Jack
I don't think anyone is having a go at ATC. A number (including me) are having a fair crack at the system that you work within, but certainly have not meant to criticise the controller.
I had dinner with an overseas based 747 captain on Saturday night. He says the only time he flies a NPA is in the SIM or when he comes to Australia. While the points about pilots should have the skill is correct, the plain fact is that many foreign pilots coming here are simply not fluent in them. He also says that the FCS of glass cockpit 747's don't deal as well the transition to NPA as the 747 classic (for reasons I don't understand). Whether or not this translates to the 777, I don't know. But its an interesting question. His comments make me wonder if modern digital FCS are optimised for NPA approaches.
My suspicion is that the problems with the approach started in the LH turn from the 11 DME arc to the VOR 346 radial, then developed.
The second point is that while the DME Arc entry VOR approach may be a good valid procedure, why should we not move with technology and strive to have systems that benchmark with other international airports in the region? Australia used to lead the world in aviation technology. Now countries that we label as third world have better, more modern infrastructure.
My beef is that the ATSB report has been used to criticise the airline flag carrier of one of our major trading partners (#7 from memory). Yet this report would fail as an undergraduate report. I count 9 discrepancies between the schedule of events in appendix B and the body copy. When I try and derive a groundspeed to fill in missing DME blanks, I get numbers that vary from 120 kts to 600 kts. There is something not quite right about the published event chronology. The crew was criticised for not reacting to the go-around instruction, but when you compile a full list of radio calls (must add the body copy reported calls to Appendix B) then it really looks like they did not hear / did not understand the first call. The report does not give this impression, but I think there was a combined total of 10 radio transmissions regarding the go-around. An initial instruction with either the response "Copied" or "Climbing"(depending on which part of the report you read), then a 35 sec break, then a further 4 transmissions & responses between 2012:26 & 2019:47. One of my only real questions of the controller, is why in the face of no positive reply of a go-around let alone the correct phraseology, did he wait 35 sec to re-issue the instruction? If we are going to criticise the crew for not affirming a command, isn't there some shared responsibility? Just asking.
If the crew was under the expectation of ICAO rules (whether rightly or wrongly), then they were operating legitimately (albeit with sloppy airmanship) from the point at which they were cleared for visual approach. At the point of actual go-around they were inside the circling area and were therefore OK according to both Australian & ICAO standards. This difference to ICAO is not mentioned in the ATSB report at all. I read the Jepp manual looking for the Australian requirement to be within the circling area and I don't think its at all clear in the Jepp document. I don't think you can blame them for not being clear.
The point at which it is unequivocal that they had breached the LSALT was during the LH turn from the DME arc to the VOR radial. This was up to 30 sec before they were cleared for visual approach and a full minute before the first go-around instruction.
This report does not mention any history of altitude infringements to RWY 34, but the active report AO 2012-120 into a US registered B747 cites that there have been 5 altitude infringements to RWY 34 in 2011/12. So, it looks like there might be a bit of an issue.
Tiger airways report (AO 2011-070) of June 2011 (vs Thai in July 2011) was another altitude infringement issue (RWY 27). It pointed to a discrepancy in operational data between the published charts and that in the flight management system. This was not considered in the recent Thai investigation. 2 incidents, a month apart, same airport, both with involvement of the FCS. It would seem to me that it should have been on the list to consider.
There is no denying that this was a pretty bad approach, but using the crew as whipping boys doesn't help explore how to improve the situation.
We should debate whether all runways should have ILS
We should debate if we should bring the cleared visual approach requirements into line with ICAO
We should debate whether a redesign of the VOR approach would make it easier for foreign crews
We should debate whether the RWY 27 ILS might be an option for longhaul foreign crews with light crosswind (about 10 kts on the night).
The ATSB report does not mention if this flight had relief crew. The BKK - MEL flight runs about eight and a half hours to nine and a quarter hours, which I understand is borderline for a relief crewmember. I understand some airlines do, some don't. It would seem to me that fatigue should probably have been discussed in the ATSB report as well. I have no idea how to weigh up the risks of landing from a non stabilised approach vs a go-around at night at the end of a 9 hour flight, but it would be an interesting discussion.