I have a shallow well, which was built before the draconian Colorado water laws were put in place. It slowly recharges, due to a heavy clay soil, but it would be perfect for some establishment watering and annual vegetables, especially if combined with a gravity pressured holding tank and a solarpump, which would only run while the sun was shining.
However, I am in Lakewood, CO, near Denver. There is an EPA superfund site only five miles away which contaminated the groundwater. We are surrounded by a suburban landscape full of traffic, sewer lines, and people spraying chemicals. So the chances of the water from a shallow well being clean are pretty slim.
What if I used the above mentioned pump to run the water very slowly through a plastic lined trench, with a section of biochar, a section of limestone, and a reed bed? Would this work? Would I need all of these components? I would be hoping the biochar could catch heavy metals, the limestone could absorb petrochemicals, and the marsh could degrade sewage organisms and pesticides. I would, of course, test the input to see exactly what I had, and the output to see if my plan worked, before pouring it on my gardens. But I don't want to do expensive testing if there is no biological way to clean it up. (I think the flow would be too low for conventional filters. Besides, those things are expensive.)
Phytoremediation (the reed bed) is good, but mycoremediation is better. If you could trickle the water over a bed containing active mushroom mycelium, and then run it through the biochar filter, it will probably be pretty clean.
Fill your plastic lined trench with wood chips and get it well inoculated with mushroom mycelium. If you drip the water in one end and let it flow through to the other end, where the biochar filter is, I don't think you would need the limestone.
Go for it!
Rocket Mass Heater Plans: Annex 6" L-shaped Bench by Ernie and Erica