You have to be careful with ...

You have to be careful with your source areas, and want to avoid major roads and industrial and commercial sites with large connected impervious areas, or even very busy parking lots at small sites.  We require that infiltration basins have a minimum six-inch liner of a soil filter medium with enough organic matter to slow infiltration rate and sorb metals and hydrocarbons.  These liners will need to be replaced and rehabilitated from time to time due to clogging with fines and organic matter, and to avoid problems with accumulation of metals.  Consider using long grassed swales or other sediment-trapping measures upstream of your basin or drywell. Drywells and similar subsurface injection methods are used for roof runoff and other relatively clean sources; we have some sites that keep roof runoff and runoff from other clean parts of the site separate from dirtier runoff from parking lots, roads, loading docks, etc.  so that they can infiltrate the clean water.   Many of these basins are neglected and tend to clog over time, so that you get low dissolved oxygen in the basin that can result in the release of metals in the accumulated sediment, or from portions of the aquifer affected by the low d.o. plume.  We have seen mobilization of arsenic in an aquifer as a result of low d.o. conditions from a wide range of anthropogenic sources. 

The attached file has some graphs of groundwater data from different sites showing the impacts of infiltration on various parameters; unfortunately it looks like two of the graphs (the first and last) were pretty distorted by the upload, but you should be able to see the general trends.  Let me know if you want more data and I'll try and get you some better files when I'm not in a rush.