Simulating Power: Fuel-Cell Design and ITBy Baselinemag | Posted 2005-12-06 Email Print
Plug Power's CEO talks about the the future of IT in his company, which relies on sophisticated modelling to research and product fuel cells that could eventually replace generators in some disaster situations.Roger Saillant, president and CEO of Plug Power, a Latham, N.Y.-based fuel cell manufacturer, likes his information systems to go with the flow-specifically, the flow of heat, hydrogen, water and coolant.
With oil prices hovering around $60 a barrel, fuel cell companies have caught the fancy of Wall Street, corporate customers and the feds.
That interest, however, hasn't translated into profit. For the nine months ended Sept. 30, Plug Power lost $35.3 million on revenue of $10.56 million.
One of the ways Saillant, who joined Plug Power in December 2000 after 30 years at Ford Motor and auto-parts maker Visteon Corp., hopes to change that is to use computer-aided design to reduce the time it takes to engineer its products, such as power generators, hydrogen backup power products and compressed hydrogen systems.
Baseline's Larry Dignan spoke with Saillant about information technology and the future of alternative energy for consumers, business and governments.
How does technology affect your business?
When I look at information technology in its broadest sense, it is research and development, graphics, drawing and computer-aided-design tools, simulations and modeling. Specifically, we view it as a time-saver. When we do modeling and simulation, we can do so many iterations that if we [built models by hand], it would be very expensive and take a long time. By the time it gets done, we have an accurate reflection without having to bend metal and build prototypes.
What exactly are you simulating?
In fuel cells, the biggest thing to simulate is flow patterns through the stack [an assembly of individual fuel cells, each producing about one volt of electricity]. How does the coolant flow? How does the hydrogen flow in and the water flow out? What does the heat transfer look like?
You need to look at combinations that give you the best flow and the most efficient conversions.
How long does it take to model a product like a system to power a home?
It depends. If it's modeling a specific heat transfer element for a radiator design, you may go through hundreds or thousands. The whole system itself is modeled for packaging, serviceability and footprint perspective maybe three or four times. Ultimately, you have to service these things and assemble them. Design in our case also includes disassembly, so you can pull the parts and reuse or remake them rather than put them in a landfill somewhere.
Do you build or buy the systems used to design products?
We buy. We have a relationship with PTC [Parametric Technology Corp., a Needham, Mass.-based product life-cycle management software company]; we are a beta site for them. Our guys are so voracious that we can sort through and find bugs to feed back in real time. We help them put their modules together, and there are variations where we co-develop.
Telecom and utility companies are your biggest customers. Are these purchases for disaster recovery?
Eventually, they will be used to mitigate the impact of outages from severe weather such as hurricanes, tornadoes and ice storms. Batteries may go out in one or two hours. Fuel cells could run 10 to 20 to 40 hours. For a telecom, that would allow cell phone usage in those areas.
Can these generators be used to power factories in outages?
There is a wide range of different technologies in fuel cells. We're in a technology that I think will be good for generating 750 watts to maybe 100 kilowatts. Right now, we're focused on generating 5 kilowatts-a level that would generate power for a home. A 7-Eleven, however, is somewhere in the neighborhood of 15 to 25 kilowatts. To really run a big business, you'd need megawatts. There is a technology by United Technologies and FuelCell Energy built on 200- and 250-kilowatt modules. They both are competing on a contract with Long Island Power [Authority in New York] for 10 megawatts of power.
When do you think fuel cells will be widely adopted?
I don't know when fuel cells will be widely deployed. The big hurdle is, what will the killer application be? What's the cost and reliability target? After you get the cost and reliability right, you have a cultural and inertia issue to overcome.