Go to the photo gallery
One of the ideas that we developed at the plasma institute was the idea of take the flat gliding arc geometry and placing it into a reverse vortex flow. The idea of doing this was to try to achieve better plasma interaction with whatever gas we put into the reactor. A reverse vortex flow can be made by inject a gas through tangential jet into a cylinder. The key feature of reverse vortex is that the exhaust hole is on the same side as the tangential jets. Here is a picture to try to explain this better:
The picture demonstrates how the gas comes in at the same side as the gas goes out. The first picture shows how the gas swirls down and the second picture shows how the gas turns around and heads for the exhaust hole. There are many benefits to this type of geometry. One feature of this flow is that it provides excellent stablization of a flame. This can be see here:
What can be seen in this picture is a flame produced by injecting methane gas into the tangential inlets. When the methane is lit the flame can be seen to become stable in the center of the cylinder. The other great feature of this type of flow is the cooling effect that takes place. We can see that Dr. Gutsol's hand is only about a centimeter away from a very hot 2kW flame. What is happening is that there is methane gas swirling around the flame which steals the heat that is trying to escape outwards and puts it back into the flame once the gas makes its turn near the bottom. There are some other key features of this flow geometry that will become more apparent in my section of H2S dissociation.
The next task, which is an interesting engineering problem, is to put plasma into this flow instead of a flame. Recall from the flat gliding arc section that the plasma arc begins by discharging across a small gap between two electrodes. The arc is then pushed and caused to elongate. In order to accomplish this in a reverse vortex flow several routes can be taken. The first idea was carried out by Chiranjeev Kalra, Dr. Gutsol, and Dr. Fridman at the Drexel Plasma Institute. The idea was to place a movable electrode in the center of the reactor. With the electrode beginning close to the top electrode an arc can form when a high potential is made across the electrodes. The bottom electrode can then be moved mechanically downward, in effect elongating the arc and creating the reverse vortex gliding arc plasma. There was also the idea of simply having an electrode that is of a spiral shape. These two designs can be seen here:
Remember there is only a single arc here, it is just moving so fast that it appears similar to a flame. Notice how the plasma is able to fill the central volume of the reactor. This is what makes the reverse vortex gliding arc design superior. There were some difficulties with this design however. One difficult task is separating high voltage from ground. Typically this is done my placing an insulating material between the two. However these materials tend to be fragile when we require the ability to go to high temperatures. When quartz is used, it becomes a challenge on how to create a good seal between the quartz and the electrodes. Another issue with this design is that there is now obstructions within the reverse vortex flow. The electrodes are in the way of the natural flow. Therefore it was the next idea to try to remove the electrode from the volume somehow.
When Mike Gallagher finished his masters work he wanted to pursue his Ph.d in fuel conversion technologies. This was of great interest to me as well. It became our task to demonstrate that the electrode could be removed somehow from the center of the reactor. The idea was to use a metal wall instead of a quartz wall. The setup was not easy to create. The basic setup had to consist of a bottom electrode with an exhaust hole in it, on top of this was a thin insulating material where the tangential jets were, and then on top of that sat a large metal cylinder. The hope was that the arc would be pushed up the wall. After much effort we were able to get it to work.. It is difficult to look at the plasma now because it is all enclosed by metal. However we placed a clear top on the reactor so that we could watch the plasma. Unfortunately I could not find any pictures of this original setup, but here is a picture of the setup we used to take some pictures after we had done several experiments:
Here is what the top down view looked like:
Remember again, this is only a single arc moving very quickly. Don't believe me? We were able to capture a picture of the arc:
Pretty cool huh! The concept was shown to work, now it was time to design a reactor.
National Energy Technology Laboratory (NETL), a branch of the Department of Energy, had expressed interest in the fuel reforming of diesel. The motivation for this project was due to truck drivers. When truck drivers drive cross country they spend a lot of time on the side of the roads when they need a break. During this time they typically will leave their engines running so that they can power their appliances inside of the truck. This is very inefficient and it causes a lot of pollution. The idea instead is to take some of the diesel fuel and reform it in our plasma reactor into hydrogen rich gas. This gas could then be fed into a fuel cell to produce the electricity the truck driver will need. This process would be more efficient and would cut back on pollution.
There were many difficult constraints on the design of this reactor making this a true engineering feat. The fuel was to be evaporated and then injected into the reactor along with a proper amount of air. Here is the design:
The complexity of this design is apparent, but what it allows for is the injection of evaporated fuel, mixed with air tangentially and the possibility to inject another gas from the top axially if desired. The entire setup is sealed inside of a long metal tube which is grounded for safety reasons. Anybody whom has worked with these National Government labs will know important safety is for them. They have tons of safety requirements that must be followed. For example Mike had to do calculations on the explosion limit of the vessel and we had to pressure test it at very high pressures as well. I cannot find any good pictures of the reactor, but here are some pictures the setup when we were running an experiment. They are dark because we had the lights out to watch the plasma through a window that we place on top.
The white strip wrapping around the tube is our heater.
Here you can see a flame in the back, this flame is our exhaust gas. Once the reforming starts we begin to produce hydrogen and carbon monoxide. We do not want to let these gases just be vented outside so we ignite them creating this flame in the background. This is also a good conformation that we are properly reforming the fuel.
Diesel is a not an easy fuel to work with, it contains a mixture of long hydrocarbon chains. Due to this fact, it is not very easy to get a proper fuel to air ratio. For this reason, it is easier to work with a simulated diesel fuel. In our case NETL said that we should work with tetradecane. This makes things much easier experimentally because we know that tetradecane has the molecular formula C14H30. With this knowledge we can determine the appropriate fuel to air ratio or more to the point the proper equivalence ratio. The equivalence ration is useful to use while we are trying to enter the partial combustion regimes for which we are aiming.
In order to know how well the fuel conversion is doing, we analyze the samples on our Gas Chromatograph/Thermal Conductivity Detector (GC/TCD). We use two columns (a gaspro and a molesieve) each with their own TCD. This allows us to detect H2, CO, CO2, O2, N2, and low level hydrocarbons. We have it calibrated to see only the C2's because we generally will not produce much of the higher hydrocarbons.
With the GC analysis we are able to perform our mass balances and see if the conversions were good. I will power some of our data soon, however I can say that we typically can have as much as 90% conversion of the fuel to useful products, meaning hydrogen and C2 hydrocarbons. The energy cost for running the plasma discharge is around only 2% of the chemical energy produced by the reforming.
Plasma Reforming Technology Project
Liquid Hydrocarbon Fuel Reforming Studies