Data Analysis and Physicochemical Modeling of the Radiation Accident in the Southern Urals in 1957

Moscow ATOMNAYA ENERGIYA , Jan 95 No 1, pp 46-50
by S. A. Kabakchi, A. V. Putilov (State Scientific Center of the Russian Federation Physicochemical Scientific Research Institute imeni L. Ya. Karpov) and Ye. R. Nazin (Military Academy of Chemical Protection imeni Marshal S. F. Timoshenko); UDC 621.039.7]

Major radiation accidents at enterprises involved in the nuclear fuel cycle are a unique source of information that can have a decisive influence on the development of nuclear- and radiation-safe plants and establishment of a system of safety standards. As a rule because there is so little objective information about radiation incidents and accidents, it is difficult to determine the true physical cause of a physicochemical phenomenon that developed at some stage of the production process at an uncontrollable rate or in an undesirable direction.

This article shows the possibility for revealing the causes of a radiation accident when information is insufficient by analyzing indirect data on its consequences. The radiation incident at the waste dump of a radiochemical plant in 1957 was the object of analysis. The article does not address the ecological consequences of the accident. Emphasis is laid on revealing the essence and laws governing development of the physicochemical processes that led to the accident and were the cause of dispersal of radionuclides over a large distance. As far as we know, this problem has not been discussed in the literature.

Analysis of Data on the Accident

The accident in the Southern Urals in 1957 attracted the attention of the world community after publication of an article in which it was concluded, on the basis of an analysis of open publications reporting research on the effects of radiation upon animals and plants of the Uralsk vicinity, that a massive release of radionuclides into the environment (a minimum of 50 megacuries and a radioactive contamination area of 400 to 900 km2) occurred in late 1957 or early 1958 in the vicinity of Kyshtym, Chelyabinsk Oblast. This conclusion created a sensation because Western experts (including intelligence services) knew nothing about a radiation accident of such a scale. Various suppositions regarding the causes of the accident were offered in a series of publications [2-12] following this report: radioactive fallout from thermonuclear weapon tests on Novaya Zemlya; deliberate dropping of an atomic bomb with a yield of around 20 kilotonnes (TNT) to determine the stability of the radiochemical plant and its infrastructure; a sudden release of radionuclides resulting from a self-sustaining nuclear chain reaction in storage trenches containing radioactive wastes from weapon-grade plutonium production; an accident occurring during assembly of a nuclear weapon; a steam or chemical explosion in a radioactive waste dump; the result of many successive releases of small quantities of radionuclides.

It was noted in [9] that hypotheses regarding exotic processes such as accidental (or intentional) explosion of a nuclear device or a reaction arising due to selective adsorption of plutonium by slurry in a waste trench are not necessary to explain the release. Radiochemical production wastes with a high relative radioactivity are chemically active systems that can boil as a result of intensive heat evolution. Because potentially explosive organic nitrates are used in some stages of radiochemical processing of irradiated uranium, scenarios involving chemical and steam explosions, fire, rupture of waste storage units and so on may be written. A chemical explosion in a liquid waste dump would cause formation of a large number of droplets that would then be dispersed and undergo evaporation in response to the heat of radioactive decay. Good conditions are created in this case for formation of aerosols containing radioactive uranium fission products. If a fire accompanies the explosion, the flow of heated air will raise dispersed aerosol particles to a considerable altitude.

The chemical explosion hypothesis can explain the large area of terrain that was observed to be radioactively contaminated, and it does not require any additional assumptions. Accidents of this type did happen at chemical enterprises before.

The possible physicochemical causes of the accident were discussed in [10]. An explosion in a storage dump containing highly radioactive wastes from acquisition of weapon-grade plutonium was recognized to be the most probable cause of the release and of contamination of the terrain. The authors of this paper suggested (and it is now known for sure [13]) that in the 1960s the USSR used an acetate procedure to process spent uranium (precipitation of uranyl and sodium acetate). A concentrate consisting of fission products from the first cycle of division, which had a relative radioactivity of 10-100 Ci/liter, was stored in underground containers in liquid form for 3-5 years. After its radioactivity dropped below 1 Ci/liter, it could be stored in other ways (solidification, burial in deep subterranean horizons etc.). The following physicochemical processes could have caused release of a large quantity of fission products into the atmosphere: a huge chemical explosion or fire after a micro-scale self-sustaining chain reaction causing superheating of the wastes; ignition of readily flammable solvents they contained; explosion or detonation of old nitrate-containing wastes; detonation of radiolytic hydrogen, and a steam explosion (due to generation of a large quantity of heat during radioactive decay) in a storage container at the radiochemical plant.

The authors of [10] explained the abnormally low 137Cs/90Sr radioactivity ratio, which is atypical of the highly radioactive wastes produced by the procedure for obtaining weapon-grade plutonium, by the fact that137Cs was isolated from the wastes by the ammonium alum procedure (apparently a procedure for isolating cesium using ammonium phosphomolybdate). What makes this procedure significantly different from others is presence of a sizable quantity of ammonium nitrate in the wastes--a substance capable of spontaneous explosion and detonation (catastrophic explosions have been known to occur with fertilizers containing ammonium nitrite). As a result the following scenario of events leading to the radionuclide release was composed in [10]: Due to damage to the cooling system or to other safety systems of a storage facility containing highly radioactive wastes, slurry or the solid precipitate in a container became superheated, causing an explosive reaction between nitrites and organic ingredients of the wastes (acetates or solids), or detonation of ammonium nitrate. In this case the energy release could have been large: If a 1,000-3,000 m3 container was two-thirds full of ammonium nitrate with a concentration of 2 to 8 moles/liter, the explosion would have been equivalent to 0.1-1 tonnes of TNT. Because the waste containers are stored in groups of several units, an explosion of one of them could have caused detonation of others, which could in principle have contained wastes of a different type with a different radiochemical composition. This in the opinion of the authors could explain the unusual composition of the release.

The major radiation incident that occurred in the Southern Urals on 29 September 1957 is described in detail in [11-13]. Following isolation of weapon-grade plutonium, liquid wastes were transferred at the radiochemical plant of the Mayak Scientific-Production Association into storage tanks with a volume of around 300 m3, made from stainless steel and situated in concrete-lined trenches dug slightly below the ground surface. The storage tanks were around 2 km from the plant. To prevent situations that could cause a chemical explosion in the wastes, the containers were cooled by means of heat exchangers on the interior wall of the storage facility. The design of the heat exchangers did not allow for their repair in case of damage. In 1956 the heat exchanger of one of the storage tanks was shut down due to a fault. Calculations carried out by plant specialists showed that the wastes would remain stable even in the absence of cooling (as was noted in [12], their calculations were wrong). As a result no attempt was made to set up heat removal from this container for over a year. Water began evaporating from wastes in the tank with the disconnected heat exchanger, and explosive nitrates and acetates concentrated on the interface between the wastes and the air. A chance spark from a faulty piece of monitoring and measuring equipment caused detonation of the salts. The explosion ruptured the container and ejected its contents.

In the opinion of the author of [13] the force of the explosion of the nitrate and acetate mixture in the storage facility was equivalent to 5-10 tonnes of TNT. This caused release of 70-80 tonnes of wastes containing radionuclides with a radioactivity of around 20 MCi. Approximately 90 percent of this quantity fell out near the place of the accident. The remaining 2 MCi formed a cloud 1 km high, which traveled over a significant amount of territory. It is estimated that 15,000-20,000 km2 were contaminated above 0.1 Ci/km2 within a radius of 300 km.

Modeling the Accident Process

Hypotheses regarding the causes of the radiation accident in the Southern Urals based on an analysis of indirect data were compared in the 1970s with actual data [11,12]. The comparison showed that in qualitative respects these hypotheses faithfully reflected the essence of the processes occurring during the accident. The results of modeling the explosion according to a scenario presented in [10-12] are presented below. They permitted estimation of the energy of the explosion and the height of release of the radionuclides.

Exothermic processes begin in mixtures of sodium nitrate and solid organic materials at a high temperature (360-400[DEG] [14]). This is because organic matter interacts with oxygen--a decomposition product of sodium nitrate, rather than with this compound itself. Pure sodium nitrate breaks down into nitrite and oxygen at 380[DEG]C. The sodium nitrite that forms as a result is stable. Mixtures weighing around 100 mg consisting of 30 percent sodium acetate and 70 percent sodium nitrate by weight, and 10 percent sodium acetate and 90 percent sodium nitrate (with molar ratios of 1:3.74 and 1:14.6 respectively) were subjected to derivatographic [?; derivatograficheskoye] analysis. It was shown that after removal of water (around 10 percent weight loss in the first case and 5 percent in the second) at temperatures of 385 and 370[DEG]C respectively, an intensive exothermic process (with 620 kJ/kg heat release) coupled with a weight loss of 15 percent begins in the first case, and a relatively slow exothermic process (with 83 kJ/kg heat release) coupled with a weight loss of 2.1 percent begins in the second. Considering the data of derivatographic analysis, the processes that actually occur when storage mixtures undergo heating, and the fact that boiled-down highly radioactive wastes are a stoichiometric mixture or one with excess oxidizer, the most probable equation for the reaction between the mixture's components would be:

CH3COONa+0.5H2O+xNaNO3=2CO2 +1.5H2O+0.5Na 2O +xNaNO2 
with release of 150.4 kcal (630 kJ) of heat.

Data from derivatographic analysis indicate that only part of the heat is actually released upon interaction of the mixture's components, depending on the composition of the mixture. In the first approximation the relationship of heat release Q[nu] to the composition of the mixture (the ratio x of the number of moles of sodium nitrate to the number of moles of sodium acetate) when x >==4 may be expressed as Q[nu]=822.6-50.6x kJ/kg. The quantity of gases formed per kilogram of sodium acetate and nitrate mixture is Vg=22.4 mg/MM, where mg is the number of moles of gaseous reaction products, and MM is their molecular mass. Using the ratio x for the sodium acetate and nitrate mixtures, and with regard for data from derivatographic analysis, when x>==4 we have Vg=(560- 22x)/(1+1.036x) cubic nanodecimeters per kilogram.

The TNT equivalent is TNTequiv=Q [nu]/4520; for these mixtures, TNTequiv=1.1|b10 -2(16.25-x). The equivalent charge is W=1.1|b10-2 mevap/(16.25-x), where mevap is the weight of the mixture boiled down to dry state, tonnes. According to data in [15], change in pressure upon explosion in a leak-tight volume depends on the ratio of the weight of the charge W to the volume of the chamber V. When W/V3, this relationship is described approximately by the equation Pexpl=1,300(W /V)0.755, where pressure Pexpl is measured in kPa, and W/V is measured in kg/m 3. For these mixtures and when x >==4, in the case of an explosion P expl=1300[1.1|b10-2m evap(16.25-x)/ V]0.755 kPa.

In the case of a ground burst, the maximum height to which the products are lifted is Hmaxground =W0. 2 km, where W is the weight of the TNT charge in tonnes. In our case of a partially underground explosion H maxunder=0.85Hmax ground or, expressing the energy of explosion of a sodium nitrate and acetate mixture by the TNT equivalent, H maxunder=0.85[1.1|b10-2 mevap(16.25-x )]0.2 km.

The results of calculating the parameters of explosion of several different mixtures, presented in the table, show that at the molar ratio of sodium acetate and nitrate characteristic of highly radioactive wastes and at a boiled- down salt weight of 30 tonnes, the equivalent charge is 2.4-4.1 tonnes of TNT. The power of such a charge is sufficient to generate a pressure of up to 10 MPa in the container, and to eject its contents to a height of 1 km. The power of the equivalent charge coincides in order of magnitude with estimates in [13].

Parameters of Explosion of a Mixture of Sodium Nitrate and Acetate (Weight of Boiled-Down Salts--30 tonnes) Composition of Mixture, % by weightHeat Release, kJ/kgQuantity of Gases Per 1 kg of Mixture, ndm 3TNT EquivalentExplosion Pressure, kg/cm2Height of Cloud, km NaAcNaNO3 198162092 0.137941.12 (x=4) 109038537 0.08365.41.02 (x=8.68) 4.495.6-234 No decomposition (x=21)


The research shows that a relatively simple physicochemical experiment provides a possibility for modeling an extreme reaction causing release of radionuclides from a storage facility, and to determine the energy characteristics of this reaction in the particular conditions under which it proceeds. By writing a plausible scenario of an accident at a radiochemical plant based on an analysis of indirect data and the energy characteristics of dangerous phenomena occurring as an accident develops, we can also pose and solve the reverse problem-- establishing the possible consequences of an accident before it occurs, and in the ideal case, before a dangerous facility is placed into operation. In other words the design of a radiochemical plant (of its individual parts) must consider both predictable and unpredictable accidents, as is done with nuclear power stations. Compiling a register of possible accidents at enterprises involved in the nuclear fuel cycle was attempted in [16]. Several scenarios of possible accidents were formulated and checked out in [17] for the radiochemical plant's department storing and processing highly radioactive liquid wastes and radioactive slurry.


1. Medvedev, Zh., "Two Decades of Dissidence," NEW SCIENTIST, Vol 72, 1976, p 264.

2. "Evidence on the Urals Incident," NEW SCIENTIST, p. 692.

3. Walker, F., "Urals Russian Incident," CRITICAL MASS, Vol 23, No 49, 1977, pp 12-13.

4. Medvedev, Zh., "Facts Behind the Soviet Nuclear Disaster," NEW SCIENTIST, Vol 73, 1977, p 761.

5. Corrigan, R., "Nuclear Disaster--Could Whatever Happened There Happen Here," CRITICAL MASS ENERGY PROJECT, 25 November, 1977.

6. Stratton, W., Stillman, D., Barr, S. and Agnew, H., "Are Portions of the Urals Really Contaminated," SCIENCE, Vol 206, 1979, pp 423-425.

7. Fremlin, J., "Nuclear Detective Story," NATURE, Vol 282, 1979, p 157.

8. Hill, J., "Uncovering the Evidence. Nuclear Disaster in the Urals," NEW SCIENTIST, Vol 84, 1979, p 369.

9. Postol, T., "The Incident in Chelyabinsk," SCIENCE, Vol 206, 1979, pp 326-327.

10. Trabalka, J., Eyman, L. and Auerbach, S., "Analysis of the 1957-1958 Soviet Nuclear Accident," SCIENCE, Vol 209, 1980, pp 345-353.

11. Nikipelov, B. V., Romanov, G. N., Buldakov, L. A. et al, "The Radiation Accident in the Southern Urals in 1957," ATOMNAYA ENERGIYA, Vol 67, No 2, 1989, p 74.

12. Smith, J., "Soviets Describe Accident at Secret Nuclear Center," THE WASHINGTON POST, 10 July 1989.

13. Bukharin, O., "The Threat of Nuclear Terrorism and the Physical Security of Nuclear Installations and Materials in the Former Soviet Union," Monterey Inst. of Intern. Studies, August 1992, ISBN 0-9633859-1-7, 1992.

14. Andreyev, K. K. and Belyayev, A. F., "Teoriya vzryvchatykh veshchestv" [Theory of Explosives], Moscow, Oborongiz, 1960.

15. Beyker, Ya, Koks, P., Uestayn, P. et al, "Vzryvnyye yavleniya. Otsenka i posledstviya" [Explosive Phenomena. Assessment and Consequences], Moscow, Mir, 1986.

16. Kovalevich, O. M., Slutsker, V. P., Kabakchi, S. A. et al., "Safety of Enterprises in the Nuclear Fuel Cycle and Possible Approaches to Standardizing It," ATOMNAYA ENERGIYA, Vol 76, No 4, 1994, p 264.

17. Kabakchi, S. A., Zagaynov, V. A., Lushnikov, A. A. et al., "Hypothetical Accidents in Facilities Storing Radioactive Liquid Wastes and Slurry," ATOMNAYA ENERGIYA, Vol 76, No 5, 1994, p 444.