I believe there is an issue with the SST alignment tables.  Below are two plots, illustrating.  The left plot shows the misaligned geometry using the (new) AgML misalignment features.  Ladder #2 exhibits a completely unrealistic orientation, but all others appear in the expected place.  There are a few transformations involved to extract the misalignment of the SFLM mother volume from the absolute position of the ladders stored in the DB.  In order to rule out an issue with the new framework, the right plot places a box at the absolute position of the ladders stored in the SST database.  It shows the same issue. 

Figure 1 -- SST in the AgML misalignment framework, timestamp = 20140214 (Valentine's day 2014).  LEFT:  Misalignments are calculated from the SST alignment tables, and applied to the ideal geometry as an additional rotation/translation of the SFLM volume.  RIGHT:  A box is placed at the position of each SST ladder, as determined by the SST alignment tables.  (Note that the axes have autoscaled between the two plots).



The off-diagonal elements in the transformation matrix in the 3rd  column are 10x larger than they are for all other ladders.  This persists in the 2015 and 2016 database as well.

General rotation/translation matrix for the 2nd ladder --

Geometry/sst/sstLadderOnSst
BFC:INFO  - -0.83786   -0.54551    0.02023 |  9.81076
BFC:INFO  -  0.54518   -0.83809   -0.01963 | 20.04990
BFC:INFO  -  0.02767   -0.00541    0.99960 |  0.21342
BFC:INFO  -  0.00000    0.00000    0.00000 |  1.00000

General rotation/translation matrix for the 12th ladder for comparison --

Geometry/sst/sstLadderOnSst
BFC:INFO  -  0.82731    0.56175   -0.00123 | -10.19040
BFC:INFO  - -0.56175    0.82730   -0.00176 | -19.84560
BFC:INFO  -  0.00003    0.00214    0.99999 | -0.03750
BFC:INFO  -  0.00000    0.00000    0.00000 |  1.00000

Conclusion:  Something is wrong with the alignment of (at least) one of the SST ladders.