Feb. 27, 2003 : Oberg quoted on bad reporting of email what-iffing

JimO: "I think this is a case of overreacting....It’s sad, because it’s not fair and it’s sad because it’s a distraction from more valid avenues of inquiry.... Sexy phrases got grabbed out under a rushed time frame and caused inaccurate impressions to be published."

MSNBC: Alan Boyle’s ‘Cosmic Log’

http://www.msnbc.com/news/750150.asp?0dm=T29BT

Feb. 27, 2003 / 8:30 p.m. ET

Decoding NASA’s e-mail:

It’s an online truism that you can’t always get the full meaning of an e-mail message just by reading the text. That also goes for the infamous e-mail exchange between NASA engineers in the days before the Columbia disaster: Some of the messages that look like dire warnings, upon closer inspection, actually were meant to refer to scenarios so outlandish that no one should take them seriously.
Unfortunately, the most outlandish "what-if" came horribly true.
NBC space analyst James Oberg points out that some of the engineers who figure most prominently in the discussion were from MMACS — the "Mechanical, Maintenance, Arm and Crew Systems" group on the flight control team at NASA’s Johnson Space Center.
"They are supposed to predict the way that all plausible failures could appear to them on data on their consoles, and then develop in advance the recommendations they would need to make to the flight director for the crew to follow," Oberg said. "They are not responsible for ‘engineering analysis’ of past events and did not take part in any debate about the seriousness of the tile damage."
Thus, their discussion wasn’t focused on whether the shuttle was in trouble or not — for good or ill, that was for others to assess. Their task was to figure out what to look for as an indication that the shuttle was in trouble, and what flight controllers and the crew should do about it. In this case, they were presented with a specific scenario: What if any damage to the shuttle’s tiles was more serious than believed? What if a breach let so much heat into the left wing’s sealed wheel well that something went wrong with the landing gear or the tires?
In this light, some of the widely reported comments take on a new cast:
"Why are we talking about this on the day before landing and not the day after launch?" William Anderson, an employee at United Space Alliance, the main contractor for the shuttle program, wrote this not because he was alarmed, but because he was wondering whether the issue was a last-minute annoyance. "If there were evidence on this flight that we were missing tiles, I might be worried," he wrote.
Anderson compared the situation to the "Burning Rocks Syndrome" — a NASA tale about the scientist who voiced concern, just before the Apollo 11 landing, that the lunar module could set moon rocks on fire. Oberg calls it "the classic last-minute cockamamie safety hazard warning that everybody learns to properly ignore."
"That’s certainly an LOCV case." The reference to LOCV — or loss of crew and vehicle — came from R.K. "Kevin" McCluney, a mechanical engineer at Johnson, in reference to one of the possible ways to deal with a landing-gear failure. He was saying that it wasn’t a good idea to try to [JimO: belly] land the shuttle if you knew that the landing gear was "toast." He wasn’t saying that a failure in the tile system would necessarily lead to a catastrophe.
"Ultimately our (MMACS) recommendation in that case is going to be to set up for a bailout (assuming the wing doesn’t burn off before we can get the crew out)." The parenthetical remark from Jeffrey Kling, a flight controller with the MMACS group, does come eerily close to what actually happened. But it was made in the context of the choice between bailing out and trying to bring the shuttle down with a burned-up landing gear.
Kling correctly predicted that if the landing gear went out due to excessive heat, it wouldn’t be merely a case of a flat tire or two. Instead, the whole system would be rendered unusable, and you’d see a wholesale loss of temperature and pressure sensors. Under those circumstances, getting set for a bailout would be the wisest advice. But he went on to note that there might not be enough time to deal with a fiery emergency — which, unfortunately, turned out to be the case.
Amid all these discussions, it was up to a different team at Johnson, called the Mission Evaluation Room, or MER, to assess the actual threat to the shuttle. On the basis of computer simulations and past experience, MER had concluded that Columbia would be OK. That’s why David Lechner, head of the MMACS group, wrote to other engineers that "we hope that the debris impact analyses are correct" and that the e-mail discussion would be moot: because the alternatives for dealing with a worst-case tile scenario were so unpalatable.
Oberg said the engineers were relying on computer simulations and past experiences with tile damage rather than real data.
"The grievous error that was made was to fail to have a solid plan to assess actual launch damage to the tiles," he said. "They should have asked for on-orbit inspection, either from remote cameras, or via a spacewalk."

http://www.galvestondailynews.com/story.lasso?wcd=8250

NASA chief: E-mail actions were correct

By Ted Streuli
The Daily News, Galveston // February 28, 2003

CLEAR LAKE — Internal e-mails released Wednesday had NASA Administrator Sean O’Keefe on the hot seat when he appeared before the House Science Committee in Washington on Thursday.

O’Keefe and other insiders said e-mail discussions among NASA engineers just before the Feb. 1 space shuttle disaster did not suggest that officials knew Columbia had serious problems before it broke apart over Texas. All seven crew members were killed.

O’Keefe told U.S. Rep. Anthony Weiner, D-N.Y., that senior engineers made the right decision by not passing their speculations about worst-case scenarios to senior managers. "It looks like that dialogue went on at exactly the right level," O’Keefe said.

Weiner, whose own temperature was clearly on the rise, said O’Keefe should have been told if engineers harbored safety concerns.

"Why was it that even if there was a hint of a footnote of a memo on a scrap of an envelope that was within this investigation’s scope, that it only made its way to you yesterday, at the same time it made its way to everyone else on the AP wire?" Weiner asked. "This is stunning to me that this is the process being followed. That’s crazy. You must have gotten these memos and hit the roof. Is that a fair characterization?"

Weiner later apologized for his demeanor. NASA flight controller Jeffery Kling told reporters Wednesday night the e-mail conversations were no more than "what-if" scenarios, but said the disaster was unnerving. "When the events started unfolding, there was a little bit of disbelief right at first," Kling said.

Among the possibilities cited by the 43-year-old flight controller in an e-mail dated Jan. 31: a serious breach in the left landing gear compartment, possibly caused by exploding tires and the compartment door blowing open, that would raise concern about the integrity of the aluminum wing.

In his e-mail, Kling went on to describe how temperatures in the wheel well would rise if superheated gases sneaked into the area. He also discussed options for having the crew bail out of the orbiter.

"We did not relay that information to (Columbia commander Rick Husband) because we did not expect it to happen," Kling said. "That thermal analysis said that there would be no burn-through and without burn-through, you won’t damage tires, you won’t even go through these things. This was just a mental exercise that we went through to what-if the whole thing."

Although Weiner and others said Thursday that the e-mails showed NASA officials knew or should have known Columbia was in trouble, at least one expert said that interpretation was wrong.

James Oberg, who worked in Mission Control for much of his 22-year NASA career, said media accounts were unfair. "I think this is a case of overreacting," he said. "It’s sad, because it’s not fair and it’s sad because it’s a distraction from more valid avenues of inquiry."

Oberg said that topics discussed in the e-mail were different than the elements that caused Columbia’s demise. "The scenario as worried about did not happen," Oberg said. "Something else they hadn’t worried about proved worse."

Oberg said the discussion was about possible landing gear and tire problems. "Sexy phrases got grabbed out under a rushed time frame and caused inaccurate impressions to be published," he said.

Flight controllers would have exercised other options if they believed the crew was in jeopardy, Oberg said. "I don’t think they were anywhere near despairing," he said. "They were nowhere near the end of their bag of tricks."

The crew could have attempted an external repair [JimO: inspection, I really said] while in orbit if the situation was desperate, he said. "Yes, it would be scary," he said. "But compared to what? Your level of acceptable risk goes up with your level of worry."

==========

This message was also posted on the usenet sci.space.shuttle discussion group

(access via Yahoo.com)

From: "James Oberg" <jamesoberg@houston.rr.com>

To: <Columbia-L@satobs.org>

Subject: Misinterpretations of the 'NASA emails'

Date: Thursday, February 27, 2003 11:41 AM

Some observations:

The AP story claimed that: "The dozens of pages of e-mails described a many-sided internal debate about the seriousness of potential damage to Columbia from a liftoff collision with foam debris — a debate that was far broader than previously acknowledged. They even considered instructing the crew to bail out."

This is a misinterpretation: First, the MMACS team NEVER debated whether the damage was serious or not. Second, they only considered various ways that possible damage to the landing gear might manifest itself to Mission Control, and various responses that their team would have to make to each possible manifestation. One option was indeed bail-out. But this was NOT in response to "the damage", it was hypothetically in response to the actual detection of symptoms of the possible damage.

AP wrote: "Engineers . . . fretted about the shuttle’s safety during its final three days in orbit, with . . .one another questioning why such dire issues had been raised so late. "Why are we talking about this on the day before landing and not the day after launch?" wrote William C. Anderson, ... less than 24 hours before the shuttle broke apart."

This is a total misreading of Anderson’s memo, which actually dismisses the concern (he calls it another case of the "Burning Rocks Syndrome", the classic last-minute cockamamie safety hazard warning that everybody learns to properly ignore – in that specific case, a warning that moon rocks brought aboard Apollo-11 might explode on contact with oxygen [or that LM exhaust could make lunar surface catch fire]). "If there were evidence on this flight that we were missing tiles," he writes, "I might be worried", as he pooh-poohs the concern.

AP wrote: "After intense debate the engineers ... decided against taking the matter to top NASA managers...." Not so. They never decided TO take it ‘over the heads’ of the existing management team – senior engineers of the MMT in Houston – who had considered and accepted the Boeing study. Those senior-level management teams were precisely the level at which such decisions were supposed to occur.

The news accounts I’ve seen so far don’t seem to pay enough attention to what happened on Friday, January 31, the day before landing. Carlisle Campbell and Bob Daugherty visited the MMACCS office and discussed their concerns – which were limited to the ‘worst case’ that the shuttle might have to land with two flat tires – with Bob Doremus and Dave Paternostro, two senior flight controllers. Since by then, at Campbell’s request, engineers at NASA-Ames in California had conducted several simulations and determined that such a landing was controllable, all present were satisfied. According to notes from Doremus, "All four agreed at the end of the discussion that we were doing a ‘what-if’ discussion and that we all expected a safe entry on Saturday." Of course, that did not occur, and something terrible did happen – but to repeat, it was NOT what Campbell and Daugherty had foreseen as THEIR worst case.

AP wrote: "Another e-mail, from R.K. ‘Kevin’ McCluney, a shuttle mechanical engineer at the Johnson center, described the risks that could lead to "LOCV" — NASA shorthand for the loss of the crew and vehicle."

This is an extremely serious misrepresentation of the actual memo. What McCluney actually wrote was to speculate on the types of readings to be expected from a breach in the landing gear door, and then he asked "what does the alert MMACS operator do in the event of such a signature?" He listed four choices: assume instrumentation error and do nothing; conclude the gear is inoperable and call for an early deploy for visual inspection in time for the crew to bail out if the deploy fails; decide that landing is impossible and bail out; or decide that the gear won’t work and attempt to land ‘gear up’. He immediately wrote that that last option is bad – "That’s almost certainly a LOCV case."

He obviously was not talking about tile damage being "a LOCV case" (as has been widely reported), he was saying that if you try a gear-up landing in response to telemetry suggestive of gear damage, THAT is "a LOCV case".

AP wrote: "But McCluney ultimately recommended to do nothing unless there was a ‘wholesale loss of data’ from sensors in the left wing, in which case controllers would need to decide between a risky landing and bailout attempt. ‘Beats me what the breakpoint would be between the two decisions,’ McCluney wrote."

Actually, McCluny recommended no such thing. Instead, his email says that if one or two measurements showed "odd signatures" he would assume instrumentation errors (a common occurrence on actual missions), but if there was a "wholesale loss of data" he would choose the second option, an early gear deploy so that it could be visually inspected in time to call for crew bailout if both wheels didn’t come down (this would be at an altitude of several thousand feet, flying slightly below Mach 1). Those were the two decisions between which he didn’t know what the breakpoint would be. He NEVER recommended "a risky landing" – the AP reporter simply misunderstood the entire meaning of the email.

It is also critically important to realize that the worst-case scenario as postulated by some individuals did not, in fact, occur. They suggested that damage to the wheel well door might lead to overheating of the landing gear and subsequent bursting of the tires, with the consequent need to land on flat tires, or if that was not feasible, to bail out. On the actual Columbia flight, there clearly was a breach in the wing, but it was not directly into the wheel well. As best as can be determined from telemetry that was received, the tires did not explode, or even begin to overheat (their temperatures and pressures were normal until the signal wires burned through). Something else very bad did happen nearby, but it was not the sequence that these emails speculated about.