TEPCO’s Akio Komori: “The Options We Had Available … Were Rather Limited”
Where were you and what were you doing on March 11, , when the earthquake and tsunami hit?
On March 11 I was in a meeting at TEPCO headquarters. It was a large earthquake, and the building shook.
The nuclear reactors automatically of course shut down, but first, how would you say, the system, from the perspective of supplying power, had undergone a major disruption. So we created an emergency general disaster response room on the second floor at TEPCO headquarters. So that is where I first went.
Then, for the whole company, from the perspective of supplying electricity and assessing damage to the power plants, I am in charge of the nuclear power plants, and so I first heard that the nuclear power plants had all shut down due to the earthquake without problem at around 2 p.m. on the 11th. I think it was around 2:40-something.
News about the tsunami arrived around five to 10 minutes after that, and as a result, since the tsunami was expected to cause additional destruction, I focused next on contingencies for the nuclear power plants.
After the tsunami hit and you heard that all power had been lost, what were you thinking?
First, warnings were being issued that a tsunami was on its way, but there was no way to assess whether the size of the tsunami being predicted would materialize until it actually hit.
… In the case of Fukushima Daiichi, all external power had been lost, so while planning to use the emergency diesel generators to achieve cold shutdown, we didn’t know at that point if seawater had affected the power. But after the tsunami hit was when I heard reports that power had been lost.
This is an extremely tenuous situation. Basically, not only is electricity used in order to cool down — control rods are used to halt the nuclear reaction — but then the reactor must be cooled, and a loss of power means a loss of cooling functions, so I was thinking about how serious the situation was.
How did you respond to the situation?
First we set up a real-time video conference link between the emergency response headquarters in Tokyo and the power plant to deal with the emergency.
We wanted to first accurately understand what the actual conditions were like at the plant, and then, because it had lost power, we thought it would be important to send emergency electricity-generating vehicles from Tokyo as rapidly as possible. Then, of course, even though power had been lost, I was thinking that we had to investigate whether or not the power could be restored.
But regardless, if we didn’t have a proper picture of what was going on at the power plant, we wouldn’t be able to sufficiently respond. So our first task in Tokyo was making sure from those at FD1 [reactor 1] that we properly understood the situation.
Of course we did what we could from Tokyo to provide support by sharing information about the plant’s parameters to help determine the stability of the plant’s reactors. That is how we started to manage the crisis.
What was your sense of the danger if the plant lost all power?
The loss of all power is to a certain extent anticipated in the design. But when AC power is lost for a long period of time, or DC power is lost, when you are talking about that kind of situation, we were entering into territory that rather exceeded what we had considered. My gut feeling was that the options we had available to respond to the crisis were rather limited.
Up until that point, you had not thought to place the power at a point where water could not reach it? …
Of course we have a variety of water-supplying devices and multiple engineered devices, but since these devices couldn’t function without AC or DC power, from an accident management perspective, I was thinking about what the next step would need to be.
So you are saying that essentially the situation was becoming very serious, which is what caused you to report to the government an emergency situation.
Right. We declared a nuclear emergency in accordance with established procedure, and as the events unfolded, we reported those circumstances to the government.
That evening, what was the condition of the reactor? … How were you thinking the situation would evolve?
In looking at the evolution of the crisis, because power had been lost at the plant, I felt that the difficulty of the disaster was that information became very fragmented.
Even the lights went out when power was lost in one central control room after the other, and devices that monitored the plant’s parameters could not be read without power.
The monitors were hooked up to batteries in order to obtain readings, requiring decisions to be made about the plant using extremely limited data, which made me feel that the way in which we were responding to this disaster was unprecedented.
In the midst of this, … the pressure at the reactors began to climb, and I believe you were faced with a venting problem. Late at night at a press conference, you said that venting would begin immediately, [but it seemed to take] a significant amount of time. What was the course of events?
The reason why we thought that it would be necessary to vent was, as I described earlier, based on plant parameters that were being exceeded.
On the early morning of the 12th — well, I think that it was the evening of the 11th, once power was restored to the devices that measure pressure, there was a report from FD1 that the pressure inside the reactor’s containment vessels was exceeding the design pressure. And since it would have been a grave matter had the containment vessel been breached, a decision was made at the nuclear power plant to vent.
Those in the emergency response room in Tokyo were in agreement with that decision. However, in order to vent the inside of the containment vessel, some radiation will be vented outside in order to prevent massive damage from occurring. Of course we had to first obtain permission from the government.
At the power plant, preparations were being made to vent. But first, as I indicated previously, it was dark — no lights, no electricity — so the central control room at FD1 was not immediately ready to vent, since workers had to go on-site and operate the valves, and when they did, it was difficult to know which valves were open and where the valves were, what kind of equipment was needed. So it took an enormous amount of time for the operator to assess the situation, create work procedures, etc.
In addition to the plant being dark, there were aftershocks. That was the conditions in which the work was being carried out, which required a significant amount of time.
… There was no manual that specified the procedures for venting.
For venting, the main tendencies were indicated in writing as part of the accident management policy. But with no power in the central control room, in order to operate the valves — and this is a part of accident management protocol — procedures indicating what needed to be done were drawn up on-site, taking into account conditions on the ground.
The procedures were drawn up at the location?
And while the procedures were being created, the team assignments were being determined. This is what the operators were hard at work doing.
… In other interviews, there were reports that you were handling not only reactor 1 but reactor 2 as well, and that in determining which one to prioritize, delays occurred. …
Implementation was not delayed. As I indicated previously, venting took a significant amount of time due to the confusing nature of the work. The decision regarding reactor 1, as I indicated before, was because the pressure in the containment vessel had become extremely high.
As for reactor 2, we were unable to see what the water level was inside the reactor and as a result did not know for sure if the fuel was covered in water. We made a conservative decision that assumed that the ability to supply water to the reactor had likely been lost, meaning we had to think about the next step to be taken. The fuel may have been submerged in water, but we made the assumption that it was not and had to start thinking about the next step to take.
We thought about that immediately, that if the fuel rods were not covered, that if the water over the fuel rods was gone, the fuel rods would be damaged, so initially we also had to think about venting. What that meant was that there was an independent decision made for each plant that venting was needed, and preparations were made in parallel for both reactor 1 and reactor 2.
However, in the case of reactor 2, the plant operators went on-site and confirmed that the devices that delivered water to the reactor were still functioning and shifted the work being done to reactor 1. That information came to Tokyo headquarters later.
Regarding the condition of reactor 1 and reactor 2, information was coming in during the middle of press conferences, so it probably did appear somewhat chaotic, but there were no delays on-site as a result.
On the morning of the 12th, the prime minister traveled to Fukushima. At that time, what were you doing, and where were you?
On the 12th, when the prime minister visited Fukushima, I was on the second floor in the emergency response room.
Did you see any of the prime minister’s visit to Fukushima Daiichi via video conference?
No, it wasn’t via video conference. [TEPCO] Executive Vice President [Sakae] Muto and the plant manager were with the prime minister in another room, so I don’t know what was happening at that time.
You weren’t present when the prime minister was at FD1, but can you tell me what was discussed, what you were told as part of internal reports? What kind of a discussion occurred between Executive Vice President Muto, [Fukushima plant manager Masao] Yoshida and the prime minister?
Well, the prime minister spoke about wanting to vent as quickly as possible, and from our perspective, we also wanted to quickly vent the reactor 1 containment vessel.
And I did hear that the plant manager did set around 9 a.m. of the 12th as a goal for manual [venting].
But in order to vent manually, this would take place in a location with very high radiation and was extremely dangerous work. What was Mr. Yoshida’s reaction to this?
I don’t now remember exactly what Plant Manager Yoshida said, but he is someone who knows the plant very well, and other experienced people such as the shift supervisor put together a plan through discussions in the emergency response room at the nuclear power plant.
I’ve heard that venting happens eventually after determining the evacuation status of local residents. In actuality, did you wait for the evacuation, or did you move forward as determined by the work being done on-site?
Regarding the evacuation — and keeping in mind that TEPCO is not the one who issued evacuation orders; it was the government that ordered the evacuation — we thought that it would be best if possible to wait until those living close to the plant had finished evacuating.
For the purposes of the work being done at the plant, rather than thinking about the status of the evacuation, we were focused on quickly moving forward with venting. But there was also talk in the plant’s emergency response room about the government’s evacuation orders, and information about the residents was also being received. So if possible, we wanted to take that information into account in determining the timing of the venting.
I think that was the decision, and at TEPCO headquarters in Tokyo, we got that information and supported that decision.
So you somehow succeed with the venting and are probably temporarily relieved, and then after that, at 3:36 p.m. on the 12th, there was an explosion-like event. What did you initially think happened?
I was at the emergency response room in Tokyo, but even those in the emergency response room at the plant could not immediately determine what the situation was. There were reports from the plant that there had been shaking and the sound of an explosion.
The situation could not even be immediately determined at the emergency response room at the plant: They didn’t have windows; no video cameras were set up. And then the injured began returning, and there was talk about their evacuation, and the situation was, how would you say, rather chaotic.
We could see from Tokyo that the conditions on-site were not initially conducive to determining what happened.
I have been to the power plant in Kashiwazaki and seen the earthquake-dampening room, and indeed with walls that thick, you indeed wouldn’t be able to determine what is happening outside.
Yes, and there are no windows facing toward the plant. And at that time, we saw from images provided by television cameras that the steel framework of the reactor building was exposed. We saw this condition from reports from television stations.
We also reported to the government that there had been talk of the sound of an explosion.
In talking to the government when you gave your report, you said that there was the sound of an explosion, that something had happened. Did you say specifically what had happened at what reactor?
I think that our preliminary report was not communicated properly. That started initially from not giving a detailed report about where the explosion had occurred.
Then, after going to the site of the explosion and assessing it, after that the information became more and more accurate.
Around when did the operator give accurate information to the prime minister’s office?
I don’t have the numbers in front of me, the exact time. … My memory isn’t that detailed.
… I imagine that the situation was difficult to understand and rather chaotic. That being the case, there were a large number of people on the plant grounds from before, contractors and the like. Were there evacuation orders for those people?
Yes. At the plant, after the sound of the explosion, orders were given to care for the wounded and for those doing work to temporarily retreat to the earthquake-dampening room that had become the emergency response headquarters.
Regarding talk of a withdrawal, tsunami warnings were subsequently issued, and there were additional aftershocks, and there were innumerable times when work was stopped and workers retreated to the earthquake-dampening room, including for instances other than the explosion at the reactor 1. That was the situation that we saw from headquarters in Tokyo.
There are a number of people who have said that directions were slow in coming, that they didn’t know what was going on.
It was probably true that it was a difficult situation for quickly communicating information. Communications inside the plant were probably insufficient, and because power had been lost, the procedures for contacting those doing the work were more from person to person, which no doubt made it a bit difficult to contact everyone at once.
About the meltdowns … on the 11th, what was your understanding of the condition of the reactor?
… If cooling is lost, that can lead to damage to the core, so we asked the nuclear power plant to closely watch the plant parameters to the extent possible — the plant was working with markedly limited data — asking them if they could tell what the water levels were after connecting DC power, whether they could read the reactor pressure, etc. We had them do such challenging tasks is what I remember.
… Once evening fell, I believe that headquarters ordered F1 to pump seawater into the reactor.
… Because there was a limited amount of freshwater to pump, regardless, the reactors needed to be cooled, so the determination that saltwater might have to be used was recognized by the plant manager at an early stage.
Headquarters in Tokyo had also started to make preparations for pumping of saltwater which were lined up by around 3 p.m. But at around 3:30 p.m., there was a hydrogen explosion which damaged the water lines that had been assembled, meaning that they had to be reconstructed. The situation at the plant was extremely difficult. …
That being the situation, partly from the prime minister’s office I believe that there was talk about whether or not to temporarily stop the pumping of seawater. What was that discussion about?
Preparations were made so that seawater could be pumped in as fast as possible.
Our intention in the emergency response room in Tokyo was also to pump in seawater, which we prepared to do, making our intentions known to the government and to NISA [Nuclear and Industrial Safety Agency].
However, there was also a TEPCO employee at the prime minister’s office, and there was a nuanced discussion about whether to wait a little while before using seawater, a discussion in which the Tokyo emergency headquarters was consulted.
At around 7 p.m. at the power plant, pumping was to begin as soon as preparations were complete, so the pumping of water began. The pumping of water was for test purposes, so Tokyo headquarters decided to temporarily suspend pumping and gave directions to Fukushima Daiichi to do so.
However, it was the judgment of the nuclear plant manager that it was important from a technical perspective to continue cooling the reactor. In other words, if pumping was temporarily halted, it might not be possible to restart pumping, a condition that would be dangerous.
So Tokyo’s guidance was what it was, and he took that into account, but in actuality decided to continue pumping in saltwater. We actually found this out after the fact when looking into the course of events.
When you look back at that decision, was it a good thing that he continued to pump water?
I think that the decision to continue pumping seawater was itself technically correct.
… On [the] evening of the 14th and into the 15th, the radiation had become extremely high. You made an emergency request of the prime minister’s office. What specifically was that request?
On the evening of the 14th and into the 15th, I had moved to the off-site center in Okuma City in Fukushima Prefecture to take over from Muto, so my recollections are not from personal experience but from later reports. Radiation had risen to very high levels and both those at FD1 and in Tokyo recognized that the condition of the plant had become extremely dangerous. So we informed the government that we were discussing temporarily withdrawing – excluding those workers involved with maintaining safety – to Fukushima Daini.
Did discussions about this with the prime minister’s office become heated?
I don’t know what the situation was at the prime minister’s office, but we were consistent, we never said that all employees would withdraw. We said that we wanted to look into withdrawing, or that we had started [to look into withdrawing] a portion of people, that is the truth. I don’t know specifically how that information was handled at the prime minister’s office. …
The plant condition had severely deteriorated, and a lot of workers had assembled in the earthquake-dampening room. We told the government that depending on the condition of the plant – excluding some workers who would maintain safety – we were discussing the possibility of withdrawal.
I imagine there are a variety of opinions about whether or not information was skillfully communicated to the prime minister’s office, but when the prime minister did visit TEPCO headquarters, what was the discussion that ensued?
… I was at the off-site center which was connected by video conference to headquarters, so what I saw was via video conference.
We never thought about withdrawing everyone, but I do think that there was a statement to the effect that we should not withdraw. I also felt a little bit surprised that the prime minister had come to TEPCO’s emergency response room.
[Editor’s Note: Read then-Prime Minister Naoto Kan’s take on the possibility of withdrawal, which differs from this account.]
… At a press conference on March 18, you indicated that the tsunami that hit exceeded what had been imagined, and as a result, the reactors could not be cooled. However, around 2009, at a variety of government-sponsored committees, there were a variety of warnings about the possibilities of an earthquake like the Jogan earthquake, but those warnings from experts were not acted on by management. Why would that be?
We took seriously what was being discussed by experts and had started researching the location and size of tsunamis that could be expected. And we also ran simulations of how high a tsunami would be if those tsunami earthquakes that happened in the past occurred off the coast of Fukushima.
However, if the wave-source model is not debated at the level of the expert, TEPCO cannot be expected to make a unilateral decision involving the overall design of nuclear power plants.
Around 2002, the Japan Society of Civil Engineers had created a model that estimated tsunamis [PDF], so they were asked to look into these new tsunami models, and this was the flow of our work.
So what you are saying is that before the research was done, the results known and changes made, the tsunami came first, right?
That was the result, yes. It was extremely regrettable, but we ignored nothing in the research. We ran simulations of the model being discussed, analyzed sediment at the site, continuously did these types of activities. That is what we did.
The current disaster is extremely large, probably the worst since the start of World War II. What do you think, as an individual, should be considered regrettable?
It’s difficult for me to say anything at the moment since we are still responding to the accident. But regardless, for preventing accidents in the future, it is important to have a robust policy to protect against tsunamis and have a better way to prevent reactor cores from being damaged.
This is also written in TEPCO’s accident investigation report, but it is of foremost importance to focus sufficiently on these types of policies. We must also create an accurate roadmap for the decommissioning of the plant that caused the accident at Fukushima Daiichi, and then safely implement that roadmap.
At a press conference in Fukushima on the 18th, in a very difficult atmosphere, as the first executive from TEPCO to visit Fukushima, it seems that you got a bit emotional. At that time, what were you feeling?
There were people there who had had to evacuate. There were many people who had a relationship with TEPCO, and I felt deeply sorry that those people had been forced to evacuate. I felt that the strongest.
You, too, including the time when you were plant manager, had spent a long time in Fukushima and probably had personal memories.
Yes. There were many people there that I knew, and I felt a strong sense of regret toward those individuals.