Robotics examination may shed light on what happened with isolation condensers following Japan quake

It is unlikely what went wrong with isolation condenser units at the nuclear reactor in Japan in 2011 will be known, if ever, until robotics can disassemble the associated valves and systems, attendees heard during the Canadian Boiler and Machinery Underwriters Association’s 40^th Annual Engineering Insurance Conference in downtown Toronto last week.

“We still to this day don’t understand why the isolation condensers were not working the way they were supposed to,” Dr. David Novog, an associate professor and director of the McMaster Institute for Energy Studies at McMaster University in Hamilton, Ontario, noted during a presentation.

“There are still ongoing investigations. Likely, we will not know until robotics can disassemble the isolation condenser valves and systems to determine what went wrong, if ever,” Dr. Novog told attendees. “We know there was either a functional failure – in risk analysis, a functional-based failure means it worked, but not well enough, or it was a complete failure of isolation condensers.”

In March 2011, a massive earthquake and subsequent tsunami resulted in three nuclear units of the reactor, a boiling water reactor, sustaining significant core damage. About 19,000 people are estimated to be deceased (some are still listed as missing), approximately 6,000 people were injured and about 400,000 buildings were significantly damaged – either during the natural catastrophes or demolished afterward, Dr. Novog noted.

“When you are shut down in a boiling water reactor, as long as you keep the fuel covered with water, you are okay. So the primary safety concern going through everyone’s mind post shutdown: keep the fuel covered with water,” he told attendees. “As long as we have external power or standby diesel generators, we have a way to make up the water level in the core.”

Power gets reduced quickly post shutdown in a nuclear reactor. “The heat load is about 6% of the normal heat load; after one day, it’s about 1%; and after five days, it’s down to half a percent. This is still enough to melt the core,” he said.

The power will reach a decay level, which persists for days. “This is radioactive byproduct in the fuel, (which) continues to decay even after we shut down, so there’s still a residual amount of heat,” Dr. Novog explained.

The systems worked as designed after the earthquake and tsunami, he reported. Following the quake, for example, the reactors shut off, went to cooling by the main residual heat removal systems, and isolated the systems. “Post earthquake, this is what’s supposed to happen; this is what did happen. All the units went into safe state, they boxed up, they all went to heat removal on residual heat removal systems,” he said.

When the earthquake prompted the loss of off-site power, the task of generating power needed to perform core cooling fell to the onsite diesel generators, which it did. “Things were cool, levels were maintained. At this point there is no significant safety concern,” Dr. Novog said.

On unit 1, the diesel generators were working and the isolation condenser system was still in place; on unit 2, systems were running and the standby diesel generators were not needed; and it was about the same for unit 3 as unit 2.

However, Dr. Novog noted the first wave hit about 40 minutes after the earthquake struck, and the largest wave about an hour after the quake hit. The latter prompted the loss of diesel generators and the flooding of all available batteries. “Batteries are lost, lighting is lost, all control instrumentation and readings are lost. The station is black,” he said.

Without battery power – meant to provide a few days of power while attempts to replenish water are under way – critical valves that are motor-operated cannot function. This spurred the beginning of emergency actions at the plant.

“The first concern should be to ensure the isolator condenser is working,” Dr. Novog said, noting that “this is the only way to remove decay.”

Plant staff managed to rig a solution to get a water level reading, finding that it was “decreasing in the core, which means isolation condensers are not removing the heat that they should be,” Dr. Novog noted.

The decision was made to close valves. “They valve out this isolation condenser and from this point, it’s in a downward spiral. Unit 1 will melt within hours.”

Not enough water could be pumped into the core to cool, necessitating venting pressure in the whole structure, but getting valves open was an issue.

With venting delayed, the vessel in unit 1 reached a pressure level beyond its design pressure, which, it is believed, caused some seals to leak. Hydrogen was among the gases that accumulated, which then escaped through some seals, “accumulated in the exterior superstructure of the building and eventually detonated,” Dr. Novog reported.

“That detonation caused a massive amount of damage in the plant,” he said, including “all the hosing they had laid to get sea water into this unit, (and) debris damaged the transformers beyond repair that were being used to hook up this diesel generator.”

The detonation “in unit 1 meant the world to this event. It highly contaminated a lot of areas in that tight-packed unit 1, 2, 3 structure,” Dr. Novog explained. “This one event, this one lack of ability to vent the system or to get cooler water in here, along with the failure of the isolation cooling system, really meant the other units are now in trouble,” he said.

The detonations “didn’t explode the reactors, they didn’t damage the reactors in any way. What they did was they hampered our response by creating a debris field, by damaging those generators, by spreading contamination all over the place so that we couldn’t get workers where we wanted,” Dr. Novog reported.

The event has provided valuable lessons with regard to emergency response and preparedness, he suggested. “A lot of time when we’re doing emergency response, we worry about one unit. We don’t often think about multi units, common mode failure affecting everything and all units; common mode failure affecting transportation and worker access,” he told attendees.

“Risk analysis tells us station blackout is the dominant risk profile,” Dr. Novog said. “We spend our time on exotic loss of cooling accidents and analysis – things that take us months and years to complete. But station blackout is the dominant risk factor for all nuclear power plants today,” he said.

“If the isolation condensers in unit 1 had been operable, or if they could have vented earlier on in unit one to avoid overpressure of the containment system, there would have been no explosion, there would have been repowering of unit 2 – or if not repowering of unit 2, at least the site would have been uncontaminated,” he noted.

“They would have been able to get water in to the other units fairly easily. The failures on unit 1 in the response functions on the isolation condensers in the containment led to a spiral that couldn’t be recovered,” Dr. Novog said.

With regard to emergency response in general, he noted, “we should not worry about the initiating event. We should make provisions to say something could happen and have emergency prepa
redness and response plans, training and drills in place.”

One positive was the early call by the Japanese government to start evacuations, he said. “This is an important part of emergency preparedness. You have to be able to pull the trigger and evacuate people,” declaring that a common mode event was taking place.