Chernobyl

The Chernobyl Nuclear Power Plant, now decommissioned, is located near the city of Prypiat, Ukraine. The V.I. Lenin Nuclear Power Station as it was known during the Soviet times, consisted of four reactors of type RBMK-1000, each capable of producing 1000 megawatts of electric power (3.2 GW of thermal power). Construction of the plant and the city of Pripyat to house workers and their families began in 1970, with reactor no. 1 commissioned in 1977. It was the third nuclear power station in the USSR of RBMK type, and the first ever nuclear power plant on Ukrainian soil. The completion of the first reactor in 1977 was followed by reactor no. 2 (1978), no. 3 (1981), and no. 4 (1983). Two more reactors, nos. 5 and 6, capable of producing 1000 MW each, were under construction at the time of the accident. Reactor no. 5 was almost complete at the time of the accident and was scheduled to start operating in the fall of 1986. However, it has since been abandoned; construction cranes still stand next to it to this day. Chernobyl 4 was a Soviet RBMK-1000 type reactor rated at 3,200 MW thermal power and 1,000 MW electric output. There were three similar operating units at the Chernobyl site. The RBMK-1000 is a graphite-moderated boiling water reactor that contains 1,661 parallel, vertical pressure tubes loaded with fuel assemblies. Reactor water flow is provided by six of eight installed main circulation pumps; two pumps are installed spares. The flow through each of the fuel channels is adjusted using regulating valves on the inlet of each channel to control heat flux margins. The water-steam mixture leaving the top of the fuel channels flows into four horizontal steam drums with moisture separators. The dry steam drives two 500 MWe turbine generators (numbers 7 and 8). Feedwater is fed directly to the steam drums, bypassing the reactor, to control water level. The core is large, 23 feet in height and 39 feet in diameter, so that regions of the core are loosely coupled to each other neutronically. As a result, individual regions of the core can be close to criticality. The reactor exhibits a reactivity increase as water density in the core decreases (positive void coefficient of reactivity). The fuel temperature coefficient of reactivity is negative. In normal operation, the overall core power coefficient is negative at and near full power but becomes positive at lower power levels. The minimum permitted power level for steady state operation is 700 MW(th) (22 percent of full power). The station was used for base-load operation. Since the plant was refueled on line, excess reactivity was not high. Unless quickly restarted after a shutdown, startup was delayed due to xenon poisoning. The system for control and protection of the reactor is based primarily on movement of 211 boron carbide absorber rods in vertical channels adjacent to the fuel channels. The rods have graphite followers attached to displace water in the rod channels. However, according to Soviet descriptions of the RBMK1000, the followers are apparently not full core length, so that when a rod is fully withdrawn from the core, about one meter of water remains in the rod channel below the follower. The rod control system provides automatic control of power level and of the flux profile in the core. In a normal configuration, the protection system can quickly reduce power by partial rod insertions; automatic scram is the ultimate response. The time to fully insert rods for a scram is 20 seconds. A minimum "operating reactivity margin" is specified. This margin is referred to by the Soviets as the equivalent of 30 inserted regulating rods. Control rods are required to be partially inserted into the core during operation to enhance the initial negative reactivity insertion rate on scram. In addition, insertion of control rods reduces the positive void coefficient. The rods are also adjusted to correct spatial power instabilities in the core. (It is our understanding that the procedural requirements for 30 equivalent rods minimum reactivity margin is a means of specifying an overall rod configuration that ensures a certain initial negative reactivity rate on scram. It also apparently prevents an initial positive reactivity insertion that can occur when rods enter the core from the top, displacing water in the rod channels near the bottom of the core.) During the performance of a turbine-generator coast-down test on April 26, 1986, Chernobyl Unit 4 experienced a severe reactivity excursion that, with the accompanying pressure surge and fire, destroyed the reactor and breached the surrounding building. The test procedure had not been adequately reviewed from a safety standpoint. Management control of the evolution was not maintained; the test procedure was not followed; several safety functions were bypassed; and control rods were misoperated. Operators lost control of the reactor during the performance of the test. The Chernobyl accident in 1986 was the result of a flawed reactor design that was operated with inadequately trained personnel and without proper regard for safety. The April 1986 disaster at the Chernobyl nuclear power plant in the Ukraine was the product of a flawed Soviet reactor design coupled with serious mistakes made by the plant operators in the context of a system where training was minimal. It was a direct consequence of Cold War isolation and the resulting lack of any safety culture. Note: The type of reactor involved in the Chernobyl accident is of an entirely different design than that from U.S. commercial reactors. It lacks important safety features that are built into U.S. plants. Additionally, the societal, institutional, and operational approaches in the former Soviet Union were quite different than the U.S. The following major operational events or errors and administrative or management control breakdowns led to the accident:

Overall management control of the evolution and its integration with plant operations was not clearly established. The test was directed by an engineer with expertise in the turbine-generator/electrical area only.
The test procedure did not receive an adequate safety review.
The operators felt a sense of urgency to complete the test. The test would have been delayed by one year if not performed at that time. The reasons for the sense of urgency were not well explained in the Soviet report but may have been caused by outside or management pressures. Since the evolution occurred over a 24-hour period, more than one operating shift was involved. The test was conducted early in the morning and just prior to a Soviet national holiday. These factors may have influenced performance.
The power reduction for the test was interrupted at the load dispatcher's request for nine hours. This changed the initial conditions in the core from those contemplated in the procedure and may have contributed to the inability to later attain the power level specified for the test.
The operators did not properly adjust the set point on the automatic power level controller. As a result, the power decreased to 30 MW(th).
The operators did not follow the test procedure:
- The reactor scram signal for the trip of the second turbine-generator was bypassed, which violated the test procedure. This was done to enable repeating the test if necessary. The test probably would not have led to the accident had the reactor scrammed at the start of the test.
- Unable to raise the reactor power back to 700 MW(th), the operators started the test at a lower power level, which violated test procedure requirements. (The test was started at six percent power instead of 22 to 31 percent as specified in the procedure.) At this power level, the reactor was difficult to control. Also, with very little steam being generated, the eight circulation pumps produced a flow rate above allowable limits. With the high flow rate and low power level, the water inlet temperature to the core was very close to saturation. Under these conditions, an increase in power caused a much greater increase in steam voids and reactivity than normal.
Other safety systems were also defeated:
- Bypassing of reactor scrams associated with steam separator pressure and water level enabled operating the reactor despite unstable conditions.
- Control rods were withdrawn well beyond safety limits specified by procedures. This was done to compensate for xenon buildup and negative reactivity resulting from void suppression in the core.
- The emergency core cooling system was deactivated for over nine hours while the plant was operating.
The design of the plant placed a heavy dependence on adherence to administrative controls and procedures for safe operation. However, the plant operators did not demonstrate an adequate understanding of the safety implications of their actions. Their willingness to conduct the test at a very low power level, with an abnormal (and unauthorized) control rod configuration and core conditions, and with safety features bypassed indicates an insufficient understanding of the reactor and its potential behavior.

On April 25, 1986, at 0100 hours, the operators started reducing power to perform a test on turbine-generator number 8, prior to a maintenance outage. The test was intended to determine how long the turbine-generator would continue to supply power near rated voltage for essential equipment as the generator coasted down. Similar experiments had been carried out twice before at Chernobyl. At 1305, turbine-generator number 7 was shut down with the reactor at 50 percent power. Much of the electric power for the plant, including four main circulation pumps and two main feed pumps, was now being provided by turbine-generator number 8. At 1400, the emergency core cooling system was defeated in accordance with the test program. Evidently this was to ensure that an inadvertent actuation would not interfere with the test. The test was then delayed due to a request from the load dispatcher to continue power generation. Operation of the plant at reduced power, using only turbine-generator number 8, continued until 2310 on April 25. (Continued blocking of the emergency core cooling system during this delay was in violation of operating procedures.) At 2310 the power reduction resumed. According to the test procedure, the coastdown of the generator was to be started with the reactor power at 700-1000 MW(th). However, in switching the automatic control systems from spatial power control to power level control, the control point had not been properly set. As a result, the power fell below 30 MW(th). At 0100 on April 26, power was stabilized briefly at 200 MW(th). Xenon poisoning of the reactor continued to increase. To compensate, the operators withdrew additional rods (more than allowed by procedure). A further increase in power to 700-1000 MW(th), as called for in the test procedure, was hampered by the small excess reactivity available. At 0103, two additional main circulation pumps were started, so that all eight were running. (Four pumps were powered by another source to provide for cooling of the reactor core, while the four circulation pumps powered by turbine-generator number 8 would coast down as part of the test.) The coolant flow rate now exceeded the maximum allowed by operating procedures. The additional flow reduced the steam content of the coolant in the fuel channels, resulting in lower steam pressure and lower water level in the steam drums. These circumstances also led to the core inlet temperature of the coolant (and probably the bulk temperature) being very close to saturation. Under these conditions, the operators experienced difficulty in controlling steam drum pressure and water level. To avoid an automatic shutdown, the operators bypassed the emergency protection signals for reactor scrams on steam drum pressure and water level. Reactivity continued to drop slowly due to xenon buildup and decreased voiding, requiring further control rod withdrawal to maintain power at 200 MW(th). At 0122:30 the operators noted that the available reactivity margin, related to the number of rods and their position in the core, had dropped well below the level requiring immediate shutdown of the reactor (six to eight "equivalent" rods vice the 30 "equivalent" rods required by procedure). Nevertheless, the operators continued with the test. The operators also bypassed the reactor scram that would be called for by the shutdown of the second turbine-generator (number 8). This was done so that it would be possible to repeat the test if the first attempt was unsuccessful. At 0123:04 the stop valves for turbine-generator number 8 were closed, starting the coast-down test. The reactor was still critical and at power. Four of the eight main circulation pumps began coasting down as the test started. Conditions had now been established for a severe transient, as follows:

The reactor was critical but at a very low power level where it was unstable and difficult to control.

The overall coefficient of reactivity was positive with the steam (void) coefficient being predominant.

Control rods were near the top of the reactor, in a region of low reactivity differential worth (low "bite"). It would take several seconds for the rods to insert appreciable negative reactivity. Initially, they apparently would insert positive reactivity as the graphite rod followers displaced water from the lower region of the core.

The steam void percentage in the core was small but the core inlet water temperature was near saturation, giving the potential for rapid voiding over a substantial region in the core.

As the four circulation pumps coasted down, the lower flow allowed more steam to form in the core. Because of the strongly positive void coefficient, the increased steam content in the core initiated a power rise as the test began. The increased power further increased the steam voids. The automatically controlled regulating rods began inserting but the power continued to rise. An operator pushed the reactor scram button at 0123:40 to scram the remaining shutdown rods but these rods were near the top of the core and power continued to rise. A loud noise was heard from the reactor. Noting that the rods had not fully inserted, an operator de- energized the control rod drives hoping this would facilitate the scram. Two to three seconds later, operators heard a second loud noise. The core reactivity exceeded prompt critical, and by Soviet calculation, the power peak exceeded 100 times nominal full power. Energy generated in the fuel by the power excursion dispersed part of the fuel into minute fragments. The disintegration of the fuel stopped the chain reaction. A rapid increase in pressure resulted as disintegrated fuel at very high temperatures contacted the water. The energy release lifted the 1000-ton reactor cover plate, severing all the fuel channels. The refueling machine and its crane collapsed onto the reactor. Upper portions of the reactor building were destroyed. Hot segments from the core were ejected from the reactor, and the graphite in the reactor was ignited. Approximately 30 localized fires started, involving roofing materials and other combustibles. This accident resulted in the release of about 50 million curies of particulate and iodine radioactivity, or about three to four percent of total core inventory, evacuation of 135,000 people from the area within 30 kilometers of the plant, gross radio- active contamination of the plant site, contamination of other portions of the evacuated area, and distribution of measurable amounts of radioactivity over other countries. Approximately 200 personnel were hospitalized from radiation exposure at the site, and 31 fatalities from this group have been reported to date.

What has been gained from the Chernobyl disaster?

Leaving aside the verdict of history on its role in melting the Soviet iron curtain, some very tangible practical benefits have resulted from the Chernobyl accident . The main ones concern reactor safety, notably in eastern Europe. (The US Three Mile Island accident in 1979 had a significant effect on western reactor design and operating procedures. While that reactor was destroyed, all radioactivity was contained - as designed - and there were no deaths or injuries.).

While no-one in the West was under any illusion about the safety of early Soviet reactor designs, some lessons learned have also been applicable to western plants. Certainly the safety of all Soviet-designed reactors has improved vastly. This is due largely to the development of a culture of safety encouraged by increased collaboration between East and West, and substantial investment in improving the reactors.

Modifications have been made to overcome deficiencies in all the RBMK reactors still operating. In these, originally the nuclear chain reaction and power output would increase if cooling water were lost or turned to steam, in contrast to most Western designs. It was this effect which caused the uncontrolled power surge that led to the destruction of Chernobyl-4.

All of the RBMK reactors have now been modified by changes in the control rods, adding neutron absorbers and consequently increasing the fuel enrichment from 1.8 to 2.4% U-235, making them very much more stable at low power. Automatic shut-down mechanisms now operate faster, and other safety mechanisms have been improved. Automated inspection equipment has also been installed. A repetition of the 1986 Chernobyl accident is now virtually impossible, according to a German nuclear safety agency report.

Since 1989 over 1,000 nuclear engineers from the former Soviet Union have visited Western nuclear power plants and there have been many reciprocal visits. Over 50 twinning arrangements between East and West nuclear plants have been put in place. Most of this has been under the auspices of the World Association of Nuclear Operators, a body formed in 1989 which links 130 operators of nuclear power plants in more than 30 countries. See also Cooperation in the Nuclear Power Industry.

Many other international programs were initiated following Chernobyl. The International Atomic Energy Agency (IAEA) safety review projects for each particular type of Soviet reactor are noteworthy, bringing together operators and Western engineers to focus on safety improvements. These initiatives are backed by funding arrangements. The Nuclear Safety Assistance Coordination Centre database lists Western aid totalling almost US$1 billion for more than 700 safety-related projects in former Eastern Bloc countries. The Nuclear Safety Convention is a more recent outcome.

In 1998 an agreement with the US provided for the establishment of an international radioecology laboratory inside the exclusion zone.

The 2005 Chernobyl Forum report said that some seven million people are now receiving or eligible for benefits as "Chernobyl victims", which means that resources are not targeting the needy few percent of them. Remedying this presents daunting political problems however.

Additional Resources