SOVIET ATOMIC ENERGY VOL. 57, NO. 1

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP10-02196R000300050001-6
• Release Decision: RIFPUB
• Original Classification: K
• Document Page Count: 84
• Document Creation Date: January 4, 2017
• Document Release Date: February 22, 2013
• Sequence Number: 1
• Publication Date: July 1, 1984
• Content Type: REPORT

Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-n ~,""~
0,001
> 0,001
> 0,001
> 0,001
> 0,(li)1
4U
> 0,001
> 0,001
> (1,001
> 0,001
> 0,001
20?
0,02
0,13
0,3
0,4
0,6
15
0,04
0,05
0,06
0,07
0,07
25
0,011
0,011
0,007
0,003
0,002
40
> 0,001
0,002
0,006
0,012
0,02 ,
50
0,003
0,05
O;OG
0,08
0,09
2,8
1,9
1,4
1,0
0,5
0,06
30
0,04
0,03
0,026
0,016
0,00:.
3U
> 0,001
> 0,001
> (1,0111
> U,001
> O,W1
14
0,1
0,7
1,3
2,1
2,6
4,2
0,7
1,1
1,5
1,9
2,2
20
0,6
0,8
1,1
1,5
i,T
20
0,7
1,1
1,5
1,9
2,3.
20
0,04
O,OG~
0,09
0,1
0,1
~;Nu
1
0,9
(1,9
0,9
0,9
(1,9
Tonal une
limi-
2,5
2,7
3,`2
4,02
4,8
Hated syst
emat-
ic error
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
Conclusions. The indeterminacy in the yield of fission products of plutonium isotopes
makes the main contribution to the total measurement error of depletion. Therefore the error
increases as the fraction of plutonium, especially its heavy isotopes, in the original fuel
increases. The contribution of the random component to the total error is inappreciable, and
there is no need to carry out a large number of parallel measurements. It is advisable to
make two measurements to eliminate the coarse error. The depletion can be determined with an
error no worse than 5~ when the total 145Nd and 146Nd content is selected as the depletion.
monitor, depending on the composition of the mixed uranium-plutonium fuel.
1.
Yu. B. Novikov, At. Energ., 43, No. 4, 240 (1977).
2.
W. Maeck, in: Proc. Panel Meeting on Fission Product Nuclear Data, JCP-1040, Bologna,
Nov. 26-30, 1973.
3.
V.
Ya.
Gabeskiriya et al., At. Energ ., 44, No. 5, 446 (1978).
4.
V.
Ya.
Gabeskiriya et al., Preprint NINAR P-24 (290), Dimitrovgrad (1976).
5.
V.
M.
Gryazev et al., Preprint NiNAR P-25 (359), Dimitrovgrad (1978).
6.
V.
S.
Prokopenko et al.., At. Snerg., 45, No. 3, 230 (1978).
7._
E.
Crouch, AERE-R 7785, Harwell, Oxfordshire (1975).
NUCLEAR-PHYSICAL INVESTIGATION OF MASS TRANSPORT OF CARBON
AND NITROGEN BY SODIUM COOLANT
0.
V.
Starkov, I. V. Istomin,
V.
A.
Karabash, M. Kh. Kononyuk,
A.
N.
Sosnin, and V. S. Shorin
The durability of structural materials (steels) in liquid sodium, used as a coolant, re-
mains an urgent problem for fast reactors. In order to make a correct prediction of the be-
havior of steels and the change in their mechanical properties during the operation of power-
equipment for (2-3)?105 hone must have an understanding of the processes of mass transport and
corrosion. Laboratory investigations [1-5] showed that the transport of the main elements -
(Fe, Cr, Ni) does not appreciably limit the operating life of steels in liquid sodium of re-
actor purity. The corrosion rate is low, amounting to less than 5 Um/yr for austenitic chrb=
mium-nickel steels at 700?C. The principal corrosion effect is the transport of interstitial
impurities, carbon and nitrogen. The transport rate devends in a fairly involved way on the
temperature as well as on the composition of the steels, the heat treatment conditions, and
geometric and other factors. Processes of absorption of the carbon and nitrogen by-the mate=~'
rial (austenitic steels) are possible and so are processes of denitration and decarburi,zatioii`~
(for unstabilized lOKh2M pearlitic steel); all of-this has a different effect on the mechanic
cal properties of the steels.
Transport of carbon through liquid sodium between steels of different composition can be
described by the diffusion equation. In order to solve the equation it is necessary to know
the boundary conditions, in particular the relation between the areas of the surface of the
steel that is the source of carbon (Ss) and the steel that is the receiver (Sr) of the carbon
-and the value of the thermodynamic activity of the carbon. The solution of.the'diffusion
equation shows that the thermodynamic activity is proportional to the surface concentration
Co of carbon. Thus, in order to study the thermodynamic properties of complex systems one
must have experimental data about the functions C(x) of the distribution of the gaseous im-
purities over depth in the range >50-100 um, which is characteristic of the width of the cor-
rosion zone. Existing methods of chemical analysis to determine the nitrogen and carbon con-
tent with a layered mechanism of removing material of-the specimen have a poor depth resolu-
tion (fix > 50 um), which does not allow Co to be measured with the necessary accuracy for..
Translated from Atomnaya Energiya, Vol. 57, No. l,. pp. 10-14,-July, 1984: Original anti-
cle submitted November 4, 1983.
0038-531X/84/5701- 0432$08.50 ? 1985 Plenum Publishing Corporation
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
E
c .50
o.
.~
v
d
H
4
Energy, MeV
0,5 1 1,5
600 650
Channel No.
Fig. 1. Differential cross section of the 'ZC{d, po) reaction
for the angle 9L = 165? (a) and amplitude spectrum o.f. protons in
the region of carbon impurities in a lOKh2M steel specimen for
E~ = 1.76 MeV (b). The arrow points to the peak of surface con-
taminants.
In the work reported here the nuclear-physical method of microanalysis from the instan-.
taneous radiation was used to study the distribution of nitrogen and carbon in specimens of
unstabilized lOKh2M pearlitic steel in contact with liquid sodium i-nder different conditions.
The method is based on the spectrometry of protons from the (d. p?) reactions which occur on
impurity nuclei when the surface of-the specimen is probed with a deuteron beam having an en-
ergy of 1-2 MeV [6]. The protons were. detected by two semiconductor detectors of the DKPs
type, set up at angles of 150 and 165? to the beam in a scattering chamber connected to the
ion guide of an EG-2.5 electrostatic accelerator. The electronics of the experiment permitted
simultaneous accumulation of. amplitude spectra from two detectors, one of -which was adjusted.
to detect the carbon impurity and the other, to detect nitrogen. The details of the method and
the experimental technique were described in [7, 8J. The method has a depth resolution 0x
1 ?m at a probing depth Ro < 10 ?m. In order to obtain information at a greater depth the
method was employed in conjunction with the sectioning technique.
In order to simplify the analytical procedure, we assumed that the desired distribution
C(x) in the region x < Ro has the form
C (x) = C? + (C arr.) S (x), (1)
where the distribution function of'surface contaminants in the specimen is approximated by a
b function and C is the average impurity concentration in a surface film of average thick-
Hess a. .The ass~imption (1) permits the method to be made a rapid method since the procedure
of analyzing the measured spectra car. be replaced by one of comparing the reaction yields
Y(E1) for several values of the incident deuteron energy E,. The yield Y(E,)
?tF'~~
y(~~)'-A J a(x)C(x)dx=ArsnrT~(~,)5~~~fC~~-i L'~,najF,)l, (2)
where the functions ~(Ei) and z(E1) have the form
r
n
433
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
Energy, MeV
1,5
240 260
Channel No.
Fig. 2. Differential cross section of the 14N(d, po) reaction fir
the angle 6L = 150? [a) points at top] and amplitude spectrum of
protons in the region of nitrogen impurities in a lOKh2M steel
specimen for E1 = 2.02 MeV [b) points at bottom].
TABLE 1. Content of Carbon (x102) and Nitrogen (x103) in Surface Layers of lOKh2M
Steel after Tests, mass
~, I ~ carbon re-
~~; ~'"~ h ceiver
5S/
ISr
n I n,nr, I o, 1 I o- o, i t
u I n, or. ~ u,i
40(1
5,1i?1(1+
lOKh18N10
1:1
1a(If1(1
12,5?0,1i
10,80,3
12f1
1(HIt10
1.i?1
181(1,7
;it(1
1111
lOKh18N10T
1 :2(N1
0,9i1?11,11
2,88~0,'L(i
3,41it11,1(1
2,~i?11,5
1,5?U,1ri
5,0a?11,42
7,8c~1,~i
Ci:>f)
5.111
QKh13
1 : 1
2 (It(I,:i2
/i,a11~11,1~1
/+,fi9fll,a/
fit1
2,115111,3
1,11111,3
8,A8~-I);:~L
(i,d1
;?1u=
~
U1KCh13
1:1
1,i)1fu,U7
i,U7f(l,ua
L,17fU,1u
511
_,1761,10
1,U9f0,10.
4,27ii1,:.
*t denotes the duration of the tests.
}Data from chemical analysis.
z (F.,) .- S?a (T ~) [Rost [~'t) ~)~ (~' i)~-~
Here o[E(x)] is the cross section of the reaction at a depth x, R(Ej is -the range. of
deuterons of energy E, Si(E) is the stopping power of the substance, ao = a(Eo), So $~(Eo),
and Eo is a certain "base " energy. The constants A and ao were determined in an experiment
with a standard, i.e., a specimen of known stoichiometry, for which we used specimens of re-
actor graphite, natural diamond (carbon), and aluminum nitride (nitrogen). The S(E) data
were taken from [9], making it possible to determine the relative value .of S(E)/So with an
error of 1-1.5%. The energy dependence?of the reaction cross section o(E)/oo was measured in
experiments with thin carbon and nitrogen (adenine CgHSNg) films with an error of no more than
The results of cross-sectional measurements are given in Figs. 1-2. The figures show
that a broad resonance at Ed = 1.2 MeV dominates in the 12C(d,.po) reaction cross section
while the cross section for the 14N(d, po) reaction grows smoothly by a ?E3'' law to a reso-
nance at 1.9 MeV. The shape of the cross section essentially determines the shape of the spec-
tra measured (see Figs. lb and 2b) and the effective depth of analysis.; calculation for
Rl = 1.8 MeV gave a value of = 4.9 um in an analysis for carbon and 2.8 um fin. an analysis
for nitrogen in steel:
This method of isolating the surface film is easily realized in the case of nitrogen~mi-
croanalysis. Measurements showed that the nitrogen film on steel specimens is fairly stable
434
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
and has a thickness of approximately 1 monolayer (0.1-0.2ug/cm2 for C = 1). The thickness
of the surface carbon film is substantially greater (0.4-3ug/cm2). andpincreases with time
[scattering-chamber pressure -3.10-5 torr (1 torr = 133.322 Pa)]. Thus, Cp for the carbon
film was determined by a simpler method, exploiting the fact that the surface film is expli-
citly isolated in the measured proton spectrum at a deuteron energy E1 > 1.7 MeV and a detec-
tor resolution dE 50 keV (see Fig. lb). For carbon analysis, therefore, it was sufficient
to carry out measurements for one value of deuteron energy, viz., E, = 1.76 MeV. This value
is optimum since interference with oxygen [the contribution of the 160(d, po:) to the measured
spectrum] is minimal at the energy indicated [8]. For nitrogen the interference of the
14N(d,.po) and 14N(d, ao) reactions is easily eliminated by placing a polyethylene filter
(~17 mg/cm2) in front of the detector; the filter removes other impurities from the spectrum
of detected vrotons (see Fig. 2b) [7].
The specimens, in the form of disks 14 mm in_ diameter, were sections made at a particu-
lar depth in the material in contact with liquid sodium. They were prepared from the wall of
an operating tube of a BN-350 steam generator as well as test-stand specimens and their sur-
face was polished. The beam diameter on the specimen was -1.5 mm. The system for moving
the specimen permitted simultaneous scanning of the beam over the surface of the target. The
final data were obtained by averaging the results of measurements at 3-5 points of the sur-
face in order to eliminate the influence of local inhomogeneities which in some cases served
as the main source of errors. The dose of an individual irradiation was 50-130 uCi at a beam
current of 0.1-0.3 uA and the statistical error of measurement of the yields Y(E,) for the
specimens was 1-5%, depending on the impurity content and the value of E,, for a measuring
time of 10-25 min. The estimated inherent (systematic) error of the method .in the case of
one-parameter analysis (of carbon) is 3%.
The results and characteristics of the conditions of the testing of specimens of lOKh2M
steel are given in Table 1. The investigations were carried out at 400-650?C for 500-56,000
h. High-alloy steels of different compositions (lOKh13, lOKh18N10, lOKh1810T) were the r.ar-
bon receivers. The presence, in the austenitic chromium-nickel steel, of titanium which is
strongly carbide-forming substantially affected the surface concentration of carbon in the
lOKh2M steel. When the testing apparatus was made of 10Kh18N10T steel, then accoxding to
chemical analysis data the average carbon content in the surface layer 50-100 um thick of
lOKh2M steel was 0.03 and 0.02 mass % after being kept in sodium at 500 and 550?C, respec-
tively. When lOKh18N10 steel served as-the carbon receiver, Co increased to 0.075-0.09 mass %.
Data from nuclear microanalysis show (see Table 1) that when lOKh2M steel is decarburized, the
value of Co is smaller than the data from chemical analysis by a factor of 2-2.5. Such a low
carbon content near-the surface can be explained not only by equalization of the thermodynamic
activity of carbon in the system lOKh2M--Na-lOKh18N10T, tending to equilibrium, but also by a
change in the stoichiometric equilibrium of the surface layers of the steels with respect to
the main elements (Cr, Fe, Ni) because of their dissolution and deposition in the sodium.
This, in turn, causes a change in the thermodynamic .activity of the nitrogen and carbon. Con-
sequently, when the surface is depleted of chromium the content of carbides and nitrides de-
creases.
1DD~
s 5D
H 30
~ ZO -
0 10~ ~~~ ~+ ~ ~
U 3
2
f~ ~
f
a,l 0,: x, mm
Fig. 3. Distribution of concentration C(x) of carbon impurity
(a) and nitrogen impurity (b) over the depth x for the wall of
the steam generator made of lOKh2M steel in the BN-350 plant.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
TABLE 2. Average Carbon Content in a 0.1-
mm Surface Layer of lOKh2M Steel after
500 h in Sodium, mass
lOKh18N10
lOKh13
lOKh18N10T
11,(/7~~
Il,llyj
n,n~5
I1,11li
11,11'_.'.11
The hypothesis put forward here can be illustrated by the data for the wall of a BN-350
steam generator operated for 56,000 h (Fig. 3, Table 1) at 390-400?C. At such a temperature
a change in the stoichiometry of the surface layers can be observed only after long tests, ow-
ing to the slow rate of mass transport of the main elements. The results of mass-spectromet-
ric analysis of specimens of lOKh2M steel reveal that near the surface the chromium content
increases from 2.25% to 3-4%, as is confirmed by the existence of a process of transport of
the main elements in a nonisothermal loop through the sodium between the lOKh18N10 and lOKh2M
steels. This process leads to carburization of the lOKh2M steel, which. is not entirely usual
for a low-alloy pearlitic steel. Nevertheless, this tendency is indicated by .chemical analysis
data which attest to a carbon concentration of up to 0.12 ? 0.01 mass % in the surface layer
(-100 ym) of the lOKh2M steel. The results of nuclear microanalysis (see Fig. 3) support this
with great reliability for carbon as well as for nitrogen, for which no other data were ob-
tained. The results of microanalysis show that carburization and nitration processes are no-
ticeable at a depth , = lc? 1. (8 )
The particle weight, taking account of .the expulsion force, is
Calculation of the Critical Velocity of Suspension-Fearing Flows with Small Particle Concen-
trations in the Liquid
A small concentration of the suspension here is understood to be such that the mutual
influence of the particles at the channel wall when u ~ uc may be neglected. In this case,
the critical velocity is taken to be equal to the breakaway velocity.
The breakaway condition for a particle at the wall of a horizontal channel (in its-lower
If the particle is nonspherical, Eqs. (1) and (6) take the form
Fp +=10.5 (k l+)=;
(11)
Fp+=5.18 (k l+)a.4~~
(12)
where ks = ZM/Z, Z = (6V/n)1/3.
Substituting FF, Fa, Fw, and F into Eq. (10) from Eqs. (1) and (6)-(9), an expression
is obtained for the dynamic flow ve~ocity corresponding to particle breakaway
u~~{k.~[k~/l-Fn6'(Pp-p)l/G]10.5p}??5/1~S (l+~r~); (13)
ld*~ ~ I?, Ilr:r-! ne ll~p _!~) l-~f; lu,~,t ~.n,ix
~ 1p (5.1libs~?~''' -~ 0.117(i/~.~1~+5n f ~osot (1/{)
(~i < l F ~ 30(1).
The critical velocity is determined from the solution of the equation establishing a
relation between the dynamic and mean-mass flow velocity [9]:
u*= uc/15.1ii lh (ucUc/r)- .G4]. (15)
For a channel with a rough surface, when the particle is on the side of a pro3ection,
the breakaway condition may be written in the form
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
FF coax-hw~ina-LlLsina ~G:.,.(F;,-!-hW CU?ct-}-1'F.~ina- I~Lco~~c)? (16)
The angle a is estimated approximately from the relation (Fig. 1)
cos x z (/~ ~ - /~~ ;'~)/ ~;r .
Using Eq. (16), the following expressions are obtained for a rough surface
ltd'.. G'~.lr,/la n?' (Dp--!') (~inry ~ G:, cn. ~) l~li ~u.:~'~S
[ '
it~~ 1-.,.J,:a-Irn~. (Pp5_` P) (~iua-)_G~.~co?~c) h;f~~u.61 ~o^sn~
{-
)
!
?
i
?
n a
-
,
T s
(cus a-
p (.>,18kS
(-u,(17fi(sina ( k,.casa)l?~',~')
(5 < 1.+~ 300).
Equations (18) and (19). are valid if hr+ < S. For cases when the particle touches a few
projections of the rough surface, the breakaway velocity is determined using Eqs. (13) and
(14).
Dependence of the Critical Velocity on -the Concentration of the Suspension
At velocities of the suspension-bearing liquid flow exceeding the critical value, there.
is an equilibrium dynamic layer of particles at the channel surface; the density of the layer
depends on the concentration of the suspended phase. If particles at the channel wall are suf-
ficiently far apart, their mutual influence in the boundary layer of liquid may be neglected.
In this case, the critical velocity is equal to the breakaway velocity. The limiting concen-
tration of the suspension at which this condition is still satisfied may be determined as fol-
lows. On average, the number of particles on unit surface at the face of an isolated element
of suspension volume is
where +p is a coefficient.
The mean distance between the centers of spherical particles at the face surface is
s = (nls!(ic)+/s (~; ~P)+/z ( 21)
It may be assumed that coincidence of the face planes and the channel surface has little
influence .on s when u > uc. There evidently always exists a line s = sl, on crossing which
the mutual influence of particles at the channel wall may be neglected. The value of. s, cor-
responds to a suspension concentration
The ratio sl/Z depends on the conditions of liquid flow around the particle. If the particle
-size considerably exceeds the thickness of the laminar sublayer, the ratio may be estimated
on the basis of the data of [10], and is -15-20. .
When c > c,, the critical velocity is determined on the basis of the breakaway condition
for a group of particles: Fora group of particles at the wall, the adhesion force and the
weight increase in proportion to the surface density of particles n in comparison with the
same forces acting on a single particle. At the same time, as a result of the mutual influ-
ence of the particles, the frontal and lift forces increase considerably less than the adhe-
sion force and the weight, For a group of particles, Eqs. (10) and (16) may be written in
the form
1%p + IcTF~ > JrT (I%a -}- Fw) lcn;
!~F(Cosa-k.rsin a)-~-I;L(sina-( Tc,.coca)%(kTFa-f ~w(sina~-l;Tcosa))1s'n.
(23)
(24)
The value of k is determined from the condition that, when c = cl, the critical velocity
is equal to the breakaway velocity, or kn = 1; hence, k = s;.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
Fig. 1. Determining the
angle of slope of a rough
surface.
Fig. 3
Fig. 2. Dependence of 'the critical velocity on the dimensions of the suspension par-
ticles: 1, 1', 1") calculation by the method here described; 2) calculation from the
data of [3]; 3) experimental values.
Fig. 3. Concentration dependence of the critical velocity: 1) by the method here
described; 2) calculation from the data of [3J.
9 90 900 L, ?m
Fig. 4. Dependence of the critical ve-
locity for iron particles in sodium.
It follows from the solution of Eqs. (23) and (24) that Eqs. (13), (14), (18), and (19)
are used to determine the critical velocity; when Z < 5, the right-hand side of the inequal-
ity is multiplied by (kn)?'S, and when 5 < Z.F. < 300+by (kn)?'?'. When c < cl, kn = 1; If
c > cl the product kn is found on the basis of the estimate of sl. When Z.~. 15, sl = 15Z. In the range 2Z < Z+ < 15, s, varies from 2Z to
15Z. In final form (taking account of the simplifications adopted), data for the calculation
of kn are given in Table 1.
The coefficient cp may be regraded as a parameter taking account of the influence of mu-
tual collisions of particles on their deposition. As the distance between particles decre~s-
es, so ~p differs more strongly from unity. Taking a linear dependence of ~ on s, cp = AJci s
may be written, where A is a proportionality factor. Calculation of A is based on the condi-
tion that, when c = 10-4, the particle collisions are practically nonexistent, i.e., ~ 1.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
Finatty
Experimental Determination of the Adhesion and Friction Coefficients
To confirm the possibility of using Eq. (8) to calculate the adhesion force of particles
in sodium and to determine the coefficients k and kT, the following procedure is used. From
the balance of forces acting on a particle, at. an inclined surface, it follows that the lim-
iting particle dimension at which breakaway sets in under the action of the gravitational
force is
_ Fikv l1/2
lL nb' (Pp -P) (cos ~3-} siti ~~kr) J '
Equation (26) is intended for the calculation of the coefficients ka and 1cT from the experi-
mental value of ZL at different angles of slope of the surface. With this aim, the function-
al dependence ZL(ka, kT) is analyzed by the least-squares method.
To determine ZL in liquid, a stand was set up in an empty vessel, with Kh18N10T steel
supports fixed at the same level, having a polished surface on one side. The polished sup-
port surfaces were inclined to the horizontal at angles of 0, 30, 60, and 90?. A carefully
mixed suspension of particles of the chosen material was poured into the vessel. Particles
of the suspension of dimension less than ZL were retained on the supports after deposition.
After evaporation of liquid from the vessel at a rate at which the velocity of level. drop is
no more than 1 um/sec, the substrate is transferred to the field of view of the microscope.
Since the adhesion force in the gas medium is several orders of magnitude higher than in the
liquid [4], transfer of the supports and their rotation does not lead to particle breakaway.
Using the ocular scale of the microscope, the longitudinal and transverse dimensions of
the largest particles were estimated with an accuracy of up to 2.5 um. The vertical dimen-
sion of the particles was determined by scanning. The particles considered were oval in form;
the deviation from the mean dimension was by no more than a factor of 1.5 for each of the three
measurements. For each support, the mean dimensions of the three or four largest particles
were averaged. The-value obtained was taken to be equal to the limiting particle dimension
ZLi in the calculations.
To estimate the accuracy with which the coefficients ka and kT are determined, the stan-
dard deviation of the values measured for ZL from the results given by Eq. (26) was calculated.
Experimental and theoretical values of the parameters foz particles of different materials ob-
tained in water, alcohol, and sodium are shown in Table 2. Taking into account that the ad-
hesion force is probabilistic in character, the value of ka corresponding to the maximum ad-
hesion force is determined by the given method.
Theoretical Estimates of the Critical Velocity. Comparison with Experimental Data
The dependence of the critical velocity on the dimension of tungsten particles in sodium,
calculated in accordance with the data in [3] and the method here proposed, is shown in Fig.
2. The following initial parameter values are adopted: k = 5.7.10-5 N/m, kT = 1.25, ks =
0.7, a = x/12, D = 0.02 m, c t 0, and a sodium temperature of 300?C. Curve 2 is a general-
ization of a large number of experimental points (shaded region) for suspended. particles with
dimensions of more than 100 um. The segment of straight line3 denotes the limits of experi-
mentally measured values of the critical velocity. Theoretical values are enclosed between
curves 1' and 1 " which correspond to the limiting deviation of the adhesion and friction
coefficients from the mean.
Curves of the concentration dependence of the critical velocity for a tungsten suspension
in sodium with Z = 100 um are shown in Fig. 3. The shaded region and curve 2 correspond to
calculation by the data of [3].
On the basis of the results obtained, the critical flow velocity of sodium with suspended
iron particles is estimated, in conditions close to those for flow in the active-zone chan-
nels of the BN-600 reactor. The sodium temperature is taken to be 600?C, Dc = 2.42 mm, k =
0.7; a = x/12. The results of calculation are shown in Fig. 4. Hence it follows that, for
the given case, with a mean-mass flow velocity of 4.75 m/sec, all particles of size >1 um are
transported.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
TABLE 1. Dependence fcir Determining the
Product kn
1. 75
(n/~~)/m'/~
~ JJ ? ~ n'4 ~1~3/ `
(48c/~c)2/~q~
(fir/n)2/[t1+W
_ 34Gc~/~w
TABLE 2. Results of Experimental Deter-
mination of Adhesion and Friction Coeffi-
cients of Particles in Liquids
The given relations allow the critical velocity of suspension-bearing flows to be cal-
culated over a broad range of initial parameters. Using the data of Table 2, the transports-
bility of suspensions of particle size up to 1 um may be estimated, the most dangerous points
may be determined from the viewpoint of clogging of the cross section, and the propagation
of some active and inactive impurities in loops with a sodium heat carrier may be estimated.
NOTATION
Z, particle dimension, m; Z1,1, minimum particle dimension, m; ZL, limiting particle di?men-
Sion, m; Dc, channel diameter, m; h microprojection height of surface roughness, m; P.r, rad-
ius of roughness projection, m; s, rdistance between particle centers; s,, limiting value of
the distance between particles, m; V, particle volume, m3; FF, frontal force, N; Fa, adhesion
force, N; Fw, weight, N; FL, lift force, N; u critical velocity, m/sec; u*, dynamic velocity,
misec; c, bulk concentration of suspension, m~/m3; c,, maximum concentration at which the
critical velocity is equal to the particle breakaway velocity, m3/m3; p, liquid density, kg/m9;
p , particle density, kg/m3; n, number of particles per unit channel surface, m-2; v, kine-
m~atic viscosity, m2/sec; ZT, sliding friction; ka, adhesion coefficient, N/m; Z+ = Zu*/v,
FF+ = FF/pv Fi,~ = FL/pv dimensionless parameters; ks, sphericity coefficient; k, propor-
tionality factor; g, acceleration due to gravity, m/sect; a, angle of slope of surface of a
roughness projection; ~, angle of inclination of the support to the horizontal; Re, Reynolds
number.-
LITERATURE CITED
1.
R.
Brown, Trans. Am. Nucl. Soc., 30, 471 (1978).
2.
V.
S. Knoroz,in: Izv. Vses. Nauk.-Issl. Inst. Gidrotekh., 40, 37 (1949).
3.
D.
Thomas, AIChE J., 8, No. 3, 373 (1962).
4.
J.
Happel and G. Brenner, Hydrodynamics at Small Reynolds Number [Russian translation],
Mir, Moscow (1976).
5.
P.
[in
G. Romankov and M. I. Kurochkina, Hydromechanical Processes of Chemical Technology
RussianJ, Khimiya, Moscow (1974).
6.
J.
Cleaver, and B. Yates, J. Coll. Interf. Sci., 44, 464 (1973).
7.
A.
D. Zimon, Adhesion of Dust and Powders [Russian translation], Khimiya, Moscow (1976).
8.
A.
[in
D. Zimon and G. A. Serebryakov, in: Abstracts of a Conference on Aerosol Adhesion
RussianJ, Frunze (1974).
9.
L.
G. Loitsyanskii, Mechanics of Liquids and Gases [in Russian], Fizmatgiz, Moscow
(1970).
10.
V. I. Subbotin et al., Preprint No. 672, FEI, Obninsk (1976).
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
USE OF DETECTORS EXTERNAL TO REACTOR TO DETERMINE THE REACTOR POWER
AND THE-MEAN ENERGY DISTRIBUTION OVER THE HEIGHT OF THE ACTIVE ZONE
A.
N.
Kamyshan, A. M. Luzhnov,
A.
S.
Makhon'kov,. V. V. Morozov,
N.
.S.
Orekhova, and S. G. Tsypin
Systems based on detectors positioned outside the reactor have been successfully used
to monitor the power and mean energy distribution over the height of the active zone in non-
Soviet atomic power stations with PWR [1-3]. The development of systems with monitoring from
outside the reactor is also of great importancd in the USSR [4J. As a rule, these .systems,
permitting continuous take-up of information characterized by inertialessness and high re-
liability, are associated with emergency-protection systems of the reactor.
A typical setup of detectors external, to 'the reactor (DER) is shown in Fig. 1. Three or
four detector units are positioned around the reactor, each with-two to four detectors at dif-
ferent heights [1, 3, 5]. Since the neutron .field outside the reactor depends on the energy
distribution. over the volume of the active zone, the ,form of-the energy distribution over the
.height may be recovered using the appropriate method. of analysis of the detector readings. The
:.integral of this distribution will be proportional to the reactor power.
In the vresent work, an attempt is made to determine some characteristics of systems of
DER optimal from the viewpoint of a reasonable compromise between the error of the results
and complexity of construction. One of the basic characteristics of this system is the number
of detectors in each unit. Usually, it oscillates from two to four, but there is no sound ba-
sis for the choice of any specific number [1, 3, 5]. For a specified class of functions, in-
cluding a set of possible forms of height distributions, how may the minimum number of detec-
tors sufficient for recovering any function from this class with satisfactory accuracy be
found? To estimate the accuracy of recovery, the following criteria may be used
fir = I max F (a) - max f (z) I / maa F (z). (2)
zE[~~, 1/] zE[~~, II] zE~il, H]
Here F(z) is the true energy distribution over the height; f(z) .is the energy distribution
recovered; fi is the active-zone height.
The, reading of detector number k in the unit is related, to the energy distribution over
the volume of the active zone FV(r) through the space-dependent weighting function of this
.detector Sk(r)
v
where V is the volume of the active zone; Sk(r) is the space-dependent weighting function,
which specifies the response of detector k to a point fission source of unit power at point
r.~
It is assumed that FV(r) may be written in the form
where the function Fl(r,~p) is normalized to unity
*Methods of determining the weighting functions were considered in detail in [6-8].
Translated from Atomnaya $nergiya, Vol. 57, No. 1, pp. 18-21, July, 1984. Original arti-
cle submitted August 22, 1983.
0038-531X/84/5701- 0443$08.50 ? 1985 Plenum Publishing Corporation 443
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
Fig. 1. Positioning of detectors: 1) detector units; 2) .active
zone; 3) shield; 4) channel and three-detector unit; 5) channel
and four-detector unit..
'l>< R
J dcp S rF~ (r, q,) dr -- 1,
0 n
.and R is the radius of the active zone.
Combining Eqs. (3) and (4) gives
rr
Dh = ~ F (z) SA (z) dz,
U
2n R
Sh (z) _ ~ d~ ` r drSk (z, r, ~) Ft (r, cp).
J
Q ~~
Below, Sk(z) will be simply called the weighting; functions. .The height of the active
zone is divided into N equal sections, and Eq. (5) is written in the form
:, zrr+t
Dk = W f ~ ~f lZ),Sp (Z) dZ ~' ... -~- W p
~ ~N (Z) 'Sh (Z) d2,
zj -
zN
- .
where the sequence {zn; n = 1, N} specifies the division of the active-zone height; and
/Zn+t'
tPn \Z) - F.(Z) I ~ F \Z) dZi
zn+1
zn
444
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
nk = Wi'Skl'~" ~~2'Sk2+ ? ? ? -{- WN'SkN+
=n+i
Skn = ~ ~n {Z) Sk {Z) IZZ.
=n
If the number of detectors in the unit is N, and.the reading of each detector is written in
the form in Eq. (7), the following system of equations is obtained:-
S21 s22
? ? ?
S2N
`'W 2 `
D2
SN1 SN2
? ? ?
SNN
WN
DN:
solution of which yields {W n = 1, N}. -From the values of { n} (called the incomplete in-
teerals of the distribution, the integrand function F(z) may be recovered.
Oae method of recovery-was described in [5]. The function F(z) is written. in the form
of a series of sines with as many terms as there are detectors in the unit
F'(z)=Cisin(H ~+Czsin( H~)+- ...-i-CNSin l H ~?
(9)
In this case, recovery of F(z) reduces to finding the coefficients {Cn; n = 1, N}. These co-
efficients are related to {Wn; n = 1, N}.
Uzi Aza ... AzN 1 1 CZ 1-- ~W2 I .
N1 ~N2 ANN ~~NI ~WN
Here A is a matrix of constant coefficients, which-may be obtained by substituting Eq. (9)
into Eq. (10) successively for all.n from 1 to N.
Since the number of incomplete integrals from-which F(z) is recovered is equal to the
.number of detectors in the unit, determining the optimal number of detectors entails finding
the minimum number of incomplete integrals sufficient for the recovery of any distribution
from the specified set of possible forms with the required accuracy.
As an example, the results of recovery for two ,. three, and four incomplete integrals of
the distribution given in [9] are .shown in Fig. 2. It is evident that the recovery of the
initial distribution from two integrals (i.e., from the readings of two detectors) may lead
to considerable distortions in-form of the .distribution, while increase in the number of in-
tegrals from three to four does not facilitate significant decrease in error according to the
estimates in Eqs. (1) and (2). On the basis of an analysis of the distributions in [1] for
PWR, it may be concluded that a construction with three detectors in a unit is acceptable.
Two detectors allow .the form of the distribution to be approximately estimated only in the
case when the function which describes it is unimodal.
Determining the number of detectors in the unit, it is natural to pose the question of
how to position these detectors in order to have minimum error in recovering the distribution
by the method in Eqs. (8)-(10). Change. in configuration and physical proverties of the pro-
tection and the position of the detectors appearing in the unit only influence their weight-
ing functions. Therefore, the initial problem reduces to finding the form of the weighting
functions ensuring minimum error of recovery. The weighting functions of the detectors are
used in calculating the elements of the weightini; matrix S, which specifies the coefficients
of the system of linear equations in Eq. (8). The relative error of the incomplete integrals
which are the solution of this system depends on the conditionality of the matrix S and may
exceed by a factor of m (m is the conditionality number of the matrix) the relative
error of the elements of the matrix S and the detector readings {Dk; k = 1, N}.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
1, o
0
1~ 0
0, 5 1~0
_ z, rel. units
FiQ. 2. Results of recovery of the distribution from two (a),
three (b), and four (c) values of the incomplete integrals .with.
8F = 11, 3.5%, 2.4% and ds = 19%, 4.8%, 2.7%, respectively - see
Eqs. (1) and (2): the continuous curves correspond to the ini-
tial distribution and the dashed curves to the distribution re-
covered.
One form of matrix with the minimum possible conditionality number, equal to unity, is
a diagonal matrix with identical elements on the diagonal
S 0 ... 0
IlSisll= o s :.. 0 .
oo...s
A weighting matrix of this form corresponds to the "ideal" weighting functions shown in
Fig: 3. In this case,, each of the detectors of the unit records only the radiation of-the
section of .the active zone opposite it (called "its section"), and does not react to the ra-
diation of other sections. In existing systems, the detector units are most. of ten positioned
in vertical channels in hydrogen-containing protection. The weighting functions for one such
.construction with four detectors per unit are shown in Fig. 3 [6]. The conditionality num-
ber-corresponding to the weighting matrix is three, which may lead to threefold elevation of
the error in the parfial integrals in comparison with .the error for a construction-:with an
"ideal".form of the detector weighting functions with the same accuracy of the initial data.
Therefore, in developing systems of DER it is expedient to choose a construction in which the
weighting function has the.form closest to ideal: The development of such a construction is
a separate problem, and will not be considered here.
Systems of DER are-also used in monitoring reactor power [1', 5]. To measure the power
W, it is sufficient to have one detector, if its weighting function is equal to a nonzero,:
.constant when z E [0, H]. .However, the creation of a construction in which the detector
would have such a weighting function is a complex scientific and engineering problem. It is
.obvious. that it is better to use a construction with several detectors in a unit, since in
this ease such rigorous requirements-need not be imposed .on their weighting functions.
-446
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
detector Active zone z~ rel. units z, rel, units
unit _
S4 (Z)
So 0 3 2 > 0
S, rel. units S, rel, units
Fig. 3. Weighting functions of detectors: a)
"ideal"; b) in a real system [6].
D erector
unit
Active
zone
Fig. 4. Weighting functions ensuring that
al and az are equal in Eq. (11).
The relation between the power and the readings of the detectors in each unit is now de-
termined. Suppose that a two-detector unit is used to measure the power. The power W is
proportional to the sum of the incomplete integrals W, and W2, which are written in accordance
with Eq. (8), with N = 2
W _ !V t - t- 6V~ = a,D~ -} a~D2. (11)
Here al = (Szs._ Sz,)/(S1iSzz "` S1zSzi); as = (S11 - S~z)/(S,,SzY __. SisSai).
It follows from Eq. (11) that the power is proportional to a linear combination of the
detector readings; the coefficients a, and a2 are not equal, in the general case, and depend
on the weighting functions of the detectors. In practice, it is expedient for the. coeffi-
cients to be equal, since in this ease the power is proportional.to the sum of the detector
readings. With .the weighting functions shown in Fig .. 4, al and az are equal. It is obvious.
that such weighting functions of the detectors may only be present in complete symmetry (ref-
ative.to the central plane) of the geometry and physical properties of the space separating
the detectors and the active zone of the reactor. If there is no such symmetry, the condi-
tion of proportionality of the reactor power to the sum of detector readings of the unit is
observed only with considerable error.
Thus, it is expedient to use units with no less than three detectors for the correct de-
termination of the power and mean energy distribution over the height of the active zone. The
position of the detectors is chosen for a specific construction of the reactor and shielding
in accordance with the given requirements on the weighting functions.
1. P. Sipush et al., Nucl. Technol., 31, 12 (1976).
2. J. Humpfries and R. Knapp, in: Proceedings of an IAEA Symposium on Nuclear Power Plant
Control, France (1978).
3. H. Neuschafer and J. Humpfries, Trans. Am. Nucl. Soc., 27, 856 (1977).
4. I. S. Krasheninnikov and V. V. Matveev, At. Energ., 50, No. 2, 110 (1981).
5. US Patent No. 4,079,236 (1978).
6. C. Mildrum and J. Easter, Trans. Am. Nucl. Soc., 27, 676 (19.77).
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
7.
M.
Crump and J. Lee, Nucl. Technol., 41,
87 (1978).
8.
H.
Tochihara et al., Nucl. Technol., 58,
310 (1982).
9.
V.
A. Sidorenko, Problems of Safe Operation of Water-Cooled-Water-Moderated Reactors [in
Russian], Atomizdat, Moscow (1977).
BUILDUP OF GASEOUS NUCLEAR REACTION PRODUCTS IN CHROMIUM AND NICKEL
CAUSED BY HIGH-ENERGY ELECTRON IRRADIATION
A.
G.
Zaluzhnyi, 0. M. Storozhuk,
M.
V.
Cherednichenko-Alchevskii,
N.
L.
Emets; L. Z. Ozhigov,
Yu.
N.
Ranyuk, and V. A, Yamnitskii
In performing experiments involving irradiation of materials with accelerated particle
beams for the purpose of simulating and investigating the phenomena occurring during irradia-
tion in reactors, it is necessary to maintain the similarity numbers with respect to certain
factors, such as the total number of primary defects, the spectrum of primary knocked-out
atoms (PKA), the amount and mass distribution of nuclear reaction products, the damage pro-
file, etc.
While the required similarities pertaining to the numbers of primary defects, their dis-
tribution profiles, and the PKA spectra can readily be secured for simulation of neutron dam-
age in heavy-ion or proton irradiation of materials [1], certain difficulties are encountered
with regard to the amount and mass distribution of nuclear reaction products. The problem of
helium buildup in simulator experiments is especially critical, since helium actually causes
many radiation phenomena, for instance, high-temperature embrittlement.
Simulator experiments on helium buildup involve alpha-particle irradiation in cyclotrons,
where the so-called tritium trick [2] or the (p, a) reaction [3] is used. However, irradia-
tion with a beam of high-energy electrons and photons [4] also makes it possible to-study the
helium buildup in materials for different ratios of the number of helium atoms to the number
of primary radiation defects. We shall present here the results of theoretical and experi-
mental determinations of the buildup of gaseous nuclear reaction products ,. primarily helium
and hydrogen, resulting-from irradiating various materials with 200-MeV electrons.
The following basic processes occur as a material is irradiated with a high-energy elec-
tron beam: development of electron-photon showers; formation of PKA by the shower electrons;
interaction between shower photons and nuclei of the target material along with the formation
of PKA due to photons, light nuclear reaction products (up to helium inclusively), and residu-
al nuclei; decay of radioactive residual nuclei. The simultaneous occurrence of many mutually
related processes makes it difficult to obtain numerical results, and, therefore, we used
mathematical simulation of the interaction between radiation and the material, utilizing the
IMITATOR program system [5] and B~SM-6 and ES-1040 computers.
Several simplifications are presently used in simulating the electromagnetic shower: We
neglect the angular divergence of the electron and photon beams as well as the Compton scatter-
ing, due to which the results obtained hold only for targets with a thickness of up to two
radiation lengths. The spectra of PKA produced as a result of electron scattering on the
target nuclei are calculated by using the total electron spectrum (primary and secondary elec-
trons resulting from the development of the electron--photon shower) and considering the Fermi
form factor.. The number of primary point defects is determined by means of the TRN-standard
cascade function. The formation of nuclear reactor products in the material is calculated
on the basis of the photon spectrum at a given depth in the target without considering the
electronuclear processes.
Translated from Atomnaya ~nergiya, Vol. 57, No. 1, pp. 21-25, July, 1984. Original
article submitted May 30, 1983.
448 0038-531X/84/5701-0448$08.50 p 1985 Plenum Publishing Corporation
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
The multistage process of interaction between a photon and a nucleus has been divided
conventionally into two stages. At the first stage, the nucleus is considered as an assembly
of free nucleons and quasi-deuterons. while the energy pulse is transmitted to the nucleus by
the oncoming photon during collisions with nucleons; forced absorption is simulated first,
and then the cascade stage of nuclear interaction proper [6]. At the time when the energy of
individual nucleons is lower than the emission threshold, the second simulation stage is ini-
tiated: the vaporization fission model, which accounts for the process of "cooling" of resi-
dual nuclei as competition between the emission of individual particles and the fission pro-
cess [7].
The calculations are performed by using the Monte Carlo method for a set of fixed photon
energies. The items determined in calculations are the number and the space-energy distribu-
tion of the following particles emerging from the nucleus: ~+'-; ~?-, ~ -mesons, neutrons,
protons, deuterons, tritons, and 3He and `'He nuclei. The fission events, the charge, the
mass, and the residual nucleus energy are recorded separately.
If we know the photon energy spectrum at a certain given depth in the material dNY/dEY
and the probability of development of the i-th nuclear reaction products pi(EY) during the
photonuclear process for photons with the energy_EY, the yield of the product in question
from a thin target layer at the assigned depth is calculated from the equation
6, = I At (~v) 6vtot [~v) ~ dGv > clL'~? (1)
En
where dytot (EY) is the total cross section of absorption of a photon with the energy EY by
the target material, determined by means of the semiempiric.al expressions interpolating the
experimental data from [8`]. Integration is performed from the neutron emission energy En
(approximately 8 MeV) to the maximum bremsstrahlung photon energy EY~x. The energy spectra
and the charge and mass distributions of residual nuclei are calculated by means of equations
similar to (1). In connection with the fact that a considerable part of the residual nuclei
are unstable, the experimentally obtained radioactive decay chains [9] are considered in simu-
lation.
The described simulation method [10] was used for calculating the distribution of pri-
mary radiation damage and the buildup of hydrogen and helium resulting from chromium and nick-
el irradiation with 225-MeV electrons (Fig. 1). The buildup of helium and hydrogen is deter-
mined almost entirely by photonuclear processes, whose contribution to the development of
primary defects does not exceed 15%, since the predominant role is played by secondary elec-
trons of the electron-photon shower. The mass distribution of residual nuclei resulting from
electron irradiation that has been found by simulation differs fundamentally from such a dis-
tribution resulting from fast neutron irradiation.
Figure 2 shows the histograms of the yields of light nuclear reaction products and res-
idual nuclei for chromium and nickel at a depth of 2 cm in the case of irradiation with 225-
MeV electrons and irradiation with neutrons from a BOR-60 reactor. This figure also shows
the fission cross section of the nucleus in question (denoted by f). In accordance with Fig.
2, electron irradiation produces, besides ?He, other gaseous nuclear reaction products: 3He,
hydrogen isotopes (up to tritium, inclusively), argon, chlorine, and possibly, nitrogen and
oxygen,, which are generated as a result of fission of-the residual nuclei from the photonu-
clear reaction (the fission barrier after Nix is used in calculations).
It should be mentioned that a considerably greater buildup of helium is produced by elec-
tron irradiation than by irradiation with neutrons from the BOR-60 spectrum (the yield of
nuclear reaction products in the latter case has been calculated by using the ALICE statisti-
cal nucleus model, which has also been included in the IMITATOR system [11]). It is also re-
markable that, during electron irradiation of chromium, the buildup of argon exceeds the hel-
ium buildup resulting from neutron irradiation.
Simultaneously with simulation of the buildup of gaseous nuclear reaction products caused
by electron irradiation, we performed an experiment on direct determination of the amounts of
helium formed in chromium and nickel under the action of an electron beam. For this, we ir-
radiated in a KhFTI LUE-300 linear electron accelerator a package of consecutively arranged
targets with an electron fluence of 1019 cm-Z at 225 MeV.
449
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
4U F-
i"
Q ? ~ X, cm
Fig. 1. Theoretical distributions of defects (a), hydrogen (b),
and helium (c) in chromium (-----) and nickel (----) developing
under the action of a 225-MeV electron beam.
;~-f~i_ ~LL_I]_~ ~l
;t d ~He f Ne Mg Si. S ar Gu T~ Grv
u t `'He F Nu AL P G6 K SG V
x
r'~
~~~~
n d 'Ne f Si S Ar Cu -1"i fr Fe Ni.
p t `'I{e A4 P Cb K Sc V Mn Ca
Fig. 2. Theoretical histograms of the yields of light nuclear reaction products
and residual nuclei at a depth of 2 cm, produced by irradiating chromium (a) and
nickel (b) with a 225-MeV electron beam ( ) and neutrons from BOR-60 (-?-?-?-).
The helium percentage in the irradiated specimen, which has a thickness of 0.2 mm, is
determined by vaporization in vacuum with direct determination of the amount of helium by
means of an IPDO-2A partial-pressure gauge. The specimen is placed in a molybdenum glass flask
with a tungsten heater, which ensures temperatures of up to 2300?K. The flask walls are in-,
tensively cooled during the vaporization process. A forevacuum is created in the volume by
means of a TsVN-1 adsorption pump, while an NORD-100 magnetic discharge pump is used both for
preparation for the experiment (when all the units of the degassing device are heated to 600-
700?K) and for creating the high vacuum immediately before the experiment. During the vapor-
ization process, the atomization volume is evacuated by means of a GIN-05M1 ion=getter pump,
which pumps out only chemically active gases under sorption conditions. An RMO-4S omegatron
data unit is connected directly to the branch pipe of the ion-getter pump.
The sensitivity of the device with respect to the partial pressure of gases is equal to
4.10-8 Pa, which corresponds to 1011 helium atoms in the volume of the device, i.e., the
atomic fraction of the helium percentage in the vaporized specimen (-5 mg) is approximately
equal to 10-e%.
~~y
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
o --- ~-- - -L--- ! ---~
~, s ~ (~ x, cm
Fig. 3. Experimental helium distribu-
tion, determined by irradiating a pack-
age of nickel targets with 225-MeV elec-
trons at 90?C, using a fluence of 1019
cm 2.
The maximum helium percentage is observed at a depth of roughly one radiation length,
while its atomic fraction amounts to 3.10-6% for chromium .and -2.5.10-5% for nickel. Ac-
cording to calculations, for an electron fluence of 1019 cm 2, the atomic fraction of the pe.r-
centage amounts to 0.6.10-6 and 3.2.10-6% for chromium and nickel, respectively. Moreover,
the theoretical position of the maximum helium percentage is located at a distance of 1.5
radiation lengths for chromium, and 2 radiation lengths for nickel. The experimentally deter-
mined helium distribution in the package of nickel targets is shown in Fig. 3. The discrep-
ancy between the calculated and the experimental positions of .the maximum can readily be ex-
plained by the fact that the angular divergence of the electron-photon shower leading to a
shift of the maximum toward the surface of the target package has been neglected. It is much
more difficult to explain the discrepancies between the absolute values of the maximum (amount-
ing to factors of 5 and 8 for chromium and nickel, respectively).
First o'f all, it should be mentioned that the maximum recorded for chromium when using
an electron fluence of 1019 cm 2 is commensurable with the experimental error of the IPDO-2A
instrument. In connection with this, the chromium specimens were additionally irradiated with
275-MeV electrons to a fluence of 1021 cm 2. The atomic fraction of the maximum helium per-
centage in this experiment amounted to 26.7.10-5%. With linear extrapolation of this value
to the electron fluence of 1019 cm 2, the atomic fraction of the helium percentage amounts to
0.67110-6%, which is in good agreement with the theoretical value (0.6.10-6% for electrons
with an energy of 225 MeV).
The discrepancies between the theoretical and experimental data on the. helium buildup
in nickel may be caused by the following factors:
accumulation of additional amounts of helium due to the yield of ?He, produced directly
by photonuclear reactions as well as by the decay of the tritium formed in these reac-
U
buildup of helium due to photonuclear reactions on small amounts of impurities (commer-
cially.pure nickel containing certain amounts of carbon and oxygen was used in the ex-
periment);
formation of helium due to secondary (n, a) reactions on neutrons occurring as a result
of photonuclear processes;
helium yield in the a decay of certain unstable nuclear reaction products (radioactive
residual nuclei);
possible inadequacy of the simulator description of the helium buildup process.
Allowance for the additional amount of 3He yields a correction of not more than 10% (see
Fig. 2) and cannot explain the discrepancy between the theoretical and the experimental data.
In addition, in the range of low mass values (up to 29 inclusive), the RMO-4S omegatron tube
has maximum resolving power.
A considerable helium yield is observed in photonuclear reactions on carbon and oxygen
in connection-with the high probability of the photodisintegration processes (y, 3a) and (y,
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
4a), respectively. Thus, the theoretical helium yield from 12C with a photon spectrum char-
acteristic for the development of electromagnetic showers in nickel at a depth of 2 cm cor-
responds to a cross section of 20.10-31 m2. Hence, for a 0.1% mass concentration of carbon
in nickel, the additional helium buildup amounts to approximately 10%, which cannot explain
the discrepancy between the theoretical and experimental data either.
The largest correction - up to 30% with respect to the amount of stored helium - is pro-
vided by an allowance for the secondary (n, a) reactions on photonuclear neutrons. The spec-
trum of photonuclear neutrons that are formed in nickel at a depth of 2 cm as a result of
irradiation with 225-MeV electrons lies in the range from 0.1 to 100 MeV, while the mean .
cross section of the (n, y) reaction for such a spectrum amounts to 35.10-31 m2. Considering
that the dimensions of the target package are much smaller than the mean free path of neutrons
in nickel, we can calculate the additional helium yield by means of the expression
,,,.
~,,,, = 41d,~'rr ~ _
V 4n~r~a
F~n mas
6 (IZ, a) (~~,r) ~Ihn
/'n min
where ~t is the electron fluence, No is the atomic density of the target package, V-is the
total target volume, r is the distance between the point at which the helium yield is .calcula-
ted and some other point in the volume V, dNn/dEn is the differential energy distribution of
photoneutrons, o(n, a)(En) is the excitation function for the (n, a) reaction, and. En is the
neutron energy. Expression (2) has been derived on the assumption of an isotropic three-
dimensional photoneutron distribution. Consideration of the effect of a decay results in an
increase of not more than 1% in the overall amount of stored helium.
Thus, additional helium sources may increase the buildup of helium by 40-50%, however,
even then, there is a discrepancy between the theoretical and experimental data amounting to
a factor of 4-5. Therefore, it would be advisable to perform further investigations in order
to refine the mathematical models and discover other sources of helium buildup.. It should be
mentioned that anomalies in the buildup of helium are also observed when nickel is irradiated
with fast and slow neutrons.
The threshold of the 58Ni(n, a)SSFe reaction on fast neutrons amounts to approximately
2 MeV (it is equal to -4 MeV for the 56Fe(n, a)53Cr reaction), in connection with which, for
a typical neutron flux spectrun in a fast reactor, the (n, a) reaction cross section is much
larger for nickel than for iron, especially if we take into account the competing (n, 2n)
and (n, p) reactions, the thresholds of which are higher for nickel than for iron. For a
neutron energy of 8 MeV, the cross section of the. 58Ni(n, a)SgFe reaction is equal to 57?
10-~1 m2, while the cross section of the S6Fe(n, a)53Cr reaction amounts to only 10.10-ai
z
m
Helium buildup in nickel due to thermal neutron irradiation occurs as a result of the
twa-stage reaction 58Ni(n, y)59Ni(n, a)S~Fe [1]. However, the half-life of S9Ni is equal
to -5000 years, so that it can be neglected. Using the equation for determining the isotope
percentage in radioactive decay chains [9], we obtain the relationship
Nu..lNnl = ~ - ~ [vie-nrryl -Q1e-o,rDl] (3)
rt~-rte
where o, is the cross section of the SHNi(n, ~)'9Ni reaction, oz is the cross section of
the S4Ni(n, a.)SfiFe reaction, ~t is the neutron fluence, and Nge and NNi are the atomic
percentages of helium and nickel, respectively. The value of of in the thermal region of the
spectrum is well known (4.2.10-29 m2), while the value of 6z varies in the (12-14)?10-ze mz
range. The large cross section of the g9Ni(n, a)56Fe reaction is due to its exoergic na-
ture. For the above values of o~ and az, relationship (3) yields a quadratic dependence of
helium buildup on the neutron fluence in the range up to 1022 cm-2, while, for higher fluence
values, the dependence becomes linear. The atomic fraction of helium buildup in nickel reach-
es 0.1%-for a thermal neutron fluence of 1022 cm 2.
In conclusion, it would be useful to indicate the possible scope of application of the
simulation experiments on irradiation with an electron beam. A highly characteristic param-
eter of radiation damage is the ratio of the helium percentage to the number of primary radia-
tion defects (atomic fraction of the helium concentration in percentages to the number of
displacements per atom). Figure 4 shows the behavior of this ratio along the target thick-
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
Fig.. 4. Behavior of the ratio of the number of stored helium
atoms to the number of primary defects during irradiation of
a package of nickel targets with 225-MeV electrons in a BOR-60
reactor (-----) and in a thermonuclear reactor (-?-?-?=),
Hess and provides its characteristic values for a thermonuclear reactor and a BOR-60 fast re-
actor. The. curve was plotted on the basis of experimental data on the helium buildup and cal-
culations of the number of the defects generated. It is evident from Fig. 4 that the same.
package of targets can be used for simulating the effect of different neutron flux spectra.
In conclusion, it should be emphasized that irradiation with high-energy electrons and photons
produces a larger buildup of gaseous nuclear reaction products than irradiation of the targets
by charged particles of any other type.-
1. V. V. Gann, V. V. Rozhkov, and o. V. Yudin, Problems of Atomic Science and Engineering,
Physics of Radiation Damage and Radiation Metallography Series [in Russian], Vol. 3(11),-
(1979), p. 10.
2. R. Blackburn, Met Rev., 11, 163 (1966).
3. V. A. Kuz'menko,B. A. Shilyaev, and V. A. Yamnitskii, Problems of Atomic Science and Tech-
nology, Physics of Radiation Damage and Radiation Metallography Series [in Russian], Vol.
1(12) (1980), p. 18.
4. V. F.. Zelenskii et al., Problems of Atomic Science and Technology, Physics of Radiation
Damage and Radiation Metallography Series [in Russian], Vol. 1(2) (_1975), p. 8.
5.. V. V. Gann, V. A. Yamnitskii, and A. M. Vaisfel'd, Problems of Atomic Science and Tech-
nology, Physics of Radiation Damage and Radiation Metallography Series [in Russian], Vel.
1(5) (1980), p. 39.
6. N. L. Emets et al., KhFTI AN USSR Preprint 72-37, Kharkov (1972).
7. N. L. Emets and Yu. N. Ranyuk, Problems of Atomic Science and Technology, Physics of Ra-
diation Damage and Radiation Metallography Series [in Russian], Vol. 1(9) (1979), p. 31.
8. B. Bulow and B. Forkmen, Rep. from Technical Reports, Series N 156, Handbook on Nuclear
Activation Cross Section, Vienna (1974).
9. N. G. Gusev and P..P. Dmitriev, Radioactive Chains (Manual) [in Russian], Atomizdat,Mos-
cow (1978).
10. V. V. Gann et al., Problems of Atomic Science and Technology, Physics of Radiation Damage
and Radiation Metallography Series [in Russian], Vol. 2(16) (1981), p. 14.
11. V. A. Kuz'menko, B. A. Shilyaev, and V. A. Yamnitskii, Problems of Atomic Science and Tech-
nology, Physics of Radiation Damage and Radiation Metallography Series [in Russian],
Vol. 2(10) (1979), p. 43.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
G. V. Samsonov, S. V. Seredkin, UDC 669.296:669.018.2:620.172.251.296
and V. N. Shulimov
Increased importance attaches to research on the corrosion resistance of zirconium alloys
in superheated steam on account of the increasing use of RBMK reactors in nuclear power, which
make extensive use of these alloys [1].
The numerous data obtained under laboratory conditions cannot finally determine the vi-
ability in reactors, since it has been found [2] that radiation substantially affects the ox-
idation. A major requirement in the tests is that the medium should be constant. However,
the composition of the medium in a closed autoclave alters continuously because corrosion
products accumulate, which means that the. number of specimens tested simultaneously is restrict-
ed by the relation
where m is the mass of a specimen in g and S is the total surface in cm2. This condition is
even more difficult to meet when there is radiolysis.
Therefore, the medium must be constantly changed in order to obtain reliable results, and
this is usually feasible in full-scale reactor loops, but it involves considerable expense.
Also, it is difficult to carry out experiments at elevated coolant parameters because of the
restricted viability of the constructional materials. The safety of the entire reactor system
is also adversely affected.
Ampul devices are more widely used in routine tests at certain stages in radiation-in-
duced corrosion[3]. The main advantages are that there is good measuring equipment, together
with reliability and economy. The safety of the entire installation is also not reduced [4].
Conditions (1) are met in such experiments by means of devices for pumping the corrosion med-
ium continuously through an ampul at a low speed, which is justified by the low corrosion- _._
product accumulation rate.
The Nuclear Reactor Research Institute has devised a suite of ampul devices of this type
for use with swimming-pool-type research reactors RBT, and here we consider some of them.
Open Stem Loop [5]. There are two major components: the ampul containing the specimens,
which is located in a channel in the .core, and a system for supplying superheated steam, which
usually lies outside the core.
As the steam flow is small, the device can be built as an open circuit without adverse
effect on the radiation environment in the reactor, which receives the medium after passage
through the ampul. This eliminates the complicated equipment characteristic of a full-scale
reactor loop.
The apparatus (Fig. 1) consists of two steam generators (SG), a superheater SH, a con-
denser, a line supplying a distillate to the SG, a sampling device, and means of flushing
the ampul with inert gas (helium) before the reactor is run up. All the pipelines are they-
molly insulated and electrically heated.
The working conditions in the heaters in the SG are determined by the pressure and tem-
perature and are controlled automatically. In the SH (autoclave), one can place reference
specimens for testing under identical conditions but without irradiation. The flowrate is
estimated from the rate at which the measuring vessel at the exit from the circuit fills up
and from the change in liquid level in the SG, which is monitored by a level gauge.
The steam is supplied to the ampul when the reactor has been run up and the temperature
in the circuit is higher than the condensation temperature at the working pressure. When the
Translated from Atomnaya nergiya, Vol. 57, No. 1, pp. 25-28, July, 1984. Original
article submitted October 31, 1983.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
Fig. 1. The open steam loop: 1) ampul, 2) re-
actor channel, 3) condenser, 4) steam super-
heater; 5, 6) reserve and working steam genera-
tors; 7) vessel.
Fig. 2: Natural-circulation steam loop: 1) evap-
orator; 2) TET; 3) condenser; 4) liquid; 5) riser
tube; 6) descent tube; 7) ampul; 8) specimens.
distillate has been used up in the working SG, the reserve one is switched on. The apparatus
has been operated for some years with the RBT-6 reactor. .
Compact Natural-Circulation Loop. The device is completely contained in a reactor chan-
nel and does not require additional equipment in the reactor bay. The loop provides for
continuous change in the medium in the ampul, so the composition remains constant. The spec-
imens are tested at a normal steam pressure.
Figure 2 shows the design. The evaporator and condenser are partially filled with liquid
and are at some distance from the ampul. The descending and rising parts of the circuit and
the evaporator are equipped with electric heaters. Under working condition's, natural circula-
tion occurs in the two-phase medium, with the ampul receiving only superheated steam. The
condenser is a vessel connected to a tube brought out from the reactor via the upper flange.
The tube is cooled by natural circulation of helium in the channel. After condensation, the
liquid trickles into the evaporator. The passage of steam through the ampul can be judged
from the readings of the thermoelectric transducer TET: The temperature at this point should
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
Fig. 3 Fig. 4
Fig. 3. Ampul with free specimen setting: 1) heat input tube; 2) central tube; 3)
ampuls 4) TET; 5) specimen; 6) filling tube (CC is core center).
Fig. 4. Special ampul with ring specimens; 1) steam outlet tube; 2) steam supply
tube; 3) heater; 4) upper collector; 5) aluminum block; 6) container; 7) specimen;
8) lower collector.
rise to 100?C. The radiolysis products are vented via the tube into the reactor bay. The loss
of water by radiolysis or leakage is made good via a capillary tube.
In tests-with the RBT-6, it was found that the system was highly reliable and that the
circulation was stable. Topping up was not required during the experiment. The instability in
the specimen temperature was not more than t5?C. The amount of corrosion medium in the cir-
cuit (70-100 cm3) was 10-20 times the amount of medium in a sealed tube, which ensures that
(1) is obeyed with an unaltered number of specimens.
.Design Features of Ampuls for Corrosion Tests in a Flowing Medium. Ampules for irradiat-
ing specimens in a corrosion medium can be divided into two types: universal (Fig. 3), in
which specimens of various shapes can be accommodated (rings, tubes, flat corrosion-test spec-
imens, and flat tensile-test ones), and special ampuls, which are adapted only to specimens
of a particular type, e.g., ring or flat ones. A universal ?ampul does not provide identical
temperatures for all the specimens. The temperature pattern is set up mainly by the natural
circulation of the steam, with its low thermal conductivity (one cannot control the specimen
temperature within wide limits by varying the conductivity of the thermal gap filled with.
helium). Such an ampul is best used in high-temperature tests (T > 500?C) when it is not
necessary to have the entire volume isothermal. One way of reducing the temperature nonuni-
formity is to use a device with specimens of the same type. In such an ampul (Fig. 4), there
is a small gap between the specimens and the walls, and the heat is transmitted mainly by
conduction. To equalize the temperatures, the ampul is placed in an aluminum block separated
from the walls of the channel by an exactly set gas space. In this design, the temperature
difference between a specimen and the wall of the ampul is small (-50?C), so there is improved
scope for adjusting the specimen temperatures by altering the thermal impedance between the
block and the channel (range up to 300?C). These special ampuls can accommodate specimens
of various sizes, for example, rings of two different diameters but at the same temperature.
For this purpose one chooses appropriate widths for the gas and steam gaps. A similar ampul
has been devised for testing flat specimens (Fig. 5). The specimens are placed in a rect-
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
Fig. 5. Special ampul for flat and ring specimens: 1) aluminum block; 2) rect-
angular container; 3) flat specimens; 4) cylindrical container; 5) ring specimens.
Fig. 6. Ampul for examining corrosions under variable temperature conditions: 1)
ampul; 2) specimen; 3) copper leak; 4) conical 12Kh18N10T steel sleeves; 5) reac-
tor channel; 6) conical aluminum sleeves; 7) radiant heater.
angular 12Kh18N10Tsteel container inserted in a longitudinal slot in the aluminum block. An
advantage of this design is that one can equalize the temperature over the height, which is
done by profiling the steam gap in accordance with the power distribution over the height of
-the core.
In addition to the ampuls for use in flowing media at constant temperatures, we also
devised a special ampul providing specimen temperature regulation i.n the range 400-1000?C.
This ampul can be used in examining the corrosion resistance of zirconium alloys under con-
ditions approximating those of emergencies leading to overheating of the fuel-pin cladding.
The device (Fig. 6) consists of an ampul containing ring specimens, a copper leak, and
conical 12KhI8N10T steel sleeves fitted on the leak. The gas gap contains conical aluminum
sleeves whose internal surfaces are matched to the outer surfaces of the sleeves, while the.
outer surfaces match the inner surface of the channel. The specimen temperature is controlled
by varying the gas gap between the sleeves by displacing the aluminum sleeves along the verti-
cal axis. The ampul has been tested in the RBT-6. The temperature control range was 300-
1000?C with a rate of change of 2-4?C/sec.
This suite of equipment for use with swimming-pool research reactors for examining cor-
rosion in zirconium alloys in superheated steam has been operated for several years. It is
planned to use the system also with other reactors..
LITEP.ATURE CITED
1. A. S. Zaimovskii, A. V. Nikulina, and N. G. Reshetnikov, Zirconium Alloys in Nuclear
Power [in Russian], Ener.goizdat, Moscow (1981).
2. A. V. Byalobzheskii, Radiation-Induced Corrosion [in Russian], Nauka, Moscow (1967).
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
3. P. G. Aver'yanov et al., in: Proceedings of the All-Union School on Research Methods
within Reactors (in Russian], Dmitrovgrad (1978), p. 419.
4. B. V. Samsonov, S. V. Seredkin, and V. N. Shulimov, NIIAR Preprint No. 20(428), Dmitrov-
grad (1980).
5. P. G. Aver'yanov et al., NIIAR Preprint No. 56(509), Dmitrovgrad (1981).
PURIFICATION OF THE FIRST CIRCUIT OF NUCLEAR POWER SYSTEMS
WITH A TWO-COMPONENT COOLANT-GAS FLOW
A. V. Beznosov, P. N. Martynov, UDC 621.039
S. Yu. Orlov, and V. E. Serov
During prolonged use, deposits of contaminants develop on the inner surfaces of the first
circuit of reactor systems and in stagnant zones thereof. In general, the contaminants re-
sult from soiled equiment undergoing repair work, contamination by repair work, impurities
.contained in the gas and the coolant of the first filling or introduced in refilling of the
circuit by these media, erosion products of the construction materials of the circuit, prod-
ucts originating from wear of the circuit components by erosion, cavitation, or mechanical
effects, fission products of the nuclear fuel and other fuel-element materials (when their
shells leak), and impurities introduced in the circuit in the recharging of the core or in
repair work. The quantity, the physicochemical composition, the thickness, and the properties
of the contaminant deposits on a particular circuit section depend upon the corrosion of the
construction material at the surface of contact between the solid wall and the coolant, the
crystallization of the impurities from a saturated solution onto the surface, the adsorption
of impurities by the deposits from the coolant flow, the separation of dispersed impurity
particles from the flow toward the walls, the solution of precipitates, and the destruction
of deposits and the subsequent removal of dispersed contaminant particles by hydrodynamic
forces. In the general case, several zones which can be treated in different ways can be
distinguished in the distribution of the contaminants at a wall (Fig. 1).
As a consequence of these processes, the surface of the circuit can be considered a fil-
ter removing impurities from the coolant toward the walls; the absolute mass ratio of the im-
purities retained by the circuit surfaces and the system of coolant treatment deserves the
greatest attention. It has been observed in the operation of reactor circuits of the primary
coolant of nuclear power systems [l, 2] that contaminant deposits develop on the inner sur-
faces in continuous operation of-the coolant-treatment system. This negatively influences the
operational features of such circuits. The surfaces can be purified from the deposits (in-
cluding precipitates weakly adhering to the walls) by mechanical means, i.e., by washing the
surfaces with a large quantity of coolant or with chemical agents, but these methods are ei-
ther little efficient or technologically complicated. It is therefore interesting to search for
and to investigate other techniques of cleaning the circuit surfaces from deposits, to sim-
plify the cleaning of the circuits, to reduce the cost of the cleaning process, and to reduce.
the volume of the liquid radioactive waste material obtained. One of the techniques is based
on cleaning the circuit by introducing a small amount of a gaseous component into the coolant
flow, which gaseous component is neutral or chemically active vis-a-vis the deposits. Sub-
sequently, the two-component mixture enriched with the impurities is discharged onto a filter.
Techniques of flotation cleaning of liquids from dispersed impurities are being used. In these
techniques, gas is bubbled through a large volume of the mixture [3] or gas flows with a high
gas concentration (up to 50% by volume) are employed in special cleaning operations with to-
tal or partial shutoff of the equipment (e.g., in maintenance work). The introduction and
dispersion of a small amount (up to 3% by volume) of the gas into the coolant flow, the de-
struction of deposits more or less strongly adhering to the walls, the attachment of parti-
cles to the bubble surface (wetting and adhesion on the particle-liquid interface and the
particle-liquid-gas boundary), and the transport of the contaminant to filters are the impor-
tant stages of the proposed process.
Translated from Atomnaya Energiya, Vol. 57, No. 1, pp. 29-31, July, 1984. Original
article submitted January 24, 1983..
458 0038-531X/84/5701- 0458$08.50 ? 1985 Plenum Publishing Corporation
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
The proposed introduction of a gas component into the coolant flow can be obtained via
an external gas supply, with subsequent release of the gas into a special gas-treatment sys-
tem (for this purpose one can use compressors--gas blowers), with circulation of the gas within
the circuit (without release of the gas), or with ejectors introducing the gas into the flow
with the energy provided by a jet of the flow proper and gas circulation within the coolant
circuit. It is not desirable to introduce the gas from an external supply (tank, compressor)
because the amount of radioactive products discharged from the circuit is increased, the re-
quired flow of the gas to be introduced is technologically hard to maintain, and the disper-
sion of the gas in the case of real, admissible pressure fluctuations in the circuit is diffi-
cult owing to difficulties in monitoring the development of the purification process. Some
of the above shortcomings can be eliminated by introducing into the circuit a compressor which
makes the gas separated from the flow (e.g., on the free surfaces) return into the coolant
flow. But such compressors still have to be built and the problem of dispersing the gas in-
troduced still must be solved. The ejection of the gas by the coolant flow offers itself as
a better solution (Fig. 2). In this technique basically all systems connected to the circuit
(air vents, treatment of the coolant) are involved; an ejection device, a filter on the mani-
fold of the air vents, and lines connecting these two components are introduced in addition.
The surfaces can be cleaned intermittently or continuously. The coolant which is gathered
in the coolant-purification system is partially directed into the ejector, whereupon the two-
component flow is returned into the main circulation circuit. In this fashion the gas con-
centration is substantially reduced by dilution with the main current. The gas component in-
teracts with the deposits, destroys the deposits, and tears them off and into the coolant
flow by purely physical processes (gas cavitation, etc.) in the ejection of a neutral gas, or
by physicochemical effects in the case of an added chemical agent. Part of the deposit par-
ticles gets close to the gas bubbles and adhesion of the particles to bubbles takes place in
the gas current, whereby the specific surface energy at the interphase boundary decreases.
Another part of the dispersed contaminant particles is caught by the coolant flow proper. The
coolant flow is purified from the contaminants by filters mounted in the manifold of the air
vents and in the coolant-treatment system. The gas (or the vapor-gas mixture) which was sep-
arated on the free level of the coolant in, say, the volume compensator is returned into the
circulation circuit by the ejector. If necessary, a chemical agent which can interact with
-the deposits is introduced into the gas (or vapor~as) volume of the circuit or into the gas
supply line to the ejector.
The introduction of a small amount of gas (up to 3% of the flow volume) does not notice-
ably influence the performance of the main equipment. Even a slight improvement of the per-
formance is possible: The anticavitation resistance of the materials of the pump components
through which the flow runs is increased, the heat dissipation coefficient is slightly in-
creased by additional swirling of the flow [4, 5], and the pressure pulsations in the circuit
(and its vibration characteristics) assume a higher frequency with a reduced amplitude.
An analysis of the changes in the coolant level in the compensator volume is important
for determining the performance of the circuit. The increased coolant level after putting
the system into operation can be determined with the formula
Ggdti-clv-{- Jac vcli, (1)
when we assume that the bubbles are homogeneously mixed within the coolant volume and that
the gas is fully separated; the notation is interpreted as follows: Gg denotes the gas sup-
plied. (mZ/h) to the circuit by ejection, with the supply reduced to the conditions in the
circuit; v (m3) denotes the volume of the gas component in the circuit, with the volume re-
duced to the conditions in the circuit; Jvc (m3/h)denotes the release of two-phase coolant
through the volume compensator; and Vc denotes the volume of the coolant in the circuit. When
we solve this equation, we obtain
C 1'c Jvc C lb .1vc . ( 2 )
NHS-- Jvc ~~ v?e~pf -- Uc i)- ~ eap(-- v~ i),
where S denotes the free surface area (mZ) in the volume compensator, and vo denotes the vol-
ume (m3) of the homogeneously mixed gas in the circuit at the time at which the system is put
into operation. ~ .
The possible gas filling Vn in the upper volumes of the circuit, which are not ventilat-
ed by the air vents, was disregarded in F..q. (2).
a
459
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
It follows from an analysis of Eq. (2) that after sufficiently long time intervals (which
are given by the ratio of the coolant flow through the .volume compensator to the coolant flow
through the circuit), the increase in the coolant level is
AI7S=Cg~cl?jvc (2a)
when the gas ejection is terminated, the time required for reducing the coolant level and for
the practically complete degassing of the circuit (with Vn being disregarded) can be deter-
mined with the formula
T'-~Vc lJ~c (2b)
The hydraulic characteristics of the tube duct and the equipment change only insignifi-
cantly when the gas concentration is as high as 5% and, as expeiments made by the authors on
models of circuit components have shown [6], the changes do not exceed 3%.
The equation for the balance of contaminants in the circuit after putting the system
shown in Fig. 2 into operation can be stated in the following form
clA/di = ~ve-~ ~'wl -- (Axf /Vc -h?a) Jvc - (tl/Vc -!ei) JpG. (3)
where w denotes the amount of contaminants entering as a result of the ejection treatment;
Ewi, total output rate of the contaminant 'sources in the circuit; Jps, flow of the coolant
through the treatment system; kl, concentration of contaminants in the treated coolant behind
the treatment system; Xf, flotation coefficient, with the attachment of contaminants to the
gas bubbles taken into account; ks, contaminant concentration in the coolant behind the filter
on the air-vent manifold; and Jvc, coolant flow through the air-vent system. The solution of
this equation (wherein the output rate of the sources of contaminants is assumed to be con-
stant in time) can be stated in the form
.1 __ 6~ (ue -~- tt~~ -{- /r,J~c -F k~.Ips -I- ~? esh ( -- c ~ i ) . (4)
t'
When the gas concentration in the coolant is constant, we have
ueldi =lcsn,
where k denotes the coefficient of contaminant-mass removal from a unit of surface; s denotes
the area of the deposits in contact with bubbles; and n denotes the number of gas bubbles,
which, on the average, is defined by the ratio of the total gas concentration to the average
static bubble volume. An analysis of Eqs. (3) and (4) reveals that the removal of the con-
taminants towards the filters increases from the time at which the gas is admitted to the
circuit, with k being constant. Obviously, during the purification process, when a certain
time has elapsed, the coefficient of mass removal of contaminants from a unit surface must
decrease and when a normative value of the contaminant mass in the volume of the circuit
(contaminant concentration) has been reached, the purification system can be switched off.
The we value is given by the flow conditions of the two-component mixture and the solid- .
ity of the contaminant deposits, with the solidity depending upon the type of the coolant,
the contaminant sources, and the physicochemical composition of the contaminants. Since the
time of the purification process and the periodicity with which the purification system is
put into operation depend upon the real form of the circuits, the materials used therein, the
conditions of operation of the circuits (power rating, impurities), etc., definite rules for
purification can be established only experimentally on the real circuit. The gas phase in
the coolant flow is of great importance for the purification with the proposed technique.
When the authors made experiments in models with optically transparent materials and
with ejection devices of different orientation, intense coagulation of the bubbles formed by
the ejector was observed on the section of hydrodynamic stabilization [6]. But with bubble
dimensions for which the rate of bubble surfacing is equal to the speed of the pulsating com-
ponents of turbulent flow, a release of the gas to the lower points of the horizontal chan-
nels in the models of the circuit components was observed:
460
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
Fig. 1. Scheme of the qualitative distribution of impurities near
a wall: 1) construction material; 2) oxide film; 3) dense layer of
deposits adhering to the wall; 4) porous impurity layer soaked with
coolant; 5) laminar sublayer enriched with contaminants; 6) core of
the flow.
Fig. L. Scheme of purifying surfaces with a two-component cool-
ant-gas flow: 1) reactor; 2) volume compensator; 3) heat exchang-
er; 4) filter; 5) circulating pump;- 6) inlet of chemical agent;
7) ejecting device.
where vx~y~z denotes the pulsating component of the speed; p', p" and u', u" denote the den-
sity and the kinematic viscosity of the coolant and the gas, respectively; g denotes the grav-
ity acceleration; and R~r denotes the critical bubble radius. When visual observations on
the bubble flow in long channels and in models of the circuit components (curves, narrow pas-
sages, etc.) were compared, a greater dispersion of the gas phase was found in the latter
case.
The authors made experiments in which they purified steel tubes from contaminant deposits
with the aid of an ejection device developing a two-component flow [6]. The experiments were
made at 15?C in an isothermal water circuit with relatively clean surfaces. An experimental
tube section of steel 3 (Z = 300 mm, dint = 20 mm) with extensive deposits of iron oxide was
fit into the circuit. In order to determine the amount of contaminants which are carried
away, a retaining filter was mounted behind the monitoring section; the filter was weighed
before it was inserted, after operation of the test setup without ejection, and after operation
tion with ejection of a gas phase.- The volume concentration of the gas in the two-component
flow amounted to 5% and the water speed to 1.5 m/sec in the experiments. The difference be-
tween the contaminant particle masses carried away by the water flow without admission of gas
and with admission of gas through the ejector was (1.5-3.0)10-3 g/(m2?h). These tests, as
well as other experiments made by the authors (on test stands and on setups under conditions
ruling inside reactors), have shown [6] that a rather .intense purification of the circuit
surfaces from contaminant deposits takes place when the circuits are .treated with a two-phase
flow.
LITERATURE CITED
1. I. K. Morozova et al., Removal and Deposition of the Corrosion Products of Reactor Ma-
terials [in Russian], Atomizdat, Moscow (1975).
2. P. Cohen, The Technology of the Water in Power Reactors [Russian translation], Atomiz-
dat, Moscow (1973).
3. V. A. Glembotskii, Principles of the Physical Chemistry of Flotation Processes [in Rus-
sian], Nedra, Moscow (1980).
4. Heat Transfer in a Two-Phase Flow [in Russian], $nergiya, Moscow (1980).
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
5. K. Sekoguchi et al., ISME, 23, No. 184, 1625 (1980).
6. V. E. Serov et al., The Characteristics of a Two-Component Coolant-Gas Flow and Its Use
for Cleaning Equipment of Atomic Power Stations and Accelerators [in Russian, Reports
of the Gorki Polytechnical Institute No. 81,100,755, Gorki (1983).
CONSTRUCTION GF AUTOMATIC SYSTEMS FOR RADIATION MONITORING OF THE
ENVIRONMENT OF NUCLEAR POWER STATIONS
E.
P.
Volkov,
A. I. Glushchenko,
V.
N.
Durnev,
V. V. Zhabo,
M.
I.
Saparov,
and L. P..Kham'yanov
Providing safety for the population and protecting the environment from contamination
are one of the important problems of modern power engineering, including nuclear energetics.
The accident in the Three-Mile-Island Atomic Power Station (March 30, 1979) has led in many
countries to a reassessment of the safety concepts and technical solutions in the planning,
building, and utilizing of atomic power stations. It has been one of the conclusions of the
report issued by. the presidential commission on the investigation of the accident that the
services for an operational evaluation of the radiation consequences of emissions for the
environment and the population were insufficiently .prepared.
Though the emissions from atomic power stations are usually smaller than those of thermo-
electric stations, the specific form of the operation of atomic power stations necessitates a
careful observation of tl-!e radiation conditions in the environment. Establishing an opera-
tional link between emissions and environmental contamination is an important aspect in a
properly organized radiation-monitoring system. When the radiation conditions in the environ-
ment are determined, two conditions of operation of atomic power stations must be distin-
guished, namely the normal operation or the accident conditions in which the radioactive emis-
sions are monitored by state equipment for radiation monitoring or accident conditions of an
atomic power station with uncontrollable arrival of radioactive materials. In the latter
case, the dose of the radiation received by the population can be determined only at previ-
ously established points in the environment, which are provided with the corresponding equip-
ment.
Such equipment must work automatically and must be connected to a single automatic sys-
tem with transfer of the entire information from the periphery to the center in order to ob-
tain operational monitoring and fast reception of the pertinent information. Since the prop-
agation of contaminants in the atmosphere depends upon the state of the atmosphere, a subsys-
tem for measuring meteorological parameters which define the thermal stratification of the
atmospheric layer close to ground mustform part of the equipment. Besides that, a subsystem
of state radiation monitoring, which checks for gaseous and aerosol emissions from an atomic
power station, must be operationally connected to an automatic "accident-monitoring" system
of the environment of an atomic power station.
The presently used radiation-monitoring systems, which were developed on the basis of
AKRB-03 "Seival-1" and AKRB-06 "Gorbach-1" instruments, do not comprise automated subsystems
for radiation monitoring of the environment. These functions have to be performed by exter-
nal dosimetry services of the atomic power station; these services are autonomous as far as
the territory and the technical means are concerned. The division of the functions causes
difficulties in the operational monitoring of the radiation conditions outside an atomic pow-
er station.
Systems for the automatic monitoring of the environment of atomic power stations have
been developed and are presently used in a number of countries. The best of the foreign.sys-
tems comprise the following components:
measuring instruments mounted at various points inside and outside *_he atomic power sta-
tion for determining radiation parameters and meteorological and hydrological parameters (the
Translated from Atomnaya Energiya, Vol. 57, No. 1, pp. 32-34, July, 1984. Original
article submitted October 31, 1983.
462 0038-531X/84/5701- 0462$08.50 ?1985 Plenum Publishing Corporation
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
monitoring techniques and the points of the measurements depend upon the type of the atomic
power station and the requirements of the user service);
a monitoring center with a reliable system reserve; the center is computer controlled to
read out the measured values, to process and store the data, and to output the required in-
formation;
a display on the control panel for representing operational information on the radiation
conditions in the area close to the atomic power station; and
a spectrometer measuring emission from humans so that incorporated gamma emitters can be
determined.
The authors of [2] have described an emergency system for recording and analyzing the
irradiation doses in and around atomic power stations in the direction of the possible propa-
gation of accidental emissions. The system comprises nine remote dosimeters with a measuring
range of 1 uR - 10 R/h (1 R = 2.58.10-A C/kg); the dosimeters are mounted at distances of 8-
1G km from the atomic power station. The system also comprises a remote meteorological station ,_
a minicomputer, and a set of color displays for projecting isotope maps along the propagation
path of the radioactive cloud. This system is connected to the main radiation-monitoring sys-
tem of the atomic power station. A mobile radiometric laboratory with a microcomputer on
board is provided for additional operational monitoring.
An automated system for monitoring the radioactivity of the environment has been devel-
oped and implemented in the atomic power station "Paksh" in the Hungarian People's Republic[3].
The experience gathered in the use of such systems has shown that in addition to current
data, efficient monitoring of the operation of an atomic power station requires average data
on the dose load near the station in the form of daily, monthly, and annual values. The devel-
opment of a corresponding computational model is an important task in the mathematical work of
emergency-monitoring systems. In normal operation, extensive calculations are made once per
day and the daily dose load is determined. Computed daily, monthly, quarterly, and annual
dose loads are kept in the computer memory. Curves of the isodose distribution around the
atomic power station are constructed on tiie basis of those data for various time interv~~i.s.
From the above considerations we can formulate the following tasks'of an automated en-
vironmental radiation_monitoring system of an atomic power station:
.During normal operation of the atomic power station, the system must sample, record, and
process information on the radiation conditions in the region of the station, determine its
normal state, and, if necessary, transmit the results of the information processing to a com-
puter; and
in the case of "n accident, the system must output instantaneous data for taking opera-
tional measures, among them measures providing for the radiation protection of the population.
In addition to the tasks listed, statistical data which are accumulated in the operation
of such systems make. it possible to refine the calculation of the concentration of c.ontamin-
nants entering the environment under various conditions of operation of the atomic. power sta-
tion; in these calculations, the changes in the meteorological parameters and the influence
of the buildings of the industrial area upon the propagation of emissions are taken into ac-
count under the condition that the ventilation tubes of the atomic power station are not very
high.
The first experimental automated system which in the USSR has been taken into operation
to monitor the contamination of the atmosphere by emissions from a thermoelectric power sta-
tion [4] was developed by the Moscow Power Institute in collaboration with Institutes of the
Academy of Sciences of the Ukrainian SSR [5] and the Board for the Protection of Nature at
the Department of Energy of the USSR and other organizations. The system comprises three main
subsystems: one for monitoring the emissions from the thermoelectric power station, one for
monitoring the meteorological parameters, and one for monitoring the contamination of the en-
vironment. Information obtained from these subsystems is automatically transmitted with an
averaging interval of 20 min to the computer of the information-processing/computing system
located in the electric power station. This system performs automatic sampling of information
on the gas pollution of the air by the emissions of the Zaporozhe Electric Power Station with-
in a radius of 25 km and transmits data to a center controlling the system and monitoring both
the emissions proper and the meteorological parameters in a 300-m-thick layer.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
The system for external-radiation monitoring of an atomic power station must be organized
in analogous fashion. The subsystem for the emissions from the atomic power station must moni-
tor radioactive rare gases, long-lived aerosols, and radioactive iodine; the subsystem moni-
toring the environment must measure the dose rate of the y radiation, and the timewise inte-
gral concentration of radioactive iodine and long-lived aerosols.
The meteorological state of the atmosphere close to ground is an important factor in the
operational monitoring of the radiation conditions around an atomic power station. The tem-
perature gradient and the wind speed at variouslevel;~are important characteristics in this
case. Resistive thermometers or thermocouples mounted at various heights are conveniently
used to determine temperature gradients. Special meteorological towers are set up to deter-
mine the wind speed or a net of meteorological ground observations with readings of the wind
speed on the level of a wind vane are employed. In the latter case, for the purpose of cal-
culating the wind speed in high layers and for estimating the wind speed on the level of the
ventilation-tube's mouth, one must assume some dependence of the wind speed upon the altitude,
and this assumption reduces the accuracy of the input information. However, rather great dif-
ficulties in building automated systems for monitoring the environment can be encountered in
the technical realization of continuous meteorological observations made at various altitudes
in the large number of areas of atomic power stations.
The mathematical provisions of the system are based on the algorithm expounded in [6, 7]
for calculating the radiation conditions around an atomic power .station, because this algorithm
requires only observations made by a net of ground s*ations for the meteorological informa-
tion input. In order to increase the efficiency in the use of the data, the system monitoring
the environment must include an automatic weather station relaying the meteorological informa-
tion obtained directly to a peripheral device (magnetic tape or disk) of a microcomputer.
Such weather stations are now serially produced. The monitoring of the meteorological parame-
ters obtained by ground observations must include at least the wind speed and the wind direc-
tion, precipitation data, and the air temperature at two or three points at the altitude (10
m) of the wind vanes.
The serially produced "Post-1" stations are conveniently used for external radiation mon-
itoring. Information obtained from sensors monitoring the emissions and discharges of the
atomic power station and meteorological data obtained from weather stations can be transmit-
ted over conventional cable channels, whereas the information from the external stations is
transmitted through (telephone) lines or radio links to the center which samples and processes
the information. But a reliable radiation monitoring of a possible contamination of the envir-
onment makes it necessary to increase both the sensitivity of the detectors and the amount of
information they provide. For example, one must develop sensors for continuously monitoring
the accumulation of 131I in all its physicochemical forms (molecular iodine, aerosols, methyl
iodide, etc.).
The construction of automatic radiation-monitoring systems and their implementation in
nuclear power generation in the USSR, first of all in atomic thermoelectric centers and atomic
heating stations near huge cities will help to determine within extremely short times the ra-
diation conditions around these objects during their normal use and in possible emergency
situations when radiation effects on the nearby living population may exceed the admissible
levels which are laid down in valid norms.
Emergency-monitoring systems, whose development and implementation are provided within
a joint special-purpose scientific-technological program of producing automatic monitoring
systems for the thermal power of atomic power stations, will increase .the reliability and
safety of nuclear power generation. This will expand the scale on which atomic power genera-
tion can be introduced in the national economy of the USSR and of the member-nations of the
COMECON.
1. Nuclear Power Generation, Man, and Environment [in Russian], Bnergoizdat, Moscow (1981).
2. D. P. Serpa,A. M. Walker, and T. A. Jenckes, IEEE Trans. Nucl. Sci. NS-28, No. 1, 236
(1981).
3. Sh. Dehme, in: Abstracts of.the Reports of the 2nd Int. Conf. of the COMECON Member-
Countries on Radiation Safety' of Atomic Rower Stations [in Russian], Press of the Inst.
of Physics of the Academy of Sciences of the Lithuanian SSR, Vilnius (1982).
4. E. P. Volkov, V. V. Zhabo, and M. I. Saparov, Teploenergetika, No. 11, 913 (1980).
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
5. A. N. Shcherban', A. V. Primak, and V. I. Kopeikin, Automatic Systems Monitoring the Con-
tamination of Air [in Russian], Tekhnika, Kiev (1979).
6. N. G. Gusev and V. A. Belyaev, in: Atomic Power Stations [in Russian], No. 4, Energoiz-
dat, Moscow (1981), p. 137..
7. A. I. Glushchenko et al., in: Atomic Power Stations [in Russian], No. 6, Energoizdat,
Moscow (1983), p. 47.
EFFECT OF ALUMINUM, BERYLLIUM, AND TRIVALENT CHROMIUM NITRATES ON
THE EXTRACTION OF URANYL NITRATE WITH A 10% SOLUTION OF TRI-n-
BUTYLPHOSPHATE IN n-PARAFFINS
B. S. Zakharkin, T. A. Rumyantseva, UDC 541.123:4:546,791.6:546.45.621.7.63.
and D . P . Adaev
The effect of salting-out in liquid-liquid extraction systems is caused by the fact that
hydration of the ions of the salting-out-agent leads to a reduction of the amount of free wa-
ter in the aqueous phase, an increase of activity of the substance being extracted (activity
'coefficient), and, consequently, to an increase of. the distribution coefficient I). .The in-
dividuality of the salting-out agents is.manifested in their effect on the activity coeffi-
dent of the distributing salt. .A detailed review of the problem of salting-out was published
previously [1, 2]. .
In the present paper, the effect is considered of nitric acid salts of beryllium, alum=
inum, and trivalent chromium on the. extraction of uranyl nitrate with a 10% solution of tri-
butylphosphate (TBP) in n-paraffins. According to [l, 2], the relation between the thermody-
namic properties of aqueous solutions of electrolytes and the salting-out action of salts dur-
ing the distribution of the substance between the organic and aqueous phases can be described
quantitatively by the equation
D = KE'o (Pe~3,
where K is the extraction constant (for uranyl nitrate in the case of extraction with TBP,
K = 1.4.10'); p' effective surface density of water molecules in-the first coordination
e
layer of all cations present in the aqueous solution; 3+'e =2[H+]pg~03 + Ep'~(Ci]Zi; P~HN09 -
0.06; Z, cation charge; [G], salt concentration; i, Al Be +, Cr Ea, active concentration
of extractant in the organic phase (Ea = Eo)/1 + Eo[H+]aq Eo is the concentration of "free"
extractant,-for nitric acid solutions and 10% TBP calculated by the equation.
0.35-2 [U1 org
E0 -1-}-8.1 [HNO9)undiss.-}-0.8 [HNOs~ndiss.
where [U]org is the equilibrium concentration of uranium in the organic phase).
The degree of dissociation of nitric acid a is found, according to the data of [1, p.
100]; for the effective acidity:
[H+leff? (IH+laq [NOs]aq )'2[31;
[HN03]undiss- a[Hl\03J aq?'
[HN09]aq is the equilibrium concentration of nitric acid in the aqueous phase; q is the ef-
fective solution number, calculated by the e~uation q = 2 - (0.06[H+]aq/1 + 2.0.06[H+]aq);
[H+]aq is the equilibrium concentration of [H ] ions in the aqueous phase.
In this case, when the values of p' of-the cations present in the solution are known,
the effect of salting-out on D can be .described by means of the equation quoted above. In
[1, 2], the value of p' is quoted only for A13+ (0.055); values of p' for Be2+ and Cr3+ are
absent. In order to determine p'Be and p'Cr, the effect of the nitrates of these elements
Translated from Atomnaya ~nergiya, Vol. 57, No. 1, pp. 34-36, July, 1984. Original arti-
cle submitted August 18, 1983.-
0038-531X/84/5701-0465$08.50 ?1985 Plenum Publishing Corporation 465
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
TAELE 1. Distribution of Uranium in the
466
1,0
1,0
i,U
1,0
1,0
1;0 i
i ,0 ',
1,4
1,0
1,4
1,4
1,4
1,4
1,4
2,0
2,0
2,0
1,0
1,0
1,0
0,5-
0,5
0,7
0,7
1,0
1,2
0,25
0,5
0,75
0,92
0,5
0,75
1,0-
0;83
1,0
1,3
Be(N03)2--Cr(N03)3-ii20 (values of D for
uranium during extraction from solutions
containing aluminum are taken from [5])
System 10% TBP-~IN03-UOz(N03)z A1(NO9)3
Aqueous phase mole/liter
Distribu-
tion coeff.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
on the extraction of uranium with a 10% solution of TBP from nitric acid solutions was studied.
The experiments were conducted in separatooy funnels at room temperature (20-22..?C), times of
contact of the phases 3-5 min and with a phase volume ratio equal to unity. After separation
the organic and aqueous solutions were analyzed for the uranium content by the method of j4],
and for nitric acid (potentiometric titration). It was established beforehand that beryllium
and trivalent chromium-are almost unextracted with tributylphosphate; for example, the dis-
tribution coefficient of Cr(III) _ (2-6)?10-' with [HN03] = 1-4 mole/liter and [Cr(N03)3_] _
1-1.5 mole/liter.
From the results obtained on the distribution of uranyl nitrate in the systems UOz(N03)z-
Be(N03)z~N03~iz0-10% TBP, and U0Z(NOs)z-Cr(N03)3-HNOs-Hz0-10%-TBP, the values of p' were cal-
- culated for beryllium and chromium, equal to 0.08 and 0.048, respectively.
Fromm a comparison of the-known values of p' for a number of cations [1, 2], it follows,
that beryllium is the most powerful salting-out agent of all cations, for which the value of
p' is known, and chromium is a less powerful salting-out agent than beryllium and aluminum
(the extraction potentials of beryllium, aluminum, and chromium are 6.45, 6.0, and 4.7, respec-
tively.
It can be seen from. Table 1 that the experimental values of D for uranium and the values
calculated by the equation-given above, coincide satisfactorily (deviations do not exceed
?12%). This corresponds to the results obtained in [6]. Hence it follows that with the con.-
di:tions of accuracy of the experimental values of D, the method of [1, 2] allows the ability
of-.the cation toward hydration (p') to be determined quantitatively with sufficiently hi4h
accuracy. If the values of p' present in the solution of cations are known, then the distri-
bution of uranium during extraction with TBP can be calculated over a wide range of compositions
of 'the aqueous phase. The equation can be used. for forecasting the extraction equilibria liy
means of a computer, during the extraction of uranium with tributylphosphate or other neutral
extractants from aqueous solutions of a complex salt composition.
LITERATURE CITED
1. A. S. Solovkin, Salting-Out and the Quantitative Description of Extraction Equilibria
[in Russian], Atomizdat, Moscow (1969)..
2. A. S. Solovkin, Salting-Out and the Quantitative Description of Extraction Equilibria [in
Russian], Results of Science and Technology. Series Inorganic Chemistry, All-Union Insti-
tute of Scientific and Technical Information, Vol. 3 (1972).
3. A.. S. Solovkin, Zh. Neorg. Khim., 15, 1914 (1970).
4. V. K. Markov et al., Uranium, Methods of Determination [in Russian], Atomizdat, Moscow
(1964).
5. V. V. Revyakin, V. V. Chubukov, and N. A. Korableva, Zh. Neorg. Khim., 13, 3090.
6. I. Shilin and T. Rumyantseva, in: Proceedings of the Tripartite Symposium "Reprocessing
of Spent Nuclear Fuels," Held a Mol, 17-19 May, 1978, p. 1.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
NEUTRON PARAMETERS WHICH CAN BE REACHED IN FAST URANIUM BLANKETS
OF HYBRID FUSION REACTORS
Great possibilities can be expected from the use of a source of thermonuclear (fusion)
neutrons for the fission of heavy-element nuclei and the simultaneous production of nuclear
fuel. On the one hand, depleted or natural uranium and thorium can be used in the blanket
of a fusion reactor so that the cost is reduced. The problems associated with the criticality
of the blanket do not arise. Mainly 238U or z3sTn burn in the blanket and, owing to the im-
proved neutron balance, much more nuclear fuel is produced in the blanket of hybrid fusion
reactors than in nuclear reactors. On the other hand, the multiplication of the energy in the
fusion reaction improves the energy balance of the reactor so that the complexity and, accord-
ingly, the cost of the fusion part of the unit is justified.
Attempts to make use of these advantages are often based on the limit parameters of the
blanket neutrons, though such parameters are not compatible with all the limitations of real
units. The goal of the present article is to evaluate the neutron parameters which must be
reached in real blankets and to compare the nuclear fuel under various requirements to the
blanket .
In order to reduce the number of possible versions, one must immediately introduce sev-
eral assumptions which do not essentially affect the conclusions but imply restrictions on the
treatment of the main points. The assumptions are as follows.
1. We do not consider symbiotic systems of fusion and fission reactors in which 239Pu
or z33U is produced in the fusion reactor while tritium is produced in the fission reactor.
It follows from the considerations below that the parameters of a hybrid fusion reactor are
adequately characterized by the total production of useful isotopes and that the system clos-
ure obtained through tritium breeding does not substantially change the optimal parameters
and limit parameters.
2. The considerations are referred to the initial .moment of time and the changes of
the parameters during the operational period of the reactor are disregarded. Corresponding
corrections can be made separately and can be important for certain versions (for example,.
in the case of a thorium fuel). The changes of the basic parameters amount to 20-30% in the
majority of blankets with uranium fuel during an operational period corresponding to a neutron
load of 1.5-3(MW?yr)/m2.
3. Blankets with enriched fuel, which resemble critical blankets, are disregarded. In
such blankets, the coefficient of breeding fissile material and the energy multiplication in
the blanket, both referred to a single fusion neutron, can be much greater. But the breeding
of fissile materials per unit of thermal power of the blanket is reduced, as will be shown
through a comparison of fuel consisting of natural uranium and .enriched uranium. The specific
load of the fuel elements of the blanket has values close to the limit value for the presently
assumed plasma parameters of fusion reactors when natural or enriched fuel is considered.
An enrichment can make sense in the blankets of fusion units in which the specific neutron
load on the first wall amounts to 0.3-0.5 MW/m2 in accordance with the plasma parameters which
can be reached. But few such units are possible and as far as the economic aspects are con-
cerned, these units are inferior to units with higher specific load values.
4. We exclude from our considerations the so-called blankets with suppressed fission
(see, e.g., [1]) in which the neutron multiplication takes place mainly through (n, 2n) reac-
tions on beryllium or lead and .where-the breeding of the fissile isotopes takes place on a
thorium salt with continuous removal of Zg3U (or protactinium).. Such blankets have a very
small energy multiplication coefficient and fissile-material breeding coefficient which is
Translated from Atomnaya Energiya, Vol. 57, No. 1, pp. 36-41, July, 1984. Original arti-
cle submitted August 1 ,1983.
0038-531X/84/5701- 0468$08.50 ? 1985 Plenum Publishing Corporation
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
two times smaller than that of the fast uranium blankets with nonenriched tuel consiaerea
in the present article.
G_eo_metry of the Systems under Consideration. Calculations have been made for a number
of different blanket compositions in one-dimensional geometry. An infinite cylinder with a
radius of 300 cm of the central zone, uniformly filled with distributed neutron sources of
the D-T fusion reaction, was assumed in the calculations. The source was assumed to be iso-
tropic and monochromatic with an energy of 14.1 MeV. This geometry rather closely describes
both tokamak-type fusion reactors and linear systems.
Calculation Method. The calculations were made with the Monte Carlo method in combina-
tion with the P, approximation; the BLANK program [2j with a 52-group [3] or 21-group [4]
system of combined constants was employed. For the majority of elements, the 52-group system
which describes the interaction of the elements with neutrons at high energies was based on
the ENDL-76 library of assessed data.
Calculations with the Monte Carlo method were usually made for 5000 neturon histories.
The relative error of the majority of the integral parameters (total neutron source, breeding
coefficients of tritium and the fissile isotope) amounted to 1-2%; the relative error was 2-4%
for the rate of the (n, 2n) and (n, 3n) reactions.
As far as neutron multiplication and the increased output energy per neutron of the
source are concerned, the layers of the material undergoing fission by thermonuclear neutrons
(i.e., z3eU or 232Th) are most conveniently disposed in the region adjacent the plasma. The
thickness of that region must be chosen so that a neutron of the source makes at least one
inelastic collision with a probability close to unity. The probabilities of the first 2-3
inelastic collisions determine the extent of the expected neutron source in mixtures of vari-
ous materials. Actually, all basic characteristics of the blanket are proportional to the
effective source value: the total coefficient of. isotope breeding, the number of fissions,
and the energy multiplication coefficient. The basic dependencies of the integral parameters
of blankets with a complicated structure can therefore be established by investigating blan-
kets of infinite thickness and a composition corresponding to the composition of the first
zone of real blankets.
The goal of the present work is to analyze the integral characteristics of a blanket.
The following parameters are employed: KT, which denotes the coefficient of tritium breeding
per neutron of the thermonuclear source; Kpu and Kzs31J., which denote the coefficient of breed-
ing fissile nuclei per source neutron in media containing 238U and z3zTh, respectively; KE _
KT + Kpu + K1,~U - Kdepl+ which denotes the yield of useful isotopes per thermonu2~eac ne9-
tron, with Kdepl denoting the number of capture reactions (including fission) on U, Pu,
and z3sU referred to a single source neutron; E denotes the energy liberated in the blanket
per source neutron;
4 (rrFiee ~"fiss ~
/ t
rCa P~ 1: C3p1 _. rr n. ?u\w~, 3u --lr u. sttQu.:; ~,),
where nl z nl 3 nl and nca t denote the number of. (n,,2n), (n, 3n), fission, and cap-
ture reactonsnatnthefisth isotopep respectively; Qn zn and Qn~3n. energy of the (n, 2n) and
(n, 3n) reactions, respectively; E~iss and Ecapt+ energies liberated in fission and capture
reactions, respectively; M = E/Eo, coefficient of energy amplification in the blanket; Eo,
energy of a source neutron; y = (KE - 1)/E, coefficient of breeding useful isotopes, with the
coefficient referred to the unit total power of the blanket and the need for breeding tritium
in the same blanket and the burnup of zssU in the initial fuel composition taken into consid-
eration; and Q, total number of neutrons in the blanket, referred to a single thermonuclear
neutron [Q = 1 + nn,~n + 2nn,3n + (v - 1)nfissl?
The neutron-physics characteristics of the reactor are usually employed in the initial
economic estimates. The reactor characteristics are divided into two groups on this occasion.
The first group comprises KT, Kpu, and E which describe the yield of the isotopes and the
energy in the normalization to unit power of the thermonuclear reaction. These quantities
are of overriding importance in systems in which the cost of the thermonuclear part of the
system forms a significant fraction of the total cost or in systems in which the limitations
are given by obtainable specific power of the plasma string and, accordingly, by the neutron
load on the first wall. In such a case, the cost of the electric energy generated by the
system and of the nuclear fuel is inversely proportional to Kpu and E.
~, Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
The second group comprises the coefficient y of breeding useful isotopes; this coeffi-
cient characterizes the yield of nuclear fuel per unit of thermal power of the reactor.- The
actual meaning of this coefficient resembles that of the breeding coefficient
Kb--1
13G =_ ,
nfiss
where Kb denotes the plutonium-breeding coefficient normalized to one. fission event. The BG
value is usually employed to characterize the quality of a fast breeder reactor. In blankets
in which nfiss ' 0.3, y and BG differ by at most 15%, whereas in systems in which nfiss is
small (e.g., in thorium blankets), the difference reaches 100% and y should be used as the
characteristic quantity. In evaluating the economic aspects of systems in which the cost of
the thermonuclear part is low or the specific energy intensity of the. blanket rather than the
power of the plasma string is the limiting quantity, y is of prime importance. In the limit,
the cost of nuclear breeders is inversely proportional to y.
Idealized Blankets. The parameters of infinite single-component blankets with a fissile
material are listed in Table 1. The best characteristics in regard to the parameters Kfiss
and E are found. in a blanket containing metallic uranium fuel. In 238U, .1.02 fission events
take place and 4.2 neutrons are captured with subsequent transition into 2sePu per source
neutron. The energy amplification coefficient M is 16. In 23gU the maximum value of the condi-
tion y = 1.4.10-2 MeV-' is reached among all uranium blankets; the maximum value corresponds
to a 239Pu-breeding in the amount of 1.1 kg/(MW?yr) of the thermal power of the blanket. The
coefficient y decreases during the operational period because the average number of fission
neutrons decreases while the enrichment of the fuel by plutonium proceeds.
Data for infinite uranium and thorium. media have been cited in [5, 6]; in the case of
2saU [6], the energy m-iltiplication coefficient is 16.5 and the total neutron capture coef-
ficient. is 4.4. These coefficients are in rather good agreement with the results of our work.
One can try to improve, in a certain sense, the blanket parameters by using natural
uranium in place of z3eU. In this case the plutonium yield is increased to 4.75 nuclei per
source neutron and the total number of fissions is raised to 1.44. But a simultaneous burnup
of z3sU takes place and K amounts to 4.4,:whereas M is increased to 22.5. Thus, the use of
natural uranium increases the yield of the fissile isotopes from the blanket by 5% and the
energy liberated in the blanket by about 30%. At the same time the coefficient y decreases
to 1.07.10-2 MeV-' which implies an isotope breeding of 0.83 kg/(MW?yr), i.e.. the breeding
is smaller than that in the case of z3eU by 17%. The question of whether uranium should be
used and the question of the possible initial enrichment of the fuel in the blanket can be
solved only when all other limiting factors of a specific thermonuclear system are taken into
account.
The total neutron flux at the boundary with the source is 30 cm-2?sec-' at a neutron flux
from the source amounting to 1 cm 2?sec-' on the inner surface of the blanket. It is inter-
esting to note that the fraction of neutrons with more than 5 MeV amounts to ?5%, i.e., the
number of displacements of atoms in the blanket materials of hybrid fusion reactors differs
only slightly from the corresponding number in fast breeder reactors when the neutron flux
is the same. However, 5% of the source neutrons suffice for increasing the rate of hydrogen
and helium formation in the blanket by 1-2 orders of. magnitgde.
Uranium compounds are inferior to the metallic fuel in regard to the parameters which
are referred to the neutron, source. When nuclei of elements which do not undergo fission
are introduced in the blanket composition, the probability of a primary collision of source
neutrons with zseU nuclei decreases and this, in turn, linearly reduces the number of fission
events and the breeding of fissile isotopes in the blanket. The smallest gain (..10%) i_n all
parameters is observed in the case of uranium silicide. 'The characteristics of uranium oxide
are below those of metallic uranium by a factor of. 1.7-.1.8. In units in which the limita-
tions are associated with the specific thermonuclear power, the use of uranium oxide leads
to a sharp deterioration of the parameters of the fusion reactor.
The yield y of useful isotopes per unit thermal power of the blanket changes only slight-
ly in all the uranium compounds. The maximum advantage obtained in the case of U02 over me-
tallic uranium is 12% and when the specific energy intensity of the blanket is the limiting
factor, the advantages of the various uranium compounds depend upon the limitations which
characterize the corresponding fuel elements in regard to temperature and intensities.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6 _~~
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
TABLE 1. Neutron Balance of Infinite Single-Component Media of Fissile Materials
Reaction and output energy
I Unat
l ,,{.;~~
IUgatSi
I ~.{"?"~~
IUnatC
I s:teUC
I UnatC2
I 2ssUpz
zszTh
ii (u, f)
aself(n, ,~)--~xsxPu-h'pi~
11,44
/f,75
9,12
4,2f1
9,37
4
44
0,93
87
3
1,07-
3
96
0,80
3
57
0,78.
3
!.5
0,60
77
2
ssTh (n, O
-
--
,
-
,
,
,
,
,
-
ssxTh (n, ~,) -r ~salJ- hsaaU
--
-
-
--
--
-
-
-
_
_
0,'20
sasl; (n, V)
Capture by nonfissile material
0,06
-
--
0,0(i
-
0,05
-
-
0 ~
,46
All (n, 2n) reactions
-
0,3:3
--
0,3:i
0,0,3
t',33
0,076
0,3,?
0,!!1
O,:i9
11,09
0
31
0,18
0
23
0,18
(1
33
0
59
All (n, 3n) reactions
/)
's"~J (~1
0,18
0,18
0,1;>
0,91;
0,94
,
0,14
,
0,11
,
0,11
,
0,25
,
ssNU (n, j) at an energy > 6 MeV
l,9ti
0,57
1,i 12
O,Ci7
1,07
__.
0,93
--
0,89
O
S,!
0,8U
0
49
0,(i4
0
38
0,(i;)
0
38
aasU (n, 1)
Total energy E (MeV)
0,28
0
:3
--
0,3U
-
,
(1;18
,
--
,
0,14
,
zsaL! (12. },)ssr,li (n. /)---_ssJl(~~, p)._ h".?
'
~
y
2
4,41
230
4,20
300
4,n8
290
:3,87
240
3,75
990
:3,57
18n
2,89
145
L,77
60
46
2
pi~ 9)/L
(h
, 10'- McV-t
(tiz.._'~)/h', 4!!' McV-t
1,1G
1
07
1,39
I
;f!!
i.ICi
u3
I
1,37,
:37
1
1,23.
1
15
1,35
1
35'
9.13
1
1,22
'
,
2,46
Total source
,
fi,24
,
5,21
,
I;,iiU
,
4,88
,
5,l!.v
,
4,39
,!.5
4,04
1,
LL
3,54
2,46
2,66
Shield thickness (cm)
Fig. 3
ExZU, EFB
-Fig. 1
0,1 0,J 0, J -0 7 0, 9
eLl , ENtO
Fig. 2
Fig. 1. Parameters of infinite media of 23BU + Fe (-----) and of 23aU + H2O ( ).
Fig. 2. Parameters of an infinite medium consisting of a mixture of 23BU and lith-
ium: ) 23BU + natural lithium; -----) zseU + 6Li.
Fig. 3. Parameters of an infinite medium of 23BU with a shield of natural lithium
( ) or of Fe (-----) between the medium and the source.
An even higher y value is observed in the case of thorium. In a blanket of metallic
thorium, 1.9 kg%(MW~yr) 233U are generated by breeding at the beginning of the operational
period. But the KE and E values of a thorium blanket are much lower than those of an uranium
blanket and thorium cannot be used for the blankets of utiits in which a substantial amount of
energy is spent for maintaining the fusion reaction.
Figure 1 illustrates the influence of the lithium-containing materials, construction ma-
terials and of the coolant upon the blanket parameters. Figure l shows the dependencies of
the parameters Kpu and nfiss upon the fractions of the various materials in a 23BU mixture for
a homogeneous blanket of infinite thickness. The greatest effect results from the incorpora-
tion of iron in an uranium blanket. Kpu and nfiss change at a rate of til% per volume percent
of iron in the mixture. This is a consequence of the fact that the macroscopical cross sec-
tions of -the inelastic processes of uranium and iron are approximately equal at the energy of
the neutron source. The blanket parameters decrease in almost the same fashion in an uranium-
water mixture but in this cage the mechanism of the nfiss decrease depends considerably upon
the decrease in the probability of the 23BU fission by .the neutrons of the fission spectrum.
A homogeneous mixture of uranium with lithium (Fig. 2) can be used in compositions of
hybrid blankets in which the maximum value of KT must be obtained. The KE and nfis values
change in about the same fashion for the natural mixture of the lithium isotopes an~ pure 6Li.
An increase in the capture by lithium causes the maximum KT in a mixture consisting of equal
volume fractions of 23BU and 6Li. The maximum KT value is 2.93 and seems to be the absolute
value for the nonenriched compositions. When lithium is introduced in an uranium blanket, KE
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6 ~i
- i
and n change much more slowly than when iron is introduced. Even with a 70% volume frac-
tion o~sslithium K~ decreases in comparison with pure uranium by 20% in the case of natural
lithium and by 30% in the case of 6Li. Thus, a homogeneous dilution of the uranium by litfii-
um does not greatly reduce the properties of the blanket.
In order to assess real blankets, one must also know how the blanket parameters are in-
fluenced by the layers of the various materials located between the neutron source and the
uranium region. Data for shields consisting of iron and lithium and inserted before an infi-
nite uranium medium are illustrated in Fig. 3. The decrease in K .amounts to .1.5%, on the aver-
age, per 1 mm thickness of an iron layer and to 0.3% per 1 mm o~ natural lithium. For a 10-
cm-thick izon shield, the number of fission events per thermonuclear neutron is still 0..19,
whereas the coefficient of breeding useful isotopes, K~ = Kpu, amou nts to 1.53; this means
that it is basically possible to manufacture a breeding blanket even in the case of a thick
steel shell before the uranium region. The coefficient y remains rather high (0.88.10-z MeV-'l,
and the breeding of plutonium amounts to 0.7 kg/(MW?yr) at the beginning of the operational
period.
The coefficient y of a shield of natural lithium increases with increasing shield t'tiick-
ness up to 50 cm. This effect is caused by the formation of tritium in inelastic scattering
of neutrons at 'Li nuclei, whereby the neutron balance in. the reactor is improved. In the
case of a 50-cm-thick shield, y amounts to 2.8.10-z MeV-t, which, in principle, makes it pos-
sible to reach a plutonium breeding of up to 2.0 kg/(MW?yr) in the blanket under consideration
when some depletion of lithium in regard to 6Li has taken place. The KE value is two times
smaller than in an uranium medium without shield. This is still acceptable, but the E value
is reduced by a factor of 6. Such a blanket is obviously economically adequate in systems
with a high neutron load on the first wall. The blanket can be particularly interesting for
pulsed fusion reactors.
The limited thickness of the uranium region has an important influence in real blankets.
Thee scale of the effect can be inferred from Fig. 4 which shows the dependence of the parame-
ters of a blanket (consisting of a 238U layer and an infinite 6Li reflector) upon the thick-
ness of the uranium layer. A 6Li reflector for uranium blankets with a hard neutron spectrum
provides the greatest KT value which is practically equal to the outflow of neutrons from a
layer of pure uranium. The highest KT value (2.65) is obtained at a thickness of 7 cm of
the uranium region. But at the maximum of KT, the coefficient K~ is 3.3 and smaller than that
of an infinite uranium blanket by 20%.
Limit Parameters of Real Blankets. Let us consider the changes of the parameters of
-idealized schemes in the transition to-real blanket designs of fusion reactors. As has been
explained in the preceding section, the maximum breeding of useful isotopes (4.2 nuclei per
source neutron) was obtained in an infinite 23gU blanket. Let us observe how this quantity
changes when ttie construction elements of the blanket are taken into account. Three basic
factors reduce the Kg value.
1. A certain quantity of construction materials must be present in the uranium region::
these materials are the shells of the fuel elements and the housings of the coolant channels.
In the optimal design of fuel elements designated for a heat dissipation of 0.2-0.3 MW/liter,
the fraction of the shell volume amounts to 10-15% of the fuel element volume. The housing
of the cooling channels, the spacing grids, and other design elements at least double the
above value and one can hardly hope for a reduction of. the fraction of the design materials
below 20% by volume. The effects of the various materials do not strongly differ and we will
ruse the data for iron in our estimates (see. Fig. 1). The change in KE amounts to -1% of iron
in the blanket and the introduction of 20% of a construction material would reduce KE to 3.4.
2. The first wall of the plasma-chamber is situated between the iron zone and the urani-
um region in the designs of the majority of fusion reactors. It follows from Fi?g. 3 that the
-wall reduces KE by 1.7% per 1 mm of wall thickness at low thickness values. In the projects of
.tokamak reactors which have been developed so far, the thickness of the first wall is assumed
as 10-15 mm; plasma sputtering of the wall and the requirements to the rigidity of the struc-
ture have been taken into account. In fusion reactors of the open-trap type [5, 6J, where the
flux of particles from the plasma to the wall is smaller than in the tokamak, a thickness of
5-6 mm was assumed. When we consider this as the real lower limit, the KE value which can be
reached must decrease by 8-10% and amount to 3.1.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
y ~
F
y
Thickness (cm) of the
uranium region
Fig. 4. Dependence of the parameters of
a medium of 238U with an infinite Li re-
flector upon the thickness of the uranium
region.
TABLE 2. Parameters of Blankets in Some
Fusion Reactor Projects
Open trap ?
, ~~
1,ti
1,1
u,7
Tokamak [4]
:.' Sf
1.51
1.ii5
Ii, S!1
Tokamak [4]t
'!,71i
I , ~i;i
1, ''~
11,`3;5
*At the beginning of the operational per-
iod.
tAt the end of an operational period of
2.25 years at a neutron load of 1 MW/m2
on the first wall.
3. It is not advantageous tq make the uranium region of the blanket very thick because
the energy liberation and the plutoniuto breeding in this region are nonuniform and because it
is hard to obtain KT > 1 (see Fig. 4). When the thickness of the uranium region is reduced
to the optimal value of 7-10 cm, losses of -20% occur in the breeding of useful isotopes.
The value KT > 1 can be obtained (though this is not always desirable) in quasihomogeneous
lithium--uranium mixtures (see Fig. 2), but in this case the K~ losses amount to 10-20%. Thus,
we arrive at an effective total breeding coefficient of 2.5-2.8 which also has been assumed
in-the majority of promising developments of hybrid blankets of fusion reactors. The parame-
ters of two of these reactors, in which metallic waste uranium is used in the blanket, are
listed in Table 2. The expected breeding can amount to 1.4-1.8 plutonium nuclei per fusion
event.
The real yield values of the fissile nuclide are reduced to 0.5-0.8 per fusion event
when other uranium and thorium compounds are employed. The production o.f these nuclides, re-
ferred to the total thermal power, either slightly decreases (to 0.9 kg/(MW?yr)) or increases.
The yield of the nuclear fuel reaches 2 kg/(MW?yr) in the case of metallic. thorium fuel and
certain cgmpositions of uranium blankets.
During the operational period, Kpu is almost constant and the specific breeding decreases.
In the case of metallic uranium, the plutonium yield decreases by a factor of 1.3 and amounts
to -0.65 kg/(MW?yr) of thermal power by the end of an operational period of -2 years and a
neutron load of 1 MW/m2 on the blanket surface.
Conclusions. The above considerations have shown that when the real requirements to
blanket design are taken into account, a production of 1.4-1.8 tons of plutonium nuclei per.
fusion event can be expected in a blanket in. which metallic uranium fuel is employed. Ac-
cordingly, in a reactor with a fusion power of 500 MW, 2.5-3 tons of plutonium can be produced
per calendar year; when the total thermal power of the reactor is 2.5-3 GW at the beginning
of the operational period, the breeding amounts to -1 kg/(MW?yr).
LITERATURE CITED
1. J. Lee and R. Moir, Fission-Suppressed Blankets for Fissile Fuel Breeding Fusion Reactors,
Preprint LLL UCRL-84104 (1980).
2. S. V. Marro, D. V. Markovskii, and G. E. Shatalov, Problems of Atomic Science and Tech-
nology, Series Physics and Technology of Nuclear Reactors [in Russian], No. 9(22) (1981),
p. 26.
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
3. L. N. Zakharov et al., Problems of Atomic Science and Technology, Series Physics and
Technology of Nuclear Reactors [in RussianJ, No. .8.(21) (1981), p. 42.
4. S. M. Zakharova, B. N. Sivak, acid G. I. Toshinskii, Byull. Inf. Tsentra Yad. Dannym.
(Bull. of the Information Center for Nuclear Data), No. 3, Atomizdat, Moscow (1967)
(Appendix: Constants of Nuclear Physics for Reactor Calculations).
5. R. Moir et al., Progress in_the Conceptual Design of a Mirror Hybrid Fusion-Fission Re-
actor, Preprint LLL UCRL-51797 (1975).
6. J. Lee, Mirror Fusion-Fission Hybrids, Preprint LLL UCRL-80720 (1978).
474
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
REVIEWS
~. M. Krisyuk and V. I. Parkhomenko UDC 614.876+613.5:539.16.08+539.16
Introduction. Natural ionizing-radiation sources make the main contribution to the pop-
ulation dose. That contribution to the collective effective equivalent dose (EED) is 60-70%.
Medical irradiation constitutes 30-40% of the collective dose. The overall contribution from
all other sources, including environmental contamination from weapons tests, professional ir-
radiation, discharges from nuclear power stations, etc., is not more than 1%.
In recent years there has been a particular increase in the interest in population ir-
radiation research. This has occurred because the International Commission on Radiation Pro-
tection (ICRP) [1] suggested that there is a linear threshold-free effect from ionizing radi-
ation. On that basis, the frequencies of somatic-stochastic genetic effects are proportional
to the mean EED to the population. The largest doses from natural sources arise in living
accommodation and working buildings, where the radiation background arises from cosmic radi-
ation, which is partly attenuated by the material between floors, the Y rays from natural
radionuclides (NRN) present in the building materials, and from radon decay products entering
the air in the building from the building materials and the soil.
The Y background in a building is dependent ~on the specific activity of the NRN in the
materials. The vast scale of building of living accommodations in the country has led to a
search for new sources of traditional building materials and to the use of products made from
the wastes from mining, metallurgy, and the chemical industry, as well as ash and slag from
the thermal power stations. The use of these wastes has also been stimulated by the need for
low-waste production technology. However, such building materials frequently have elevated
NRN activities, which increases the population dose.
The elevated NRN activities in the materials also raise the concentrations of radon and
its daughter decay products DDP in the air. Other reasons for elevated concentrations of ra-
don and DDP may be that the nuclide enters the building from the soil and passes upwards be-
tween floors; low rates of air exchange are also another factor. There has been a tendency to
reduce air-exchange rates in living accommodations because of the need for power economy in
heating.
Until recently, there have been no standards for radioactivity in building materials,
nor have there been building regulations that restrict the_cencentrations of radon and DDP,
which has meant that in some countries (Sweden, the Federal German Republic, and the USA)
there have been considerable rises in the doses to large groups of the population [2]. In
recent years, there have been large-scale researches in most developed countries on the irra-
diation of the population from natural sources, with emphasis on the origins of the dose, and
there have been discussions on restricting population irradiation.
Research Methods. Thc: NRN contents of building materials are usually examined by y spec-
troscopy, where the single measurement enables one to determine the specific activities of
ZZ6Ra, ZZeTh, and ?OK. If there is radioactive equilibrium in the uranium and thorium fam-
ilies, the activities of these nuclides completely characterize the radioactivity of the ma-
terial. We have developed [3] the SGS-2G0 high-sensitivity y spectrometer, where the detec-
tor is an NaI(T1) crystal of size 150 x 150 mm containing a well (sample volume up to 212 cm3).
The minimum measurable specific activities of NRN are 23 Bq/kg for `iOK, 5 for ZZ6Ra, and 9
for zZeTh with a measurement time of 20 min and an error of 25%. A relatively small sample
is.used (250 g). If there are deviations from equilibrium in the uranium and thorium families
(as in specimens that have undergone chemical processing), it is best to use radiochemical
methods for ~~ spectrometry with a Ge(Li) detector, which enables one to determine the activ-
ities of several nuclides in each family, although the method is more laborious.
High-pressure chambers are mainly used to measure the y background. The first such cham-
ber was described in [4]. The apparatus is not readily portable, which has led to difficul-
Translated from Atomnaya ~nergiya, Vol. 57, No. 1, pp. 42-48, July, 1984. Original arti-
cle submitted November 14, 1983.
0038-531X/84/5701- 04.75$08.50 ? 1985 Plenum Publishing Corporation 475
Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050001-6
ties. This disadvantage is absent in a scintillation dosimeter as described. in [5]. The de~
tector here is an air-equivalent plastic scintillator (C,4H,o) covered with zinc sulfide. The
instrument has a virtually constant (?10%) dose sensitivity for radiation of energy >25 keV.
The photomultiplier dark current and other features give rise to a very .low inherent back-
ground, which is equivalent to a dose rate of 0.04 1!R/h (1 R = 2.58.10-4 Ci/kg). The instru-
went has been used in large-scale research on the y background in the Federal German Republic.
In recent years, extensive use has also been made of LiF, CaF2(Dy), and CaSO,,(Dy) thermolumi-
nescent dosimeters to measure the y background [6]. The advantages of these are that the
detectors can be sent by post and the measurements made in a single laboratory, which ensures
that the results are comparable.
We have calibrated geophysical exploration apparatus for irradiation from models for
saturated beds containing uniformly distributed potassium with uranium and thorium in equi-
librium [7], which has shown that the ratio of the dose sensitivities for these models and
a point zzeRa source is 1.0 ? 0.1 if one uses instruments containing gas-discharge counters.
The only disadvantage of such instruments is that it is necessary to correct for the counter
background.
The concentrations of radon and DDP may be determined in the air and .livingaccommodati.ons
by means of various instruments and methods.. In particular, the radon concentration can be
determined with an SAS-R-2 alpha scintillation counter [8]. In determining the DDP concentra-
tions, one usually employs the aspiration method, whose sensitivity is dependent on the
throughput of the air-sampling device and the method of counting the filter. Various methods
have been compared [9]. A deficiency common to them all is that one can determine the concen-
trations of radon and DDP only at the time of sampling; as these concentrations vary over time,
one requires repeated measurements to test a single building.
In recent years, integral methods have been developed for determining the concentrations ~
of radon and DDP averaged over, large time intervals. These take passive and active forms. in
the active method, the air is pumped through a filter and the a particles from the DDP depos- i
ited from it are recorded continuously [10]. This method is used to measure-DDP concentra- ~
tions and the sensitivity is determined by the throughput of the sampler (usually