One might assume that the ground has shifted -- and that the run is damaged enough to create an open wound.
This should progress until the conductors are faulted short.
I'd strongly suspect that the original transformer pad was set upon uncompacted soils.
Out my way, the Pocos have been so chronically burned from such fiascos that they now are hypervigilant WRT base compaction.
After the inevitable repair -- perhaps years from now -- it would be advisable to grout the secondaries in slurry. The run can't be too long, and its now plain that you've got a serious water table issue.
The asphalting of the interior tells us that there is a serious need for French drains around the exterior footing, too.
In the meantime, figure on the CT box being trashed by rust.
It's imperative that remediation occur before catastrophic failure, perhaps this Summer?
Tesla: There have been a few attempts to 'fix' this leak over about 1 to 1-1/2 yrs. The story seems to have started by someone hitting the padmount xfr and shifting is. The POCO raised it and the PVC was cut and replaced (3' down).
That did not help; the POCO UG crew injected epoxy stuff thru the CT end (HaHa)...still leaking. The xfr end was duct sealed, spray foamed & whatever....still leaking.
Then ditch down (manually) near the perimeter foundation...
And install a French drain/ rock pocket down and below the feeder entry...
With a sump pump and float switch...
Itself powered by a GFCI circuit off of the 'house' panel from some above grade location and a Bell box with an in-use cover...
All to be done before the CT cabinet gets grossly rusted out in the near term future.
Because of the sump pump's location, some provision would be required for continued access...
I'd recommend using the super robust PVC pipe used for fresh water delivery, the bluish-green stuff...
Which would be vertically jammed down into the fresh hole...
With weep holes all around drilled in...
So that a common 120V sump pump can be lowered from above by rope...
Powered by a cord connection above grade...
With the ejected water flowing through a flexible pipe/ garden hose, into a common drain/ rain leader reasonably far away from the problem zone.
The rest of the backfill would be 3/4" crush-- with no fines -- and some earth screening wrap to hold back native fine materials -- keeping the zone highly porous.
Out my way, anything that permits rainfall to enter a building -- even in -- especially in -- the basement is prohibited by the building code.
It's a red-tag event for new construction or re-hab.
If freezing weather is a risk, then some sort of insulating cork will have to be inserted above the drain portal so that the 'well' doesn't freeze -- and ruin the sump pump.
To make sure that water doesn't flow by this set-up, it is necessary to CAREFULLY ( i.e. un-energized ) port the feeder conduits with some weep holes WITHOUT damaging the conductors nor providing rodent entry.
The only other option would be to ditch back along the PVC run, creating a French drain all the way back to the transformer pad.
The stuffing placed up at the transformer pad is useless.
The water is coming in from the surrounding soil.
That it's getting past any foam 'cork' at the CT cabinet tells you that the hydraulic pressure is at least 20" of water column.
I have to ask: were there no bollards to protect the pad mount transformer?