2006 Q2 AVAILABILITY

[signalling survey]

Another attempt, 3/4 hour with reference to material. let me know your thoughts

A collection of material that is not very much focussed on answering the specific question, but a lot of the building blocks are actually there.
In some bits you seemed to be a bit confused and so I modified in "track changes" and added some comments to attempt to clarify and explain.

I didn't do much to target the material better to the actual question though! as for a lot of it this was not really the direction that it was taking.

The first part said "calculated" and that to me suggests an estimate before it is actually built, using base data of components and similar sub-systems and then arriving at a composite figure having evaluated the system architecture, use of reliability block diagrams, fault trees etc.
You seemed to concentrate more on the definition and "measurement" of an existing completed system having a service history. I can't say that you are definitely wrong as I accept there is an element in ambiguity in the question wording, but I don't think you gave what the examiners were really looking for. This impression would have been reinforced because you certainly wandered around related topics- very understandable considering that you are expanding your knowledge and thus writing yourself some notes as you read and assimilated material, but not what you'd want to do in the exam.

I think that for the main part of the question you should have considered how much diversity, duplication, fall-back, override etc could be justifiably provided within such a control centre.
You were right to consider UPS power supplies, probably should also have mentioned diverse communications.

The material under the R A M S headings was good stuff, althoug you did stray too far sometiemes from the pure design elements I think. Also I felt sometime it was too generic and could have been more targetted to the Control Centre that you were supposed to be discussing.

Thinking about what I'd be assuming to be a VDU control system could then talk about:
a) whether there should be a completely duplicated signallers workstation and if so within a separate building.
b) If not, then which parts of a single physical workstation should be duplicated; is a cold standby arrangement good enough?
c) If route setting by tracker ball, do there need to be two of them, or is it adequate that a replacement one can be quickly provided should it fail?
d) Would it be better to provide some diversity another way; e.g. the keyboard which is needed for entering TD dscriptions etc could potentally be used as a less convenient but diverse means of setting and cancelling routes.
e) Does there need to be a completely independent means of performing a high importance action, such as the replacement of signals to danger in an emergency?

In this way you are being far more specific about the scenario in the question, rather than giving information that is rather generic. Also by discussing the effects of particular failures on system availability and safety, you can work into your answer demonstration that you can make an assessment re how much it is worth paying to get that increase in dependency from the system. I think that is where the examiners were hoping that you'd be taking your answer.

Peter, thanks for the feedback, when i was first looking at the question i was divided as to answer by referring to specific issues or look at the whole generic process, it looks like i have chosen the wrong one.