AS Avies JS31 at Sveg on May 3rd 2013, both engines fluctuating in initial climb

An AS Avies Jetstream 3200, registration ES-PJR performing flight U3-2071 from Sveg to Mora (Sweden) with 14 passengers and 2 crew, had overnighted at Sveg in a hangar. The aircraft was subsequently towed to the apron for boarding. During engine start and taxi the crew observed all normal operations without any anomaly. The aircraft departed Sveg's runway 09 with normal acceleration during the ground roll and normal rotation. The aircraft was in the initial climb through about 500 feet AGL when the left hand engine began to lose power, recovered, then the right hand engine lost power and recovered, both engines continuing to oscillate that way causing the aircraft to alternatively yaw left and right. The crew had difficulty to keep the aircraft flying with the overall reduced power and needed to focus on maintaining airspeed, maintaining altitude and directional control. Both crew were concerned the aircraft would go down and looked out for a possible landing spot outside the aerodrome, however the terrain, forested and with water streams, did not offer any open spot for a landing. The crew thus decided to wage a right hand turn to turn back to the airfield and land on the runway. At the same time, while attempting to return to the aerodrome, the crew undertook several attempts to stabilize the engines, the captain reduced power on both engines and switched off the Torque Temperature Limiter. After the aircraft had rolled out of the right hand turn both engines stabilized and operated normally again. The crew completed a normal downwind and landed safely on runway 09 about 6 minutes after becoming airborne.

Sweden's Statens Haverikommission (SHK) released their final report concluding the probable causes of the serious incident were:

The incident was likely caused by a too low RPM during take-off. A contributing factor was that the aircraft type has no warning system for take-off with an incorrect engine configuration.

The SHK reported that the commander (41, ATPL, 5,146 hours total, 3,203 hours on type), pilot flying, had long experience on the aircraft type and was also an instructor on the aircraft type. The first officer (26, CPL, 630 hours total, 175 hours on type) was pilot monitoring.

A post flight ground run revealed no anomaly with both engines. However, corrosion was noted at the aircraft plumbing. The tube fittings for the total air pressure at the engine inlet (PT2) were found damaged at both engines, the left hand tube had been provisionally repaired with a piece of rubber hose, the right hand tube was leaking on a coupling, when disconnected the pipe burst due to corrosion. In addition the plumbing connections for the static air pressure at both engine outlets were found leaking and contained some water. Analysis however showed neglible effect on the engine operation by both defects and could not explain the chain of events.

The SHK reported that the propeller speed should be about 96-100% during takeoff, requiring the RPM lever to be moved into the HIGH position prior to advancing the throttle levers for full power. In this configuration the RPM would be held constant and power changes would occur through fuel flow changes and adjustment of the propeller blade angle automatically performed by the propeller governor. The torque would be measured in the propeller shaft. The power management of the engine would permit only a certain fuel flow depending on the current engine RPM according to a acceleration curve as integral function of the power management.

The cockpit voice recordings were overwritten, both flight crew and maintenance failed to power down the cockpit voice recorder after landing. A passenger however recorded the takeoff with his mobile phone, the SHK used the recording to analyse the sound during the event.

The operator lacked necessary and mandatory documentation to convert the digitally stored data on the flight data recorder into engine units. The standard documentation by the aircraft manufacturer also produced unrealistic results. The investigation commission thus failed to locate such documents. The investigation therefore developed a special correction polynom to convert the recorded parameters into engine parameters (see below for the graph). The corrected parameters show the engines accelerated to 100% RPM at the time the torque increased for takeoff as well and reduced again settling at 95% with the left hand engine slightly below.

The SHK reported ambiguity in their checklists prior to takeoff. Initially the term "HIGH" is used to describe the setting for the RPM levers, later the term "FLIGHT" is being used. In both cases the procedures make clear however, that a minimum of 96% RPM is to be achieved on both engines with the power levers in ground idle positions.

The SHK analysed: "In connection with take-off, severe oscillations of the power occurred in both engines at low altitude, which entailed a serious flight safety risk. No technical fault which could explain the engine oscillations has been found. Neither the defects established in some of the piping nor the pollutants found in the fuel and oil filter are deemed by SHK to have had any significance in this context. It is unlikely that there would have been temporary external conditions of some sort that affected the function of the engines during take-off. According to SHK's experience, both engines have also functioned after the incident, with no remarks. SHK establishes that the sequence of events and the similar oscillations in both engines is very much in line with what has occurred in previous incidents with this aircraft type when a high engine power is set whilst the engine RPM is too low; see Figure 21. This is supported by the fact that the amplitude of the oscillations increased drastically after a certain time; this indicates that the engines' TTL system was activated. This “characteristic” of the engine/propeller installation has been verified and is well known by engine and aircraft manufacturers."

The SHK concluded the flight analysis: "SHK establishes that the specific characteristics of this aircraft type - i.e. that severe oscillations in the engine power can occur if the RPM is too low when the power output is high - can entail a serious flight safety risk if the pilots do not have full knowledge of the phenomenon."

The SHK expressed dissatisfaction in their further analysis of the operational aspects: "From a flight safety viewpoint, it cannot be considered satisfactory that a system in which the consequences of a malfunction or mismanagement can be so serious that oscillations occur on both engines simultaneously does not feature a safety system which warns the pilots."

With respect to reading emergency checklists the SHK analysed: "The malfunction which occurred during the incident in question was likely caused by an incorrect engine configuration for take-off. The consequences - serious engine oscillations just after take-off - occurred in a critical phase of the flight when the aircraft was at low altitude during acceleration from a low speed area. During this phase of the flight, the crew's focus must be on the flight continuing in a safe manner. In such a situation, the crew cannot be expected to take out an emergency checklist in order to look up the most appropriate measures. Such measures should be included in memory items. The fact that the commander still carried out virtually all of the prescribed measures at the time is likely attributable to his long experience – including his service as an instructor – on the aircraft type. Recently trained pilots, or pilots with low experience on the type, cannot be expected to possess the equivalent knowledge."

The SHK stated: "SHK considers this incident to be so serious that the conditions for the crew's handling of this problem need to be revised."

The acceleration curve in stable configuration (Graphics: SHK):

The reconstructed engine data off the flight data recorder (Graphics: SHK):

The flight path, yellow indicates flight with engine malfunction (Graphics: SHK):