Detection in Analytical Chemistry - ACS Publications - American


Detection in Analytical Chemistry - ACS Publications - American...

1 downloads 121 Views 526KB Size

Chapter 14

Downloaded by UNIV OF MASSACHUSETTS AMHERST on May 23, 2018 | https://pubs.acs.org Publication Date: December 9, 1987 | doi: 10.1021/bk-1988-0361.ch014

Radioactivity Analyses and Detection Limit Problems of Environmental Surveillance at a Gas-Cooled Reactor James E. Johnson and Janet A. Johnson Department of Radiology and Radiation Biology, Colorado State University, Fort Collins, CO 80523 The Lower Limit of Detection (LLD) values required by the USNRC for nuclear power facilities are often difficult to attain even using state of the art detection systems, e.g. the required LLD for I-131 in air is 70 fCi/m . For a gas-cooled reactor where I-131 has never been observed in effluents, occasional false positive values occur due to: counting statistics using high resolution Ge(Li) detectors, contamination from nuclear medicine releases and spectrum analysis systematic error. Statistically negative concentration values are often observed. These measurements must be included in the estimation of true mean values. For this and other reasons, the frequency distributions of measured values appear to be log-normal. Difficulties in stating the true means and standard deviations are discussed for these situations. 3

The Fort St. Vrain High Temperature Gas-cooled (HTGR) power reactor, operated by Public Service Company of Colorado i s located approximately MO miles north of Denver at the confluence of the South Platte r i v e r and St. Vrain creek. I t i s the only gas-cooled power reactor i n the country and while i t has had operating d i f f i c u l t i e s , the nuclear aspects of the design have great promise f o r the following reasons: 1. Net e l e c t r i c a l e f f i c i e n c y i s 39.2$ 2. Extremely low in-house worker radiation dose rates 3. Extremely low r a d i o a c t i v i t y release to the environment The l a s t two c h a r a c t e r i s t i c s are primarily due to the unique HTGR fuel element design and the rather innocuous environment of the core. Table I i l l u s t r a t e s this point with a comparison of HTGR effluents with those from B o i l i n g Water Reactors (BWR) and Pressurized Water Reactors (PWR). 0097-6156/88/0361 -0266$06.00/0 © 1988 American Chemical Society

Currie; Detection in Analytical Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 1987.

14.

JOHNSON & JOHNSON

Table I:

Radioactivity Released by Reactor Type H-3 (Ci/MWe-year) Gaseous Liquid 0.05 0.1 0.2 1.2 0.1 6.5

BWR PWR HTGR

Downloaded by UNIV OF MASSACHUSETTS AMHERST on May 23, 2018 | https://pubs.acs.org Publication Date: December 9, 1987 | doi: 10.1021/bk-1988-0361.ch014

Radioactivity Analyses & Detection Limits

Cs-137 (mCi/MWe-year) Liquid 25 1 .7 2.8 E-4

Since the r a d i o a c t i v i t y release i s so low, the r a d i o a c t i v i t y con­ centration i n a l l environmental sample types i s e s s e n t i a l l y a l l due to natural radiation background sources with f a l l o u t and releases from other industries p r i n c i p a l l y medical, superimposed upon i t . The background source term due to primordial radionuclides i s e s s e n t i a l l y constant and the frequency d i s t r i b u t i o n can be normally or log-normally distributed. The f a l l o u t source term i s time de­ pendent and highly variable and generally log-normally d i s t r i b u t e d . Since the objective of a reactor environmental monitoring program i s to document the presence or absence of radioactive materials due to reactor effluents, i t i s imperative that background i n the reactor environs be documented. For Fort St. Vrain the preoper­ ational period occurred during periods of s i g n i f i c a n t Chinese weapons test f a l l o u t and e s s e n t i a l l y a l l r a d i o a c t i v i t y data shows lower values after reactor start-up. This makes comparison of operational periods to preoperational periods of no value. Thus i t i s necessary to compare s i t e data near the reactor or i n the predominant wind directions to control data. Figure 1 shows gross beta p a r t i c u l a t e concentrations i n a i r . Gross beta measurements include both background (due to Κ-Η0 and the U-238 and Th-232 series) with f a l l o u t superimposed. Shown are the half yearly arithmetic means f o r the four stations located i n the predominant wind directions from the reactor and the means f o r the reference (control) a i r sampling stations. There i s no difference i n the means between the two station sets. The ex­ tremely large temporal v a r i a t i o n i s due to f a l l o u t from the Chinese atmospheric weapons tests during the period. (Chernobyl f a l l o u t measured during the spring of 1986 reached weekly maxima of 560 fCi/m ). Figure 2 shows that f a l l o u t Cs-137 concentrations i n milk are log-normally distributed. (The geometric mean (median) multiplied by and divided by the geometric standard deviation includes 68.1 % of the area under the frequency d i s t r i b u t i o n . ) Figure 3 shows gross beta p a r t i c u l a t e concentrations i n pre­ c i p i t a t i o n at the two f a c i l i t y s i t e c o l l e c t o r s . The t o t a l data ( s o l i d c i r c l e s ) , i f assumed to be log-normally distributed, would be very misleading. There are c l e a r l y two separate frequency d i s ­ tributions with very d i f f e r e n t geometric means and geometric stand­ ard deviations. The high a c t i v i t y concentration d i s t r i b u t i o n i s assumed to be f a l l o u t and the lower one due to natural background. Iodine-131 i s certainly the key radionuclide i n reactor environmental monitoring. I t i s an indicator of release since the radionuclide i s almost always gaseous and i t i s the predominant f i s s i o n product contributor to radiation dose to the general public due to i t s food chain mobility. Reactor license requirements therefore put great emphasis on 1-131 measurements i n a i r and 3

Currie; Detection in Analytical Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 1987.

267

DETECTION IN ANALYTICAL CHEMISTRY

Downloaded by UNIV OF MASSACHUSETTS AMHERST on May 23, 2018 | https://pubs.acs.org Publication Date: December 9, 1987 | doi: 10.1021/bk-1988-0361.ch014

268

Figure 1985.

1.

Gross

beta concentrations i n a i r for 1974

through

Currie; Detection in Analytical Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 1987.

Downloaded by UNIV OF MASSACHUSETTS AMHERST on May 23, 2018 | https://pubs.acs.org Publication Date: December 9, 1987 | doi: 10.1021/bk-1988-0361.ch014

14.

JOHNSON & JOHNSON

Radioactivity Analyses & Detection Limits

Figure 2. Cs-137 in milk, p C i / l i t e r , area, 197^ and 1975.

composites from adjacent

Currie; Detection in Analytical Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 1987.

269

270

DETECTION IN ANALYTICAL CHEMISTRY

milk. The detection l i m i t (LLD) f o r 1-131 i n a i r required by the USNRC i s 70 fCi/m . Using activated charcoal as the c o l l e c t i o n medium and a fixed a i r sampler flow rate, i t i s possible to achieve this value with Ge(Li) spectroscopy and counting times of approximately 300 minutes. (The net count rate due to 1-131 i s obtained by subtracting the sum of the count rates i n four channels adjacent to the peak on each side from the sum of the count rates i n the eight channels i n the peak.) During a l l of 1985 the Fort St. Vrain reactor was shut down, therefore no 1-131 could possibly have been released. 1-131 i s released from l o c a l hospitals after patient diagnostic procedures and treatment with 1-131 but this release i s confined to surface waters. Figure H shows the measured 1-131 concentration at a l l seven of the a i r sampling stations f o r a l l of 1985. I f , i n f a c t , there was no source term f o r 1-131 i n a i r , the expected frequency d i s t r i b u t i o n of the measured values would be due only to methodo­ l o g i c a l uncertainty and would be expected to be normally d i s t r i ­ buted. The arithmetic mean would be expected to be zero. Figure 4 shows the observed mean was actually -1.5 fCi/m indicating a negative systematic error, probably due to bias i n the subtraction method used to obtain net count rate. The frequency d i s t r i b u t i o n was indeed normal. This i l l u s t r a t e s the necessity of including a l l negative values i n data analysis. I f only positive values were averaged, the mean would obviously be biased toward a f a l s e higher value. I t also i l l u s t r a t e s a method of determining systematic error. Using the s i m p l i f i e d expression of currie 0_) f o r S ( S = 2.33 s^) the a - p r i o r i , S f o r our measurement parameter was 33 fCi/m . For the data shown i n Figure 4 (n =356), only 1.2$ of the values exceeded the S value where 5% f a l s e positive values would be expected. I f the negative bias i s taken into account and the d i s t r i b u t i o n normalized to a mean of zero by adding 1.5 fCi/m , more of the values would be greater than 33 fCi/m bringing the f a l s e positive percentage closer to the expected 5$.

Downloaded by UNIV OF MASSACHUSETTS AMHERST on May 23, 2018 | https://pubs.acs.org Publication Date: December 9, 1987 | doi: 10.1021/bk-1988-0361.ch014

3

3

Q

Q

3

3

3

The a - p r i o r i S determined from the sum of the count rates i n c channels adjacent to the peak can be compared to the S based on σ determined f o r the net peak count rates for the 356 t r i a l s : ο q

σ

7

ο

2

= S.Ε.M. (η/ * = (0.63) (356)^ = 12 fCi/m S = 1.65 σ = (1 .65) (12) = 20 fCi/m c ο

3

3

As stated above, detection of effluent releases depends upon comparing mean a c t i v i t y concentrations over a period of time with those i n a reference or control area, as close i n c h a r a c t e r i s t i c s as possible to the reactor area. To properly compare means the appropriate variances must be used. When reporting standard devi­ ations therefore, i t i s important to remove the f r a c t i o n of the t o t a l uncertainty which i s due to the method. Since the precision (random uncertainty) to be attached to the method i s commonly determined with spiked samples, only the c o e f f i c i e n t of v a r i a t i o n may be used. Figures 5, 6 and 7 i l l u s t r a t e an approach to sub­ t r a c t i n g the c o e f f i c i e n t of v a r i a t i o n of the method from Sr-90 measurements i n s o i l . Figure 5 shows that the observed t o t a l

Currie; Detection in Analytical Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 1987.

Downloaded by UNIV OF MASSACHUSETTS AMHERST on May 23, 2018 | https://pubs.acs.org Publication Date: December 9, 1987 | doi: 10.1021/bk-1988-0361.ch014

14.

JOHNSON & JOHNSON

Radioactivity Analyses & Detection Limits

CUMULATIVE % Figure 3. Total gross beta i n p r e c i p i t a t i o n , F-1 and F-i» combined July 1976 through June 1977 (F-1 and F-4 were the sample locations f o r two large precipitation collector funnels). _ X = -1.5fCi/m 3

S.E.M. -

0.63

normal dist., α

=

0.01

η = 356

Se = 33 fCi/m

40

25

10

0

fCi/m Figure 4.

1-131

5

20

35

3

concentrations i n a i r , 1985.

Currie; Detection in Analytical Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 1987.

3

271

272

DETECTION IN ANALYTICAL CHEMISTRY 2

2 =

°Total

2 +

°Env.

Method

If Independent, 2

2

β

Downloaded by UNIV OF MASSACHUSETTS AMHERST on May 23, 2018 | https://pubs.acs.org Publication Date: December 9, 1987 | doi: 10.1021/bk-1988-0361.ch014

^Method

2

+

^Sampling Method

Since

2

+

^Sample Prep.

. 2

+

Chemical Sep.

Counting

often determined with spiked sample,

2

q

method

Use:

V X /Total

*

Figure 5. Determination environmental v a r i a t i o n .

x

v

' Env.

of

x

Method

uncertainty

due

to

method

and

From a single homogenized sample η

=

10 2

(-2) V

X

=

0.22

(i)

'method

=

0.05

method

This includes counting uncertainty

Method:

89.90

9 0

S r

Sr

b

y

by

SrC0 i,SrN0 i 3

9 0

3

Y(OH) i 3

Sr chemical yield by AAS of carrier.

Figure 1979.

6.

Precision

(random

uncertainty)

of

Sr-90

method,

Currie; Detection in Analytical Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 1987.

Downloaded by UNIV OF MASSACHUSETTS AMHERST on May 23, 2018 | https://pubs.acs.org Publication Date: December 9, 1987 | doi: 10.1021/bk-1988-0361.ch014

14.

Radioactivity Analyses θ Detection Limits

JOHNSON & JOHNSON

η

=

X σ

Figure 7. Sr-90 Vrain environs.

38

220

pCi/kg

- 140 pCi/kg

measurements i n surface s o i l ,

1980, F t . St.

Currie; Detection in Analytical Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 1987.

273

Downloaded by UNIV OF MASSACHUSETTS AMHERST on May 23, 2018 | https://pubs.acs.org Publication Date: December 9, 1987 | doi: 10.1021/bk-1988-0361.ch014

274

D E T E C T I O N IN A N A L Y T I C A L C H E M I S T R Y

c o e f f i c i e n t of v a r i a t i o n term squared i s e q u a l t o the c o e f f i c i e n t of variation of the method squared and the square of the c o e f f i c i e n t of v a r i a t i o n due o n l y t o e n v i r o n m e n t a l f a c t o r s . Since i t i s o n l y t h i s l a s t term t h a t s h o u l d be used t o d e s c r i b e the r e s u l t s f o r a g i v e n environment, the u n c e r t a i n t y due t o the method must be s u b t r a c t e d . I n t h i s case 10 r e p l i c a t e subsamples from a homogenized large sample were analyzed identically and the c o e f f i c i e n t of v a r i a t i o n determined t o be 0.22 ( F i g u r e 6 ) . T h i s i s not l a r g e c o n s i d e r i n g the number of c h e m i c a l and c o u n t i n g steps involved i n the method. F i g u r e 7 shows t h a t i f the method u n c e r t a i n t y i s s u b t r a c t e d (as the c o e f f i c i e n t of v a r i a t i o n squared) from the t o t a l c o e f f i c i e n t of v a r i a t i o n squared of a s e t of 38 s o i l samples a n a l y z e d f o r S r - 9 0 , the r e s u l t i n g c o e f f i c i e n t of v a r i a t i o n was 61$. A l t h o u g h i t i s t r u e t h a t o f t e n the m e t h o d o l o g i c a l uncertainty i s small compared t o b i o l o g i c a l or environmental v a r i a t i o n , the approach must s t i l l be used. When the u n c e r t a i n t y f r e q u e n c y d i s t r i b u t i o n of the method i s normal and the observed d i s t r i b u t i o n i s l o g - n o r m a l , t h i s approach i s not r i g o r o u s . I n c o n c l u s i o n , i t i s o b v i o u s t h a t e n v i r o n m e n t a l measurements a t r a d i o a c t i v i t y l e v e l s n e a r , or i n the case of F o r t S t . V r a i n , below preoperational backgrounds present a complex problem. A t t e n t i o n must be g i v e n t o d e t e r m i n a t i o n of the p r o p e r f r e q u e n c y d i s t r i b u t i o n of the d a t a and t o the p r o p e r s t a t i s t i c a l t e s t t o compare f a c i l i t y a r e a d a t a t o t h a t from background locations. Statistical methodology cannot always be r i g o r o u s l y applied and must be combined w i t h a common sense approach t o the available data. N e g a t i v e v a l u e s must always be i n c l u d e d i n the c a l c u l a t i o n of means and s t a n d a r d d e v i a t i o n s i n o r d e r t o a v o i d b i a s i n g r e s u l t s towards h i g h v a l u e s .

L i t e r a t u r e Cited 1.

C u r r i e , L. A.

RECEIVED May

19,

Anal. Chem. 1968,

40,

586-693.

1987

Currie; Detection in Analytical Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 1987.