Vermont Yankee Root Cause Analysis
On June 22, 2010, Vermont Yankee released the results of its root cause analysis of the radioactive material leak that occurred in the Advanced Off-Gas (AOG) building.
The report itself was not made available to the public by Vermont Yankee. Based on a reading of the final draft, the Health Department has prepared a summary.
Health Department Summary
This report describes the course of events, beginning with the discovery of the tritium leak that was reported by Vermont Yankee in January 2010, the search for the source or sources of the leak, the identification of the AOG pipe tunnel leak and the soil contamination that resulted as nuclear reactor water passed from the failed pipes, out the pipe tunnel into the soil, and then into the ground water.
The report also describes the collection of wells on site that have been used since 1988 for testing ground water to show compliance with its Indirect Discharge Permit from the Vermont Department of Environmental Conservation. A total of 34 wells existed before January 2010. Many of them were used to verify that radioactivity and other contaminants did not pass from two septage spreading fields, one at the northern end of the site, and one at the southern end of the site. Of these 34 wells, three of them (GZ-1, GZ-3 and GZ-5) were specifically installed as part of the Nuclear Energy Institute’s Groundwater Protection Initiative.
Major Points of Interest
The following summarizes major points of interest from the report that relate to public health and environmental protection.
- Root Causes
- Contributing Cause
- Organizational/Programmatic Causes
- Extent of the Problem and Conditions
- Operating Experience
- Safety Significance
- Corrective Action Plan
- Event Timeline
- Root Cause and Failure Mode Analysis
- How will Entergy Vermont Yankee take action to terminate, remediate and monitor the situation for the future?
- Safety Culture Comparison
- Inadequate construction and housekeeping practices employed when the AOG Building was constructed in the late 1960s and early 1970s, and when the AOG drain line was added in 1978
- Ineffective monitoring and inspection of vulnerable structures, systems and components that eventually leaked radioactive materials into the environment
The inadequate construction practices were revealed in the lack of a watertight seal where the concrete encasement for the AOG drain line and AOG pipe tunnel were joined, and where the drain line joins the AOG drain pit. It was at these points that most of the contaminated water leaked from the pipe tunnel. The construction debris and silt that had been allowed to accumulate in the pipe tunnel was cited as inadequate housekeeping. This condition eventually clogged the AOG pipe tunnel floor drain, causing reactor process water to fill the floor of the tunnel until the contaminated water leaked from the tunnel from the two points along the AOG drain line concrete encasement.
Ineffective monitoring and inspection was evident in the fact that, of the 11 Groundwater Protection Initiative criteria, most were only partly met. Among those criteria only partly met, the fact that only three groundwater monitoring wells had been installed by the end of 2009 was especially noteworthy. In addition, the underground buried pipes program was found to be insufficient in that the pipes that leaked were not part of the required periodic monitoring.
Corrosion found in two pipes in the AOG pipe tunnel was considered a contributing cause of the leak. The report stated that pipes should not fail. If pipes do fail, the contents should be contained and kept from the environment, and any leaks that occur should be identified promptly.
- Inadequate definition of groundwater protection responsibilities between Entergy and Vermont Yankee
- Inadequate management commitment to fully implement the Nuclear Energy Institute’s Groundwater Protection Initiative
According to the report, Vermont Yankee’s Chemistry Department did not champion the groundwater protection initiative. Plant documents in 2008 described that the site hydrogeological conceptual model and risk assessments needed updating. The report also stated that a groundwater environmental monitoring plan had yet to be written at the time the AOG leak was first detected. Although additional groundwater monitoring wells were discussed at weekly management meetings beginning as early as May 2009, the first of these new wells were not drilled until January 2010.
- underground pipes
- below grade pipes in trenches or in guard pipe or pipes retired from service
- process tanks and storage pools
- sumps, drain, basements and tunnels
- from previous spills
Sites with floor drains were identified as requiring careful consideration. Several floor drains within the main power plant buildings had recently been found to be clogged. Despite this, modern design, construction and maintenance practices especially stringent exclusion of foreign materials around open systems, were cited as limiting the extent of the specific root causes leading to a similar leak from structures, systems and components.
The report described a search of three years’ worth of Vermont Yankee operating experience reports for experiences related to the AOG leak. Of 14 reports, three involved tritium leaks into the environment:
- 2001: tritium from air conditioning condensate found in Manhole 13 in the storm drain system
- 2008: tritium found in House Heating Boiler #2
- 2008: tritium found in a gap in the turbine building wall
The report also described leaks at other nuclear power plants, including a series of leaks at the Braidwood nuclear power station in Illinois, leaks at Indian Point 2 in New York, the Perry plant in Ohio, the Brunswick Steam Electric Plant in North Carolina, at D.C. Cook in Michigan, and at Millstone Unit 2 in Connecticut. Various aspects of each of these leaks are described, although their relevance to cause and corrective action related to the Vermont Yankee AOG leak was not seen as providing any particular lessons learned. Of the more than two dozen plants that have had tritium leaks, only these six were mentioned in the report.
According to the report, there was “no nuclear, radiological or personnel safety significance.” As evidence of this, it was pointed out that the AOG system is not safety-related and therefore the protection of the reactor and fuel was not jeopardized. The calculated dose from the methods of Vermont Yankee’s Off-site Dose Calculation Manual was used to demonstrate the lack of radiological safety significance. This dose – 0.00095 millirem per year – was compared to the Nuclear Regulatory Commission annual dose limit of 100 milirem per year and the Environmental Protection Agency annual limit for the maximally exposed individual of 25 millirem per year, as evidence that there was no radiological safety significance.
- By July 1, 2010 - Revise the Flow Accelerated Corrosion Program to include an “analysis of susceptibility and consequence” that may have led engineers to include the pipes in the AOG pipe tunnel in the Buried Pipes and Tank Inspection and Monitoring Program.
- By July 2010 - Eliminate the means by which excess condensate created a pool of highly tritiated water in the radioactive waste trench.
- By July 2010 - Incorporate all appropriate plant components into the Buried Pipes and Tank Inspection and Monitoring Program. This is to be followed by an inspection plan for each by October 1, 2010.
- By August 1, 2010 - Schedule inspections of the two steam traps that had failed drain lines. By September 1, 2010, complete inventory of steam traps in other vulnerable systems. By October 1, 2010, incorporate industry benchmarks for steam trap preventive maintenance.
- By September 1, 2010 - Incorporate floor drain preventive maintenance in each operating cycle, and complete hard to detect analysis on the soils taken from the AOG excavation.
- By September 1, 2010 - Complete remaining elements to meet the 11 criteria of the Nuclear Energy Institute’s Groundwater Protection Initiative.
- By October 1, 2010 - Install a run timer on the AOG drain pit sump to signal potential problems, and conduct an AOG pipe tunnel drain flow test.
- By December 1, 2010 - Redesign watertight joint of the AOG drain line encasement.
- By February 1, 2011 - Develop means to monitor and inventory water.
An event timeline is presented in the report, starting with the first positive groundwater monitoring well sample from GZ-3 obtained on November 17, 2009 through Vermont Yankee’s March 24, 2010 announcement that the original leak had been found and stopped. Each of the events in the timeline has been described in Health Department updates. One previously unpublished collection of events is the date that each groundwater monitoring well went into service:
November 2007: GZ - 1, 3 and 5
January 27, 2010: GZ – 6
January 31, 2010: GZ – 4 and 14
February 3, 2010: GZ – 3
February 4, 2010: GZ – 7 and 9
February 5, 2010: GZ – 11
February 6, 2010: GZ – 10, 12 and 13
February 13, 2010: GZ – 15
March 1, 2010: GZ – 14D
March 3, 2010: GZ – 19
March 4, 2010: GZ - 17
March 11, 2010: GZ - 16, 20, 13D and 19D
March 21, 2010: GZ - 21
Root Cause and Failure Mode Analysis
A widely used root cause analysis process in the nuclear industry is known as Kepner-Tregoe. The root cause analysis report lists four questions that are its foundation:
- How does groundwater move on site?
- Where is tritium in the groundwater?
- What is/are likely sources of tritium that could contaminate the groundwater?
- What actions will be taken by the plant to terminate, remediate and monitor the situation for the future?
A particular application is the Failure Mode Analysis. According to the timeline of events, this application was initiated on January 13, 2010 to find the source of tritium discovered in November 2009. This report states that one of the first steps of this process in 2010 was again looking at a 2005 Entergy Vermont Yankee review of “all plant systems to proactively determine which characteristics might result in a release to the environment.” That study led engineers to focus on inspecting, testing and considering the following:
- The condensate storage tank
- The AOG drain pit
- AOG ground surface subsidences, sometimes referred to as sink-holes
- The AOG bypass valve vaults
- The condensate storage tank pipe tunnel
- Conduit access manholes and handholes
- The radioactive waste pipe tunnel
- The AOG pipe tunnel
- The stack sump
- The effects of transporting spent fuel from the reactor building to the independent spent fuel storage installation (ISFSI)
- Whether the plant manages its configuration adequately
The condensate storage tank (CST) is discussed extensively in the report. In 1976 there was a review of an overfilling of the tank, and in 1986 a leak was discovered and repaired. There was also an assessment of two episodes where water accumulated in the CST tell-tale in 2007 and 2008. In both cases on the basis of tritium concentrations, the water was determined to come from precipitation washing down the sides of the tank, not from water leaking out of the tank. The water found in the tell-tale was about one-sixth the concentration of CST water, which is about 3.9 million picocuries per liter (pCi/L). Two inspections of the CST were also described, in 1993 and 2008. Another inspection is planned during the refueling outage scheduled for 2011. While never ruled out as a possible source of tritium groundwater contamination, there appeared to be no substantial evidence that tritium was leaking from the CST in the first four months of 2010.
Most of the other potential sources in the failure mode analysis provided no evidence of tritium leaks to the environment. Exceptions were the radioactive waste pipe trench and the AOG subsidences. The radioactive waste pipe trench became a very serious candidate source of leaks early in mid-January, when it was opened and inspected and found to have accumulating water with a tritium concentration of about 2 million pCi/L. Evidence, especially the negative samples at well GZ-6 when that well was commissioned on January 27, proved this was not a viable source of groundwater tritium contamination. With regard to the AOG ground subsidences, those were taken as possible evidence that there may have been some source near the AOG pipe tunnel. These subsidences were noted in the report to have been identified as far back as July 2008.
- A long-term monitoring plan
- Repairing physical components
- Developing trigger points for each well that would initiate remedial actions
- Monitoring for potential future leaks on site
The failure mode analysis ends with structures, systems and components grouped into four failure modes. Mode 1 is a collection of 32 of the more vulnerable buried pipes runs. Mode 2 contains 15 of the plant’s process tanks and storage pools. Mode 3 captures the building sumps, drains, basements and tunnels. Mode 4 is where a host of five significant site spills reside.
Safety Culture Comparison
The report concludes with an assessment of the safety culture at Vermont Yankee. Thirteen separate elements of an effective safety culture at a nuclear power station were analyzed, and by all elements were judged not to have contributed to the AOG leak.