Narrative:

The crew was on the last (5TH) leg of a 12 hour 47 min duty day. The aircraft was approximately within 10-20 NM of abb VOR. Passing 13000 ft MSL, on descent into cvg airport, smoke began to enter the cockpit. Consulting the checklist, the smoke cleared when the packs and bleeds were shut off. Maintenance and dispatch were consulted on our situation. An emergency was declared with cvg approach. Following the checklist, we landed as soon as possible at cvg as it was the closest airport at the time the decision was made. The landing was uneventful. Later the aircraft was found to have a faulty bleed air valve that had let oil into the pack system. The moment an oil smell and smoke began to appear, the captain (seeing that the packs were in normal instead of low where they were known to give off oil smells at low power settings), shut the packs and bleeds off. This immediately stopped the smoke. With no smoke or fumes present, the crew did not don oxygen masks. The captain had the first officer run the smoke emergency checklist during which time they contacted the flight attendant and maintenance. After getting to the section 'smoke suspected from the air conditioning system' and the note that this could be the warning of an internal engine fire, the crew carefully ensured that there were no other confirming indications. The next statement was land as soon as possible, the captain already talking to dispatch considered both cvg and sdf. He decided with the time needed to prepare the aircraft, run checklists, and descend that cvg was the best airport at that point in time. At this point the crew informed dispatch of the decision and declared an emergency with cvg approach. The emergency checklist continued on to isolate the 'bad' bleed or pack. The checklist was not continued at this point because the captain had already shut both packs and bleeds and no smoke or odor was present. This left the crew hesitant to bring smoke back into the aircraft. Descending to 8000 ft MSL, the aircraft slowly depressurized as it continued to cvg. With the smoke halted, the crew was more concerned about getting on the ground. As the captain, I was in charge of the flight. At the time of this incident, with the smoke stopped and cleared, the aircraft below 10000 ft MSL, I judged that we were out of immediate danger of fire and smoke and it was in our best interest to get to cvg and land as soon as practical. Before and after conferring with maintenance, I believed the problem to be the pack switch position as the EMB120 flight standards manual stated it would produce an oil smell from the vents with a low power setting and packs in normal. The oil passing the seals would give the smell, it seemed natural for a puff of oil smoke to accompany it. I believed that we would recycle the packs and be smoke free. I wanted to confer with maintenance to confirm this. When the first officer read the note that it could be an internal engine fire, I realized that we might have a much worse situation. Once we ensured that we did not have an internal engine fire we finished talking with maintenance. Next we talked to dispatch concentrating on where to land. I wasted valuable time that considering cvg and sdf. Furthermore while we prepared to land, I continued to second-guess myself as to whether sdf airport might have been a better destination (given the problem had started near aab VOR). But I felt that the descent, checklists, and coordinating with the flight attendant, ATC, dispatch and notifying the passenger placed cvg as the better destination. Had I, as the captain, handled this situation differently, checklists would have been complied with more efficiency. The left engine bleed would have been isolated without securing the right bleed or the packs. Oxygen would have been used prior to beginning the checklist. Although this would have allowed more smoke to enter the cockpit and even the passenger cabin, it would have achieved a greater level of safety and let the checklist flow in a more rehearsed manner. I do not believe that we would have finished the checklist quicker, but given our position it may have led to vectors to cvg or sdf sooner. Callback conversation with reporter revealed the following information: the reporter stated the crew had been on duty in excess of 13 hours andwere tired. The reporter said maintenance found the high pressure switch had failed in the low pressure compressor mode. The reporter said there is no indication for what bleed valve is open at any particular time. The reporter stated it is known to pilots who fly the EMB120 that the low pressure compressor when in a low power situation and supplying bleed air cannot prevent lubrication oil from entering the bleed air. The reporter said the bleed air with oil enters the air conditioning system causing the smoke. Supplemental information from acn 436724: at one point in the checklist there is a caution that smoke from the air conditioning system could be the first indication of an internal engine fire. However, the next step in the checklist said 'land as soon as possible.' investigation by maintenance personnel found that a pressure switching/regulating valve for bleed air used in the left engine's lubrication system had failed and allowed engine oil to enter the compressor section of the engine, therefore causing the smoke. Maintenance history showed the same problem had occurred on this airplane 3 days earlier, but maintenance could not duplicate the problem at that time. Third is the format of the cabin fire or smoke emergency checklist. My concern is over the formatting of the checklist. Following the section on the possible internal engine fire, is the statement 'land as soon as possible.' this is not indented like the steps for the possible internal engine fire. This leads a crew to believe that landing is the next step in the checklist and a step that would logically conclude the checklist. Callback conversation with reporter revealed the following information: the reporter stated he discovered the smoke and the captain immediately shut off both air conditioning packs. The reporter said an emergency was declared and ATC was advised of the nature of the emergency. The reporter stated the smoke emergency checklist is somewhat misleading and needs revision.

Related reports:

Google
 

Original NASA ASRS Text

Title: AN EMBRAER 120 AT START OF DSCNT AT 13000 FT DECLARED AN EMER DUE TO SMOKE IN THE COCKPIT CAUSED BY A FAILED HIGH PRESSURE BLEED SWITCH.

Narrative: THE CREW WAS ON THE LAST (5TH) LEG OF A 12 HR 47 MIN DUTY DAY. THE ACFT WAS APPROX WITHIN 10-20 NM OF ABB VOR. PASSING 13000 FT MSL, ON DSCNT INTO CVG ARPT, SMOKE BEGAN TO ENTER THE COCKPIT. CONSULTING THE CHKLIST, THE SMOKE CLRED WHEN THE PACKS AND BLEEDS WERE SHUT OFF. MAINT AND DISPATCH WERE CONSULTED ON OUR SIT. AN EMER WAS DECLARED WITH CVG APCH. FOLLOWING THE CHKLIST, WE LANDED ASAP AT CVG AS IT WAS THE CLOSEST ARPT AT THE TIME THE DECISION WAS MADE. THE LNDG WAS UNEVENTFUL. LATER THE ACFT WAS FOUND TO HAVE A FAULTY BLEED AIR VALVE THAT HAD LET OIL INTO THE PACK SYS. THE MOMENT AN OIL SMELL AND SMOKE BEGAN TO APPEAR, THE CAPT (SEEING THAT THE PACKS WERE IN NORMAL INSTEAD OF LOW WHERE THEY WERE KNOWN TO GIVE OFF OIL SMELLS AT LOW PWR SETTINGS), SHUT THE PACKS AND BLEEDS OFF. THIS IMMEDIATELY STOPPED THE SMOKE. WITH NO SMOKE OR FUMES PRESENT, THE CREW DID NOT DON OXYGEN MASKS. THE CAPT HAD THE FO RUN THE SMOKE EMER CHKLIST DURING WHICH TIME THEY CONTACTED THE FLT ATTENDANT AND MAINT. AFTER GETTING TO THE SECTION 'SMOKE SUSPECTED FROM THE AIR CONDITIONING SYS' AND THE NOTE THAT THIS COULD BE THE WARNING OF AN INTERNAL ENG FIRE, THE CREW CAREFULLY ENSURED THAT THERE WERE NO OTHER CONFIRMING INDICATIONS. THE NEXT STATEMENT WAS LAND ASAP, THE CAPT ALREADY TALKING TO DISPATCH CONSIDERED BOTH CVG AND SDF. HE DECIDED WITH THE TIME NEEDED TO PREPARE THE ACFT, RUN CHKLISTS, AND DSND THAT CVG WAS THE BEST ARPT AT THAT POINT IN TIME. AT THIS POINT THE CREW INFORMED DISPATCH OF THE DECISION AND DECLARED AN EMER WITH CVG APCH. THE EMER CHKLIST CONTINUED ON TO ISOLATE THE 'BAD' BLEED OR PACK. THE CHKLIST WAS NOT CONTINUED AT THIS POINT BECAUSE THE CAPT HAD ALREADY SHUT BOTH PACKS AND BLEEDS AND NO SMOKE OR ODOR WAS PRESENT. THIS LEFT THE CREW HESITANT TO BRING SMOKE BACK INTO THE ACFT. DSNDING TO 8000 FT MSL, THE ACFT SLOWLY DEPRESSURIZED AS IT CONTINUED TO CVG. WITH THE SMOKE HALTED, THE CREW WAS MORE CONCERNED ABOUT GETTING ON THE GND. AS THE CAPT, I WAS IN CHARGE OF THE FLT. AT THE TIME OF THIS INCIDENT, WITH THE SMOKE STOPPED AND CLRED, THE ACFT BELOW 10000 FT MSL, I JUDGED THAT WE WERE OUT OF IMMEDIATE DANGER OF FIRE AND SMOKE AND IT WAS IN OUR BEST INTEREST TO GET TO CVG AND LAND AS SOON AS PRACTICAL. BEFORE AND AFTER CONFERRING WITH MAINT, I BELIEVED THE PROB TO BE THE PACK SWITCH POS AS THE EMB120 FLT STANDARDS MANUAL STATED IT WOULD PRODUCE AN OIL SMELL FROM THE VENTS WITH A LOW PWR SETTING AND PACKS IN NORMAL. THE OIL PASSING THE SEALS WOULD GIVE THE SMELL, IT SEEMED NATURAL FOR A PUFF OF OIL SMOKE TO ACCOMPANY IT. I BELIEVED THAT WE WOULD RECYCLE THE PACKS AND BE SMOKE FREE. I WANTED TO CONFER WITH MAINT TO CONFIRM THIS. WHEN THE FO READ THE NOTE THAT IT COULD BE AN INTERNAL ENG FIRE, I REALIZED THAT WE MIGHT HAVE A MUCH WORSE SIT. ONCE WE ENSURED THAT WE DID NOT HAVE AN INTERNAL ENG FIRE WE FINISHED TALKING WITH MAINT. NEXT WE TALKED TO DISPATCH CONCENTRATING ON WHERE TO LAND. I WASTED VALUABLE TIME THAT CONSIDERING CVG AND SDF. FURTHERMORE WHILE WE PREPARED TO LAND, I CONTINUED TO SECOND-GUESS MYSELF AS TO WHETHER SDF ARPT MIGHT HAVE BEEN A BETTER DEST (GIVEN THE PROB HAD STARTED NEAR AAB VOR). BUT I FELT THAT THE DSCNT, CHKLISTS, AND COORDINATING WITH THE FLT ATTENDANT, ATC, DISPATCH AND NOTIFYING THE PAX PLACED CVG AS THE BETTER DEST. HAD I, AS THE CAPT, HANDLED THIS SIT DIFFERENTLY, CHKLISTS WOULD HAVE BEEN COMPLIED WITH MORE EFFICIENCY. THE L ENG BLEED WOULD HAVE BEEN ISOLATED WITHOUT SECURING THE R BLEED OR THE PACKS. OXYGEN WOULD HAVE BEEN USED PRIOR TO BEGINNING THE CHKLIST. ALTHOUGH THIS WOULD HAVE ALLOWED MORE SMOKE TO ENTER THE COCKPIT AND EVEN THE PAX CABIN, IT WOULD HAVE ACHIEVED A GREATER LEVEL OF SAFETY AND LET THE CHKLIST FLOW IN A MORE REHEARSED MANNER. I DO NOT BELIEVE THAT WE WOULD HAVE FINISHED THE CHKLIST QUICKER, BUT GIVEN OUR POS IT MAY HAVE LED TO VECTORS TO CVG OR SDF SOONER. CALLBACK CONVERSATION WITH RPTR REVEALED THE FOLLOWING INFO: THE RPTR STATED THE CREW HAD BEEN ON DUTY IN EXCESS OF 13 HRS ANDWERE TIRED. THE RPTR SAID MAINT FOUND THE HIGH PRESSURE SWITCH HAD FAILED IN THE LOW PRESSURE COMPRESSOR MODE. THE RPTR SAID THERE IS NO INDICATION FOR WHAT BLEED VALVE IS OPEN AT ANY PARTICULAR TIME. THE RPTR STATED IT IS KNOWN TO PLTS WHO FLY THE EMB120 THAT THE LOW PRESSURE COMPRESSOR WHEN IN A LOW PWR SIT AND SUPPLYING BLEED AIR CANNOT PREVENT LUBRICATION OIL FROM ENTERING THE BLEED AIR. THE RPTR SAID THE BLEED AIR WITH OIL ENTERS THE AIR CONDITIONING SYS CAUSING THE SMOKE. SUPPLEMENTAL INFO FROM ACN 436724: AT ONE POINT IN THE CHKLIST THERE IS A CAUTION THAT SMOKE FROM THE AIR CONDITIONING SYS COULD BE THE FIRST INDICATION OF AN INTERNAL ENG FIRE. HOWEVER, THE NEXT STEP IN THE CHKLIST SAID 'LAND ASAP.' INVESTIGATION BY MAINT PERSONNEL FOUND THAT A PRESSURE SWITCHING/REGULATING VALVE FOR BLEED AIR USED IN THE L ENG'S LUBRICATION SYS HAD FAILED AND ALLOWED ENG OIL TO ENTER THE COMPRESSOR SECTION OF THE ENG, THEREFORE CAUSING THE SMOKE. MAINT HISTORY SHOWED THE SAME PROB HAD OCCURRED ON THIS AIRPLANE 3 DAYS EARLIER, BUT MAINT COULD NOT DUPLICATE THE PROB AT THAT TIME. THIRD IS THE FORMAT OF THE CABIN FIRE OR SMOKE EMER CHKLIST. MY CONCERN IS OVER THE FORMATTING OF THE CHKLIST. FOLLOWING THE SECTION ON THE POSSIBLE INTERNAL ENG FIRE, IS THE STATEMENT 'LAND ASAP.' THIS IS NOT INDENTED LIKE THE STEPS FOR THE POSSIBLE INTERNAL ENG FIRE. THIS LEADS A CREW TO BELIEVE THAT LNDG IS THE NEXT STEP IN THE CHKLIST AND A STEP THAT WOULD LOGICALLY CONCLUDE THE CHKLIST. CALLBACK CONVERSATION WITH RPTR REVEALED THE FOLLOWING INFO: THE RPTR STATED HE DISCOVERED THE SMOKE AND THE CAPT IMMEDIATELY SHUT OFF BOTH AIR CONDITIONING PACKS. THE RPTR SAID AN EMER WAS DECLARED AND ATC WAS ADVISED OF THE NATURE OF THE EMER. THE RPTR STATED THE SMOKE EMER CHECKLIST IS SOMEWHAT MISLEADING AND NEEDS REVISION.

Data retrieved from NASA's ASRS site as of July 2007 and automatically converted to unabbreviated mixed upper/lowercase text. This report is for informational purposes with no guarantee of accuracy. See NASA's ASRS site for official report.