Track residuals offset from 0 in SciFi layers

Development of FTTrackMonitor ( !2608 (merged) ) has shown that the track residuals in the SciFi are biased, the X1 and X2 layers in each station having opposite biases.

I suspect this bias in tracking could be related to the effects we have been seeing in alignment, where neighbouring layers make large movements in opposite directions even when no original misalignment is applied to the system.

There was also some brief discussion of this topic in the lhcb-rta-reconstruction list (see: https://groups.cern.ch/group/lhcb-rta-reconstruction/default.aspx?RootFolder=%2Fgroup%2Flhcb%2Drta%2Dreconstruction%2FLists%2FArchive%2FTrack%20residuals%20offset%20from%200%20in%20SciFi%20X%20layers&FolderCTID=0x01200200174B6A666AE92D4DB07AD52D621EEE5B&View={E18010BE-A180-4D13-85C5-D1AABEC9C4C1} )