Before a new Southern Water pipe could be laid between Otterbourne and Southampton, UCL archaeologists dug the route and found five wooden wells still sunk into a natural spring; one was built between 719 and 396 BC.

Five wooden water wells sunk into a natural spring north of Southampton have been excavated ahead of a new Southern Water pipeline, and radiocarbon dating places the construction of one of them between 719 and 396 BC, UCL announced on 5 November 2026.

The wells were found by UCL Archaeology South-East, part of the UCL Institute of Archaeology, during excavation along the route of the Southampton Link Main, a pipe being laid between Otterbourne and Southampton. Four of the five are thought to date to the Iron Age. The fifth is Roman, and pottery from the site shows that the spring was still being drawn on into the later Roman period. The timbers have since been lifted and are being kept in tanks of water at UCL Archaeology South-East’s laboratory in Portslade.

The date of 719 to 396 BC is a radiocarbon range rather than a year of construction. It places the earliest well so far dated in the Early Iron Age, and it comes with a span attached: on the evidence available, the well was built at some point within those three centuries, and the figures mark the ends of that range rather than a single event.

Wooden structures of this age are an unusual thing to meet on a dig, as the excavation team noted. Here the timbers lasted because the ground around the spring stayed waterlogged, and the wet soil starved them of the oxygen that decay requires.

“It is rare and exciting to find wooden structures from this period,” said Rachael Geoghegan of UCL Archaeology South-East, who led the excavation. “The oxygen-starved conditions caused by the wet soil around the spring prevented the wood’s decomposition, preserving the ancient timbers in exquisite detail. We can even see axe marks on some of the timbers.”

Oak trunks hollowed out and slotted into pits

The four wells thought to be Iron Age were built to a single idea. An oak trunk was hollowed out and slotted into a pit. The largest of those hollowed trunks measures 70 centimetres in diameter.

One of the four had been supported by a ring of stakes. Another carries evidence of a working life long enough to need maintenance: it was repaired while in use, with a wooden shovel blade used to support a broken section.

Details of that order are the dividend of a waterlogged site. The axe marks noted by the excavation team, the shaping of the trunks and the improvised repair all depend on the survival of the wood itself.

What links the five is the spring. All of them were sunk to reach the same natural source of water, and the condition of the ground around that source is also the reason the timbers lasted long enough to be recorded.

The Roman well on the same spring

The fifth well was built on a different plan: a shallow rectangular pit lined with timber planks and stakes. The wood is primarily oak, and the builders combined large timbers with roundwood stakes cut from both saplings and older trees.

Hannah Street, a waterlogged wood specialist at UCL Archaeology South-East, said that the team “carefully lifted the Iron Age hollowed logs in as complete a condition as possible”. The Roman well, according to the announcement, required careful excavation to untangle the phases of timber construction within it.

The wells were found in position, still sunk into the spring they tapped, and recorded there. The Iron Age hollowed logs have since been lifted for study, so those timbers are now with the specialists rather than in the ground.

Taken together, the pottery and the radiocarbon date describe a long relationship with one source of water: reached in the Early Iron Age, and still in use in the later Roman period.

Water tanks at Portslade and a £100m pipe

Analysis is continuing at UCL Archaeology South-East’s laboratory in Portslade, where the timbers are being kept in water tanks. Specialists are examining pollen, plant remains and tool marks, and the team hopes that dendrochronology, the matching of tree-ring patterns, may be possible on the Roman well. The announcement reports that work as ongoing.

The excavation took place because the pipe had to be laid. Southern Water describes the Southampton Link Main as a 19-kilometre pipeline representing investment of more than £100 million; those figures come from the company as reported in the announcement, rather than from independent verification.

Phil Maunder, Southern Water’s project manager, drew attention to the contrast between an ancient water source and a modern water project following the same ground, as reported from the company’s release of the find.

The sequence is a familiar one in British archaeology: the route of a new utility is investigated before construction, and the work along it turns up sites that had not been recorded. In this case the route crossed a spring that had been found, and used, a very long time before.

Iron Age builders reached that water with a hollowed oak trunk. Roman builders lined a pit with planks and stakes. The twenty-first-century answer runs between Otterbourne and Southampton in a pipe, and the laboratory work at Portslade is still under way.