Airbus near CFIT in Paris.

Sluggo63

Pattern Altitude
Joined
Oct 9, 2013
Messages
1,897
Display Name

Display name:
Sluggo63
Airbus A320 came within 6 feet of the ground on an RNAV approach into LFPG. Controller gave them a QNH of 1011mb versus the correct 1001mb. The crew read back and set the incorrect QNH that was given to them.

I do have some questions based on the article, like the continued descent after hitting the DA, seemingly not hearing the RA 1000, 500, etc. callouts (could definitely be task saturation), and disconnecting the autopilot for the missed approach.

https://www.flightglobal.com/safety...n-after-pressure-setting-error/149352.article
 
Last edited:
It looks like they applied the go around at the point they THOUGHT the DA was. Big aircraft often sink well below the DA on the go around made at the DA.

I'm confused as to why a bad QNH figured so much into the glidepath though. You would think that the GPS altitude itself would be better than 300' off (notwithstanding the bogus baro assist).

You're right, the RA should have been talking away there. There should be callouts as soon as you go below 1000, 500, and 100 and then pretty continuous from 50 to RETARD RETARD!
 
Baro VNAV uses the altimeter for glidepath calculations. If the altimeter is off, the glidepath will be also.

what I read is that they pushed the GA button and went takeoff thrust well below DA.
 
Baro VNAV uses the altimeter for glidepath calculations. If the altimeter is off, the glidepath will be also.

what I read is that they pushed the GA button and went takeoff thrust well below DA.
10 millibar is about 100 meters?
 
Not to try to get picky, but this isn't CFIT. Maybe CFRRCTT - really, really close to....
 
It looks like they applied the go around at the point they THOUGHT the DA was. Big aircraft often sink well below the DA on the go around made at the DA.
They sink a bit, about 50 feet or so. Not 300 though.
 
Airbus A320 came within 6 feet of the ground on an RNAV approach into LFPG. Controller gave them a QNH of 1011mb versus the correct 1001mb. The crew read back and set the incorrect QNH that was given to them.

I do have some questions based on the article, like the continued descent after hitting the DA, seemingly not hearing the RA 1000, 500, etc. callouts (could definitely be task saturation), and disconnecting the autopilot for the missed approach.

https://www.flightglobal.com/safety...n-after-pressure-setting-error/149352.article
Wonder what Approach it was. Did it have LPV. Kinda looks like it didn't. Those low altitudes seem to have happened to far out if on Glidepath. Must have been doing LNAV/VNAV. Or maybe I just have it pictured wrong.
 
Airlines also do CANPA… so are on a descent like a glide slope, without a glide slope.

A bad altimeter setting could EASILY cause this problem.

CANPA is Constant Angle Non Precision Approach. It being a method of doing a non precision approach, really wouldn’t cause any different result than diving and driving with a bad altimeter setting. Don’t mean to imply that.
 
They sink a bit, about 50 feet or so. Not 300 though.
The didn't sink 300'. They executed the DA at the altitude they thought was the DA. It just happened to be 300' too low

Wonder what Approach it was. Did it have LPV. Kinda looks like it didn't. Those low altitudes seem to have happened to far out if on Glidepath. Must have been doing LNAV/VNAV. Or maybe I just have it pictured wrong.
RNP to LNAV/VNAV minimum.

Ther article seems to be behind a paywall now. Here's the actual BEA report. https://bea.aero/fileadmin/user_upl...minary_report_for_publication_EN_finalise.pdf
 
The Swiss cheese was really trying to line up here. Glad nobody got hurt.
 
So if the article is correct (big IF) then there should be brush/tree branches in the undercarriage...here is a screen capture of google maps where they say the aircraft got within "6 feet of ground", .8nm from 27R:

upload_2022-7-13_19-28-2.png
 
Don't make sense to me. Those are pretty sophisticated planes. I'd think they would have all the goodies.

The lack of SBAS approaches isn't too surprising to me. But I'd love to learn more about why their EGPWS/TAWS didn't alert to this
 
Seems odd an A320 wouldn't be equipped to do LPV. @Sluggo63 , whadda ya know about A320's?
I've never flown, or heard of, an airliner that did LPV approaches.

We do RNP approaches (down to RNP 0.11) and we do GLS approaches which are very similar to an LPV approach but aren't currently widely available. EWR, IAH, and SFO are the US airports that I'm aware of that have GLS approaches.
 
I've never flown, or heard of, an airliner that did LPV approaches.

We do RNP approaches (down to RNP 0.11) and we do GLS approaches which are very similar to an LPV approach but aren't currently widely available. EWR, IAH, and SFO are the US airports that I'm aware of that have GLS approaches.
OK. How hard would it be to just stick a GPS in an Airbus?
 
OK. How hard would it be to just stick a GPS in an Airbus?
The Airbus has GPS. At least two, probably three. (I've never flown the Airbus other than commuting in the jumpseat)

A modern airline has a flight management system (FMS). The names may differ, but the functions are all very similar.

Two, or three, Inertial Reference systems are the heart of the FMS. They need data from their inertial platforms to the FMS. Each inertial reference unit is aligned prior to flight and each then maintains a position (where it thinks the airplane is) based on the accelerations each IRU senses when the airplane is in motion. Inertial platforms aren't perfect so, as the flight progresses, the IRS positions drift away from the aircraft's actual position.

The FMS takes inputs from the IRUs as well as from one of several other inputs. These inputs include DME/DME and Rho/DME from VORs, TACANS, and DME ground stations. GPS updating was added in the 1990s. The FMS calculates the accuracy of each input based on its capabilities and uses the most accurate updating to keep the FMS position as accurate as possible. Since the airplane has two, often three, FMSs, the airplane has two, often three, FMS positions.

The combined accuracy of the FMS systems is displayed to the pilots as ANP or Actual Navigation Performance. The ANP in the 737s (two IRU/two FMS/ two GPS) is typically 0.02 to 0.04 while in flight. The procedures and airspace in which we fly has established RNP or Required Navigation Performance. RNP can be as high as 10.0 in oceanic airspace to as low as 0.11 on the RNP approaches with the lowest landing minimums. If our ANP exceeds the RNP for the airspace or procedure we are flying, we receive alerts which tell us that our RNAV is no longer able to provide the required navigation performance.

RNP approaches are available at a lot of airports and allows for very accurate curved paths, including radius-to-fix or RF legs, and vertical paths. Eugene, OR has a nice RNP approach that leads you down a curved path between mountain peaks to land north (Rwy 32?). There is no ILS to that runway because of the terrain.

GLS is a form of Local Area Augmentation System (LAAS) which uses a single ground station which transmits a correction value, to the nearby airplanes flying one of many GLS approaches that can be available at the airport, that improves the accuracy of the GPS position in that local area. GLS provides minimums similar to CAT I ILS approaches and is intended to eventually provide for autoland.

But, we don't do LPV.
 
I’m more likely to ‘preset’ the altimeter setting an hour or two out, still flying the flight levels. One gets it from electronic ATIS or a recent observation. Then again the controller will offer the altimeter setting as you start down, often an airport 50 miles or so from the destination.

Yes it could change a bit that last hour, usually not by much.
 
The Airbus has GPS. At least two, probably three. (I've never flown the Airbus other than commuting in the jumpseat)

A modern airline has a flight management system (FMS). The names may differ, but the functions are all very similar.

Two, or three, Inertial Reference systems are the heart of the FMS. They need data from their inertial platforms to the FMS. Each inertial reference unit is aligned prior to flight and each then maintains a position (where it thinks the airplane is) based on the accelerations each IRU senses when the airplane is in motion. Inertial platforms aren't perfect so, as the flight progresses, the IRS positions drift away from the aircraft's actual position.

The FMS takes inputs from the IRUs as well as from one of several other inputs. These inputs include DME/DME and Rho/DME from VORs, TACANS, and DME ground stations. GPS updating was added in the 1990s. The FMS calculates the accuracy of each input based on its capabilities and uses the most accurate updating to keep the FMS position as accurate as possible. Since the airplane has two, often three, FMSs, the airplane has two, often three, FMS positions.

The combined accuracy of the FMS systems is displayed to the pilots as ANP or Actual Navigation Performance. The ANP in the 737s (two IRU/two FMS/ two GPS) is typically 0.02 to 0.04 while in flight. The procedures and airspace in which we fly has established RNP or Required Navigation Performance. RNP can be as high as 10.0 in oceanic airspace to as low as 0.11 on the RNP approaches with the lowest landing minimums. If our ANP exceeds the RNP for the airspace or procedure we are flying, we receive alerts which tell us that our RNAV is no longer able to provide the required navigation performance.

RNP approaches are available at a lot of airports and allows for very accurate curved paths, including radius-to-fix or RF legs, and vertical paths. Eugene, OR has a nice RNP approach that leads you down a curved path between mountain peaks to land north (Rwy 32?). There is no ILS to that runway because of the terrain.

GLS is a form of Local Area Augmentation System (LAAS) which uses a single ground station which transmits a correction value, to the nearby airplanes flying one of many GLS approaches that can be available at the airport, that improves the accuracy of the GPS position in that local area. GLS provides minimums similar to CAT I ILS approaches and is intended to eventually provide for autoland.

But, we don't do LPV.
Thanks for all the info. Ok, ya got GPS. My guess is without WAAS though, that's why no LPV. How'd I do
 
Weird ATC never saw a wrong level off altitude. Although once in a while you just don’t have one, and the “checking in descending through 8.5 for 4000” really never confirms altitude accurately enough.
 
Weird ATC never saw a wrong level off altitude. Although once in a while you just don’t have one, and the “checking in descending through 8.5 for 4000” really never confirms altitude accurately enough.
I think I saw someplace that the transition altitude is only 5000 ft in that area, so they may not have been leveled off very much at all.
 
How many airports do they go to that don’t have ILSs to most, if not all, runways?

I think the ILS for 27R was out of service. So they were given the RNP.

One other thing that stands out is this: Both the Airhub and EasyJet exchanges were in English. The controller subsequently gave descent instructions to an Air France crew, in French, using the correct QNH figure.

Apparently, it is common in certain regions for the aircrews and ATC to converse in their native languages. From chatting with a buddy who flies for a low-budget European carrier, this is the case in France. I also believe this is the case in some South American countries. Wonder if the Airhub flight crew would have questioned the QNH they were given if they heard the Air France crew get a different setting from the controller in English?

I’m more likely to ‘preset’ the altimeter setting an hour or two out, still flying the flight levels. One gets it from electronic ATIS or a recent observation. Then again the controller will offer the altimeter setting as you start down, often an airport 50 miles or so from the destination.

Yes it could change a bit that last hour, usually not by much.

That's another surprising thing, here. I know some airline SOP have the PM preset the QNH just after TOC... Again, this is from a buddy in the UK who flies 737 NGs so not sure if you can do that on an Airbus, but I assume there is some way. Or at least to note it for reference? Then, if it differs significantly, it can be questioned. Interesting that the EasyJet crew was given the QNH of 1011, yet read back the correct setting of 1001. Wonder if they had it preset... and they just read back what they had set off of a QNH they got from datalink weather? And EasyJet operates Airbus-es (Airbi?), so the SOP would be similar, I would think.

I'm also surprised there isn't a warning that sounds if on an RNP approach and descending well below the generated glide path - like the glide slope warning that you get on an ILS?

But yeah, the holes in the queso were lining up big-time on this one. Glad everything worked out ok. Wonder if any changes will come out of this.
 
I've never flown, or heard of, an airliner that did LPV approaches.

We do RNP approaches (down to RNP 0.11) and we do GLS approaches which are very similar to an LPV approach but aren't currently widely available. EWR, IAH, and SFO are the US airports that I'm aware of that have GLS approaches.
Our A220s can do LPV approaches.
 
Better and a lot quieter to be 6' up than 6' too low. To me a reminder that the tolerances for missed approaches aren't real high. Like the UPS pilot and the smokestack.

I know it's common for private pilots to have "personal minimums" that are more conservative than the regs. Does anyone pad the go around points for instrument landing when flying part 91?
 
The didn't sink 300'. They executed the DA at the altitude they thought was the DA. It just happened to be 300' too low


RNP to LNAV/VNAV minimum.

Ther article seems to be behind a paywall now. Here's the actual BEA report. https://bea.aero/fileadmin/user_upl...minary_report_for_publication_EN_finalise.pdf
You're right, it wasn't 300 feet, I miscalculated based on the article.

They hit the DA on their altimeters at 122 RA, then 6 seconds later at 52 RA they disconnected the autopilot and pitched up, then 3 seconds later at 6 RA they put the thrust levers in TOGA.

Seems odd an A320 wouldn't be equipped to do LPV. @Sluggo63 , whadda ya know about A320's?
Absolutely zilch. But we aren't allowed to do LPVs at my airline. We had to do the LNAV/VNAV going into CDG while the ILS for 27R was OTS. But, again, I know nothing about the A320s and their capabilities.

So, I'm not sure what their Go Around procedures are. And I don't know if them disconnecting the autopilot has something to do with what happened to the Airbus at the Paris air show years ago when it went into the trees. If I remember correctly, in that accident, the plane didn't respond to the TOGA because it was too low RA, so the plane was in "land mode" and wouldn't allow them to go around. I wonder if the same type of thing happened here, where the crew toggled TOGA, the plane didn't respond because of some flight control law/mode it was in, and they had to disconnect the AP to get it to go around. But, again, because of my zero-knowledge of anything Airbus, I can't say.
 
Last edited:
Wonder if the Airhub flight crew would have questioned the QNH they were given if they heard the Air France crew get a different setting from the controller in English?
So much this! You win the astute award for this thread. We've been saying this for decades. The fact that French controllers speak to French aircraft in French is a huge safety issue. I was based out of De Gaulle for years and still fly there regularly, so I'm pretty familiar with the operations there. There is such a lack of SA because of the split language there. It's near impossible to clear on the radio or build a mental picture of what's going on around you because half the radio calls aren't in English. There was a survey years ago sent out by ICAO about this exact issue, and the responses were overwhelmingly that it should be English only in France, but the French just disregarded this and continue with the French/English mix.

Same thing happens in China (which is worse), but that's a tougher nut to crack.



That's another surprising thing, here. I know some airline SOP have the PM preset the QNH just after TOC... Again, this is from a buddy in the UK who flies 737 NGs so not sure if you can do that on an Airbus, but I assume there is some way. Or at least to note it for reference? Then, if it differs significantly, it can be questioned. Interesting that the EasyJet crew was given the QNH of 1011, yet read back the correct setting of 1001. Wonder if they had it preset... and they just read back what they had set off of a QNH they got from datalink weather? And EasyJet operates Airbus-es (Airbi?), so the SOP would be similar, I would think.
We (Boeing, not Airbus) can preset the local altimeter while still having 29.92/1013 set in the altimeter. We'll usually do it before TOD after we get the latest ATIS that we brief off of. If the controller gives us something significantly different than what was on the ATIS, I'll definitely query them on it. Surely a 1001/1011 difference would get my attention.
 
I think I saw someplace that the transition altitude is only 5000 ft in that area, so they may not have been leveled off very much at all.
Transition Altitude is 5000 in France. Transition Level (which would be what's pertinent here) depends on the local altimeter and is given on the ATIS, but typically it's around FL060.
 
Thanks for all the info. Ok, ya got GPS. My guess is without WAAS though, that's why no LPV. How'd I do

Safe assumption that they have WAAS. ADSB essentially mandates WAAS for a position source
 
Safe assumption that they have WAAS. ADSB essentially mandates WAAS for a position source

Not a safe assumption that WAAS is available for navigation just because they have ADS-B.
 
Not a safe assumption that WAAS is available for navigation just because they have ADS-B.

True. And in this case they absolutely did not have WAAS, since they're in Europe. They're likely equipped with an SBAS receiver, which may or may not be used for navigation. And even if the SBAS receiver is connected to the FMS/FMC, it doesn't guarantee they have LPV capability
 
We (Boeing, not Airbus) can preset the local altimeter while still having 29.92/1013 set in the altimeter. We'll usually do it before TOD after we get the latest ATIS that we brief off of. If the controller gives us something significantly different than what was on the ATIS, I'll definitely query them on it. Surely a 1001/1011 difference would get my attention.
Is this a disadvantage of metric? 1001/1011 seems easy to be a bit dyslexic on and miss a 0 or a 1 vs only needing to really listen to the final decimals after a 30 or a 29?
 
Thanks for all the info. Ok, ya got GPS. My guess is without WAAS though, that's why no LPV. How'd I do
I don't know any of the specifics of the GPS systems that provide position updates to the FMSs. All of our RNAV is based on RNP. We can do RNP down to 0.11. We don't see the GPS or GPS position anywhere. Everything updates the two (or three) FMS and each FMS maintains a position based on the best available information as I described above.

Apparently, it is common in certain regions for the aircrews and ATC to converse in their native languages.
That's the rule, not the exception. The only place I can think of where local crews didn't routinely use their local language was Japan.

I assume there is some way. Or at least to note it for reference?
Every modern airliner I've flown (i.e. not the DC9 or DC8) has had the ability to preset the QNH while staying on 29.92. And some of the DC8/DC9s had newer altimeters installed which also had that capability.

Our A220s can do LPV approaches.
That's not surprising. The design is new enough that it was designed into the avionics from the start.
 
Looks like SLS is pretty new to the 320 fleet and requires updates to the receiver (SLS needs a multi-mode receiver or MMR), updates to the autopilot as well as the flight warning computer. Also requires updates to the (Honeywell only) FMS - I guess the FMS has to be changed so the database can store and make available these type of approaches. Long story short - I bet this particular A320 did not have the upgraded SLS capability that would have allowed them to shoot the straight-in LPV approach.
 
Back
Top