Hydraulic investigation of Hadrian aqueduct

C. Tsiouri, Hydraulic investigation of Hadrian aqueduct, Diploma thesis, 260 pages, Department of Water Resources and Environmental Engineering – National Technical University of Athens, Athens, October 2018.



Hadrian aqueduct was constructed on 140 AD, by the method of sequential shaft excavation, in order to provide water supply at the ancient roman distinct in Athens, Greece. It begun at the Olympic Village region and ended, after approximately 20 km, at Dexameni in Kolonaki region. Until its operational end, a lot of modifications were performed and it was connected to many supplementary aqueducts in order to increase its capacity. The scope of this Diploma Thesis is the hydraulic investigation of Hadrian Aqueduct, namely the confirmation of its theoretical discharge value as well as its response to various analysis scenarios. In particular, a parametric study was conducted focusing on the discharge, the topography, the cross-section dimensions as well as the entrance points of additional water supply. After digitizing any available historical records (i.e. 2 sets of longitudinal sections, consisting of ten drawings each) and finalizing the longitudinal section and the plan view of the aqueduct, the aforementioned scenarios were applied, in order to investigate hydraulically the aqueduct, using the hydraulic simulation software HEC – RAS. Initially, the segment Agios Dimitrios – Dexameni was ignored from the plan view of Hadrian Aqueduct, as there were no sufficient evidence regarding the shape of its longitudinal section. For this assumption, two scenarios regarding the shape of the longitudinal section were modeled: the former with constant cross-section along the aqueduct and the latter with cross-sections of different dimensions, as derived from the final longitudinal section of the aqueduct. As for the discharge scenarios, scenarios of steady, permanent discharge along the aqueduct were initially analyzed, with discharge values ranging from 10 to 150 L/s, while the value of 150 L/s corresponds to the maximum discharge transported at the aqueduct. In the sequel, scenarios with different discharge values in the segments of the longitudinal section that consist connection points with the supplementary aqueducts were analyzed. The supplementary aqueducts, that were considered at the respective discharge scenarios, are located at Kokkinara region and at Halandri, Attiki. For the first supplementary aqueduct, different inflow scenarios are considered in order to achieve a discharge of 150 L/s after the connection point, while, respectively, for Halandri the additional inflow discharge is Q=50 L/s, in order to achieve 180 L/s discharge downstream of the connection point. The same discharge scenarios were also applied, when the segment Agios Dimitrios – Dexameni is included in the plan view, and led to similar analysis results as before. From the sum of the results, four regions along the Hadrian Aqueduct were defined, where the aqueduct works under pressure even for the lowest discharge value (i.e. Q=10 L/s). This is attributed to the abrupt inclination changes of the longitudinal section, which must be modified accordingly so that the Aqueduct to work as an open channel. It is noted that for all the aforementioned scenarios, a Manning coefficient equal to n=0.020 was selected. The modification of the longitudinal section was exclusively performed at these four regions, aiming to create smooth inclinations and was verified by steady discharge scenarios along the aqueduct, with discharge values ranging from 10 to 150 L/s. In addition, at the modified Aqueduct, the scenario of additional discharge from Halandri that results to total discharge of Q=180L/s at the downstream segment was considered as the worst case scenario. For this scenario, the inclinations were properly modified until the Hadrian Aqueduct worked as open-channel with free surface flow. Finally, different values of Manning coefficient (i.e. n=0.017, n=0.020, n=0.025) were considered at the analyses, in order to simulate all possible types of roughness in the inner wall surface of the Aqueduct. The results of these analyses revealed that, for n=0.025, the Hadrian Aqueduct works locally under pressure in different regions from those that were initially modified. After the end of the analyses, important conclusions have been drawn regarding the first approach of the hydraulic investigation and many topics of future research on the Hadrian Aqueduct have been revealed.

PDF Full text (14715 KB)