“Hanford Determines Double-Shell Tank Leaked Waste From Inner Tank.”
This headline on a Department of Energy press release from last October was bigger news than it appears on first glance. For the first time, the type of massive storage tank built to hold some of the most radioactive waste in the world was found to be slowly leaking.
The press release -- dated October 22, 2012 -- stressed that no waste had escaped the outer wall of AY-102, a million-gallon-plus tank built in the late 1960s to store waste collected from decades of plutonium production at Hanford. The Energy Department assured the public that no other double-shell tank at Hanford – there are 28 in all -- was leaking.
Tom Fletcher, a top official in charge of the Hanford tank farms, was quoted in the release saying the Energy Department was “working collaboratively with Washington State Department of Ecology to determine the most effective path forward for AY-102 and ensure that we are effectively protecting the public and the environment.”
The press release said the leak was found “during a regularly scheduled” inspection of AY-102 in August.
But the KING 5 Investigators found that the first evidence that AY-102 was leaking came nearly a year earlier, in October 2011.
Why it took DOE and the private company hired to oversee the tank a full year to conduct a thorough investigation into signs of a leak of highly radioactive waste raises a larger question about the clean-up effort at Hanford – a 586 square mile patch of Central Washington that’s the most contaminated place in the Western Hemisphere.
That question: How effective is the oversight of the private contractors doing much of the complicated work at Hanford?
When Mike Geffre arrived at his job at Hanford on Oct. 10, 2011, he was told his scheduled duties would have to wait. A specialist with years of experience working with the equipment used to monitor Hanford’s most dangerous waste tanks, Geffre was dispatched to check out a leak detection alarm on AY-102.
One of the instruments monitoring the tank’s annulus – the two-foot space separating the inner and outer shells of AY-102 – had triggered an alarm the night before. Since AY-102 is a double-shell tank, the assumption was that the instrument had failed: None of the 28 double-shell tanks had ever leaked at Hanford. They were thought to be the most robust tanks, capable of keeping the worst waste safely contained until the project for disposing it for good is completed.
Geffre expected to confirm the false alarm and make the necessary fix.
“We never guessed in a million years that we’d have a double-shell tank leak,” said Geffre.
But when he arrived at the tank accompanied by a health physics technician – an expert in monitoring radiation levels – Geffre couldn’t find anything wrong with the device, called an ENRAF. Computer diagnostics tests showed it was working properly … and it was showing one-half-inch of liquid in the tank annulus.
The next step was to raise a part of the ENRAF that was suspended from a shaft that technicians use to lower instruments into the tank space from the ground above. A plummet attached to the ENRAF hangs through the shaft to just above the annulus floor. The physics of displacement allows the ENRAF to detect liquids that come into contact with the plummet.
If no liquid was present in the annulus, Geffre and his colleague would see a clean plummet after they raised it, confirming a false alarm.
That didn’t happen. Instead, Geffre and his co-worker saw something troubling through the leaded glass window -- liquid and white crystalline debris on the plummet, which Geffre said appeared to be nuclear waste (Geffre took a photo two days later when he was working again at AY-102). The health physics technician then held up her Geiger counter and recorded radioactive contamination.
“We knew that we had a real problem,” said Geffre. “I knew it was serious because that’s what my career has been for the last 25 years -- to monitor these tanks, to check for leaks, to check for temperatures on these tanks, that’s what I’ve done. I’m very thorough, very meticulous about this type of equipment.”
The technician’s Geiger counter recorded a dose-rate of 4 millirem (mrem) of radiation through the glass. In most environments that would be a small amount of radioactivity. According to the U.S. Department of Energy, the average person receives approximately 360 mrem a year from natural radioactive sources and through man-made sources like x-rays, smoke detectors and older TV sets.
But to experts who monitor Hanford’s waste tanks day in and day out, the reading was a shock. According to Geffre and other Hanford sources, they’d never seen a reading of radioactivity coming from equipment placed in the annulus space of a double-shell tank.
“To receive 4 millirems was far exceeding what we thought we’d ever see. It was very shocking for us,” Geffre said, adding that the experience was “surreal.”
“Hundreds of times we’ve calibrated this equipment, hundreds of times we’ve done radiation level checks in the annulus [and] nothing is there, nothing,” Geffre said. “I can’t imagine how many times they’ve checked, and then all of a sudden I’m the person there.”
“Four millirems is high. You don’t expect that in an annulus. That’s a lot for a clean space,” said another Hanford worker, who requested anonymity before speaking with KING 5.
The annulus area of Hanford’s double-shell tanks is expected to be radiation free. The fact that the plummet was emitting radiation was the first major indication that waste from inside AY-102 had leaked.
Of particular concern was that AY-102 is unlike any other underground storage tank at Hanford. It’s filled with 850,000 gallons of thermally hot waste considered the most hazardous material at Hanford and, in fact, in the entire country.
“This tank is holding the nastiest of all the nasty stuff at Hanford,” said Geffre. “The chemical radioactivity in this tank is so high that it actually boils by itself. If you could pick one tank you didn’t want to leak, this would probably be the one.”
Geffre has 26 years of experience working at Hanford, currently as an employee for Washington River Protection Solutions, a private company that has a contract to manage all the tank farms at the site.
After he spotted what appeared to be waste on the AY-102 ENRAF plummet, Geffre reported his findings to two superiors at the tank farm. According to Geffre, the two men insisted that the instrument must have malfunctioned.
"They just kind of sat there and said, ‘Are you sure that thing's calibrated?’ I said, ‘Yea, I'm sure. I'm the one who does the calibrations on it. I'm the one who takes care of it. Of course I'm sure,’" Geffre said.
WRPS followed protocol and reported the AY-102 ENRAF alarm to both the federal government and the Washington Department of Ecology, the state agency that shares regulatory authority over Hanford with the U.S. Environmental Protection Agency.
WRPS told Ecology that rainwater most likely seeped into the AY-102 annulus space, not tank waste. The company also said the radioactive readings were probably the result of “legacy contamination” deposited years ago.
KING 5 requested on camera interviews with the two managers but a company spokesperson declined that request on their behalf.
“As for the October 2011 events, experience gained over decades of tank farm operations led us to believe that a small amount of rainwater, not waste, was collecting in the AY-102 annulus,” according to a written statement provided by WRPS to KING 5. “This was based on recent heavy rainfall, the discovery of water intrusion pathways, known low levels of radioactive cross-contamination between the primary tank and the annulus, and readings from the leak detection system.”
Geffre said he argued with his bosses that their assumptions were not based on scientific evidence.
“I knew it wasn’t rainwater because of the color of the liquid on the plummet and the dose rates were no indication this was rainwater,” said Geffre. “There was no doubt in my mind that we had a leak, that something was wrong down there.”
Geffre wasn’t the only WRPS employee to urge a detailed inspection. Gary Tardiff, a WRPS engineer at the tank farm, verbally recommended a video inspection of the AY-102 annulus on Oct. 12 or 13. It didn’t happen, even after Tardiff filed a formal request with his company to do so.
“At this point in time it would have been beneficial to insert a camera into the annulus to see if the water intrusion could be pinpointed. When Engineering requested a camera to be installed, the AZ (tank farm) Team indicated it would take several days just to plan the work package and would most likely not be in time to determine the source of the water intrusion,” wrote Tardiff on Oct 19, 2011.
Officials from WRPS said company managers decided to wait for the next regularly scheduled inspection of the tank in August 2012, ten months after the first alarm sounded.
“There was no threat to the environment, the public or our workers. Because of the small amount of any liquid indicated by the ENRAF, its location within the sealed annulus, and the double-filtered ventilation system on the annulus, waiting until a scheduled video inspection in August posed no appreciable risk,” WRPS said in its response to KING 5.
More red flags
Company managers stuck to their theory of rainwater and legacy contamination as the cause of the AY-102 alarm, despite more red flags, including the following:
* October 26, 2011 - An alarm on an annulus air monitor went off. The radioactive reading of air particles spiked to the highest levels ever seen in this space by current employees.
* March 10, 2012 – The AY-102 ENRAF plummet became glued to the bottom of the annulus, suggesting it might be stuck in dried waste.
* May 24, 2012 - The plummet’s wire broke, after which a health physics technician got a reading of 20,000 disintegrations per minute (dpm) on the wire and reel (dpm is the measure of the intensity of the source of radioactivity).
* June 4, 2012 – A small camera lowered into the annulus to observe the stuck plummet recorded video of what looked like radioactive waste to many of the employees who viewed it.
“(On that video) we could tell it wasn’t rainwater. It had a shine to it and you could see a pattern to it so it didn’t look like a (clean) metal bottom,” said Geffre. “But they stuck to their story, which was strange because they stuck to a story they could not prove.”
“It looked like tank waste to me,” said another source who asked to remain anonymous. “All the managers did not want that tank to have a leak.”
“This is evidence that the company was bending over backward to not find the bad news,” said Tom Carpenter, executive director of the Hanford Challenge, a watchdog group based in Seattle. “This is the mindset at the Hanford site -- of denial.”
Of all the pieces of evidence of a leak in the 10 months before a full-scale video inspection was initiated, one stands out above all others, according to nationally recognized experts in the fields of radiation and nuclear physics who examined the evidence for KING 5. They said the contamination reading off a tiny piece of wire on May 24, 2012, should have stopped all the managers in their tracks.
“If your equipment’s screaming at 20,000 dpm, I really don’t know what it’s going to take. What’s it going to take to wake up and start dealing with the problem?” said Marco Kaltofen, a civil engineer and veteran environmental scientist who’s conducting a wide-scale radiation research project at the Worcester Polytechnic Institute in Massachusetts.
Kaltofen is currently collecting radioactively contaminated items from around the world to study their affects on human health.
“Of all the stuff we’ve collected from Fukushima, Hanford, Los Alamos, Oak Ridge, Savannah River [nuclear reservation], even the Soviet Union, I don’t think any of the stuff we have would read 20,000 dpm. So that screams, ‘Do something!’ It literally screams,” Kaltofen said. “It’s hard to believe you could get this much advance warning of a problem and not deal with it.”
Media relations representatives from WRPS said their experts disagree. They told KING 5 that a reading would have to be in the millions of dpm to cause alarm.
“Contamination readings found on the drum and reel were well below what would have been expected from tank waste,” the company said in its statement.
Mixed messages in follow up report
It was not until after the August 2012 scheduled video inspection that WRPS officials notified Washington’s Department of Ecology that evidence of a leak had been discovered. It took another two months for the leak to be confirmed to the public.
After the video investigation, WRPS assembled a team of company experts and consultants to compile a Leak Assessment Report for the Department of Energy. In the 445-page report, released on November 7, the authors site several examples of why the company’s original assumption -- that rainwater and legacy contamination were the source of the liquid in the AY-102 annulus, not nuclear waste – was defensible.
The panel said there were signs that rainwater had entered the tank annulus in the past. In a 2006 examination of AY-102, rust streaks and mineralization were observed in the annulus.
“It is likely that soil channels created by…water supply leaks or the annulus ductwork may provide a path for rainfall and snowmelt seepage onto the concrete dome of the tank,” wrote the authors.
The Leak Assessment Report also discussed the potential of legacy contamination.
“The primary tank headspace and the annulus ventilation system are cross-connected via floor drains in the annulus pump pit and the leak detection pump pit. The cross-connections provide a pathway for contamination to enter the annulus,” wrote the authors.
But a close review of the report shows no hard scientific data exists to support these possibilities. In their own report, the authors wrote that there’s never been evidence to show rainwater has collected in any measureable way on the AY-102 annulus floor.
“Several studies have examined the possibility that a seasonal moisture accumulation and evaporation cycle occurs within the annulus, but the possibility of a net water accumulation cycle has not been demonstrated,” wrote the authors.
They also wrote, “Water intrusion from rainfall was the suspected cause [of the ENRAF alarm]; however this explanation seems problematic because of the plummet radiation readings where none would be predicted.”
The experts also said events in 1976, 1985 and 1986 resulted in low-level contamination in the annulus pump pit and probably the annulus itself, although that has never been scientifically proven. They also noted no contamination has been found in the annulus since an extensive sampling in 1999.
“In 1999, contamination surveys of the annulus determined the surfaces were contamination-free,” wrote the panel. “Generally, equipment removed from the annulus risers has been contamination free.”
“Nothing added up that said this wasn’t a problem, and they just chose not to do anything,” said Geffre. “That has been my job for 25 years, to keep track of every drop, every ounce, every piece of radiation that is out there. And all of the sudden when we’re getting to the point where we’re not keeping track of it, something’s wrong.”
'Making the problem go away'
On October 27, 2011, 18 days after the first leak detector alarm sounded, Geffre’s managers instructed him to take on another job – to conduct what’s called a zero-reset on the ENRAF that sent up the signal that a potential leak was underway. Technicians perform a zero-reset when the equipment has been proven to have drifted out of position and is not accurately detecting any accumulation of liquid.
But that was not the case this time. According to the company’s own Leak Assessment Report and documents obtained by KING, the ENRAF was doing its job correctly.
“On October 12, 2011, it was determined (the) ENRAF was working properly,” wrote WRPS experts in the follow up report.
Despite that, Geffre’s managers directed him to program the ENRAF into a less sensitive state, via the zero-reset. This meant the leak detector would no longer sense what was already sitting at the bottom of the annulus. In essence, the equipment was starting from scratch, as if whatever had leaked into the space already didn’t exist. In the WRPS Work Order for this action obtained by KING, Geffre wrote in the comment section: “Reset zero per engineering direction.”
“I did it under protest. I tried to tell the managers, to educate them; I even drew diagrams on the white board,” said Geffre. “This was simply a way of making the problem go away, to make sure the ENRAF didn’t go back into alarm unless a lot more liquid piled up.”
KING 5 asked WRPS representatives why the instrument technician was ordered to perform this task. The company said the worker had to re-calibrate the equipment to put it back into service. When pressed further by KING, spokesman Jerry Holloway said, “We’ll have to agree to disagree on this one.”
“That doesn’t make sense.” said Geffre. “The calibration is a two part step. I had already verified that the ENRAF was in calibration. Doing a zero reset just tells the equipment at what level you want it to start reading a liquid. I told the managers at the time, that by doing this we could miss something important. They didn’t want to hear the details. It was a way to stop that alarm from going off again soon-a way to make the problem go away.”
A deadly and costly mess
Overseen by three government agencies, operated by private companies, guarded from the public by fences and its own security force, Hanford is a $2 billion-a-year cleanup operation funded by U.S. taxpayers. The project won’t be complete until 2040, at the earliest, and some cost projects show the cleanup costing $114 billion by the end of the century.
Contamination at the Hanford Site is the result of the U.S. nuclear bomb program – begun during World War II and active until the end of the Cold War. Plutonium production over the course of five decades produced millions of gallons of radioactive waste, the most dangerous of which is contained in the double-shell tanks and more than a hundred single-shell tanks. None of the tanks was expected to contain waste permanently; even double-shell tanks like AY-102 had an estimated lifespan of 25 to 50 years [AY-102 was put into operation in 1971] and were expected to hold the bulk of Hanford’s waste until it could be processed by the Waste Treatment Plant, expected to begin operations in 2019.
"The fact that a double-shell tank is leaking at the Hanford site is a game changer,” said Hanford Challenge’s Tom Carpenter.
The leak from AY-102, if it were to corrode the tank’s outer shell, could ultimately contaminate groundwater and make its way to the nearby Columbia River.
"If the Columbia River gets contaminated it eventually finds its way up here [to Western Washington]. If there's an explosion or fire, it can still come up here. So we're not safe from it, but we need to keep our eye on what we're being told out there,” said Carpenter.
But the AY-102 leak is problematic for another reason – the multi-billion-dollar plan to permanently dispose of Hanford’s radioactive waste depended on using this specific tank as a “feeder” for the Waste Treatment Plant. A leaking tank can’t be used because the process of channeling waste into it from other tanks and agitating it to mix uniformly for transfer to the treatment plant could make any leak larger.
Government agencies support contractor’s actions
KING 5 asked the government officials responsible for the operation and oversight of Hanford about the months-long delay in investigating the AY-102 leak.
Both the Energy Department and Washington Ecology stand by the contractor’s initial actions.
“All responses were factual and displayed that WRPS appropriately generated the needed field planning documents quickly to trouble-shoot, recalibrate, and put back into service the annulus ENRAF,” wrote Dieter Bohrmann, an Ecology spokesman. “They had radiological monitoring in place as they brought the ENRAF up to visually inspect it. And should waste material have been present at that time, the radiological readings would have been much higher, as indicated in their response.”
The Washington State Department of Ecology told KING 5 that it received no additional information until August 2012 about potential warnings of a leak from WRPS beyond the first alarm in October 2011.
On October 13, 2011, WRPS notified the Department of Energy – the agency that granted the tank waste contract to WRPS -- that an alarm had sounded. The contractor reported “Rainwater intrusion is thought to be the cause of the wetness. A video inspection will be taken…”
The video inspection didn’t happen until the regularly scheduled review took place 10 months later. There’s no mention in this notification to the federal government that the equipment from this space gave off an unexpected reading of radioactivity.
“WRPS notified ORP [Dept. of Energy’s Office of River Protection] in August 2012 that possible tank material was seen during a visual inspection of the annulus of AY-102. The contractor completed a comprehensive leak assessment of AY-102,” wrote Lori Gamache, ORP media representative.
The U.S. Department of Energy said its Hanford operation “takes its oversight role seriously.”
“We are now assessing the adequacy of the double shell tank monitoring program for the underground storage tanks and applying lessons learned from this event to improve processes and procedures,” said Tom Fletcher, assistant manager for the Tank Farms Project.
How much has leaked?
The Leak Assessment Team estimated in October 2012 that between 190 and 520 gallons of waste had leaked into the annulus of AY-102. Last week WRPS media representatives Jerry Holloway and John Britton repeated those exact numbers to KING when asked how much had leaked so far. Holloway and Britton said that the tank was still leaking but that there has been “no appreciable change” since last year. In addition they wrote “a significant portion of the liquid has evaporated, leaving about 20 – 50 gallons of drying waste.”
But KING 5 has obtained documents written this month by Washington Department of Ecology’s Michelle Hendrickson stating that the amount of waste has increased by a measurable amount.
“The area within the annulus that waste occupies has increased approximately 25 percent since Ecology’s visit (to the tank farm) on December 27, 2012. This area increased vertically and horizontally where it extends outward in ‘finger channels’ and near the smaller piece of what appears to be orange insulation foam,” wrote Hendrickson in an April 3 letter. Hendrickson is Ecology’s Tank System Operations and Closure Engineer, Nuclear Waste Program.
“The edges of the waste material are very wet and are changing color. The waste material that was white (indicated dry waste) is now yellow or light green.”
Hendrickson sent the letter and accompanying photographs taken by Ecology employees to the Department of Energy. Four WRPS managers were copied on it.
“Not only do I think they lied flat out, I think whoever authorized that WRPS statement to the news media- maybe someone from the Department of Energy or the company- should be held accountable and should no longer be working there,” said state Rep. Gerry Pollet, D-Seattle.
Pollet is the legislature’s top expert on Hanford, a public interest attorney and the executive director of the Hanford citizen watchdog group, Heart of America Northwest. “It turns out there has been appreciable change, which should add urgency to the leaking tank and a plan to do something about it. The public needs to know the truth to adequately weigh in on the policy issue here,” said Pollett.
In her April 3 letter, Hendrickson urged the federal government to take action.
“Ecology strongly encourages the United State Department of Energy to remove the contents [of AY-102] as soon as it is practicable,” said Hendrickson.
Experts interviewed by KING 5 said making assumptions at Hanford instead of conducting thorough investigations doesn’t cut it at a place with so much at stake.
“They’ve wasted money, they’ve wasted time. We can’t afford to do those things out at Hanford. It’s too urgent of a problem. All of us have too much invested in that clean up to succeed for that kind of mentality to prevail,” said Carpenter.
"Every day that you wait (to investigate a leak), it's more material that you have to dispose of, more stuff gets contaminated and the problem gets worse and more expensive and harder to control,” said Kaltofen. “Time is not our friend. This waste actually gets worse as time goes on. It's hot, it's corrosive, it eats through its container. It can't sit there, or it's going into the Columbia River. There's no other choice."
"If your alarm clock goes off, hey, it's time to go to work. If your alarm in the tank goes off, apparently it's time to wait another year,” Kaltofen said.