It was just before Thanksgiving, 2008, when Mines doctoral student Matthew Walsh sat down at a keyboard in the chemical engineering building, typed in a few final instructions for the nearby supercomputer Ra, and took off for his holiday vacation.
By the time he returned in January, he and his 7-foot-tall, 14-footlong virtual colleague had made history.
“I’m lucky,” says Walsh, 29, whose resulting scientific research paper, “Microsecond Simulations of Spontaneous Methane Hydrate Nucleation and Growth,” appeared in the November 2009 issue of the prestigious journal, Science. “I am just a guy who pressed go.”
Humility aside, Walsh’s groundbreaking study (co-authored by faculty advisors Carolyn Koh, E. Dendy Sloan, Amadeu Sum, and David Wu) marked the first direct simulation of the nucleation, or birth, of a natural gas hydrate, a historic breakthrough in the 70-year-old field of hydrate research, and a shining example of how the power of supercomputing has begun to make the impossible possible in energy research.
L to R: Amadeu Sum, Matt Walsh '03 and David Wu
For decades, the oil and gas industry has been at once intrigued and frustrated with hydrates – mysterious crystals of burning ice that form spontaneously on the sea floor, in the permafrost, and inside oil and gas pipelines. On the one hand, they are a colossal nuisance, slowing flow inside pipelines to a halt and, in a few dire cases, causing deadly explosions. But on the flip side, they are an enormous untapped resource, with the tiny cages of water molecules containing a wealth of methane and other gases. “By some estimates, there is more energy trapped in these natural gas hydrates worldwide than all the conventional fossil fuels combined,” explains Walsh.
Until now, researchers using conventional laboratory technologies and computers have been able to learn much about the circumstances under which hydrates form (under high pressure and low temperatures) and their physical make-up (a crystalline structure in which water molecules form polyhedral cages around methane molecules). But because their birth is a rare event that happens in a few nanoseconds
at a random location, two questions have eluded scientists: What happens at the molecular level to spark hydrate creation, and how fast do they grow?
“We wanted to study, not where and when, but how? How do the water and the methane molecules re-arrange themselves to form a hydrate,” explains Walsh.
The 2008 arrival of Ra – a supercomputer capable of processing 23 trillion operations per second – brought the unanswerable question within reach.
Over the course of several months, Walsh and his colleagues designed a complex simulation in which they would provide Ra with the number of water molecules and methane molecules, the temperature and pressure, and Ra would use Newton’s Second Law of Motion (force equals mass times acceleration) to paint a tantalizing
mathematical glimpse at what would happen as those thousands of molecules bounced around each other.
In real time, Ra calculated for two months, simulating just two microseconds (two one-millionths of a second) of molecule interactions. But in the world of molecular computer simulation, two microseconds is a relative eternity. By the time Walsh had returned from winter break, something extraordinary occurred. A hydrate was born.
“It was amazing,” recalls Walsh. “The conventional wisdom had always been that you can’t simulate it.”
Within weeks, fascinated engineering students and curious oil and gas industry experts were logging on to YouTube to witness Walsh’s water and methane molecules engaged in a chaotic dance, culminating with the former enslaving the latter in an elegant cage, and more cages blossoming from its faces as time stepped on. Within 10 months, Walsh was basking in a career coup.
“It is very unusual for someone at this stage of their career to get a publication in Science,” says co-author and advisor Wu.
Adds Carolyn Koh, a study co-author and co-director of the Mines Center for Hydrate Research. “His creativity, innovation, and perseverance really pushed us toward this discovery.”
Ultimately, Koh says, such research will advance the path toward helping oil and gas companies devise a chemical inhibitor to prevent the formation of hydrates in pipelines, scientists create hydrates to use as vessels for transporting hydrogen fuel, and alternative energy seekers devise ways of extracting the dormant untapped resource from our ocean floors.
But for Walsh, there is not time to bask in the glory. There is work to be done. In the coming year, he and his advisors hope to perform more simulations that more closely reflect the temperature and pressure inside a pipeline, and mine the data even further to see just how fast a hydrate forms under those revised circumstances.
“We now have a qualitative understanding of how a hydrate forms. I would like to reach the quantitative level,” he says. “These movies are admittedly cool, but they are not quite the hard science I am after yet.”