Path: utzoo!utgpu!jarvis.csri.toronto.edu!db.toronto.edu!hogg
Newsgroups: sci.space.shuttle
From: hogg@db.toronto.edu (John Hogg)
Subject: Re: Why 3 SSME's?
Message-ID: <1989Oct11.185113.26962@jarvis.csri.toronto.edu>
Organization: University of Toronto, CSRI
References: <538.252A3A3B@mamab.FIDONET.ORG> <4900@tekig5.PEN.TEK.COM>
Date: 11 Oct 89 22:51:15 GMT

In article <4900@tekig5.PEN.TEK.COM> robina@tekig5.PEN.TEK.COM (Robin Adams) writes:
>>[Proposal of ``PES'' system to explosively separate shuttle cabin from
>>rest of craft in Challenger-type emergency, and land it under parachute.]
>
> -- I agree. I've always felt NASA's first obligation is to ensure that all 
>aspects of Challenger should not be able to repeat. 
>
>[Description of additional ``safety'' devices.]

I myself have always felt that NASA's first obligation is to carry out
research into aeronautics and space.  This can't be done without risk;
the key is risk management.

Each ``safety feature'' that is added to a shuttle carries at least three
costs:
	1) Its dollar cost.  Would you rather spend $10 000 000 (random
	   figure) on Shuttle safety, where it *may* save seven lives, or
	   on highway safety?  Or AIDS research?  Or supplying a proper
	   diet to expectant mothers?  Or flood control?  Add your own
	   life-saving program...

	2) Its safety cost.  As Henry Spencer points out every so often,
	   military ejection seats regularly kill careless maintenance
	   workers.  Adding any sort of escape system to a shuttle incurs
	   a non-zero risk of the system being deployed at the wrong time,
	   with potentially disastrous consequences.  Many KSC engineers
	   were *very* worried that the recent dousing of Columbia by a
	   (safety!) sprinkler system might have created some hidden
	   danger.

	3) Its opportunity cost.  Every kilo of ``safety'' equipment is a
	   kilo less of payload.  That's a kilo more that has to be taken
	   up by another flight.  Is it safer to fly n times without the
	   escape system, or n+1 with?

I'd start ranting at this point about how Mary Shafer is dead on, and I'd
jump at a shuttle berth, but we went through that post-51L.  Let's let it
lie this time.
---
John Hogg			hogg@csri.utoronto.ca
Department of Computer Science, University of Toronto