after oil wrote: the fuel (heated first) starts strong with a brand new 10 micron and before half a barrel it slows to a tortuous trickle.
I can't explain why your procedure worked previously and doesn't now, but I'm in the same boat. I'm lucky if one of my Goldenrod or Baldwin 10u paper filters lasts 1/2 a barrel before it clogs up when I filter hot, so I've switch to filtering cold. It is slower, but the filters last at least a barrel.
I've got 25 or so of these 10u paper filters that have been clogged by a 1/2 barrel or less of hot oil. I held on to them because I just couldn't believe they were completely used after 75-100L of filtering. I've tried cleaning them by letting them soak in the hot oil (180-200F) during my initial heating, but I'm not sure if this help much or not. These filters just seem to work again after letting the dry out a bit.
My new procedure is pour the oil through a 100u screen and heat a full barrel of WVO until I'm convinced all the water is gone, usually up to 200F. Then I let it cool and settle for as long as I can, which is usually 1 week, but sometimes I can stretch it to 2. Then I filter cold through a 30u cleanable redline filter and through a 10u paper filter, and into one of my storage totes. The already used filters are still working after a barrel of cold WVO. It takes longer to filter, but I just turn it on and come back in 2 hours.
My theory for my success of my cold filtering over my warm/hot (140F/180F) filter is that the fine sediments that clog the filters are not stirred up by the convection currents of the heated oil. My setup is heated the same way as yours, with a 120V 1500W hot water tank element at the bottom of the tank. I used to do the initial dewatering at 180F, let it settle, then warm again to 140F for filtering. Originally I thought that the fine sedments would fall below the level of the heating element and not be disturbed by the second heating, but but now I'm thinking that the fine stuff does get stirred up by the pre-filtering warming session. Anyway, that's my theory and it seems to be working out so far.