(Readers may have noticed that I have not started quoting eyewitnesses or inserting commentary sub-sections yet. General background information continues below, before the clock starts ticking with Part 1: 25 April 1986)
A few final words to set the scene. On 25 April 1986, Unit 4 was slated for planned maintenance. Having reached steady-state operation, the reactor had figuratively reached maturity, where it would maintain similar core parameters and fuel burn-up for the remainder of its long and promising career. Meanwhile, Pripyat was preparing for the May holidays (International Workers Day, followed by the Eighth of May, or V-E Day). Some of the shift personnel, such as Senior Reactor Control Engineer (SIUR) Leonid Toptunov, were looking forward to vacations.
As plant director Viktor Bryukhanov would insist in one of his post-prison interviews, nothing out of the ordinary was afoot, other than the usual battery of equipment checks and tests that accompanied any planned shutdown. In fact only one test was particularly noteworthy, scheduled to take place the morning after the shutdown. Several specialists from the Kurchatov Institute had flown in to carry out a delicate experiment on air-cooling the reactor as an additional safety feature.
Two more tests were scheduled leading up to the shutdown: a 'live fire' of the main steam discharge valves (ГПК), and an assessment of a new gas mixture for cooling the graphite stack. Both of these tests had to be cancelled when reactor power dropped unexpectedly just after midnight on April 26th.
Then there was the most important test to be carried out, on the No. 8 turbogenerator (turbine). This was not the infamous test that readers were no doubt expecting, but rather a recording of the vibrations produced as the turbine coasted down. The results would be used to fix a design defect which led to dangerous oscillations and leaking oil. Three employees of the Kharkov Turbine Factory parked a unique mobile laboratory (manufactured by Mercedes-Benz) near the turbine and deployed their equipment. This test, with its direct relevance to the safety of the plant, is almost entirely ignored in English sources, but potentially played a key role in the lead-up to the accident. While the discharge valve and gas mixture tests mentioned previously were cancelled because they required a higher reactor power, the vibration test needed low power, putting it on a collision course with the ill-famed 'rundown' test on the 8th turbine.
Rundown
Anyone who has read this far is no doubt familiar with the concept of the turbine rundown test: using the inertia of the spinning turbine to provide coolant during a black-out. However, this common explanation somewhat misrepresents the nature of the test. In actuality the RBMK was already equipped with a 'rundown' capability that would bridge the gap to diesel generator startup. The eight main circulating pumps (MCPs) were themselves fitted with counterweights designed to increase their inertia and continue pumping for several minutes even when electrical power was lost. In other words, the RBMK was already protected from core damage if off-site power was lost.
Rather, the reactor's designers had an entirely different scenario in mind when they concocted the turbine rundown scheme: a Maximum Design-basis Accident. This referred to the rupture of a large-diameter coolant pipe leading to the reactor, resulting in a sharp drop in water supply. In such an event the Emergency Core Cooling System (ECCS) was meant to save the reactor by injecting cold, pressurized water into the core. However, the ECCS required assistance from the feedwater pumps during this process. What would happen if the station suffered a blackout at the same time as a Maximum Design-basis Accident? This turn of events may sound unlikely, but no doubt the designers anticipated that the destructive failure of a 600mm coolant pipe could knock out power to the building at the same time. Therefore, the intention was for the inertia of the coasting turbine to produce a few minutes of electricity, enough to run the feedwater pumps and perhaps support the main circulating pumps as well.
The rundown scheme only applied to 'second generation' RBMKs, such as ChNPP Units 3 and 4. Chernobyl's managers have variously been presented as shirking their duty to perform the test in 1983, or as recklessly staging an unauthorized experiment on their own initiative. The former admonishment can only be aimed at the reactor designers, who never actually designed any such capability as 'turbine rundown'. This safety feature existed only on paper, and rather scanty paper at that: a single reference in the generic operating regulations of the RBMK, and another mention in a textbook written for general audiences. In practice, extensive alterations to the electrical system and the turbine itself were required to make rundown a reality. Every other RBMK plant simply re-wrote their regulations to omit references to this capability rather than embark on a major engineering project, no doubt sensing that NIKIET really had little interest in the idea. But this was not good enough the flagship plant of the Energy Ministry.
There is evidence that NIKIET initially had some involvement in the belated design of a rundown capability for Chernobyl's reactors. But the design bureau soon washed its hands of the idea, leaving the plant's managers to carry on alone. Prodded by compliance audits that highlighted the missing test results, Chernobyl's Electrical Division spearheaded several attempts to demonstrate the concept. These ended in failure, so the plant eventually contracted another state enterprise (DonTekhEnergo) to install a specialized 'rundown block' on the 7th and 8th turbines. A test procedure (program) was written by Grigori Metlenko of DonTekhEnergo in consultation with Anatoliy Dyatlov. The program was never cleared with external regulatory bodies, although in the permissive atmosphere of the time this was apparently not required. In November 1985, a rundown test was carried out on Unit 4 using a program almost identical to that of April 26, 1986. As far as the personnel were aware, the rundown procedure succeeded. However due to either human error or a hardware failure, the oscilloscope failed to record any results. DonTekhEnergo was then scheduled to assist the plant once again, during the planned shutdown the next year.
Source for information in this section:
 |
| Photo credit: chidori@emptytriangle.com |
Comments
Post a Comment