CRM

28.10.2004

Submission:

Skyguide is designing a new ILS CAT I approach procedure to a major airport in Switzerland. There are a large amount of obstacles in the final part of the approach. In Switzerland the determination of the OCA/H for an ILS approach is usually done with the ICAO Collision Risk Model (CRM). We do not use a stand-alone CRM software but one which is embedded into a PANS OPS software.

When using more than 785 obstacles in the CRM run the software does not take into account all the obstacles. Some obstacles are just ignored during the run - without any error message!

Therefore I have to reduce the obstacle input in order to have less than 785 walls/spikes. Is the assessment of the effect of obstacle density still sufficient when reducing the amount of obstacles? Is there any risk to get a false OCA/H or false risk when reducing the obstacle to 785? How should the reduction be done? Did anybody else encounter similar problems with the CRM?


Romano Germann, Skyguide, Zurich

Answer or Commentary:

(I.W.

I am surprised that the ICAO CRM just ignores some obstacles. If the obstacles are outside the precision segment the CRM will not analyse them but it will advise you of the fact in the report document. If your CRM version accepts an obstacle file (i.e. a "???.obs" file), I suggest that you open the file with notepad and check if the missing obstacles are in the file. If the obstacles are not in the file then it is the fault of the software you use to create the "???.obs" file not the CRM. If on the other hand, you find the ignored obstacles in the "???.obs" file please send all the relevant data to me and I will check it against the version of the CRM that I have.

Obstacle density:

Ideally one should have as many obstacles as possible. However previously (and still in some countries) the CRM was run after manually entering the obstacle data. In those days (when the vast majority of ILS approaches world wide were designed) even the most fanatical procedure designer would not sit down and enter more than 50 -> 100 obstacles. So I think that it is safe to assume that 785 obstacles should cover the obstacle density case. The risk in reducing the obstacle number is that you may miss the critical obstacle. I personally find that entering only the obstacles contained within the OAS areas is an efficient method. As obstacles outside this area will be outside the precision segment.

Has anyone else any thoughts on this matter)