B737 QRH. What happened to Electrical Smoke or Fire? A potted history.

[Centaurus]

On 3 November 1973 a Panam B707 freighter crashed near Boston USA following heavy smoke in the cockpit which the crew interpreted at the time as due to an electrical fire.
See : ASN Aircraft accident Boeing 707-321C N458PA Boston-Logan International Airport, MA (BOS)

In those days the procedures for electrical fire or smoke required the crew to turn off various electrical systems in order to isolate the site of the fire or smoke. These procedures were essentially a trouble-shooting checklist and were often lengthy as a certain amount of time was called for between each selection to see if smoke eventually ceased or not.

In the case of the Panam 707, the flight engineer commenced trouble shooting action at cruise altitude. While he was doing this the captain maintained cruise. After some time and as the engineer continued to read through the checklist item by item trouble shooting, the smoke got worse until it was so dense the captain was forced to make an emergency descent from cruise altitude to get the aircraft on the ground as quickly as possible. The smoke was so thick that the first officer was unable to read radio frequencies. This all took time which in the event was fatal.

Control of the aircraft was lost at very low altitude, when the engineer continuing with his QRH trouble-shooting checklist switched off the battery switch causing the yaw dampers to be inoperative. Smoke was seen pouring from the copilot's window at the same time control was lost. The combination of spoilers, landing flap and no yaw damper made the aircraft uncontrollable. Investigation revealed it was not an electrical problem at all. Instead it was found that bottles of acid in the front cargo hold had broken releasing their contents on to wood chips surrounding the bottles to protect them from damage. The airflow carried the smoke from burning wood chips through cooling tubes to the circuit breaker panel on the flight deck. This led the crew to diagnose an electrical fire somewhere.

Boeing later changed the QRH for electrical fire or smoke and deleted reference to trouble shooting since smoke may not always have electrical origin. The checklist was reduced to bare essentials with the advice to land at the nearest suitable airport if the problem was not solved. The B737-300 QRH dated January 1999 for example had this checklist under the heading "Electrical Smoke of Fire"

Oxy Masks and Regulators (if required) ON, 100%
Smoke Goggles (if required) ON
If smoke source can be determined: Electrical Power....REMOVE.
If smoke source cannot be determined:
Bus Transfer...OFF
Galley Power....OFF

Land at nearest suitable airport.
............................................................ ........

Despite this revision to the QRH based upon the experience of the B707 accident and it's trouble shooting checklist, at least one European B737 operator elected to keep the trouble-shooting checklist for electrical fire or smoke. The reason given was the airline had over-water routes and the nearest suitable airport could be some distance away. The operator missed the whole point that smoke could quickly incapacitate the crew and that in the worst case an immediate ditching may be required before the crew succumbed to smoke fumes.

During the aftermath of the Swiss Air MD11 accident near Halifax where smoke and fire in the cockpit (electrical origin) caused the crew to become incapacitated, there was speculation that needless delays occurred as the crew went through a lengthy trouble shooting checklist and then circled to jettison fuel.

Since those days, more research by Boeing (737) resulted in further QRH changes and the words "electrical fire" were deleted. Possibly because it may be difficult to differentiate between electrical smoke and smoke from other causes, the Boeing 737 Non Normal checklist Section 6 Electrical, says nothing about Electrical smoke or fire but instead refers the reader to Section 8.8 under the heading of "Smoke, Fire or Fumes"

The reader is now confronted with no less than four pages of checklist items designed to trouble shoot the source of the problem and eventually directs the reader to a further page in Section 8 dealing with Smoke or Fumes Removal.

Yet only on page 3 of the "Smoke, Fire or Fumes" checklist do we see the first reference to consider an immediate landing if the smoke, fire or fumes situation becomes uncontrollable.

My point is this: After the Panam 707 accident referred to earlier, Boeing saw fit in the QRH to remove all reference to trouble shooting to isolate electrical system items in event of an electrical smoke or fire. In other words - forget trouble shooting. Turn off essentials and if that doesn't work, get the aircraft on the ground asap. Reason? Danger of incapacitation due smoke.

Later on, Boeing had another think and removed all reference in QRH Section 6 Electrical, to electrical smoke or fire. Instead transferred the scenario to Section 8 of the QRH under the heading Fire Protection - under the catch-all item "Smoke, Fire or Fumes". To adequately encompass that lot requires several QRH pages of trouble shooting.

Of course, there is always the possibility a crew can incorrectly diagnose the cause or origin of smoke. But it seemed to me when Boeing revised the old "Electrical Smoke or Fire" QRH item, and not only reduced it to bare essentials but left it under its rightful place in Section 8 Electrical, it was a good decision. The theory being if the revised shortened checklist did not fix the problem immediately, it was essential to land immediately. Incapacitation was recognised as the real danger.

Now we have gone the full circle and the lessons of the Panam 707 and Swiss Air MD11 accidents lost in time. Trouble shooting is back in vogue. Why the change? Is incapacitation now considered unlikely?

Comments welcomed.

[Sciolistes]

Centaurus,
A great insight into the design of the QRH, thank you for posting.

You're core point is that the during an onboard smoke/fire event, the biggest risk is time to land, even and off airport landing/ditching may be required whilst control is still possible.

In the QRH, Checklist Instructions section, it lists the minimum items for which landing at the nearest suitable airport must be performed. There is a reference to "fire or smoke continues", but important item is "any other situation determined by the flight crew to have a significant adverse effect on safety if the flight is continued"

Later it says that if the smoke, fire or fumes becomes uncontrollable then even an off-airport landing or ditching maybe required.

The way I see it, all the airmanship advice contained in the text in the QRH and FCTM are elements that should be considered before during and after actioning specific non-normal checklists. The non-normal checklist should not be robotically called without considering the possible ramifications if other measure are not considered first.

Therefore I think Boeing intend us to depart from the script as and when we deem it necessary, it says so right there, in dark grey and light grey (as distinct from black and white). Last time I did this procedure in the sim as PF, we immediately diverted without waiting for the checklist. Naturally, the checklist must be started when the safest course of action has been determined. Boeing give us the tools to think like this and that is what I hope we would do for real.

It is my personal opinion that any fire, smoke or fumes situation is probably unquantifiable to any reasonable degree of certainty and you won't know till it is too late anyway, so I classify the situation as uncontrollable until we know better. Like I said, that is my opinion and it can only be my opinion because if you prioritise initially in favour of attempting to complete the checklist, you'd probably be justifiable in most cases.

Perhaps seems to me that Boeing have performed a neat ACM (Ar5e Covering Manoeuvre) in providing a comprehensive checklist with additional advise to for more extreme situations. Perhaps it very much depends on the quality and nature of the training as to how individuals handle such a situation, what kind of weight they give the advice in the FCTM and QRH relative to specific non-normal checklists and ultimately whether they repeat past mistakes or land safely and at worst unnecessarily.

In interesting link:
Skybrary: In-flight fire guidance to to flight crew

An interesting table from Skybrary on in-flight fires:

[Centaurus]

The way I see it, all the airmanship advice contained in the text in the QRH and FCTM are elements that should be considered before during and after actioning specific non-normal checklists. The non-normal checklist should not be robotically called without considering the possible ramifications if other measure are not considered first.

You are quite right of course. But in my experience and particularly with certain cultures, the checklist becomes God and certainly in a court case, lawyers would likely seize on checklist omissions as proof of incompetency and conveniently disregard the small print or amplification in manuals such as the FCTM.

[Denti]

Depends probably a lot on flight ops and training culture in each airline. In "my" outfit the primary target in case of fire (and smoke usually has a fire as source) is to get it on the ground (or water) within 10 minutes. That is pretty fast especially if at cruising altitude. The checklist can be done if we have some time for it, but it is not the primary thing in case of fire, smoke, fumes on board. So yes, the checklist is now a really long thing, one of the most cumbersome procedures in the whole QRH, but the primary thing is to save as many lives as possible and statistically that is done by landing ASAP.

[RAT 5]

Becoming a little specific in this matter, I have asked a question to my training dept. and failed to have an answer, even after an inflight fire training scenario. We applied the new QRH fire/smoke checklist. It says isolate the sourse if possible, (electrical). Switch off the Bus Tfr switch etc. Now, in previous Boeings, after this was done, each Gen was switched off in turn and you waited to see if smoke reduced. In the NG this doesn't happen; you just open the Bus Tfr. How then can you determine which Bus is producing the smoke and isolate it? Help & understanding will be appreciated.

[EEngr]

NG Procedures

I can't read the mind of the people who write the procedures, but here's a possible line of thought.

In the NG this doesn't happen; you just open the Bus Tfr. How then can you determine which Bus is producing the smoke and isolate it? Help & understanding will be appreciated.

Back in the 'old days' it was assumed that manual intervention would be necessary to clear a large set of fault types. Particularly if you have a flight engineer available to go through a longer list of procedures while the capt/f.o. fly the plane. But now a goal of systems design is to reduce crew workload by the use of 'smarter' systems*. If the fault progresses to a state of imminent danger, automated protection should be able to clear it. The 'open Bus transfer' step is a quick procedure to ensure that a remaining good generator will not inadvertently close into the faulted bad side should the automated systems not be smart enough.

*Why the designers of the 777/787 power systems saw fit to eliminate a type of automatic bus/breaker protection which has proven itself over time and what they think they've replaced it with, I can't say. I wasn't there when that paper was blank.

[Hirem J Trashcan]

Parallel situation:
If I were driving my car and found smoke suddenly filling the cabin, I would find a safe place to stop ASAP and investigate.

Surely in an aircraft, at the first sign of smoke, an emergency should be declared and plans made to locate and land at the nearest safe haven. Investigations into the possible source of the smoke could be made whilst this is going on. I find it somewhat unusual that the first steps would appear to be "carry on as if all is normal whilst the problem is investigated" and only when nothing obvious can be found to eliminate the smoke would an attempt be made to find soewhere to land