AAIB Bulletin: 7/2007 D-IAFC EW/C2006/09/05

At 27 hrs, the plot was nformed by ATC that the coast was now 23 nm away and they asked hm f the arcraft’s engnes had stopped. He repled sayng that the rght engne had stopped and only the left was workng. Three mnutes later the plot was transferred to the Distress and Diversion (D&D) radio frequency of 2.5 MHz4. On ntal contact wth D&D, the plot nformed the controller that “both engnes are dead now” and that he was plannng to dtch n the sea near to one of three shps that he could see. They advsed hm that a Search and Rescue (SAR) helcopter had been scrambled and, upon enquiring how far away it was, the pilot was nformed 5 mns. He repled that he thought he had only three mnutes before he would have to dtch. At 225:30 hrs, the plot reported he was at ,500 ft amsl, to whch D&D nformed hm he was 9 nm from the coast. Shortly thereafter, they nformed hm radar contact had been lost. He repled that he expected to dtch n about 30 seconds and removed hs headset n preparaton. No further communcatons were receved from the arcraft. Although the arcraft’s Operatng Manual contans a secton on ‘Ditching’, and the plot later stated that he carred out the dtchng check lst from memory. As the arcraft approached the sea, the plot postoned it to fly parallel to the heading of a ship he had seen durng the descent, plannng to dtch near t n order to Footnote 4 Ths was to enable rado contact to be mantaned for as long as possible in the descent, as this frequency had a better low level capability than the en-route frequency. mnmse the tme taken to be rescued. At 00 ft amsl, he opened the emergency hatch located n the rght door. Just before the arcraft touched down, the plot flew the aircraft level, until the stall warning sounded. At a speed of about 80 kt, the tal of the arcraft ht the sea, followed by the fuselage. The arcraft survved the mpact wthout breakng up and, when t came to rest, the plot unstrapped, abandoned the arcraft through the emergency hatch, clmbed onto the rght wng, took off hs shoes and got nto the water. Due to the swell of the sea, water entered the arcraft cabn through the open hatch and t sank after approxmately three mnutes. The accdent occurred at 228 hrs. The dtchng was wtnessed by personnel on the shp, who mmedately launched a lfeboat. Just before ths reached the plot, the SAR helcopter arrved on the scene and wnched hm on board. Once n the helcopter, the plot was found to be sufferng from the effects of immersion in cold water. He was flown to a hospital in Ipswch, Suffolk, where he was found to have suffered a fractured vertebra. Search and Rescue At 26 hrs, D&D were nformed by London ATC that an arcraft had run out of fuel and was gong to dtch n the sea. Control of the SAR operation was subsequently transferred to the Aeronautcal Rescue Control Centre (ARCC) at RAF Knloss, Scotland. As a result, at 240 hrs, two lfeboats, one from Harwch and one from Aldeburgh, Suffolk, were launched. The SAR helcopter from RAF Wattsham, near Ipswch, was scrambled at 226 hrs and, by 245 hrs, had wnched the plot on board. The Harwch lfeboat was then stood down but the Aldeburgh lfeboat contnued to the scene to search for wreckage and to check for possble polluton. On arrval on the scene, at 32 hrs, no evdence of ether was found. 55 © Crown copyrght 2007 AAIB Bulletin: 7/2007 D-IAFC EW/C2006/09/05 Durng D-IAFC’s gldng descent, the crew of a commercal passenger arcraft nbound to London Heathrow Arport (LHR) from Brussels, was asked by ATC f they could offer assistance in locating the aircraft during the final stages of descent before radar contact was lost; at ths pont D-IAFC was 5 nm ahead. Havng assessed that ther arcraft had an excess endurance of approxmately 5 mns, the crew were able to help. ATC cleared the arcraft to descend to FL00 and vectored t n the drecton of D-IAFC. Once above D-IAFC’s poston, the crew became vsual wth the arcraft and watched t dtch approxmately one nautcal mle ahead of a shp. They then nformed ATC that the aircraft was floating and had not broken up, and passed on ts poston before contnung to LHR. Another commercal passenger arcraft, also nbound to LHR, had fuel avalable to reman n the area for two hours and offered further assstance. Ths arcraft proceeded to the poston passed by the prevous arcraft and, wth ATC clearance, descended to 3,000 ft asml. Once overhead, the crew observed a statonary shp, but no arcraft, and passed the descrpton of the shp to ATC. At ths tme, the SAR helcopter was approxmately 5 nm west of the poston. Upon arrvng at the scene the helcopter crew became vsual wth the shp and ts lfeboat before locatng the plot n the water. On hearng ths over the rado, the passenger arcraft contnued to LHR. Survival aspects The plot was fortunate to locate a shp pror to dtchng and to have commercial air traffic in the vicinity to expedite hs rescue. He was dressed n a long sleeved shrt and trousers. At the tme of the accdent the sea temperature was 7ºC and, at ths temperature, wthout approprate survival equipment, he had an expected survival time of just over one hour5. He was n the water for 8 mnutes. Footnote 5 Data from the Royal Ar Force. Weather An aftercast covering the duration of the flight was obtained from the Met Office. This stated that the wind at FL00, was from 230º at 25 to 30 kt. At the dtchng locaton there was scattered cumulus cloud at 3,000 ft amsl, the surface wnd was from 270º at 5 to 20 kt and the vsblty was n excess of 0 km. Addtonally, the sea swell was 0.5 to 1 m with a period of five seconds. The average track for the flight was 260º. The wind at FL00 would have gven a headwnd component of approxmately 22 kt. Pilot’s comments The plot was ntervewed by the AAIB n hosptal the day after the accdent. He was also ntervewed by the German Federal Bureau of Arcraft Accdents Investgaton - Bundesstelle für Flugunfalluntersuchung (BFU) - after hs return to Germany. He stated that he had planned the flight on the basis of an average fuel consumpton rate of 04 ltres/hr. He added that there was a ‘Digi-Flow’ meter fitted to the arcraft but that t had not been servceable for “some tme”. Because the manufacturng company had ceased tradng t had not been possble to obtan any spare parts. Addtonally, he stated that the arcraft’s fuel gauges were “pretty naccurate” and so he usually trusted hs own calculations. The aircraft was not fitted with low fuel quantity warning lights. The plot also stated that there was no way of drectly establshng the contents of the fuel tanks on the ground, due to the dhedral of the wngs and the fact that the refuellng caps are located close to the wng tps. Also, there were no dip sticks fitted. The only indication of fuel quantity on board the aircraft were the readings from the fuel gauges located close to the cockpit floor. He added 56 © Crown copyrght 2007 AAIB Bulletin: 7/2007 D-IAFC EW/C2006/09/05 that he had not checked these durng the cruse untl the right engine started to run roughly, as they were difficult to see as well as beng naccurate6. The aircraft did not carry a liferaft, but two lifejackets were located n the rear of the cabn. The plot stated he had not worn one as he had not remembered that they were there. Aircraft fuel system The Cessna T303 has one fuel tank n each wng, wth a total useable fuel capacty of 579 ltres (98 lb). There are two fuel gauges, one for each tank, marked left and right, located at the rear of the centre console just above floor level, Figure 1. The lower sector of the fuel level scale s marked n red, to ndcate when only unusable fuel s remanng n each tank; unusable fuel is quoted in the manual as 12 lb. A separate low fuel level warnng system, ncorporatng two warnng lghts on the nstrument panel, was avalable as a customer opton on the T303, but ths was not fitted to D-IAFC. Each light illuminates when the fuel remanng n ts respectve tank reduces to 38 ltres (60 lb) or less. The aircraft had been modified to incorporate a ‘Digi-Flow’ digital fuel flow meter, but this had been noperatve for some tme. Figure 1 Footnote 6 Another operator of ths arcraft type has also reported that the fuel gauges are naccurate. 57 © Crown copyrght 2007 AAIB Bulletin: 7/2007 D-IAFC EW/C2006/09/05 Company information The arcraft’s operatng company had an Ar Operator’s Certificate (AOC) issued by the German authorities. Ths gave approval for the company to transport cargo and passengers. The AOC was vald at the tme of the accdent. The company’s flight planning documentation states that 25 lb of fuel should be allowed for durng start-up and taxi, and a fuel flow rate of 174 lb/hr should be used for calculatng the trp fuel. Ths ncludes the fuel used durng clmb and descent. Addtonally, there are columns to add to the trp fuel for contngences such as diversion to an alternate airfield and 45 mins holding. Aircraft performance The flight planned track from EDVE to EGTK was 468 nm and the head wnd component was approxmately 22 kt. When ths data s plotted on the Fuel and Tme Required (60% power) graph in the aircraft’s Operating Manual, a flight time of approximately 3 hrs 21 minutes and fuel required of approximately 585 lb is predicted. Ths ncludes fuel for engne start, tax, takeoff, normal clmb, descent and 45 mnute reserve. The tme required includes that for normal climb and descent, all of which equates to a fuel consumption rate of 74 lb/hr. The Holdng Tme table states that 80 lb of fuel is required for 45 minutes holding at 45% power. Fuel plan All pre-flight and in-flight paperwork was on-board the arcraft and was not recovered. However, the refuel certificate for D-IAFC was obtained from LZIB, and this showed that 200 ltres (37 lb) of fuel was uplfted pror to the arcraft returnng to EDVE. Table  compares the plot’s assumed fuel load and endurance wth the (AAIB) estimated fuel load and endurance, for the accident flight and the two previous flights. The estimated departure fuel figures were derived with reference to the aircraft’s Depart EDVE Arrve LZIB Depart LZIB Arrve EDVE Depart EDVE Arrve EGTK Plot’s assumed fuel on board (lb) 98 (a) 406 98 (b) 548 548 (d) 3 AAIB estmated fuel on board (lb) 98 (a) 406 723 (c) 353 353 (d) -07 Plot’s assumed endurance 5 hr 30 mn --- 5 hr 30 mn --- 3 hr 30 mn  hr AAIB estmated endurance 5 hr 8 mn --- 4 hr 0 mn ---  hr 53 mn -37 mn Flght tme 2 hr 48 mn ---  hr 59 mn --- 2 h 30 mn (e) --- Table 1 Pilot’s assumed and AAIB estimated fuel figures and endurances Notes: (a) Arcraft departed EDVE wth full tanks (b) Plot assumed the tanks were full pror to departng LZIB (c) Fuel records show only 200 ltres (37 lb) uplfted at LZIB, gvng 723 lb on departure (d) Arcraft was not refuelled agan pror to departng for EGTK (e) Flght plan estmated elapsed tme 58 © Crown copyrght 2007 AAIB Bulletin: 7/2007 D-IAFC EW/C2006/09/05 refuelling records. The arrival fuel figures and the endurances were calculated usng the fuel consumpton figures quoted in the operator’s flight planning logs. The estimated endurance figures are to fuel exhaustion and make no allowance for reserve and alternate fuel. Aircraft’s weight The MTOW of D-IAFC was 5,50 lb. When ntervewed after the accdent, the plot stated to the AAIB that there was 200 lb (90.9 kg) of cargo on board the arcraft. When ntervewed by the BFU, on hs return to Germany, he amended ths to 540 lb (242.5 kg). He later explaned that he had expected a cargo of ths weght and, as ths would have placed the arcraft close to ts MTOW, he decded not to refuel pror to departng for EGTK. Documents recovered from the suppler of the cargo ndcate that ts total weght was actually 262 lb (119 kg). As this flight was an ad hoc charter flight, the cargo suppler s convnced there was no addtonal cargo on board. Furthermore, the operatng company’s insurer has not been notified of any other loss other than the 262 lb of documented cargo. The followng tables compare the arcraft’s takeoff weight for its flight from EDVE to EGTK (using the two dfferent cargo weghts) wth the maxmum permtted takeoff weght, wth the plot’s assumed fuel on board (Table 2) and the AAIB estmated fuel on board (Table 3). These tables use a basc arcraft weght of 3,654 lb, a pilot weight of 187 lb and the fuel figures from Table 1; all weghts are n pounds. 540 lb cargo 262 lb cargo Arcraft’s weght wth plot 3,84 3,84 Plot’s assumed fuel 573 573 Fuel used durng start up/tax -25 -25 Plot’s assumed takeoff weght 4,929 4,65 Fuel weght avalable, but not used, lmted by the MTOW 22 499 Additional flight time if fuel weight available had been used, assumng fuel consumpton at 74 lb/hr  hrs 6 mns 2 hrs 52 mns 540 lb cargo 262 lb cargo Arcraft’s weght wth plot 3,84 3,84 Plot’s assumed fuel 378 378 Fuel used durng start up/tax -25 -25 Plot’s assumed takeoff weght 4,734 4,456 Fuel weght avalable, but not used, lmted by the MTOW 45 694 Additional flight time if fuel weight available had been used, assumng fuel consumpton at 74 lb/hr 2 hrs 23 mns 3 hrs 59 mns Table 2 Table 3 59 © Crown copyrght 2007 AAIB Bulletin: 7/2007 D-IAFC EW/C2006/09/05 Joint Aviation Requirements - Operations (JAR-OPS) 1 ‘JAR-OPS 1 Subpart B - General, Appendix 1 to JAR-OPS 1.005(a) paragraph (12) 1.255 Fuel Policy’ states: ‘(ii) For A to B Flights – An operator shall ensure that the pre-flight calculation of usable fuel required for a flight includes; (A) Taxi fuel - Fuel consumed before take-off, if significant; and (B) Trip fuel (Fuel to reach the destination); and (C) Reserve fuel - (1) Contingency fuel - Fuel that is not less than 5% of the planned trip fuel or, in the event of in-flight re-planning, 5% of the trip fuel for the remainder of the flight; and (2) Final reserve fuel - Fuel to fly for an additional period of 45 minutes (piston engines) or 30 minutes (turbine engines); and (D) Alternate fuel - Fuel to reach the destination alternate via the destination, if a destination alternate is required (E) Extra fuel – Fuel that the commander may require in addition to that required under subparagraphs (A) – (D) above.’ Joint Aviation Requirement - Flight Crew Licensing (JAR – FCL) 1 ‘JAR – FCL 1, subpart A – General Requirements’, states: ‘JAR–FCL 1.060 Curtailment of privileges of licence holders aged 60 years or more (See Appendix 1 to JARFCL 1.060) (a) Age 60–64. The holder of a pilot licence who has attained the age of 60 years shall not act as a pilot of an aircraft engaged in commercial air transport [CAT] operations except: (1) as a member of a multi-pilot crew and provided that, (2) such holder is the only pilot in the flight crew who has attained age 60. (b) Age 65. The holder of a pilot licence who has attained the age of 65 years shall not act as a pilot of an aircraft engaged in commercial air transport operations. (c) Any national variant to the requirements in (a) and (b) above are given in Appendix 1 to JAR-FCL 1.060.’ There are no German national variants to the requirements n (a) and (b) n Appendx . Analysis Licensing The plot was 67 years old at the tme of the accdent. He held a Commercal Plot’s Lcence. Ths was vald until 18 June 2007, but this only permitted him to fly in Germany, which he also did as a flying instructor. His 60 © Crown copyright 2007 AAIB Bulletin: 7/2007 D-IAFC EW/C2006/09/05 licence, however, was not valid for him to fly as a pilot of an aircraft engaged in CAT operations, as he was over 65 years old. Fuel planning The fuel flow rate (of 104 litres/hr) that the pilot stated he used in his calculations, equates to 164 lb/hr. With the aircraft’s fuel tanks full, this equates to a flying time (to fuel exhaustion) of 5 hrs 26 mins, allowing 25 lb for start up and taxi. This figure is effectively the same as the pilot’s assumed full tanks endurance of 5 hr and 30 mins, if slightly optimistic. Although the pilot believed that he had filled the aircraft’s fuel tanks to full at LZIB, it was estimated that they may have been less than full by some 195 lb (123 litres). If this were so, and the tanks had been refuelled to full, the aircraft should have been able to fly for, approximately, an additional one hour seven minutes before running out of fuel. The fuel on board prior to taking off from EDVE, estimated by the AAIB, was 353 lb. This equates to an endurance of two hours. The fuel on-board became exhausted after 1 hr 53 mins and 314 nm, approximately 160 track miles from EGTK. At a ground speed of 160 kt in the cruise, the aircraft was, therefore, approximately one hour short of its destination. Thus, although the aircraft should have been able to reach EGTK, had the fuel tanks been full on departure from LZIB, it would not have had any reserve or fuel to fly to an alternate airfield. If the aircraft’s fuel gauges were accurately indicating three-quarters full prior to takeoff for EGTK, there should have been approximately 688 lb (434 litres) in the fuel tanks. This equates to an endurance of around 3 hr 50 minutes to fuel exhaustion, using the operating company’s fuel burn figure of 174 lb/hr, allowing 25 lb for start up and taxi. The flight planned route was 468 nm and, at a ground speed of 160 kt, this should have taken 2 hr 55 minutes. In order to fly to EGTK in the flight planned time of 2 hrs 30 minutes, the aircraft would have had to fly at a groundspeed of 187 kt; this is an unrealistic speed for the aircraft with the headwind at the time, given that average groundspeed of D-IAFC from takeoff to fuel exhaustion was 149 kt. At this average speed, it would have taken 3 hr and 8 mins to fly from EDVE to EGTK, and would have required 545 lb of fuel. The AAIB estimate of fuel on-board at the start of the flight, from Table 1, was 353 lb. It appears the pilot may have thought the weight of his cargo was 540 lb, not its actual weight of 262 lb. Using the pilot’s assumed weight of the cargo and his assumed fuel on-board (573 lb, see Table 2) prior to departure, this would have allowed him to load an additional 221 lb of fuel to take the aircraft up to its MTOW. This would have given the aircraft an additional 1 hr 16 min of flying time to fuel exhaustion. Thus, flying at an average groundspeed for the trip of 149 kt, the aircraft would probably have reached EGTK, but would not have had any alternate or reserve fuel. Using the AAIB estimated fuel quantities from Table 3, and the pilot’s assumed weight of the cargo, he should have been able to load an additional 415 lb of fuel without exceeding the aircraft’s MTOW. This would have given an additional 2 hrs 23 min flying time to fuel exhaustion. Thus, the aircraft should have been able to reach EGTK with sufficient fuel remaining to reach an alternate airfield, plus reserve fuel. Survivability Although the pilot’s survival time, at a water temperature of 17ºC, was theoretically predicted to have been about one hour, his useful conscious time would have been significantly less. It was fortunate that the SAR 6 © Crown copyrght 2007 AAIB Bulletin: 7/2007 D-IAFC EW/C2006/09/05 helcopter and the rescue boat from the nearby shp were on scene so quickly as, after only 18 minutes in the water, he was dagnosed as already sufferng from the effects of cold. It was also fortunate that the sea state was slght, gven that the plot was not wearng a life jacket. Had it been greater, his chances of surviving n the open sea long enough to be rescued would have been much reduced. Although the arcraft was not carryng a dngy, had the plot been able to deploy and board one, hs survval tme would have been greatly ncreased. It would, therefore, seem prudent for sngle and multi-engine aircraft not required to carry a dingy, to do so when transtng large areas of water, and for plots to be traned n ther deployment and operaton. The CAA have published Safety Sense Leaflet 21, titled ‘Ditching’, whch contans comprehensve nformaton on this subject. Conclusion The accdent occurred as a result of the arcraft runnng out of fuel approxmately 60 nm short of ts destnaton. Although the wreckage of the arcraft was not recovered, all the evdence suggests that ths occurred due to insufficient fuel being on-board the aircraft prior to departure, rather than because of a techncal problem. The pilot’s lack of awareness of the fuel quantity and the actual weght of the cargo on board D-IAFC pror to takeoff, are considered to have been significant causal factors n the accdent. A contrbutory factor was that the plot dd not montor the reportedly ‘unrelable’ fuel gauges, thus mssng a chance to notce the arcraft’s low fuel state and divert to a suitable airfield before the stuaton became crtcal.