![[Go to Home page]](../images/smallseg.gif){kind=small} |
Australian Academy of Science Science at the Shine Dome Canberra, 3-5 May 2006 |
|
Full listing of papers
|
SCIENCE AT THE SHINE DOME ANNUAL SYMPOSIUM Science on the way to the hydrogen economy 5 May 2006
Hydrogen car prospects
(Click on image for a larger version) I would also like to thank the Academy for inviting me to talk about one of my favourite topics. Thank you to the audience as well for being ‘the stayers’. Indeed, I would like to introduce myself as ‘a stayer’. I was privileged enough as a child to build my first internal combustion engine, and I have managed to stay with these systems throughout my career. At Imperial College, my doctoral thesis was about the homogeneous charge compression ignition of hydrogen-air, and hydrogen-oxygen mixtures. So some of you may know that Homogeneous Charge Compression Ignition (HCCI) is the ‘in’ area of research, and the automotive industry is spending as much on HCCI combustion as they are on hydrogen vehicles. Thus I think it is a privilege to be ‘a stayer’ and to have some awareness of technology, because not many of students present today are going to find opportunities for being able to work in the one field for so long.
I am now going to tell you about some issues with which you are already familiar. This is because, being the last speaker, I find that just about all of my slides have been presented already. So I thought that I would just do a bit of stirring and put in a few extra slides.
First, I want to tell you a bit about the fuel supply side in summary form. I will take this approach because I will be pulling together lots of different components from what has already been presented before. I will particularly focus on my second-last bullet point here, stated as comparing the performance of fuel cells and hydrogen combustion systems. Although there is a lot of allusion to the fact that fuel cells are so efficient, now with 500 systems running at various locations around the world, it is very difficult to find on-the-road performance data for these systems. I just wonder why this is the case.
Now I would like to reflect on the fact that for the area that we are examining, we are really only talking about the right-hand end part of this diagram that outlines possible fuel scenarios. Here it is important to recognise that there will be a growing number of activities that emerge as the price of the left-hand market increases and as people begin to consider some of the intermediates as well. Actually, I think it is in some of the intermediate areas that we are going to see the next generation of energy options emerge - and I have believed this for a long time.
Accurate predictions for long-term fuel supplies have not always been easy to forecast. [At one point when various alternative vehicle systems were being trialled, the oil industry was forecasting that oil production would peak in 1986. Such trial vehicles were then built but not sold to anyone. So when we talk about some of these future projections, we have to bear it in mind that historically we have not been too good at knowing where we are in terms of resource supply.
I think that one of the big issues that has been hidden in our discussions today relating to fuel cell vehicle systems is whether the reformer is best located on or off the vehicle, Plainly, there are some advantages to having the reformer located off the vehicle (if you can find a way of storing the hydrogen).
This slide presents a review of the available fuels that can be put in the marketplace at the moment. For this table, we can see that natural gas is probably the best fuel for on-board operation. However, of course, if we can move to reforming and, as we have seen, manage to take advantage of the option of efficiency improvements, then we could even attain hydrogen production with an efficiency that encompasses about 70 per cent of the energy.
George Crabtree, this morning, said that we could reclaim all of the energy involved in the process. Of course, he didn’t include any of the chemical engineering in these calculations. So I think we do need to bear this in mind when we use steam reforming with a water shift reaction because it is not a perfect process. Ultimately, there is some energy loss involved in what we are able to get into the hydrogen. Of the available fuels, something that I think needs to be looked at more carefully is the ratio of hydrogen at 200 atmospheres that is illustrated here in this slide. We have seen 700 atm tanks enter into the public arena for use recently, but ammonia stacks operate pretty well, and petrol is still the fuel by design if we are looking for energy on a volume basis.
Now one of the things that we have to think about when these hydrogen tanks blow up is that they have probably very much less energy in them than a comparable petrol tank when it blows up. So the ultimate accident (and you can expect it to happen somewhere at some time) is that between a gasoline tanker which has an accident with a liquid oxygen tanker, and then we would see two city blocks wiped out. So we really have been lucky that we don’t manage to get that sort of scenario.
One of the things that I would like to point out here too, is the energy involved in getting the fuel onboard the vehicle. Although we have been talking about the route represented by the central two columns here as being the one that is most tenable, we already lose more than 20 per cent of this energy in the compression process. However methane, because of its high energy density per unit volume, looks to be more attractive too. So I think that for transition, we should be looking at this consideration and for the longer term we are obviously going to be looking at the opportunities shown in red in the centre. The metal hydrides have a particular advantage because their refuelling energy losses are a lot less.
I think the images on this slide have been up several times.
There has been some talk regarding of the need to set appropriate targets for the range of power outputs for fuel cells.
The NEBUS or the Daimler Chrysler buses, of which three are currently running in Perth, look to be pretty efficient on this chart in terms of power versus lifetime in years that is expended by the technology.
In this slide, I wanted to illustrate a gradual trend for improvement in terms of power from fuel cells. This is a necessary consideration, because inevitably we have got past the few dedicated users, people are going to expect performance for their vehicle that is similar to that which they are accustomed to with petrol or depending on which part of the world you come from – diesel , if you are from Europe.
Now we can see in this slide, some of the energy densities and even the fuel stacks at the bottom of this graph are relatively poor compared with the DoE targets for 2005–2010, when we look at a weight to power ratio.
Here, represented at the bottom of this graph are ‘stacks-only’, and these here are the targets. Now, when we overlay some internal combustion engine data then we see that there is at least a factor of 2 in difference between the targets and what is achievable. These are for the NECAR 3, 1997, and some of the later models are still located in the right-hand region. So there is still quite some way to go yet in terms of improving the power to weight ratio (that is, dealing with a heavy fuel tank, a heavy power train that has to be put in the vehicle). So lots of aids that are designed to produce lighter and more aerodynamic vehicles are being added to address this issue. Thus these types of vehicles are then not really suited to ‘apples with apples’ comparisons with ordinary vehicles that we currently use to drive around.
We have seen some of the bus energy storage issues here, and this provides another view of the same technology.
Here are some cars that you have seen earlier, so I will skip over them – except to say that perhaps between the top left picture and the top right picture you can see the improvement in aerodynamics. I thought you would also like to see an image of the refuelling process for gaseous cars since this looks something like an LPG hose. It makes use of a more positive integrator (if you have never seen an LPG vehicle being refuelled) because it has a screw-on connector. This can be a bit of a nuisance because the threads sometimes don’t match, but this is a lock-on connector. Obviously, with 700 atmospheres of pressure coming through here, you can’t afford to have the nozzle flying off in your hand.
Another point that I will highlight is that hybrid drives are a necessary part of good technology. Actually, this is a particularly desirable feature of the fuel cell due to the use of a secondary battery to provide the transients. This means that the fuel cell itself can operate closer to those high-efficiency regions that we saw in some of the graphs earlier. The batteries also provide, as they do in the Toyota Prius, an extra source of energy for doing the transient driving that is essential (even, I discover in Canberra morning traffic). (Indeed I thought you would have the ideal trip from the airport to here, with the politicians running late, but no, you have your peak hour issues too.)
Another consideration that is not often mentioned is the amount of energy it takes to start and stop these systems. In the internal combustion engine, you can just switch it off and switch it on, and it is both ready to stop and ready to go. In a lot of fuel cell technology you need actually to get the temperatures up for the various processes to work efficiently. Thus the difficulties of getting fuel cells to start in sub-zero conditions can be quite significant. For example, in the fuel cell bus that is illustrated here – you can see this is a methanol fuel cell – and about 10 litres of methanol is needed for each start and stop in this system. Indeed this is the only data I could find in the literature about this rather delicate issue, and this was an in-service trial. It will therefore be good to see the experimental results from Honda, General Motors and Toyota that currently undertaking studies in California at the moment.)
So what about the hydrogen Internal Combustion Engines (ICE)? Well, one of the difficulties with using hydrogen as a fuel is that if you put it into an LPG conversion type system then you discover that you have a major backfiring problem. Then very soon there is a meltdown under the bonnet and this occurs because hydrogen has very wide flammability limits. One of the systems that we developed at the University of Melbourne in the middle 1970s was a patented system called delayed port admission (DPA). In this system, hydrogen gas goes into the middle of the intake flow to the engine, so that the first entering gas which goes in at the front is free from the combustible mixture. This means that the residuals left from the previous combustion cycle can be cooled. Then the hydrogen goes into the system which is followed by mixing of the gases that takes place in the cylinder. Finally, the last gas that is entered into the engine is air which means that there will be no hydrogen left in the manifold. A simple configuration to deliver this process is schematically shown on this slide.
It is interesting that BMW today use a similar system in the latest of their engines which they have manufactured for cars. They also have some different technology for their Siebel 8 design, which is not yet in the public domain.
This was our first hydrogen car, a converted Ford Cortina. Under the bonnet, some variations have been specially tuned in the system to try and improve the ‘breathing’ of the engine. This is necessary because hydrogen displaces about one-third of the air that normally enters the engine. Indeed this ratio occurs because to get a stoichiometric mixture, a chemically correct mixture, you need about 30 per cent of the charge to be fuel.
At the left you see some of the controls – in this case, gas regulation and so on, and the use of our delayed port admission system (DPA).
The graphs illustrated here show load and torque. There is engine twisting power but roughly you can talk about load here that occurs under full load conditions. Thus in this case, there is some loss in power because of the displacement of air, and also the oxygen displacement that I mentioned before. Then under the light load conditions here, which are critical, there is a doubling of efficiency. So this is an important factor. Even so, hot surfaces at high power outputs are a significant issue, and some internal cooling of the combustion surfaces is necessary in these systems as well as the use of special lube oils. The synthetic oils, when they came along, were a gift for hydrogen engine technology.
Illustrated in this slide is a BMW’s V12 engine. This car has been available in Australia on demonstration since 2002. When you look under the bonnet it looks very much like a 7 Series BMW. Of course, not too many of us are in a position to own one of those.
So illustrated in this slide is a comparison between some of the technologies, and one of the major issues is the loss in power associated with the hydrogen displacement. BMW, in the 745 model that I showed, were fairly cautious, but then their base engine has quite a high power output. A 45 per cent loss in power is shown here. By comparison, Mazda’s rotary engine is very suitable for separate admission of the hydrogen to air. This is possible because one of the ports that normally would be admitting air is devoted to hydrogen. In this way, air can be injected into the Mazda rotary chamber before the hydrogen is admitted thereby avoiding the problem of residual gas ignition backfiring. Amongst these comparisons then, our Ford Cortina is looking quite respectable. Now turning to comparing the performance of hydrogen cars, we have some emission data here for the Ford Cortina. We recall that this is 1986 technology and so there are emissions of oxides of nitrogen which are not tolerable today. Nevertheless, this is significant on the base and another important issue is that there is a 72 per cent increase in energy economy.
So when we look at hydrogen spark ignition engines, as compared with the petrol counterparts, then we have to bear in mind the reduction in power and the benefits of reduced emissions which obviously include fewer hydrocarbons and less carbon monoxide. These principles remind me that one of my colleagues said for Open Day that: “the exhaust was so clean owing to such low emissions, that Professor Watson would drink the condensation coming out of the exhaust!” Well, I was overseas at the time that he put this in the university program. So when I returned, they gave me this glass of water and I was very concerned about drinking this stuff – especially since it contained a very special lubricating oil containing lots of additives. Thus I was wondering how much of this material would be present in this water that I was drinking. It looked pretty clean, but it was still an issue.
An area that the last speaker covered was transitions and so this slide shows transitional technology that is based on using much smaller amounts of hydrogen than would ordinarily be requireded for a pure hydrogen internal combustion engine. Now in place of the spark plug we instead have an assisted ignition system which consists of a gas injector to inject hydrogen. There is also a smaller spark plug fitting that can be fitted in the normal spark plug hole, and beside this is a pre-chamber, which typically has only about 2 per cent of the energy delivered as hydrogen.
As you can see from the image in this slide, there are turbulent jets from the little chamber that ignite the main chamber fuel. In such systems, the fuel is typically petrol although it can involve a number of other fuels. Also, in these systems, the active species of the rich mixture combustion from the pre-chamber enhances the burning speeds by a factor of 3 to 6.
In this slide, a little image sequence is presented that shows an optical access engine that is burning a very lean mixture. The λ represents 100 per cent excess air. At this point we note that a normal spark ignition engine is not able to run with 100 per cent excess air. However, under the conditions shown here, we are able to burn mixtures that in fact exist at λ 3 or more. Indeed, we can achieve a significant improvement in efficiency without producing nitrogen oxide emissions because the combustion temperatures are so low with these lean mixtures.
Perhaps somewhat surprisingly, we had a thought about five years ago that we could apply this hydride-assisted ignition process to hydrogen main fuel combustion. It can be seen from these results that the peak efficiencies of burning mixtures situated at the top of the upmost curve have a high compression ratio and mixtures as lean as λ 3.5. This typically means then that about 250 per cent excess air, with no nitrogen oxide emissions, provides very good efficiency that is in excess of 40 per cent. So, when we look through the literature for some actual measured fuel cell data, we can find some examples produced by an independent group rather than a fuel cell producer. Fro example, here we have some results from FEV, a German company who are adding improved compressors to the fuel cell to improve the efficiency of the system. In this particular design, we can see that the move from using a blade (or roots) to a screw type compressor system resulted in an increase in peak efficiency. Nevertheless, the driving of these auxiliaries is also very important. We notice in the NEBUS, which has a 250kW fuel stack that 60kW is required to drive the auxiliaries. This is why the stack efficiency may well be up high but in reality the combined system functions at a lower level of efficiency.
Now when we overlay the results from the previous slide, we can see that the differences in efficiency, that occur between the conventional or modified conventional combustion system and the fuel cell, are not as great as the protagonists would want to say.
I don’t know whether you can see Shell as an independent arbiter in relation to well-to-wheel energy analysis but together with General Motors and Daimler Chrysler this company worked (in 2002) on producing some comparative information on well-to-wheel lifecycle analysis and efficiency. Some results are illustrated in this slide and the base A-Class Mercedes can be seen down at the fourth line from the bottom labelled - gasoline. Now the best technology that they came up with was a diesel hybrid, where you can see about a 50 per cent reduction in energy consumption. The fuel cells also performed well and the CNG fuel cell is located in a position that is third from the top, followed by the gasoline fuel cell, and located three lines down from this, we have the compressed hydrogen fuel cell. So in this way it is possible to compare the differences in efficiency.
I would now like to mention that some of technologies that were examined were hybrids, and we were told this morning that a hybrid was a ‘no-brainer’. So I would just like to think how a more integrated vehicle system could perhaps better improve the drivability of the vehicle (involving the ability to look ahead) – where the stops and starts for hybrid vehicles are very important.
This slide shows a comparison that was made between a standard four-speed automatic Falcon and a fully-optimised hybrid vehicle. The ‘look ahead’ change is represented by the red line estimated by driving using the usual Australian Design Rule test process, so avoiding the stops.
The sorts of fuel savings that can be achieved by integrating the vehicle into the traffic are very considerable, and the hybrid’s advantages start to disappear.
Now if we are look at the well-to-wheel greenhouse emissions, then we see that CNG hybrids and CNG fuel cell are located neck to neck and so in terms of greenhouse gases, there is no clear winner.
Next, I will say that if we look at Argonne’s long-term forecast that was made in the late 1990s where the relative prices of fuel cell technology are compared with hybrid technology, then the Toyota Prius turns out to have a purchase cost ratio of about 2.0.
In this slide we can see some of the points made this morning about higher maintenance costs.
More recently for light trucks (and I am not sure why this is a military light truck except to guess that some of the funding was sourced from the military) there was some analysis that suggests some improvement in this type of vehicle with regard to fuel cell technology.
In terms of the lifecycle analysis for a light truck, the WtW shows up in the data here as marginally better than diesel engines, but this is still perhaps 30 per cent better.
In conclusion, it looks as though the jury is still out on hydrogen for fuel cell or combustion engines. There are some improvements for internal combustion engines, and with any such further advances, there will be a need for fuel cells to proceed ahead to remain competitive.
Despite the huge investment in new technologies, we are probably going to stay with the internal combustion engine at least until 2020. This is due to the costs that we saw earlier.
The forecast market for hybrids in 2020, according to the survey shown in this slide on the proportion of market percent assigned to each technology, is about 3 per cent for hybrid fuel cell technology and a growing proportion for conventional hybrids.
So is there a 300 km/h fuelled hybrid vehicle on the road today? The second hydrogen car was expected to beat the world speed record for hydrogen vehicles at 200 km/h, but I think that has gone beyond our grasp with the latest work from BMW. Now I think the topic that we have not yet mentioned here today is the possibility of fusion. Devlopments in this area are still awaited. Thank you.
Sukhvinder Badwal – Harry, thanks for a very entertaining talk, which I enjoyed tremendously. I just wondered: some of the data that you presented on your slides seemed to be fairly old. For one of those slides you mentioned a phosphoric acid fuel cell bus. They were used some time ago. The high temperature fuel cell operated on 200°C. However today people use polymer electrolyte membrane fuel cells, and they can convert power instantaneously at room temperature. For this process, there is no need to use too much energy. Harry Watson – Yes, however I don’t believe they can go to full power as an ordinary internal combustion engine can do from the start. Sukhvinder Badwal (cont.) – They may not, but I am commenting specifically on your energy data. They do not in fact do what you said they could do. The second point I will make is that regarding well-to-wheel efficiency and CO2 emissions. You need to look at the fuel cell hybrid, not just the fuel cell vehicle. Harry Watson – Those were fuel cell hybrids in there, some of them, if you looked at the data carefully. Sukhvinder Badwal (cont.) – I produced the data directly from Toyota this morning, and in fact that quite disagrees with what you are presenting here. Harry Watson – Well, people will have the opportunity to compare our data. I actually did a web search last week to see if Argonne had released their very latest study. When I was in America four weeks ago, they said it would be out any day. However the fuel cell is starting to look better than the internal combustion engine with each study. As I say, however, there is not a lot of information about in-service performance. If this information is not available, then one has to be a bit suspicious that some of the tests might look good but the actual in-service performance is probably not as good as they would like it to be, otherwise I am sure these results would be widely publicised. Bob Watts – Thanks very much. I think we have a fine set of speakers this afternoon. It confirms one of my long-term prejudices, for want of a better word. I have never been terribly worried about global warming, because I feel that the human race is self-regulating. The view towards the future together with efforts made towards dealing with the problem before it becomes too dramatic, I think has been shown in the talks today. Thanks to all the speakers.
Other speakers
Dr John Wright
Dr George Crabtree
Professor Cameron Kepert
Dr Sukhvinder Badwal
Professor Andrew Dicks
Dr Evan Gray
Dr Ben Hankamer Dr Catherine Grégoire Padró
Professor David Trimm
Dr Wes Stein |
© Australian Academy of Science