Narrative:

During our approach to land on runway 6 at taipei, I misread the approach plate and departed an altitude early. There were no conflicts with other traffic and ground contact was never an issue. The rest of the approach and landing were uneventful. Backgnd: crew duty day began at approximately XA30Z after about 24 hours of layover. We flew to ZZZ with about 2 hours on the ground in ZZZ and then continued to taipei. It was the second leg of the day for all 3 crew members and the second leg of the day flown by the first officer. The altitude deviation I'll describe occurred about 14 1/2 hours into the crew duty day. We left ZZZ1 while the afternoon sun was still up, flew all through the night and approached taipei with the morning sun. WX was VMC, so we had planned a visual approach which was backed up by the ILS. Avionics were a factor in this case for their absence. I took the very rare opportunity that presented itself -- daylight, VMC, light traffic -- to practice an approach without using any automation at all. The descent, approach, and landing were hand-flown without using autoplt, autothrottles, or flight director guidance. A typical approach into taipei under VMC involves vectors to intercept localizer final with subsequent clearance for the visual. On this VMC day, however, we were cleared the full sober 2 arrival into taipei and then subsequently 'cleared the approach' from sober. The plan view for the approach is somewhat difficult to read due to the quantity of information that is presented in a box (that is not to scale) in the lower left corner of the page. Among other things, sober (the IAF) is displayed in this box along with its 4000 ft restr. Significantly, sober is not displayed on the profile view. In brief, the procedure calls for leaving the IAF (sober) on a heading of 090 to intercept the localizer course. The next fix is called hukou, which is the first fix found on the profile view, and is located at 14 DME on the ILS. Hukou has a crossing altitude of 4000 ft. The next restr occurs at 11 DME and that altitude restr is 3000 ft. Flown correctly, the procedure requires you to maintain 4000 ft from sober to hukou, then you can descend to 3000 ft. I misread the approach plate while dividing my attention between flying the approach and clearing outside the cockpit for other traffic. We flew past sober at 4000 ft and then gradually descended to 3000 ft. Passing sober, I shifted my inside-the-cockpit concentration to the profile view and thought that I had met all restrs prior to the 3000 ft restr at 11 DME. That is, hukou is presented on the profile view at 4000 ft, but I was thinking that this was the sober restr that I had already met. We were in between hukou and 11 DME, stable at 3000 ft, when the approach controller asked us why we weren't at 4000 ft. We were clearly legal to be at 3000 ft at this point and were puzzled by his question. He thought he had cleared us to maintain 4000 ft until established on the localizer. He had not and the captain told him the clearance he had given us ('cleared the approach'). This exchange went back and forth a couple of times. The rest of the approach and landing were uneventful. Summary/analysis of contributing factors: as always, many factors were contributing -- fatigue, changes in habit patterns (manual approach, approach clearance atypical of the airport, etc). The most significant of these, however, remains fatigue. The likelihood for making these errors increases exponentially as a crew member exceeds 8 or 10 hours of duty when operating on a normal circadian rhythm. During non circadian flying, the likelihood is certainly higher even with shorter duty days. With the type of international flying we do, where you literally fly all around the world, we must expend a great deal of concentration just trying to understand what the controllers are saying. With that in mind, many of us don't even attempt to develop proficiency hand-flying prior to short final because we all need to be listening to clrncs and trying hard just to understand the controllers, searching for unexpected fixes we've been cleared to, and clearing outside for other airplanes (many of which speak to the local controllers in the local language, so knowing their expected location in the pattern is impossible unless you see them). With the level of automation available to us, it's a given that most of us use the automation when we know we're tired or when WX conditions or traffic conditions make that the obviously prudent choice. So all that being said, proficiency at flying the raw data takes a back seat. In the future, I'll likely sacrifice the opportunity for proficiency in flying raw data in favor of automation whenever my duty exceeds about 8 hours -- regardless of flight conditions or my assessment of whether or not I'm fatigued. I believe that if I had been flying this approach using the autoplt, I would not have misread the approach plate. While I accept responsibility for this altitude deviation, my additional recommendations would include shorter duty days, better on-board rest facilities, scheduled layovers that consider circadian rhythms, and better chart depictions. Supplemental information from acn 528706: the controller later stated that we were to maintain 4000 ft till established on the localizer, then descend in accordance with approach procedures. Our clearance, as I recall, was to 'cross sober intersection at or above 4000 ft, cleared for approach.' further review of the approach plate showed that there was an MEA of 4000 ft between sober and a point along the localizer for runway 6. This information is contained in a box in the lower l-hand corner of the plan view. I did not catch this during our briefing.

Related reports:

Google
 

Original NASA ASRS Text

Title: A B767 FLC DSNDS PRIOR TO THE LOC DURING A VMC APCH TO RWY 6, VIOLATING THE MEA PUBLISHED FOR THE APCH TO RCTP, FO.

Narrative: DURING OUR APCH TO LAND ON RWY 6 AT TAIPEI, I MISREAD THE APCH PLATE AND DEPARTED AN ALT EARLY. THERE WERE NO CONFLICTS WITH OTHER TFC AND GND CONTACT WAS NEVER AN ISSUE. THE REST OF THE APCH AND LNDG WERE UNEVENTFUL. BACKGND: CREW DUTY DAY BEGAN AT APPROX XA30Z AFTER ABOUT 24 HRS OF LAYOVER. WE FLEW TO ZZZ WITH ABOUT 2 HRS ON THE GND IN ZZZ AND THEN CONTINUED TO TAIPEI. IT WAS THE SECOND LEG OF THE DAY FOR ALL 3 CREW MEMBERS AND THE SECOND LEG OF THE DAY FLOWN BY THE FO. THE ALTDEV I'LL DESCRIBE OCCURRED ABOUT 14 1/2 HRS INTO THE CREW DUTY DAY. WE LEFT ZZZ1 WHILE THE AFTERNOON SUN WAS STILL UP, FLEW ALL THROUGH THE NIGHT AND APCHED TAIPEI WITH THE MORNING SUN. WX WAS VMC, SO WE HAD PLANNED A VISUAL APCH WHICH WAS BACKED UP BY THE ILS. AVIONICS WERE A FACTOR IN THIS CASE FOR THEIR ABSENCE. I TOOK THE VERY RARE OPPORTUNITY THAT PRESENTED ITSELF -- DAYLIGHT, VMC, LIGHT TFC -- TO PRACTICE AN APCH WITHOUT USING ANY AUTOMATION AT ALL. THE DSCNT, APCH, AND LNDG WERE HAND-FLOWN WITHOUT USING AUTOPLT, AUTOTHROTTLES, OR FLT DIRECTOR GUIDANCE. A TYPICAL APCH INTO TAIPEI UNDER VMC INVOLVES VECTORS TO INTERCEPT LOC FINAL WITH SUBSEQUENT CLRNC FOR THE VISUAL. ON THIS VMC DAY, HOWEVER, WE WERE CLRED THE FULL SOBER 2 ARR INTO TAIPEI AND THEN SUBSEQUENTLY 'CLRED THE APCH' FROM SOBER. THE PLAN VIEW FOR THE APCH IS SOMEWHAT DIFFICULT TO READ DUE TO THE QUANTITY OF INFO THAT IS PRESENTED IN A BOX (THAT IS NOT TO SCALE) IN THE LOWER L CORNER OF THE PAGE. AMONG OTHER THINGS, SOBER (THE IAF) IS DISPLAYED IN THIS BOX ALONG WITH ITS 4000 FT RESTR. SIGNIFICANTLY, SOBER IS NOT DISPLAYED ON THE PROFILE VIEW. IN BRIEF, THE PROC CALLS FOR LEAVING THE IAF (SOBER) ON A HDG OF 090 TO INTERCEPT THE LOC COURSE. THE NEXT FIX IS CALLED HUKOU, WHICH IS THE FIRST FIX FOUND ON THE PROFILE VIEW, AND IS LOCATED AT 14 DME ON THE ILS. HUKOU HAS A XING ALT OF 4000 FT. THE NEXT RESTR OCCURS AT 11 DME AND THAT ALT RESTR IS 3000 FT. FLOWN CORRECTLY, THE PROC REQUIRES YOU TO MAINTAIN 4000 FT FROM SOBER TO HUKOU, THEN YOU CAN DSND TO 3000 FT. I MISREAD THE APCH PLATE WHILE DIVIDING MY ATTN BTWN FLYING THE APCH AND CLRING OUTSIDE THE COCKPIT FOR OTHER TFC. WE FLEW PAST SOBER AT 4000 FT AND THEN GRADUALLY DSNDED TO 3000 FT. PASSING SOBER, I SHIFTED MY INSIDE-THE-COCKPIT CONCENTRATION TO THE PROFILE VIEW AND THOUGHT THAT I HAD MET ALL RESTRS PRIOR TO THE 3000 FT RESTR AT 11 DME. THAT IS, HUKOU IS PRESENTED ON THE PROFILE VIEW AT 4000 FT, BUT I WAS THINKING THAT THIS WAS THE SOBER RESTR THAT I HAD ALREADY MET. WE WERE IN BTWN HUKOU AND 11 DME, STABLE AT 3000 FT, WHEN THE APCH CTLR ASKED US WHY WE WEREN'T AT 4000 FT. WE WERE CLRLY LEGAL TO BE AT 3000 FT AT THIS POINT AND WERE PUZZLED BY HIS QUESTION. HE THOUGHT HE HAD CLRED US TO MAINTAIN 4000 FT UNTIL ESTABLISHED ON THE LOC. HE HAD NOT AND THE CAPT TOLD HIM THE CLRNC HE HAD GIVEN US ('CLRED THE APCH'). THIS EXCHANGE WENT BACK AND FORTH A COUPLE OF TIMES. THE REST OF THE APCH AND LNDG WERE UNEVENTFUL. SUMMARY/ANALYSIS OF CONTRIBUTING FACTORS: AS ALWAYS, MANY FACTORS WERE CONTRIBUTING -- FATIGUE, CHANGES IN HABIT PATTERNS (MANUAL APCH, APCH CLRNC ATYPICAL OF THE ARPT, ETC). THE MOST SIGNIFICANT OF THESE, HOWEVER, REMAINS FATIGUE. THE LIKELIHOOD FOR MAKING THESE ERRORS INCREASES EXPONENTIALLY AS A CREW MEMBER EXCEEDS 8 OR 10 HRS OF DUTY WHEN OPERATING ON A NORMAL CIRCADIAN RHYTHM. DURING NON CIRCADIAN FLYING, THE LIKELIHOOD IS CERTAINLY HIGHER EVEN WITH SHORTER DUTY DAYS. WITH THE TYPE OF INTL FLYING WE DO, WHERE YOU LITERALLY FLY ALL AROUND THE WORLD, WE MUST EXPEND A GREAT DEAL OF CONCENTRATION JUST TRYING TO UNDERSTAND WHAT THE CTLRS ARE SAYING. WITH THAT IN MIND, MANY OF US DON'T EVEN ATTEMPT TO DEVELOP PROFICIENCY HAND-FLYING PRIOR TO SHORT FINAL BECAUSE WE ALL NEED TO BE LISTENING TO CLRNCS AND TRYING HARD JUST TO UNDERSTAND THE CTLRS, SEARCHING FOR UNEXPECTED FIXES WE'VE BEEN CLRED TO, AND CLRING OUTSIDE FOR OTHER AIRPLANES (MANY OF WHICH SPEAK TO THE LCL CTLRS IN THE LCL LANGUAGE, SO KNOWING THEIR EXPECTED LOCATION IN THE PATTERN IS IMPOSSIBLE UNLESS YOU SEE THEM). WITH THE LEVEL OF AUTOMATION AVAILABLE TO US, IT'S A GIVEN THAT MOST OF US USE THE AUTOMATION WHEN WE KNOW WE'RE TIRED OR WHEN WX CONDITIONS OR TFC CONDITIONS MAKE THAT THE OBVIOUSLY PRUDENT CHOICE. SO ALL THAT BEING SAID, PROFICIENCY AT FLYING THE RAW DATA TAKES A BACK SEAT. IN THE FUTURE, I'LL LIKELY SACRIFICE THE OPPORTUNITY FOR PROFICIENCY IN FLYING RAW DATA IN FAVOR OF AUTOMATION WHENEVER MY DUTY EXCEEDS ABOUT 8 HRS -- REGARDLESS OF FLT CONDITIONS OR MY ASSESSMENT OF WHETHER OR NOT I'M FATIGUED. I BELIEVE THAT IF I HAD BEEN FLYING THIS APCH USING THE AUTOPLT, I WOULD NOT HAVE MISREAD THE APCH PLATE. WHILE I ACCEPT RESPONSIBILITY FOR THIS ALTDEV, MY ADDITIONAL RECOMMENDATIONS WOULD INCLUDE SHORTER DUTY DAYS, BETTER ON-BOARD REST FACILITIES, SCHEDULED LAYOVERS THAT CONSIDER CIRCADIAN RHYTHMS, AND BETTER CHART DEPICTIONS. SUPPLEMENTAL INFO FROM ACN 528706: THE CTLR LATER STATED THAT WE WERE TO MAINTAIN 4000 FT TILL ESTABLISHED ON THE LOC, THEN DSND IN ACCORDANCE WITH APCH PROCS. OUR CLRNC, AS I RECALL, WAS TO 'CROSS SOBER INTXN AT OR ABOVE 4000 FT, CLRED FOR APCH.' FURTHER REVIEW OF THE APCH PLATE SHOWED THAT THERE WAS AN MEA OF 4000 FT BTWN SOBER AND A POINT ALONG THE LOC FOR RWY 6. THIS INFO IS CONTAINED IN A BOX IN THE LOWER L-HAND CORNER OF THE PLAN VIEW. I DID NOT CATCH THIS DURING OUR BRIEFING.

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.