A300 Flight Crew reported descending below FAF altitude prior to the fix due to FMS error and continuing the approach briefly with an indication of 4 red PAPI lights. Flight crew leveled off and intercepted glide slope for descent.
Synopsis
A300 Flight Crew reported descending below FAF altitude prior to the fix due to FMS error and continuing the approach briefly with an indication of 4 red PAPI lights. Flight crew leveled off and intercepted glide slope for descent.
Narrative
Incorrect RNAV VNAV path on visual approach; Runway XXL not recognized until low at ZZZZZ FAF. We had set up for the RNAV (GPS) Y Runway XXL. Prior to arrival we had briefed arrival and approach options including the possibility of ILS XYL; but with multiple departures were assigned XXL. We received slight vectors off of the ZZZZZ1 RNAV arrival downwind track (5-10 degrees) and cleared descent to 6000 ft. but left the arrival navigation track in the FMS until passing the endpoint of ZZZZZ2 >8100 ft. (instead of 'cleaning up' the FMS); to keep a visual reminder of terrain in that sector; to not descend too early. The new hire First Officer (FO) was newly consolidated in the A300; and in the approach briefing; when I asked how comfortable they were with RNAV approaches; FO indicated; as normal for folks new in the A300; that they felt not entirely fluent with them. So I wanted to give plenty of room to turn base and get established about 3 NM prior to ZZZZZ; Soon after ZZZZZ2; just prior to base turn; I asked the FO to clean up the approach. We had left in the published hold at ZZZZZ3 on the approach sequence; in case of necessary holding; and had briefed that it would be necessary to delete the hold prior to ZZZZZ3. Upon cleaning up the approach; prior to being abeam the FAF ZZZZZ; I noticed that the waypoint had sequenced past ZZZZZ3 to ZZZZZ. Expecting to turn base between the two; this appeared that the TO waypoint was correct; per RNAV procedures; and I mentioned that. We were cleared the visual approach; and I mentioned that we would not fly too tight an approach; to have time to stabilize and get set up on RNAV final; which we did. I was hoping to give a good RNAV experience to help the FO feel more comfortable with using RNAV approaches. When I asked the FO to delete the hold on base turn (had not yet); while descending from 5000 ft. to 4100 ft. (ZZZZZ altitude) I noticed that ZZZZZ3 disappeared entirely from the Navigation Display (ND); so there was no extended runway centerline past ZZZZZ. To TO waypoint was then Runway XXL; and I was surprised to see the vertical indicated +2290 and well below us. This was incorrect for approaching 4100 ft. well prior to ZZZZZ. We continued turn to final in NAV; and had the field and PAPI visually. Since the VNAV path was incorrect; I asked the FO to resequence direct to ZZZZZ; which ran into problems doing since ZZZZZ was now the FROM waypoint and obscured/unavailable on the page. FO had to type it back in; which FO did. I flew Vertical Speed during this time; planning to level at 4100 by ZZZZZ. When the VNAV deviation indicator reappeared; we were very close to path; and I reselected PROFILE and we went into final P SPD/PDES mode as desired; speed 160. I thought all was good at this point and we had a good RNAV setup. It was not until we were passing 3700 ft. on the VNAV path that I realized ZZZZZ was still in front of us; where we should have been 4100 ft. I selected ALT HOLD to level off at 3600; and indicated that I had the PAPI; and no longer trusting the VNAV path; would hold that altitude until getting on the PAPI visual glideslope; we did so; slowing to final flap configuration and before landing checks. At a point on final the autopilot clicked off (3200 ft. as per approach page entered MDA); which surprised me; and I hand flew the rest of the approach on the PAPI; in FlightPath Vector vs the flight director. FO did a nice job with vert speed callouts on the final hand flown segment.I think when the FO went to clean up the approach and delete the turn point; the waypoint sequence had already sequenced past ZZZZZ3; and instead of deleting the T-P; deleted ZZZZZ3. This may have caused the VNAV path to initially be incorrect. Resequencing to ZZZZZ again did not solve the issue; but it was not recognized. The painfully long ALT* at 4100 ft. level off and transition at that point; contributed to not recognizing that descending out of 4100 ft. prior to ZZZZZ was not correctand that the VNAV path we were intercepting was still incorrect. I'm not sure the level off took; either. We both felt behind and did not realize we were actually on an early descent. Accepting all red on the visual PAPI for a while; as normal prior to the FAF/glide path intercept; instead of ensuring we maintained level until getting 2 white contributed to late recognition. We should have transitioned to the visual approach references on the PAPI earlier; upon initial recognition that the RNAV vertical path was not correct; or discontinued the RNAV attempted approach with asking for a hold on base/final to sort it out; or a full go-around. If it had been IMC certainly we would not have continued descent or inside ZZZZZ without having the FMS behaving to our satisfaction. For FMS handling; pay careful attention when deleting points (T-P) on an approach; sometimes a sequence does not need to.
Second reporter narrative
During arrival into ZZZ; we were on downwind and cleared the visual approach after calling airport in sight. The Captain asked me to clean up the approach (RNAV XXL) to give a clean white line to final. When I did that; I believe the FMS sequenced past the point I had used to clean up then led to inaccurate glide slope indication. I attempted to fix the approach in the FMS; but was unable to get it to provide what we believed was accurate glide slope indication. We noticed that it was inaccurate based on being low on the PAPI indication. We leveled off at 3600 ft. until we regained proper glide slope indication for the visual. No terrain alerts or Tower notifications were received. Landing was uneventful. I believe it was an issue with FMS sequencing and I was unable to get it reloaded in time to get good FMS glide slope indication.
Source: NASA Aviation Safety Reporting System (public domain). Reports are voluntary submissions and are not verified by NASA.