After many experiments with solar filament driers, the best solution has been an enclosed thing on the bed.  The filament goes on the bottom.  A tub of calcium chloride goes on top.  There's a piece of cardboard to keep the desiccant from getting hot.  It's creating a distillation column.  The filament gets hot & the water condenses in the desiccant.  It seems to require a bed temperature of 100C for 1 hour with degraded levels of effectiveness as the temperature goes down to 70C.  

Others have tried circulating fresh air in an open container, but it's too humid.  An improvement might be a fan in the enclosed container.

It's believed the solar filament drier wasn't effective because it didn't get hot enough.  Below 100C, TPU doesn't release any water.  PLA & PETG did better in the solar drier.  They might do better at 70C in the bed drier.

An improvement on this would be gcode to turn it off after 1 hour.

The trick with this is in hot weather, you don't want to run the bed heater at all & a solar drier becomes a better proposition.