Dramatic advancements in software efficiency, hardware speed, and modeling accuracy have helped scientists assess a huge database of some 4 million CO2 absorbing minerals, pointing the way to new, lower-cost carbon capture methods.
The researchers have identified a large group of extant and new zeolite materials that could help lower, by as much as a third, the parasitic energy costs associated with removing CO2 from power plant emissions.
Published online in May 2012 in the journal Nature Materials, the anlysis was developed by scientists from three institutions: the Electric Power Research Institute (EPRI), the Lawrence Berkeley National Laboratory (LBNL) at the University of California, Berkeley, and Rice University in Houston, Texas.
In the new study, researchers identified dozens of zeolites — many commonly used in industrial processes — that could significantly improve the energy efficiency of carbon capture technology.
“We believe we can beat current state-of-the-art carbon capture technology by about 30 per cent,” study co-author Michael Deem, Rice’s John W. Cox Professor of Bioengineering and professor of physics and astronomy said in a phone interview. Reducing parasitic power losses during CO2capture could increase a generator’s sellable electric power “by 10, or maybe 15, per cent” based on a back-of-the-envelope estimate, said Deem. “That’s a lot of money.”
The predicted performance gains are relative to current methods, where CO2 is bubbled through a bath of amines. To release the CO2, the amines are heated to boiling, and then the CO2 is compressed into a liquid to be sequestered or used otherwise. Up to one third of the power plant’s steam output is diverted to boil the amines and liquefy the CO2 for shipment.
Computer rendering of the carbon-capture characteristics of a zeolite structure. The arrangement of (red) oxygen atoms and (tan) silicon atoms influences the pore spaces, depicted as green, blue, grey colored surfaces, where CO2 can be captured. Credit: B. Smit/UC-Berkeley.
The new study used computational techniques to identify zeolites that promise to absorb and release CO2 using less energy. Deem explained that Zeolites are a good candidate for this role because they have long been studied, and are used industrially to refine gasoline, as well as to make laundry detergent and other chemically engineered products.
Speaking with GHG News, co-author Berend Smit, a professor of Chemical and Bimolecular Engineering at the University of California-Berkeley, explained:
This round of testing focused exclusively on parasitic load, or the energy penalty needed to separate the CO2 from flue gas, currently considered one of the major cost barriers to the commercial deployment of CCS.
[…]“If the parasitic energy doesn’t go down significantly, then it’s not worth looking at other properties,” Smit said. Of those 5 million materials tested, Smit said roughly 500 turned out to be “promising”…
Comprised mostly of silicon and oxygen, the performance of Zeolites varies by their nanoscale porosity. Made up primarily of silicon and oxygen molecules, the size and shape of pores vary by the geometric linkages made between molecules. In a chemical reaction, each pore acts like as microscopic reaction vessels, bonding and interacting with molecules that fit into the cavity.
The work of Deem et al., focused on sorting through a huge database of zeolite compositions. The roots of the work date back to 2007, when Deem and his colleagues used computers to calculate millions of atomic formulations for zeolites. Adding to this catalog since then, Deem’s database now contains some 4 million structures of zeolite.
In this latest study, the researchers pushed zeolite analysis to a higher level, using a new computer model designed by a team at Berkeley/LBNL to identify candidates well suited for CO2 capture. This model was refined with the addition of technical criteria of ideal carbon capture material, provided by technical experts at EPRI.
Coordinating Deem’s existing zeolite data, with new computational methods, add the additional CO2-capture characteristics, the team predicted the energy demands to capture and release CO2 for all the materials in the zeolite database.
Hardware advancements played a big role too. Given the complexity of the analysis, conventional computational methods using central processing units — or CPUs, the costly, complex chips that serve as the brain of most PCs — would have taken roughly five years to simulate each of the millions of zeolite models in Deem’s database.
Instead, the Berkeley/LBNL team adapted the model to run on graphics processing units, or GPUs, which are specialized, lower-cost processors typically used to render graphics in computers. Switching to GPUs, Deem explained, was integral to the project’s success: “It would have been unfeasibly large to do the old way. Instead of years, the calculation took about a month.”
The graph (below) summarizes the researchers’ findings. The green line gives the current cost of CO2capture through amine recovery. Red dots represent commerciall- available zeolites, while blue dots (both solid and circles) are predicted materials.
“The black curve is the envelope of the best possible performance, within zeolites,” explained Deem. “So if yours is close to this, you can be confident you’re close to best performance.”
The model suggests there are dozens of currently available Zeolites that promise to capture carbon at lower costs than current amine processes. Speaking with GHG News, Smit explained the top candidates will go through further analysis to assess other criteria, such as diffusion limitations and reactions with water.
And if those fail to yield viable performance characteristics or are unavailable because of patent issues, Deem added, the model predicts a large number of viable compounds that can begin to be explored.
As industry begins to sort through the candidates his research has helped identify, Deem is looking forward to add still further attributes to his zeolite database, to further refine for CO2 performance as well as mapping out other potentially valuable chemical interactions.
The team’s research was supported by the Department of Energy (DOE), the Advanced Research Projects Agency-Energy (ARPA-E) and EPRI’s Office of Technology Innovation.
Check out the original post here: http://www.globalccsinstitute.com/community/blogs/authors/adamaston/2012/07/18/computer-modeling-identifies-optimal-zeolites-could