Chimgu nuclear accident

The Chimgu nuclear accident, also known as the Chimgu disaster, was a nuclear accident that occurred at the Chimgu Nuclear Power Plant, located in Chimgu County, a part of Chimchŏn Prefecture (today Yŏng'an Municipality) in Gangwŏn Province, Menghe on April 3rd, 2003. It consisted of a partial nuclear meltdown in the Number 3 reactor building, brought on by a loss-of-coolant accident which resulted from a broken relief valve and several faulty pressure sensors. It is classified as a "serious accident," with a score of 6 on the International Nuclear Event Scale.

Subsequent investigations concluded that many of the problems leading up to the accident could be traced back to cost-cutting measures imposed during the construction of the Chimgu Nuclear Plant, including the installation of outdated electronic equipment and the omission of redundant safety systems. Workers at the plant were given insufficient training on how to handle a loss-of-coolant accident, and poor decisions in the first two hours after the accident intensified the problem. Human error also compounded the initial recovery effort, as the Gangwŏn Provincial Government, and later the Menghean central government, attempted to conceal the severity of the accident from the general public.

The accident resulted in three separate releases of radioactive material beyond the containment chamber. The first was a venting of contaminated coolant steam into the atmosphere as the incident was in progress. The second was a hydrogen explosion inside the Number 3 containment unit, which released radioactive gases but did not expose the core. The third was a continuous dumping of coolant water from the reactor facility into the Chim river, which empties into the Meng river and from there runs through several major population centers. On April 10th, the authorities imposed a 10-kilometer exclusion zone around the nuclear plant, and ordered that crops and livestock in the downwind area be destroyed.

Estimates of the death toll vary. Official state sources report that two plant workers and one emergency worker died of radiation exposure, with no statistically significant increase in cancer or birth defects downstream or downwind. Unofficial estimates place the number of radiation-related deaths as high as 200,000 over the course of 10 years, though there is extensive debate over how to distinguish Chimgu-related cancers from cancers due to industrial pollution, also high in the area.

Background

Construction

Construction work on the Chimgu Nuclear Power Plant (침구 원자 력발 전소 / 沈溝原子力發展所, Chimgu Wŏnja Ryŏkbal Jŏnso) began on September 9th, 1994, as part of the Second New Five-Year Plan. The plant was commissioned on February 3rd, 1999, close to a year ahead of schedule. It was Menghe's sixth nuclear power plant built for the purpose of generating electricity for civilian needs, and the third to be opened after the Decembrist Revolution. Five more plants were still under construction at the time of the accident, part of a broad state-led campaign to expand the country's nuclear power sector.

As the province's cadre promotion system placed a strong emphasis on economic growth and infrastructure expansion, officials overseeing the plant's construction had strong incentives to complete the project below budget and ahead of schedule. This led to a number of cost-cutting measures which would prove disastrous in the future. Many parts were sourced from the lowest bidder, leading to a heavy reliance on small private enterprises with unscrupulous records. Economic growth targets also led officials to favor bidders from Gangwŏn province or from Chimgu county itself, even where better-quality foreign equipment was available. Migrant laborers hired for construction had little experience in reactor engineering, and worked exhausting 12-hour shifts with minimal breaks.

There were also ample opportunities for corruption. Ri U-hyŏn, the Party Secretary of Chimchŏn Prefecture and future Party Secretary of Gangwŏn Province, awarded the contract for the site to the state-owned Yŏng'an Oho Construction Company, which was run by his cousin. The company had no prior experience designing and building nuclear power plants. Oho Construction also bribed inspectors from the national Nuclear Regulatory Commission, offering them large sums of money to expedite the approval process. Similarly, many staff positions at the plant were filled through family and school-network channels, and in a few cases qualification documents were forged or altered in order to ensure that the employees held the necessary qualifications.

In a final effort to bring the plant online ahead of schedule, the facility manager, Yun Jae-sŏng, ordered that some of the pre-activation safety tests and employee drills be replaced by a series of on-paper exercises simulating equipment failure scenarios. Ironically, the engineer leading these exercises, Bae Chang-min, realized that the lack of redundancy in coolant systems could trigger a loss-of-coolant accident and meltdown, but Yun sealed his report in a file drawer and declared the facility safe for operation.

Layout of the plant

The Chimgu Nuclear Power Plant used a conventional pressurized water reactor design, the most popular design for civilian nuclear plants. A "primary loop" of water circulates directly through the reactor core, where it acts as a neutron moderator and coolant. Hot, pressurized water is drawn from the core into a steam generator, where it transfers its heat to water flowing in a "secondary loop." This water, heated into pressurized steam, powers a turbine, which generates electricity. Steam from the turbine flows through a condenser, where it is cooled through heat transfer to water in the "tertiary loop." This water does not enter the reactor chamber itself, and may be safely vented from cooling towers as part of the plant's regular operation.

An important consideration in such a design is that the water in the primary loop must be kept circulating and under pressure at all times. If pressure falls, or if circulation stops, the water may boil, creating corrosive steam bubbles inside the reactor core itself. The narrow pipes inside the heat exchanges must also be kept free from blockage and corrosion, which could interrupt the regular flow of water. In the case of Chimgu, the plant's main water source was the mineral-rich Chim river, which flows southwest from the Chŏnsan mountains before joining the Meng river in southern Gyŏngnam province. As such, the purification of the water depended on a set of condensate polishers, thin polymer-mesh filters designed to keep the water in the secondary loop at a high level of purity.

At the time of the accident, the Chimgu Nuclear Plant had a total of four reactors. Reactor 1 began operation in 1999, Reactor 2 in 2001, and Reactor 3 in 2003. The fourth reactor had been completed, but it was not yet operational, and was scheduled to come online within 18 months. All reactors used the same fundamental design, and each was hooked up to two turbines, numbered 1 through 6.

Technical problems at Number 3 Reactor

The Number 3 Reactor came online on February 14th, 2003, and had been fully operational for less than two months at the time of the accident. During this time, the engineering team had already become aware of a few problems. The Number 5 turbine hooked up to the reactor was observed to exhibit unusually high levels of shaking and knocking during operation, particularly when running at close to full capacity. This was most likely the result of poor tolerance levels between parts, though the exact cause remains unknown.

Full replacement of the turbine was expensive, and would force the plant to run at reduced capacity until reconstruction was finished. It would also place a negative mark on the construction team's performance assessment. As such, Yun Jae-sŏng ordered the work teams at Reactor 3 to keep the Number 5 turbine running at reduced capacity until their on-site engineers could determine the source of the problem.

Accident

The secondary loop shuts down

On April 3rd, 2003, at around 10:00 AM, knocking in the Number 5 turbine abruptly increased, possibly as a result of a loose part coming detached. Vibrations from the turbine traveled up and down the secondary coolant loop, setting off an emergency safety valve which tripped the turbine at 10:04. Under normal circumstances, a turbine trip should re-route the turbine steam feed directly into the condenser, continuing circulation in the secondary loop but cutting off the flow through the turbine itself. In this case, however, the pipe linking the turbine to the condenser violently ruptured, possibly due to faulty components, a valve left open, or prior damage from the vibration surge nearby. Compressed steam and boiling water surged into the turbine room, damaging pressure sensors and short-circuiting electrical lines. This caused the main pump and condenser pump shut off.

In the control room, the reactor operators received an alert that pressure in the return side of the secondary loop was falling rapidly, and noticed that the main pumps were not operational. Yun Jae-sŏng, who was present at the time, ordered them to turn on the backup pumps to keep the secondary loop flowing. The wiring to the backup pumps, however, had also been damaged in the turbine trip, and electronic commands failed to activate them. Yun dispatched an engineer to turn the valves by hand.

The relief valve fails

Because secondary-loop water was no longer flowing into the steam generator, the primary loop of reactor coolant no longer had a way of transferring heat out of the system. The control team responded by putting the Number 3 reactor into emergency shutdown, inserting all control rods to stop the reaction. Even with all control rods inserted, however, the fuel rods still generated a large amount of decay heat. Pressure and temperature in the primary loop continued to climb as a result.

At 10:08, the pilot-operated relief valve attached to the primary loop pressurizer automatically blew open to relieve the buildup of pressure. Radioactive steam from the primary loop was vented into the reactor building, though it did not escape outside into the environment. The pilot-operated relief valve should have closed automatically after pressure fell back within acceptable levels, but due to a mechanical failure, it remained stuck in the open position. Because the solenoid used to pull open the valve was set to the "off" position, the workers in the control room mistakenly believed that the valve was closed.

Shortly after this time, the engineer sent to open the valves to the emergency pumps returned to the control room, where he reported that the valves were inaccessible because a ruptured pipe was venting high-pressure steam. At this point, the reactor engineers became fixated on the question of how to access the emergency pumps in the secondary loop. Complicating any response, any new water pumped into the system would have to pass through either the turbines or the ruptured diverter pipe. None of the workers present at the time had trained for this particular scenario, and there was heated disagreement over how to proceed. In the midst of this debate, one of the workers noticed that pressure in the primary loop was falling, and brought this to Yun's attention; Yun attributed this report to faulty equipment, noting that the relief valve was shut and the temperature was still high.

At around 10:15, the pressurizer relief tank hooked up to the stuck-open relief valve overflowed, dumping radioactive primary loop water onto the floor of the containment building. From there, it flowed into a sump pit on the floor of the containment building, setting off an alarm at 10:16. Despite indications of rising temperature and pressure in the containment building, Yun remained adamant that the pilot-operated relief valve was closed, and attributed the alarms to an overflow out of the flooded turbine room. Automatic sump pumps began pumping water from both buildings into a holding tank outside the containment building, transferring radioactive coolant into the open.

Depressurization and partial meltdown

Meanwhile, the constant leakage of coolant worsened conditions inside the already overheated reactor. As pressure fell and temperature rose, the primary loop water began to boil, creating steam bubbles inside the reactor and the coolant pipes near it. Steam bubbles also interfered in the flow of coolant into the reactor itself, causing a backlog of water inside the pressurizer. Yun Jae-sŏng and the reactor engineer team interpreted the rising coolant levels in the pressurizer as further evidence that the relief valve was closed, even as alarms linked to the sump pump continued to sound. At this point, Yun was more concerned that coolant levels in the pressurizer were too high, and he ordered that the backup heat exchange be shut off to create more steam in the system. In reality, the growing steam voids inside the reactor had already left large sections of the fuel rods exposed.

The temperature of the primary coolant steadily increased over the course of the next hour, until the main reactor coolant pumps were pumping a steam-water mixture rather than pressurized water alone. This caused dangerous levels of cavitation, sending all four pumps into emergency shutdown between 11:24 and 11:28. Yun again dismissed the seriousness of the issue, confident that natural convection would continue to circulate the coolant until the pumps could be restored.

In fact, the steam voids inside the reactor quickly brought primary coolant circulation to a halt. This accelerated the boiling process, as fresh water was no longer being drawn into the reactor. Large sections of the fuel rods were now exposed, and their zircalloy cladding began to react with the high-pressure steam, in a chemical reaction that produced additional heat. Once the cladding melted off, the fuel pellets inside the rods were exposed, releasing particles of radioactive material into the coolant water leaving the reactor chamber. The melting reaction also produced large quantities of flammable hydrogen gas, which along with the radioactive material was channeled into the primary loop and vented out of the stuck-open relief valve into the containment building.

It took another hour for the vented primary coolant water to reach radiation detectors in the containment building, setting off an alarm at 12:41. Distrusting the sensors, Yun sent an engineer to the containment building to manually check the radiation levels. The engineer returned reporting that coolant water was pooled on the floor, and radiation levels in the spilled coolant were 400 times greater than normal. This was firm evidence that the cladding inside the reactor had degraded catastrophically.

Realizing that the high temperature readings in the tailpipe of the pilot-operated relief valve were not in fact the result of faulty equipment, the reactor operators activated a backup valve to cut the flow of coolant out of the system. By this time, the pilot-operated relief valve had been open for two hours and 37 minutes, and 150,000 liters of radioactive coolant had been released from the primary loop. The containment building was highly contaminated, and the sump pumps had drawn contaminated water into a holding tank which lacked proper radiation shielding. Pressure inside the containment building had also reached dangerous levels, as a result of the venting of steam and hydrogen gas. Fearing a rupture or explosion, Yun ordered the venting of gases from the containment building into the atmosphere - the first release of radioactive material to result from the accident.

Aftermath

Death toll

The exact death toll from the Chimgu nuclear accident is still heavily disputed, and much of this dispute is political. The General-Directorate for Energy, under the Ministry of Economic Development, has a strong incentive to expand the country's nuclear power sector, and its detractors have accused it of covering up the true death toll in favor of its own less severe statistics. Government officials contend that the higher death figures are the product of biased science and anti-nuclear alarmism, and have refused to allow any independent studies of radiation levels and their health effects.

Other issues stem from the question of how to set the baseline for the number of deaths. Apart from the handful of high-exposure cases around the plant itself, which were well-documented and acknowledged by the Menghean government, most radiation exposure would have occurred in moderate doses, through the consumption of air, food, and water contaminated by radioactive material. Cancer from these sources could take several years to be diagnosed, and even then might not cause death until several years later. Adding to measurement difficulties, the Chim river also experiences high water pollution from non-radioactive industrial waste, which in combination with air pollution has contributed to an elevated baseline rate of cancer in southeast Gangwŏn Province. Many working-age adults from Chimgu county also migrated to Menghe's coastal cities to seek work, often in workplaces with lax safety standards. Finally, news of the radiation leak led individuals in the Chimgu area to consult oncologists and report suspicious symptoms at above-average rates, resulting in a rise in the share of cancer cases which are officially diagnosed. All of these problems have contributed to an elevated background level of reported cancer and illness, making it hard to distinguish how much of that increase stems from the radiation releases themselves.