptions for current and future projections for the cost of hydrogen? Maybe we can go to Marc Melaina first.

Marc Melaina:
Thank you Sunita. I think we have a lot of different cost results because we focus on cost quite a bit. We've already done the footwork to commonize all the assumptions and different means of estimating the cost of hydrogen, but we do want to explore a broad range of technologies. So that work has already been done and there are various reports on it. I don't know if Fred Joseck or anybody wants to point to some of the particular records but it has been evaluated extensively.

Sunita Satyapal:
Yes and so I think as the questioner also points out there are different assumptions that are based on the volumes, the projection of future technology versus current technology, and so forth. So based on the models and the timeframes across which the analysis was done there are some different—and also of course how the hydrogen was produced—there are some differences in terms of the cost. And we can provide more information on the assumptions that are available on our website.

One other question relates to the interest in H2USA. There we can state that again the partnership is just being launched now. And there will be a website available with more information about H2USA. In the meantime if there is specific interest feel free to email—to send us an email through the website and we'd be happy to respond. And as soon as the H2USA website is up of course that would be disseminated widely. There's one more question.

Fred Joseck:
A question came in asking what forecourt means. Forecourt means the gasoline station or the fueling station. If you had distributed refueling it would be at the forecourt meaning that it would be right at the station. And another question was when the slides will be available. They'll be available in approximately ten business days.

Sunita Satyapal:
One more question relates to onboard hydrogen production and if that's possible. The questioner states that fuel reformers have evolved sufficiently and are we looking at onboard reforming for fuel cell electric vehicles. Here I would like to point out that we did fund onboard reforming significantly in the past. And then in the 2004 timeframe we did make a no-go decision on reforming complex fuels such as gasoline, which was the initial focus before the program shifted more than ten years ago to hydrogen onboard.

There were a number of issues such as startup time, transient response, durability cost, size, weight, and so forth. So the current approach is in the near term really looking at high pressure, hydrogen compressed tanks onboard, and then in the longer term materials-based or hybrid approaches. However, looking at complex reforming onboard is not the current focus within DOE and is not the focus of the analysis that we've been discussing today. So it's really providing the hydrogen at the station for dispensing into the vehicle directly.

Fred Joseck:
Another question is what is the predicted least cost way to produce hydrogen in the next ten years? I'm going to deflect that over to Marc Melaina. You've done some of the modeling in the results you've shown.

Marc Melaina:
Thanks Fred. So we've used the SERA Model to look at a lot of different production types and compete them against each other. Several years back we used to see more heterogeneity. And now with the low natural gas prices that's really the dominant production means. In most of the scenarios we run, unless we introduce some kind of market constraints that ends up being the dominant production type.

Fred Joseck:
Another question came in about support for running some of the analysis. A lot of the tools that are publically available have user guides such as H2A Models and HDSAM. Some of the more complex would be limited as to the amount of materials that are available. But at least on the publically available models there should be some user guides such as the VISION Model and H2A and HDSAM.

Sunita Satyapal:
One more question was, is a station considered only a new green field site or could a station be the addition of a dispenser or dispensers at an existing petroleum station? And how is this distinction handled? There I would—it depends on the model or the tool. I would say that both can be considered a station. It does not need to be only a new green field site. And then I'll turn it over if anyone else on the phone would like to respond to that.

Amgad Elgowainy:
With the model HDSAM it would be just an additional dispenser with refueling equipment or it would be as a different hydrogen station. Both would be a model.

Sunita Satyapal:
And then one additional question was, has anyone looked at CNG or LNG distribution, storage, and dispensing costs and compared them to these models?

Marc Melaina:
This is Marc Melaina. We did do a study comparing those costs over the long term. It was the Transportation Energy Futures study—the infrastructure expansion report. And we have ongoing efforts to collect more information on retail station costs for all alternative fuels.

Joan Ogden:
This is Joan Ogden from UC-Davis. We also have work ongoing looking at natural gas infrastructure, both the passenger car but also in terms of trucks and the LNG infrastructure that's going in now. We're also doing some comparative looks at those things.

Sunita Satyapal:
It looks like those were most of the questions. There are quite a few that relate to, unfortunately, the voice transmission during the webinar. So once again we do apologize for the technical difficulties. Again we will post the slides to our website as soon as possible. And it will take some time to post the recording. But within a week to 10 days we'll have the recording archived as well. Once again I would like to thank all of the presenters, the expert technical stakeholders and analysts who have been developing and implementing these models over the years.

We at the Department of Energy greatly value their participation and their expertise and thank them for participating in the webinar and sharing information in detail. And then we would like to once again thank the audience. We will go ahead and sign off now. We look forward to your participation in our next webinar series. Thank you.