| Level Busts – ACEing
the hazard
by Tim Atkinson
Since September 2001, a UK National Air Traffic
Services Action for Continuous Excellence (ACE) project team has been examining level busts, searching
for their causes, and planning strategies for mitigation. At the 346th
meeting of the UK Flight Safety Committee (UKFSC) in May 2002, team members gave a presentation,
describing the work done to date, and explaining future plans…
 Level busts have a real potential to claim
lives. The mid-air collision near New Delhi in 1996 killed 349 people,
and was the result of a simple level bust. In the UK, there are
approximately 300 level busts a year, many of which do not result in
losses of separation, though all have the potential for very serious
outcomes.
What is a level bust?
A level bust occurs when
an aircraft deviates from the correct level by more than 300ft.
Normally vertical separation between aircraft is 1000ft. This project
did not address level busts caused by TCAS events, nor instances of
‘late re-clearance’.
Modern ATC equipment, including Short Term
Conflict Alert (STCA), high quality radar, well-trained and
experienced controllers, and up-to-date aircraft fitted with Traffic
Alert & Collision Avoidance System (TCAS) flown by competent crews,
should afford better protection. Despite this, a mid-air collision
(not, it seems, the result of a level bust) occurred near Ueberlingen,
earlier this year. This proves that reliance upon present collision
avoidance techniques is not sufficient; accepting level busts and
trusting that a collision will be avoided is not enough; new ways of
eliminating level busts must be found.
Recognising the nature and potential severity of
the hazard, UK National Air Traffic Services set up an ACE project in September 2001. The Level Bust ACE
team, chaired by a Terminal Control Watch Manager, included ATC
managers and safety investigators, representatives of Safety
Regulation Group, a human factors scientist, and pilots from British
Airways and easyJet. The project is sponsored at NATS board level, and
has an unrestricted brief. An ‘ACE agent’ oversees and facilitiates
the group’s work, providing guidance and administrative support, and
ensuring the group retains its focus and that effective progress is
made.
ACE projects address difficult problems that
need to be solved. The process, developed from an analysis of industry
best practice in problem solving, uses multi-disciplined teams and a
structured approach, focusing on taking action and implementing
solutions.
Over a series of one-day workshops, the team
carried out a practical analysis of the root causes of level busts,
and prioritised these causes, before discussing and identifying forty
possible solutions. These solutions were themselves prioritised and
matched against causes, in a matrix, before further work to specify
solutions for action took place.
Solutions identified for action ranged widely in
their simplicity of development and implementation, cost, and
potential benefits in addressing issues other than level busts.
Education and Awareness
The most important solution identified is the
need for an education and awareness programme for ATCOs and pilots. It
became clear that many, if not all, level busts would be avoided if
crews and controllers complied with current regulations and ‘best
practice’. A new working group, incorporating members of the ACE
project team, together with ATC training experts, is developing this
campaign. The campaign will be rolled out through summer and autumn
this year. It is hoped that a concerted effort to improve SOP
adherence, RTF discipline, and the use of ‘best practice’, may bring
about a significant improvement in the level bust statistics.
Alphanumeric Callsigns
Data from ATC safety investigations identified
that callsign confusion, a significant cause of level busts, is much
reduced by the use of alphanumeric callsigns. This solution is being
further examined, in order to establish that there are no unknown
risks associated with alphanumeric callsigns. Once this is complete,
consideration may be given to seeking regulatory action to mandate the
use of alphanumeric callsigns.
RTF Phraseology
Much discussion took place regarding some of the
RTF phraseology in current use. Whilst it was felt that most
communication difficulties arose from poor technique or assumption,
rather than from questionable phraseology, some opportunities for
improvement were identified. The words ‘Flight Level Wun Hundred’ have
been used in place of ‘wun zero zero’ on a trial basis for some time,
and with success. This phraseology has been extended to other levels
(200, 300, and 400). Other phraseology changes, such as the use of the
word ‘degrees’ after headings, in order to differentiate from flight
levels, are still under consideration.
The UK air traffic controller’s ‘bible’, the
Manual of Air Traffic Services Part One, gives a full account of
standard RTF phraseology as used in the UK. Other states use slightly
different phraseology. The Manual is available online
here , and Appendix E is the appropriate section.
The UK AIP also gives
relevant information
here
(You will have to register - free).
CAP413 gives an overview of RTF techniques
here.
Chart Deficiencies
Many pilots reported level busts as a
consequence of misreading SID charts. In particular, one chart
manufacturer’s depiction of step climb SIDs was felt to be unclear.
Representations have been made to the manufacturer concerned, and work
is ongoing to achieve changes.
FMC Software Modification
 Modern aircraft such as the Boeing 737-NG and
Airbus family have an alerting feature, which warns crews of climbing
or descending through transition altitude without re-setting the
altimeters. Investigations have been carried out to ascertain whether
the FMC on older aircraft (such as the Boeing 737 EFIS, 757, and 767)
could be modified to provide a similar alert. Initial indications are
that this is feasible, and presentations will be made to FMC
manufacturers in due course. One major operator has found that their
Boeing 737 EFIS fleet experiences ten times more level busts with
altimeter setting errors as their cause, than their A320 family fleet.
(Industry support for this proposal is sought – interested parties
should contact the author
tim.atkinson@easyJet.com).
Distraction-Free Flight Deck
The project group identified that distraction,
causing breakdown of SOPs, was a common factor in level busts, and
concluded that a distraction-free flight deck is an environment in
which level busts are less likely. Of course this is a matter of
discipline, and some companies already operate a ‘sterile’ flight
deck. Most operators are believed to be aware of the issue, and when
the security implications of ‘locked door’ policies were considered,
it was decided not to pursue this solution further for the moment.
Risk Analysis
It was decided that further formal risk
assessment should be carried out. NATS Safety Analysis experts have
studied substantial amounts of radar data, evaluating the possible
consequences of level busts involving aircraft passing through or
levelling at incorrect levels, this evaluation being carried out for
various bands of flight levels. Whilst this work is ongoing, early
indications are that the conclusions will enable better prioritisation
of mitigating techniques. The work undertaken so far has already
identified that some classes of level bust are far riskier than others
– some of the riskiest being those involving altimeter mis-setting or
mis-read SID charts. This knowledge allows greater priority to be
given to eliminating these errors.
Mode S Implementation, and sub-scale setting
Consideration of the manner in which Mode S
might help to reduce the level bust risk led to the project group’s
identifying that Mode S should be implemented without delay. Moreover,
it was decided that Mode S selected altitude data could be of
considerably more value, if altimeter sub-scale setting were also a
down-linked parameter. Although the present Mode S plans do not
incorporate sub-scale setting, NATS has expressed a desire that it
should.
London TMA Design
The design of the London TMA was often cited as
being critical to the number and nature of level busts. In particular,
the fact that most SIDs climb under the holding stacks, makes certain
types of level bust particularly hazardous. Whilst it is clearly a
long-term objective, the project group felt that a recommendation to
re-design the TMA should be made, and that any such re-design should
have, as its basis, the need to avoid the types of interaction between
arrival and departure routes that presently exist. Such a re-design
might not reduce the number of level busts, but would certainly reduce
the likelihood of a level bust causing a collision.
Datalink Communications
Datalink communications (CPDLC) provide an
opportunity to eliminate many human errors in the communication chain.
However, other forms of human inter-action with data received by
electronic means will take place, and careful assessment of other
areas of risk will need to be carried out. CPDLC trials are ongoing in
various locations worldwide, and it is hoped that a robust and
reliable form of technology will be arrived at. The project team
agreed that CPDLC had significant potential.
Transition Altitude
The transition altitude in the UK varies between
3000ft and 6000ft, depending upon location. These different values,
and the fact that our transition altitude is relatively low, not only
add to the risk of level busts occurring, but make those which do
occur, more likely to result in an encounter with another aircraft.
For these reasons, the project group recommended the implementation of
a 24000ft Transition Altitude throughout the UK FIRs, in compliance
with the European Single Skies concept. A further working group will
study this issue in due course, and its work will be monitored.
Conclusions
Level Busts represent a significant hazard. Much
of the mitigation relies upon human endeavour – there is no
technological remedy. The education and awareness programme mentioned
above will aim to ensure that everyone exposed to the level bust
hazard has a clear understanding of the problem, and knows how reduce
to their risk. This programme, together with the other solutions
raised by the ACE team, may go some way to improving the statistics,
though greater awareness will also lead to more widespread reporting
of non-safety-significant events. The only true measure of success
available relates to level busts involving losses of separation, which
are almost invariably detected. This statistic will be closely
monitored, and will prove a measure of success of the ACE project’s
work, and other work presently being undertaken elsewhere.
Tim Atkinson, easyjet.
Download this article as a
 PDF file |
