既然有人提起，就开个新贴讨论。首先，ALARP只适用于英国，所以其他国家有权不采用，当然法德都有类似的原则如GAMAB和MEM。其二，ALARP有法律含义：就是相关人士有义务把风险尽可能降低，只要在合理可行的前提下。这种义务制是英国特有，就像我们义务教育一样。因此如何定义某风险已经达到ALARP就是要靠义务者和权威人士的判断。这里必然带有个人的主观性。这个制度之所以通行，是基于西方发达国家内在的信任和职业道德体系上。说白了，我雇用你，你就有‘义务’在力所能及的前提下把所负责的事情办到最好。
至于如何判断所谓的合理可行，则由good practices/best practices这个概念决定。也就是说，当我们要对某个风险进行控制时，要考虑到现有已知的各种被行业接受或认同的技术或方案。最新但未成熟的肯定不是good practice。然后决定哪些该采用。这里的决定要考虑到费用和相应的安全得益的权衡比较。
补充一点，要降到ALARP，该风险必须可以容忍(tolerable)。这里容忍界限的定义考虑因素很多。不同国家和地区肯定不一样，每个地方的人都有自己可接受的安全模型，一般都是基于过去的安全数据。也就是说如果一个国家的事故率高，相应容忍界限应该要低些。所谓国情不一样，不能一概而论。其次不同行业也不一样，因为涉及的系统类型不一样。一般都是由国家定义一个可容忍的指标（如每年只能容许若个人员伤亡），然后分配到各个行业（如铁路），最后分配到不同的系统类型（如铁路信号）。必要时还要针对不同的人群来考虑容忍界限：如公众，用户，和公司职员等，因为他们的风险暴露不一样。而公司可以根据这些指标为参考来定义相应产品的风险容忍度。实践上可以使用risk matrix/graph来简化，细节以后再谈。
简单说，ALARP是一种安全风险模型，比较有名的一种。
Subscribe to:
Post Comments (Atom)
ALARP as a RAC principle is widely accepted and used in the world. Like you said, it is not mandatory in countries like Norway and Netherland for example. The definition of the RAC could be made by government or companies themselves.It varies from country to country.
ReplyDeleteThe essence of ALARP is to reduce the risk to a level where cost and potential risk reduction is grossly disproportional. And in the UK regime, it's the company responsibility to demonstrate this in the safety case.
The ALARP principle itself is easy to understand, however, it is rather hard to apply in practice, given the difficulty of demonstrating the gross disproportion between cost and risk reduction
If you look at other RAC approaches such as F-N model and risk buble, they all assume the same classification: unacceptable risk, broadly acceptable risk and tolerable risk. To determine the class of a particular risk, you can use risk matrix or compare with historical data. Once the risk is determined to be tolerable, you can adopt 'good practice' (qualitative method) or conduct cost-benefit analysis (quantitative method). Different industries also have different guidelines. For example, railways safety in the UK will publish annual strategic safety plan which instructs the budget of preventing a fatality
ReplyDeleteNot sure which one (risk classification, or determination of good practice, or cost benefit analysis) you think is difficult.
Risk classification and cost benefit analysis are difficult.
ReplyDeleteFor instance, if it is necessary to spend 2M dollars to save a potential fatality as long as the risk in the ALARP region. This is the same as asking if 2M dollar to 1 fatality is grossly disproportional. Again, here we unavoidable encounter the problem of pricing human life.
Under the UK system, it's the goverment who gives the unacceptable risk, broadly acceptable risk and tolerable risk. However, this is not the case in other countries. Take the norwegian oil industry for example, the boundaries of these three regions are decided by the companies themselves, so does how much they want to invest to reduce the risk by 1 unit, say one fatality.
The ALARP concept is well developed in the UK, and the British government has relevant regulations or guidelines to instruct the applications. But this is, again, not the case in other countries, eventhough the ALARP principle is used from time to time. It may be a good choice to follow the british way, or it may be not. The acutally situation needs to be take into account. The most obvious case is China, I am pretty sure strictly follow the british practice of ALARP will not work in China.
May be I should not raise so much doubt?