Police helicopter crashes onto Glasgow pub: final AAIB report
Yes really.
There was ample time to crosscheck the flicking of the right switch. A possible reason not to switch on the XFER pumps but switch on the PRIME pumps is if there is no valid reason to, that is, that 75Kg fuel remaining was presented as 0kg or unusable - which we will never know.
There was ample time to crosscheck the flicking of the right switch. A possible reason not to switch on the XFER pumps but switch on the PRIME pumps is if there is no valid reason to, that is, that 75Kg fuel remaining was presented as 0kg or unusable - which we will never know.
Originally Posted by AAIB
Since the helicopter had 76 kg of fuel on board at the time of the accident, the pilot might have been expected to make a PAN call, upgrading it to a MAYDAY on reaching the Final Reserve Fuel IFR [i.e. 85kg], if he was aware of the fuel state. Again, due to the lack of evidence that might otherwise have been provided by cockpit voice and flight data recorders, the investigation was unable to determine the reasons for this apparent omission.
Originally Posted by AAIB
The FLIR system may have been used during this final task and verbal exchanges probably took place within the helicopter but no evidence was recovered which could explain the reason for the 5 minutes spent in that location or provide any indication of the crew’s awareness of the fuel situation. It was calculated that, as the
helicopter left Bargeddie, a total of 87 kg of fuel remained on board.
helicopter left Bargeddie, a total of 87 kg of fuel remained on board.
Last edited by PAX_Britannica; 23rd Oct 2015 at 12:48. Reason: typo,punctuation
My vision of possible occurrences after reading the final report:
- both supply tanks did falsly indicated full on the CAD
(a 0.5ml (!) water globule at the capacitance based fuel sensors can do the trick, see p. 54, G-NWEM)
- "Low Fuel" Warnings (driven by correctly functioning thermistors), were considered as acting up "shi**y EC".
The were acknowlegded as being irrelevant, because of (incorrect) triple "plenty fuel" indication on the CAD.
That is the reason why he flew that final task, although he just had acknowledged the Low Fuel warnings, multiple times!
- right engine dies, still with wrong fuel indication of CAD or suddenly correct emtpy indiction
(as the waterglobule was finally gone)
- stress level rises: "damn EC, lets do the OEI drill, we've got at least 4 min to go till trouble"
pilot commences OEI setting checklist, switch check, etc, when
- he was scared by 2nd engine spooling down 32 secs later
coupled with lights out and radalt off, right where he'd never planned to land
That leaves the following:
pilot errors:
- not being aware of sensor reliability differences between CAD tank indication graphics and thermistor driven LOW FUEL warning lights (e.g. low fuel override any CAD fancy graphics)
- botching the autorotation due to high stress when at night his ac goes dark, that was showing sufficient fuel in all three CAD graphics regions, just seconds ago.
It does need some cold heartedness to put the sole blame on him.
His actions do not line up with an +5500h pilot flying with NVG and all, from radio comm (or lack thereof) he appeared very relaxed:
He might have felt to be sitting in an ac acting up with the insider-known unreliable fuel indicators.
Operators have been sending in "bad fuel sensors" for repair and mostly got them back with "all within sepc".
Therefore he showed no noticable stress on the radio.
Therefore he did the final sortie(?) near Bargeddie, with LOW FUEL warnings, w/o being concerned or feeling negligent.
Checking xfr switches after right engine stopped wouldn#t feel very urgent had the CAD indicated ok supply tanks.
eurocopter error: stupid fuel pumps setup
The fuel system features "a pump runs dry" sensor. Only after the sensor is on for three minutes uninterruptedly the "F PUMP AFT/FWD" warning comes on. AND the warning does come on too, if the pumps are off, but immersed in fuel for three minutes! See p.16
This super manual fuel pump operation regime, coupled with varying fuselage attitude and unergonomic switch placement, presents an engineering wise totally unnecessary work load for the pilot.
The Reely recommended setup would be:
Install transfer pump switches featuring three positions: off, auto, on
auto will start the pump after it is immersed in fuel for 10 seconds straight, and will stop the pump after it was "dry" for 10 seconds straight.
Having one of those "reely transfer pump switches" to "OFF" should be indicated on the CDS as red warning text (starvation risk)
Having it on "ON" would be an amber caution text (dry run risk).
And only with the switches on auto there'd be no message.
The nice part: all you need is already in place: There is a software that controls the warning light. It maintains timers, resets them to 0 when ever the pump gets splashed at from time to time, and only fires the warning light after sustained three minutes of a "dry" or "immersed" condition. All they need is two switches, two relais and a slight amendment of the current transfer pumps monitoring software.
That'd preserve the pilot's old style manual (override via off-on) control over the pumps,
and, when set to auto, he could totally forget about transfer pumps mgmt.
- both supply tanks did falsly indicated full on the CAD
(a 0.5ml (!) water globule at the capacitance based fuel sensors can do the trick, see p. 54, G-NWEM)
- "Low Fuel" Warnings (driven by correctly functioning thermistors), were considered as acting up "shi**y EC".
The were acknowlegded as being irrelevant, because of (incorrect) triple "plenty fuel" indication on the CAD.
That is the reason why he flew that final task, although he just had acknowledged the Low Fuel warnings, multiple times!
- right engine dies, still with wrong fuel indication of CAD or suddenly correct emtpy indiction
(as the waterglobule was finally gone)
- stress level rises: "damn EC, lets do the OEI drill, we've got at least 4 min to go till trouble"
pilot commences OEI setting checklist, switch check, etc, when
- he was scared by 2nd engine spooling down 32 secs later
coupled with lights out and radalt off, right where he'd never planned to land
That leaves the following:
pilot errors:
- not being aware of sensor reliability differences between CAD tank indication graphics and thermistor driven LOW FUEL warning lights (e.g. low fuel override any CAD fancy graphics)
- botching the autorotation due to high stress when at night his ac goes dark, that was showing sufficient fuel in all three CAD graphics regions, just seconds ago.
It does need some cold heartedness to put the sole blame on him.
His actions do not line up with an +5500h pilot flying with NVG and all, from radio comm (or lack thereof) he appeared very relaxed:
He might have felt to be sitting in an ac acting up with the insider-known unreliable fuel indicators.
Operators have been sending in "bad fuel sensors" for repair and mostly got them back with "all within sepc".
Therefore he showed no noticable stress on the radio.
Therefore he did the final sortie(?) near Bargeddie, with LOW FUEL warnings, w/o being concerned or feeling negligent.
Checking xfr switches after right engine stopped wouldn#t feel very urgent had the CAD indicated ok supply tanks.
eurocopter error: stupid fuel pumps setup
The fuel system features "a pump runs dry" sensor. Only after the sensor is on for three minutes uninterruptedly the "F PUMP AFT/FWD" warning comes on. AND the warning does come on too, if the pumps are off, but immersed in fuel for three minutes! See p.16
This super manual fuel pump operation regime, coupled with varying fuselage attitude and unergonomic switch placement, presents an engineering wise totally unnecessary work load for the pilot.
The Reely recommended setup would be:
Install transfer pump switches featuring three positions: off, auto, on
auto will start the pump after it is immersed in fuel for 10 seconds straight, and will stop the pump after it was "dry" for 10 seconds straight.
Having one of those "reely transfer pump switches" to "OFF" should be indicated on the CDS as red warning text (starvation risk)
Having it on "ON" would be an amber caution text (dry run risk).
And only with the switches on auto there'd be no message.
The nice part: all you need is already in place: There is a software that controls the warning light. It maintains timers, resets them to 0 when ever the pump gets splashed at from time to time, and only fires the warning light after sustained three minutes of a "dry" or "immersed" condition. All they need is two switches, two relais and a slight amendment of the current transfer pumps monitoring software.
That'd preserve the pilot's old style manual (override via off-on) control over the pumps,
and, when set to auto, he could totally forget about transfer pumps mgmt.
Last edited by Reely340; 23rd Oct 2015 at 15:38.
Join Date: Aug 2002
Location: Craggy Island
Posts: 69
Likes: 0
Received 0 Likes
on
0 Posts
One little extra point, if the supply tanks were faulty showing full (my personal theory), the CAD Amber fuel caution does not operate when supply tanks go below 32ish. See NWEM pictures for proof. The board 'found no evidence of CAD fuel illuminating'.
Sadly my good friend ignored a critical warning and continued on task. He must have been misled by the gauges. To say there was no evidence to show a fuel system fault is a poor conclusion. That moves blame from 50:50 pilot:manufacturer to 100% pilot. Sad, but not unexpected
Sadly my good friend ignored a critical warning and continued on task. He must have been misled by the gauges. To say there was no evidence to show a fuel system fault is a poor conclusion. That moves blame from 50:50 pilot:manufacturer to 100% pilot. Sad, but not unexpected
Join Date: Apr 2010
Location: London
Posts: 7,072
Likes: 0
Received 0 Likes
on
0 Posts
Looks like he switched the pumps off and just forgot to switch them back on - all the warnings he was getting he just fixated as error warnings
"Up to the accident to G-SPAO, the EC135 had accumulated more than three million flying hours, over a period of about 20 years, and there had not previously
been a reported instance of fuel starvation"
there but for the grace of God go ................
"Up to the accident to G-SPAO, the EC135 had accumulated more than three million flying hours, over a period of about 20 years, and there had not previously
been a reported instance of fuel starvation"
there but for the grace of God go ................
Join Date: Sep 2015
Location: London
Posts: 78
Likes: 0
Received 0 Likes
on
0 Posts
How hard is it to judge height above Roof Tops in the Dark?
There are Street Lights and other light sources visible to aid in judging height.
Add that to the definition of angle by looking at building shapes, streets, vehicles (moving and parked)....there is still a lot to see that would aid in determining height.
Somehow I find it hard to accept that the Pilot would misjudge aircraft height by Hundreds of Feet even without a RadAlt or Landing Light being functional.
There are Street Lights and other light sources visible to aid in judging height.
Add that to the definition of angle by looking at building shapes, streets, vehicles (moving and parked)....there is still a lot to see that would aid in determining height.
Somehow I find it hard to accept that the Pilot would misjudge aircraft height by Hundreds of Feet even without a RadAlt or Landing Light being functional.
The roof of that pub, at night and in rain could easily be mistaken for tarmac, unfortunately we will never know for 100% certainty, the facts are that an experienced high time pilot got in to a situation that snow balled very quickly resulting in a serious accident.
We will all have our own opinions but I think it is disingenuous to start portioning blame on the pilot in light of how much is still unknown about the root and subsequent causes. In all likelihood it would have been a chain reaction sequence of events that ultimately caused the crash.
I note that the target date for the development of a low cost FDR is December 2000, Already missed. The main work to install a FDR is the wiring and fitting of sensors, also some instruments may not be compatible. A CVR fit is relatively easier.
Join Date: Dec 2013
Location: Warks
Posts: 71
Likes: 0
Received 0 Likes
on
0 Posts
I thought very little was made of the error in the manufacturers manual regarding the length of time between engine flameouts. Stated as 4 mins, yet even with the supply tank contents as prescribed in ideal conditions, it was nearer 1 min 30, MAX. Manual has now been corrected, but if I had been over a city centre and that close to base, I know how I would have been thinking of using those 4 minutes ... And it doesn't involve an emergency landing in the city centre.
A very good pilot simply ran out of time. 30 seconds is too short a time for the second engine to fail for the same fuel starvation reasons.
The report states:
"Corrective action is being taken by the manufacturer to amend the potential
engine flameout interval to a more accurate figure in the next revision of the
SDS manual, scheduled for publication in December 2015." The manual says that the 4.5kg difference between tanks allows 3-4 minutes between engine flame outs.
It is my belief that give the pilot 3-4 minutes rather than an unexpectedly short 30 seconds and he would have had a very good chance of getting that craft and its occupants down intact.
It is not the manual that needs updating, it is the design that needs changing to increase the differential in tank size, ensuring that the pilot gets his full four minutes. Maybe it is not in the AAIB remit to make this recommendation but it seems an obvious fix and provides a safety net to the same situation developing again.
A further thought on the supply tank capacities. We know that the right tank holds 4kg less than the left tank, the idea apparently being to prevent both engines flaming out at the same time due to fuel starvation.
Given that the aircraft would have been flown exclusively from the right hand seat, would there not be a natural tendency for orbits to be flown to the right, as was the case here? In such circumstances, unless the turns are perfectly balanced, there would be a tendency for fuel to cross the tank divide from the tank which was supposed to hold more, to the tank which was supposed to hold less?
For as long as there is sufficient fuel in the main tank to keep the supply tanks topped off, this situation would correct itself, but once you have insufficient fuel to top them off, and/ or the transfer pumps are switched off, there is no way to correct this situation, so all bets should be off as to which engine would flame out first, and how long a gap there would be between flameouts.
Would the situation not be better if the supply tanks were in fact completely separate from each other, giving greater predictability over what their relative contents would be?
The manual said he had 4 minutes from first flame out to get the aircraft on the ground, and it is very hard to imagine the shock when the second engine flamed out so quickly thereafter. The odds against a successful autorotation from that height in those circumstances were always significant; that degree of shock made them overwhelming.
Which does nothing to address the elephant in the room: why does there seem to be no sign of fuel anxiety in any of the pilots actions? Everything in hindsight screams "land" and yet he was taking on additional apparently routine tasking beyond what would seem any point of reason. I must confess to being a little underwhelmed by the report in its lack of exploration of this aspect. We learn that the initial tasking was to do with an incident on a railway, but what were these routine tasks at the key stage of the sortie which seem to have cropped up at such short notice? Could they have affected the pilots judgement, for example relating to a threat to life, sufficient to persuade him to stretch the elastic a bit further?
I appreciate it is not the AAIBs remit or style to speculate, but the report just seems to fall short.
Given that the aircraft would have been flown exclusively from the right hand seat, would there not be a natural tendency for orbits to be flown to the right, as was the case here? In such circumstances, unless the turns are perfectly balanced, there would be a tendency for fuel to cross the tank divide from the tank which was supposed to hold more, to the tank which was supposed to hold less?
For as long as there is sufficient fuel in the main tank to keep the supply tanks topped off, this situation would correct itself, but once you have insufficient fuel to top them off, and/ or the transfer pumps are switched off, there is no way to correct this situation, so all bets should be off as to which engine would flame out first, and how long a gap there would be between flameouts.
Would the situation not be better if the supply tanks were in fact completely separate from each other, giving greater predictability over what their relative contents would be?
The manual said he had 4 minutes from first flame out to get the aircraft on the ground, and it is very hard to imagine the shock when the second engine flamed out so quickly thereafter. The odds against a successful autorotation from that height in those circumstances were always significant; that degree of shock made them overwhelming.
Which does nothing to address the elephant in the room: why does there seem to be no sign of fuel anxiety in any of the pilots actions? Everything in hindsight screams "land" and yet he was taking on additional apparently routine tasking beyond what would seem any point of reason. I must confess to being a little underwhelmed by the report in its lack of exploration of this aspect. We learn that the initial tasking was to do with an incident on a railway, but what were these routine tasks at the key stage of the sortie which seem to have cropped up at such short notice? Could they have affected the pilots judgement, for example relating to a threat to life, sufficient to persuade him to stretch the elastic a bit further?
I appreciate it is not the AAIBs remit or style to speculate, but the report just seems to fall short.
Join Date: Jan 2010
Location: Center of the Universe
Posts: 14
Likes: 0
Received 0 Likes
on
0 Posts
Been following this accident from a great distance for quite awhile. Easy to have missed this, but the question that keeps me thinking about this is, "why did this helicopter depart on the mission with so small a fuel reserve?" It seems that modern technology encourages fuel planning to the last drop. Numerous accidents have shown that this is fraught with peril.
Reely340
Please tell me there aren't other pilots out there, who having just done 1:37 in a 135, would do this!! 
Therefore he did the final sortie(?) near Bargeddie, with LOW FUEL warnings, w/o being concerned or feeling negligent
Last edited by Sky Sports; 23rd Oct 2015 at 16:48.
Join Date: Apr 2008
Location: In a house
Posts: 56
Likes: 0
Received 0 Likes
on
0 Posts
That is an interesting point.
Hopefully there will now be a Fatal Accident Inquiry (required by law in Scotland when someone is killed at work). The Sheriff will want to know details of any police communications.
Hopefully there will now be a Fatal Accident Inquiry (required by law in Scotland when someone is killed at work). The Sheriff will want to know details of any police communications.
I know the AAIB got a lot of flak for the time it took to publish this report, but given the amount and depth of testing they did on the aircraft systems, it now seems justified.
Join Date: Oct 2008
Location: South West Scotland
Posts: 8
Likes: 0
Received 0 Likes
on
0 Posts
Yes if there were any communications between the crew and a Police control then it would be recorded, but obviously there was nothing or it would have mentioned it in the report.
Join Date: Jan 2010
Location: Center of the Universe
Posts: 14
Likes: 0
Received 0 Likes
on
0 Posts
It departed with the required full reserve, but flew past it.
The fact the pilot had 20 audio or visual warnings of his low fuel state, prior to engines shutting down, (table 3, page 33), leads me to believe that he thought they were spurious warnings and decided to carry on with taskings.
But even still, would you risk it?
But even still, would you risk it?
"Financially catastrophic for BOND."
What an odd and very unpleasant comment. 10 people died. Why would you even start thinking about financial consequences?
If the AAIB have not apportioned blame, I fail to understand why anyone on these boards would wish to, or try to.
What an odd and very unpleasant comment. 10 people died. Why would you even start thinking about financial consequences?
If the AAIB have not apportioned blame, I fail to understand why anyone on these boards would wish to, or try to.
You seem determined to make out the pilot to be some sort of arrogant, thoughtless idiot. The fact that he was none of the above suggests there was a little more going on than thoughtless, idiotic arrogance, the reasons for which we are very unlikely to find out.