A publication of the Archaeological Institute of America
How a Roman Aqueduct Works
Volume 65 Number 2, March/April 2012
Ancient aqueducts were essentially man-made streams conducting water downhill from the natural sources to the destination. To tap water from a river, often a dam and reservoir were constructed to create an intake for the aqueduct that would not run dry during periods of low water. To capture water from springs, catch basins or springhouses could be built at the points where the water issued from the ground or just below them, connected by short feeder tunnels. Having flowed or filtered into the springhouse from uphill, the water then entered the aqueduct conduit. Scattered springs would require several branch conduits feeding into a main channel.
If water was brought in from some distance, then care was taken in surveying the territory over which the aqueduct would run to ensure that it would flow at an acceptable gradient for the entire distance. If the water ran at too steep an angle, it would damage the channel over time by scouring action and possibly arrive too low at its destination. If it ran too shallow, then it would stagnate. Roman aqueducts typically tapped springs in hilly regions to ensure a sufficient fall in elevation over the necessary distance. The terrain and the decisions of the engineers determined this distance. Generally, the conduit stayed close to the surface, following the contours of the land, grading slightly downhill along the way. At times, it may have traversed an obstacle, such as a ridge or a valley. If it encountered a ridge, then tunneling was required. If it hit a valley, a bridge would be built, or sometimes a pressurized pipe system, known as an inverted siphon, was installed. Along its path, the vault of the conduit was pierced periodically by vertical manhole shafts to facilitate construction and maintenance.
Upon arrival at the city’s outskirts, the water reached a large distribution tank called the main castellum. From here, smaller branch conduits ran to various districts in the city, where they met lower secondary castella. These branched again, often with pipes rather than masonry channels, supplying water under pressure to local features, such as fountains, houses, and baths.
Rabun Taylor is associate professor of classics at the University of Texas at Austin.Share