ISSN 10637788, Physics of Atomic Nuclei, 2010, Vol. 73, No. 14, pp. 2296–2300. © Pleiades Publishing, Ltd.

, 2010.
Original Russian Text © E.I. Grishanin, 2009, published in Voprosy Atomnoi Nauki i Tekhniki. Seriya: Fizika Yadernykh Reaktorov, 2009, No. 3, pp. 20–26.

The Role of Chemical Reactions in the Chernobyl Accident

E. I. Grishanin
Russian Research Center Kurchatov Institute, pl. Kurchatova 1, Moscow, 123182 Russia
email: [email protected]
Received February 12, 2009

Abstract—It is shown that chemical reactions played an essential role in the Chernobyl accident at all of its
stages. It is important that the reactor before the explosion was at maximal xenon poisoning, and its reactivity,
apparently, was not destroyed by the explosion. The reactivity release due to decay of Xe235 on the second
day after the explosion led to a reactor power of 80–110 MW. Owing to this power, the chemical reactions of
reduction of uranium, plutonium, and other metals at a temperature of about 2000°C occurred in the core.
The yield of fission products thus sharply increased. Uranium and other metals flew down in the bottom water
communications and rooms. After reduction of the uranium and its separation from the graphite, the chain
reaction stopped, the temperature of the core decreased, and the activity yield stopped.
DOI: 10.1134/S1063778810140073

There is a vast technical literature devoted to the Runaway on prompt neutrons up to a power of
first stage of the catastrophe at unit IV of the Cherno 47000% led to evaporation of all fuel and graphite in
byl Nuclear Power Plant (NPP), that is, an uncon the core. After that, nothing essential took place in the
trolled increase in the reactor’s power [1]. There have silo of the reactor as nothing else could burn up. Then,
not been any publications on the study of the processes three concrete blocks of the destroyed box of the drum
which occurred after the explosion of the reactor. divider fell down into the empty silo, followed by the
Elena upper metal structure falling down on its edge.
Research into the condition of the accident at unit
IV by drilling holes has led to surprising results; However, this theory fails to explain many well
namely, the reactor lacks graphite stacking and fuel but known and documented facts, namely, a relative “lull”
has concrete structures which have remained from the for 20 h after the explosion when the worst seemed to
destroyed room where the drum dividers were located. be in the past; a sharp increase in the yield of radioac
Traces of the impact of high temperature are absent tivity after 20 h; active burning of graphite; a fourfold
and, in bottom rooms, there is no more than 10% of rise of radioactivity from May 1 to 6; a decrease in
the fuel in the form of lava fuelcontaining masses [2]. radioactivity release by 1 million times after May 6.
It can be easily shown that the graphite could not
In [2], to explain the facts above, an interesting ver evaporate. First, it produces only 6% of the power of
sion has been proposed. According to it, after the the core. Second, its weight is 10 times greater than
increase in reactor power to 10000%, evaporation of a that of the fuel, while its thermal capacity is three
portion of the fuel and destruction of the zirconium times greater than the heat capacity of the fuel. There
highpressure tubes took place. The authors state that fore, upon runaway of the reactor, the graphite is
the part of the core in the southeastern sector “worked hardly heated.
as nonoptimal nuclear rocket engine,” which led to For instance, if the fuel is heated by 1000°C
the damage of the support structure there and precon because of reactor runaway, the graphite will only be
ditioned the presence of the socalled lava masses, heated by 2°C; if the fuel is heated by 3000°C and
containing 10% of the uranium load of the reactor. evaporates, it will fly out of the technological channel
High vapor and gas pressure within the graphite as if the latter were the barrel of a gun, and the graphite
stacking destroyed the steel case of the reactor, which temperature rise only by 6°C. This property of the
was designed to allow for the breakage of only one highpower channeltype reactor (RBMK) is well
highpressure tube (it was able to hold an excess pres known to nuclear plant operators. A portion of the
sure of no more than 3.5 bar). Then, upon the impact graphite could fly out of the reactor owing to mechan
of pressure, the graphite stacking, together with the ical coalescence with the zirconium tube of the tech
remaining fuel, was ejected out of the underground nological channel.
concrete silo. This is the explanation of the first explo The goal of this study is to clarify the events that
sion. When it flew out, the core instantaneously lost all took place during the failure with chemical and
water, because of which a positive reactivity on the nuclear reactions taken into account. It is widely
order of 4.5 fractions of delayed neutrons appeared. accepted that an uncontrolled chain fission reaction

2296
 THE ROLE OF CHEMICAL REACTIONS IN THE CHERNOBYL ACCIDENT 2297

occurred in unit IV of the Chernobyl nuclear power should be applied at RBMKs, since an RBMK is able
plant, which was preconditioned by flaws in the reac to work on natural convection at a power of up to 40%.
tor’s structure and breach of the procedures by the The fact that the regulating bodies permitted the oper
operators [1]. What caused the catastrophe is the same ation of the unit without this system speaks for the
as the reasons for the failure at unit I of the Leningrad extent of its futility.
NPP in 1975 [3, 4]. During testing, it was planned to bring the reactor
All RBMKtype reactors have the following inher to the power of 200 MW, which is approximately
ent feature: there is a spatial instability of the neutron equivalent to the power of the residual heat increase,
field owing to the large size of the reactor compared to then deenergize half of the MCPs, and test their elec
the migration length and the presence of a positive tricity supply by means of the turbine generator’s run
feedback with respect to the power (via the steam coef down.
ficient of reactivity). Many critical masses exist in the Prior to the tests, the reactor was strongly poisoned
core of a large reactor and the connection among them with xenon, which, according to our estimates, tempo
is weak; the RBMK is large both in diameter and in rarily decreased the reactivity margin by 2.9% and sig
height. The effect of the neutron field’s instability was nificantly increased the steam effect of reactivity. This
not taken into account when unit I of the Leningrad is of fundamental importance. Such a test conducted
NPP was designed. These spatial fluctuations of the under the same conditions will result in a similar
neutron flux had to be neutralized by means of moving catastrophe. In the process of chemical reactions, an
control rods manually by the operator. additional release of positive reactivity took place.
It is no surprise that the moment came when the Before the explosion, the temperature of the
operator could not handle the control rods. As a result, graphite stacking was approximately 500°C. The con
in December 1975, more than 42 fuel assemblies over sumption of coolant exceeded the nominal value
heated. The temperature of the claddings of fuel ele owing to the switchingon of the reserve MCP. For the
ments contained in the fuel assemblies hit 1600°C. large loss of reactivity under conditions of the xenon
Only two highpressure casings were destroyed, how poisoning, in order to bring the reactor to the required
ever; otherwise, the effects of this accident could have power, the operators had to withdraw from the core a
been like those at unit IV of the Chernobyl NPP. The substantially larger number of control rods than the
failure at the Leningrad NPP and the Chernobyl procedure allowed. As a result, some sites in the core
nuclear accident have the same nature and differ only were not regulated by an LAR. Therefore, from the
in the scale of the consequences, which, in turn, were viewpoint of the neutron flux instability, the situation
due to different initial conditions (mainly, xenon poi at unit IV was close to the conditions at unit I of the
soning). Leningrad NPP in 1975. Seemingly, it was the south
Under the decision of a government commission eastern sector where the majority of control rods were
which examined the causes of this accident, a set of removed. Hence, the initial stage of the development
technical solutions increasing the stability of the spa of the accident at unit IV was virtually the same as at
tial power distribution was developed. This set the Leningrad NPP in 1975.
included a system of local power governors (LAR). Some channels overheated owing to the uncon
Every relatively large spot in the core had its own auto trolled increase in the neutron flux in the absence of
matic power governor. The system supported at a sat local governors. Compared with the accident at the
isfactory level the required power distribution, i.e., the Leningrad NPP, the situation was further aggravated
ratio of powers between nearly independent areas of by the fact that the reactor was subject to xenon poi
the core. It is for this reason that the large reactor had soning, where the steam effect of reactivity became
the procedural limitation on the minimum number of significantly higher, as well as that the button for drop
control rods, which was violated by the operators dur ping all control rods was pushed (in this case, it pro
ing tests. duced a positive reactivity at the start of the motion).
On April 16, 1986, unit IV of the Chernobyl NPP As a result, the number of overheated channels was
was subjected to scheduled tests of a standard security several times higher than in the accident in 1975. The
system before it was put into operation. Analogous power increase was accompanied by boiling of the
tests were conducted at other RBMK units. This sys coolant.
tem was not timely tested and it was commissioned at When the crisis of heat exchange arose and the
the startup of unit IV in Chernobyl. It relies on the temperature of the zirconium claddings of fuel ele
rundown of the turbine generator for electricity supply ments in part of the core reached 1000°C, hydrogen
to the main circulation pumps (MCPs) upon their de started to form through the steam–zirconium reaction
energizing. Zr + H2O = ZrO2 + H2, producing 1000 normal cubic
The system was developed for reactors of the meters. This led to an acute increase in hydraulic resis
VVER440 type, which have sealed pumps with a very tance in the circulation loop of the coolant and to a
low rundown value; this was the reason for the danger boost in the reactivity for the reason of displacement
ous overheating of fuel elements upon complete zero of water out of the core. As a result, in the fourth sec
ing of the MCPs. It is doubtful whether such a system ond after the actuation of the emergency protection

PHYSICS OF ATOMIC NUCLEI Vol. 73 No. 14 2010

2298 GRISHANIN

rods (AZ5), MCPs with working electric drives power and the heating of the core before April 30;
decreased the consumption twofold in accordance (3) destruction of the claddings of fuel elements, con
with their force–consumption parameter, which is a figuration anew of the core accompanied by a further
reliably documented fact. At the same time, MCPs increase in reactivity and temperature of the core, and
working in rundown stopped immediately, which is an increase in the release of fission products starting
also documented by way of unreliable readings of their on April 30; and (4) destruction of the zirconium high
sensors of consumption (backflow of water through pressure tubes and direct contact of the graphite with
MCPs). Two seconds later, all MCPs resumed their fuel pellets from May 4 through 6.
consumption, which should be unambiguously inter During the first hours after the explosion of the
preted as multiple destruction of zirconium tubes of reactor, it seemed that the worst things had already
the technological channels, bearing the pressure of the happened and the reactor would cool since the resid
coolant, since in this case, the hydraulic resistance in ual heat output was falling rapidly. From the air, the
MCPs sharply fell. glowing core was seen at night and steam coming out
The manifold destruction of highpressure tubes is of the destroyed reactor was visible in the daytime.
preconditioned by that fact that, with the further At the same time, there was the following reaction in
increase in the reactor’s power, the temperature of the the graphite stacking: 2C + O2 = 2CO. Other reactions
superheated steam exceeded 1000°C and the tubes with the formation of CO2 were insignificant because
made of the zirconium alloy lost their firmness. This of the excess of graphite or lack of oxygen. At the time,
happened in the southeastern sector of the reactor. At the initial temperature of the graphite was on the order
the multiple nearly simultaneous rupture of the high of 500°C, at which the rate of the reaction was low.
pressure channels, a powerful hydraulic shock took A small amount of CO mixed with the cold air in the
place to deform the reactor’s support and drop it 4 m space over the reactor and had a relatively low temper
lower. Such a shock could have caused corresponding ature. Therefore, there was no fire over the reactor
seismic disturbances and underlies the hypotheses during the first day.
about the seismic reasons for the Chernobyl catastro The second stage began on the second day after the
phe. explosion. The reactor started to attain power as a
After the destruction of many highpressure chan result of the decay of iodine and xenon135 (termina
nels, steam and hydrogen with a temperature of tion of xenon poisoning). The reactivity excess was
1000°C begun to penetrate the graphite stacking, compensated by the corresponding increase in the
where the endothermic chemical reaction H2O + C = temperature. The release of radioactivity rose sharply.
CO + H2 occurred. The role of other chemical reac Because of the heatup of the reactor, the rate of the
tions was insignificant. At the same time, the reactor chemical reaction 2C + O2 = 2CO increased sharply
worked as giant gas generator with a nuclear source of and a fire burst out over the reactor, in which the car
heat. When the excess pressure in the graphite stacking bon monoxide was burning. Approximately on
reached 3.5 kg/cm2, the vessel of the reactor was April 29 after practically complete decay of iodine and
destroyed. This was the first steam explosion which xenon, the maximum reactivity Δk = 2.9% was
overturned the upper structure of the reactor and released, and the reactor reached the maximum tem
ejected part of the fuel and graphite. perature, which can be assessed as follows.
Hydrogen and carbon monoxide (CO) came from If we assume that the Doppler coefficient is equal
the reactor into the Central Hall in an amount of over approximately to dk/dt = 2.9 × 10–5 K–1, then the tem
10000 m3; they mixed with the air to produce an perature of the fuel averaged over the volume of the
explosive mixture. The explosion of the mixture was core should increase by the value of ΔT = Δk/(dk/dt) =
initiated by pieces of graphite which had a temperature 0.029/(2.9 × 10–5) = 1000°C. Owing to extensive nat
on the order of 500°C. This was the second massive ural air circulation, the bottom of the reactor had a rel
explosion which provided the appropriate nature of atively low temperature on the order of 30°C. Hence,
the destruction of the reactor building and caused the the maximum temperature of the fuel must have been
presence of soot at the moment of the CO explosion 2000°C in order to produce the average fuel tempera
and the dark red color of the fire. Thus, the following ture of 1000°C. The release of positive reactivity after
chemical reactions played an important part in the the termination of xenon poisoning determined the
explosion of the reactor of unit IV: the steam–zirco condition of the failed reactor, which was called later
nium reaction in the fuel channels and the formation the “active phase.”
of CO in the graphite stacking. The level of power established after the heatup of
Only part of the core exploded, in which control the reactor was determined by the excess reactivity of
was lost because of the removal of the control rods in 2.9% and the consumption of air via the destroyed
violation of the procedure. The remainder was in a communications of the reactor by means of natural
subcritical condition. convection. Air circulation in the ruptured reactor
Further events during the accident can be conven explains many of the processes that occurred. It took
tionally divided into the following stages: (1) the first place at every site where heat was extracted. It is
24 h after the explosion; (2) the reactor’s attainment of important to note three parallel profiles of air circula

PHYSICS OF ATOMIC NUCLEI Vol. 73 No. 14 2010

 THE ROLE OF CHEMICAL REACTIONS IN THE CHERNOBYL ACCIDENT 2299

tion. The first of them includes sections of channels in greater extent. A chaotic strewing of the fuel pellets
the southeastern sector, whose bottom was destroyed, and, probably, of a certain amount of fragments of the
and distributing headers and other channels destroyed oxidized zirconium channels took place in the upper
in the ceiling. The cold air arrived via the destroyed part of the active zone.
channels of the southeastern sector to distributing Such a change, which occurred from about April 30
headers and channels ruptured in the ceiling. The air to May 1, led to immediate contact of UO2 pellets with
cooled the fuel. Air consumption in this profile is esti graphite, and intense reduction of uranium oxide and
mated at G = 40 kg/s, and its thermal capacity is C = other oxides of fission products began, resulting in a
1.0 kJ/(kg K). fourfold increase in emission of radioactivity.
With such air consumption, the power of the reac In essence, the reactor worked as if it were a nonopti
tor is calculated as N = GCΔT = 40 × 1.0 × 2000 = mal blast furnace for the smelting of metal uranium,
80000 kW. At the average temperature of the fuel, its plutonium, and fission products from their oxides. As
maximum value must equal 2000°C. The difference in the reduced metal uranium continued to flow down,
temperature between the fuel and the air coolant is the hightemperature zone dropped down. The pro
small owing to low power. Therefore, the heatup of cess of reduction was like the burning of a candle.
the air in the channels with the fuel amounts to Upon immediate contact, the reduction reactions
approximately 2000°C. Heat exchange in the active took the form C + UO2 = U + 2 CO2 and MeO + C =
zone due to endothermic reactions of metal reduction Me + CO2. We made calculations of the change in iso
is estimated at 3 MW. The increase or decrease in heat barisothermal potential (Gibbs’ energy) at atmo
exchange via chemical reactions will lead to the spheric pressure and various temperatures in order to
increase or decrease in power, respectively. assess the fundamental feasibility of the chemical pro
The second loop of natural air circulation is formed cess of uranium reduction by graphite. There, we used
by the gap between the casing of the reactor and the the methodology and initial data on change in
graphite stacking and the gaps in the graphite stacking. enthalpy and entropy of chemical reactions taken
Cold air arrives via the gap between the casing and from [5, 6]. These estimates showed that the reaction
graphite stacking, then in the gaps in the graphite of uranium reduction with graphite was possible at a
stacking, and cools the latter and leaves via the upper temperature above 500°C in the absence of oxygen.
part of the destroyed reactor to the space above. The The absence of oxygen was ensured by intense oxida
third loop of air circulation in the premises beneath tion of graphite at high temperature.
the reactor cools the tubes feeding water to the reac Reduced uranium (melting point of 1100°C) flew
tor’s channels in the normal mode. There, the air down to the bottom steel tubes, through which the
starts to circulate extensively when melted uranium coolant was fed and which remained in good condition
begins to reach these tubes. since they were cooled by air. The metal uranium
Air, entering the channels with the fuel, actively hardened in these tubes, and at the same time, ura
oxidizes, at high temperature, the zirconium claddings nium gradually left the graphite stacking. When the
according to the following chemical reaction: Zr + O2 = process of reduction was over, the fission chain reac
ZrO2. As a result, the claddings of fuel elements were tion ended since the uranium was isolated from the
destroyed. graphite. The graphite burned out, and the release of
The third stage of the process of the accident com radioactivity decreased first by a 1000 times and then
menced after the complete destruction and shedding by a million times, which occurred on May 6. The
of the fuel element claddings in the channels. Accord metal plutonium sublimated at high temperature;
ing to calculations, removal of the zirconium clad therefore, its concentration was low in the uranium.
dings of fuel elements from the core leads to approxi In the southeastern sector of the reactor, the first
mately a 1.2% increase in reactivity. The average tem explosion damaged the support of the reactor and
perature of the fuel increases by approximately 350°C coolantfeeding communications. Therefore, the
and the reactor’s power increases to 110 MW. As a reduced uranium flew down to the bottom premises as
result, the air temperature in the outlet of the reactor well, where it reacted with the concrete at a tempera
and the fuel temperature in the core ceiling increased ture of 1100°C, forming socalled fuelcontaining lava
to 2700°C. masses. Their amount tallied well with the estimate of
At temperatures so high, the zirconium tubes of the the superheated portion of the core obtained in [1].
reactor channels started to oxidize and be destroyed The estimate of the time of the process after the explo
rapidly. In the upper part, they could even have sion nearly coincides with the real duration of the pro
melted. This was the beginning of the fourth stage of cess. One should look for missing uranium in the hard
the accident process. Complete oxidation, shedding, ware of the water feed of the reactor, which was not
and melting of these tubes could have led to the drilled. If metal uranium is discovered in the hardware,
removal of a large amount of zirconium from the core this fact will confirm our theory.
and the appearance because of that of additional reac Precisely because of chemical reactions owing to
tivity on the order of 1.5%. Therefore, the power of the the heat produced by the fission chain reaction, most
reactor and its temperature could have risen to an even of the volatile fission products, i.e., gases, iodine,

PHYSICS OF ATOMIC NUCLEI Vol. 73 No. 14 2010

2300 GRISHANIN

cesium, strontium, and plutonium, flew out of the Metal uranium is present in the bottom hardware of
reactor and spread over the world. Only fission prod the water supply to the reactor, and metal plutonium
ucts with a high boiling point remained in the sublimated, contaminating the 30km zone around
destroyed reactor. That is why the release of radioac Chernobyl.
tivity decreased by a million times after May 6. High In the process of development of the accident, the
heat fission products primarily remained under the overwhelming majority of fission products with a boil
sarcophagus. ing point lower than 2000°C left the reactor. Only
There ought to be a natural explanation of why highheat fission products remained in the sarcopha
there are practically no traces of the hightemperature gus. Therefore, the contents of the sarcophagus do not
impact in the reactor and how concrete blocks with present any hazard from the viewpoint of a significant
traces of paint ended up there. The concrete silo has a release of radioactivity.
gap over 0.5 m wide between the concrete and the The author hopes that the publication of this paper
graphite stacking. After the explosion, cold air arrived will inspire the interest of experts in a more careful
via the gap to further penetrate the graphite via cracks study of the role of chemical reactions in the Cherno
and channels in the graphite stacking; in the graphite byl accident.
layer, the chemical reactions mentioned above
occurred. The mixture of nitrogen and CO formed as
a result of these reactions at a temperature of approxi ACKNOWLEDGMENTS
mately 450°C came out into the space above the reac The author is grateful to E.V. Burlakov for valuable
tor, where it mixed with air and formed a fire torch, comments. Special gratitude goes to Candidate of
which was visible to everyone. Chemical Sciences V.K. Popov for calculations justify
ing the possibility of the chemical reactions men
tioned in this paper.
CONCLUSIONS
A new version of initial events which led to the
explosion of the reactor of unit IV at the Chernobyl REFERENCES
NPP is proposed based on the instability of the neu 1. E. O. Adamov, Yu. M. Cherkashev, L. N. Podlazov,
tron flux in the largesized reactor. The character of et al., in Proceedings of the Intern. Seminar on Chernobyl
destruction of the reactor hall speaks for the presence Lessons, Technical Aspects, Desnogorsk, Smolenskaya
of a voluminous explosion of the mixture of air, hydro AES, 15–19 Apr., 1996, pp. 19–35.
gen, and carbon monoxide. 2. Yu. M. Cherkashev, O. Yu. Novosel’skii, N. I. Zhukov,
The decay of xenon and iodine resulted in resump et al., in Proceedings of the Intern. Seminar on Chernobyl
tion of the fission chain reaction, the heat of which was Lessons, Technical Aspects, Desnogorsk, Smolenskaya
AES, 15–19 Apr., 1996, pp. 224–251.
spent to reduce uranium oxide and oxides of pluto
nium and fission products. The attainment of the 3. A. N. Kuz’min, in History of Atomic Energetics of Soviet
Union and Russia, No. 3 (IzdAT, Moscow, 2002),
power of about 80 MW by the failed reactor, with a fur pp. 156–170 [in Russian].
ther increase to 110 MW, caused the extremely grave
4. E. B. Burlakova, V. M. Kuznetsov, V. A. Moskalenko,
radiation effects a full day after the explosion. et al., Unknown Chernobyl: History, Events, Facts, Les
Chemical reactions played a crucial part in the pro sons (MNEPU, Moscow, 2006), pp. 331–336 [in Rus
cess of the catastrophe at unit IV of the Chernobyl sian].
NPP at the stage prior to the explosion, during the 5. M. Kh. Karapet’yants, Introduction to Theory of Chem
explosion, and after it; taking them into consideration ical Processes (Vysshaya Shkola, Moscow, 1981) [in
can explain important details. At the final stage, the Russian].
reactor worked as a nonoptimal blast furnace, where 6. V. A. Ryabin, N. A. Ostroumov, and T. F. Svit, Thermo
oxides of uranium, plutonium, and fission products dynamical Properties of Substances, The Handbook
were reduced to metals. (Khimiya, Leningrad, 1977) [in Russian].

PHYSICS OF ATOMIC NUCLEI Vol. 73 No. 14 2010