Thursday, May 03, 2007

Things I Know #12 & British Midland Flight 092

"Believed to be" and "is" are not synonymous.

Example: The CO2 increase is generally believed to be due to the combination of increased burning of fossil fuels and (mostly before 1905) deforestation.

The CO2 increase is to be due to the combination of increased burning of fossil fuels and (mostly before 1905) deforestation.

Not the same statement. "Has been shown to be" would be much closer to being synonymous with is.

If we believe that the earth's climate is in some moral, stable state that is just because that's the way it's supposed to be -- and we see it change -- we are likely to believe that we had something to do with it. Not that it's a bad thing to suspect, but you might see then how this might cause us to go back and see what it was that we had done to upset the mythical balance. Unfortunately, we have a tendency to fit things we find to what it is we're looking for. So we find things we did that correlate to the time that the balance was disrupted, and we say "aha!" We've been burning fossil fuels which put off C02. CO2 levels have been rising in the atmosphere. Ipso fatco!

Wait, somebody says. That doesn't explain the increases in CO2 and temperature before 1905. Hmmm. Hmmm... it must have been something ELSE we did. We cut down TREES! Trees consume CO2 -- there were fewer of them to do it.... Ipso fatco!

I watched a show last night on National Geographic called "Seconds From Disaster". This one concerned a British Midland Flight 092 in 1989 in which somewhere during the flight the plane started shaking. Smoke started coming into the cabin. The pilots knew an engine was acting up. They believed that the air conditioning ran through the right engine -- therefore, if smoke were coming in it had to be the right engine that was acting up. What they didn't know was that on this newer 737, the air conditioning actually ran through both. They shut the right engine down. The vibrations stopped. The smoke stopped. They must have been right! Ipso facto!

Unfortunately, they weren't. It turns out that to shut one engine down, they had to disengage what would be the equivalent of the plane's cruise control -- a system that automatically delivered more fuel to an engine to keep the plane going the same speed if it sensed the plane was slowing down. Only it really wasn't sensing if the plane was slowing down, it was sensing the fan speed and assuming that if the fan speed slowed, the plane must be slowing -- so it was delivering more fuel to the left engine than it could handle in its crippled state.

When they disengaged the cruise (autothrottle), this stopped happening as the fuel rate was reduced to the manual throttle setting at that position.

See Things I Know #4. Correlation does not mean causation.

Of course, on final approach they manually increased the throttle, the vibration started again, the engine failed, and they were going too slow to start the other one up by airspeed alone this time, though they tried. Thanks to heroic efforts by the pilots, the plane didn't crash in a nearby neigborhood, but rather on a highway embankment right near the start of the runway. 47 people died. (Happily, 79 lived!)

Tragically, several passengers had noticed that after the pilot said everything was ok, there were still flames and sparks coming from the left engine. Nobody tried to inform the pilots. Why not?

Because they trusted the experts (this was confirmed in interviews with the survivors). They trusted the pilots' expertise dispite clear evidence that something about what the pilot had just told them wasn't right. The experts had spoken. The debate was over.

No comments: