Part 4: Maximizing Operating Time and Solving the Cracking Issues Industry-Wide

Gentilly Nuclear Generating Station. (Image: Rowshyra Castañeda)

Gentilly Nuclear Generating Station. (Image: Rowshyra Castañeda)

Operators of nuclear power plants using Canadian technology joined together to study the cracking issues observed at Point Lepreau and later at Gentilly-2, and ultimately solved the underlying issue for all Canadian reactors.

The events at the Point Lepreau Generating Station (PLGS) had repercussions for Canada’s fleet of reactors, beyond demonstrating safety for their licensing hearings. The need for on-going inspections threatened to reduce their operating capacity, adversely affecting their revenue stream.

Carrying out extensive inspections of the hundreds of feeders in each reactor is not a simple task.[1] The feeders connect directly to the reactor fuel channels and many of the bends are located close to the face of the reactor where radiation is significantly higher when the reactor is operating. Thus, inspections are done when the reactor is not operating to reduce the radiation exposure to workers. Further, the inspection procedures are time-consuming and complicated by different methods required based on crack orientation, or whether it is on the inside or outside surface.

Since a power station loses money every day its reactor doesn’t generate electricity, precise knowledge of the type and location of potential cracks on the feeders that were at the highest risk of cracking was of immense importance to reduce the scope of inspections. Focused inspections would minimize the stations’ down time.

Research using neutron beams provided critical knowledge needed to understand the phenomenon of cracking in feeder pipes, which was impacting some of Canada’s nuclear power plants. This understanding allowed inspections of feeders across the industry to be targeted to areas of vulnerability. As a result, radiation dose received by plant inspection staff was significantly reduced, and plant downtime was also decreased.
– Paul Spekkens, (former) Vice-president – Science and Technology Development, Ontario Power Generation, 2004-2016

The industry pooled money through the CANDU Owners Group (COG) to study the cracking mechanism and to develop more effective inspection techniques.[2] The COG program funded significant efforts to measure the distribution and magnitude of residual stresses in feeder bends using the CNBC frequently from 2001 to 2008, and most recently in 2011, to further understand the role of stress and to help determine which feeder bends were most likely to be affected. The stress measurements covered a wide array of specifications representing the diversity of the hundreds of feeders at each reactor, and the differences between the feeders at each station. These specifications vary in factors such as pipe thickness, angle of the bend, or the methods used to manufacture the bend, install the feeder, or make repairs.

Because welding is known to produce stress, some of the first such studies at the CNBC were to measure stress in repair welds of the type used to replace the cracked feeder bends at PLGS, and in the original welds used to install the feeders. While this research was taking place, a leak through a repaired weld on a feed tube was discovered at the Gentilly-2 Nuclear Generating Station (GNGS) in Quebec in June 2003. In contrast to the leaks at PLGS, the leak rate was low enough for the reactor to operate until the next planned two-month outage on August 31. This outage was extended to four months due to greater-than-expected complexity of the repair of the cracked weld on the feeder pipe and as well as to address other unrelated maintenance projects undertaken at the same time.[3] Hydro-Québec extended inspections to include all feeder hub welds considered at risk for cracking, but no additional cracked welds were found.[4] Examination of the cracked weld after the outage pointed to high stress as a major underlying factor.

The incident at GNGS heightened the need for stress data from feeders under various conditions of welding, and the CNBC was soon employed for these studies, including but not limited to, stress in an ex-service feeder weld from GNGS and in test feeders before and after applying the same repair method used on the cracked weld at GNGS.

By 2005, the stress data from the CNBC had helped the industry produce a relative ranking of feeder bends in order of highest to lowest likelihood of cracking, thus: tight radius bends of angle greater than 45°, repaired welds, tight radius bends of angle less than 45°, long radius bends, and non-repaired welds.[5] These findings informed the fitness-for-service guidelines produced by COG, thereby providing a framework within which the utilities could operate with approval from the CNSC. Importantly, the stations were able to successfully demonstrate that the frequency or scope of inspections could be reduced, for example, by focusing on the bends at higher risk.[6]  Reduced scope of feeder inspections has resulted in significant time and resource savings in during outages, thereby minimizing station down time.

These guidelines continued to be informed by subsequent stress measurement in the following years. The final examinations at the CNBC in 2011 were ex-service feeders that provided data showing how stable the stress had remained over time. One of these feeders contained a partial crack, providing a unique opportunity to compare the stress distribution with un-cracked feeders and providing evidence that could support predictive studies of how such cracks would evolve over time.[7] Such a predictive study could inform decision-making about whether to replace a feeder when inspections discovered partial cracks.

The cumulative results of the decade of COG-funded research and operating experience following the initial cracking incidents at PLGS have effectively solved the problem of feeder cracking in two ways:

  1. Better materials and maintenance methods have been qualified using neutron stress measurements that provide high confidence that feeder cracking of the types seen at PLGS and Gentilly-2 will not be significant concerns for new reactors or for the existing reactors after the feeders are replaced during the planned refurbishments.[8]
  2. In the meantime, safety is assured through the use of standard fitness-for-service guidelines for feeders that are now in place across the industry. These guidelines have enabled the stations to keep the scope of inspection programs down to manageable levels.


The availability of neutron beams at the CNBC to easily obtain stress data non-destructively has been of immense benefit to the entire industry, helping it to demonstrate safety, assure export clients, and operate reliably. While many other research studies and contributed to these impacts, stress data from the CNBC played a critical role. The value to Canada of the impacts of this one line of research can be estimated to be in the range of hundreds of million dollars, exceeding all of Canada’s direct investments in the neutron beam laboratory at Chalk River since neutron scattering was pioneered there in the 1950’s.

Next: Back to Series Summary


[1] CBC. Lepreau may soon be online again. April 6, 2001.
See also: CNSC. Significant Development Report No. 2001-8. 2001-11-09. Accessed from Energy Probe Research Foundation:

[2] John P. Slade, and Tracy S. Gendron. Risk-Reduction Strategies used to Manage Cracking of Carbon Steel Primary Coolant Piping at the Point Lepreau Generating Station. 12th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Salt Lake City, USA, The Materials, Metals & Materials Society, 2005 August.

[3] There was also unplanned corrective maintenance on the turbine.

[4] CNSC. Annual CNSC Staff Report for 2003 on the Safety Performance of the Canadian Nuclear Power Industry. November 2004.

[5] John P. Slade, Tracy S. Gendron. Risk-Reduction Strategies used to Manage Cracking of Carbon Steel Primary Coolant Piping at the Point Lepreau Generating Station. 12th International Conference on Environmental Degradation of Materials in Nuclear Power Systems – Water Reactors, Salt Lake City, USA, The Materials, Metals & Materials Society, 2005 August.

[6] John C. Jin and Raoul Awad. Regulatory perspective on CANDU feeder pipe degradation due to flow accelerated corrosion (FAC) and intergranular stress corrosion cracking (IGSCC).  2009.

[7] D. Banks, R. Donaberger, B. Leitch, R.B. Rogge. Stress Analysis of Feeder Bends Using Neutrons: New Results and Cumulative Impacts. 2014. Pacific Basin Nuclear Conference. Vancouver, BC, Canada. PBNC2014-186.

[8] Slade and Gendron. A Last Look At PLGS Life-Limiting Feeder Degradation. 8th CANDU Maintenance Conference. 2008.

Series Overview

An unexpected incident at a Canadian nuclear power plant led to an urgent need for greater scientific understanding of the underlying cause. Stress measurements with neutron beams played a critical role, both in the immediate failure analysis and in subsequent activities over a decade to assure the safe, reliable and economic operations of Canada’s nuclear power plants. This series of articles reviews the resulting impacts on the industry, which include informing safety evaluations, providing confidence for a multi-billion export project, minimizing down time, informing lifetime management, and ultimately solving the underlying issue for the long term:

Summary: A Decade of Feeder Studies for Canada’s Nuclear Power Technology

Part 1: Responding to Cracked Feeders at Point Lepreau

Part 2: Managing Risk of Feeder Cracking at Point Lepreau

Part 3: Assuring Exports, Relicensing of the Reactor Fleet, and Qualification of Innovation

Part 4: Maximizing Operating Time and Solving the Cracking Issues Industry-Wide

feeder-tubesMock-up of a CANDU reactor face. The feeders are pipes (black) attached to each end of the hundreds of channels (yellow) through the reactor core holding the nuclear fuel. The feeders circulate heavy-water coolant from the fuel channels to the steam generators. Each feeder contains multiple bends. (Image: NA Engineering)