Automatic System For Crack Detection in Nuclear Power Plants Developed
It is common knowledge that accidents in nuclear power plants lead to hazardous consequences with long-term implications. Examples include the Chernobyl catastrophe in Russia.
One of the reasons for accidents is the failure to detect the cracks on components well in advance. Routine inspections may not be able to identify the cracks and they evade detection because the inspection methods may not keep pace with an aging plant's structural problems.
Thanks to a new automated system developed at Purdue University, the steel components of nuclear power plants can be examined more accurately compared to other existing systems.
This was revealed in a study published in the Computer-Aided Civil and Infrastructure Engineering journal. The paper was authored by Fu-Chen Chen, a doctoral student.
Periodic Inspection Essential
As nuclear plants age, they face problems of fatigue, wear and tear, erosion, embrittlement of metal components, corrosion, and oxidation. This calls for a stepped-up periodic foolproof inspection to ward off any future calamities.
"Periodic inspection of the components of nuclear power plants is important to avoid accidents and ensure safe operation," noted Mohammad R. Jahanshahi, an assistant professor at Purdue University's Lyles School of Civil Engineering and the paper's co-author.
According to the expert, the current inspection methods have many defects, including lack of objectivity that takes away a lot of time and the faults of operators who try to manually locate cracks in metallic surfaces.
The new automated system called CRAQ uses advanced algorithms and machine learning to detect cracks on the basis of changes in the texture that appear on steel surfaces.
Detecting cracks in metallic surfaces is a challenge because many of the automatic crack detection algorithms cannot trace them as they are too tiny and hard to distinguish from welds, scratches, and grind marks.
Remotely Recorded Videos
At the moment, remotely recorded videos are used for inspection. The complexity of the inspection process of nuclear plants is aggravated by the fact that nuclear reactors are submerged in water for cooling purposes.
"Consequently, direct manual inspection of reactor internals is not feasible due to high temperatures and radiation hazards," Jahanshahi said.
He noted that cracking-led degradation would lead to hazardous accidents and huge financial costs.
"For instance, the Millstone nuclear power station in Connecticut had an accident in 1996 that was caused by a leaking valve, and the accident cost $254 million," the co-author added.
According to the researchers, greater reliability needs to be maintained in accepting the results of videos taken during inspection because they are recorded at the underwater reactor surface with scope for many imperfections.
The Purdue researchers, in evolving their new system, used videos taken by an underwater camera system and scanned more than 300 stainless steel specimens that had cracks, grinding marks, scratches, and weld marks.
Their analysis went beyond the conventional single-image processing as they used multiple video frames to arrive at the best results. The new system showed itself as superior to many other systems.
In the new method, the cracks have been identified by a method called "Bayesian data fusion," which tracks cracks via video frames from the information coming from multiple frames.
Certainly, the new system will benefit the American nuclear plants, which recently received an overall D+ rating by the American Society of Civil Engineers.
Meanwhile, Toshiba's decision to quit the business of nuclear power plants has hit the sector, as innovation and research on reactor designs will be a casualty.
A majority stake in Westinghouse Electric by the Japanese conglomerate had raised hopes that new generation power plants that are safer, smaller, cheaper will be in the offing.
However, rising cost overruns, technical problems, and regulatory challenges in many projects led to Toshiba announcing a $6.3 billion write-down in the nuclear business and a planned offloading of its stake in the company.
"It looked like a big deal at the time, but it's turned into a mess," said Michael Golay, a professor at MIT.
The retreat of Toshiba comes when the International Energy Agency estimated that nuclear energy capacity would need to double by 2050 to prevent worldwide temperatures from increasing more than 2 degrees Celsius.