B38M, Helsinki Finland, 2019

B38M, Helsinki Finland, 2019


On 18 January 2019, two aircraft taxiing for departure at Helsinki were cleared to cross the landing runway between two landing aircraft. Landing clearance for the second was given once the crossing traffic had cleared as it passed 400 feet in expectation that the previous landing aircraft would also shortly be clear. However, the first landing aircraft was slower than expected clearing the runway and so the second was instructed to go-around but did not then do so because this instruction was lost in the radar height countdown below 50 feet and the runway was seen clear before touchdown.

Event Details
Event Type
Flight Conditions
Flight Details
Type of Flight
Public Transport (Passenger)
Flight Origin
Intended Destination
Take-off Commenced
Flight Airborne
Flight Completed
Phase of Flight
Location - Airport
Airport Layout, ATC Training
ATC Clearance Cancelled
Incursion after Landing, Runway Crossing
Damage or injury
Non-aircraft damage
Non-occupant Casualties
Off Airport Landing
Causal Factor Group(s)
Air Traffic Management
Safety Recommendation(s)
Air Traffic Management
Investigation Type


On 18 January 2019, a Boeing 737-900 (TC-JYH) being operated by Turkish Airlines on an international passenger flight from Istanbul to Helsinki as THY1XE was slow to clear the landing runway at Helsinki after completing a landing in night VMC and the next landing aircraft, a Boeing 737-MAX 8 being operated by Norwegian Air International on an international passenger flight from Krakow to Helsinki as IBK351, was instructed to go around when on very short final but did not do so after the crew had confirmed that the runway was completely clear.


After a Serious Incident was notified by ATC, an Investigation was carried out by the Safety Investigation Authority of Finland (SIAF) into the circumstances which had led to the issue of the very late go around instruction and the landing which was then completed contrary to it. ATC radar and voice recordings were available and FDR and CVR data relevant to the event were provided by Norwegian Air.

The crew of the second landing aircraft filed an internal report in line with company procedures after learning about the event from ATC. Although it was not assessed to have had any bearing on the event, the TWR controller position was being operated by a trainee under the supervision of an On-the-Job-Training Instructor (OJTI) albeit, the trainee was an experienced controller undergoing validation for the Helsinki TWR position.

What Happened

It was noted that separation between the two aircraft landing on the active runway 22L was approximately 4 nm, which was an equivalent to a one and a half minutes gap between them. The general traffic situation at the airport was normal for the time of the day. Once the first of these two aircraft had landed, the TWR controller cleared two aircraft which were taxiing north from the terminal area towards the active departure runway 22R - SAS 1713, a Bombardier CRJ 900 going to Stockholm and on taxiway ZD and SAS 719, an Airbus A320neo going to Copenhagen and on taxiway Y - to cross runway 22L (see the illustration below).

Once the TWR controller had verified that both departing aircraft had cleared the landing runway and that the previous landing aircraft was turning off onto exit taxiway ZJ, he cleared the second aircraft to land as it was passing approximately 400 feet. It subsequently became clear that the aircraft expected to vacate had slowed markedly during its rollout and after entering taxiway ZJ at a groundspeed of approximately 9 knots, its speed had then reduced to approximately 4-5 knots as it entered the taxiway. Having instructed this aircraft to vacate the 60 metre-wide runway and contact GND, the second landing aircraft was approaching the threshold and by the time it arrived over the threshold, the previous landing aircraft had vacated the 60 metre-wide runway and was entirely on the taxiway.

However, although the TWR controllers were monitoring the situation, the vacating aircraft’s slow speed prevented them from positively determining whether it was moving or not and having concluded that the runway was still occupied and with the next landing aircraft now between 50 feet and 30 feet agl over the threshold, the OJTI controller prompted the trainee to issue a go-around instruction to the aircraft which had previously been cleared to land. There was no response to the go-around instruction which was obscured by the fixed high volume of the automated callout system annunciating radar height at 10-foot intervals from 50 feet agl.

The Helsinki Aerodrome Diagram. [Reproduced from the Official Report]

On reaching the end of taxiway ZJ, the first landing aircraft turned right onto taxiway Z and was led by a ‘Follow-Me’ car to its designated parking stand. During the second aircraft’s taxi-in, the TWR controller asked the crew to establish contact on a discrete frequency and there advised them that they had been told to go around and that the event would be reported. The Captain, who had been PF for the flight involved, also subsequently called ATC on the phone and made it clear that they had not ignored the go around instruction, they had not heard it.

The traffic situation with the first landing aircraft just clear and the second over the runway. [Reproduced from the Official Report]

Some Pertinent Observations

  • The prevailing weather conditions at the time were being reported as CAVOK (cloud and visibility OK).
  • The flight crew of the second landing aircraft were both wearing Bose A20 headsets certified for aviation use which incorporated a noise reduction feature that effectively attenuated background noise which would otherwise be audible in the flight deck. The automated callout system which enables the 10 foot interval height calls below 50 feet agl (as well as less frequent height calls earlier in an approach) may be annunciated through the roof panel speakers but when headsets are in use these would normally be (and were) set to mute. These calls were thus (routinely) heard at a relatively high (and pre-set) volume in the pilots’ headsets.
  • Communication systems used in civil aviation do not incorporate technical solutions to supplement voice messaging when critical instructions, such as go-around calls, are issued to arriving aircraft and do not incorporate a selective calling method such as that used to communicate alert, urgency, or emergency messages between maritime stations nor any other audio tones that would alert a receiving station to an impending critical message.
  • Until a landing clearance has been received, pilots are “go-around minded,” whereas after it has been given, their attention shifts to landing the airplane and they become “landing minded”. There was no communication with ATC or any other direct visual indication that conditioned the pilots to expect an instruction to go-around. Since a go-around call is rarely transmitted when the aircraft is already over the runway, pilots with a landing clearance are unlikely to anticipate one unless the controller has pre-advised the possibility or they have already recognised themselves the possibility on account of a change in weather conditions, an unusual traffic situation or some other circumstance. Once over the runway (the threshold should be crossed at 50 feet agl) touchdown will happen in a matter of seconds.

The specifically ATC perspective

  • The second aircraft was cleared to land on the still-occupied runway on the basis that it could be reasonably expected that adequate separation would exist when the aircraft reached the runway threshold. The purpose of the principle of reasonable expectation is to facilitate efficient use of a runway and it involves a built-in option to cancel the landing clearance and instruct the aircraft to go around if the runway remains occupied.
  • An alternative is to notify a crew that they should “expect a late landing clearance” or by use of an information call such as “continue approach, another 737 vacating” with a landing clearance then issued once the runway is clear. This option also enables the crew to anticipate a go-around which they will then automatically initiate in the absence of a landing clearance at a time which they can decide.
  • The location of the position on the runway at which the departing aircraft needed to cross and that where the first landing aircraft was to exit both posed a visual observation problem for the controllers because of their angular displacement of approximately 90° which was exceeded for the just landed aircraft as it entered the taxiway. This meant that in the latter case, it had been impossible to determine whether the aircraft was continuing to move or had come to a stop whilst still partly on the runway. However, aircraft which had landed both before and after this one were found to have then exited the runway and proceeded along the same taxiway at approximately twice the speed.
  • When issuing the late go-around instruction, the trainee controller maintained the tone and pitch of speech that he had used in preceding communications which gave no hint of any abnormality. This was understandable given that no potential for collision existed and especially so given that the controller in position was a supervised trainee. Had there been a real collision hazard, the tone used “would very likely have been more forceful and surprise would definitely have been reflected in the controller’s voice”.
  • During interview in connection with the Investigation, both controllers “mused on the reasons why the crew had disregarded the go-around call” after initially establishing that the go around instruction had not been masked by a transmission from another aircraft. It was concluded from these interviews that the controllers did not understand what had happened on the second landing aircraft’s flight deck. Their view was purely technical and experience-based and did not demonstrate an understanding of factors related to flight crew operation, the human factors in play during the final approach and landing, or the impact of automated callouts in the flight deck.
  • Prevailing ATC SOPs stated that an aircraft may be cleared to land on an occupied runway when there is reasonable assurance that the runway will be clear no later than the time when the aircraft crosses the runway threshold. The Investigation revealed that risks resulting from issue of a landing clearance for an occupied runway and its consequences had not been assessed by either the safety regulator or the ANSP and parameters such as but not necessarily limited to weather conditions, surface friction and lighting relevant to ‘reasonable assurance’ have not been established. The safe application of the principle of reasonable assurance “therefore hinges on appropriate action by controllers exercising operational responsibility”.

Five specific Conclusions in respect of the Investigation which “include the Causes of the investigated event which means the factors underlying it and the direct and indirect factors that had an effect on it” were documented as follows:

  1. The principle of reasonable assurance is recognised internationally, and its purpose is to enable the efficient use of runway capacity. If the situation changes, safety will be ensured by cancelling the landing clearance. This instruction shall be transmitted in a timely manner to allow time for its execution.
  2. Runway crossings affect tower operations at Helsinki-Vantaa airport in several ways and may result in the breakdown of de-confliction. Controllers should be aware of the associated risks and assume a proactive stance to the problem during all work shifts.
  3. There can be a multitude of justified reasons for slower-than-normal taxi, such as procedures at an unfamiliar airport, winter conditions, or waiting for a taxi clearance or a ‘follow-me’ car. The traffic situation and prevailing conditions may change abruptly, in which case ATC will need to decide quickly whether the conditions of a landing clearance continue to exist, or should the succeeding airplane be directed to go around.
  4. Controllers shall be sufficiently conversant with flight crew activities and the sound environment in the flight deck during various phases of the flight. Sometimes the controller may need to transmit an urgent instruction (such as a go-around call) to ensure safety in an evolving situation. These calls shall be readily distinguishable among other radio communications in terms of both tone and volume.
  5. An underlying safety management principle is that actions shall not contradict rules and regulations. Either the actions and procedures shall be aligned with internationally recognised regulations, or the Finnish interpretation - that in this case offers a wider safety margin - shall be endorsed and incorporated in aerodrome-specific publications.

Three Safety Recommendations were made as a result of the Investigation as follows:

  • that Traficom (State Safety Regulator) and ANS Finland (ANSP) launch a joint long-term risk analysis of potential hazards associated with landing clearances that are based on reasonable assurance. The analysis would look at challenges to the controllers, the potential consequences of non-executed go-around instructions and other contingencies. The results would be used to revamp the existing procedures by eliminating the possibilities of inconsistent interpretations resulting from the way the conditions are prescribed in the present documents. [2019-S61]
  • that ANS Finland pays attention to the way controllers speak during communication and the tone and volume of speech in the delivery of critical messages. The principles shall be clearly documented and observed during site-specific and other training. [2019-S62]
  • that Traficom and ANS Finland jointly see that the controllers’ actions are in compliance with the existing guidelines and instructions and clarify the definition of runway vacation in the national documents if necessary. [2019-S63]

The Final Report was published on 19 December 2019.

Related Articles

SKYbrary Partners:

Safety knowledge contributed by: