Measurements are made between signal cables in the DOM main cable and the over-all shield immediatly after deployment, before the water in the bore hole freezes. Usually the leakage currents driven by an 18V battery are unmeasurable.
Measurable currents suggest that water with ionic content has formed a path between the shield and a signal conductor where the conductor is spliced. (Splices occur in the cable where break-outs for DOM connection are made.)
The snow and ice which is melted in the drilling processs is very pure, much like deionized water. Interestingly, the absence of metal ions in the water makes it a more agessive solvent. The water picks up Iron ions by attacking the cast iron water heater chambers which operate at elevated temperature and pressure.
Water analysis carrried out on samples taken in January 2007 shows that the iron content of the water in the bore hole is roughly ~0.1 grams/cubic meter, based on samples collected in tanks which receive water pumped from the top of the bore hole.
No other ions or inorganics were reported in the analysis. (Either the traces were below the measurement floor, or were not requested in the analysis order)
Weight: 55.85 grams/mol
1/2Fe2+ conductivity: ~54m2S/mol
1/3Fe3+ conductivity: ~68m2S/mol
[CRC Handbook of Chemestry and Physics, 72nd Ed.: P. 4-64, 5-96]
(For reference, salt, NaCl, has conductivity ~126 m2S/mol)
Let's assume a leakage path 0.1mm2 x 1cm long.
[(m3/0.1g)(55.85g/mol)/(54x10-4m2/Ωmol)](1x10-2m2/10-7m) = ~1x1010Ω
18V/10-10Ω = ~2nA
High leakage current cannot be blamed exclusively on water paths with the reported ionic iron contamination, or the foregoing analysis is flawed.
If the analysis is not fatally flawed, then other sources of leakage account for high leakage currents measured post deployment/pre freeze-in.