There is also the 1853 Arrival from Buxton 2N99 which arrives in the platform the Liverpool working goes from.
Observation says that 2N99 goes on to form the 2121 BPN-Manchester Victoria, as per RTT.
Given you say there are no 'acrobatics' in the morning end. I do think this is a bit of reforming to ensure the units going to Allerton are cycled through so they all get their maintenance. I often end up on that services heading into Liverpool after work and noted it was regularly a 4 carriage working but is now usually only 2.
If, as you say, the 1803 BPN-LIV is now usually a 2-car set, it makes a nonsense of the 1840 BPN-MIA being short-formed, since it means there is no dmu shortage! The 1623 Victoria to BPN could split into the 1803 Liverpool and 1821 Huddersfield. Observation actually says that the rear unit off the 1629 MIA-BPN forms the 1821 BPN-Huddersfield, as suspected.
If you travel on the 1803 again, notlob.divad, could you report the unit number(s)?
The diagrams in Post #3 do not currently show any stock rotation to/from Allerton, although I do not know if any more 156s overnight at Blackpool, in addition to the four that form Diagrams 1, 2, 4 & 5 the following morning. If so, there could be overnight swapping/reforming in the CMD.
There are definitely other 156s which stable at Blackpool, since the units which finish up there as diagrams 3+8 and 6+7 don't necessarily reappear as 1+2 and 4+5 the next morning. About 18 dmu sets commence their day's work at BPN, and this will include a fair few doubles. Some we know will not be 156s (such as 158s for Yorkshire and 142s for the Colne line), but there is still scope for a few 156s to be around, in addition to the four which we know go on to 1+2 and 4+5. I'm pretty sure, for instance, that 5N50, 0504 BPN CMD to Blackburn, is a 156.
Nevertheless, Bovverboy's theory that 2F65 to Liverpool forms part of Diagram 4 or 5, to rotate the unit back to Allerton, seems highly plausible.
No, I suggested that the reason for the swaps was that there was no suitable unit available to form the 1803 Liverpool. It was notlob.divad who suggested the point of the exercise was to get a 156 back to Allerton. Personally I think it highly unlikely that a complete round trip BPN-MIA-BPN would be short-formed for that reason, since there are other ways in which 156s can get from Blackpool to Allerton.
1U86, the 2245 Blackpool North to Airport, needs to be a double, because this set overnights at Piccadilly then forms Diagrams 6+7 the following morning. I agree that either 2N99 from Buxton or 2N93 from Victoria, or both, could provide a unit for this service.
Observation says that neither actually does, since yesterday evening (21/7) neither contained a 156, yet the 2245 BPN-MIA was a double 156 as usual.
Alternatively, a unit could be split from 1N71, the 2103 arrival from the Airport, leaving 1U84, the 2140 to the Airport, and 1N77, the 2330 return working, as 2-car - but perhaps Bovverboy has eyeballed one of these?
Whenever I have seen 1U84 or 1N77, which will be a few times since the 11/7 changes, they have always been formed of a double 156.
I believe the 2N99 Buxton arrival is formed from Sprinters that are stabled all day at Buxton after the morning peak, but I do not know if it normally includes a 156.
Probably not, since yesterday it was formed of a double 150.
The Diagram 9 Windermere unit can be rotated at Barrow overnight, whence it can depart as Diagram 3 or 8 and be replaced by a unit off Diagram 1 or 2.
There are more 156s stable at Barrow overnight than are necessary to provide for diagrams 3, 8, and 9, since, even before diagram 9 existed, the units coming off diagrams 1+2 didn't always reappear as 3+8 the next morning, although they usually did.
I'm sure I have read in 'Rail' that some Cumberland Coast diagrams are covered by 156s, sometimes on their own, and sometimes combined with a 153.
I suggest that the reason why the units off 1+2 usually reappear as 3+8 is simply that they don't need to be split, and there's no point in doing that without good reason.
RTT for 13 July shows a curious anomaly that I am at a loss to explain.
1N25 from Barrow arrived at Preston Platform 5 at 10:35, 12 late. But
1C72 then departed for Windermere at 10:32:30, only 3 late. So the reversal, for which 6 minutes is nominally allowed, actually took minus 3 minutes - Northern seems to have invented time travel!
Both the arrival and departure times at Preston seem reasonably consistent with those at preceding and subsequent timing points, so this cannot be explained away as a recording error. And RTT showed no STP/VSTP ECS workings to swap the unit.
I notice that 1N25 ostensibly managed to lose three minutes between Lancaster and Garstang & Catterall (which, in itself, seems suspicious) and a further three between there and Preston. If, in reality, it didn't lose any time at all, that would leave three minutes turnaround time at Preston, which is about the time you would expect a turnaround to take, given that the incoming train was late.
Unfortunately since the full schedule has disappeared from RTT I can't now investigate further (conflicting movements, etc).