U.S. NRC Blog

Transparent, Participate, and Collaborate

An Inspector’s Perspective On the Control Rod Drive Mechanism Housing Flaws At Palisades

Elba Sanchez Santiago
Materials Engineering Inspector
NRC Region III
 
Elba Sanchez Santiago is a Materials Engineering Inspector in the NRC's Region III.

Elba Sanchez Santiago is a
Materials Engineering Inspector in the
NRC’s Region III.

There has been a lot of interest lately in the flaws that were recently found in the control rod drive mechanism (CRDM) housings at the Palisades nuclear plant, near South Haven, Mich. I want to share my direct experience with the NRC’s thorough and independent evaluation of this issue.

First, some background. The control rod drive mechanism moves control rods inside the reactor to control the level of nuclear chain reaction. The housing is a metal tube around the control rod drive mechanism, which is connected to the control rod and prevents leakage of reactor water into containment.

According to a commitment made in 2012, the plant conducted inspections of 45 CRDM housings in this reactor and found flaws in 17 of them. Palisades committed to these inspections after the discovery of a crack in one of the housings resulted in a plant shutdown in 2012.

Because of my expertise as a materials engineering inspector, I was dispatched to Palisades after it shut down in 2012. I was to evaluate the plant’s response to the discovery of the through-wall crack. As a member of a special inspection team that further evaluated this issue, I reviewed the plant’s testing of eight additional CRDM housings and their corrective actions. Even though no other cracks were found, the plant committed to further evaluate the condition of the housings during the 2014 refueling outage.

I came to Palisades before the current outage started to evaluate the site’s inspection methodology, work procedures, tooling and personnel qualifications. When the examinations started, I observed some of the actual testing and evaluated the results. To date, there is no evidence of leakage resulting from the flaws. I will remain onsite providing oversight over the plant’s actions until the issue is resolved.

Since the issue first came to light in 2012, I have been working with a team of other inspectors and specialists in Region III and the headquarters office in Rockville, Md., to make sure we ask the necessary questions to understand the plant’s methodology and assessments, and independently verify the conclusions.

Our in-depth independent reviews will continue until the plant completes the necessary repairs and takes proper actions to make sure the CRDM housing flaws do not lead to a significant safety concern. The results of our inspections will be documented in a publicly available inspection report.

7 responses to “An Inspector’s Perspective On the Control Rod Drive Mechanism Housing Flaws At Palisades

  1. Mike Mulligan February 10, 2014 at 6:57 pm

    Come on, in the 2012/2013 assumption in this outage is you would test 25% of the CRDM and find no cracks. Then test 25% thereafter. It is the normal condition that a plant finds no cracks on their CRDM throughout the life of the plant.

    So far you found 17 CRDM having cracks …that is 37% of the rods. You usually find lots more cracks after the all the is results come in. You we can’t trust the agency to anticipate this blossoming level of degradation…how can we trust the agency to anticipate a leak? How come the agency didn’t see this coming? You know, there is a astonishing number of violations going back years with the CRDMs and quality. The 2012 leak violations was just a repeat of the 2001…and 2001 was violations repeat of other prior incidences.

    Just saying, why wasn’t the state of the art with metallurgy able to predict in 2012 that Palisades would find “at least” 37% of their CRDMs had cracks in them. Believe me, based on all the past reports I read on this, you won’t disclose the full numbers of cracks in the CRDMs (more than one per CRDM) till after restart. Why can’t these licensee PhD metallurgist predict future flaws and cracks, instead of justifying past flaws?

    It is illegal to start-up and operate if they had evidence CRDM unidentified leakage was increasing…they assumed they didn’t have leaking CRDM because they didn’t have full vision of the CRDM. They did have perfect evidence that the unidentified leakage wasn’t a CRDM leak.

    Lets remember with the incomplete information accident in the Davis Besse head event. The licensee and the NRC assumed the CDRM flanges were leaking when it was a crack through the CRDM nozzle and eating the head. Who in a new different accident is talking about they were “refusing to communicate uncertainty“.

    For decades there is a widespread mindset in these organization that piping cracks and flaws don’t lead to leaks. That is what is behind this.

    You get it, these CRDM leaks at Palisades tend to show up within a month or so of a start-up and the leak worsens quickly. I’ll bet you both leaks (2001 and 2012) actually started before start-up. Palisades has a pattern of calling a prohibited CRDM leaks not a leak….cold bodily waiting to the CRDM leak gets to .3 GPM unidentified leakage or more according to their procedures. They have a requirement not to operated with pressure boundary leak and they chose to ignore the rules!

    I am just saying next operating cycle, how can you trust these guys to follow the rules? How can you trust these guys with a pattern of behavior like this…when these guys have abnormal or increasing unidentified leaking, when they don’t have absolute proof a pressure boundary “is not leaking“…how can you trust them to do the right thing with incomplete information. Honestly, how can you trust these guys to meet the commitment they won’t operate with pressure boundary leakage again.

    How can we trust the agency to make sure a plant like Palisades is keying on accurate, up to date and real time information about pressure boundary leaks. When Palisades and the agency is knowingly keying off incomplete information…that they act “super” conservatively.

    Next operating cycle will they immediately shutdown the plant when a CRDM begins to leak like regulation requires. Will it take weeks and months to shutdown the plant over leaking CRDMs like in the two times in the past? In the leaks in 2001 and 2012 Palisades did that. That is a pattern.

  2. Andrew K. February 9, 2014 at 8:41 am

    Thanks for you hard work and incite on this issue. This is not an easy issue or inspection to make and it seems you are very knowledgeable and protecting the public is a responsible way. Thanks!

  3. CaptD February 6, 2014 at 5:39 pm

    Interesting article and I really enjoyed the included links, although I would suggest that the NRC at least BOLD their font to make them more noticeable, if they cannot be shown in another color.
    +
    I’d be interested in learning how many other Regions have seen similar problems in the NPP that they monitor? Perhaps that information would help put this particular problem in context to your Top 10 examples of “Material Problems” that you face.
    +
    I’d also suggest that a future article should be about testing for stress fatigue inside steam generator tubes, since that still is a major issue in the on-going investigations into the replacement steam generators debacle that caused the early decommissioning at San Onofre NPP in California. Because of the different inspection methods available, the NRC should have a procedure that requires inspections by the most through technique, rather than accepting whichever method the Utility chooses, so that they can then say they have “inspected” their tubes, while in effect using a technology that is far from rigorous.

    • Moderator February 27, 2014 at 4:46 pm

      We have changed the “style sheet” for the blog so now all links show up in blue and underlined to make them more obvious. Thanks for the suggestion.

      Moderator

  4. Mike Mulligan February 6, 2014 at 5:27 pm

    So Palisades destroyed their CRDMs through or a result of all the recent startups and shutdown?
    Is the Pressurizer weld flaw today connected the all the CRDM cracks?
    Does the NRC really think those pristine inspected rods without flaw in 2012 really didn’t have flaws…when two of them were discovered with cracks this outage?
    And the industry says the incubation period for developing a crack is over ten years?
    These vulnerable eight CRDMs not replaced this outage…is anyone thinking about the loss of NRC and industry credibility if any of them come up with cracks or leaks within the next operating period…maybe something worst.

    • Moderator February 7, 2014 at 4:21 pm

      In response to your questions:

      Q. So Palisades destroyed their CRDMs through or a result of all the recent startups and shutdown?

      A. There are various factors that have been determined to impact the initiation and propagation of flaws in the CRDM Housings: susceptible material, stresses in the weld, and the environment inside the housing. We will be evaluating if the startups and shutdowns could have additional impact on these flaws during the current inspection.

      Q. Is the Pressurizer weld flaw today connected the all the CRDM cracks?

      A. The pressurizer weld flaw is not related to the flaws identified in the CRDM housings. These components are made of different materials, have different configurations, are exposed to different environments, and the flaws themselves differ as well.

      Q. Does the NRC really think those pristine inspected rods without flaw in 2012 really didn’t have flaws…when two of them were discovered with cracks this outage?
      And the industry says the incubation period for developing a crack is over ten years?

      A. It is possible that the cracks in these housing were present in 2012 but were too small to be identified by the inspection techniques used. However, the NRC and the licensee performed an analysis which concluded that if additional cracks existed at the time but were too small to identify, they would not grow fast enough to cause through- wall leaks prior to the present inspection. However, we are looking at this again during our current inspection.

      Q.These vulnerable eight CRDMs not replaced this outage…is anyone thinking about the loss of NRC and industry credibility if any of them come up with cracks or leaks within the next operating period…maybe something worst.

      A.The licensee is replacing CRDM housings with a newer design that eliminates the problem weld. If the licensee leaves any of the older style CRDM housings in place, they will need to perform additional testing and analysis to support that decision, and we will inspect those activities to ensure they are performed correctly and the conclusions reached are technically sound and well supported.

      Elba Sanchez Santiago

      • Moderator February 10, 2014 at 4:05 pm

        To clarify our previous answer, the NRC did a confirmatory analysis, to verify that under the different potential growth scenarios, the flaws would still not grow fast enough to cause through wall leaks prior to the present inspection.

        Elba Sanchez Santiago

%d bloggers like this: