In my opinion, the ancient water also receives constant recharge though it is so slow. So, it is difficult to dataing the age of continuously mixing water.
It depends. In the case of the Roman quanat of Metz, ZAC des Hauts des Queuleu (Moselle, France), it was possible to obtain a radiocarbon date on wickerwood within funnel 48. At the Roman quanat of Cutry, La Hache (Meurthe-et-Moselle, France) stratigraphical observation showed that it was build together with a road between 150-300 A.D.
(Hassan 2011, Mays 2010, Ortloff 2009, Viollet 2007 or Garbrecht 1991) said : Those investigating the history of water technology face two major problems: (i) techniques are often difficult to date and published ages are often imprecise and (ii) artifacts or specific sites are not depicted in their entirety in the archaeological record (they are either buried or reshaped). The chronology of the presented examples must, therefore, be treated with caution – which is why we have indicated age and location only on a broad scale, i.e. on the regional and millennial scale.