Water stored underground was better protected and more hygienic, and therefore more suitable for consumption. Drawing water was done sometimes in a similar way as drafting water from a well, but through a shaft.
They were predominantly brick-built (with some exceptions), and the interior was divided into bays. Supports were spaced generally 204 m apart; the roof supports were almost always columns. Building subterranean had another advantage: the surrounding earth contained the outward forced exerted by the water within. If a part of the cistern was above the ground, that part would need to be reinforced with extra buttresses. Sockets above the captials in the cisterns attest to the use of wooden tie-beams between the supports, possibly to ensure that the structure would move uniformly in an earthquake. Water in underground cisterns also needed to be areated. This could be done by inserting windows, which also provided a suitable overflow. The walls wee plastered to just above the height of the opening. Many subterranean cisterns possibly also served as the substructures of buildings (Crow et al., 2008, p. 137). The columns in underground cisterns could be recent manufactures, as well as reused spolia (mostly salvaged from damaged/destroyed buildings after i.e. fires and earthquakes). The majority of capitals (re)used date back to the 4th-6th centuries, when the Proconnesian marble quarries were in operation (Crow et al., 2008, p. 138).
Buondelmonti commented how many enclosed cisterns were ruined and had been converted for the growing of vines. Furthermore, by the Palaeologan period, many cisterns had lost their names (Crow et al., 2008, p. 22).
A consequence of out of order cisterns in decay was the accumulation of rubbish and sediment, reducing the original capacity of the citern, or hindering its entrance (Crow et al., 2008, p. 125).