U.S. Department of Energy - Energy Efficiency and Renewable Energy

Fuel Cell Technologies Office

National Hydrogen Learning Demonstration Status Webinar (Text Version)

Below is the text version of the webinar titled "National Hydrogen Learning Demonstration Status," originally presented on February 6, 2012. In addition to this text version of the audio, you can access the presentation slides and a recording of the webinar (WMV 124 MB).

Moderator:
Well, welcome to the Fuel Cell Technologies Office webinar. Today you will be hearing about the National Hydrogen Learning Demonstration status. And our speaker today is Keith Wipke, from the National Renewable Energy Laboratory. And introducing Keith will be our technology development manager, Jason Marcinkoski. Here you go, Jason.

Jason Marcinkoski:
Hi. This is Jason Marcinkoski from the Hydrogen and Fuel Cells Program at the Department of Energy. I'm the acting technology team lead for technology validation. We've had a change in personnel recently. John Garbak has retired. So I'm filling his shoes.

So, introducing Keith – he's the senior engineer and manager of hydrogen analysis at the National Renewable Energy Laboratory, where he has worked in the area of advanced vehicles for almost 19 years. The hydrogen analysis work includes hydrogen infrastructure analysis as well.

Keith is also NREL's representative at the California Fuel Cell Partnership and sits on the board of directors at the Fuel Cell and Hydrogen Energy Association. Today, Mr. Wipke is here to talk to us about work on hydrogen fuel cell vehicles and the Controlled Hydrogen Fleet and Infrastructure Demonstration and Validation Project, which is culminating this year. He has been leading NREL's role on this project since 2003.

So, without further ado, Keith, you can take it over.

Keith Wipke:
Great. Thank you, Jason. It's a pleasure to talk to such a large audience today. I think we had over 400 people register, and it looks like 245 are online right now. So it's a good audience there. And please do send in your questions through the chat. We'll have some time at the end of the talk to go through those. So before I get started, I just wanted to thank the rest of my team here at NREL – Sam Sprik, Jennifer Kurtz, Todd Ramsden, Chris Ainscough, and Genevieve Saur, as well as to acknowledge the support of DOE in making this possible.

[Next Slide]

So I'm going to divide the talk today up really into three parts. The first part is going to talk about the learning demonstration overall, which DOE has been supporting with the industry, and then also to give a station status and fuel cell vehicle status. Then the middle part there is really the meat of it, which are technical highlights of the vehicle and infrastructure analysis results and progress.

And for those non-technical folks on the phone, I do have little boxes on each of the slides that are kind of the key takeaways, so you don't have to worry about getting lost in technical details. And then, finally, I'll wrap up with some next steps and a project wrap-up.

[Next Slide]

So these are the objectives and relevance of the project. The objectives are to validate the fuel cell vehicles and infrastructure in parallel in a real world setting, as well as to identify the current status and evolution of the technology as it evolves over the five-to-seven-year period. In terms of relevance, we were objectively assessing the progress towards targets and market needs, as well as providing feedback to the hydrogen research and development activities, not only within the U.S. Department of Energy, but also within the industry companies themselves. And then, finally, to publish the results for the key stakeholder use and investment decisions.

You can see in the lower left is a table of the key targets from this project. This project is really focused on the second column, which is the interim or 2009 targets. You'll see the asterisk there. This project was originally scheduled to end in 2009, and it was extended for another two years for those participants who wished to continue on. The stack durability target was 2,000 hours. Vehicle range was 250+ miles, and hydrogen cost produced at the station was $3.00/gallon gge or $3.00 per kilogram. And as you can see – the checkmarks – the first two were demonstrated from this project, and the third, the cost, was actually done through an outside review panel. I'll touch on all three of these with a detailed slide as we go through this.

[Next Slide]

Stepping back in history a little bit, back in 2003 the RFP was put together. Actually, in 2003 the RFP came out, and we had four teams that were competitively selected and awarded projects through this FOA. And you can see from the boxes there are the teams.

[Next Slide]

As I go through the names, we acknowledge that some of the company names have changed as well. So, we had Chevron Texaco, which later became Chevron, with Hyundai Kia using UTC fuel cells. We had Ford and BP using Ballard stacks. DaimlerChrysler and BP, which later – DaimlerChrysler demerged to become Daimler and Chrysler separately, so Daimler continued on. And they used Ballard stacks as well. And then we also had GM and Shell using GM's own proprietary stack technology.

In terms of funding, this was a 50/50 cost-shared project with DOE putting in $170 million dollars. Industry cost here was actually greater than that—$189 million dollars, for a total of about $360 million dollars, and NREL received about $6.6 million dollars since the project started in 2003 going through the end of this year.

[Next Slide]

On the top there, you can see a timeline showing the project as it was formulated the first couple of years. After the first five years of the project, at the end of 2009, Ford, BP, and Chevron Hyundai Kia concluded their projects as planned (at the end of the five-year period).

You can see the Gen 1 and Gen 2 vehicles from Ford in the photos there. Hyundai Kia used the same vehicle platform for both their Gen 1 and Gen 2 stack technology. Daimler and GM opted to continue on for another two years. We also had Air Products through the California Hydrogen Infrastructure Project, or CHIP, also add data to this on the infrastructure. You can see the Gen 1 and Gen 2 vehicles from Daimler and the Gen 1 and Gen 2 vehicles from GM as well. Then there are a couple of examples of fueling stations that Air Products has put in place. So those teams continued to provide data for the next two years, through September 30, 2011.

[Next Slide]

What's NREL's role? I've given you an overview of the project itself, which was primarily an industry project. NREL's role was to support the DOE and public, as well as fuel cell developers through our involvement in the project. You can see, in the upper left, we basically received data from the companies directly. We received that either monthly or quarterly, and it included detailed on-road performance data, as well as summary information about the stations and fueling events.

That data then came to NREL, into our Hydrogen Secure Data Center, which is a room here in Golden, Colorado. We perform our internal analysis every quarter and then generate results. Those results really took two forms. One was detailed data products, which would go directly back to each of the companies that provided us the raw data. That was considered protected data, and it would basically be our analysis of their individual stacks and vehicles and stations. No other companies would get that data. Then we would have composite data products that were actually the public results and a subset of what I'll be showing you here today. Those were aggregated data across multiple system sites and teams. We had a rhythm where we would publish every spring and fall. This last quarter, because the project was wrapping up, we did it after three months.

[Next Slide]

This graph just shows the quantity of data that was received here at NREL, dating back to September of 2004. You can see we accumulated 3.5 million miles and over half a million vehicle trips. Each of those trips had second-by-second vehicle data, and at each second, we typically would have between 20 and 40 different columns of information about the batteries, fuel cells, vehicle speed, and so on. So that amounts to about 420 million seconds of data. And so it was quite a challenge to process that. And we actually created in-house tools – I'll talk about it a little bit later – to accomplish that. Also on this graph, you can see some slope changes there, particularly at the end of 2009. You'll see that the slope is not quite as steep. And that's when the two teams exited the project. And then, also, you'll see an uptick in the slope toward the right, in September of 2011, as we had more vehicles added into the project.

[Next Slide]

So here we have a graph that shows just thumbnails of the 99 composite data products. I'll refer to those as CDPs. You can see the varied structure and makeup of the different types of results we created – not only line graphs and pie charts, but various things that look like fuel tank gauges and clocks and days of the week and so on. Because of the large number of results, I'm just going to cover some highlights here today, dating back to the whole seven-year duration of the project.

[Next Slide]

I'm starting off here with the vehicle deployment status. This shows the number of vehicles deployed cumulative by quarter starting back in 2005 Q2. The solid colors indicate the vehicles are still on the road, and the hashed symbol indicates those vehicles have been retired. Then it's broken down into three categories by the hydrogen storage system – either 700 bar in the dark green, 350 bar in the blue, or liquid hydrogen in the red. A couple of things you can see here. The liquid hydrogen were retired back in 2007. Then in 2009, as I mentioned, when we had two teams exit the project, we basically had a large number of vehicles retired. Then we had a steady increase in the number of vehicles from the remaining two teams.

Also, there may be some question about why there are so many vehicles? Why were there 183 vehicles to validate performance? That was necessary, especially when there are four different manufacturers involved and two generations. We needed a large number of vehicles to get enough statistical confidence and significance in the results and also to allow for unexpected things to happen.

[Next Slide]

I want to highlight a couple of differences between the Gen 1 and Gen 2 fleets. Gen 1 were not freeze-capable. They were from about 2003 stack technology. They encompassed a number of different storage technologies for the hydrogen and had medium range, between 100 and 200 miles. The efficiency of the fuel cell system net was between 51 and 58 percent at about quarter power.

For the Gen 2 vehicles, the stacks were all freeze-capable, which is a significant achievement. The stack technology is a few years later, and all the storage systems were 700 bar. In terms of range, the range was a little bit higher at 200 to 250 miles. That's enabled by the 700 bar storage. The efficiency was 53 to 59 percent at quarter power, which is just a little bit higher than the Gen 1. However, that's still an accomplishment, given all the other improvements that were done, especially the freeze-capability. Finally, the fuel cell durability was longer, as I'll talk about later.

[Next Slide]

On the infrastructure side, which is the overall status, this graph shows that there were three different types of delivered hydrogen technologies and two of on-site production. And the pink diamonds basically show the numbers of each of those types at the peak of the project in 2009. After that, some of the stations were decommissioned or transferred to others, and so we wanted to reflect that there was really significant investment and research going on in terms of understanding on-site production and for reforming and electrolysis in particular. The light green shows the stations that continue to operate within the project and the dark green shows stations that have continued to operate outside the project and no longer provide data to NREL.

I also want to note that many of the demonstrations were taken offline as planned at the conclusion of the demonstration. But some have stayed open and/or received upgrades, particularly in California and New York.

[Next Slide]

This slide shows the four regions that still have stations operating as part of this project. There was actually a fifth one down in Florida whose stations were shut down in 2009. You can see in the top graph the roll out of the stations, which was almost linear throughout the project. The red indicates the stations that have been retired, and the green at the top are the ones that are continuing outside the project.

We basically have seven stations still operating within the project, 12 retired, and six operating outside. Out of the 25 project stations, about half are still operational, and half of those are outside of the DOE project. Northern and Southern California, D.C. to New York corridor, and the Detroit area are still active.

Most of the activity is in the Los Angeles area. You can also see by the symbols that the triangles are the ones online and the circles are future. Those are primarily the future ones that are funded by the State of California through CEC and CARB and they will be coming online in the next few years.

[Next Slide]

I want to really highlight the three main objectives of this project that was established back in 2004, and how we measured up against those metrics. The goal for fuel cell stack durability was 2,000 hours. For our Gen 1, we had 1,807 hours from the best team and 2,521 in the Gen 2 which equals 2,521 and exceeds the 2,000, so that goal has been met. You can also see other statistics there on the number of hours accumulated and the average projections. In terms of driving range, the goal was 250 miles. From Gen 1, we had between 103 and 190 miles. Gen 2 was 196 to 254, so you can see that goal has also been met.

While there wasn't a target for fuel economy, I have listed those numbers here as well, along with the fuel cell efficiency at quarter power and full power. And, finally, we have the hydrogen cost to the station, and that target was $3.00/gge. That was a pretty challenging target to meet through a demonstration project.

From this project, the cost of the hydrogen was estimated to be along the lines of between $7.00 and $13.00 per kilogram or per gasoline gallon equivalent. Outside this project, a DOE independent panel concluded that at 500 stations per year, distributed natural gas reformation would be about $3.00 per kilogram, and distributed electrolysis would be about $5.00 per kilogram.

[Next Slide]

I'm just going to go through each of those with one slide each. This shows the fuel cell durability for Gen 1 and Gen 2 which is a bit historical and goes back to March of 2010. We weren't able to replicate this exactly (once we had two teams exit the project) because we had to protect the confidential nature of whose data is whose.

I've already mentioned the numbers in Gen 1 and Gen 2. We had about 1,800 hours from Gen 1. This is a projection to 10 percent voltage degradation. In Gen 2 we had about 2,500 hours. You can also see because of the amount of time some of these vehicles were on the road (Gen 1) we actually had some vehicles accumulate almost 2,400 hours. Gen 2 was about half that, about 1,200 hours, simply because of the amount of time the vehicles were on the road.

[Next Slide]

In terms of vehicle range, you can see Gen 1 is in the dark green on the left pair, and Gen 2 is the right one of the pair – the light green. We have dyno range, window-sticker range, which is adjusted by 78 percent on the highway and 90 percent on the city. Then the on-road range, which is what we actually gathered from the data as the vehicles drove around under real world conditions. As you can see, here in the middle, the window-sticker range is the one that we used to compare to the DOE targets and the Gen 2 just exceeded that 250-mile goal.

[Next Slide]

Finally, on the third one this is the hydrogen production cost at the station. What we did here is we actually used inputs from the project partners, and we had them provide inputs on facility capital costs, capacity utilization, O&M, maintenance, repairs, and so on. We then put those into the H2 model and calculated what that range of hydrogen costs would be. You can see the median with the red and the 25th and 75th percentile is the box. The 10th and 90th percentile is with whiskers. You can see that in general natural gas reforming was cheaper than electrolysis.

Each of these technologies continues to evolve, but certainly this project provided a good benchmark of where the technology was today. Also just to note that these stations were not meant to emulate the high-volume replicate stations of the future. Permitting was also in transition, which caused things to go a little bit slower and cost more than they would in more of a mainstream operation.

[Next Slide]

Onto some of the technical results from the project in a more detailed level. I mentioned efficiency already. This graph shows in light green the Gen 1 efficiency range. On top of that, we have the Gen 2 efficiency range. You can see the DOE goals at quarter power and full power by the two blue symbols.

You can see we almost met the 60 percent target at quarter power. We had 58 percent from Gen 1 and 59 percent from Gen 2. In terms of 100 power, we exceeded the target with both Gen 1 and Gen 2. A critical result here is that the efficiency, as we went from Gen 1 to Gen 2, was not sacrificed in order to achieve improved durability and freeze capability.

You might expect sometimes you have to make compromises in efficiencies. Sometimes it is the thing that's compromised as you move toward the ultimate product.

[Next Slide]

On fuel economy, I mentioned already there wasn't a fuel economy target from this project, so there's nothing to compare it to as a benchmark. But this shows Gen 1 in dark green and Gen 2 in light green – again, the same categories as before: vehicle dynamometer, window-sticker, and on-road.

A significant thing to note here is that Gen 2 is just slightly higher on the dyno results and the window-sticker, but in terms of the on-road, it is significantly higher than the Gen 1. This indicates that the Gen 2 on-road fuel economy is much more robust to real world driving than the Gen 1 was, so it's not as much of a drop from the window-sticker to the on-road, as it is for Gen 1.

[Next Slide]

In terms of driving range, we saw a significant improvement throughout this project of the real world driving range in-between fueling. This is a histogram or a series of histograms that show the distance each of the vehicles drove between refueling. A vehicle would get filled up with fuel, drive around, and come back, and get more fuel. This is the histogram of those collective distances.

You can see the median distance between refuelings for Gen 1, back in the gray histogram, was 56 miles. Gen 2 was a 45 percent improvement, up to 81 miles. This is the darker gray. Then the final two years of the project, it actually came out to be 98 miles for the median, with a 71 percent improvement in the real world driving range, with the latest advanced technology vehicles.

Also, remember that the actual range possible from these vehicles is over 200 miles. This is just what we were actually observing – vehicle/driver behavior – as far as when they refueled.

[Next Slide]

Another activity we participated in, relative to range, was a special experiment that NREL did with Savannah River National Laboratory, where we drove from Torrance, California, up to L.A., and then down to San Diego, had a pause for lunch, and then came back from San Diego in a reverse manner, and actually drove 331 miles without refueling.

There was enough fuel left in the tank – about 1.5 kilograms – that we calculated the range of this vehicle for that day at 431 miles. This was from two separate vehicles driving after each other. There is a nice video that Toyota put together on their website, as well as a full-on 17-page report on our website if you want to read more about that.

[Next Slide]

On to fuel cell stack durability. Once we had two of the teams complete their projects, we had two teams remaining – two OEM teams. So we decided to do things a little bit differently to protect the confidential nature of the fact that there were only two teams left. We have a series of different results that I'm going to cover here.

This shows the percentage of stacks as a function of operating hours. As you can see in blue are the in-service stacks. Red is retired, and orange is not in service. You can see here there's basically a median of 620 hours, so while we had some stacks operate up over 1,400 hours, half of them were still below 600 hours. This is in terms of hours accumulated and a quarter of the stacks were above 937 hours.

[Next Slide]

In terms of durability and looking at how the power would degrade of these systems over time, we actually created a bar graph that shows the collective focus of where the power is from these systems, since it degrades with time. So you can see with a nominal kind of nameplate power normalized here to 100 percent, in the first 200 hours, we saw maybe about a 10 percent drop.

And then in the next 1,000 hours, we saw about another 4 or 5 percent drop. And so in conclusion, on this slide, we have – between Hour 0 and Hour 1300, we saw an 18.2 percent drop. But most of that occurred during the first 200 hours. And I'll indicate, on the next slide, why I'm focusing on that – that 200-hour point.

[Next Slide]

So if we do our normal voltage degradation to 10 percent, as I had talked about with our previous results, which showed Gen 2 at 2,500 hours, with the data from the last two years, the projection is 1,748 hours. You might wonder immediately why that's lower than the previous result. The primary reason is that we have only accumulated the hours as I showed on the previous slide, and so we have actually a large number of stacks that don't have much degradation. We have actually put an artificial limit on what that projection is, so we don't have projections out to 50,000 hours, when we know that's not a real answer. Our limit here was arbitrarily set at 2x, so if a stack had accumulated 1,000 hours, we would limit the projection of that team to 2,000 hours. That's why we have a clustering up here at around 3,000 hours. This is using all the data from Time 0. If we allowed the Fit to be started at 200 hours, after that initial degradation, we would have a fleet average projection of about 2,261 hours. This is just to indicate the impact of that first degradation on our projections.

[Next Slide]

Now I'm showing the same data, but just in a little bit different way. This is the scatter plot of each individual stack and how many hours those stacks have accumulated, and for each of those stacks, what our projected hours are, based on that 10 percent voltage projection. Obviously, the line here going up, one-to-one, is when the projected hours equals the operation hours.

I want to point out a couple of things here. You can see, on the top, these triangles or stacks would have been limited by that 2x factor, to minimize the extrapolation. Also, you can see there have been a number of stacks that have operated beyond the 10 percent voltage degradation.

[Next Slide]

Switching over to infrastructure, we looked at hydrogen production infrastructure efficiency, and we did it again with natural gas, reforming on-site, and on-site electrolysis. Each of these little x's is an individual quarterly efficiency data point from one station. You can see a smattering of those points.

We also have a probability distribution function – the dashed curve over to the right. And then, finally, the average station efficiency is the big blue diamond, and the red star is the highest quarterly efficiency. You can see the targets up here at around 70 or higher percent efficiency.

The highest quarterly efficiencies, these red stars, did approach the DOE target. However, the average station efficiency was about 20 percent points lower than that. Again, realize this was a demonstration project and a lot of these technologies were being implemented in this way for the first time.

[Next Slide]

The next slide here shows the greenhouse gas emissions of vehicles from this project, using the hydrogen from these on-site production stations. On the left, we have on-site natural gas reforming again, and on the right we have on-site electrolysis. The band at the top is the baseline – mid-size SUV to passenger car range for greenhouse gas emissions and CO2 equivalent grams per mile.

And so what you can see is that these vehicles from this project, using the hydrogen produced in this project, have significantly lower greenhouse gas emissions than the conventional gasoline vehicles. And this is significant, because these are just demonstration systems, and we're still seeing almost a 2x drop in greenhouse gas emissions. And you can see there are a number of footnotes here, which you can read later, on all the details of what went into those calculations.

[Next Slide]

Also, in terms of infrastructure we looked at the station capacity utilization. This is for the last two years of the project, so we had seven stations included. We basically took the nominal capacity of each of the stations, as reported by the operators and owners of the stations, normalized that to 100 percent, and then we looked at the maximum daily and quarterly utilization in the blue and red symbols, and then also the bar – the average daily utilization.

You can see many of these stations were utilized, according to the nameplate calculation at least, at about 20 to 30 percent, with the highest station being used about 60 percent of the nameplate capacity. Again, a lot of these stations were put in place for coverage and availability of fuel for the vehicles, and not designed for high-station utilization by vehicles. There simply weren't enough vehicles, at the beginning to make that possible.

[Next Slide]

Looking at the same data slightly differently, which is – the bars were by day of week, so Sunday through Saturday. You can see most of the refuelings occur Monday through Friday. Each of these blue curves is the amount of hydrogen dispensed on average by day for each of these stations.

For example, this top curve has a peak of 27 kilograms per day on average, which is a peak on Thursday. This is the average fuelings at that station every day of the week. You can see several of the stations dispensing a reasonable amount of hydrogen and some not providing a whole lot of hydrogen. Several stations are serving about five vehicles a day or more on average and a few of them are fueling a couple of vehicles a week. They're still providing significant value through the coverage they afford.

[Next Slide]

In terms of infrastructure reliability, we looked at reliability growth curves. The Y-axis here is the shape parameter. It's a little bit technical, so I'm not going to get into that.

You can look at the right axis that shows that the higher the number means the failure rate is increasing, and the lower the number indicates the failure rate is decreasing, with one being this green line that is neutral. It's not increasing or decreasing.

In the first 120 days, that's the star symbol in red, you can see that several of the stations had very high growth and failures at the beginning of the project. In the last 20 percent, which is the yellow bar, you can see what changed. So the instantaneous mean time between failure has actually improved for five out of the seven sites in the last 20 percent of events.

This is going from the blue bar, which is the overall average of the whole time, to the last 20 percent of events. You can see all these white arrows indicating that the failure rate was decreasing, except for two stations where it actually did increase.

[Next Slide]

Moving on to fueling rates, we gathered data from both the station and the vehicle to figure out how fast the fuel gets put into the vehicle once it's connected and it starts fueling. You can see this through time, starting back in 2005 to 2011. The dash curves are the first five years.

The solid curves are the last two years – 2010 and 2011. You can see the average here, if you follow the cursor, increasing and kind of settling in by 2009 at about 0.8 kilograms per minute. The last two years we had some of the higher throughput stations close down, so we actually saw a slight decrease in the fueling rate, back down to about 0.63 to 0.67 kilograms per minute.

[Next Slide]

This shows us more clearly – the first five years are in gray, and the last two years are in yellow, where the average in the first five years was 0.77 kilograms per minute, and it dropped down to 0.65 kilograms per minute. You can see very easily this gray area over here, where the higher capacity stations are no longer in service providing data to this project at least.

Actually, back on this last slide – that was a result of the average hydrogen per-fill increasing, so more fuel being put into the vehicles, with the average fueling time actually increasing by a larger amount – by 37 percent.

[Next Slide]

If we look at that two different ways, we can look at fueling rate as a function of the pressure level – 350 bar versus 700 bar (that's in the upper left); and communication versus non-communication. Communication, in this sense, is when the vehicle hooks up to the station to get a fill. Is the vehicle talking back to the station and giving it a status, particularly about the temperature of its carbon fiber composite tanks?

In the upper left, you can see that the 700 bar fueling rates were holding constant at about 0.63 kilograms per minute, while the 350 bar fueling rates dropped from 0.82 to 0.7 kilograms per minute, again, because of some of those higher capacity 350 bar stations being decommissioned.

Then on the lower right, we have communication fill rates dropping, while the non-communication fill rates are increasing. So this is the first five years of the dash curves, and the last two years are the solid curves. Yes, that's right. The first five years are the dashed, and the solid are the last two years. Blue is communication. Red is non-communication.

[Next Slide]

We have taken what we have learned from this project, especially the tools we have, and we have actually applied it to many different applications. So, I've been talking all about fuel cell cars and stations today, but we also did get data from buses and forklifts and laboratory data. So, we've actually used our fleet analysis tool to analyze all that data.

As an example of that, I show here refueling rates for three different technologies. We have the cars that I just talked about. That's this green curve. Forklifts are on the left in orange, and buses are on the right in the blue. You can see, basically that the fueling rates are different, and they're driven by constraints on nominal pressure, volunteer, and tank material. So, in the case of buses and forklifts, the pressure is only at 350 bar. In forklifts they're often using steel tanks. So, while they could fill faster, since they're only putting between 0.5 kilogram and 2 kilograms in there, they don't really need to worry about doing it so fast.

[Next Slide]

Finally, we have one more technical result here, which is really pretty detailed technical stuff, so I just wanted to show it as an example of an analysis result that's been informing the R&D activities and the codes and standards development.

Here is a little bit of background. When a car pulls up to a station to fill up, the temperature of the tank may be different from the ambient temperature. This is because as the fuel comes out, it thermodynamically interacts and can have a different temperature than the surrounding temperature. So, what we showed here was that the fuel cell vehicles arrived at the station with a tank temperature that's 3.8 degrees centigrade colder than the ambient temperature. This was actually used by the SAE committee, J2601 as they formulated that standard. They were using a computer model to do various simulations about how much the temperature would rise during a refueling event, and they wanted to know what the initial condition was to set that computer simulation. This provided a distribution for them to use and feed into that model. Here is an example of how – while most of the results are geared kind of a higher level summary, we have done and have the capability of doing much more detailed analysis, based on individual requests.

[Next Slide]

I'm going to wrap up here with a technical summary. The project has now completed about seven years of real-world operation and validation, and a lot of data has come out of that. In terms of vehicle operation, we had 183 vehicles, over 150,000 hours, 3.5 million miles, and a half million trips.

On the station side, 25 stations were deployed, over 150,000 kilograms produced or dispensed, and over 33,000 refueling events. In terms of DOE's key technical targets, they've been validated and met. The fuel cell durability is greater than 2,000 hours, and the range is greater than 250 miles.

[Next Slide]

Here's the wrap-up – the winter 2011 CDPs were just posted on NREL's website and those are available now for you to take a look at. We will have a draft final report written at the end of next month, and it will be published the following month in April. I'll discuss any other results at the Annual Merit Review in Washington, D.C. in May. We are continuing to receive data on hydrogen infrastructure with support from DOE.

These are stations, primarily in California, that are funded by CEC and ARB, so there will continue to be new results to follow and see on the station side. We are in discussions with how to continue to assess fuel cell vehicle progress in the coming years, so that there continues to be a pipeline of information about the status of the technology.

Finally, this project is really the first time such a comprehensive data set was collected by an independent third party and consolidated for public dissemination. We also have seen this successful framework being used for other projects.

[Next Slide]

This slide just shows graphically those six-month period outputs of composite data products that came out of this project, like a drumbeat, for every six months for the last seven years. Those have led to objective, credible evaluations, and helped lead to informed decisions. As the demo concludes here, I will be publishing the final public report.

[Next Slide]

To learn more on your own, all of these results and all our publications, including the presentation I'm giving here today are online for you to take a look at later and share with your colleagues.

[Next Slide]

I just want to pull up the website here, just to walk you through that a bit. So, this is our website, and you can see basically all of our publications and presentations, including the final set of results here. This presentation includes the 1999 CDPs I mentioned, but we have presentations going back all the way to the beginning of the project, back in 2004.

Also, you can individually select results by topic. So, for example, if you want to look at fuel cell stack durability, you can see all the results that relate just to durability. Or if you're interested in the infrastructure or fueling rates, you can go directly to those results, and then pull up individual results that way.

[Next Slide]

Here is something new here that we just added, and it will be available online from our main site soon. But right now, it's only available on our development server and the website I've listed there at the bottom. This is basically access to the same information, those 1999 CDPs, but organized in a bit more of a consolidated way.

Rather than scrolling down through a number of static pages, you can actually click on this wheel and see the results in a more organized way. So, for example, if you want to look at hydrogen storage, you can click on that. It spins the wheel around. You can then, with hover-over, see thumbnails of those individual results. Let's say there's one that you want to look at. You can click on it, and it will then pull up that individual result. So similar – it's the exact same result, just a different way to get to it, and you can see all the results on one individual screen. OK?

[Next Slide]

And that is the end of my presentation here. We've got some time for additional questions, and I'm going to turn it back over to the moderator here to figure out how to deal with those questions that have come in online.

[Crosstalk]

Question:
Are you working on fuel cell CNG breaking regenerator combined systems?

Keith Wipke:
I'm not familiar with what that is exactly, so I would have to say no.

Question:
Another question is: can we get copies of these slides?

Keith Wipke:
Yes. I believe within the next 10 days. Not only will be the slides be available, but also an audio recording of my talk here today will be available on DOE's website. We'll certainly make the link available from the website that we have here as well.

Question:
What is bar versus liquid storage?

Keith Wipke:
So sorry for not being clear on that. Liquid storage is liquid hydrogen. So it's stored at cryo temperatures – very cold temperatures – to keep it as a liquid. So, basically, you have a liquid container surrounded by a vacuum jacket, to keep it from warming up too quickly. And bar just refers to barometric pressure. So that's one barometric pressure at sea level is one bar or one atmosphere. And so these tanks are at high pressure. So the medium pressure there is a 350 bar. A higher pressure is a 700 bar. So bar just refers to the units of the pressure.

Question:
Next question is: can you explain the difference between the max team projected hours and the average fuel cell durability projection listed on Slide 13.

Keith Wipke:
So, let me go back to that slide here. So, the max team projected hours – if you think about – actually on the previous slide here, with the graph. Let's see. So this shows it graphically. So the maximum here is – of the four teams, the maximum is the best of the four, and the average is the average of the four. And the durability projection, as I briefly mentioned here – and it's listed at the bottom as well – is we looked at the voltage degradation of each stack over time, and then weighted those according to the Fit, and came up with an overall Fit, and determined when that voltage degradation dropped by 10 percent. And that's what we used. That crossing point was the amount of time it took to get to that 10 percent voltage drop.

Question:
Another question: on average, how many hours per year does a car operate?

Keith Wipke:
So, 5,000 hours on a stack is meant to be about 150,000 miles. So you've got about 30 miles-per-hour average vehicle speed. So 150,000 miles – what's the question? On average, how many hours per year? So 500 hours per year would be somewhere in that neighborhood. A number of questions here about whether the presentation's going to be available. We already talked about that

Question:
Which of these vehicles were fuel cell only or used some battery to operate as hybrid with no plug-in?

Keith Wipke:
We did have, in the first generation, some vehicles that did not have a battery. They were non-hybrids. All of the Gen 2 systems were hybrids with batteries. We also did have a few vehicles that had a plug, and those would be considered a plug-in fuel cell hybrid vehicle. OK. Let's see.

Question:
Why only compare fuel cell vehicle well-to-wheel greenhouse gas emissions to mid-sized vehicles? FCVs need to be compared with other advanced vehicles – HEVs, PHEVs, and VEVs.

Keith Wipke:
And, yes, for simplicity, we just compared to the baseline here. However, if you look at the GREET model results from Argonne National Lab, they do a very comprehensive analysis, where they compare all the different technologies with many, many different bars. We were just trying to do one here, for simplicity.

Question:
Another question: how would a solid oxide fuel cell work in vehicles, particularly in trucks, for service, durability, and reliability?

Keith Wipke:
These are being looked at for long-haul trucks, to prevent idling of the stack or idling of the diesel engine at truck stops. Often they will spend the night and idle their engines, so APUs can basically remove those emissions. But for cars, which really only operate, on average, between a half an hour and an hour per day – so most of the time, they're sitting there, off – you don't really want to have a system that's got to be on most of the time, and takes a long time to start up and cool down. OK, so we've got a number of other questions here. My team is helping filter through them here.

Question:
The question is how do these results compare to similar projects in Europe and China?

Keith Wipke:
So, there are activities going on in Europe that I'm familiar with. This is the largest demonstration of its type, at least when it was in full force. I'm not sure about demonstrations in China. Although in Europe, there is a H2 mobility activity that involves a number of OEMs and will be supporting a large number of fueling stations, as well as a large number of vehicles. So most of the other countries that have a large carbon-manufacturing industry are pursuing these technologies and getting support from their governments to support the infrastructure as well, which is really, at this point, not necessarily a money-maker, but is really critical to enable the vehicles to get deployed.

Question:
Any recommendations for further improvements at the stack level, to ensure a higher reliability and better overall performance?

Keith Wipke:
That's a good one. Jen, do you have any comments on that one?

Jennifer Kurtz:
I think what we're seeing with the durability results – they're actually maybe a technology generation behind the latest and greatest stack technology. So, there is a lot of work going on in the lab, which we should see getting deployed in the next few years. We didn't focus very much into specific technology improvements. But certainly durability of the stack is critical to market acceptance.

But we will see an improvement, and we've studied that in our lab durability tests that Keith mentioned earlier in the presentation. Those results are available online as well – that show you state-of-the-art durability results.

Question:
A question on tank temperatures. The question is: what did you see the max vehicle tank temp was offset, compared to the ambient conditions at the station? And why was the tank temp offset from the ambient conditions of the station?

Keith Wipke:
So, as the vehicles drive around, you basically have your high pressure gas expanding. So you have the tank cooling down. However, there are also other effects going on, such as vehicles that are parked for a while. And so maybe you're parked in a shopping center's parking lot, parked on top of asphalt that might be 130 degrees Fahrenheit. So, you could be essentially baking the bottom of your vehicle. The tanks are normally mounted from below, and so the tank temperature could be significantly hotter than the ambient temperature. Or if you had just been driving pretty hard in the vehicle and using a lot of fuel, the tank temperature could be quite a bit cooler. So, that's really why the standards committee was looking at how much does that temperature of the tank vary? And, under what conditions?

Question:
Have any of the vehicles been in accidents? And were there any problems as a result?

Keith Wipke:
That's a good question. So, you don't like to see accidents. But in this case, we do like to see accidents, where everything works as it did. There were several traffic accidents with fuel cell vehicles, and all the safety systems performed as designed, with the tanks being cut off from the rest of the system and deactivated. And, there were no hydrogen releases associated with those accidents.

So, there were only minor injuries not having to do with the fuel cell system, just from the actual impact itself. So, I think the systems have demonstrated that they are safe and can definitely be managed with good engineering upfront.

Moderator:
At that, we're going to let them ask one more question. And then we're going to wrap up the call.

Question:
What can we tell the public about the feasibility of buying fuel cell technology in the future?

Keith Wipke:
Okay. You can, if you're in a certain part of Southern California, get on the list to lease a fuel cell FCX Clarity from Honda or a B-Class from Daimler. Both those fuel cell vehicles running on hydrogen, using the fueling stations in Southern California. Other OEMs have announced vehicles coming out in the next few years. So, Hyundai Kia will have a relatively large set of vehicles – I think about 2,000 vehicles going out between 2012 and 2014. And then in 2014 to 2016, you'll have probably six or seven OEMs, all having a fuel cell hydrogen vehicle on the market, either to buy or lease in select areas that have a sufficient hydrogen infrastructure to support those vehicles. So, certainly follow up with the companies and express your interest in the technology, and you can get on their list.

Moderator:
Okay. Great. Well, Keith and team, thank you very much. Just to reiterate. This webinar has been recorded and will be available on the Fuel Cell Technologies website – the recording and these slides, approximately 10 business days from today. So definitely check back and take a look, and also keep in tune to learn about our upcoming webinars. We have just about one every couple of months. So, thank you.