SHARP could slash nuclear reactor design costs

November 16, 2012 by Renee Nault
Coolant-flow pressure distribution in a 217-pin wire-wrapped subassembly, computed on P=65536 processors of the Argonne Leadership Computing Facility's Blue Gene/P. Credit: Argonne National Laboratory

Back in the earliest days of nuclear energy, Argonne physicists and engineers used slide rules and their own basic knowledge of reactions and physics to design nuclear power plants. Then, beginning in the early 1960s, they enlisted computers to develop designs with data from experiments and actual reactor testing. Over this entire span, Argonne built over 85 experimental reactors to test its reactor designs and computer programs – each a costly and time-consuming endeavor.

Despite all this, many of the highly complex that affect reactor performance and safety remained somewhat of a mystery. It wasn't possible to "see" what was taking place inside this very —until now.

Researchers are using some of the world's most at the Argonne Leadership to take a leap forward in design, analysis and engineering. Their efforts could shave millions of dollars off the cost of reactor design, development, preparation for licensing, and construction.

Researchers have developed a suite of computer tools, called the Simulation-based High-efficiency Advanced Reactor Prototyping (SHARP) Reactor Performance and Safety Simulation Suite, that numerically mimic and allow researchers to "see" the physical processes that occur in a nuclear reactor core. SHARP users can build complex virtual reactor models, which can run the reactor through a variety of operational or accident scenarios that would be impractical or impossible in the real world.

SHARP builds upon existing computer codes used to conduct safety evaluations of today's aging nuclear power reactors. When you want to use those older codes to virtually test new reactors, however, you run into a few problems. They're well calibrated for evaluating the safety of new reactor designs, but they aren't so great for optimizing designs for efficiency or cost. SHARP has been written specifically to simultaneously look at the safety, lifetime, and performance of different advanced design ideas.

This is a perfect fit for Argonne's nuclear program, which has a half-century of experience with reactors under its belt and one of the world's largest concentrations of researchers involved in advanced reactor design. For example, early in the program, the code creators wanted to test how accurate the codes could be. They built a virtual model of one of Argonne's historical test reactors, the Zero Power Reactor, ran "experiments" on it and checked how well the computer results matched up against the actual data.

Modern codes like SHARP are already successfully replacing some types of nuclear experiments. SHARP's next step is refining the code to make it run more efficiently on powerful supercomputers like Argonne's new Blue Gene/Q. More efficient codes mean scientists can run more tests during the same computer time with fewer problems, making SHARP a more useful tool for reactor designers around the world.

This story was originally published in volume 6, issue 1 of Argonne Now, the laboratory's biannual science magazine.

Explore further: Sustainable nuclear energy moves a step closer

Related Stories

Sustainable nuclear energy moves a step closer

December 11, 2006

In future a new generation of nuclear reactors will create energy, while producing virtually no long-lasting nuclear waste, according to research conducted by Wilfred van Rooijen, who will receive his Delft University of ...

Using supercomputers to explore nuclear energy (w/ Video)

January 22, 2010

Ever wanted to see a nuclear reactor core in action? A new computer algorithm developed by researchers at the U.S. Department of Energy’s (DOE) Argonne National Laboratory allows scientists to view nuclear fission in much ...

GE defends nuclear plant design

March 18, 2011

General Electric defended its 40 year old Mark 1 reactors at the center of Japan's nuclear crisis Friday, saying that early questions about reactor's safety had long been addressed.

Recommended for you

For these 'cyborgs', keys are so yesterday

September 4, 2015

Punching in security codes to deactivate the alarm at his store became a thing of the past for Jowan Oesterlund when he implanted a chip into his hand about 18 months ago.

How to curb emissions? Put a price on carbon

September 3, 2015

Literally putting a price on carbon pollution and other greenhouse gasses is the best approach for nurturing the rapid growth of renewable energy and reducing emissions.

Magnetic fields provide a new way to communicate wirelessly

September 1, 2015

Electrical engineers at the University of California, San Diego demonstrated a new wireless communication technique that works by sending magnetic signals through the human body. The new technology could offer a lower power ...


Adjust slider to filter visible comments by rank

Display comments: newest first

1 / 5 (1) Nov 16, 2012
What does it cost to design a reactor? If it is in the region of hundreds of millions, then to "shave millions of dollars off the cost" might be just one percent or less. Every little bit helps, but it would be nice to know how much.

I expect that many of the people who trust climate change models will be reluctant to trust computer models that predict a reactor design is safe. Conversely, I think many of those who say computer modeling is worthless when they criticise climate change will be happy to trust computer models when it comes to reactor safety.
not rated yet Nov 16, 2012
Its billions.
not rated yet Nov 16, 2012
Although the cost of building a reactor runs into the billions (as does any type of power plant in the gigawatt range), the design is a relatively small part of it. If the French model of 'design once, build many' was used in the States, the proportion of design cost would drop even lower.
not rated yet Nov 16, 2012
Sigh - I agree with your observation about some of the contradictions - we humans are complex - no? I think there is a convergence of reality and models. As the models help us understand these complex systems - we tweek the models - and they reflect the reality better - etc. etc. As time goes on - modeling takes over more and more of the task of design and testing - development accelerates...
1 / 5 (2) Nov 17, 2012
Does it model the cost of cleaning the fallout as well?

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.