ML19322C394
| ML19322C394 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 06/22/1979 |
| From: | Hajnal F ENERGY, DEPT. OF |
| To: | |
| References | |
| TASK-TF, TASK-TMR NUDOCS 8001160930 | |
| Download: ML19322C394 (21) | |
Text
I N-M
/
INTERNAll0NAL ATOMIC ENERGY AGENCY OECD NUCLEAR ENERGY AGENCY
~ z'r '
g g
x INTERNATIONAL SYMPOSIUM ON OCCUPATIONAL RADIATION EXPOSURE IN NUCLEAR FUEL CYCLE FACILITIES Los Angeles, USA,18-2 2 June 1979 I AE A-SM-242/
24 STRAY NEUTRON FIELD 6 IN Tile CONTAINMENT OF PWIls Ferenc Ilajnal Environmental Measurements Laboratory U. S. Ik partment of Energy New Yorp, NY 10011 e
B D
M Q
~
N J
g W
G n
(
')
p0r o
O 4
g3F-4 d
WO W
4 mWO
+
4 mh$E" E
W Ci NwE Ni E
?
F o w: Z -
m a
Z 4g" F30 Oz$mg m
O W W 0g b22 3 M]
$$i[o y
oral Presentation 2en3 w
M O
WW Z
e Q I k
? N i[O 8
mm 8001160 9 3 o This is a preprint of a paper iriterkfod for presentation at a scientific risoeting. Because of the provisional nature of its content and since charujos of sutistance or slutail riiay have to be rnado beforo publir.ation, the proprint is rrwulo avail.nlilo on the uvulinstaniling that it will not be citud m the hturature or in any way 1.o reperulut ud in its present fortn. The vinws empronud anti the statnnu,nts in.ulu eenn.pn thu susponulee: sy of the nanual unilun h), the vouws d.e not nui nn. inly influe t ilunu ni tho quvuon truent of the 'synatsty Moenlier Statch) or ut the dougnalny uit: surationh). In partsrular, amither thu 14l A n<n any ot/mr wi.e.orratsun no Innly spee.nonoog this owvring e an In h,*ht u,:p.oa s
- r, fon any nuatm oal oopenahn mi nu strin vo. vernt
3 l
.: ~
l, I
J 1.
INTRODUCTION l
There have been few systematic investigations of the stray penetrating radiation fields to.which workers and instruments are exposed inside the containments of nuclear j
power reactors. The recent concern about neutron exposure in these mixed ficid inside i*
pressurized water reactor containments and how adequately neutrons are routinely moni-I
{
tored has motived the development of data needed to evaluate the distribution of doses to workers and to determine levels of exposure (I).
Gamma radiation monitoring of operations and maintenance staff is performed relatively casily. The determ nation of neutrua dose or dose-equivalent values and their distributions with neutron encigy in the presence of significant gamma-ray levels inside i
and near PWR containments with available instrumentation is more difficult. The un-pleastnt if nei. hostile conditions of high ambient temperature and humidity and possible i
airborac and surface radionuclido contamination are barriers to the required spectro-i metric investigations.
Recently, measurements have been made with various devices, such as low energy l
resolution, moderating sphere systems and with fission counters using 235U, 238U and 237Np in conjunction with various thermal neutron absorbers (
- Polycarbonate track i
238 etch detectors with U, 237Np and Pu fission foils also have been used to measure 239 neutron flux and to obtain dose-equivalent rates f
s The present paper reports some results of a collaborative study performed in the containments of six PWR's of somewhat different design and construction. Both passive 1
LIF thermoluminescence detectors (TLD) and active Lil(Eu) scintillators were employed as thermal neutron detectors with the multisphere neutron spectrometer systems. The I
total neutron flux and dosimetric quantities, such as absorbed dose and dose equivalent,
{
as well as values of average energy, E, and quality factor, QF, were obtained from the spectral determinations. Tentatively, the results indicate that rapid estimates of quality factors and average energies may be obta'ined, from the ratios of measure-
~
ments performed with different moderator spheres 0 i
t l
d
-1_
i e
1 4
2.
THE MULTISPIIERE NEUTRON SPECTROMETER SYSTEM A multisphere neutron spectrometry system was used in the experiments ( ).
Six different diameter (2, 3, 5, 8,10 ad 12 in.) polyethylene moderating spheres along with a bare neutron detector and one covercd with a 0.032 in. thick cadmium absorber l
G 7
form the system. The neutron detectors are paired LIF and LIF TLD's, and 4 x 4 mm and 12.7 x 12.7 mm cylindrical Lil(Eu) scintillators. The energy response functions of the detectors were calculated in 20 evenly spaced logarithmic intervals from thermal i
to 2G MeV.
Each thermoluminescence detector has four each of LIF and LIF chips stacked to form two separate 3.2 x 3.2 x 3.G mm columns whose surface areas are the same as that l
of the equivalent sphere used in the calculations of the 4 x 4 mm right cylindrical scintil-(
6 lator response function. The net signa { due to the Li (n,o) reactions are extracted from the TLD measurements after individual chip responses are corrected for observed loss of sensitivity due to neutron irradiation damage. The TLD stacks were positioned inside i
l the cadmium covers in the same geometry to minimize readout differences due to the neutron capture gamma radiation from the cadmium. The 4 x 4 mm Li responses, after normalization, were used in unfolding both types of spectr'ometric measurements.
The use of highly enriched Lil(Eu) scintillator results in good resolution (about 9 percent) and in high light output which makes gamma-ray background differentiation l
relatively easy and reliable. In practice, the measurements are re' corded on a multi-channel analyzer and the background as represented by the area remaining after a straight line background under the neutron peak is subtracted.
The measurements were unfolded with an iterative unfolding method that successively l
corrects trial solutions finding only non-negative values, while the deviation between the measured and computed detector responses is minimized It is inherent of iterative least-squares unfolding techniques that sometimes only a reasonably smooth unfolded spectrum can be obtained from measurements with poor statistics. Care must be taken to assure that the measurement is good before the unfolded spectrum is accepted. Tests -
.v are being made to resolve this problem, but reliance on inter-laboratory comparison measurements is necessary at present.
3.
REACTOlt MEASUREMENT METIiODS The measurements in the containments of the PWR's were usually made at one meter from the floor level. Data were acquired by switching moderator spheres be tween the two scintillators, and the acquisition of a complete data set of two spectra re-quired about 30 minutes. Data acquisition with the TLD system required a minimum of several hours, while in low neutron fields, the TLD system was exposed for two or three days.
The scintillation detectors performed very well in gamma-ray fields up to tens of mrad /h combined with neutron contributi,ons of hundreds of mrem /h.
The 4 x 4 mm scintillator was more useful for higher neutron dose rates and the 12. 7 x 12.7 mm scit.illator in higher gamma-ray dose rate fields.
As mentioned earlier, data from the active system were accumulated with a multi-channel analyzer and stored on magnetic tape for later analysis. Due to high ambient temperature and humidity in the containment, the electronics were enclosed in an in-sulated, air tight container cooled with dry ice and dried with silica gel. The dry ice kept the inside temperature below about 40 C for two days, anc nu closed container protected the instrumentation from contamination.
4.
RESULTS AND DISCUSSION SLIDE #1 The energy spectra derived by unfolding the multisphere measurements were used to calculate the integral quantitica, such as neutron flux, cp l
average energy (E), absorbed dose and dose-equivalent rate and quality factor (QF). We shall show only one example, Figure 1, of the dif-forential energy spectra, measured on the operating floor near the re-actor cavity and the control drive mechanism., This neutron spectrum is relatively "hard"; E is 90 kev and the quality factor ic G.4.
The absorbed dose rate is 3 mrad /h, the dye-equivalent rate is 10 mrom/h, and the total neutron flux is 1060 n/cm s.
Eighteen percent of tne
[
neutrons are themal and about 13 percent have energies greater than
^
!.i-d 4
l 300 kev, that is over the threshold energy of the track etch detectors.'
Generally, the differential energy spectra have thermal groups of various intensities followed by an approximately 1/E distribution. Then, depend-j ing on the E and QF, the spectra may decline rapidly. In the case of
_t E = 7 kev, for example, a rapid decline sets in at about 4 kev and
~!
virtually no neutrons are found over 200 kev.
j' The 30 spectral measurements have a wide range of E', from 0.4 kev i
i to about 1 MeV, with the E of 26 of these measurements being from
- 10. kev to 1 MeV.
I The single E >l MeV measurement point has a spurious peak in the unfolded spectrum at about 10 MeV, which might be attributed to the limitation of the unfolding procedure.
4
[
SLIDE #2 The quality factor distribution showed two peaks; 24 were from 4 to 7 and i
j 6 from 2 to 3 as shown in Figure 2.
Larger quality factors were dominant i
on the top or operating floor and somewhat small values, QF < 3, for the i
middle levels of the reactors. 'For the very large E change. from 0.1 to 3.
100 kev, the quality factor changes only from 2 to 4.
For the 24 measure-ments which have QF equal to 4-7, E changes only from 40 kev to 1 McV.
-[
The neutron flux varies moderately, from 2.5 x 102 to 2. 6 x 104 n/(cm2 g),
P No unusually high thermal or fast flux fields were encountered, t
To extract the kind of information from the measurements which might help us to understand the in containment stray neutron fields of the PWIPs we used graphical and correlation analysis techniques. The graphical display of the measured and evaluated data aided us in understanding the numerical relationships reflected in the data.
This method helped us to reveal of some peculiarities in the observed data and helped us to identify subsets. The graphical data display uncovered features of the data that were totally unanticipated prior to the data analysis. Generally, the absence of precedents of dealing with reactor measurements, the graphical representation of the data wa,
chosen on trial and error basis, until representations have been f
' which seen lofy
^
simplified the data and seemed good candidates for correlation analysis.
The dain i
were plotted and crossplotted in various ways using the total flux, E, QF, specific c'(. -;,
' specific dose equivalent, sphere responses, sphere response ratios, their products or t
their ratios as variables.
l p
SLII3E #3 The logarithm of the specific dose equivalent, rem per n/cm, versus the quality factor QF exhibits a linear dependence as shown in Figure 3.
The 4
two points on the top of the graph at 3.5 and 3.G x 10-8 rem per n/cm2 (QF are 3.0 and 0.1) were measured with our primary detector, a 12.7 x 12.7 mm (Lil(Eu) scintillator detector in low scattering environment The measures E are 2.8 and 2.0 McV. The theoretical values for bare Cf sources are E = 2.13 MeV and QF = 9.3.
There appears to be two subsets, the reactor measure ments plus the D O moderated 252Cf source form 2
one subset and the rest of the calibration sources form the second subset.
This kind of separation even more pronounced if E is used as variable in-stead of QF. The average specific dose equivalent values of the two sub-sets differ by about a factor of ten. The straight line represents the fit to a linear regression equation. As expected the fit is very good.
f SLIDE #4 The variables of the previous two figures are evaluated quantities based on the derived differential energy spectra and the appropriate conversion factors. At this point we introduce the sphere responses, response ratios and the combinations of these with the evaluated gutmtities as new variables.
In Figure 4 the logarithm of the ratios of the 10 in, and 3in, sphere responses,
/Il, versus the quality factors, QF, are shown. The resulting graph is Il 3
10 similar to the one in Figure 3 albeit the specifle dose equivalent values are replaced with the response ratios of the 10 in, and 3 in spheres, suggest-Ing that the response ratios may take place of at least one of the derived quantities.
~Again the separation of the reactor and calibration measurements is app-4t.
The line is linear regression fit to all the measurements, and using thi,,
linear fit the quality factors of the reactor measurements in' half of the cases will be underestimated. The bare 252Cf and Al - moderated 252Cf fission sources show the largest deviation.
~
SLIDE #5 4
The quality factors and the logarithm of the sphere response ratios are in-I terchangeable as shown fin Figure 4, therefore next the behaviour of the specific dose equivalent as function of the different sphere response ratios were investigated. In Figure 5 the specific dose equivalent values versus the ratios of the 8 in, and 3 in. sphere responses, Ilg/Il3 are shown on a log-log scale. The distinct separation of the reactor and calibration l
measurements is evident.
1 1 l
The dynamic range of the response ratios, which indicate how well this f
method could determine the specific dose equivalent, of all the measure-n.wts is 25 and of all the reactor measurements is 2.2, which is not ruita a large range. Nevertheless these response ratios could be used
!c distinguish reactor measurements from calibration ones. We must emphasize when we talk about calibration measurements of the ones l
used in the present experiments are in that category.
t SLIDE #6 i
The functional dependence of the specific dose equivalent values versus the ratio of 10 in. and 3 in. sphere responses,11
/N, plotted on a log-i 3
10 log scale are shown in Figure 5.
This seems to be more linear than the previous plot. The dynamic range of the response ratios of all the i
measurements also increased to 80 and the reactor measurements to 7.
This increment of dynamic ranges makes the Il /II ratios a better 10 3
measure of the specific dose equivalent values than the R /II3 ratios.
8 The linearity of the plot increase 4however the separation of the reactor i
and calibration source measurements is still evident.
FIGURE 7 In the present studies the quality factor (QF) was almost exclusively used in place of the average energy, E, a well lamwn and extensively used quantity in neutron spectroscopy. The functional dependence of the measured and evaluated quantities in terms of E are under investigation and only one example of this ongoing study presented in Figure 7, where the E versus ratio of 10 in. and 3 in. sphere responses, R /R3 are shown on a semi-10 log scale (Note the break in the abscissa at 2.9). This figure illustrate a clear cut difference between the presently used calibration sources and reactor measurements. It also indicates that if the average energy cali-bration was based on bare 252Cf, Al - and I! 0 moderated fission spectra 2
measurements then the average energy of a reactor spectrum can not be determined from sphere response ratio measurements. The average
{
cnergy of a reactor neutron spectra can not be determined from the average energies and ratio of 10 in, and 3 in, sphere responses of cali-bration sources.
SLIDE #8 The term " rem meter" is quite familiar to anf practioner of health physics.
Hem meters are usually single moderator and single detector instruments.
Many attempts have been made in the past to design a'.J build such instrument
_c_
t l
. _-=
M fo.' neutron dosimetric purposes from thermal to say 20 MeV energy range.
Bas.ed on the present experiments an 8 in. " rad meter" was " constructed".
In F'gure 8 the ratio of the response of the 8 in. sphere, I1, and the 8
mearured absorbed dose rate values versus the quality factors are shown.
The use of the quality factor separates the reactor and calibration source measarements into the now familiar two subsets. A bare 252Cf fission source seems to be a good calibration source however the measured ab-sorbed dose rate values would be 40 percent accurate.
SLIDE #D Next the behaviour of simple " rem meters", based on the responses of the 1
8 and 12 in. ' spheres were investigated. The independent variable, the abscissa, is again the quality factor, and the ordinate the sphere responses R8 and 1112, divided by the dose equivalent rate values. Calibrating with bare 252Cf fission neutron sources the 8 in. " rem meter" will over-respond to stray neutrons of a P,WR by as much as a factor of 5 and the 12 in.
sphere will under-respond in most cases. Visual inspection of the 8 in.
" rem meter" response suggests that there might be a quality factor dependent correction factor which possibly could flatten the response of this detector. The sensitivity of the 8 in. " rem meter" is good and 6 counts per minute corresponds to 4 prem/h.
SLIDE #10
~
i A " rem meter" constructed from the 10 in. sphere has the flattest response.
In Figure 10 the response of the 10 in. sphere divided by the dose equivalent values verius the quality factors are shown. A simple device, such as this may be usable for reactor measurements. If this instntment is calibrated with a bare 252 Cf source the dose equivalent rates might be overestimated by about a factor of 2.
The sensitivity of this device is good, G counts per minute correspond to 10 prem/h. It may be possible to obtain a calibration
/Il3 vs. QF) to make factor from the data presented in Figure 4 (log R10 the response of this detector linear.
SLIDE #11
?-
The measured dose rate of bare point sources of'different source strength versus the total neutron flux in the same geometry are in constant proportion such as line a_ in Figure 11. Ilowever it came as some of a surprise that the reactor measurements, including the D 0 moderated 252Cf fission 2
source, are also in near constant proportion as shown in line b.
Linear
^
regression analyses indicate that the slope of the bare 252Cf mcasurement t.l.
1i[
curve is 1, as it should be. The slope of the curve for the reactor measure-j ments plus the D 0 calibration is 1. 05 + 0. 03. The correlation coefficients 2
are over G. DU. The equation for the dose rate in rad /h in terms of the.
191 l
total flux, 9, in n/(cm2 s ) is : dose rate (rad /h) = 1. D l
SLIDE #12 I
Finally, the evaluated dose-equivalent rates versus the total flux values are l
shown in Figure 12. Note that the application of the quality factors separates the data into three distinct groups. One is the calibration measurements, line n_, where the specific dose equivalent is greater than 2 x 10-8; the second group, line b, has specific dose equivalent ranging from 2 x 10-0
~
l to 1 x 10-8, and the third group line c, where the specific dose equivalent are less than 2 x 10-9 The slope of the regression lines for each of the separate groups are very near to one, and the correlation coefficients are l
over 0. DG.
The dose equivalent rates at the operating level of the reactors are in terms 1
of the total neutron flux s
d.c. rate (rem /h) = 14 g I
and at the middle level of the reactor i
d.e. rate (rem /h) = 4.7 9 5.
CONCLUSION We showed that the neutron differential energy spectra, average energy, quality factor and the dosimetric quantities of the stray neutron radiation fields in the contain-i ments of the PWR's can be determined using low resolution moderating sphere neutron l
spectrometers. The physical and dosimetric quantities are estimated to be accurate to i
within 10 to 15 percent.
-[
4 Based on 31 measurements at six PWR's of somewhat different design, simple linear relations were found for the dose and dose-equivalent rates as a function of the total flux, 9 These indicate that in containments of PWR's the problem of dose equivalent rate E;
determination might be reduced to the measurement of the total neutron flux.
,p
' n e
i
).'
t
_s_
i l
1 REFERENCES l
1.
Rossi, H. H., Afays, C. W., Health Phys. 34 (1978) 353
-2.
Hankins, D. E., Griffith, R. V., "A Survey of Neutrons Inside the Containment I
of aPressurized Water Reactor", in Radiation Streaming in Power Reactors, j
USDOE Itep. OltNL/RSIC-43 (1970) 114 t
3.
Bricka, AI., " Neutron Aleasurements Near PWR Power Reactors", in Seventh 4
DOE Workshop on Personnel Neutron Dosimetry, USDOE Itep. PNL-2807 (1978) 103 4.
Butler, AI. H., Ohnesorge, W. F., Auxier, J. A., "Aicasured Distribution of Neutrons Inside Containment of a PWR", in Itadiation Streaming in Power lleactors, USDOE Itep. OltNL;ItSIC-43 (1970) 110 5.
Hankins, D. E., A Afodified-Sphore Neutron Detector, USDOE Rep. LA-3505 (1967) 6.
AfeLaughlin, J. E., O'Brien, K., " Accelerator Stray-Neutron Dosimetry:
Spectra of Low-and Intermediate-energy Neutrons", in Neutron Atonitorint, IAEA, Vienna (1967) 335 7.
Sanna, R. S., "Afodification of an Iterative Code for Unfolding Neutron Spectra from A1ultisphere Data", USDOE Rep. HASL-311 (1976) i 8.
Casson, R. AI., Benton, E. V. " Nuclear Track Detection",
i 2 (1978) 173 t
~.
-D-i
4 I
(A)
(B).
> 10s 4
1 4 10s E
e Neutron Flux 4 Absor%d Dose Rate
/f x-4 Dos
. g 104 80 auivalent Rate E
2 s
E102 4 60 5
8 y
s 2
/. +
g g40 3 1M 7
a, y
h10-2 20 I
I I
I I
I I
1 0-10 0
10-a 10-s 10-4 10-2 100 102 to-a 10-s 10-4 10-2 1 00 102 E,MeV E.MeV Figure 1.
Differential neutron energy spectra measured in the vicinity of the pressure vessel of a PWR on the operating floor:
4 a) differential energy spectra; b) cumulative percentage of neutron flux, absorbed dose and dose-equivalent rates versus the neutron energy.
_f ;
e s
.e-
. - -. ~.. _ - n
+.
W v
v v
i s
104 :
y E
~
i
- . 8
~
S
~
o B
%C i
3*
1 103 :
o
}
.=
102 Reactor measurements
. Unmoderated c D O moderated, Cf 252 calibration 2
e H O moderated measurements 2
Al moderated 10 1
2 3
4 5
6 7
8 9
10 11 OF 2
Figure 2.
The measured total neutron flux, n/(cm -s) versus quality factor, QF.
Reactor measurements
+ Unmoderated
~
c D2O moderated Cf 252 calibration measurements
=
E a H2O moderated o
c Al moderated
++
a E
5 i
o.
a Ee b
t 5 10.
m
.3
- s g
t*
O t
o 5g c..
m.
en 1.
1 I
i I
2 3
4 5
6 7
8 9
10 OF Figure 3.
The specific dose-equivalent values versus the quality factors, QF. The solid line is the linear regression curve.
a I
~
. Reactor
+ Unmoderated c D20 moderated Cf 252 calibrat. ion measurements H2O moderated A
c Al moderated n
E S 1.
E s c e
8.
~
g m
2 e
.g
. - g.-
n a.
m 1
O o
m E
.01 I
i I
i f
2 3
4 5
6 7
8 9
10 11 OF Figure 4.
The ratios of 10 in. and 3 in, sphere responses, R /R '
10 3
versus the quality factor, QF. The solid line is the linear regression curve.
F 10.
v-x On Y..
^
E g
c.
o E
a c
c 1.0 e
g o7 0)
=
e m
o y
S_
o Reactor measurements
.8
+ Unnioderated S
- D O moderated, Cf 252 calibration 2
- H O moderated measurements 2
c Al moderated 0.1
' ' ' '. 0 10.
.1 1
Ratio of sphere response, RgIR3 2
8 Figure 5.
Evaluated specific dose equivalent, rem per n/cm x 10,
versus ratio of S in. and 3 in. sphere responses, Rg/Il -
3
100 I
Reactor measureme nts e Unmoderated
- D O moderated Cf 252 calibration 2
x e H O moderated measurements 2
C c Al moderated g
t E
eo I
e 10-
.E
=7
~
e h
Y'
.o Y.'.
,g u) 1, 0.1 1.0 10.
Ratio of sphere responses, RdR3 9
Figure 6.
Evaluated specific dose equivalent, rem per n/cm2 x 10,
versus ratio of 10 in, and 3 in. sphere responses, R o/Il -
i 3
The solid line is the linear regression curve.
.y.
4-m,
+am e.
pu,
--sum e imm.
4--ema m-b
+
s 6
2 T
e
,~
d
+
100 o
S.
~
10-1 l a[..
e lE g
gg 10-2 :
~
Reactor measurements
. Unmoderated 10-3 E
- D O moderated Cf 252 calibration 2
H O moderated measurements c
2 f
- Al moderated i
0
.2 5
.6
.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 3.0 3.2
~
Ratio of sphere responses, R o/R3 i
Figure 7.
Evaluated average energy, E Q1eV) values versus the ratio of 10 in, and 3 in. sphere responses. Note the
/R3 = 2. 4.
discontinuity in the abscissa after R10
.m
.ma-.
e
_wg
,m
-e-eman.
---ma
'w6eu ep e-4*
~h
- m hwp& em#e+wN 9M h
"*N*
e*6-*
200 4
d:
g g 150 a
E g
E cc a n e100 m
o O
C C
8.
g e
E Reactor measurements
.c 50
. Unmoderated S
e D O moderated, Cf 252 calibration d
2 c H O moderated measurements 2
c Al moderated 0
2 3
4 5
6 7
8 9
10 11 OF Figure '8.
The evaluation of a " rad meter" : The response of the 8 in.
sphere divided by the measured dose rate, mrad /h, versus the quality factor, QF.
l l
l
_ _ _ =
_. = _ _
l
loo Reactor measurements E
- Unmoderated SE D O moderated Cf 252 calibration 2
C!
H O moderated measurements 2
hE Al moderated e
$g g
e m ig
- 8
". E
~
.E g g
- ,. * + +
, e o.
,+
10 10 c
E 3,4,'* e.3
.c g,
.e E
cg
(.
- a. 5 mE o
@y W
Cs
.c
- c. m m
.d_ E
~
mo
- c.
1.0 1
2 3
4 5
6 7
8 9
10 11 OF Figure D.
The evaluation of two " rem meters";
Upper diagram: the response of an 8 in. sphere, Lower diagram: the response of a 12 in. sphere, each of them divided by the measured dose equivalent rate, mrem /h, versus the quality factor, QF.
x N
E Reactor measurements h30
+ Unmoderated
~
o D O moderated, Cf 252 calibration u
2 E
c H O moderated measurements 2
25 c Al moderated E
i20 s2 ct i 15 g
o a
o o
$"*+*g*+,
E
^ ^
10 m
8 E"
5 ci b
t t
t i
1 9
I a
f 2
3 4
5 6
7 8
9 10 QF Figure 10. The evaluation of a 10 in. spherical " rem meter";
The response of a 10 in. sphere divided by the measured dose equivalent rate, mrem /h, versus the quality factor, QF.
I
'o w
d
- b
/
/
Reactor
+ Unmoderated pg
- /
c D 0 moderated Cf 252 calibration 2
8 s H2O moderated measurements o Al moderated,
a 10.
~
r 3
e s
.s E
a 3
13 1.
^
.1 I I ! l ill!
i
! I l i f !!!
I
' I 11111!
I I I Iliff 1 01 102 103 1 04 Total flux, nl(cm2-s)
Figure 11. Dose rate versus the total flux of neutrons: line a - from 252 unmoderated Cf fission sources, line b_ - in thc con-tainments of the reactors plus D 0-moderated fission source.
2 m
t Z
b I
Reactor measurements
+ Unmoderated o D O moderated Cf 252 calibration t
2 a H2O moderated measurements c
c Al moderated
- E E
/
e100.
a E
3 c
e
~
/
3
/
So 5
2 A
@ 10.
t g
/
1.
t I '
!I!'
f f ' ' ' ' ' ' '
'il I
1 01 1 02 103 1 04 10s Total flux, nI(cm2.s)
Figure 12. Dose-equivalent rate versus the totcl flux of unmoderated 252Cf fission sources, line a; the reactor measurements on the operating floor levelplus the D O moderated fission 2
source, line b; and the reactor measurements on the medium floor level, line c.
=
U _S6,b INTERNATIONAL ATOMIC ENERGY AGENCY l
OECD NUCLEAR ENERGY AGENCY 4
INTERNATIONAL SYMPOSIUM ON OCCUPATIONAL RADIATION EXPOSURE IN NUCLEAR FUEL CYCLE FACILITIES Los Angeles, USA,18-22 June 1979 IAEA-SM-242/
24 STRAY NEUTRON FIELDS IN TiiE CONTAINMENT OF IMRs j
t l
Ferenc Itajnal and Robert S. Sanna i
Environmental Measurements Laboratory U. S. Depart. ment of Energy New York, NY 10014 i
Robert M Ryan and Elizabeth II. Donnelly Rensselaer Polytechnic Institute Troy, hT i
J
{
i
}
t This is a preprint of a paper intended for presentation at a scientific rnesting. Because of the provisional nature of its 4
content and sina changes of substance or detail may have to be made before publication, the preprint is made available on the understanding that it will not be cited in the literature or in any way be reproduced in its present form. The views expressed and 4
the statements made remain the responsibility of the named author (s); the views do not necessarily reflect those of the govern-ment of the designating Member State (s) or of the designating organization (s). In pJrticular, neither the /AEA nor any other l
organization or body sponsoring this meeting can be held responsible for any material reproduced in this preprint.
LAEA-SM-242/24 STRAY NEUTRON FIELDS IN THE CONTAINMENT OF IVRs ABSTRACT Multisphere neutron spectral measurements were perfonmed at six pressurized water reactors. (PWRs). The measured differential neutron energy spectra and flux were used to determine average energy, quality factor, dose and dose-4 equivalent rates. The maximum quality factor found was 7.
The neutron spectra varied from a highly moderated one with E = 1 kev to a less moderated one with 5 = 0.7 MeV.
Simple linear relations were found to calculate the dose and dose-equivalent rates from the total neutron flux. The dose-equivalent rates on the operating floor icvel are fourteen times and on the middle level five times the total neutron flux.
1
?
r ;
r
1.
INTRODUCTION There have been few systematic investigations of the stray penetrating radiation fields to which workers and instruments are exposed inside the containments of nuclear power reactors.
The recent concern about neutron exposure in these mixed fields inside pressurized water reactor containments and how adequately neutrons are routinely monitored has apparently motivated the development of data needed to evaluate the distribution of doses to workers and to determine acceptable levels of exposure [1].
Gamma radiation monitoring of aperations and maintenance staff is performed relatively easily. The determination of neutron dose or dose-equivalent values and their distributions with neutron energy in the presence of significant gamma-ray levels inside and near PWR containments with availabic instru-mentation is more difficult. The unpleasant if not hostile conditions of high ambient temperature and humidity and possible airborne and surface radionuclide contamination are barriers to the required spectrometric investigations.
Recently, measurements have been made with various devices, such as low energy resolution, moderating sphere systems and with fission counters using 235U, 238U and 237Np in conjunction with various thermal neutron absorbers [2,3].
Polycarbonate 239 u fission foils 238, 237Np and P
U track etch detectors with also have been used to measure neutron flux and to obtain dose-equivalent rates [4]. While neutron spectra can be found by unfolding the multisphere and fission counter data, the track etch system provides approximate spectral information that is only sufficient to set an upper lhnit on the applicable quality factors.
Properly calibrated neutron survey meters supplement such spectral measurements to obtain rapid estimates of neutron dose-equivalent rates [2]. One type of survey instrument is equipped with both 3 and 9 in. diameter moderator spheres and the sphere response ratio is also used as an indicator of the spectral shape and to interpret personnel neutron dosimeter readings [5].
The present paper reports some results of a collaborative study performed in the containments of six PWRs of somewhat different design and construction. Both passive LiF thermo-luminescence detectors (TLD) and active LiI(Eu) scintillators were employed as thermal neutron detectors with the multisphere systems. Dosimetric quantities, such as absorbed dose and dose equivalent, as well as values of average energy, E, for the neutron field and the quality factor, QF, were obtained from the spectral determinations. Tentatively, the results indicate that rapid estimates of quality factors and average energies may be obtained, as Hankins has indicated [5], from the ratios of measurements performed with dif ferent moderator spheres.
t 2; THE MULTISPHERE NEUTRON SPECTROMETER SYSTEM 2
The multisphere system was adapted from a system used in.
previous studies around particle accelerators [6]. Six 2
different diameter (2,.3, 5, 8, 10 and 12 in.) polyethylene moderating spheres along with a bare neutron detector and one 1
covered with a 0.032 in, thick cadmium absorber form the system.
6LiF and 7LiF TLDs, and he neutron detectors are paired 6Li1(Eu) scintillators.
4 X 4 m and 12.7 X 12.7 mm cylindrical he. energy, response functions of the detectors were calculated in 26 evenly spaced logarithmic intervals from thermal to 26 MeV.
Due to the wide' energy bins, resonance effects are smoother and there is no line structure in the response function.
It is I
I unlikely that any fine structure can be observed in the spectra.
Higher resolution spectrometers are needed to resolve resonance scattering and absorption peaks, such as those at 25 and 28 kev,'
produced by selective filtering in the iron of the reactor pressure vessel [7].
Each thermoluminescence detector has four each of 6LiF and j
7LiF chips stacked to form two separate 3.2 X 3.2 X 3.6 m columns, whose surface areas are the same as that of the equivalent sphere j
used in the calculations of the 4 X 4 m right cylindrical scin-6
]
tillator response function. The 4 X 4 m Li responses, after i-normalization, were used in unfolding both types of spectrometric measurements [8]. Resonance scattering and absorption by the 6Li (n,a )
fluoride were neglected. The net signal due to the reactions are extracted from the TLD measurements af ter indi-vidual chip responses are corrected for observed loss of sensitivity due to neutron irradiation damage. The TLD stacks were positioned inside the cadmium covers in the same geometry to minimize readout differences due to the neutron capture gamma radiation from the cadmium.
l The use of highly enriched Lil(Eu) results in good resolu-i tion (about 9 percent) and in.high light output which makes f
gama-ray background differentiation relatively easy and reliable.
In practice, the measurements are recorded on a multichannel
^
analyser and the background as represented by the area remaining after a straight line background under the neutron peak is sub trac ted,
he measurements were unfolded with a modified version of the BON coh,- an iterative unfalding method that successively corrects trial solutions finding only non-negative values, while the deviation between the measured and computed detector responses is minimized. The process is usually terminated after 1000 iterations [8].
It is inherent'of iterative Icast-squares unfolding techniques that sometimes only a reasonably smooth j
unfolded spectrum can be obtained from measurements with poor statistics. Care must be taken to assure that the measurement
'is good before the unfolded spectrum is accepted. Tests are
- 1 g
g+
,-------.c,-,,,-.,,.,,-,.-.,wy-,-,,-,m
,.p a
.-g-.+
--, -., - o
M P
m-,
<2 g
+
2.
Je m-+
Ja
-.d,,
, - - ~
+
4-4 baing mada.to resolva this problem, but reliance on inter-
- laboratory comparison measurements is necessary at present.
3.
REACTOR MEASUREMENT METHODS The scintillation detectors performed very well in gamma-ray fields up to tens of mrad /h corbined with neutron contribu-tions of tens of mrem /h. The 4 X 4 mm scintillator was more useful for higher neutron dose rates and the 12.7 X 12.7 mm scintillator for higher gamma-ray dose rates.
Both spectrometer systems were calibrated with bare and 252Cf fission sources. The D 0, H 0 - and aluminum-moderated 2
2 calibration factors obtained with three differ ent bare sources a'
agreed to within *4 percent.
The measurements in the containments of the PWRs were usually made at one meter from the floor level.
Data were 2
j acquired by switching moderator spheres between the two j
scintillators,and the acquisition of the complete data set of two spectra required about 30 minutes. Data acquisition with I
the TLD system required a minimum of several hours, while in low neutron fields, the TLD system was exposed for two or three days.
Data from the active system were accumulated with a multi-channel analyzer and stored on magnetic tape for later analysis.
Due to high ambient temperature-and humidity in the containment, the electronics were enclosed in an insulated, air tight container cooled with dry ice and dried with silica gel. The dry ice kept the inside temperature below about 40 C for two days, and the closed container protected the instrumentation from contamination.
The reactor pressure vessel, in the center of the contain-l ment, rests on the vessel support structure and is surrounded by the primary shield. The reactor vessel support structure rests on the lowest elevation of the containment. Above this, there are two or three more floors that are accessible to personnel. The. radiation levels at different floors will vary according to the power level of the reactor and the shielding l
design and construction.
I The neutrons from the core encounter different amounts of shielding due to the exact design. There may be additional l
shielding on the top of the reactor. In case of older PWRs, neutron streaming is not of major importance.
In newer aesigns, i
to reduce the possible high asymmetric pressure load on the reactor vessel and biological shield which could result from a
. major malfunction, the upper part of the region was enlarged and resulted in an increased neutron and gamma-ray streaming.
On the lowest and medium containment levels, gamma radia-
-tion usually dominates,'and on the top or operating level, neutrons are dominant. The radiation levels may vary from a i
r.---
few mrem /h to many rem /h, therefore, a detailed area survey is necessary prior to neutron measurements. Most of the present measurements were performed on the top levels where radiation fields often have large spatial gradients. Also surveyed were important areas where personnel and equipment are transferred into the containment through an airlock or hatchway.
4.
RESULTS AND DISCUSSION The energy spectra derived by unfolding the multisphere measurements were used to calculate the integral quantities, such as neutron flux, average energy (5), absorbed dose and dose-equivalent rate and quality factor (QF). An example of the differential energy spectra, measured on the operating floor near the reactor cavity and the control drive mechanism is shown in Figure 1.
This neutron spectrum is relatively "hard"; E is 90 kev and the quality factor is 6.4.
The absorbed dose rate is 3 mrad /h, the dose-equivalent rate is 19 mrem /h, and the total 2
i neutron flux is 1060 n/cm s.
Eighteen percent of the neutrons are thermal and about 13 percent have energies greater than 300 kev, that is over the threshold energy of the track etch detectors [9]. Cencrally, the differential energy spectra have thermal groups of various intensities followed by an approxi-mately 1/E distribution. 1 hen, depending on the 5 and QF, the spectra may decline rapidly.
In the case of 5 = 7 kev, for example, a rapid decline sets in at about 40 kev and virtually no neutrons are found over 200 kev.
The 31 spectral measurements have a wide range of E, from 0.1 kev to about 1 MeV, with the 5 of 26 of these measurements being from 10 kev to 1 MeV.
The quality factor distribution showed two peaks; 24 were from 4 to 7 and 6 from 2 to 3.
The summary of 5 and QF is in Table I.
There seems to be no simple 5 versus QF dependence.
Larger quality factors were dominant on the top or operating floor and somewhat small values, QF < 3, for the middle levels of the reactors. For the very large 5 change, from 0.1 to 100 kev, the quality factor changes only from 2 to 4.
For the 24 measurements which have QF equal to 4-7, 5 changes only from 40 kev to 1 MeV.
The single 5 > 1 MeV measurement point has a spurious peak in the unfolded spectrum at about 10 MeV, which might be attributed to the limitation of the unfolding procedure.
2 to 2.6 X 104 The neutron flux varies moderately, from 2 5 X 10 2
n/(cm s), for all measurements which were performed evenly over this flux range. No unusually high thermal or fast flux fields were encountered. The frequency distribution of the neutron flux is shown in Table II.
The average energies can only be determined with lesser accuracy. The 10 and 3 in, sphere response ratios correlate very weakly with 5, having a correlation coefficient of only 2 I
O.2.
Better results can be obtcined by using ths 8 and 3 in, response ratios to obtain 5 = 7.9 X 10-3 (R /R ) with an 0 5 8 3 correlation coefficient.
A hypothetical " rem" counter was constructed from the responses for the 10 in. sphere divided by the evaluated dose-equivalent rates versus the quality factors.
The smallest 252Cf calibration sources. A simple device ratios were for the like this is usabic for reactor measurements, but if it is calibrated with a bare 252Cf source most of the dose-equivalent rates will be overestimated by a factor of up to 2 [5]. A device constructed from the 8 in, sphere will over-respond by as much as a factor of 5 and the 12 in, sphere will under-respond by as much as a factor of 3.
Since the slope of the specific dose versus 5 curve is small, 0.18, it is not quite unexpected that the evaluated dose rate versus the total neutron flux gives essentially a straight line, as shown in Figure 4.
Linear regression analyses indicate that the slope of the bare 252Cf measurements curve is 1, as it should be.
The slope of the curve for the reactor measurements plus the D20 calibration is 1.05 0.03.
All correlation coefficients are close to one.
Finally, the evaluated dose-equivalent rate versus the total neutron flux curves are shown in Figure 5 Note that the appli-cation of the quality factors separates the data into three distinct groups. One is the calibration measurements where the specific dose equivalent is greater than 2 X 10-8; the second group has specific dose equivalent ranging from 2 X 10-9 to 1 X 10-8, and the third are less than 2 X 10-9. group where the specific dose equivalent The slope of the regression lines for each of the separate groups are very near to one, and the corre-lation coefficients are over 0.96.
To identify possible simple relationships, linear regression analyses were performed on the measured quantities, the detector responses and response ratios, and the evaluated quantities such as the flux and dose rate. For example, by plotting the specific absorbed dose against the average energy, on a log-log scale, the correlation analyses indicate that the reactor data plus the D 0 2
252Cf measurements fit a line with a small slope and moderated 252 252Cf, H 0 and aluminum-moderated Cf measurements the bare 2
fit a line which has a much larger slope. Because the slope of the linear regression of the reactor data is small, the depend-ence on 5 is weak. An example of good correlation between the specific dose equivalent and quality factor is shown in Figure 2.
2 The specific dose equivalent in rem /(n/cm ) in terms of the evaluated quality factors, QF, is equal to 4 X 10-10 X (QF)0.19, and the correlation coefficient is 0.96.
As expected, the quality factor, average energy, dose or dose equivalent per neutron per cm2 are not measurable directly with a single instrument. Correlation analyses show that the E
ratios of the responses of different sizes of detectors may provide an indirect method to determine these values. The quality factors may be determined from the response ratios of the 10 and 3 in. sphere data. -Figure 3 shows these response ratios versus the quality factor. The quality factor, QF, may be calculated from the linear regression equation in terms of the response
/R, and the ratios of the 10 and 3 in, diameter spheres, R10 3
/R ). The correlation equation is QF = 8.63 + 4.72 log (R10 3
coefficient is 0.92.
Similarly one may obtain the sgecific dose equivalent from rem /(n/cm ) = 1.9 X 10-10 (R /R )0.1, and the 2
10 3 correlation coefficient is 0.97.
5 CONCLUSIONS i
Since the operation of most of the instruments and the equipment of a PWR are monitored from the console, no routine in-containment inspections and surveys are necessary. Entry may be required in case of equipment failure,such as the breakdown of pumps or valves. Occasional in-containment checks of pressure
~
and temperature gauges ' might be necessary. Extended area surveys, gamma and neutron, are usually performed in preparation for outages before major repair or refueling.
To verify the " rem" survey meter measurements at selected sites, the neutron differential energy spectra, average energy, quality factor and the dosimetric quantities of the stray neutron radiation fields in the containments of the PWRe can be determined using low resolution moderating sphere neutron spectrometers. The physical and dosimetric quantities are estimated to be accurate to within 10 to 15 percent. Simple relationships amongst the measured and evaluated quantittas can be obtained using' linear regression analysis.
Based on 31 measurements at six PWRs of somewhat different design, simple linear relations were found for the dose ind dose-2 s).
equivalent rates as a function of the total flux, g, in n/(cm The equation is g1 l
dose rate (rad /h) = 1.9 Similarly the dose-equivalent rate at the middle level of the reactor is
- d. c. rate (rem /h) = 4.7 g and at the operating level
- d. e. rate (rem /h) = 14 g.
Since these equations are the result of all the measurements, good and bad statistics, the dose and dose-equivalent rates in some cases might be off by a factor of two.
The equations shoply.
state that there are one-to-one relationships between the total neutron flux dose rates and the dose-equivalent rates.
ACKNOWLEDGHENTS Special thanks are extended to J. E. McLaughlin who assisted in the measurements and in the preparation of this paper. We would also like to give out thanks to John Gulbin who did the TLD measurements, to Michael Boyle who provided excellent technical support, and to Edward Arrington who performed some of the computer programming, processed the large quantity of data and prepared the illustrations. We are also grateful to the management and technical staff of the six PWRs we visited for their assistance in performing these measurements.
REFERENCES
[1] ROSSI, H. H., MAYS, C. W., Health Phys. 34 (1978) 353.
[2] HANKINS, D. E., GRIFFITH, R. V., "A survey of neutrons inside the containment of a pressurized water reactor",
in Radiation Streaming in Power Reactors, USD0E Rep.
ORNL/RSIC-43 (1979) 114.
[3]
BRICKA, M., " Neutron measurements near PWR power reactors",
in Seventh DOE Workshop on Personnel Neutron Dosimetry, USDOE Rep. PNL-2807 (1978) 103.
[4] BUTLER, M.
H.,
OHNESORGE, W.
F., AUCIER, F.
A., " Measured distribution of neutrons inside containment of a PWR", in Radiation Streaming in Power Reactors, USDOE Rep. ORNL/
RSIC-43 (1979) 110.
[5] HANKINS, D. E., A Modified-Sphere Neutron-Detector, USDOE Rep. LA-3595 (1967).
[6] McLAUGHLIN, J. E., O' BRIEN, K., " Accelerator stray-neutron dostmetry: spectra of low-and intermediate-energy neutrons,"
in Neutron Monitoring, IAEA, Vienna (1967) 335
[7]
BENNETT, E. F., YULE, T. J., Techniques and Analyses of Fast-Reactor Spectroscopy with Proton Recoil Proportional Counterd',
USAEC Rep. ANL-7763 (1971).
[8] SANNA, P., S., Modification of an Iterative Code for Un-folding Neutron Spectra from Multisphere Data, USDOE Rep.
HASL-311 (1976).
[9] CASSON, R. M., BENTON, E. V. Nucleac Track Detection, 2, (1978) 173.
TABLE I.
DISTRIBUTIONS OF AVERAGE ENERGY, 5, AND OF QUALITY FACTOR, QF, FROM PWR NEUTRON MEASUREMENTS 5 (MeV)
Frequency QF Frequency 1egu iva l en t rate versus the total flus of unmoderated M
252Cf fission sources, line a; the reactor measurements on the operating floor level plus the D 0 moderated fission 2
source, line b; and the reactor measurements on the medium floor level, line c.