Grounding is the big problem at LF as I am sure you know.Radial systems that would be effective on 40, 80,or even 160 are impossible (100 X quarter wavelength makes the mind boggle)Ground penetration and the extent of the near field means that modest radial systems and short ground rods are virtually useless,and I have tried,I currently have about 500m of buried wire and 30 four foot rods.I recon I could disconnect most of this with no effect.I think that the same would apply to burried sheets of metal unless they were several hundred feet square ! Re. the water table Fresh water is an insulator so no use. In the UK this winter the water table was on the surface,with no measurable difference to loss resistance. I live near salt water and a connection to the sea does help but only a little. For the Transatlantic tests I ran out 3 radial wires of about 150 feet to 4 ft rods in sea water . One helped but the others added nothing. So use the biggest longest system you can find ie mains earth, water supply etc. but dont waste too much time with buried wires! far better to put more WIRE IN THE AIR.I have a theory that the top load should have as large a Footprint as possible.In my case the top load is 10X the length of the vert section.It does not need to be in a straight line and can zig-zag if necessesary. Rik ON7YD has written this up in his "Antennas for 136kHz . see www.qsl.net/on7yd/