Caerwent is lined up along the valley, slightly off the traditional north-south-east-west alignment of most Roman towns. Thomas A. Walker, the chief contractor on the Severn Tunnel construction (Biography), noted in his book on the 1870s' project that*...
The Roman walls remain in fair preservation, and it is believed that when this station was held by the Roman Legions, the tide from the Severn flowed up to the base of the southern wall, and that the rings to which the boats were moored still remain.
Where these tides flowed is now a rough piece of marsh land, through which the little river Neddern passes to join the Severn.
It is hard to imagine any reason for ring fixings on a defensive wall other than for mooring - note that on the 1887 Map the southern wall is identified as the "port wall", though on what evidence is not clear.
In addition, notice on the map that there is a road that leaves the southern wall and travels down to meet the alluvium for no apparently good reason. While this may be Roman (and modern maps of Caerwent do note it as leaving the village's "South Gate"), it may equally be a construction road put it as part of the efforts of John Hawkshaw (Biography), chief engineer of the Severn Tunnel project, to control the waters of the Troggy (Details).
The authors of Kelly's Directory for Monmouthshire dismiss the notion of ships coming up to Caerwent, however they do note the local belief that the name of the village of Crick is derived from "Creek", having once stood on one*.
CHT's Steve Farrar has identified a possible Roman dock system 200-1000 yards (182-914m) downstream from Caerwent's Brook (apparently the linear earthworks are most noticeable in moonlight or the high summer when the differences in ground colour show up*).