IAEA Report Admits Fukushima Unit 5 Worse Than Previously Stated

Details found in the recent IAEA report on the Fukushima disaster admitted some key information about the events at unit 5. These details show that the events at unit 5 were much more dire than the impression given to the public.

During the initial disaster little was mentioned about unit 5 as other units spiraled into catastrophic meltdowns. TEPCO’s reporting was limited, purposely avoids key events and obfuscates others. The Japanese government official report did provide more detail but it used carefully crafted language to downplay and avoid admitting the real risks unfolding at this unit.

Unit 5 was in a maintenance outage during the initial disaster. This configuration was in preparation for pressure testing of the reactor pressure vessel (RPV) that holds the fuel assemblies. The test was likely a cold hydrostatic test, similar to this one done recently at Watts Bar in the US. To conduct this test the reactor vessel was over filled with water. The concrete containment shield blocks that normally cover the reactor well during operation were removed. The large yellow steel containment cap was also removed. In addition, the lower containment air lock doors were both left open. This left unit 5 in a configuration with NO CONTAINMENT. The reactor vessel had direct interface with the refueling floor and out the lower air lock doors into the reactor building. This also allows the suppression chamber (SC) to have air exchange between it and the containment structure drywell. This matters as the drywell area is the portion where the air lock doors are. All of this would give any reactor steam dumped into the suppression chamber a direct route out of containment into the rest  of the reactor building. Any release of radioactive steam through the top of the RPV or through the steam relief valves into the suppression chamber would then be part of the air inside the reactor building. A potential major failure of the RPV would be un-contained. The only defense would be the reactor building itself. As seen at units 1, 3 and 4, the reactor buildings at Fukushima Daiichi were no match for a hydrogen explosion. Such an explosion it known to be an expected risk in a meltdown if the hydrogen can not be released, as seen at the other units.

 The illustration above shows a concrete cover (5) partially removed along with the location of the yellow containment cap that was removed in relation to the reactor vessel below.

In addition to disabling the containment structure, workers had disabled half of the reactor steam relief valves (SRV) on the RPV. These are the last line of defense to release the build up of pressure and steam inside the reactor vessel. This was done to prevent them from opening during the pressure test. Since disabling these required disconnecting multiple sets of equipment, re-enabling them would be considerable work. This left the RPV with only three SRV valves potentially available for use. With so many systems knocked out, any effort to open the SRVs would require hands on work inside the containment structure next to the over pressure RPV. The containment structure is already a cramped and complicated location to work in. Any work would have to be done by flashlight.

The over filling of the RPV to conduct the pressure test created a unique set of problems when the disaster hit. With the RPV over filled with water there was little room to accommodate the expansion of water or steam generated as the fuel heated up the reactor water. When the cascading losses of electrical power occurred, unit 5 lost the ability to circulate water in to the RPV. This allowed it to begin to overheat.

Normally in a situation where regular cooling function is lost a reactor could switch to the high pressure coolant injection (HPCI) or the reactor core isolation condenser (RCIC). Both of these systems depend on the considerable steam generated by an operating reactor as it is abruptly shut down. Since unit 5 was not actually operating and it was over filled with water, it was unable to power either of these systems. As the tsunami hit any cooling system that also relied on a sea front pump to exchange and lose heat would also be non functional.

These events left TEPCO with no viable way to cool the reactor. In later accounts of the disaster, plant supervisor Masao Yoshida said he would not abandon Fukushima Daiichi, if they did units 5 and 6 would also melt down. These findings of unit 5’s precarious situation makes that statement even more profound.

As pressure in the RPV rose, workers looked for any way to relieve pressure. Initially the vessel head vent nozzle was used beginning at 06:00 on March 12. This vent is located on the top cap of the RPV. Any venting through this route would have direct interface with the refueling floor. This valve was left open but was insufficient to lower the RPV pressure. This would have allowed any radioactive gasses generated and potentially hydrogen to escape into the refueling floor. Valve N7 on the illustration below is the vessel head vent nozzle.

 

As it became apparent that the vessel head vent nozzle was not going to be sufficient, workers tried dumping water via other systems. Valves on other systems such as the RCIC were opened to try to allow water rather than steam to flow out those lines. These attempts did not work to lower the water level or the reactor pressure. Workers then moved on to attempting to open some of the SRV vents. One of the high pressure set point SRVs that was still operable opened automatically 10 hours (1:40 on March 12) after the unit lost all power.  This managed to limit pressure but was not enough to reduce the pressure inside the reactor. At 20:48 on March 13 power was restored to unit 5. This gave them enough resources to re-establish a means to inject cool water into the reactor. At this point the pressure inside the RPV was too high, water could not be injected until the pressure was reduced.

Another SRV was opened using DC power and nitrogen supplies. This reduced the RPV pressure and allowed water injection into the reactor at 05:30 on 14 March. They continued to do a “bleed and feed” where they used the SRV to dump steam and pressure to the suppression chamber while injecting water via the MWUC system. During these events the Standby Gas Treatment System (SGTS) was offline until 20:48 on March 13 when AC power was available using the unit 6 diesel generator. For two days after the station blackout there was no filtration system on the building. At the same time they had left the vessel head vent nozzle open for an extended period of time and had repeatedly allowed the SRV valves to open, dumping steam into the suppression chamber. This increased the risk for a dangerous hydrogen build up and also for releases of radioactivity to the environment. After the SGTS was brought back online they would have had the ability to apply negative pressure and filter the air before purging it from the building. At some point during the early events at Fukushima Daiichi, workers had cut holes in the roof of units 5 and 6 in an attempt to prevent a dangerous hydrogen build up. These more detailed accounts of the problems with unit 5 show why that effort to put a hole in the roof may have been a prudent action at the time.

This set of three different reports of what happened at unit 5 clearly illustrate how different parties have chosen to not just frame the disaster but disclose or not disclose key information about what happened. The events of unit 5 could have easily taken a much worse direction. If it had, the outcome could have been much worse due to the lack of closed containment. This also shows that even a reactor in a cold shutdown mode can be a serious risk if both power and water access are knocked out at the same time.

Below are key passages from the new IAEA report and the Japanese government report on the disaster. Important sections were underlined to identify them.

The new details from the IAEA report:

Unit 5 had fuel assemblies loaded in the reactor core. The fuel had relatively low decay heat due to the period elapsed since power operations. The RPV was filled completely with water and isolated (bottled up) and was being pressurized by a pump in preparation for the RPV pressure (leak) test. Its confinement structure, the PCV, was open, with its lid removed.

In Unit 5, the reactor pressure, which was kept at an elevated level by the use of a pump for pressure testing at the time of the earthquake, initially dropped when the pump stopped as it lost power as a result of the LOOP. The pressure started to rise in the water filled RPV, as a result of the decay heat, but, unlike in Units 2 and 3, it remained for some time well below the levels that would activate the SRVs.

LOOP = Loss Of Offsite Power aka AC power from the grid

DC power was also available in Unit 5. The residual heat removal by a high pressure cooling systems — the RCIC or high pressure coolant injection (HPCI) — was not possible, since the reactor was not generating steam to drive dedicated turbines to run the pumps. The reactor had to be depressurized to enable coolant injection by low pressure injection systems, such as the makeup water condensate (MUWC) and residual heat removal (RHR) systems. However, depressurization using the SRVs was not possible, since the depressurization function of the SRVs (except for three SRVs with the highest opening pressure) was disabled. Alternative options, such as discharging water from the RCIC and HPCI lines, were tried unsuccessfully, and the RPV, which was pressurized and filled with water, continued to heat up and pressurize. 

1.1.3.5. Heating up of Unit 5 and restoration of AC power
During the time when core cooling and confinement functions in Unit 1 were the primary concern, the Unit 5 reactor had continued to heat up in the absence of heat removal measures. An SRV in Unit 5 automatically opened for the first time approximately 10 hours after the SBO, because the reactor pressure reached its opening set value, at 01:40 on 12 March. The valve automatically opened and closed several times to maintain the pressure in a range determined by the design. The SRVs were operating automatically to limit pressure but could not be used to reduce pressure, since most of them had their depressurization function disabled for the test carried out before the accident. Reducing pressure by opening a small valve (the head vent nozzle) on the RPV was considered as an alternative because DC power was available for this purpose. 

1.1.3.20. Establishing core cooling and confinement functions of Unit 5
At around 20:48 on 13 March, the power from the Unit 6 EDG was connected to the pump of Unit 5’s normal, low pressure, non-safety heat removal system, the MUWC, and the pump was started at 20:54. The MUWC–RHR interconnecting pipe valves were remotely opened, completing a reactor water injection line to Unit 5’s reactor via a train of the RHR system at 21:00, approximately 53 hours after the SBO. However, the reactor pressure had gradually risen and exceeded the RHR pump discharge pressure, which prevented water injection to the reactor. In response, the shift team, acknowledging that the pressure reduction by the head vent nozzle was not sufficient, decided to use the SRV set that was not locked, for additional RPV pressure relief. An SRV was opened, making use of available DC power and nitrogen supplies, reducing RPV pressure and allowing the water injection into the reactor at 05:30 on 14 March. This bleed (by SRV) and feed (by MUWC pump) operation continued afterwards. In addition, the AC power to operate the standby gas treatment system (SGTS) for controlling RB pressure was supplied from Unit 6’s EDG. The system was started up a little over two days after the SBO and kept the pressure inside the RB below the atmospheric pressure, ensuring secondary confinement.

Unit 5 was the first unit to reach cold shutdown mode. Its shutdown heat removal system was put into service as the RHR system was aligned to the shutdown cooling (SHC) operation at 12:25 on 20 March. The reactor temperature decreased to below 100°C in approximately two hours, placing Unit 5 in the cold shutdown mode at 14:30 on 20 March 2011, nearly nine days after the earthquake.

 

From the Japanese government official reports on the disaster, they paint a much different picture of the events at unit 5:

(1) Overview of the response to the accidents at Units 5 and 6 of the Fukushima Dai-ichi NPS When the Tohoku District off the Pacific Ocean Earthquake occurred on March 11, 2011 the reactors of Units 5 and 6 at the Fukushima Dai-ichi NPS had been shut down for routine inspections. Compared to the plants that were in operation, those plants had much less decay heat generated in reactors and maintained adequate water levels. After the arrival of the tsunami, Unit 5 lost all AC power, while the adjacent Unit 6 had a single emergency diesel generator (emergency DG) operating continuously to supply AC power. Thanks to this emergency DG, AC power was interconnected from Unit 6 to Unit 5, enabling shift operators check the reading of the various monitoring instruments not only for Unit 6 but also for Unit 5 in the main control room for Unit 5 and 6 (hereinafter referred to as the “Units 5 & 6 main control room”). Furthermore, this emergency DG allowed shift operators to take actions necessary for plant control such as depressurizing the reactors and injecting water into the reactors. Nevertheless, at Units 5 and 6, their seawater pumps suffered damage due to the impact of the tsunami, and so it was impossible to start their residual heat removal (RHR) systems. The course of action for handling the accident was to take steps to restore the RHR systems while controlling the reactors by depressurizing the reactors and continuing to inject water. After restoring their capabilities, the RHR systems removed the heat of the SFPs whose water temperature had been rising, and subsequently cooled down the reactors. Both units reached cold shutdown conditions on March 20 (see Attachment II-3-1).

(a) Situation at Unit 5 Due to the routine inspections, the Unit 5 reactor had been shut down with fuels still inside the reactor since January 3, 2011, and the reactor was in a state of cold shutdown. On March 11, the day when the earthquake occurred, from around 8:30 the shift team (meaning all members including the shift supervisor and the other shift operators; hereinafter referred to “shift team”) had been filling the pressure vessel with water and had been increasing its pressure180 in order to carry out a leak and hydrostatic test on the pressure vessel181. At the moment of the earthquake the plant status of Unit 5 was: the reactor pressure was approximately 7.15 MPa gage; the reactor water level was at about 8,700 mm on the shutdown range water level indicator182 (hereinafter referred to as the “reactor water level indicator (shutdown range);” see Attachment II-3-2); the reactor water temperature was approximately 90.6℃; and the SFP water temperature was approximately 23.7℃

The Unit 5 pressure vessel head had been closed for the leak and hydrostatic test, while the containment vessel cover was left open. Furthermore, the main steam isolation valve (MSIV) outside the containment vessel had been left open at Unit 5.

The ERC had grasped the information that heavy oil tanks had been washed up along the seaside area at Units 1 through 4. Supposing that the seaside area at Units 5 and 6 was in similar conditions, they thought that it would take time to restore the seawater system pumps. For this reason, the ERC recognized that, during the period in which reactor cooling was interrupted at Unit 5 & 6, it would have to depressurize reactors and inject water into them as needed. In addition, since both Units 5 and 6 were undergoing routine inspections, it was hard to conceive that there would be striking changes such as a sharp rise in reactor pressure or reactor water level. Furthermore, the reactor water level at the moment of the earthquake was adequate. Moreover, despite the fact that water must be injected into the reactors to make up for the dropping reactor water levels due to the decay heat generated in the their fuels, the ERC believed that there would be a relatively long deal of time to spare before the reactor state reached such a stage.

To start with, the Unit 5 pressure vessel was filled with water, and so it was not possible to start the RCIC system and the HPCI system, both of which are steam-driven. Since all AC power to Unit 5 had been lost, power supply had to be restored to run the systems related to the alternative means for injecting water

On the other hand, at the moment of the earthquake, at Unit 5 the reactor pressure was as high as approximately 7.15 MPa gage191, with a strong possibility that the pressure increased thereafter. In order to inject water into the reactor from the MUWC system, the reactor pressure had to be reduced to below the maximum discharge pressure for the condensate transfer pump, which was 0.98 MPa gage. Therefore, for Unit 5, the shift supervisor concluded that it was necessary to control the reactor pressure until it became possible to inject water into the reactor from the MUWC system. In general, in order to control the reactor pressure, the SR valves are remotely opened to depressurize the reactor from the Units 5 & 6 main control room, letting the steam be discharged from inside the pressure vessel into the S/C. However, before the earthquake, the shift team had already taken measures to ensure that none of the SR valves be remotely operated from the Units 5 & 6 main control room in order to conduct the leak and hydrostatic test on the reactor pressure vessel at Unit 5. Specifically, to ensure that the SR valves were not opened due to operational error, the electrical power fuse had been removed from the electronic circuits on the back of the control panel in the Units 5 & 6 main control room. In addition, the nitrogen supply line valve had been closed and, in parallel, the accumulator blow valve had been closed, so that nitrogen would not be supplied to the SR valves, which were nitrogen-driven valves. Accordingly, in order to remotely open the SR valves from the Units 5 & 6 main control room, the shift operators had to connect the electrical power fuse terminals to the electronic circuits on the back of the control panel in the main control room. In addition, the shift operators also had to operate the manual valves installed inside the containment vessel in order to configure the line for supplying nitrogen to the SR valves. However, the shift supervisor wanted to avoid field work in the containment vessel—which had poor footing without light—as much as possible. He thought that he had to first secure another means of reducing the reactor pressure without using the SR valves to control the reactor pressure and decided to carry out work procedures designed to configure the line for injecting water into the reactor from the MUWC system. 

(b) Reactor depressurization through the vessel head vent nozzle Beginning in the evening of March 11, the NRC ERC Operation Team and the shift team examined the means by which the reactor could be depressurized without actually entering the containment vessel in Unit 5192. Then the ERC and the shift team figured out that the reactor would be depressurized if water could be discharged from the reactor pressure vessel, which was full of water, by opening the vessel head vent nozzle. They decided to act on this idea and, from early in the morning of March12, started to examine how to put this idea into practice. In order to open the vessel head vent nozzle, nitrogen had to be supplied to the vent nozzle, which was driven by nitrogen, from the nitrogen tank installed outdoors. However, an electromagnetic valve sitting on the nitrogen supply line was an obstacle to the nitrogen supply, because it was impossible to excite the electromagnetic valve due to the loss of power supply. As such, from about 5:00 on March 12 the ERC Operation Team inserted a tool into this electromagnetic valve to force it open on the first floor inside the Unit 5 R/B and successfully configured the nitrogen supply line. After this, at about 6:06 that same day the shift team remotely opened the vessel head vent nozzle from the control panel in the Units 5 & 6 main control room193. After the above-mentioned operation completed, the reactor pressure was reduced from approximately 8.3 MPa gage at around 06:00 that day194 to about 0.2 MPa gage by around 06:30 that day. And then, in order to keep the reactor pressure low, the shift team kept the vessel head vent nozzle open thereafter. 

(d) Depressurizing the reactor by opening SR valves and injecting water into the reactor As was described in (b), after the shift team opened the vessel head vent nozzle to depressurize the reactor at around 06:06 on March 12, it kept the vent valve open and continuously monitored indicators like the reactor pressure indicator. Although the vessel head vent nozzle had been kept open, the reactor pressure began to rise gradually 201. The shift team attempted to depressurize the reactor by discharging water from the pressure vessel through the RHR piping and the main steam pipes from March 13. But the shift team was unable to lower the reactor pressure by either of these means. As such, the shift team thought that it had no other option other than to carry out work inside the containment vessel. As was described in (a), in order to operate the SR valves for Unit 5 from the Units 5 & 6 main control room, the shift team had to connect the terminal for the electrical fuse to the electrical circuit on the back of the control panel inside this main control room and had to enter the containment vessel to configure the nitrogen supply line. Moreover, the shift team had locked with tools the SR valves whose set pressure were low for functioning as safety valves, excluding three valves (Valves A, G, and H)202 whose set pressure were high for functioning as safety valves, in a position to ensure that those valves would not automatically open due to safety valve function while the pressure was kept high for conducting the leak and hydrostatic test on the reactor pressure vessel. Because of this, the shift team decided to use one of the three valves (Valves A, G, and H), none of which required the task of removing the tools from the valves, and concluded to use Valve A for depressurizing the reactor on the grounds that this valve required the least amount of work to configure its nitrogen supply line. From around 02:25 on the same day the shift team had started to configure the nitrogen supply line to this SR valve (Valve A) and, at around 05:00 on the same day, remotely opened this SR valve (Valve A) from the Units 5 & 6 main control room203. As a result, the Unit 5 reactor pressure, which had been approximately 2.0 MPa gage at around 05:00 that day, fell down to about 0.8 MPa gage at around 05:20. Afterwards, since the reactor water level dropped as a result of reactor depressurization by opening this SR valve, the shift team remotely opened the RHR system discharge valve from the Units 5 & 6 main control room at around 05:30 on the same day and injected water into the reactor from the MUWC system via the RHR Train B204. From then onwards, the shift team monitored the reactor pressure indicator and the reactor water level indicator and maintained the reactor pressure and water level by setting 2 MPa gage as the reference criteria for opening the SR valve. That is to say, whenever the reactor pressure exceeded the reference criteria, the shift team depressurized the reactor to the pressure lower than about 0.8 MPa gage by operating the SR valve and injected water from the MUWC system

200 The eight members of the NPS ERC Recovery Team and two employees from a contractor laid and connected approximately 220 m of cable. What is more, thanks to the power interconnection, power supply also became available for the Unit 5 SGTS, and so at around 21:00 on March 13 the shift team started up the SGTS for the purpose of maintaining negative pressure within the Unit 5 R/B.

TEPCO’s unit 5 report can be found here but gives such a vague narrative it seems to be almost a completely different event.

 

 

This article would not be possible without the extensive efforts of the SimplyInfo research team
Join the conversation at chat.simplyinfo.org

© 2011-2016 SimplyInfo.org, Fukuleaks.org All Rights Reserved Content cited, quoted etc. from other sources is under the respective rights of that content owner. If you are viewing this page on any website other than www.simplyinfo.org (or www.fukuleaks.org) it may be plagiarized, please let us know. If you wish to reproduce any of our content in full or in more than a phrase or quote, please contact us first to obtain permission.

Nancy Foust

Editor, SimplyInfo.org

Leave a Reply