Inter-Risk Comparisons - ACS Publications - American Chemical Society

Energy and Environmental Policy Center, Jefferson Physical Laboratory, Harvard University, ..... College Football .... Living in Colorado compared wit...
0 downloads 0 Views 1MB Size
7 Inter-Risk Comparisons

Downloaded by EAST CAROLINA UNIV on January 12, 2018 | http://pubs.acs.org Publication Date: January 13, 1984 | doi: 10.1021/bk-1984-0239.ch007

Ε. A. C. CROUCH and RICHARD WILSON Energy and Environmental Policy Center, Jefferson Physical Laboratory, Harvard University, Cambridge, MA 02138

The comparison of different actions or processes for their risk content is contingent on the performance of some sort of risk assessment, which consists of the evaluation of some measure(s) of risk for those actions or processes. The particular measure(s) will depend on the reason for the assessment, for no single measure of risk is known which can encompass all aspects of risk. The need to evaluate risk measures usually requires an extrapolation of observations to new situations, a task performed by adopting models to describe how the meas­ ures vary. Such a procedure introduces various uncer­ tainties which should be incorporated into any state­ ments about risk. To put health risks from chemicals into perspective we compare some measures of risk for various aspects of everyday life with similar measures of risk from chemicals. In both cases we outline the models used in the risk assessment and the uncertain­ ties in the values obtained.

The d i s c u s s i o n o f r i s k s from any p a r t i c u l a r a c t i o n , process, or system often procèdes i n s p l e n d i d i s o l a t i o n , u s u a l l y with protagonists and antagonists ranged on two s i d e s o f an unbridgeable g u l f quoting c o n t r a d i c t o r y and alarming r i s k estimates a t each other. What we t r y t o do i n t h i s paper i s c u r s o r i l y point out why the c o n t r a d i c t i o n s may be only apparent, the g u l f b r i d g e a b l e , by i n d i c a t i n g where such apparent c o n t r a d i c t i o n s o f t e n a r i s e , and then go on to help remove the i s o l a t i o n and alarm by p r o v i d i n g a few examples from everyday l i f e with which to provide comparisons. For i t i s o f t e n the i s o l a t i o n o f r i s k estimates that make them seem alarming.

0097-6156/ 84/ 0239-0097506.00/ 0 © 1984 American Chemical Society

Rodricks and Tardiff; Assessment and Management of Chemical Risks ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

98

ASSESSMENT A N D MANAGEMENT OF CHEMICAL

RISKS

Comparisons o f r i s k s r e q u i r e s the e v a l u a t i o n i n various cases of s i m i l a r measures o f r i s k , a task that r e q u i r e s the model l i n g of r i s k y a c t i o n s o r processes i n order to e x t r a p o l a t e to new s i t u a t i o n s . We w i l l describe some of the problems of doing t h i s , together with some of the procedures used i n t h i s paper and the approximations i n v o l v e d . Against the background provided by everyday r i s k s , and with an a p p r e c i a t i o n o f the approximations and u n c e r t a i n t i e s i n v o l v e d , we can extend our " r i s k l i s t " (at l e a s t p a r t i a l l y ) to some o f the products o f the chemical i n d u s t r y and sketch a procedure f o r d e a l i n g with the r i s k s which a r i s e .

Downloaded by EAST CAROLINA UNIV on January 12, 2018 | http://pubs.acs.org Publication Date: January 13, 1984 | doi: 10.1021/bk-1984-0239.ch007

Measures of Risk In making comparisons between r i s k y a c t i o n s or processes, as i n making any comparisons, i t i s d e s i r a b l e to avoid attempting to compare u n l i k e q u a n t i t i e s . A q u a n t i t a t i v e r i s k assessment ends up with some number o r range o f numbers which d e s c r i b e ( s ) r i s k . A c l o s e r look w i l l r e v e a l the p o s s i b i l i t y of f i n d i n g a whole s e t of such numbers, each o f which d e s c r i b e s some p a r t i c u l a r aspect of the r i s k . Each member o f such a s e t i s j u s t one measure of that r i s k , and must be so t r e a t e d . Comparison with a d i f f e r e n t measure o f some other r i s k may be m i s l e a d i n g — i n d e e d comparison with a d i f f e r e n t measure o f the same r i s k can be confusing. Figures 1 and 2 show how two d i f f e r e n t measures of r i s k of a c c i dental death f o r the U.S. c o a l i n d u s t r y v a r i e d over the 20 year p e r i o d from 1950 to 1970. One f i g u r e seems to i n d i c a t e that the i n d u s t r y got s u b s t a n t i a l l y " s a f e r " over that p e r i o d , while an opposite c o n c l u s i o n may be i n f e r r e d from the other. Each measure represents a d i f f e r e n t aspect of the r i s k o f a c c i d e n t a l death, and whether they support or deny any conclusions as to the s a f e t y of the c o a l i n d u s t r y depends, i n t e r a l i a , upon a d e f i n i t i o n o f " s a f e t y " i n t h i s context. S i m i l a r apparently c o n t r a d i c t o r y measures o f r i s k may be constructed i n other cases, and they are u s e f u l f o r emphasizing the n e c e s s i t y of c l e a r d e f i n i t i o n . The purpose of the r i s k assessment has to be w e l l - d e f i n e d before s u i t a b l e r i s k measures can be constructed, and comparisons between d i f f e r e n t r i s k measures can e a s i l y be ambiguous. System Boundaries Another "apples and oranges" comparison can a r i s e when r i s k assessments have been performed on two or more p u t a t i v e a l t e r n a t i v e s , f o r example c o n s t r u c t i o n and operation o f " c o n v e n t i o n a l " versus "renewable energy" e l e c t r i c power p l a n t s . Even i f the same r i s k measure i s used i n each case, l i t t l e i s gained i n t h e i r comparison i f the a l t e r n a t i v e s are not e q u i v a l e n t (or d i f f e r e n t ) i n some w e l l - d e f i n e d way. Changing the d e f i n i t i o n o f equivalence may a l t e r any conclusions t o be drawn from a r i s k assessment.

Rodricks and Tardiff; Assessment and Management of Chemical Risks ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

Downloaded by EAST CAROLINA UNIV on January 12, 2018 | http://pubs.acs.org Publication Date: January 13, 1984 | doi: 10.1021/bk-1984-0239.ch007

CROUCH A N D WILSON

Inter-Risk Comparisons

O.o-I—.—.—.—ι—.—.—.—ι—.—.—.—ι—.—.—.—i—*—.—.—I 1950

1954

1958

1962

1966

1970

DATE

Figure 1. U.S. Coal Industry, 1950-1970. l i o n tons output.

Deaths per m i l ­

Rodricks and Tardiff; Assessment and Management of Chemical Risks ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

100

A S S E S S M E N T A N D M A N A G E M E N T OF C H E M I C A L RISKS

Downloaded by EAST CAROLINA UNIV on January 12, 2018 | http://pubs.acs.org Publication Date: January 13, 1984 | doi: 10.1021/bk-1984-0239.ch007

This i s apparent i n the example quoted, where the d i f f e r e n t types of power p l a n t s may be constructed to be equivalent i n d i f f e r e n t ways. Conventional p l a n t s are u s u a l l y designed to meet s p e c i f i c a v a i l a b i l i t y ( p r o b a b i l i t y of being able to supply power when c a l l e d upon to do so) and power output goals, whereas a more sens i b l e way of designing "renewable energy" p l a n t s , e s p e c i a l l y those powered by wind or sun which provide energy a t times not d i c t a t e d by man, might be f o r maximum energy output with l i t t l e account taken of a v a i l a b i l i t y . Once again the reason f o r the r i s k assessment i s of paramount importance i n d e c i d i n g what measures of r i s k to compare and f o r which systems those measures of r i s k have to be evaluated. Modelling and u n c e r t a i n t y Any r i s k assessment r e q u i r e s the i m p l i c i t or e x p l i c i t use of modelling to d e s c r i b e the process which i s being assessed and a s s o c i a t e r i s k with i t . The model i s then used to e x t r a p o l a t e to the s i t u a t i o n of i n t e r e s t i n the r i s k assessment. A c a r e f u l c o n s i d e r a t i o n of even the simplest assessments w i l l show t h i s general p a t t e r n and a s s i s t i n i n d i c a t i n g the u n c e r t a i n t i e s which n e c e s s a r i l y a r i s e as a r e s u l t . As an example, we take the r i s k of death i n auto a c c i d e n t s . F i r s t we need to s e i e c t some u s e f u l measure of r i s k . For t h i s example we w i l l use the United States population average annual p r o b a b i l i t y of dying as our r i s k measure. There are abundant data f o r the past behaviour of t h i s measure, some of which are p l o t t e d i n F i g u r e 3. I t appears that t h i s r i s k has been f a i r l y constant i n the past few years, with o c c a s i o n a l jumps such as that i n 1973-1974, but no s i g n i f i c a n t long term trend i s appar e n t . On t h i s b a s i s we might propose that t h i s measure of r i s k i s a constant, with random annual v a r i a t i o n s . Such a proposal would then c o n s t i t u t e our model, which we would f i t to the data and f i n d that the r i s k i s , on average, 24 per 100,000 per year, with a random year to year s c a t t e r of about 10%. On t h i s b a s i s we might then suggest that i n f u t u r e t h i s same measure of r i s k would a l s o be 24 per 100,000 per year plus or minus ten per cent. The procedure used here was to propose a p l a u s i b l e ad hoc model, obtain the parameters of that model by f i t t i n g to h i s t o r i c a l data, and then e x t r a p o l a t e (to the future) u s i n g the model. The two parameters obtained were the average value (24 per 100,000 per year) and the average annual v a r i a b i l i t y (10%). There are two sources of u n c e r t a i n t y i n t h i s procedures, the f i r s t easy to handle but the other very d i f f i c u l t . The f i r s t a r i s e s i n f i t t i n g the model to a v a i l a b l e data i n order to estimate the model parameters. The values obtained f o r the parameters w i l l be subject to the usual s t a t i s t i c a l u n c e r t a i n t i e s a s s o c i a t e d with f i t t i n g

Rodricks and Tardiff; Assessment and Management of Chemical Risks ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

7.

CROUCH AND WILSON

Inter-Risk Comparisons

101

Downloaded by EAST CAROLINA UNIV on January 12, 2018 | http://pubs.acs.org Publication Date: January 13, 1984 | doi: 10.1021/bk-1984-0239.ch007

t h e o r e t i c a l models to observed p o i n t s , but such u n c e r t a i n t i e s can themselves be estimated and d e a l t with by standard procedures. The second source of u n c e r t a i n t y i s i n the choice of model and the v a l i d i t y of the e x t r a p o l a t i o n process. Although i n t h i s case the model we chose i s p l a u s i b l e when l o o k i n g at F i g u r e 3 , a l i t t l e r e f l e c t i o n w i l l show that i t c o u l d be completely w r o n g — the constancy of t h i s measure i s c e r t a i n l y not fundamental and may have a r i s e n i n the past f o r t u i t o u s l y . In other words, while our model may adequately f i t (or describe) the data i n the past, i t does not f o l l o w that i t gives a mechanistic d e s c r i p t i o n of what a c t u a l l y happened then nor of what may happen i n the f u t u r e . A l t e r n a t i v e models can be p o s t u l a t e d which may be b e t t e r repres e n t a t i v e of the world, yet behave completely d i f f e r e n t l y when e x t r a p o l a t e d i n t o the f u t u r e . Figure 4 shows another measure of r i s k f o r auto a c c i d e n t s , the average number of deaths per v e h i c l e mile t r a v e l l e d , which shows a d e c l i n i n g trend with time, which trend may continue i n t o the f u t u r e . I f we are i n t e r e s t e d i n e s t i m a t i n g the average r i s k of death i n the p o p u l a t i o n , using t h i s l a s t measure a l s o r e q u i r e s an estimate of v e h i c l e miles t r a v e l l e d . E x t r a p o l a t i n g to new s i t u a t i o n s (e.g. the future or to a d i f f e r e n t country) may be more s a t i s f a c t o r y using models of Figure 4 together with models of how v e h i c l e miles t r a v e l l e d w i l l vary, r a t h e r than using a simple f i t to F i g u r e 3 , since such an approach autom a t i c a l l y contains the i n t u i t i v e l y o b v i o u s — t h a t gross v a r i a t i o n s i n the t o t a l amount of d r i v i n g w i l l have some e f f e c t on the numbers k i l l e d . There i s no way i n which the s i z e s of the u n c e r t a i n t i e s i n troduced by f a i l u r e to choose the " r i g h t " model may be r i g o r o u s l y estimated. To use any model one u s u a l l y has to make a l a r g e number of i m p l i c i t or e x p l i c i t assumptions, many o f which cannot be t e s t e d with a v a i l a b l e data, although obviously i t helps i f any models chosen agree with what data i s a v a i l a b l e . E x t r a p o l a t i o n s based on models thus have to be made on the b a s i s of p l a u s i b i l i t y , and u n c e r t a i n t i e s due to i n c o r r e c t choice of model can only be guessed at i f one i s prepared to accept some assumptions—such as by accepting that a c e r t a i n c l a s s of models encompasses the only p o s s i b i l i t i e s , and f i n d i n g the spread i n e x t r a p o l a t e d r e s u l t s for every member of that c l a s s . Everyday

Risks of L i f e

Bearing i n mind the dangers of m i s i n t e r p r e t a t i o n and the l i k e l i hood of e r r o r s which we have j u s t discussed, i t i s u s e f u l to appreciate the magnitudes of some of the r i s k s we face i n everyday l i f e . Table 1 presents a few such values f o r o c c u p a t i o n a l r i s k s of death i n U.S. i n d u s t r i e s . These values w i l l provide a u s e f u l anchoring p o i n t f o r comparison with some of the values we obtain l a t e r f o r other r i s k s . Notice that a r i s k of one i n a

Rodricks and Tardiff; Assessment and Management of Chemical Risks ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

ASSESSMENT A N D MANAGEMENT OF CHEMICAL

RISKS

Downloaded by EAST CAROLINA UNIV on January 12, 2018 | http://pubs.acs.org Publication Date: January 13, 1984 | doi: 10.1021/bk-1984-0239.ch007

30

•1 • 1950



» 1 » » •

1954

I



1958





I

1

••

1962 DATE

ι ... I 1966

1970

1974

1978

Figure 3. U.S. Motor-vehicle accident deaths, 1950-1970. Deaths per 100,000 population.

8.0-

o ο or

£ 2.0-

α O.ol 1950

I 1954

ι

1958

••ι

ι

1962

. . I . . 1966

I 1970

1974

1978

DATE

Figure 4. U.S. Motor-vehicle accident deaths, 1950-1970. Deaths per 100 m i l l i o n v e h i c l e m i l e s .

Rodricks and Tardiff; Assessment and Management of Chemical Risks ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

7.

C R O U C H A N D WII S O N

103

Inter-Risk Comparisons

m i l l i o n per year a p p l i e d to the whole U.S. population would r e s u l t i n an annual death t o l l of about 240, but the r i s k s shown i n Table 1 are a p p l i c a b l e only to various subpopulations covered by the designated i n d u s t r y group.

Table I.

U.S. Occupational Risks i n 19 78 or i n the Year Shown

Downloaded by EAST CAROLINA UNIV on January 12, 2018 | http://pubs.acs.org Publication Date: January 13, 1984 | doi: 10.1021/bk-1984-0239.ch007

Industry Group Trade Manufacturing Service and Government Transport and Public U t i l i t i e s Agriculture Construction Mining and Quarrying

Annual Occupational Risk of Death (1978) 5.3 χ 10J! 8.2 χ 10

Variability (percent)

Trend

15 8

Yes Yes

χ 10"

4

8

Yes

3.7 χ 1 0 ~ 6.0 χ 10 6.1 χ 10"^

4

16 9 6

No No Yes

9.5

χ 10"

4

22

No

3.6

χ 10"* (1977)

5.9

χ 1 0 " (1971)

2.2 2.4

χ 1 0 ~ (1978) χ 10 (1977)

1.0

More f i n e l y d i v i d e d grouping: Farming Stone q u a r r i e s and m i l l s Police Officers ( i n l i n e o f duty) R a i l r o a d Employee Steelworker (accident only) Firefighter

7

4

20

4

19 7

4

? ?

2.8 χ 1 0 " (19 72) 8.0 χ 10 (1972)

The values obtained i n t h i s t a b l e were obtained by a p p l y i n g a p a r t i c u l a r model. For each year from 1955 to 19 78 the measure shown was. computed by d i v i d i n g reported o c c u p a t i o n a l deaths i n the i n d u s t r i e s l i s t e d by the reported average work f o r c e . I t was then assumed that t h i s measure v a r i e s l i n e a r l y with date (and i s i n d e ­ pendent of v a r i a t i o n s i n o c c u p a t i o n a l populations, average hours worked, average experience o f the working p o p u l a t i o n , e t c . , except i n s o f a r as these things vary l i n e a r l y with date), so that a simple l i n e a r time trend could be e x t r a c t e d from the raw data. The value of the r e s u l t a n t f i t t e d model i n 1978 i s l i s t e d i n Table Τ i f there was a s i g n i f i c a n t time trend, otherwise the average value over 1955 to 1978 i s l i s t e d . The v a r i a b i l i t y recorded represents the standard d e v i a t i o n o f the observed values about the t h e o r e t i ­ c a l model, as a percentage of the 1978 estimate. For r e p o r t i n g

Rodricks and Tardiff; Assessment and Management of Chemical Risks ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

104

ASSESSMENT A N D M A N A G E M E N T OF C H E M I C A L RISKS

Downloaded by EAST CAROLINA UNIV on January 12, 2018 | http://pubs.acs.org Publication Date: January 13, 1984 | doi: 10.1021/bk-1984-0239.ch007

on h i s t o r i c data the model adopted i s o f t e n adequate, but f o r ext r a p o l a t i o n purposes i t may be gravely inadequate—examples of e x t r a p o l a t i o n hwere would be f o r p r e d i c t i o n of f u t u r e r i s k s , pred i c t i o n of r i s k s i n i n d i v i d u a l i n d u s t r i e s (components of the i n d u s t r i a l groupings l i s t e d ) , p r e d i c t i o n s of r i s k s i n s i m i l a r i n d u s t r y groups i n other c o u n t r i e s . There i s c l e a r l y c o n s i d e r a b l e v a r i a t i o n between i n d u s t r i a l groupings (by a f a c t o r of ten) i n t h i s measure of the occupational r i s k of death borne by employees. The v a r i a t i o n i s probably l a r g e r between i n d i v i d u a l i n d u s t r i e s , s i n c e the values given are averages over i n d u s t r y groups which o f t e n contain s u b s t a n t i a l l y d i f f e r e n t components. Some i d e a of t h i s v a r i a t i o n may be seen from the second p a r t of Table 1, which shows s i m i l a r l y computed measures of r i s k f o r a few subgroups of employees. Comparison of the two t a b l e s shows v a r i a t i o n s up to a f a c t o r of ten w i t h i n the groupings of Table I, and i t i s almost c e r t a i n that larger v a r i a t i o n c o u l d be found with f u r t h e r s t u d i e s of other subgroups. Nevertheless a general c o n c l u s i o n i s that o c c u p a t i o n a l r i s k s of death l i e i n the range of one i n ten thousand to one i n a thousand per year, with the o r d e r i n g of i n d u s t r i e s being approximately as one would expect. Table II l i s t s a set of commonplace r i s k s of a c c i d e n t a l death i n the United States. As i n Table I, s i g n i f i c a n t time trends have been f a c t o r e d out of these v a l u e s — t h e y represent a value estimated f o r 1977 and 1978, based on a sequence of s e v e r a l years. For comparison with these a c c i d e n t a l r i s k r a t e s , the r i s k of death by homicide i n the U.S. i n 1976 was about 9 per 100,000. Risks of a c c i d e n t a l death i n various sports i s shown i n Table I I I . Perhaps these could be i n t e r p r e t e d as showing what we are prepared to do to o u r s e l v e s , compared with what we are prepared to have imposed upon us. These values correspond to the annual average r i s k of death f o r those p a r t i c i p a t i n g i n the sport. There i s a l a r g e unc e r t a i n t y i n the value of most of these r i s k s , corresponding to a f a c t o r of 2 or 3, s i n c e although the number of deaths i s u s u a l l y a c c u r a t e l y known, the number of people p a r t i c i p a t i n g i n each sport i s h i g h l y u n c e r t a i n . From t h i s l i s t i t would appear that going up i n t o the a i r i n almost any (noncommercial) way or down i n t o the deapths of the sea (scuba d i v i n g ) i n any way are both a s s o c i a t e d with comparatively high r i s k s ! One might immediately wish to s t a r t comparing the values i n the v a r i o u s t a b l e s , and t h i s i s p o s s i b l e . But r e c a l l the e a r l i e r d i s c u s s i o n on comparisons of d i f f e r e n t t h i n g s . Although the r i s k measures are s i m i l a r i n the v a r i o u s tables, the purpose of any such comparison must be made c l e a r before attempting i t , f o r a d i f f e r e n t r i s k measure might be more a p p r o p r i a t e . For example most employees are at work f o r a l a r g e f r a c t i o n of the time throughout the year, whereas s p o r t s are played only i n t e r m i t t e n t l y . A measure of r i s k which took t h i s d i s p a r i t y i n t o account may be of greater value.

Rodricks and Tardiff; Assessment and Management of Chemical Risks ACS Symposium Series; American Chemical Society: Washington, DC, 1984.

7.

CROUCH AND WILSON

Table

II.

105

A c c i d e n t a l R i s k s o f D e a t h i n t h e U.S. i n 1977

Accident

Motor V e h i c l e A l l Home A c c i d e n t s Fall Drowning Fire

Downloaded by EAST CAROLINA UNIV on January 12, 2018 | http://pubs.acs.org Publication Date: January 13, 1984 | doi: 10.1021/bk-1984-0239.ch007

Inter-Risk Comparisons

Inhalation/Ingestion of Objects Accidental Poisoning Firearms (accidents)

Annual Average R i s k o f Death

2.4 1.1 6.2 3.6 2.8

Flood Lightning

1.5 1.4 1.0 5.3 6 6 5

T r o p i c a l Cyclone/ Hurricane Bite/Sting

3 2

Electrocution Tornado

X X X X X

1

0

~

io"! -5 10

10

5

5

X 10 io X io X X X X X

1 1 1 10 'I io

1

0

io

1

- 7

Variability

o r 1978 Trend

10 5 6 7 5

No Yes Yes No Yes

8