The Symphony of Catastrophes

Right аafter the tragedy, specialists from the Siberian Branch of the Russian Academy of Sciences (SB RAS) set out for the dam. The first hypothesis about the reasons for the accident was that of a hydraulic impact that disturbed the turbine. The geophysicist s suggested using the data recorded at the nearby Cheriemushki seismic station of the SB RAS Geophysical Service. Interpretation of the records showed no hydraulic impact but revealed small vibrations coming from the station 2.7 seconds before the accident. Investigation into the reasons that caused the accident established that the arrival of the first impulse coincided in time with the breaking-off of the first nuts on the hydroelectric set cover. In 2.7 seconds the cover broke loose.

During further studies, as the geophysicists were analyzing the events that had taken place at the dam, they came to the conclusion that the system used for the observation over this kind of facilities was inappropriate. At present, hydroelectric sets are surrounded by a large number of meters that measure either the displacement of the set or displacement derivatives and record the maximal change in a certain time interval.

A hydroelectric set is a complicated system of vibrations having many sources. The vibrations can be generated by the turbine fins, which vibrate under the flow of the water; by the guide blades, which re-direct the flow coming to the turbine; and by the vortex core, arising in the whirlpool of the water turning in the turbine. Also, there can be other sources, which should be multiplied by the number of sets.

Observing a running hydroelectric set with the help of such meters is the same as recording a symphony played by an orchestra using a sound level meter instead of a microphone.

An orchestra generates a great many vibrations varying in frequency and intensity. Having listened to a fragment of an orchestral performance, a good musician can note down what has been played and on which instruments. Each instrument sounds at its own frequency, and the aggregate sound picture is recoded by notes.

This vision is easy to translate in the language of physics: A spectral analysis of vibrations is carried out; some “notes” (frequencies) are selected for further study. After the spectral analysis has been conducted and different sources of vibrations have been established, the seismic wave recording can help to reveal the contributors to the general vibration picture: a particular source of seismic waves, its frequency, intensity, direction, and period of activity.

And though these vibrations are not within the audio range, they look the same as the sound vibrations, which means that they also can be expanded in terms of frequency and their different spectral components can be examined. Specialists of the SB RAS Geophysical Service have managed to develop a seismic record-based technology for analyzing various modes of hydroelectric set operation, controlling the operation, and creating a “black box” system.

When, after the accident, the sixth set was being installed at the Sayano-Shushenskaya station, geophysicists mounted twelve seismic stations in the main generator hall and in the body of the dam and studied what the set radiated at different operation modes. Also, seismic vibrations radiated during the sixth hydroelectric set tests were recorded at the Cheriemushki seismic station. As a result of these tests, a more reliable interpretation of the recording made at the seismic station prior to and at the time of the accident was obtained.

It should be noted that even when the Sayano-Shushenskaya dam was constructed and its turbines were being installed, it became clear that it would be impossible to make them work at their design capacity. At low energy production, there were virtually no vibrations but as production was increasing, vibrations were becoming much stronger. After that, the production decreased, and at 500 megawatt the plant began to operate properly. When the production grew to 600 megawatt, vibrations increased dramatically, and so the decision was made to avoid this situation. The nature of this phenomenon is yet to be established.

The new technique gives us hope that scientists will be able to find the reason behind the strong vibrations – if the vibrations can be classified in terms of their sources, their force and the reason why they were generated can be established. At present, specialists have learned to “listen to” low-frequency vibrations of a working hydroelectric set .

What conclusion should be made from the accident? There are two ways to solve this problem. The first one is continuous control. If vibration exceeds certain values, the operation of the machinery should be stopped and reasons should be investigated. The second way is to do research and find out why such vibrations occur and how they can be eliminated. In principle, the technology proposed allows doing it.

Now, the Sayano-Shushenskaya station is being restored to its state before the accident., which means that the reasons for the accident will still be there. We suspect that in certain operation modes noticeable acoustic waves are radiated, resulting in strong vibrations of the hydroelectric set and water flushing pipe, and these vibrations coincide in frequency with the eigen-frequencies of some elements of the set. Such coincidence of radiated and eigen frequencies cannot be tolerated. This rule should be spelled out in the laws regulating industry safety.

References
Kurzin V. B., Seleznev V. S. On the mechanism of the occurrence of strong vibrations in the turbines of the Sayano Shushenskaya hydroelectric power station // Prikladnaia fizika i tekhnicheskaya fizika, 2010. No. 4, V.5, P. 166—175.
Lobanovskiy Yu. I. Technical reasons for the accident at the Sayano-Shushenskaya hydroelectric power station (investigation results) // www.synerjetics.ru