D J & N a Shanks Limited v Heli Support New Zealand Limited
[2019] NZHC 1615
•11 July 2019
IN THE HIGH COURT OF NEW ZEALAND INVERCARGILL REGISTRY
I TE KŌTI MATUA O AOTEAROA WAIHŌPAI ROHE
CIV-2017-425-67
[2019] NZHC 1615
BETWEEN D J & N A SHANKS LIMITED
Plaintiff
AND
HELI SUPPORT NEW ZEALAND LIMITED
Defendant
Hearing: 6 - 10 May 2019 Counsel:
P Davies and S F W Learmonth for Plaintiff N S Gedye QC for Defendant
Judgment:
11 July 2019
JUDGMENT OF THOMAS J
Table of contents
Introduction[1]
Background[7]
The Accident[11]
Post-Accident inspections[18]
First inspection [18]
Second inspection [27]
Third inspection [33]
The Helicopter[34]
Mr Hobday’s investigations [41]
Air intake system[42]
Fuel system[46]
Fuel system diagram[48]
Heli Support’s work on the Helicopter [54]
Was the B-nut loosened during the work?[56]
Pre-flight checks[63]
Written records[68]
Check flight[73]
Opinion evidence[75]
Sealed union but loose B-nut [82]
D J & N A SHANKS LIMITED v HELI SUPPORT NEW ZEALAND LIMITED [2019] NZHC 1615 [11 July 2019]
Hobday clear tube test[97]
Buick test 24 June 2013[111]
Could snow ingestion have caused the flame-out? [116]
Wet Michigan filter[133]
Heli Support’s submissions[138]
Shanks’ submissions[149]
Legal principles[155]
Factual findings[157]
Result[182]
Introduction
[1] On 27 May 2013, Sean Mullally and Sarah de Reeper were flying a McDonnell Douglas (Hughes) MD369 (now MD500) series helicopter, registered ZK-HNA (the Helicopter), from the workshop of the defendant, Heli Support New Zealand Limited (Heli Support), to Mirror Lakes in the Eglinton Valley, Fiordland National Park. They flew through a light snow shower lasting about a minute. During the descent, when the Helicopter was approximately 150 feet above the ground, the engine flamed out and stopped (the Accident).1 Mr Mullally managed to control the Helicopter so that it touched down to land, although it slid along the ground for 10 to 15 metres. During this slide, one of its skids caught on a tussock, tipping the Helicopter forward. Although Mr Mullally was able to bring the Helicopter to a stop, the main rotor blades severed the tail boom, causing damage.
[2]The question addressed in this decision is why the engine stopped.
[3] DJ & NA Shanks Limited (Shanks), sues Heli Support in negligence, claiming as damages the cost of the repairs to the Helicopter. The factual issues for determination are whether Shanks has proved on the balance of probabilities that:
(a)after installing the engine in the Helicopter, Heli Support released it to service on 27 May 2013 with the AN818 alloy tube nut, which secured the outlet elbow joint between the fuel filter and fuel line leading towards the engine (the B-nut), loose; and
1 An engine flames out when the flame in the combustion chamber is extinguished.
(b)air ingestion from the loose B-nut caused the engine failure.
[4]It is not in dispute that:
(a)Heli Support owed a duty of care to Shanks (as owner of the Helicopter) when it worked on the Helicopter between 24 and 27 May 2013;
(b)the engine suddenly failed close to landing;
(c)the B-nut was one-quarter turn loose when inspected after the Accident on 29 May 2013;
(d)otherwise there was no defect in the fuel system which caused engine failure;
(e)some liquid was observed after landing, sufficient to drain to the ground;
(f)a properly torqued B-nut could not work itself loose either during flight or when the Helicopter was being transported after the Accident to Te Anau or Queenstown;
(g)the cost of repair of the damage to the Helicopter was USD 390,643.91.
[5]Heli Support does not accept that:
(a)Anyone at Heli Support loosened or undid the B-nut during the work Heli Support carried out on the Helicopter between 24 and 27 May 2013.
(b)There was any deficiency in Heli Support’s checks of the fuel system prior to release of the Helicopter to service. Instead, it says the tightness of the B-nut was checked and a blue dot placed on it to record that fact.
(c)The B-nut was loose when the Helicopter left Wanaka.
(d)The loose B-nut caused engine failure.
[6] Heli Support disputes that air ingestion caused the engine failure and contends it is more likely than not that the Accident was caused by snow ingestion.
Background
[7] Shanks owned the Helicopter and leased it to Milford Helicopters Limited (MHL). MHL provides helicopter services for lifting and supply operations, passenger transport, aerial photography, avalanche control and scenic flights.
[8] Shanks had used Heli Support for many years to carry out maintenance, repairs and servicing on all its helicopters.
[9] In April 2013, when Mr Mullally was piloting the Helicopter, a length of rubber sealing strip became partially ingested by the compressor section of the engine (the April incident). Heli Support was engaged to remove the engine for inspection and repair. The engine was sent to Oceania Aviation Limited (an approved overhaul facility), which carried out the repair. Heli Support installed a rental engine as an interim measure to allow the Helicopter to continue in operation.
[10] On Friday 24 May 2013, the Helicopter was returned to Heli Support for removal of the rental engine and reinstallation of the original engine. It was to be collected by Shanks on Monday 27 May 2013.
The Accident
[11] Mr Mullally, who has held a commercial pilot’s licence since 1991 and at the time of the accident had 4,575 hours of flying time in the Helicopter, was the main pilot during the Accident. He had flown the Helicopter since it was leased by MHL in September 2011, mostly on tourist and venison recovery work.
[12] His co-pilot, Ms de Reeper, has held a New Zealand commercial pilot’s licence since 2004 and a Canadian commercial pilot’s licence since 2009. At the time of the
Accident, she had around 700 hours’ flying time, of which about 100 hours was in the Helicopter. She was keen to accompany Mr Mullally to pick up the Helicopter on 27 May 2013 because the weather was forecast to deteriorate and she wanted to learn from Mr Mullally about alternative weather flying routes back from Wanaka. There was no snow on the ground when Mr Mullally and Ms de Reeper drove to Wanaka that morning.
[13] When Mr Mullally and Ms de Reeper arrived at Heli Support, the Helicopter was still in the hangar and Jason Buick, shareholder and director of Heli Support, was working in the engine compartment. This meant Mr Mullally and Ms de Reeper had to wait for the repairs to be completed.
[14] When they eventually left Wanaka at just after 11.00 am, there was fresh snow on the hills and low cloud. Mr Mullally turned on the anti-ice and pointed out to Ms de Reeper how the turbine outlet temperature (TOT) gauge increased by about 15 to 20 degrees. As they were passing Elfin Bay, a light snow shower appeared ahead and Mr Mullally opened the alternate air door (described in more detail below). He and Ms de Reeper discussed this, Mr Mullally explaining to her it was best to open the alternative air door before flying through falling snow. They both said the snow shower was brief and did not cause any problems. Ms de Reeper was very conscious of the possibility of snow build-up and paid attention during the flight, enabling her to be sure that there was no snow build-up during the flight.
[15] The Helicopter performed normally and satisfactorily for the duration of the flight until the engine failed on the approach to land, when the Helicopter was in a turn about 150 feet above ground level. The engine failure was described as a smooth rundown, similar to what would occur in a normal shutdown procedure. The engine did not make any noise suggestive of a mechanical failure. There was adequate fuel on board which had been checked and found free of contamination prior to departure from Wanaka. No adverse weather had been encountered in the 15-minute period immediately prior to the engine failure.
[16] Four people observed liquid inside the bottom of the engine after the Accident; both pilots, Daryl Palmer, the intended passenger, and David Shanks, director of
Shanks.2 Both Mr Palmer and Mr Shanks thought it was fuel and commented on the strong fuel smell. Mr Mullally recalled seeing a clear liquid but did not know whether it was water, fuel or both. Ms de Reeper described clear liquid inside the engine bay and an oily looking greenish yellow mark on the snow under the engine bay. She also noticed a slow drip from the engine bay drain with what seemed to be water, although she noted a faint smell of fuel. She said, later on, there was a noticeable amount of grease or fuel on the snow.
[17] Mr Shanks, Mr Mullally and Mr Palmer left the scene while Ms de Reeper remained. Mr Shanks contacted both the Civil Aviation Authority and Mr Buick of Heli Support. Having been informed by the Civil Aviation Authority that the wreckage had to be made secure, Mr Shanks and Mr Mullally flew back to the Accident site. Covers and tie-downs were placed on the Helicopter, which was then transported and stored in a shed at Mr Shanks’ home, probably on the same day.
Post-Accident inspections
First inspection
[18] An inspection took place on 29 May 2013, two days after the Accident. It was attended by Mr Shanks, Mr Mullally, Mr Buick and Mr Anderson from Heli Support, Aaron Shaw and a colleague from Oceania Aviation, an insurance broker and Peter Webb, an aviation and engineering loss adjuster.
[19] Mr Shaw, then the chief engineer at Oceania Aviation, said in evidence that one of the first things he noticed was that the fuel line to the filter head was loose.
[20] Mr Webb gave evidence as an expert witness. He said, when the engine compartment of the Helicopter was opened, Mr Buick and Mr Shaw very quickly focused on the fuel filter. They pointed out that the B-nut was loose and the blue stripe (the torque stripe) that had at some stage been applied to one flat on the nut was no longer visible. He explained that, as a method of detecting when threaded fasteners have worked loose, aircraft engineers commonly employ a technique known as torque
2 Mr Shanks arrived at the scene shortly after the Accident.
striping. This involves drawing a line in ink or paint across an assembly so that any subsequent misalignment of the stripe signals that the fastener has moved. He noted some engineers prefer simply to apply a mark to the visible face of a nut or bolt with a marker pen.
[21] Mr Webb and Mr Mullally both observed more than one blue mark on the B-nut. Mr Webb was surprised at this, describing the mark as more of a tick or confirmation that the joint had been tightened rather than a true torque stripe. He did not know whether the marks were put there at the time of the work just completed by Heli Support prior to the Accident or at some earlier time.
[22] Mr Webb described the blue mark on the B-nut being in the five o’clock position. That is one and a half flats, or one-quarter turn, loose from the position in which it would clamp the pipe joint firmly. Mr Buick pointed out that what he described as the blue witness dot was one quarter turn loose from the position he said it was in when he checked it on 27 May. He said no other nuts had their blue dots displaced.
[23] Mr Shanks and Mr Mullally recalled Mr Buick telling Mr Shanks that Heli Support must have failed to tighten the B-nut. Mr Buick denied saying this.
[24] Mr Buick considered that both Mr Shanks and Mr Mullally behaved uncharacteristically towards him after the Accident. Mr Shanks agreed that he and Mr Buick had always got on well, but explained he was extremely angry with Mr Buick and Heli Support after the Accident. He also blamed Mr Buick for the April incident, which he believed was caused by Heli Support’s failure to repair the foam seal. Mr Buick categorically denied any responsibility for the April incident.
[25] Mr Buick had visited Mr Shanks’ premises the day before, on 28 May, and viewed the Helicopter from a distance. He noted Mr Shanks did not then mention either the loose B-nut or the fuel wetness around the top of the Michigan filter which he commented on at the first inspection. Mr Buick did, however, notice the top of the fuel filter was wet at the first inspection.
[26] The Oceania Aviation personnel examined the air inlet to the compressor section of the engine and noted it was free of any foreign matter or obvious obstruction. The checks they performed confirmed to Mr Webb that internal engine damage may not have caused the Accident and it was more likely the engine had stopped because of interruption to the fuel supply.
Second inspection
[27] On 5 June 2013, the Helicopter was transferred to Oceania Aviation’s premises in Queenstown for a full engineering inspection, which took place on 6 June 2013. John Hobday, a licenced aircraft maintenance engineer, and Derek Emeny, an aviation loss surveyor, attended.
[28] Mr Hobday was engaged in June 2013 on behalf of MHL to conduct a technical investigation to determine the cause of the Accident and was the main expert witness for Shanks. He has been a licenced aircraft maintenance engineer since 1977 and is experienced and qualified in the maintenance of the Allison 250 engine and the Hughes MD500 series helicopter.
[29] Mr Hobday did not observe any damage or areas of concern which may have contributed to the sudden shutdown of the engine mid-flight. He concluded the only way the engine could have failed was if it had been starved of fuel, air or both.
[30] Mr Emeny was engaged by Heli Support’s insurers to advise on the claim. He is a licenced aircraft engineer familiar with the MD369 series fuel system and gave evidence as an expert.
[31] On 6 June the B-nut was tightened to the correct position. The blue witness dot was then in the correct place. The start pump was activated for three minutes and no fuel leaks were observed. The B-nut was loosened one-quarter turn and the start pump again activated but without the engine running. There was an immediate substantial leakage of fuel from the loose B-nut.
[32] That day, Mr Hobday also conducted what he referred to as a “clear tube test”. This involved attaching a clear tube to the engine to show what was in the fuel line
between the fuel pump and the fuel nozzle at the time of the Accident. I discuss this test in more detail below.
Third inspection
[33] A third inspection took place on 26 June 2013, attended by Mr Shanks, Mr Buick, Mr Emeny, Mr Shaw, Mr Webb and a representative of the Civil Aviation Authority. The purpose was to check the fuel tank to ensure it had not caused or contributed to the Accident. Mr Shaw carried out this work in accordance with a Technical Directive issued by Mr Hobday. No issues were identified with the fuel tank.
The Helicopter3
[34] The Helicopter is a five-seat light utility machine powered by a single Rolls Royce Allison 250–C20B gas turbine. The engine consists of a front mounted air compressor section that provides high pressure to the combustion chamber at the rear where the gas flow is reversed and the high-pressure gases produced by the combustion of fuel and air expand through the turbine disks, causing them to rotate.
[35] Fuel is introduced through a nozzle in the rear end of the combustor and the turbine runs as long as an adequate supply of fuel and compressed air is maintained.
[36] Fuel is piped through the engine compartment firewall to a filter and then to the fuel control unit on the turbine. An electric fuel pump (the start pump) is fitted in the fuel tank, typically used only when the engine is being started. Once the turbine is running, fuel supply is maintained by a small suction pump on the engine. Reliance on a suction system to maintain fuel supply means that any leaks in the fuel pipework can allow air to be sucked into the fuel.
[37] Intake air for the engine is drawn in and ducted through an air filter and a plenum chamber into the mouth of the compressor section of the turbine. The
3 The description of the Helicopter is taken from Mr Hobday’s evidence. There was no challenge to this aspect of his evidence.
incoming air is filtered by a particle separator which forces suspended solids to be flung out and ducted away.
[38] An accumulation of wet snow against the face of the particle separator is well-recognised as a cause of air starvation in this type of helicopter. To avoid this, the plenum chamber (through which air enters the engine) is fitted with an alternate air door. This is a simple hinged trap door in the side of the chamber on the top of the helicopter cabin which houses the air intake, particle separator and plenum chamber. The pilot can open the alternate air door by pulling a handle and the door then stays open until the helicopter lands and it can be re-latched. No air filtration is provided when the alternate air door is open.
[39] The Helicopter is fitted with systems which detect engine stoppage and alert the pilot by way of warning light on the instrument panel and a horn. An automatic re-ignition system senses activation of the engine-out warnings and activates the ignitor to relight the combustor. Attempting a relight is not recommended when the aircraft is at low altitude because the pilot has little time to recognise the problem and take appropriate action. The pilot flight manual for the MD500 indicates that, in the event of engine failure at between 12 and 500 feet above ground level, the pilot should concentrate on making an auto-rotational landing.
[40] The Helicopter was fitted with a Michigan filter which allows the Helicopter to be operated in low temperatures by filtering ice crystals (and other foreign matter) out of the fuel. Fuel passes through the filter element, leaving unwanted solids on the outside. The filter is fitted with inlet and outlet pipe connections. The outlet fitting on the Helicopter was a 90-degree elbow secured with the B-nut. The B-nut must be securely tightened on the elbow fitting to ensure that the joint will not leak fuel or allow the ingress of air. It was this B-nut that was loose after the Accident.
Mr Hobday’s investigations
[41] Mr Hobday conducted an investigation to determine the cause of the Accident as he explained when he gave evidence.
Air intake system
[42] Mr Hobday said that the most reliable warning sign of a restriction in the air intake system is an increase in the outlet temperature at the gas turbine, TOT. He explained that, given the low tolerances present in the fuel/air ratio of the turbine, any obstruction of the air intake and subsequent reduction in the air across the turbine will result in an instant spike in the TOT.
[43] Mr Shaw obtained the data from the memory of the TOT gauge. It did not record any over-temperature events on 27 May.
[44] In Mr Hobday’s opinion, while a restriction of the air intake will cause a flame-out, this will not be instantaneous and there will be prior warnings such as the running of a high TOT or power loss.
[45] Mr Hobday concluded that the Accident was unlikely to have been caused by an issue with the air supply because:
(a)there were no over-temperature events recorded for the flight;
(b)there was no obstruction or evidence of an obstruction having existed in the air intake;
(c)the pilots did not describe a high TOT or loss of power; and
(d)the alternate air door was opened some time prior to the Accident.
Fuel system
[46] During his mechanical inspection, Mr Hobday found no evidence of any defects in the engine or fuel system apart from the loose B-nut and fuel line joint.
[47] Mr Hobday carried out testing on the engine of the Helicopter on 6 June, the “clear tube test” discussed in some detail at [97] to [110] below. The test showed, to him, that a small amount of air had remained in the fuel line following the activation of the start pump when the B-nut was loose, establishing, in his opinion, that air can
be ingested into the system as a result of a loose B-nut. He accepted this was not an exact replica of the environment when the Accident occurred because, at the time of the Accident, the suction pump was drawing fuel through the system rather than being pushed by the start pump. The suction pump runs at a much higher pressure than the start pump and would place more pressure on a loose joint, he said. With the throttle still open, fuel would continue to run through the system and eject into the combustor.
Fuel system diagram
[48] Mr Hobday noted that the fuel system check valve stops air escaping when it is closed but does not close until fuel pressure drops to a low pressure. Mr Hobday explained that during the Accident, the engine did not stop instantly. It ran down as it would in a normal shutdown, meaning air would have bled out of the fuel line. In Mr Hobday’s opinion, it was therefore not surprising that small amounts of air only were present in the fuel line. It would, however, have been sufficient to cause the engine to shut down, in his opinion.
[49] The engine failure occurred when the Helicopter was in a turn after a long period of fairly straight inter-level flight. Mr Hobday concluded that the reduction in forward speed during the landing approach would have caused a reduction in the tank pressure and increased the suction asserted by the mechanical suction pump. As the
Helicopter slowed down, different forces would have been exerted on the B-nut and joint. The change in forces would, in his opinion, have allowed an already loose joint to leak air.
[50] Mr Hobday said, when a turbine is starved of air, it will falter and flame out or run at a high TOT. However, if starved of fuel, it will flame out instantly and wind down in the same way as in a normal shutdown. Mr Hobday said an ingestion of air large enough to cause a bubble within the fuel line is a well-known and documented cause of flame-outs with the Allison 250 engine.
[51] Mr Hobday said that a small quantity of fuel remaining in the fuel line to the engine will drain back when the engine stops and leak from any loose union in the pipework. He noted reports from the witnesses of a fuel leak from the engine compartment after landing, fuel collecting in the lower part of the engine compartment doors and a smell of fuel.
[52] In summary, Mr Hobday’s post-Accident engineering investigation found no defect in any of the major components of the Helicopter drive train and no defects in the fuel system except for the loose B-nut. He found no evidence of pilot error, no evidence the weather affected the pilot’s ability to fly safely and no evidence of any restriction or blockage of the air intake. This led to his conclusion that the Accident was caused by air entering the fuel system, most likely through the loose B-nut.
[53] Given the B-nut was one-quarter turn loose when inspected after the Accident on 29 May 2013, and it was agreed that a properly torqued B-nut could not work itself loose either during flight or when the Helicopter was being transported after the Accident, the question is whether the B-nut was loose when the Helicopter left Heli Support on 27 May 2013.
Heli Support’s work on the Helicopter
[54] Mr Buick is a licenced aircraft maintenance engineer, certified under the Civil Aviation Rules and has some 30 years’ experience in maintaining helicopters. He described the work undertaken by Heli Support beginning when the Helicopter arrived
at Heli Support’s Wanaka maintenance facility for the removal of rental engine modules and reinstallation of the Helicopter’s own compressor and turbine.
[55] Two main areas of Mr Buick’s evidence assumed significance during the trial. The first was whether the B-nut was loosened or interfered with during the work on the Helicopter. The second was whether, had the B-nut been loose on completion of the work, that would have become obvious during the pre-flight checks carried out by Heli Support.
Was the B-nut loosened during the work?
[56] Mr Buick, assisted by two employee engineers, carried out most of the work on the Helicopter between 24 and 27 May 2013. Daegal Schmidt, a licenced aircraft engineer employed by Heli Support, also had some involvement. The work required removal of the fuel control unit and associated plumbing, and the outer combustion case. The engine was reassembled and then reinstalled in the air frame. Although four people were involved in the work on the Helicopter, Mr Buick said none of the others would have had any occasion to touch the B-nut as they were not involved in the engine replacement or checks.
[57] Mr Buick’s evidence was that the B-nut was not loosened or undone, or otherwise worked on beyond checking its tightness. He said the engine was removed and replaced without interfering with that part of the fuel system. Mr Buick said the disconnection of the fuel line was done at the engine side and not at the filter side (meaning the B-nut was not interfered with). Mr Hobday and Mr Shaw agreed that, when the engine is removed, it is standard practice and logical to disconnect the fuel line at the engine side rather than the filter side. Mr Buick accepted that, if the fuel line is disconnected at the engine side, it falls down and typically he tucks it behind a vertical rod to keep it out of the way. Mr Schmidt did not agree that tucking it away like this was best practice.
[58] In Mr Hobday’s opinion, however, there were still reasons for the B-nut to have been loose:
(a)the fuel line was undone by someone who did not know what they were doing;
(b)the fuel line had been disturbed at some time in the past and under-torqued, allowing it to be further loosened when it was disconnected from the engine and moved to one side; or
(c)the fuel line forward end was loosened for some maintenance activity not stated.
[59] Mr Mullally said he has seen the engines in MD500 series helicopters being changed many times and said the B-nut is always undone because otherwise the fuel line gets in the way. He recalled seeing staff from Heli Support doing exactly that when they changed the engine in April 2013 after the April incident.
[60] Mr Shaw confirmed he has on occasion seen the fuel line disconnected at the fuel filter end.
[61] In any event, Mr Buick said that, once the work was completed, he did check the tightness of the B-nut by applying a wrench to it. He agreed that, after an engine change, approximately 20 nuts of about five different sizes, requiring different sized spanners, need checking. Mr Buick described the process he followed. He said he placed a blue witness dot on the face of each nut with a blue marker pen. Then, after checking the nut was correctly tightened with a wrench, he would apply a blue witness dot to the outward facing nut. He described doing this on a nut by nut basis to all the nuts on the fuel line, in accordance with Heli Support’s policy. He confirmed it was standard practice to check and mark all nuts in the fuel line, even those which had not been loosened off or removed.
[62] Mr Buick was certain the B-nut was not loose when the Helicopter was returned to service on 27 May 2013. He said, because the Accident occurred on the day the Helicopter had left his facility, he turned his mind immediately to the checks he had carried out. He said he clearly recalled a torque check on each B-nut in the
fuel unions from the bulkhead to the engine fuel inlet. Furthermore, he said it was his invariable practice to carry out a check of each nut in the fuel line.
Pre-flight checks
[63] After checking the nuts, Mr Buick carried out a bleeding of the Michigan filter, the engine fuel pump and the engine nozzle. He said he noticed a clean, free flow of fuel after initial trapped air bled out. The same occurred when the nozzle bleeding exercise was carried out with the engine being cranked over. In Mr Buick’s opinion, had the B-nut been loose, it would have manifested itself during the bleeding process. The fuel would have run aerated and, because the start pump was being run, fuel would have squirted out around the B-nut had it been loose. This did not occur.
[64] Mr Mullally did the pre-flight checks and carried out a ground run which involved starting the engine and then winding it back down. The ground run lasted for about 10 minutes. Mr Buick described standing at the back of the Helicopter with the engine cowlings open while the Helicopter engine was started. This enabled him to closely inspect all areas of the fuel line, the fuel filter and the fuel engine interface, and the engine itself. This would have required activation of the start pump which pushes fuel from the fuel tank under pressure in the fuel lines. Fuel would have squirted out of the B-nut had it been loose, in Mr Buick’s opinion.
[65] Mr Buick agreed that, when the engine is running, fuel would not squirt out of a loose B-nut. This is because the start pump is switched off and the engine pump then operates. The engine pump has a sucking action, drawing fuel through the fuel lines. If there were a loose B-nut, the engine pump would suck in air and fuel could not leak out.
[66] Mr Buick and Mr Schmidt looked at the catchment area under the engine and saw no fuel had collected in that area. They both paid particular attention to the Michigan filter, including the filter head, and fuel line fittings, of which the B-nut was part. At no time did they observe the engine display any abnormal running characteristics.
[67] The engine was then shut down and the Helicopter was released for a check flight.
Written records
[68] Heli Support’s work on the Helicopter was detailed on a Maintenance Worksheet. The Maintenance Worksheet is a form with a number of boxes to be completed for each item of maintenance carried out. The left-hand side of each box sets out the maintenance required in respect of each item and the right-hand side records the details of the maintenance performed. Each box is to be signed and dated by the person performing the maintenance. The Maintenance Worksheet recorded Mr Buick’s name and signature in respect of the ground run/leak check.
[69] The Helicopter’s log book maintenance record was then completed. That detailed the discrepancy/work required and the action taken. Relevantly it recorded:
Discrepancy/Work Required Action Taken Ground runs & leak checks required
Ground runs & leak checks performed. Engine pneumatic control system
leak check performed. Found satisfactory.
[70]Underneath all of the entries the following was typed:
We certify that a Duplicate Safety Inspection has been carried out and the identified control system of the aircraft functions correctly, and in respect of the maintenance performed, the control system is assembled and locked correctly.
[71]This was signed by Mr Buick and Mr Schmidt.
[72] Mr Buick then signed that the Helicopter was released to service for an operational check flight and that the maintenance recorded had been carried out in accordance with the requirements of the Civil Aviation Rules.
Check flight
[73] Mr Mullaly started the engine again and Mr Buick observed no fuel squirting out of the B-nut during the second start. Mr Mullally then piloted the Helicopter on the check flight, accompanied by Mr Buick and Ms de Reeper. This involved lifting off, hovering for a few seconds, climbing to 3,000 feet and carrying out a range of power checks. Mr Mullally described the Helicopter performing well. On return from the check flight, the engine was left running at flight speed. Mr Buick said he again opened the engine cowl doors and carried out another thorough check using a torch. He was satisfied and signed the release to service document, which Mr Mullally also signed. Mr Mullally then taxied to the fuel bowser, where the fuel tank was filled and he carried out a fuel drain check which involved draining a small amount of fluid from the fuel tank to check for water or contaminants.
[74] Mr Mullally confirmed that he did not detect any problems prior to the Accident.
Opinion evidence
[75] The total running engine time prior to the Helicopter departing the airfield was around 40 minutes. The pre-flight checks involved:
(a)the engine running for approximately 30 minutes;
(b)two engine starts;
(c)a take-off and a landing;
(d)changes in air speed and altitude; and
(e)the engine running at ground idle, flight idle, full power and (on approach to landing) low power and low fuel flow, with high power and high fuel flow in the landing flare, then, after landing, back to ground idle.
[76] In Mr Buick and Mr Schmidt’s opinion, had the B-nut been loose, then this would have been apparent.
[77] Christiaan Minnee, a retired aircraft engineer, has extensive maintenance experience on helicopters, including the Hughes MD369 series (now MD500 series), having worked on them since 1975. He gave evidence as an expert and was called by Heli Support. His expert evidence was based on reviews of the briefs of evidence, the videos (discussed below) and other documents relevant to the case. In Mr Minnee’s opinion, based on the descriptions of the checks carried out and operation of the Helicopter prior to its departure from Wanaka on 27 May 2013, the B-nut could not have been loose prior to departure. His evidence confirmed that of the other engineers as to normal practice, both in relation to whether the B-nut would have been undone to enable reinstallation of the engine and the steps an engineer would carry out to check the fuel line unions.
[78] In Mr Minnee’s opinion, the number of inspections of the engine bay and fuel system which Mr Buick said he carried out made it highly improbable that he would have missed a loose B-nut. He said, if the B-nut had been loose, fuel would have squirted out around the nut when the start pump was activated during the bleeding process. Significant quantities of fuel would have been apparent. Mr Minnee considered all the pre-flight checks and was confident that a loose B-nut would have evidenced itself. His evidence was qualified in cross-examination. I discuss this in more detail below at [93].
[79] Mr Shaw’s opinion was that, if the B-nut had been one-quarter turn loose during the pre-flight checks, particularly under start pump pressure, fuel leakage would have been apparent. He accepted that the engine was likely to start but said it would not run smoothly. He therefore concluded that the B-nut must have been airtight to the extent the union was sealed.
[80] Likewise, Mr Hobday agreed that, if the B-nut had been a one-quarter turn loose, fuel would have gushed out during the bleeding process carried out by Mr Buick prior to the flight. Mr Hobday therefore doubted that the B-nut had been one-quarter turn loose at the time of the Accident. He noted a number of people had looked at the
Helicopter after the Accident. He said the first thing he did was put his hands on the hose and “wriggle it”. He was aware others did the same. He therefore concluded that the position of the B-nut at the time of the Accident could well have changed given the number of people who had touched the B-nut after the Accident.
[81] In Mr Hobday’s opinion, the union of the two surfaces held by the B-nut must have been liquid and air tight in order to pass all the pre-flight checks because, during those checks, it would have been obvious if fuel was leaking.
Sealed union but loose B-nut
[82] Mr Hobday’s evidence was that the absence of a fuel leak during pre-flight checks did not mean that the B-nut was properly torqued. He said he loosened the fuel line B-nut on an MD500 helicopter with the start pump running, then finger tightened it and wiped it dry. No fuel leaked out. His conclusion, therefore, was that the absence of a leak did not say anything at all about whether or not the B-nut was properly torqued. Mr Shaw agreed with this conclusion.
[83] When the B-nut is properly torqued, the union is under compression. In Mr Hobday’s opinion the union must have been “mated” or sealed but not under compression because the B-nut was loose. This meant it was then vulnerable to forces during the flight, including G-force and vibration. He concluded that the union surfaces remained sealed through all the pre-flight checks at Heli Support’s premises in Wanaka and remained sealed throughout the flight.
[84] Mr Hobday said at some point, and it was not possible to say at what point, the differential pressure between the inside and the outside broke the seal. Once the surfaces were not touching, then there was a gap and air was ingested. Essentially, Mr Hobday said it was impossible to say why the seal failed when it did, but it had to have been because the union was in an unstable condition.
[85] Mr Hobday was asked how the Helicopter had been able to fly for some time before flaming out. He considered the various forces exerted during a flight, including air pressure from the fuel tanks. His theory is best explained by the following exchange which occurred during cross-examination:
A.… I'm saying that a sealed surface which isn't under compression at some point will leak when the differential pressure between the inside and the outside reaches the point where that sealed surface can't stand it anymore. So, as we go through the flight, there's some things that are happening which are changing the pressure and at some point, that sealed surface is going to let go.
Q. But the only thing that changed on this theory is that the helicopter descended and reduced to about 110 knots to about say 80 knots. Do you agree with that?
A. Yes.
Q. Are you saying the difference in air speed between 110 and 80 would have unsealed this join?
A.No, but the other forces taking place is that the – as the fuel level drops the engine has to draw fuel from the tank over a greater height, so the
– that drops the pressure as the speeds bleeding off and it's also dropping the pressure.
Q. Aren’t all of these forces exceedingly small and unmeasurable?
A. Well it depends what you wanted to measure them with.
Q. Well you've not been able to measure them yourself, have you?
A.No, no, and I – I've not tried to. It's easy enough to calculate the force from the fuel that's being sucked up. That's the – a simple calculation.
Q.Isn't the crux of this part of your evidence in paragraph 61 of your brief where you say, “Engine failure occurred when the helicopter was in the turn after a long period of straight, fairly late, straight and level flight. The reduction in forward speed during the landing approach would have caused a reduction on the tank pressure and increased the suction asserted by the mechanical suction pump.”
A. Yes.
Q. That's the crux of it?
A.That combined with the other forces which are happening. So, at some point, a sealed surface has to let go and it depends how tight that sealed surface is, as to what the differential pressure will be when it lets go.
[86] Mr Shaw agreed with Mr Hobday that it was possible the union was sufficiently sealed to enable the flight. He accepted the possibility that, if the B-nut was not properly torqued, it could have loosened as a result of vibration and forces during the flight, although he could not say at what rate.
[87] Mr Buick disagreed with Mr Shaw and Mr Hobday’s opinion that, if the union was sealed but the B-nut not properly torqued, the checks could have been completed without evidence of fuel seepage. Mr Buick was of the opinion that some fluid would have been visible, although acknowledged it depended on the level of tightness of the seal.
[88] In Mr Buick’s opinion, had the B-nut been loose by any margin, there is no possibility the union would have stayed tight enough to prevent air ingress throughout all the pre-flight checks, followed by a 45-minute flight. In his opinion, even a small degree of looseness would allow significant quantities of air to be entrained into the fuel line due to the suction created by the engine fuel pump. He referred to his own experiments on other aircraft which he said showed significant quantities of air sucked through even a small gap, causing the engine to run roughly, decelerate, and then flame out.
[89] Mr Buick regarded Mr Hobday’s theory as untenable. He noted that, when the fuel line exits the filter, it takes a right angle turn within a very short distance. In his opinion, the weight of the flexible line with fuel in it, the right-angle junction and the design and structure of the junction faces which sit inside the nut means that suction would not hold the junction tight when the nut is loose.
[90] In Mr Schmidt’s opinion, even if the union was sealed, the seal would not have survived the pressure of the start pump unless the union was “perfectly” mated. He said he had never been able to achieve a sealed union under finger pressure, although acknowledged it was possible to create a seal with a spanner.
[91] In Mr Minnee’s opinion, if the B-nut had been loose at the point of departure in Wanaka, there was no way the Helicopter could have flown through the 45-minute flight. He said a loose union would have leaked air all along and there would have been no significant change in forces on the union or the aptitude of the union to allow air to be ingested resulting from a descending turn and/or reduced air speed. He did not consider Mr Hobday’s theories could explain how a loose union could stay tight enough to prevent air ingress for some 85 minutes engine time through different phases of flight. He disagreed with the proposition that the suction from the engine-driven
fuel pump could have been sufficient to hold the two ends of the union together or that the fuel tank vents could exert a slight pressure on the surface of the fuel, thereby reducing the suction force required to draw fuel from the tank to the fuel control unit. He did accept that, when the helicopter is flying with significant forward speed, the inflow of air through those vents would create a slight pressure effect on the fuel in the tank.
[92] Mr Minnee’s opinion was based on the theory that the B-nut was loose to the extent it did not provide a clamping force on the union when it left Wanaka. He accepted his conclusions would have been impacted had the B-nut been sufficiently tight so as to enable a sealed union which was fluid and airtight. He accepted that, if a seal were formed, it could withstand the bleed test.
[93] Mr Minnee was asked about a test carried out by Mr Buick under which a B-nut was loosened by one full flat (60 degrees or a one-sixth turn) with the start pump on before any fuel appeared. On that basis, Mr Minnee agreed it was possible that, if the B-nut were less than a one-sixth turn loose, it would be possible to bleed the system without noticing leaking fuel.
[94] It was then put to Mr Minnee that, with the B-nut one-quarter turn loose and the engine running, the engine continued to run and it was only after the fuel line was wiggled that the fuel bled out. On that basis, Mr Minnee accepted the evidence suggested a change of force was required.
[95] Mr Minnee said the issue was the force that was required to loosen the clamping force of the union. He accepted Mr Shaw’s evidence that the vibration during flight could loosen an insufficiently tightened B-nut by as much as a quarter turn. He also accepted that there was a slight pressure effect as a result of air through the tank vents.
[96] Mr Minnee considered Mr Hobday’s theory of a single slug of air being taken in as improbable, saying that once air starts to be drawn in, he could not imagine the union resealing.
Hobday clear tube test
[97] On 6 June 2013, the day of the second inspection, Mr Hobday carried out a “clear tube test”. This involved attaching a clear tube to the engine of the Helicopter at the fuel nozzle, or point F on the diagram above at [48]. As the check valve closes when the engine winds down, it holds fuel in the fuel line. The purpose of Mr Hobday’s test was to show what was in the fuel line between points C, the fuel pump, and F, the fuel nozzle, at the time of the Accident. Because the clear tube test was not conducted until many days after the Accident, the experts agreed that any small air bubbles in the fuel would have by then joined into larger bubbles. What they expected to see therefore was not the heavily aerated bubbles seen in the similar test carried out by Mr Buick on 24 June, discussed below.
[98] The clear tube test assumed considerable importance during the trial. It was accepted that it would show whether there was any air in the fuel line at the time of the Accident. The experts agreed that, if a sufficient quantity of air had been sucked in through a loose B-nut and union to cause the engine to flame out, then air in the fuel line would be observed when its contents were expelled by cranking the engine.
[99]At the time the test was carried out, Mr Hobday recorded:
A clear tube was attached to the fuel nozzle end of the fuel nozzle hose. The engine was cranked and the ground idle fuel control setting selected. Clean fuel flowed immediately indicating no air in the fuel line from the fuel control to the fuel nozzle.
[100] Three videos were taken on 6 June, all recording the same event but from different angles. All three appeared to show some air bubbles, Mr Webb’s video being the clearest.
[101] When he gave evidence, Mr Hobday explained why his opinion had changed from that recorded at the time the test was carried out. He said, having viewed the three videos taken of the test again, he could see “significant quantities of air”. In his opinion, air in the hose was apparent in the first couple of seconds. He did not accept the test simply showed microscopic bubbles of a very small number.
[102] Mr Hobday pointed out that fuel continued to flow after the flame-out. He said that, once air was in the fuel line, it equalised the pressure a little so it was possible that air would not necessarily continue to be dragged through the fuel line.
[103] Mr Hobday accepted that, if there was a partial restriction of air, the engine would falter rather than cut out when at ground idle or flight idle. He said, when the engine is under power, it changes how the engine responds to air, saying:
… Now, the fuel nozzle is spraying a fine mist of fuel into the right-hand end and that is flowing into an airflow when it's in the air. That airflow’s doing about 200 feet per second and the flame front is burning up the fuel cone into that fast-flowing air. A very small interruption in that cone will mean that the flame is chased down through the turbine at 200 feet a second and doesn’t, so it doesn’t take very much of an air slug to break that cycle and the engine, and for the flame to disappear out of reach of the fuel nozzle.
[104] Mr Hobday was then asked why the automatic reignition system had not relit the flame as it was designed to do. Mr Hobday explained one of the triggers for the automatic reignition system is the main rotor rotations per minute (RPM) dropping. His understanding of the Accident was that they entered an auto-rotation immediately so the RPM did not drop. He referred to incidents he had investigated where engines had sucked in air because they were low on fuel and, in each case, did not relight on auto-relight. This had occurred mainly in low level operations when there was not a great deal of time before the helicopters were on the ground.
[105] Mr Hobday was recalled to view a slowed video of the clear tube test. This changed Mr Hobday’s opinion of the cause of the Accident from highly probable to conclusive. It confirmed his opinion that he could clearly see a globule of air at the very start of the test. He explained that the tube was a U-shape, meaning that, given air is lighter than fuel, air would collect at each end of the bend. He said the video showed a globule of air at the very start of the video, followed by more globules later which, in his opinion, was the air that was sitting against the check valve at the other end. Mr Minnee agreed that the air would settle at the two highest points, F, the fuel nozzle and E, the check valve, with more at the check valve end.
[106] Mr Buick and Mr Emeny described all three videos as showing a small amount of turbulent flow at the start. Mr Buick accepted there was a small quantity of air in
one of the videos (the Webb video), saying this was a result of undoing the union to attach the tube which allowed a small quantity of air to enter. He acknowledged Mr Hobday had kept the tube level when connecting it but considered some degree of manoeuvring was required to attach it. Mr Hobday rejected that proposition because of the way in which he had attached the clear tube, saying it would not have allowed air to enter the fuel line.
[107] Mr Hobday disagreed with Mr Buick and Mr Emeny’s description of there being only turbulence at the start of the test, representing, in their view, the fuel flowing into the slightly larger clear tube. Mr Hobday pointed out that, if they were correct, then the turbulence would remain at the start of the clear tube (a proposition with which Mr Minnee agreed) whereas it clearly rises up the tube, showing, he said, there was air in the tube. The fitting at the end of the clear tube was manufactured specifically for the test, making the wall as thin as possible so that the difference in size between the fuel line and the clear tube was approximately two millimetres. Mr Hobday did not regard that as significant for the purpose of the test.
[108] In Mr Buick, Mr Emeny and Mr Minnee’s opinion, the air bubbles they were able to see would have been insufficient to cause a flame-out and a much greater volume of air (evidenced by heavily aerated fuel) would have been required, Mr Emeny saying either a substantial slug of air or a continuous aeration of the fuel line content would be required. Mr Emeny was, however, unable to explain the presence of bubbles as the fuel flowed rather than simply at the beginning. Mr Buick, Mr Emeny and Mr Minnee did not consider the air bubbles were consistent with a one-quarter turn loose B-nut or a sudden engine flame-out. In their opinion, the clear tube test proved that air contamination in the fuel line did not cause the flame-out.
[109] Mr Minnee acknowledged that any visible air larger than a micro bubble was a concern. He accepted he could see in the video bubbles at the fuel nozzle (point F), the check valve (point E) and possibly some at the fuel control unit (point D). He accepted he did not know how much air had been in the line and pushed out before the check valve had shut.
[110]Mr Hobday reiterated that there should never be any air in the system.
Buick test 24 June 2013
[111] On 24 June 2013, Mr Buick carried out a set of tests on another helicopter of the same type, with a fuel system identical to that of the Helicopter. Mr Emeny and others were present. This involved trying to replicate the clear tube test as closely as possible. After a brief flight, the helicopter was shut down and a clear hose attached to the fuel nozzle and the engine cranked. Mr Buick described seeing one or two tiny bubbles of air, which he said probably came from the surface of the clear hose, but otherwise no aeration or bubbling in the fuel. This represented what could be expected from a normal and intact fuel system.
[112] Messrs Emeny, Minnee and Hobday agreed that tiny amounts of air would normally be introduced in an experiment of this kind.
[113] The second step was to loosen the B-nut to one-quarter turn loose when the engine was running at ground idle. Mr Buick said, within 20 seconds, the engine began to falter and decelerate and, two to three seconds after that, the pilot shut the engine off to prevent damage. They then ran the start pump and fuel squirted freely out of the loose junction. When the B-nut was tightened and the engine cranked with a clear hose on the nozzle, the emerging fluid was heavily contaminated with air, showing cavitation and air bubbling. In Mr Buick’s opinion, this conclusively proved that, if the B-nut is loosened one-quarter, the resultant fuel flow downstream will be heavily aerated and full of bubbles. This was not seen in the clear tube test on 6 June 2013.
[114] Mr Emeny described this test as showing heavily aerated fuel with a substantial amount of continuous air bubbling.
[115] Mr Buick accepted that, after undoing the B-nut one-quarter turn, he wiggled the hose around, saying he did so to replicate the break in the union. It was put to him, however, that the engine did not shut down until he had interfered with the seal. Mr Emeny accepted that Mr Buick’s wiggling of the fuel line would have introduced a large amount of air, resulting in highly aerated fuel. Mr Hobday contrasted the way in which he had carried out his test with the way Mr Buick carried out his test. In his
opinion, Mr Buick’s wiggling of the line would have allowed a significant amount of air to enter the system.
Could snow ingestion have caused the flame-out?
[116] I will now turn to the evidence adduced by Heli Support in support of its alternative theory that the flame-out was caused by snow ingestion.
[117] On 31 January 2000, Transport Canada issued a “Service Difficulty Advisory” headed:
McDonnell Douglas (Hughes)
369 Series Flight in Falling Snow – Optional Engine Air Inlet Deflector
It referred to an engine failure caused by snow ingestion following the dislodging of a snow deposit in the air inlet plenum.
[118] In March 2002, the Canadian Association of Geophysical Contractors (CAGC) issued a safety alert headed “MD 500 engine flameout”. It noted an incident in 2002 when an MD500 flying in British Colombia in light to moderate snow at around negative six degrees Celsius experienced a loss of power. The pilot heard a popping noise and experienced a yaw.4
[119] Canada requires a deflector kit to be installed in MD500 series helicopters when operating in temperatures below 4.4 degrees Celsius.
[120]In October 2005, Rolls Royce issued an alert as follows:
ENGINE FLAMEOUTS DUE TO SNOW OR ICE INGESTION – WARNING
Owners, operators and pilots are warned that helicopters/aircraft using the Model 250 engine in falling or blowing snow, or icing conditions, require special equipment. Snow or ice can build up on aircraft parts, inlet ducts or plenum chambers and break loose in “slugs”. Slugs of snow or ice entering the compressor of these engine models can cause flameout.
4 A yaw means the nose of the helicopter turns either to the left or the right.
[121]Notably, the alert was not limited to Canada.
[122] The manual for the Helicopter states that MD500s are permitted to fly in snow, provided they are fitted with an automatic reignition kit and engine failure warning system. Both of these items were installed in the Helicopter.
[123] Heli Support’s position is that the more likely explanation for the Accident is snow ingestion. Mr Minnee, Mr Buick and Mr Emeny referred to the MD500 series helicopter’s vulnerability to snow ingestion, saying the ledges in the plenum chamber provide spaces where snow can build up. They theorised that snow entering the alternate air door was drawn in, deposited and accumulated on a ledge behind the door in the plenum chamber. In their opinion, the tilting of the air frame during the turn or the changed angle during descent could have caused the snow to slide off, drop into the air inlet and cause a sudden and complete flame-out. Liquid found within the engine bay and underneath the Helicopter after the landing would be consistent with snow ingestion causing a flame-out, in their opinion.
[124] The particle separator in the MD500 helicopter will open the alternate air door provided it has been unlatched. Mr Hobday said the particle separator needs to be “quite substantially” blocked in order for this to occur. When the alternate air door is unlatched, it sits on the inside of the plenum chamber in its unlatched position, being held closed by air pressure until the pressure inside the plenum chamber drops, releasing the door.
[125] Mr Buick did not accept it would be normal or acceptable to open the alternate air door in snow conditions. He said the alternate air door should have been unlatched only when the warning light showed on the cockpit panel. He noted that, when the door is unlatched, unfiltered air enters the engine. By doing this, in Mr Buick’s opinion, any hindrance to snow, ice or rain ingestion was removed, allowing snow to build up in the plenum chamber and then to break off in a large clump. He said a mass of three to five square inches was required to block the air intake and cause a flame-out.
[126] Mr Buick was referred to the incidents discussed in the Canadian Safety Alert. He agreed that the pilots in those incidents reported hearing a popping noise and experiencing a yaw. He accepted Mr Mullally had given a similar description of the April incident, which had involved an air blockage. Those descriptions contrasted with the descriptions given by Ms de Reeper and Mr Mullally of a sudden wind-down of the engine during the Accident.
[127] Mr Emeny considered the fact the engine died suddenly was more consistent with snow ingestion rather than air ingestion through a loose B-nut. He referred to the test carried out by Mr Buick showing, in his opinion, that the engine began to falter once air was ingested rather than cut out instantly.
[128] Mr Buick, Mr Minnee and Mr Emeny did not agree with Mr Hobday that snow ingestion would always produce a TOT increase, for example if a clump of snow (sufficient to cause flame-out) accumulated and dislodged suddenly, falling into the engine. They accepted TOT will increase in progressive or partial obstruction but not if there is a total obstruction.
[129] Mr Hobday did not consider snow ingestion as an explanation for the flame- out. He noted that the alternate air door was opened as soon as the pilots encountered snow, the pilots said there was very little snow, it did not cause any problems, they were watching to make sure it was not building up and they were flying fast. He disagreed with Mr Minnee’s suggestion that fuel or other liquid observed in the engine compartment after the Accident indicated snow ingestion during flight. In Mr Hobday’s opinion, had the engine ingested snow during flight, it was highly improbable it would have left liquid that would be observable by witnesses after the Accident. He said, given the engine runs at between 600 and 700 degrees Celsius, any snow would have turned to steam immediately on contact. In any event, he referred to the evidence about the smell of fuel after the Accident and coloured liquid on the ground which, in his opinion, was much more suggestive of a fuel leak than of water.
[130] Mr Hobday also pointed out that many pilots fly MD500 helicopters without incident in conditions of much heavier snowfall than observed by the pilots in this case. Mr Hobday considered it unsurprising snow deflector kits are required in
Canada, given their more extreme conditions in contrast to New Zealand. He noted the Canadian incidents appeared to have occurred in a climb-out after take-off which would involve a heating phase with the result that clumps of snow which had formed in the plenum chamber before take-off would have been dislodged. In Mr Hobday’s opinion, there would need to have been a “good size lump of snow” to cause a blockage in the engine.
[131] Mr Shanks has flown MD500s since 1986 and has over 14,500 hours’ flight time in helicopters, virtually all in alpine and mountain conditions. He reported having flown the MD500 many times during snowfall and has never experienced a flame-out. Mr Mullally said not only was the snow shower light and lasted less than a minute, it was well over before the Accident. He said the Helicopter was travelling fast and the snow could not have built up because the speed would have caused it to blow off. In any event, he said he was conscious of the need to keep an eye on this and was doing so.
[132] Mr Mullally was asked about the Helicopter manual which suggested the alternate air door should be opened only once the particle separator caution light turns on, which it does when the filter becomes clogged. In Mr Mullally’s opinion, this was too late. He considered the reference in the flight manual to opening the door only when the light has been on for five seconds or more applied in the case of a malfunction and was not advice on how to use the alternate air door when not in an emergency situation.
Wet Michigan filter
[133] The final aspect of the evidence which requires some discussion concerns the Michigan filter.
[134] Mr Shanks described noticing that the top of the Michigan filter was wet when he saw it after the Accident on 28 May 2013. He made a note to that effect on 30 May 2013.
[135] Mr Hobday agreed that he would not expect the top of the Michigan filter to be wet if his theory were correct. That is because, when the engine winds down, the
check valve, point E of the diagram above at [48], closes and holds the fuel in the fuel line between points B, the air frame fuel filter, and E, the check valve, even if the B-nut were loose. Mr Hobday confirmed that a limited quantity of fuel only would escape.
[136] Mr Buick and Mr Emeny said the top of the Michigan filter could only have become wet when the start pump was activated and the last time this must have been done, to the best of Mr Buick’s knowledge, was prior to the Helicopter departing Wanaka. Mr Buick explained that, if fuel had leaked out of the B-nut, it would have accumulated around the nut and then dripped vertically down rather than travelled onto the top of the filter. He said, in contrast, if fuel were ejected under pressure, as with the start pump running, then it would reach the top of the filter.
[137] Mr Minnee agreed, saying a loose B-nut would not have resulted in fuel on the top of the Michigan filter. He said, if the Michigan filter was wet, the start pump must have been run with the B-nut loose or that it had somehow ended up there other than by normal operation or the Accident.
Heli Support’s submissions
[138] In Mr Gedye’s submission, Shanks failed to prove that Heli Support released the Helicopter to service with a loose B-nut and that the loose B-nut caused engine failure. He advanced numerous reasons why Shanks’ theory of the case could not be correct. Broadly they involved three categories: what happened prior to the Helicopter departing Wanaka; what happened during the flight; and what happened following the Accident.
[139] Mr Gedye characterised the key elements of Shanks’ case as improbable, saying they depended upon a high degree of guesswork or speculation. In his submission, the theory of “sealed but not tightened” was raised for the first time during the trial, something he described as a substantial shift in position. Mr Gedye described the theory as lacking any empirical support and being based entirely on assumption, given that the B-nut was found one-quarter turn loose.
[140] In Mr Gedye’s submission, Shanks had to establish how two metal faces, which merely sit against one another, can seal without the clamping force from the B-nut.
He said Mr Hobday did not address how and why a seal would undo. There was then, he said, no credible explanation as to how the B-nut came to be one-quarter turn loose when the Helicopter was inspected after the Accident. He referred to the evidence that air would have been ingested when the B-nut was less than one-quarter turn loose and therefore any problems would have manifested themselves well before the B-nut reached a quarter turn loose.
[141] Mr Gedye reviewed the work carried out by Heli Support on the Helicopter prior to the Accident and the evidence that the B-nut was not undone during the work. He stressed it was not normal practice to do so and this was agreed by the experts. He asked the Court to disregard the evidence of Mr Mullally as to what he had previously seen at Heli Support’s premises. In his submission, the weight of the evidence was against the possibility that the B-nut had been loosened when the Helicopter was worked on by Heli Support.
[142] Mr Gedye noted all experts agreed that, if the start pump was activated with the B-nut one-quarter turn loose, fuel would squirt out and it would be immediately apparent. He referred to Mr Buick’s evidence of checking the tightness of all nuts, Mr Schmidt’s evidence in support and the documentary records which, in his submission, supported Mr Buick’s credibility, including the presence of the blue mark on the B-nut.
[143] In Mr Gedye’s submission, had there been a loose B-nut, it would have evidenced itself during the pre-flight checks and the check flight. He then referred to the 45-minute flight time without incident, submitting Mr Hobday’s theory about vent pressure was untenable.
[144] Mr Gedye noted that the wet filter top was irreconcilable with Shanks’ case, which required a sealed union at Heli Support’s premises in Wanaka, yet the wet filter evidence conflicted sharply with this. In this regard, he questioned Mr Shanks’ credibility.
[145] Mr Gedye then referred to the fact the automatic reignition system did not relight the engine, despite mostly clean fuel being found in the line on 6 June 2013.
He noted that the system was designed to reignite and submitted Mr Hobday’s explanation, that he was aware of other cases where the automatic reignition did not work, was an unsatisfactory answer.
[146] Mr Gedye referred to the clear tube test and the tests carried out by Mr Buick. In his submission, the tests undermined Shanks’ case.
[147] While emphasising there was no onus on Heli Support, Mr Gedye identified the obvious logical possibility that someone loosened off the B-nut after the Accident and before the meeting on 29 May.
[148] Finally, in Mr Gedye’s submission, there was no dispute the Helicopter flew through snow and, given the warnings in respect of this model of helicopter, snow ingestion must be the most probable cause of the Accident.
Shanks’ submissions
[149] Ms Davies accepted Shanks does not know exactly how the B-nut came to be loose or how Heli Support went about its work. She acknowledged Shanks cannot attest to the state of the B-nut at the time the Helicopter left Wanaka. Nor could it directly establish the cause of the Accident, given the impossibility of recreating the exact circumstances. In her submission, however, Shanks had put forward a credible theory, consistent with key facts.
[150] In Ms Davies’ submission, the B-nut cannot have been properly torqued before the Helicopter left Wanaka, and this was the cause of the Accident. She referred to the clear tube test which clearly showed air to be present in the fuel line, stressing that any air in a fuel system is a matter for concern. She noted that both Mr Hobday and Mr Minnee accepted the slow-motion video showed an initial globule of air which had been sitting in the fuel line at the outlet end, followed by further globules which would have sat at the check valve and some later bubbles at the fuel control unit.
[151] Ms Davies referred to the criticisms of Mr Buick’s tests and submitted there was no evidential value in them. In her submission, the clear tube test was the best available evidence as to what was in the fuel line at the time of the Accident. She
referred to the criticism that there were not enough air bubbles in the fuel line and noted Heli Support’s witnesses’ concession that, because the Helicopter throttle was open before the Accident, fuel would have continued to be pushed through the line until the engine wind-down reduced the pressure in the line to the point when the check valve started to operate. Consistent with that concession, air in the line would have been pushed through by the ongoing flow of fuel and therefore the amount of air found after the engine shutdown did not indicate how much air was present earlier.
[152] Nor, in her submission, was it possible to determine post-Accident the rate at which the B-nut was letting air into the fuel line. The experts differed as to whether this would be a continuous flow or, as Mr Hobday theorised, a change in pressure after the initial slug of air could have slowed the rate of ingress. Ms Davies emphasised there was air in the line which could only have got there via a loose B-nut.
[153] That being the case, Ms Davies submitted the necessary implication was that the B-nut was loose (airtight and liquid tight but insufficiently torqued) when the Helicopter left Wanaka. All that could be said was that the B-nut was one-quarter turn loose when inspected following the Accident, after the impact of the crash and relocation of the Helicopter to Mr Shanks’ property.
[154] Ms Davies said Heli Support’s theory of snow ingestion should be disregarded, submitting the only factual evidence was that of the pilots, who both said there was not nearly enough snow for it to cause a problem. She noted the assertion that this model of helicopter is susceptible to flame-out from snow was supported only by bulletins from Canada but there was no evidence New Zealand conditions are bad enough to produce a similar outcome and no evidence of similar incidents in New Zealand. She stressed that the Helicopter was fitted with the systems for snow flying which are required in New Zealand and those were operational.
Legal principles
[155] Shanks claims that Heli Support was negligent in performing work on the Helicopter. The standard of care which applies to a person in a profession or trade is that reasonably expected of a skilled and informed member of that profession, judged
at the time the work was done.5 Evidence of common practice may be helpful and compliance with clearly established practice weighs heavily in favour of the person carrying out the work at issue.6
[156] This is a case where there cannot be direct evidence of what exactly happened at the time of the flame-out. I have therefore weighed the evidence carefully and, where appropriate, drawn inferences and resolved conflicts to the extent possible.
Factual findings
[157] The question is whether Shanks has proved, on the balance of probabilities, that the flame-out was caused by an interruption in the fuel supply caused by a loose B-nut and that Heli Support was responsible for that, having failed to properly torque the B-nut.
[158] Heli Support carried out work on the engine of the Helicopter which required disconnection of the fuel line. The fuel line is connected to the engine at one end and to the outlet side of the Michigan filter at the other, via a joint which is secured by the B-nut. The B-nut must be properly torqued so as to secure the joint. An inadequately torqued B-nut will allow the entry of air into the fuel line. If air enters the fuel line, it will cause the engine to flame out.
[159] On 27 May 2013, the engine of the Helicopter flamed out within 45 minutes of leaving Wanaka, where work on the engine had just been completed. Two days after the Accident, the B-nut was found to be loose by one-quarter turn. The loose B-nut was the only defect found in the Helicopter.
[160] It was not in dispute that the B-nut could not have been loose when the Helicopter was delivered to Heli Support for the engine to be installed. Furthermore, it was not in dispute that a properly torqued B-nut could not have come loose during the flight from Wanaka to the scene of the Accident.
5 Stephen Todd (ed) The Law of Torts in New Zealand, (7th ed, Thomson Reuters, Wellington, 2016) at [7.5.02].
6 At [7.4.01].
[161] I was considerably assisted by the experts in this case. They all conducted themselves impeccably and were fair and considered.
[162] I will start with the clear tube test, as all experts (and Mr Buick) considered it supported their theory of the case. If the test revealed that air was present in the fuel line, then that is strong evidence to support the cause of the Accident being an interruption in the fuel supply.
[163] Although I watched all three videos, including the slowed video, and considered the photographs, I place more weight on the evidence of the experts, given their familiarity with the subject matter. In saying that, I share Mr Hobday’s interpretation of the clear tube test. That is that air was present at the start of the test and then at a slightly later stage. This is consistent with the air having settled at either end of the tube. The obvious presence of air later in the video is not explained by Heli Support’s theory that the air at the outset was caused by the way in which Mr Hobday connected the clear tube for the purpose of the test.
[164] In many ways, Mr Buick’s tests on 24 June supported Mr Hobday’s position. First, any microscopic bubbles in the fuel line when the B-nut was properly torqued were explained, as Mr Minnee accepted, by the way in which the clear tube was connected. Secondly, it was only when the B-nut was one-quarter turn loose and the fuel line was wiggled that the engine cut out.
[165] That being the case, I return to assess the other evidence. As noted, it was agreed that the B-nut could not have been one-quarter loose when the Helicopter left Wanaka. As Ms Davies suggested, there are a range of possibilities as to why the B-nut was not properly torqued before it left Wanaka. She referred to the possibility the fuel line was undone at the filter end for the engine change and the B-nut not properly torqued afterwards. Alternatively, the fuel line was undone at the engine end but the B-nut loosened at the filter end to help in tucking the fuel line out of the way when the work was being carried out. It was accepted that neither approach would have been in accordance with best practice or even standard practice. As against that, Mr Mullally said he had seen the B-nut being loosened for engine changes on many
occasions, specifically when the engine was removed in April 2013 after the April incident. That evidence was not challenged.
[166] Mr Buick was sure he had checked the B-nut and had put a blue mark on it. The blue mark was found after the Accident to be in the five o’clock position (or one quarter turn loose), a position Mr Buick said it could not have been in when he marked it because of the awkward angle. Both Mr Mullally and Mr Webb saw what they thought was a second mark.
[167] One potential explanation is that the blue mark was placed on the B-nut by Mr Buick after he had checked it in Wanaka. The blue mark does not necessarily have to be in the eight o’clock position (consistent with the B-nut being properly torqued), as its purpose is to record that the B-nut has been checked. It is fair to say, however, it would be there or thereabouts. If Mr Buick’s evidence about torquing the B-nut and making the blue mark is accepted, it means that the B-nut must have been deliberately loosened between the time of the Accident and the inspection on 29 May. I reject entirely any suggestion of deliberate or malicious interference. Those with the opportunity to do so reject that proposition out of hand. I accept that evidence.
[168] This leads to two possibilities. Either that Mr Buick did mark the B-nut, but it was not properly torqued and loosened from its position after it left Wanaka. Alternatively, that the mark was left from a previous occasion when the B-nut had been checked. Mr Buick said he cleaned old marks off with solvent, although it was not clear when he did so and whether he cleaned the old marks off just before making new marks.
[169] The Helicopter was not ready when the pilots arrived to collect it. Mr Buick knew they had a job to do. Four engineers were involved in the repair work. We heard evidence from only two. Mr Schmidt countersigned the maintenance worksheet but he did not check the B-nut because, in theory, it should not have been disturbed for the engine replacement.
[170] There is no doubt that extensive pre-flight checks took place and, had the B-nut been one-quarter turn loose at this stage, it would have been apparent.
[171] Mr Hobday’s evidence was that he (and in his impression others) had wiggled the fuel line during their post-Accident investigations. That was his explanation as to how the B-nut came to be one-quarter turn loose when examined on 29 May. That seems to me to be a real possibility.
[172] Given the evidence of air in the clear tube test, the question of how the Helicopter passed the pre-flight checks and flew approximately 45 minutes prior to the Accident needs to be addressed. Mr Hobday’s theory was that the B-nut was sufficiently tight to hold the union air and fuel tight. The two faces of the union had to be not only touching but forming a seal.
[173] I do not accept Heli Support’s criticism that Mr Hobday’s theory changed. Rather, it could be said that it was better explained and perhaps refined as the case progressed, to the extent that Heli Support’s expert, Mr Minnee, accepted many of Mr Hobday’s propositions. Mr Minnee acknowledged, for example, the possibility that at something less than one-sixth turn loose, fuel would not leak out of the B-nut, even with the start pump on. Mr Minnee’s disagreement really came with what he observed in the clear tube test. It was only Mr Emeny who declined to accept the propositions arising from Mr Hobday’s theory put to him in cross-examination. I place greatest weight on the evidence of the two truly independent witnesses, Mr Hobday and Mr Minnee. Both were thoughtful and measured and genuinely tried to assist the Court. As an aside, I record my impression that, had the two experts met in advance of the hearing and properly discussed the case, a large measure of agreement would have been reached.
[174] Mr Hobday’s evidence that “some things are happening which are changing the pressure and at some point, that sealed surface is going to let go”, probably reflects the reality. It is impossible post-Accident to be definitive as to exactly why the seal broke when it did.
[175] The evidence about the smell of fuel and the presence of coloured liquid on the ground after the Accident supports the case that the cause of the Accident involved the fuel system.
[176] It is not necessary for me to conclude how it was that the B-nut was not properly torqued prior to departure from Wanaka. I am satisfied, on the balance of probabilities, however, that it was not. As Heli Support accepts, it breached its duty of care to Shanks if the B-nut was not properly torqued before the Helicopter left its premises in Wanaka.
[177] While I place limited weight on Mr Buick’s “admissions” to the effect that Heli Support must have failed to check the B-nut, both Mr Shanks and Mr Mullally are sure he said that and I can find no basis to reject their evidence.
[178] That Mr Shanks observed fuel on the top of the Michigan filter is a problem confronted by both sides. Given there was agreement that there were only two possible causes of the Accident – interruption of the fuel supply or interruption of the air supply
– the wet filter top is a conundrum. It does not fit either possibility. Mr Shanks noticed the wet filter and made a note to that effect. Mr Buick also says he noticed it on 29 May. As I say, the evidence does not work with either theory of the cause of the Accident and I can take it no further.
[179] I do not find the theory of snow ingestion to be a reasonable possibility. It required a sufficient amount of snow to restrict the operation of the particle separator to the point where the pressure inside the plenum chamber reduced and the alternate air door opened. Then sufficient snow had to enter through the alternate air door that it accumulated on the ledge at the back of the plenum chamber, dislodged and fell into the compressor when the Helicopter came in to land. This in circumstances where the two pilots said they flew through a light snow shower for about a minute and there was not sufficient snow to have caused a problem. I place considerable weight on the evidence of the two pilots, particularly that of Mr Mullally, given his long flying career in Southland/Fiordland. I also note the behaviour of the Helicopter when it flamed out was not consistent with the interruption of air supply. Mr Mullally was flying the Helicopter around a month earlier in the April incident when there was an air supply incident. He described the Helicopter losing power and starting to yaw. The Canadian CAGC evidence implies a yaw on flame-out when air supply issues caused by snow ingestion arise. In contrast, the evidence of the Accident was that the Helicopter
stopped suddenly, which the pilots described as the same as during an ordinary shutdown when the fuel is turned off.
[180] It is difficult to understand the implied criticism of the pilots for opening the alternate air door in advance of entering snow when the evidence was that the door is held firm until the particle separator becomes blocked. That seems to dispose of any suggestion that the decision to unlatch the door earlier led to contaminants entering the plenum chamber and building up. Mr Mullally said this was the practice he had employed over his long years of experience as a helicopter pilot.
[181] I am satisfied, having weighed up all the evidence, that Shanks has established the cause of the Accident on the balance of probabilities.
Result
[182] Judgment is given for Shanks in the sum of USD 390,643.91 plus interest and costs. Costs should, in my view, be on a 2B basis. If Shanks considers otherwise, it is to file and serve a memorandum within 20 working days from the date of this judgment. Any response from Heli Support is to be filed and served within 10 days thereafter.
Thomas J
Solicitors:
Fee Langstone, Auckland for Plaintiff Tompkins Wake, Hamilton for Defendant
0
0
0