February 28, 2006

Putting Hard Numbers on Global Warming

Arizona  -- There's a lot of talk about global warming, but not much hard data on the chemicals that cause it.

In fact, engineers who design environmentally friendly manufacturing processes often are stymied because data on a chemical's global warming potential (GWP) just isn't available.

A University of Arizona chemical engineer is working on computationally based methods that should produce the hard numbers needed by both engineers and policy makers who are trying to curb global warming.

Paul Blowers, an assistant professor in UA's Chemical and Environmental Engineering Department, first encountered the global-warming data vacuum when he set out to create an elective course in which students would evaluate manufacturing processes for their environmental damage.

"But it turned out that for nearly every chemical I picked, I couldn't find information on its toxicity, global warming potential or other factors influencing its effect on the environment," Blowers said.

Then he attended an academic conference in Berlin and found that he wasn't alone. Everyone else was having the same trouble finding data on the environmental effects of chemicals.

In most cases, the measurements have never been made and the data simply doesn't exist. For instance, toxicity data "” cancer-causing potential and other harmful effects "” isn't known for 95 percent of the chemicals in use today, Blowers said.

There's a good reason for this. Gathering the experimental data is expensive and time consuming.

Experimental Approach is Expensive

The spectroscopy equipment needed for some of the GWP analyses costs $1 million. Another $500,000 is needed for equipment that will do other measurements. "So you're looking at $1.5 million just to get the input data you need to start determining the global warming potential," Blowers said. After that, computing the atmospheric lifetime of the chemicals requires thousand of hours of expensive supercomputer time.

Just calculating how much solar energy a chemical traps in the atmosphere requires heroic experimental work. "I estimated from one of the papers I read that it took 15 researchers five years to determine the GWP for one chemical," Blowers said. By contrast, Blowers can calculate this energy-trapping potential in about 15 minutes using his theoretical models.

"When I returned from the Berlin conference, I decided that I might be able to obtain some of the environmentally related data on some chemicals entirely from theory," Blowers said.

"I didn't know if I could predict global warming or ozone depletion numbers from theory, but I told my students that all the topics we would pick for our research group meetings would deal with this subject."

After a year's work, Blowers and his students completed their first prediction of global warming potential for a chemical last November.

They modeled difluoromethane (CH2F2) "” a chemical that's commonly used in fire extinguishers "” because it's been analyzed experimentally and they could check their calculations against the existing data.

"At all the intermediate points that we could check, our predictions were nearly identical to the experimentally measured values," Blowers said.

Many Factors Influence Global Warming

He noted that several factors go into predicting a chemical's global warming potential from theory and there are millions of details involved. Existing software packages can handle some of the calculations, but none of them has everything that's needed, and plenty of hands-on data analysis is required.

Several factors have to be known before a chemical's global warming potential can be calculated, Blowers explained. These include:

Radiative forcing "” The amount of energy that can be trapped by chemicals as it is radiated from the earth. A positive number indicates the atmosphere is being warmed.

Activation energy "” The amount of energy needed to start the reaction that will break down a chemical in the atmosphere.

Reaction rate "” How fast a chemical breakdown occurs, depending on the concentration of the chemicals, energy available for the reaction and other factors like the activation energy.

Geometry "” How the atoms are physically arranged in a molecule.

Deciding When the Answers are Good Enough

In each case, Blowers needs to decide when the answer is "good enough." The quality of the answer has to be balanced against the computer resources available. "Ideally, you'd want a perfect answer for every factor involved, but it would take an infinite amount of time with a computer larger than all the super computers on Earth," he said.

While some answers can be less precise, others require a high degree of accuracy, he explained. A difference of just two kilocalories in the activation energy can lead to large changes in reaction rates for some chemicals. In some cases, this two-kilocalorie difference could change the time they would last in the environment by a factor of 10. So instead of degrading in 10 years, they might stick around for 50 or 100 years. (A kilocalorie is the energy needed to raise the temperature in a liter of water by one degree centigrade.)

Even when using "good enough" calculations, Blowers often has a laptop analyzing data at home, another running in his office, a desktop handling other calculations and a routine running on a super computer at the University of Illinois.

Although the methods used to predict the global warming potential worked nearly perfectly for CH2F2, Blowers found they don't work for all chemicals. So now his research group is focusing on refining the models. "We know all the mechanics," he said. "There are no more surprises or unknowns left. Now we just need to improve our methodology and make it more robust."

"Once I have the method perfected, it should apply to any chemical," he added. "I want this to be a robust method where I just go and say, 'Here's a new chemical. I'm going to go through this mechanical series of calculations and I'm going to get a global warming potential that's going to be right.' That's my goal."


On the Net:

University of Arizona