FLOOD08791 - Laserfiche WebLink

Home Browse Search

53 of CUR, 1990). In both those cases there is a proportionality between the acceptable probability of failure and the square of the number of deaths. In the 1994 Guidelines, the second approach, of having a progressive increase in slope, was adopted. There are reasons for thinking that the second approach more nearly reflects community values (Higson, 1990). Governments or corporations spend heavily to prevent recurrence of accidents involving many fatalities (Granville rail disaster, Kempsey bus smash). As a result, curves of national accident statistics tend to show a progressive increase in slope as N increases (see Figure I of DOP, 1992). The 1994 Guidelines interim criteria for Societal Risk were based on those proposed by Higson (1990) for the nuclear industry. So far as is known, Higson's curves have not been adopted in any other jurisdiction. In 1994, these curves (see Figure I) seemed defensible, though probably conservative, for dams because they started, at the left hand axis, at probability levels consistent with existing ANCOLD detenninistic guidelines (ANCOLD, 1986) and they showed an accelerating reduction in expected value of life loss as N increased. The concepts of Limit, Objective and ALARP apply with Societal Risk as they do with Individual Risk. t.OE.3 Unacceptable risks Z h 1.0E-4 ~ "0 " " !2 "0 " (3 1.0E.S " 0- x " € .~ ~ '" " >- 10E-6 ~ ' 0- E '" "0 ~ 0- ~ 10E.] .;; , E "0 ~ :c '" .g 1.DE-S ct Risks 10 be as row as reasonably practicable . the ALAR? principle De minimis region Acceptable risks 10 100 1000 10000 Number of fatalities due to dam failure, N (Refer ANCQlD Fi9Jfe 2.7 : Ir'lterim ANCOLD Societal Risk Criteria) Figure 1 Interim Societal Risk Criteria of the 1994 Guidelines It was acknowledged in the 1994 Guidelines that the whole issue of Societal Risk criteria is difficult and controversiaL For example, HSE (1989) and DOP (1992) had both considered the issue and had decided not to set quantitative criteria.