A long distance waterway that brought fresh water from Thrace to Constantinople. In the 4th century (completion in 373) it consisted of 39 bridges and 2 tunnels over a length of 246 km. The system was extended in the fifth century to a total length of 426 km with 57 bridges and 3 tunnels. We can therefore distinguish two channels in the Valens Waterway:
- The narrow channel (c. 0.70-0.80 m) from Damamandıra and Pınarca built by Valens in the 4th century
- The broad channel (c. 1.6 m) from Vize, added by the Theodosians in the 5th century
This made it the longest water supply line of the Roman world.
One of its sources were the Halkalı springs, which however had limited output. A more important source were the hills of Stranja Dağları: this includes Vize/Bizye, and springs of Danamandıra and Pınarca. Themistius refers to the latter in his panegyrics of the Valens system (Crow et al. 2008, p. 14). Initially after its completion in 373, there are no mentions of the construction of new cisterns until 30 years later. There is building activity between 406-459, which includes the construction of five cisterns (among which those of Aetius and Aspar). This building surge could be explained by the completion of the second line from Vize under the Theodosians (Ibidem, 15).
Before this system was built, the higher parts of the city may have gotten their water supply from wells and cisterns (Crow et al., 2008, p. 13).
Once it was fully operational (especially after the extension of the Theodosians), consuls were expected to contribute to the building/upkeep of aqueducts by means of financial support, and citizens were expected to use the water supply responsibly without diverting water from the aqueduct (see legislation): Cod. Theod. 15.2.3, 15.1.23, 15.2.4 and CIC (Justinianus) 11.42.6.
In 487, the magister militum Theodoric Strabo revolted against emperor Zeno (r. 474-475 and 476-491). In his bid for the capital, he cut the aqueduct; this may have affected the Thracian Valens line, however there is no concrete evidence for this (Crow et al., 2008, p. 16-17).
By the time of Justinian, authors such as Malalas and the anonymous author of Chronicon Paschale refer to the Thracian aqueduct as the 'city's aqueduct', which could indicate pre-eminence or deliberate distinguishment from the water supply of the Constantinian city (Crow et al., 2008, p. 18).
The reservoirs of Aetius, Aspar and Mokios, along with all the other cisterns along the Thracian line, fell out of use between 626-765 when that line was cut and subsequently repaired (Crow et al., 2008, p. 131).
The size of the workforce mobilised by Constantine V to restore 'the aqueduct' as recorded by Theophanes must also indicate that Constantine renovated the Theodosian Thrace line as well (Crow et all., 2008, p. 20)
There were minor restorations under Basil II (r. 958-1025) and Romanos III (r. 1028-1034), the latter renewed the underground conduits and the kastelloi. This system was maintained until the reign of emperor Manuel I (r. 1143-1180) in the 12th century. During Manuel's reign, it is recorded that 'the old arcades (stoai) which conveyed water to Byzantion were long since collapsed'. 'Arches' may refer to an aqueduct bridge. Instead, Manuel focused on cleaning the surviving channels and directing the flow to a newly built underground cistern outside the city, called 'Petra' (which locations is unfortunately unknown) (Crow et al., 2008, p. 21).
The Ottomans (Kırkçeşme system) would rely more on the low level supply line (c. 30 m), which roughly corresponds to the earlier Hadrianic line (Crow et al., 2008, p. 15).