House of Lords - Chinook ZD 576

Select Committee on Chinook ZD 576 Report

CHINOOK ZD 576

PART 6: CONCLUSIONS

144. When they drafted their remarks upon the investigating board's report both Air Marshals had before them evidence as to weather conditions at the lighthouse and higher up the hill, and Mr Holbrook's single answer to the board, from which they deduced that the conditions of cloud prevailing at the lighthouse would have prevented the pilots from seeing the Mull as they approached it.

145. Mr Holbrook's evidence to the FAI and to us puts a different complexion on the matter. We also saw a photograph taken by a Mrs M J Gresswell at the lighthouse minutes before the crash in which the sea appeared to be visible; but this was not altogether clear and it would be unwise to place too much reliance on it. However, in the light of Mr Holbrook's evidence together with the views of Group Captain Crawford that the crew had seen the Mull when they made the way point change and the evidence of Witness A to the same effect, we conclude that the crew had probably seen the land mass at or before the time the way point was changed. Squadron Leader Burke said that he had flown to the Mull and very similar areas and that a line of breakers could nearly always be seen even if nothing else (Q 779). Given an onshore wind of 25 knots (force 6 on the Beaufort scale) it seems more than probable that breakers would be readily visible from a low-flying aircraft out at sea.

146. Negligence, as a concern of the board, was not an abstract concept but could only be inferred from facts relevant to causation. Paragraph 1 of Annex G to Chapter 8 of AP3207 stated that causes of accidents broadly speaking fell into three categories: technical faults, natural operating or medical hazards and human failings. Paragraph 2 provided among other things:

b. Ought the person to have foreseen that their action or their failure to take action would in all probability occasion the final event?"

147. In the context of the Air Marshals' conclusion that the pilots were grossly negligent in placing the aircraft in the position in which it was at or before the way point change was made, regardless of what happened thereafter, the question to be answered is whether there is absolutely no doubt whatsoever that they ought to have foreseen that their action would in all probability occasion the final event. It must be borne in mind that it is not known at what height or speed the aircraft was flying at the way point change, nor its position in relation to cloud. However Sir William accepted the possibility that they could have seen the coastline under cloud cover (Q 364). Furthermore the Air Marshals' views as to the danger of the aircraft being at or in the vicinity of the way point change position even if the crew had intended to alter course at that point were much influenced by the high speed at which they assumed the aircraft then to be travelling - an assumption which, having regard to the deficiencies in the simulation which have now emerged, may no longer be justified.

148. We consider that Sir John's conclusions on this matter must be weakened by his reliance on matters which he treated as facts but which have been demonstrated to our satisfaction to be not facts but merely hypotheses or assumptions. It must be a matter of speculation what would have been the Air Marshals' conclusion if the Boeing simulation had not been available, or if its deficiencies had been identified.

149. Sir John stated that at the speed at which the aircraft was going the crew would have needed to start a 30 degree bank turn at the point where the way point was changed "if they were to stand a reasonable chance of not striking the ground". The necessary angle of bank would have increased as the aircraft approached the land (Q 1032). This of course assumes that the aircraft was then and thereafter under control, travelling at the groundspeed of 174 knots (150 knots airspeed plus 24 knots tailwind) used by the Boeing simulation. He further said that if they had left it to the way point change "they probably, possibly, would have hit the cliffs even in the turn" (Q 1042). He later described a turn in such circumstances as "dangerous" (Q 1060). He made no reference however to the possibility that the crew in making such a turn could have reduced speed, thereby significantly reducing the radius of turn, a result well demonstrated by the formula referred to in paragraph 138 above.

150. Sir William appeared to accept that if, having visibility of 1000 metres, the crew had altered course at the way point change and flown maintaining visual contact with the coast, they would have been "perfectly entitled" to do this (QQ 355, 394, 1039). However on his second appearance he rather departed from this view and explained that, if the aircrew had 1000m visibility, they would have seen that they were displaced from their planned track some 8 or 9 seconds before they made the way point change, and should therefore have altered course earlier (Q 1059). Both Air Marshals attached importance to the results of the simulation, and in particular to the high speed at which the aircraft was assumed to be travelling at or before the way point change, an importance which must now be considered doubtful given the deficiencies already referred to in the simulation. In any event, even if the aircraft was travelling at the assumed high speed at or before the way point change, no reason has been suggested as to why speed could not have been reduced in making any subsequent turn, thereby reducing its radius.

151. This evidence certainly does not justify a finding that there is absolutely no doubt whatsoever that an alteration of course at or very shortly after the actual point of way point change would have resulted in a crash. In this context Squadron Leader Burke's view as to the Chinook's ability to spin round "just like a top" (Q 747) must also be taken into account.

152. We have approached the foregoing question upon the basis that the way point change was made when the aircraft was some 600 metres from the cliff. This assumes that the position divulged by the TANS was precisely accurate - an assumption which may not be justified. It will be remembered that the TANS recorded the height above sea level of the point of impact as 665 ft (± 50), whereas in fact it was 810 ft (see above, para 53). If the aircraft had been on the programmed track it would probably have been 1 kilometre from land ahead at the way point change and there is no suggestion that an alteration of course at that point would have resulted in a crash.

153. If as we consider to be the case there was no justification for holding that the pilots were negligent in placing the aircraft at the actual way point change, the next question is whether nevertheless there was evidence which would justify a finding of negligence in relation to subsequent events. If, in the knowledge that they were close to high ground, whether at the assumed or actual way point change, the pilots voluntarily thereafter maintained course and speed until they reached the point where no actions on their part would avoid a crash, then there would be no doubt as to their negligence. Is this what happened?

154. Ever since the final proceedings of the Board were made known there has been much discussion in Parliament, the media and elsewhere as to why two highly skilled helicopter pilots, knowing how close they were to high ground, should have deliberately flown their aircraft into that ground. It has been suggested that this scenario is positively bizarre, particularly in view of the making of the way point change some seconds before.

155. Air Commodore Crawford, in addition to commenting in his Remarks on the significance of the way point change, stated in evidence that it would have been very unwise to have altered the way point if it was intended not to alter course but to fly over the Mull and he did not think that any normal pilot would have done it (Q 872). Witness A was asked whether he could think of any reason why having changed the way point the aircraft should have continued on its existing course at an apparently slow ROC to the Mull. He replied, "That is the crux of the matter. I cannot think of any reason why the crew would have elected to do that unless they were not doing it of their own volition" (Q 802).

156. Witness A later pointed out that the Special Forces crews exercised regularly in the area and that all of the crews were pretty familiar with the landscape and weather conditions associated with it. He continued, "I have operated with both of the individuals concerned in that area a number of times and it would be normal practice, if the visibility was poor, to remain below the cloud level and use the coastline, as we call it, as a 'handrail' and follow the coastline up towards the next turning point. So I can think of no reason why they should commit themselves to instrument flight in that area if they could see the Mull." It was put to him that the finding of negligence was based on the premise that the act of negligence was not changing to IFR 20 seconds or so before impact: was there any reason why they should change to IFR if their visibility was all right? He replied, "Absolutely none, my Lord. Everything pointed towards them continuing VFR flight, both the weather, the task, the icing limitation on the aircraft, everything pointed towards them continuing that flight VFR." He was asked whether the way point change was further evidence pointing in the same direction; he replied that it was (QQ 836-8).

157. Witness A also referred to the quite dramatic down-draughting and turbulence likely to be present on the west side of the Mull with the wind from 170 degrees as another good reason for the crew to turn away early from the land (Q 849). Squadron Leader Burke explained that if a way point was changed the pilot would not normally be expected to maintain the existing course unless he wanted to fly to a point to update the navigation system, something he could not do in bad weather (Q 779).

158. It is clear from the evidence of both Squadron Leader Burke to us (Q 779) and Witness A to the FAI[38] that if the crew had altered course at the way point change they would not necessarily have adopted the course indicated by the TANS but could have perfectly properly altered course further to port and then flown a course to Corran over the sea and parallel to the coast possibly at a reduced height. There is, of course, no evidence that the low cloud hugging the upper part of the Mull extended to any measurable extent over the sea.

159. Against that background the movements of the aircraft after the way point change must be considered. At the risk of repetition it may be worth setting out again such information as could be gathered about the last seconds of the aircraft's flight:

(ii) of 0.95 nm from the point of impact,

(b) the TANS recorded the height of the aircraft between 15 and 18 seconds before power down as 468 feet plus or minus 50, and

160. Initial impact estimations deduced from marks on the ground and the state of the wreckage which were considered by Mr Cable to be reliable (AAIB statement para 6) were:

Flight path - 20° Up relative to the horizontal

Pitch Angle - 30° Nose up approx

Roll Angle - 5-10° Left

Yaw Angle - Probably less than 10°

Track - 025°M(L) [i.e. approx 015°(G), 012°(T)]

Forward groundspeed - Considerably in excess of 100 knots

161. There is however no evidence to establish (a) the time of the way point change, (b) the height of the aircraft at the way point change, (c) the position of the aircraft when at the recorded height, (d) the course and speed of the aircraft at either of the two foregoing events or indeed at any time prior to impact, nor (e) that the aircraft was in cloud at the time of the height recording. Given the evidence of Mr Holbrook that the cloud was hugging the land, the fact that it was at or below 300 feet at the lighthouse throws no light on conditions prevailing either at the way point change or at the unknown position of the aircraft to seaward some 15 to 18 seconds before impact.

162. The Boeing simulation was prayed in aid to fill in some of the foregoing gaps but as already described it can only determine what could have happened rather than what did happen and was itself deficient in the following respects, namely (i) it did not take account of FADEC, (ii) it postulated a combined speed and ROC which have been found by Witness A and Sir John Day to be unattainable, (iii) it also produced a rotor speed of 91 per cent which was a fairly extreme position differing considerably from that found on the instrument panel and of whose accuracy Mr Cable had doubt, (iv) it produced a groundspeed during the final manoeuvre of 158 knots which exceeded by 11 knots the speed of 147 found in the ground speed indicator, and (v) it hypothesised a final manoeuvre initiated by the crew some 4 seconds before impact, and that prior thereto the aircraft had been under control on a steady course and speed.

163. Both Air Marshals accepted as a matter of fact that the aircraft was under control when the way point change was made and at the moment 4 seconds before impact when the simulation assumed that the final flare was initiated. So far as the way point change is concerned we accept that it is highly unlikely that the crew would have made a way point change if they had thought that they were not in control, but it is possible that if some loose article had jammed the controls during steady flight this would not manifest itself until the controls were moved in order to alter course. Squadron Leader Burke referred to his experience of test flying with control and engine malfunctions when after a period of steady flight dormant faults can appear when a manoeuvre is initiated or engine speed is reduced or increased (Q 705). There is no evidence that such was the case here but equally no evidence that it was not. Alternatively, the movement of the controls to alter course could have precipitated a jam.

164. So far as the aircraft being under control at the moment four seconds before impact is concerned, we do not consider that there is evidence to justify such a conclusion to the required standard of proof. Indeed, apart from the simulation, such evidence as there is - to which reference will shortly be made - suggests the contrary.

165. Both Sir John and Sir William accepted that the possibility of a control jam or engine malfunction could not be disproved. They were adamant however that the pilots were faced with no problem prior to the way point change and that their negligence in reaching that position was not mitigated by anything that might have happened thereafter (QQ 339, 1069-71).

166. If however the finding of negligence at or before the way point change has not been established to the required standard of proof, as we consider to be the case, this proposition does not stand up. The evidence before us was entirely consistent with an intention to alter course and fly VFR to Corran and equally inconsistent with an intention to continue on the same course over the Mull under IFR.

167. The AAIB were not able to exclude the possibility of a control jam given the level of system damage. Nor could they exclude the possibility of pre-impact detachment of the thrust balance spring attachment bracket and other inserts. It will be remembered that this bracket had some three weeks previously detached from the aircraft's thrust/yaw control pallet (see above, para 56). The AAIB were unable to assess the functionality of number 1 DECU owing to gross fire damage. Metallic contamination of the hydraulic system of the integrated lower control actuators found by the AAIB was thought to have been present pre-impact but not to have contributed to the accident; however, the subsequent experience of the US Army and their recommendations (see para 104 above) suggest that such contamination could cause disturbance in the normal operation of those components at the time. DASH runaways have caused temporary loss of control problems as Squadron Leader Burke explained, and UFCMs and false engine failure captions have also afflicted Chinook Mk 2s. Mr Cable accepted that it was possible that there had been an intermittent engine fault which had subsequently reverted to normal before the impact. The problems arising from the newly installed FADEC system had not all been resolved by June 1994; and the Boeing simulation has been shown to have relied to some extent on postulations which are impossible in performance and parameters some of which do not fit with what was found by the AAIB. Can it in these circumstances be said that there is absolutely no doubt whatsoever that it was the voluntary action of the aircrew - including not only both pilots but also MALM Forbes who in our view was probably assisting with the navigation - which caused the aircraft to fly into the hill?

168. Squadron Leader Burke, when asked whether he saw anything significant in the position of the rudder pedals which were at 77 per cent of full travel, replied that it was "an enormous rudder input", unthinkable at high speed (Q 719; see above, para 112). He had also referred to this matter through one of the papers which he had submitted to the Committee in the following terms: "The position of the rudder pedals on impact (almost full left rudder), the high impact speed, and the fully up, or close to fully up, lever position coupled with 100.5% N_R[39] and only 70% torque suggest that an erratic flight path typical of a partial control loss is the most likely of the many guesses as to what was happening in the cloud on ZD 576's last seconds of flight".

169. Witness A commented, "There is absolutely no reason for applying that amount of yaw pedal during forward flight and the only reason I can think of for applying that much yaw pedal would be if the aircraft was becoming extremely difficult to control" (Q 807). He went on to state that the view of the Board that the pedals had been displaced by impact could not be ignored either.

170. Squadron Leader Burke expressed the view that the most likely cause of the accident was a jam of some kind affecting the control of the aircraft, perhaps arising from displaced articles in the broom cupboard (Q 738). A UFCM resulting possibly from a DASH runaway and causing temporary loss of control was also considered by him to be a possibility (Q 739). Such a runaway could cause a temporary increase in rotor speed which the pilot would seek to contain by raising the collective lever thereby forcing the aircraft to climb perhaps unexpectedly into cloud.

171. Witness A considered a control jam to be a strong possibility for the cause of the accident but certainly not an exclusive one (Q 806). He also cited the possibility that a control problem in pitch could have produced oscillations which resulted in the 30 degrees pitch up position in which the aircraft was found (Q 844). Mr Perks expressed the view in his second memorandum that a major mechanical flight controls failure could be an explanation for the difficulty which Boeing experienced in matching their simulation to the data provided.

172. We consider the evidence of Mr Holbrook as to the probability of the pilots being able to see the lower part of the Mull to be of considerable importance - evidence which unfortunately was not before the Air Marshals when they carried out their reviews. For the reasons already given we do not think that the Boeing simulation merits the status which has been accorded to it in the past, and that even if there were some last minute manoeuvre of the aircraft it cannot be said that there was absolutely no doubt whatsoever that it was initiated by pilots who were in control of the aircraft.

173. It follows that the Air Marshals were not justified in concluding that the pilots were in control 4 seconds before impact, or at any time after the way point change. In short it has not been established to the required standard of proof that it was the voluntary action of the pilots which caused the aircraft to fly into the hill.

174. In carrying out our terms of reference, we have considered the justification for the Air Marshals' finding of negligence against the pilots of ZD 576 against the applicable standard of proof, which required "absolutely no doubt whatsoever". In the light of all the evidence before us, and having regard to that standard, we unanimously conclude that the reviewing officers were not justified in finding that negligence on the part of the pilots caused the aircraft to crash.

175. We consider it appropriate to identify those matters to which we have had regard which were not before the Air Marshals when they considered the investigating board's report:

(b) the evidence of Mr Perks, Witness A and Squadron Leader Burke;

(ii) it used a postulated speed and ROC which have been shown to be incompatible; and

(d) the possible effect of contamination in the hydraulic fluid in the integrated lower control actuators, as referred to in the US Army report of June 1997.

176. How could it be that a very experienced crew, having planned to fly VFR, having taken when probably visual with the Mull the appropriate steps to alter course, when there was nothing to prevent them flying northwards within sight of the coast, flew into the Mull? It is as Sir John and Sir William speculatively described "incomprehensible" (Q 342) and "astonishing" (Q 377). We shall never know.

38 Unpublished transcript, p 2352 C. Back

39 Rotor rotational speed. Back