Possible Cause of Reactor Building Explosions

March 18, 2011
Dave Lochbaum
Former contributor

Dramatic videos show the explosions that severely damaged the reactor buildings at first Unit 1 and then Unit 3 at the stricken Fukushima Dai-Ichi nuclear plant in Japan. The explosions are attibuted to the ignition of hydrogen gas that collected within the reactor buildings. This was early in the crisis, and before the spent fuel pools are thought to have lost water and started producing hydrogen.

The hydrogen was likely produced by damaged fuel rods in the reactor core. To reduce pressure in the reactor vessel, some of that hydrogen was released from the vessel into the primary containment structure of the reactor.

A key, unsolved riddle is how a significant amount of hydrogen escaped from the primary containment into the reactor building, and how this low-probability event would have happened in mulitple reactors.

How Hydrogen Got into Primary Containment

Figure 1 shows a cross-sectional view of a boiling water reactor with a Mark I containment like that at Fukushima Dai-Ichi. The reactor core is housed within a metal reactor vessel. The reactor vessel is enclosed within the primary containment structure. The reactor building completely surrounds the containment structure. The reactor building walls are made of 18 to 30 inch-thick concrete up to the elevation of the refueling platform. The walls are made of metal from that elevation to the roof.

Figure 1

The hydrogen gas most likely came from a chemical reaction between water and the metal cladding of fuel rods in the reactor cores when the water level inside the reactor vessels dropped low enough to expose at least the upper core regions.The hydrogen gas initially collected in the reactor vessel.

To cool the fuel in the reactor, workers attempted to pump seawater into the reactor vessel. As pressure inside the reactor vessel increased, it kept water from flowing into the reactor. Periodically, workers opened valves to vent steam and gas from the reactor vessel to into the pressure suppression chamber (also called the torus). The gas, including hydrogen, collected in the torus and periodically equalized with the air space in the drywell.

When pressure in the primary containment (the combination of the drywell and the torus) rose too high, workers vented the containment to the atmosphere. This vent piping passed through the reactor building, but discharged well outside of it, and should not have led to a hydrogen buildup inside the building.

How Hydrogen May Have Gotten from Primary Containment into the Reactor Building

The destruction of the Unit 1 and 3 reactor buildings appears to have been caused by hydrogen explosions. As noted above, an unanswered question is how the hydrogen got into the reactor buildings. A little-known test performed decades ago at the Brunswick nuclear plant in North Carolina may hold the key to answering that question.

To satisfy a requirement in the American Society of Mechanical Engineers (ASME) code for prototype containment designs, workers performed a structual integirty test on the reactor at Brunswick in the 1970s.

The primary containment structure at Brunswick was designed to withstand an internal pressure of  62 pounds per square inch (psi). The ASME code required it to be tested at 71 psi. This test involved pumping air into the containment structure until the pressure rose to 71 psi. The pumps would then be turned off and the pressure would be monitored for several hours to verify that it remained fairly constant, indicating that the primary containment was intact and not leaking. During this time, workers would record data from strain gauges and other instrumentation to verify that structural loads were properly distributed.

But as workers increased the containment pressure they encountered a problem. The pressure stopped increasing and remained constant at 70 psi. The pumps continued to push air into the containment, but its pressure just stopped increasing. This unexpected plateau started a hunt for air leaking from the containment somewhere.

A hissing sound attracted workers to the top of the containment structure. They identified air leaking through the drywell flange area (see Figure 1). The metal drywell head (see Figure 2) is bolted to the metal drywell with a rubber O-ring between the surfaces to provide a good seal fit.

                                                        Figure 2

Workers found that the containment pressure of 70 psi pushing upward against the inner dome of the drywell head lifted it off the drywell flange enough to provide a pathway for air to leak from the containment. That air leaked into the area labeled refueling cavity in Figure 1. The refueling cavity is located outside the primary containment but inside the reactor building.

At Brunswick, workers tightened the drywell head bolts beyond the amount specified in the reactor plans in order to reduce the leak rate and continue the test. While workers conducted pressure tests at all nuclear reactors prior to initial startup and periodically thereafter, these tests were performed at or below the containment design-pressure of 62 psi. So none of them reached the pressure that caused the leak around the drywell head.

In other words, had Brunswick not featured a prototype containment design, its initial and recurring pressure tests would have been conducted at 62 psi, not 71 psi. Leaking from the drywell head was not observed until the containment pressure rose to 70 psi.

How does this Brunswick containment testing experience relate to the reactor building explosions experienced at Fukushima Dai-Ichi Units 1 and 3?

Like Brunswick, the containment design at those reactors features a drywell head bolted onto the lower portion of the drywell. Workers at these reactors faced siginficant problems cooling the reactor cores. The combined effects of the earthquake and tsunami left the reactors without ac electrical power. The only dc-powered (i.e., battery-powered) backup system was lost when the batteries were exhausted. Workers turned to their only remaining option: injecting sea water into the reactor vessels to cool the reactor cores.

The pumps used to pump seawater into the vessel operated at low pressure. When seawater entered the reactor vessel, it was heated by the hot reactor core to the point of boiling. Steam produced by the boiling increased the pressure inside the reactor vessel. To prevent this rising pressure from hindering seawater from being pumped into reactor, workers periodically vented the reactor vessel. This carried steam and gas, including hydrogen, into the primary containment. This flow in turn increased the pressure inside containment. When containment pressure rose too high, workers vented the containment to the atmosphere.

The workers properly sought to minimize the amount of gas they vented from containment to the atmosphere to lessen the amount of radiation released. They did this by allowing the containment pressure to rise as high as tolerable between ventings.

It is possible that the containment pressures rose high enough to replicate the Brunswick experience by lifting the drywell head enough to allow hydrogen and other gases to leak into the refueling cavity and reactor building. If so, hydrogen could build up to an explosive mixture.

This tragedy will be closely examined for its causes. That scrutiny must determine how hydrogen got into the reactor building early in the crisis. The drywell head pathway may be that answer.

Answering this question is critical to prevent hydrogen explosions at the other reactors at Fukushima.

If this mechanism is the cause of the leak, it could be averted easily and effectively simply by changing the venting procedures so that workers vent the containment pressure to the atmosphere more frequently and do not let it build up to such high level. Taking such action might moderately increase the amount of radioactive gases vented into the atmosphere, but could eliminate a source of hydrogen inside the reactor buildings that could cause another explosion.

Authorities should launch an  investigation to pinpoint the source of the hydrogen leak to eliminate this risk in the future. But in the meantime, since the Brunswick test showed that this containment is vulnerable to high-pressure leaking, Tokyo Electric Power Co. can and should take immediate steps to avoid creating such a leak by changing its procedures to vent the containment before it builds up to such high pressure (70 psi).