Many thanks for your response.
"I flew many of his test points exploring height loss during go-around, these were mainly from 50 ft in support of the Economic Cat 3 Programme at Bedford. I suspect that the ht loss formula used in JAR AWO stemmed from this work, but it may only be relevant to rad alt operations. I think that ground effect was also considered."
I think we must have have met in another incarnation - I was at BLEU 1965-68, but I think before the economic Cat II (or was it III) was going on. I left the RAF in 1970 and joined what was then the Board of Trade, where we asked RAE Farnborough to develop a software to predict height loss and also asked Pinsker to be the technical adviser. After a month he sent me a couple of graphs and a note saying he would rather do it himself instead of watching someone else screw it up! Those graphs predicted height loss and time to lowest point for one combinations of vertical rate, applied g, period of SPO and damping in pitch. RAE Farnborough wrote a Monte Carlo program based on his graphs and flight test data on distributions of vertical rate and g (plus a correlation of 0.5 between vertical rate and g. The altimeter error was based on the 1968 report of the UK All Weather Operations Committee. Unfortunately this was sent to the Public Records Office and I have had no success at all in getting them to respond to requests for retrieval. It gave means and variances for the component errors for direct dreive, servo and corercted servo altimeters. However I want to update these and cannot find an expert who can take the RVSM spec and deduce the equivalent errors for modern peizoelectric altimeters. If you know of one we would be delighted.
"There is an example of achieving a new system ‘sub-classification’ in JAR-OPS 1 (sub part E) with the ‘super fail-passive’ autopilot... However the method for this approval requires operational demonstration of ‘a highly reliable system’ in service..."
This was the approach the FAA used in the SAAAR criteria (which I have been assisting in converting into the equivalent ICAO criteria for RNP/AR). They defined a quadratic equation and obtained confirmation from aircraft manufactueres that their aircraft systems could meet it. They then published this in the new ICAO Performance Based Navigation Manual as an aircraft approval requirement. Since this is yet to be published by ICAO I include it here:
-8.8*10E-8*(Elevation)^2 + 6.5*10E-3*(Elevation)+50
This is one option, but it will require operators to spend the money on getting the appropriate approval.
"I assume that the benefits sought from the use of modern altimetry would only apply to Cat 1 or NPA approaches. "
Correct.
"Where a pressure source is to be used perhaps the largest altitude error will come from the accuracy in setting the datum, which depends on the human’s ability or ‘desire’ (human factors) for accuracy."
You are of course corect, the setting error can be relatively large and is presumably a bias error - I do not think it was ever considered as a separate component in the SAAAR criteria. PANS-OPS Vol I does contain updated errors though the much smaller ATIS error was included in the SAAR criteria (20 ft).
"... I note that there are quantified reports of significant GPS altitude errors due to external electronic interference."
Would appreciate details if available. I have heard of the long term drift noted by the Eurocontrol RVSM monitoring but not these errors.
"... The current problem of GA altitude loss from MDA (or just above) has been well researched."
Unfortunately we are looking at extrapolating down to 10E-5 probability and that is always difficult from test data (there is never enough to obtain a sensible confidence level). Hence the need for the height loss model.
"... And even though in some instances a GA from MDA infringes current safety margins (PANS-OPS / TERPS differ)..."
We have been having an interesting debate over the TERPS vs PANS-OPS missed approach. The problem with TERPS is the large reduction in horizontal distance between initiation and SOC if the missed approach gradient is increased. The new ICAO RNP/AR criteria is thus based on the PANS-OPS concept, not TERPS.
"... many regulators judge that the benefits of a stabilised, continuous descent with GA at MDA vastly outway the hazards of a ‘dive and drive’ procedure. Thus the benefits (improved safety) come from viewing the overall operation, consisting of the regulations (‘system’ safety), and the operational flight aspects (crew human factors). Such an approach might be required for the altimeter / GA ht loss proposals."
My friend Theo van der Ven at KLM (Adviser to the IATA Member of the ICAO Obstacle Clearance Panel), also member of the TARA group and others, has been pressing this for the last three Panel meetings. He is faced with opposition from those Members who regard MDA as a never go below height. We are still working on this one.
"... I would be concerned if the arguments for a revision crept into changing more than the altitude component (OCH/A) i.e. “we can go lower, we will be closer, thus we can reduce the visibility”. As with many weather related operations there are few such direct relationships and it should be noted that the ‘transformational’ weather conditions (Cat 2, Cat 1, and NPA instrument to visual transition) often result in situations that require a GA or in which, having made a decision to land, it can subsequently be incorrect."
Thisis an interesting point of view. However, from the procedure design aspect, the end product is an OCA/H that will ensure clearance from obstacles at the desired level of probability (for ILS/MLS/GBAS). The DA/H and RVR associated with the OCA/H are operational matters. Hence the problem is one for JAR-OPS (or its Euro successor).
"... Dr Sidney Dekker has written an interesting paper on the problem of small changes, gaps, or assumptions made in regulatory methods, see
Past the edge of chaos. A quote from his paper would be good guidance for any regulator, design team, or test pilot:-
“The most important ingredient of engineering a resilient system is constantly testing whether ideas about risk still match with reality; whether the model of operations (and what makes them safe or unsafe) is still up to date. Helios 522 may suggest that we, in Europe, may still be applying models that no longer are."
Precisely -that is why we are under pressure from manufacturers to update the PANS-OPS criteria to account for "Advanced aircraft and systems" - hence this request for help in finding a definition for same!!!
Many thanks for your input, and interested in any further info.
Robert