Helicomparator
A number of interesting points arise from this discussion:
JAR-OPS 3.660 rule is objective and therefore needs additional material to provide guidance on methods of compliance. In trying to understand the intent of the rule it is worthwhile looking at the reasons for its presence; it followed the accident to the S61N in the Penzance to Scillies scheduled service which resulted from a (slow) controlled descent into water - resulting in fatalities. This resulted in the requirement for a RADALT and AVAD (an early attempt at an HTAWS for overwater flights) to be fitted for flights over water - this was then migrated to JAR-OPS 3.
The UK regulations were accompanied by guidance material which, for no apparent reason, was not carried over to JAR-OPS 3 – here are excerpts from the UK guidance material (AIL0114):
Whilst most current Radio Altimeters include a height warning light, this is not considered sufficient for alerting a pilot to an inadvertent descent as his attention may be directed away from the indicator at the critical time. It is for this reason that an audio low height warning is considered necessary. The warning must be distinguishable from other warnings and should therefore be a clear and concise voice message.
Research activity has indicated that the characteristics above can be satisfactorily met if the warning format incorporates the following features:
a) A unique tone should precede the voice message. A further tone after the voice may enhance uniqueness and attention-getting without causing undue annoyance.
b) The perceived urgency of the tone and voice should be moderately urgent.
c) The message should be compact as opposed to lengthy, provided the meaning is not compromised e.g. ’One fifty feet’ as opposed to ’One hundred and fifty feet’.
d) An information message is preferable (e.g. ’One hundred feet’). Messages such as ’Low height’ do not convey the correct impression during deliberate descents through the datum height
e) Command messages (e.g. ’Pull up, Pull up’ are not acceptable unless they relate specifically to height monitoring (e.g. ’Check height’).
f) The volume of the warning should be adequate and not variable below an acceptable minimum value.
Every effort should be made to prevent spurious warning.
Altitude Trigger
The height at which the audio warning is triggered by the radio altimeter should be such as to provide adequate warning for the pilot to take corrective action. It is envisaged that most installations will adopt a height in the range of 100 – 160 ft. It will not be permissible for the datum to be altered in flight.
The pre-set height should not be set such that it will coincide with commonly used instrument approach minima (i.e. 200 ft). Once triggered, the message must sound within 0.5 seconds.
The voice warning should be triggered only whilst descending through the pre-set height and be inhibited whilst ascending.
The fact that this advisory material is not present in JAR-OPS 3 has complicated the issues – it was suggested recently that this material be inserted but this was resisted for a number of reasons. Whilst the requirement is perfectly acceptable for over water CAT flight, there are a number of circumstances (such as SAR) where an inhibit switch might be required; it is also the case that the JAR-OPS 3.660 does not distinguish between a RADALT fitted for use over water and one for use over land (those of you who flew helicopters which were first fitted with AVAD will be reminded that, whilst it could be relied upon to provide a necessary warning when flying over water, the call occurred on every approach over land); it is optimised for use over water, not land.
It is also the case that the introduction of other warning devices into the cockpit (EGPWS, ACAS etc.) had drawn attention to the fact that there is no equivalent guidance in AC 29-2B Chg 2 to AC 25.1322 - the integration of alerts and warnings in the cockpit (in fact in the S92 and other helicopters, the EGPWS specifically assumes responsibility for other warnings). The lack of clear general (in an equivalent to AC 25.1322) and specific (in JAR-OPS 3.660) guidance therefore puts the manufacturer at a disadvantage.
Helicomparator is correct that, for offshore operations, the JOEB for the AW139 should have picked this up (as another JOEB did for the S92 – due mainly to the presence of the CAA flight department), it is clear that the construction of the AW139 JOEB invited this oversight; it is not clear that either of the two operational pilots (one Italian and one Canadian) had offshore experience, and this points to a weakness in the system. (Unfortunately, the provision of joint JOEB teams (FAA/TC/JAA) makes it certain that compliance with operational requirements (for EASA/JAA JAR-OPS 3) are reduced in importance alongside the common elements (provision of training courses, STD, MMELs etc).)
The intent of JAR-OPS 3.660 is that there will be a voice message associated with the RADALT; an ‘alert’ will be given with the attainment of a bugged height (ascent or descent) such as ‘check height’, and a ‘warning’ when reaching the ‘preset’ (by the manufacturer) height; such a warning will announce the preset height - for example ‘100 feet’ or ‘150 feet’. Both calls should be preceded by ‘attension tones’
As an additional point, AC 25.1322 makes it clear that all messages should be prioritised and integrated – it is not clear from this discussion that this is being achieved for the aircraft mentioned.
The extension of the discussion, between Helicomparator and Nick Lappos, to the protection modes of EGPWS/HTAWS is welcome. Although the functionality required by JAR-OPS 3.660 has now been introduced to mode 6 of the EGPWS, there is still a need to re-assess the relevance of protection modes 1-5 to offshore operations.
The tailoring of protection modes 1-5 can bring further safety benefits for offshore operations; additionally, if the press release from Honeywell on the further development of SVS is correct, then the addition, and integration, of real-time obstacle sensors into HTAWS/SVS will provide the protection from mobile obstacles that is missing in the current implementation of EGPWS (in the 'enhanced' mode)
Jim