Sir George Cayley
My apologies in the delay in replying – partly because student demands have been high (I like to say the students were revolting; not PC) – but mainly because I did not want to treat your important issue too lightly.
the right process should involve a calculation based on all the contributory factors that surround each runway, inc its surface, the LDA, wx, and design a/c plus most frequent a/c, to arrive at a figure that can be demonstrated as being acceptable on a LARP basis
(LARP or ALARP = as low as reasonably practicable)
UK CAP168, in Chapter 3, section 5, expands on this process, and calls for airports to review and determine on an annual basis the RESA distance required for individual circumstance taking into account, in their risk assessments, a range of such factors.
There are at least three diverse approaches to this aspect of safety, and they each have their advantages and weaknesses.
Discussing these for the case of RESAs for code 3 / 4 aircraft:
1. The ICAO approach of simple prescription (90m mandatory, 240m recommended).
I find great elegance and sophistication in the simplicity of ICAO rules like this. In the field, I am struck by how many of these simple rules (another example of such a rule is the 'average texture depth of not less than 1mm') give an excellent result even when tested by the real world variability. The result is often not improved by application of more complex methods. The rule simplicity means they can be applied by all of ICAO's many and diverse member states. If one calculates risk in societal terms, the net improvement in global safety by applying simple rules to all the Member States is greater than by applying highly sophisticated rules to a handful of advanced States. (And shopping mall managers and accountants can't fiddle them which must be good).
2. The UK CAP approach opens the door to the calculated risk assessment, ALARP school of thought.
This approach is similar to some ATC risk analyses, and derives from systematic risk assessment from nuclear, oil/gas, and chemical industries (to name but a few). In Australia, we have a relevant standard (AS 4360:1999 Risk Management), and the aviation regulator has various procedures on the Preparation of Safety Cases.
These approaches have policy/bureaucracy attractiveness because they "cover your arse". They have a real engineering attractiveness because one can then design the safety of a system in an engineering sense. They appeal to accountants because they form a major input into cost/benefit ratios. I've used this approach for some projects and found them useful in complex and very complex cases (I did an AHP Expert Choice analysis for a runway upgrade the other day – rather surprised with the result).
Despite the good bits of this approach, the cynic in me worries that they are too open to fiddling by shopping mall managers and accountants (or their paid-for consultants).
And I worry that they lack teeth. Let me ask the question - does the wording of the new CAP168 mean anything? After all, CAP168 reads wonderfully. Unfortunately for me in Australia, after listening to many years of politically correct, wonderfully reading, best world's practice, but incompetent bureaucrats have left me a bit cold to "wonderful reading". So I look for what really happens before I do the "wonderful reading". Uncomfortably for some, I know where to look so as to be able to look at what really happens. In the UK for example, I'd look at Bristol Airport (BRS?). Anyone know the RESA lengths at Bristol? I don't know them, but if they are 90m or less, then let me strongly condemn CAP168 and its "wonderful reading". If they are 150-240m, then let me say that this 'calculated risk assessment, ALARP' school of thought really works. I do look forward to finding out the RESA length.
3. The probability approach.
This approach is typified by the excellent analysis by Kirkland, Caves et al in their paper
An improved methodology for assessing risk in aircraft operations at airports, applied to runway overruns. I think there is great value here, although it is very sophisticated and very demanding of analysis. Perhaps too much for application to every airport at this stage, but in the next 10 or so years, perhaps we can do better.
There is a trio of papers which are going to be published the next 12-18 months and which will hopefully enable the probability approach to safety to be better applied to airports. The (only) one which I am not a co-author is titled 'Measures of Societal Risk and their potential use in Civil Aviation". This, I hope, gives the mathematics that will allow airports to make judgements about acceptable risk when using technology like the Kirkland approach.
Today, for me, I reckon that the lowest risk to global society is captured by the ICAO (simple) approach. But your approach is the best way forward –
improving safety thru dialogue. I'm wide open to dialogue.