Easy Street
Bowties are used to list the triggers, or threats that may lead to an undesired event (Top Event), and the consequences if the Top Event occurs. They also display the preventive measures in place in order to avoid the Top Event and the mitigations to reduce the impact should the Top Event occur.
The Top Event for the CAA and MAA is the same, "Aircraft in close proximity with another aircraft that their safety is, or may be compromised [Airprox]" That is what we are trying to prevent, because a consequence of that occurrence is a possible collision. TCAS acts as both prevention and mitigation.
You talk about Bowtie being used for probability analysis and predicted occurrences, may I suggest that you use some historical "actuals". Since the MAA started to use Bowtie we have had seven Class A near misses and one collision, so the predictive model does not appear to be working. Focus on preventing the Top Event, then the consequences will not appear.
Why do I believe that the risk is not Tolerable and ALARP? - Simple. The Risk Register shows it to be CATASTROPHIC and REMOTE, which by the H-C and CAA matrix makes it INTOLERABLE. By the way, MAA's definition of REMOTE is highly questionable; many would suggest OCCASIONAL. With regards to ALARP, once again the Risk Register makes it clear. Under the heading "Management and Mitigation Strategies and Controls to Achieve ALARP State" is listed CWS. So if we do not have CWS we do not have ALARP. Besides who in their right mind believes that you can go from not Tolerable and not ALARP in 2011 to being Tolerable and ALARP simply by initiating a programme.
DV