So much for technology

Re: "Revere, then retire the shuttle," Feb. 23 Perspective.

Regarding the figures quoted by Bruce McCandless for the probability of losing a space shuttle, the realistic estimates (not NASA estimates) made by Richard Feynman at the time of the Challenger loss ranged from about one in 100 to one in several hundred. At that time, the failure rate of launch vehicles in general was said to be about one in 25 or 30. That has not really improved much over the last couple of decades.

The recently retired Ariane-4 satellite launcher, widely considered to be the most successful launcher yet developed, had three failures in 166 launches, although it did finish up its career with 74 consecutive successes. Its replacement, the Ariane-5, has had four failures in 14 flights. So much for technological improvements.

Presumably more care is taken with the shuttle; on the other hand, it's much more complex than a one-way satellite launcher. The net result seems to be, empirically, that the increased complexity cancels the extra care - leading to a failure probability not much different from the that of conventional satellite launchers.

The numbers don't tell us that flying the shuttle isn't a worthwhile endeavor - merely that it will be a dangerous endeavor for a long time to come.

LOWELL MORGAN

Monument