Engineering Rome

The Water System of Ancient Rome

Written by Artemis Zafari, 12/06/2019. All images are created by the author unless otherwise noted.

Figure 1. The still-standing arcades of the Aqua Claudia, one of Rome’s ancient aqueducts.

It is impossible to discuss the glory of ancient Rome without including its complex water systems featuring baths, fountains, latrines and more, all supplied by the famous aqueducts. Today, access to clean water is taken for granted by millions across the globe. But this was not always the case. In fact, it was only around 470 B.C.E. when the Greek doctor Alcmaeon of Croton first theorized that people’s health may be influenced by the quality of their water [5]. Since then, civilizations like the ancient Romans have made great advancements towards improving reliable access to clean water. This report will outline the history of the water systems in Rome, from the initial use of the Tiber river through the construction of the aqueducts. This paper will also explore the variety of inner-city water distribution methods and uses once it arrived in Rome.

History of Roman Water

Figure 2. The iconic Tiber river, a key component of Rome’s advantageous founding location.

According to legend, Rome was founded by the brothers Romulus and Remus in 753 B.C.E. [6]. Rome’s location provided two key advantages: its seven hills made city defense more manageable and the Tiber river supplied a steady source of water. The first water-related project in Rome was likely the Cloaca Maxima, or the Great Sewer. The Cloaca Maxima was a drainage canal that began construction in 600 B.C.E. at the order of the fifth king of Rome, Tarquinius Priscus. Priscus’ intention was to drain the flood-prone area between three of Rome’s hills (Palatine, Esquiline, and Capitoline) which would later become the Roman Forum [7]. This area was originally 20 feet below sea level and flooded annually by the Tiber, but under Priscus’s guidance the basin was filled with soil and debris until the ground level rose by 30 feet. The surface was then paved in order to allow for the construction of the main canal, which would convey flood waters into the Tiber in order to prevent erosion in the Forum. As the city expanded over time, additional canal segments were frequently added and modified to fit the needs of the growing populace. Eventually these canals were covered to allow for structures to be built above them, creating the sewer network that is still in place today. The main outfall of the Cloaca Maxima into the Tiber river is still standing in modern-day Rome; a testament to the ingenuity of the first Roman civil engineers.

Figure 3. A modern photo of the Cloaca Maxima. It now serves as a covered shelter for the homeless community. Photo from Jeff Bondono [13].

While Rome’s initial water sources consisted of local wells and cisterns near the city, the needs of the growing population soon required a larger, more consistent supply. This is where the famous aqueducts came into play. The first order of business was to locate a reliable water source within a reasonable distance of the city. This was a sort of pseudo-science; the ancient Romans did not have advanced methods for checking water quality so they had to use more qualitative measures. Marcus Vitruvius, a civil engineer and architect, wrote about some of the techniques they used. He described the process of looking for plants in the vicinity of potential water sources, speaking with local inhabitants and observing their health, and visually judging the nearby rocks and soils [1]. Even with these precautions, the water quality from the aqueducts was not always perfect. Water sources with clay soils were often poor due to the inability to filter out the clay particles and storms in the countryside could cause the incoming water to be turbid [4].

Figure 4. A map of Rome’s aqueducts, showing their origins (where their water source was located) and paths into the city. Map from [4].

The actual process of constructing the aqueducts consisted of building intakes to catch groundwater from the source, digging tunnels and creating bridges to transport the water through the majority of its path, and distributing the water once it reached Rome. There were several different methods of obtaining groundwater including well intakes, infiltration galleries, and river intakes [3]. Well intakes consisted of rectangular chambers which had water supplied from numerous splits and openings and discharged into one outlet (which would become the aqueduct). Infiltration galleries were 20 – 100 meter long sections which ran alongside a hill and intercepted water flow. The water would trickle into the gallery through small splits in the wall and collect in a settling basin, which helped remove debris and sediments. River intakes consisted of diverting a clean river into two separate channels using dams, with one of these channels feeding into an aqueduct. River intakes were rarely used as aqueduct sources in ancient Rome due to the difficulty of finding suitably clean rivers.

Figure 5. The typical components of an aqueduct. Illustration from [4].

After the water was taken from the source through the various methods explained above and given time to sit in a settlement tank, the aqueduct would begin. The aqueducts contained different segments depending on the specific needs of the path chosen for the aqueduct. These segments included covered trenches, tunnels, bridges, and arcades [4]. Contrary to popular belief, most of the aqueduct lengths were underground. The arcade portions of the aqueducts, with their iconic arches and elevated flow paths, only consisted of around 12% of aqueduct lengths [2]. Based on the path of a specific aqueduct (from its water source to Rome), different combinations of underground and aboveground water transportation methods were necessary. In general, the aqueducts were powered by gravity and had serpentine paths similar to rivers; they would twist around mountains and hills and find paths that made for the easiest construction. If it was not possible to navigate around an obstacle, then tunnels would be used to dig through the barrier. To ensure that the aqueducts followed their designed paths, the Romans used basic surveying techniques and tools. The most common surveying tool was the groma, an instrument that comprised of a vertical shaft with a horizontal cross-piece on top. The cross-piece had plumb lines hanging vertically at four ends, each making a right angle with the adjacent side. The groma would first need be stabilized on the ground and aimed in the needed direction. A helper would then step back a certain distance and, guided by the surveyor using the groma, place a pole to serve as a guide for the desired alignment.

Figure 6. A graphic showing a groma and how it would be used in the field. Image from [14].

The slope of the aqueducts ranged from 0.07% to 3.00%, with an average slope of 0.20% [4]. There is a relatively wide range in slope because different segments of the aqueduct required different water speeds. The slope was critical because if the aqueducts were too steep, the fast water flow would cause damage to the building materials and degrade them over time. If the slope was not steep enough, the slow water flow would lack the speed to make it past the siphons . The following sections will further delve into the construction methods behind the various subsurface and above surface segments of the aqueducts.

Figure 7. The view from inside an aqueduct tunnel.

Construction: Tunnels, Trenches, and Pipes

The aqueduct tunnels were built following an ancient Persian technique called qanat [2]. This consisted of digging shafts (putei) at consistent horizontal intervals, normally around 230 feet. These shafts would be dug down until they reached a desired depth, then workers would begin excavating laterally until they connected with an adjacent shaft. Using this method, the Romans were able to connect all the shafts they needed in order to create a continuous path for the aqueducts. Cranes using pulley systems were then able to carry out excavated material and lower building materials into the tunnels. The shafts also served as maintenance holes in the future, allowing the Romans to inspect and repair the aqueducts if there were ever any issues.

Figure 8. A visual of the quanat method of digging and connecting vertical shafts to construct a continuous underground tunnel.
Figure 9. This is a photo looking up into a maintenance hole, while standing in the aqueduct tunnel. Some indents in the rocks are visible; these were created so that the workers could climb in and out of the shafts.

Once the tunnels had been excavated, the Romans then needed to install the proper structures necessary to keep the water flowing and sanitary. This consisted of a foundation and footing beneath the floor of the tunnel, a wall along the sides, and an arched vault along the top [4]. After these elements had been constructed, the Romans would then add a waterproofing mortar along the floor and sides of the tunnel. This prevented the water from permeating through the walls of the tunnel, which would degrade the material overtime and reduce the quality of the water. Even with these preemptive measures, minerals in the water would attach to the sides and floors of the aqueduct channels. This accumulation was referred to as sinter and most commonly consisted of calcium carbonate. The Romans were aware of this and conducted regular maintenance to clear the sinter from the channels and ensure the water quality was kept as high as possible. Workers would divert the flow of the aqueduct into an adjacent channel, effectively creating a bypass, and lower themselves into the tunnels using the same shafts that were used to create the conveyance path. Once in the empty channels, they could properly chip away at the sinter and restore the aqueduct to its previous quality.

Figure 10. This photo shows the layer of waterproofing mortar that the Romans used inside the tunnels. The layer degraded over time until a cross section was exposed.

This maintenance technique was effective for the large channels, but different techniques were needed for pipes since the workers could not fit inside. In these cases, workers would create a makeshift pipe cleaner by balling up rags and attaching them to the end of a chord which would then be pulled through the pipe [4]. If the sinter had accumulated too much and caused irreversible damage, then the pipes would have to be replaced. Consistent maintenance was important because if the sinter was allowed to accumulate, the cross sectional area of the channels would decrease over time. This would then cause the speed of the flow to decrease due to increased friction with the sinter’s surface.

Construction: Bridges and Siphons

Bridges were necessary when the aqueduct needed to pass over a valley, river, or other similar obstacle that required an overpass. Siphons were used when the obstacle was too deep or wide to be covered by a bridge.

Two key elements of the Roman bridges were their uses of pozzolana cement and the arch [8]. Pozzolana was a type of slag that formed naturally from volcanic rock. It was a natural cement that the Romans used to make their concrete, allowing them to create strong mortar for the supports of their bridges. The mortar acted as a glue between the building pieces of the bridge; it ensured a tight seal and equal distribution of pressure between connected pieces. Two advantages of pozzolana cement were that it grew stronger over time and it was ecologically cleaner than the cement mixtures used today [8]. The arch allowed Romans to take advantage of the superior compressive strength of their stone building materials. By stacking trapezoidal stones called voussoirs in the shape of an arch (held together by the crucial keystone in the center), the weight of the bridge was used to compress the tapered stones together. The resulting pressure created a “locked” mechanism in the arch that required a large amount of force to rupture, essentially creating a very secure supporting structure. By using multiple arches in alignment, Roman bridges were incredibly stable and many are still standing today (like the Alcántara Bridge from 104 C.E.) [8].

Figure 11. A diagram showing the fundamental components of the Roman arch: the voussoirs and the keystone. Image from [15].

Sometimes, such as when constructing bridges over bodies of water, it was not possible to construct the piers of the bridges on land. In these situations, the Romans used cofferdams. Cofferdams provided a temporary dry area in the middle of a body of water. The Romans constructed these by digging a ring of timber logs into the riverbed. Then, the gaps between the logs would be filled with clay in order to create a watertight seal. Once all the gaps were filled the water inside the ring was then pumped out. Now that the riverbed was dry, the Romans could construct the bridge piers using pozzolana and stone as before. After all work was done the logs were removed and the piers stood firmly in the water.

Figure 12. A ring of timber logs that would be dug into a riverbed to create a cofferdam. Image from [16].

When aqueducts needed to pass by a valley that was too deep or wide for a bridge, siphons were used instead. These siphons contained three main elements: an initial distribution tank, a row of lead pipes moving from the tank through the valley, and a receiving tank on the other side of the valley [9]. The distribution tank served as a transition between the open channel of the aqueduct into multiple lead pipes. These pipes had small diameters and were normally laid parallel to each other in a row. It was essentially to keep the pipes fully watertight to prevent leaks and air-bubbles within the system, which would cause the siphon to fail. Furthermore, the pipes had to be strong enough to withstand the high static and dynamic pressures due to the steep descent of the siphons. The receiving tank also needed to be lower in elevation than the distribution tank in order to provide enough head loss to maintain a functional hydraulic grade line.

Figure 13. A simplistic visualization of a siphon.

Construction: Arcades

When people think of the Roman aqueducts, they oftentimes envision the arcades. These were series of arches supported by columns that carried the flow channels when the water needed to be elevated above ground [4]. Each arch’s lateral thrust was supported by its neighbor, so these were essentially long spans of arches that were using each other’s weight as support to stay standing. They were used to convey the water in the plains around Rome where the natural dips and rises would have caused the waterline to be unsteady. Instead, with the arcades, the Romans were able to maintain the steady slope they needed to consistently deliver water to the city.

Figure 14. The remaining arcades of the Aqua Felice.

The materials used to build the arcades included stone blocks, concrete, mortar, tiles, and bricks [2]. Wooden scaffolding was used during construction to allow the workers to put the arcades together piece by piece. The scaffolding held the weight of the arcades until the final piece of each arch, the keystone, could be placed. Depending on how high the arcade needed to be, the Romans would stack multiple layers of arches on top of each other (although they rarely exceeded three layers). Massive pillars, measuring around 10 feet by 10 feet, were used at both ends of the arches in order to support their full weight. These pillars would often increase in size towards the base, giving the structure more resistance against tipping over due to the arch loads.. Finally, the water channel (specus) would be placed on top of the arcades. These were made similarly to the subsurface tunnels, with waterproofing mortar and vaulted roofs [2]. Sometimes, if multiple aqueducts were traveling near each other along the same path, the Romans would stack channels on top of each other in order to prevent the need to construct an entirely new arcade.

Figure 15. A drawing showing the scaffolding and construction framework that was necessary to build the arcades. Image from [17].

Now that the unique components of the aqueducts and how they were generally constructed have been discussed, this report will highlight one specific aqueduct: the Aqua Appia. The Aqua Appia was chosen to be highlighted because of its historical significance as the first aqueduct constructed by the Romans.

Aqueduct spotlight: Aqua Appia

Figure 16. A visualization of the Aqua Appia in ancient Rome (located adjacent to the Aqua Marcia). Image from [18].

As mentioned, the Aqua Appia was the first aqueduct built in ancient Rome. The need for the aqueduct rose from the fact that the wells and springs around the Tiber river were no longer adequate enough to meet the growing needs of the city [10]. The aqueduct began construction in 312 B.C.E. under the guidance of Appius Claudius Caucus, who was one of the two censors at the time. A censor was a civil officer who was responsible for supervising public morality and overseeing government works. Appius Claudius was already working on the Appian Way (one of the first ancient Roman roads), so he decided to take on the aqueduct project as well.

The source for the Aqua Appia was approximately 24 meters below ground level, at a series of springs discovered by the Roman statesman Gaius Plautius Venox [10]. The total length of the aqueduct, from its source to Rome, was around 10 miles. It took several years to fully complete the aqueduct and it was almost entirely underground, with only 0.1 miles of arcades residing above the surface. To finance the project, money was furnished both by public and private sources through the treasuries, town councils, and citizens. The cost of the Aqua Appia is estimated to have been around 400,000 sesterces, which is equivalent to approximately $1,200,000 in today’s currency.

Although the Aqua Appia was an incredible feat of engineering, it was not without its faults. The aqueduct developed leaks over time and required consistent maintenance. It was also frequently targeted by Rome’s enemies as a means to cut off water supply to the city. Regardless, the Appia was monumental as the first aqueduct and paved the way for more advancement in Roman water engineering.

Water Distribution and Uses in Rome

To fully discuss the extent of Rome’s water uses, it is important to first understand the way Roman citizens lived. This is because different citizens could have different water accessibility based on their societal positions and wealth. In ancient Rome, only the wealthy and powerful could afford their own homes (a domus). The majority of citizens in the middle and lower classes lived in three to six story apartment buildings called insulae [12]. In these apartments, the ground floors were frequently occupied by commercial shops such as restaurants, clothe sellers, and craft stores. The quality of life and sanitary levels of these buildings would be considered poor by modern standards; they were often lacking proper heat, water, and light supply. However, the common Roman tenant spent little time in their rooms. Instead, they were out in the streets going to shops, baths, and attending city events. The rooms were mainly only used for sleeping and storing private possessions.

Figure 17. A visualization of a Roman insulae, in which the majority of citizens lived. Illustration from [19].

When water arrived to Rome via the aqueducts, it would first flow into large covered storage tanks. Here the water was given further time to sit and allow any remaining sediment to settle. Then the water would leave the tank via a combination of canals, lead pipes, and clay pipes [12]. Water would flow along these pipes into more storage tanks distributed throughout the city, spreading out like a web. From these smaller tanks, the water would continue through lead pipe (called fistulae) to reach their final destinations. Nowadays, scientists know that lead pipes can be extremely damaging due to the mental damage that lead can cause once it enters the body. The Romans, unfortunately, were not aware of this and relied on the material heavily to make most of their pipes. To give an idea of how much lead the Romans were using, one pipe found by archaeologists in the 19th century was over 1750 meters long and contained 232,750 kilograms of lead [12]. This is the equivalent of a pipe that is 15 football fields long and weighs as much as 42 elephants!

It was very rare for a pipe to supply water directly to the home of a private citizen, since Romans would have to acquire an official authorization to validate the direct tap. Water mostly serviced the ground floor in buildings, rarely supplying the upper floors due to the difficulty this would provide in the gravity-powered system. Residents of apartment buildings who lived in the upper floors would have to carry water upstairs and store it in their rooms for sanitary uses. However, Romans would frequently try to bribe water officials in order to obtain direct access. If that failed, then they would often engage in an illegal act called “puncturing”. This was when secret pipes would be installed underground that led directly to the business or localities that desired the direct supply. Such networks were against the law and the water officials constantly checked water paths to try and prevent their existence. The Roman government was strict on water-theft because it threatened the city’s water supply intended for its main public uses: latrines, baths, and fountains.

Latrines, which were communal bathrooms, were a progressive advancement in city sanitation. Before the use of latrines, Romans would dump their human waste out of their windows and onto the streets [4]. This was obviously unsanitary, as being exposed to human excrement increased the likelihood of obtaining diseases and infecting food with dangerous bacteria. Cholera, dysentery, and typhoid were some of the diseases that ran rampant in ancient Rome. Latrines allowed citizens to dispose of their waste in much more preferable conditions. The water used for latrines was low quality, likely already used in baths and fountains, which was acceptable since it was not being consumed.

Figure 18. An example of what a public latrine looked like. This photo was taken in Ostia, a small town southwest of Rome that was a once a bustling port city in ancient times.

Baths were another common use of water in Rome. Many public baths were constructed inside and around the city, the most famous being the Baths of Caracalla (named after the emperor who had them built). These baths required massive amounts of water and some aqueducts, such as the Aqua Marcia, were constructed solely to provide for baths [4]. They contained latrines inside them and incorporated advanced heating mechanisms, using fires below the floor to heat the water to comfortable temperatures in order to provide a spa-like experience. Using the baths was part of everyday life for the Romans; admission was inexpensive and it was a relaxing way to socialize and freshen up.

Figure 19. Some preserved mosaic decorations from the famous baths of Caracalla. Roman baths did not only serve as a place to bathe; they were a place where citizens could socialize, play games, and discuss current events happening throughout the city.

Rome is famous for its beautiful fountains. They were an impressive monument to Rome’s wealth and power; not many ancient civilizations had access to so much clean water that they could afford to construct fountains purely for luxurious purposes. In addition to opulence, fountains served practical purposes as well. Public fountains were the most common form of potable water for Rome’s citizens, the majority of whom did not have private taps in their homes or apartments. In the peak of the Roman empire, it was said that a public fountain could be found within a 50 meter radius anywhere in the city [4]. These fountains were free and available for everyone to use. Many Romans were advised to fill buckets of water at the fountains and store them in their homes for later use.

Figure 20. A water fountain in current-day Rome. Fountains similar to this one were common in ancient times as well.


Rome has a rich history of water engineering, from its humble beginnings with the Tiber river through its construction of the aqueducts. Having consistent access to clean water through services like public baths and fountains allowed Rome to keep its population healthier and happier. The diligent engineering that was required to create Rome’s water system is a testament to the capabilities of human innovation; the fact that Romans were able to accomplish so much in ancient times should serve as motivation for current society to keep pushing the limits of engineering.

Figure 21. The pinnacle of Roman water engineering: some of the fountains featured in Villa d’Este, a 16th century villa in Rome famous for its gorgeous (and numerous) fountains.


[1] Aicher, Peter. “Watering Ancient Rome.” PBS, Public Broadcasting Service, 21 Feb. 2000,

[2] Richter/GTRES, Juergen, et al. “Aqueducts: Quenching Rome’s Thirst.” National Geographic, 15 Nov. 2016,

[3] “ROMAN AQUEDUCTS.” Introduction on Roman Aqueducts and Water Supply Systems,

[4] Muench, Steve. “Water and the Development of Ancient Rome.” We’re Never Far from Where We Were, 23 Jan. 2018,

[5] “IWA Publications.” IWA Publications,

[6] “Rome Early Settlers.” HISTORY’S HISTORIES You Are History. We Are the Future.,

[7] “The History Blog.” The History Blog RSS,

[8] Sheehan, Jake. “Bridge Construction Methods: Why Are Roman Bridges So Stable?” Machines4u Magazine, 31 Mar. 2017,

[9] “Greek and Roman Siphons.” Siphons in Roman (and Hellenistic) Aqueducts,

[10] “Aqua Appia.” Roman Aqueducts: Rome Aqua Appia (Italy),

[11] “Aqua Anio Novus.” Roman Aqueducts: Rome Aqua Anio Novus (Italy),


[13] Cloaca Maxima,

[14] Muelaner, Jody. “Make a Simple Groma!” Dr Jody Muelaner, 27 Mar. 2015,

[15] Roman Construction Techniques,

[16] MAMcIntosh. “Pons Fabricius: Rome’s Timeless Bridge.” We’re Never Far from Where We Were, 23 Jan. 2018,

[17] Roman Aqueducts: Technical Introduction,


[19] Jasi, Jakub. “Mieszkaniec Kamienicy Rzymskiej ” IMPERIUM ROMANUM.” IMPERIUM ROMANUM, IMPERIUM ROMANUM, 8 Oct. 2018,

Follow us

Don't be shy, get in touch. We love meeting interesting people and making new friends.