I'd venture to suggest that you will have a pretty reliable indication of this situation when you're halfway through the reversal turn. I certainly watch the needle movement through the turn, to get the earliest possible indication of overshoot, or even undershoot. For an overshoot, I then have the flexibility of going to 30-40 degrees of bank quite safely at this stage, to minimise the amount of overshoot. I know that some companies limit bank angle to 25 degrees, but I am not so limited.
If you're prepared to break the model of maximum 3 deg/s turns, then I think the problem is significantly mitigated. At 150 kt a 3 deg/s 180 deg turn has a diameter of 1.6 miles. Thus at 5 deg/s a 180 deg turn has a diameter of about a mile, and gives you 0.6 mile lateral correction that can be applied within the turn. If my trig is right, that's 34 degrees bank angle for 5 deg/s if 22 degrees gives you the 3 deg/s, which doesn't seem unreasonable.
For the pedantic like me, is that permitted by PANS-OPS? It says that procedures
are based on 3 deg/s (25 deg max bank) and that holding turns
are to be made at that rate. It doesn't mandate that rate for reversal turns.
Perhaps for those of us without Ops Manual limitations, tightening the turn is the pragmatic answer. But I'm not looking forward to explaining that to my examiner...
This whole argument about tracking outbound in a severe crosswind is the exact same as doing say an NDB/DME base turn when there is say a 50Kt tailwind on final approach........would anyone seriously considder exceeding the outbound DME to provide suficient time to descend on final approach to circling minima?
A good analogy, DFC, but I think the problems are less pronounced along-track than across-track. Most teardrops (in the UK at least) seem to provide for completion of the reversal at least a mile before the FAF. Thus, although things happen more quickly with a tailwind, there shouldn't be a problem with being in a position to descend with the procedure at the FAF.
By contrast, someone performing an NDB approach to an on-airfield beacon who rolls out of the turn a mile off the FAT and, say, at 6 DME with a 5 DME FAF, is 10 degrees off the FAT, and needs to take a 25 degree upwind cut to get to 5 degrees by the FAF. Add to that the probably 15+ degrees of drift and it starts to conform to my vision of "pear-shaped"
For an ILS approach where a LOC is involved, it's worse because half-scale deflection is much closer to the FAT.
I wonder if we're starting to get too technical in responding to the initial enquiry?
Pack2's question was about "permitted procedures". I think it's within scope to discuss the wording of PANS-OPS, its interpretation, the consequences of different interpretation for risk management and pragmatic solutions. (But then again this is PPrune, so when has the scope of the original question ever mattered a jot...
)