As part of the preparations for this year’s Olympic games EYDAP, the water and sewerage corporation of Athens, was tasked with finding out how the 28th Olympiad would affect its ability to meet required levels of service for water supply and wastewater provision in the city and nearby regions.
The corporation commissioned Atkins Consultants in the summer of 2000 to help address the supply/demand and sewerage issues associated with staging an event of the magnitude of the Olympics.
The objectives of the Atkins study were to develop a hydraulic model, using an internationally renowned software package, which could encompass all of the trunk sewers with enough complexity to be able to identify major hydraulic deficiencies during the Games, set for August 2004. The study was also to identify outline options for alleviating known deficiencies and develop a model suitable for further future development and use as a planning tool. Wallingford Software’s InfoWorks CS software solution was chosen for this role.
EYDAP’s 1997 Master Plan for Athens’ Sewerage Division was used as a primary source of information for the study, alongside many other studies commissioned by EYDAP and the ministry of public works.
The model was to be a skeletal representation of the Athens sewerage system - further work would be required if the model was to be used as a full Drainage Area Planning (DAP) tool. Because of time and funding constraints the model targeted five core areas:
- Development of the Olympic Village and its impact on downstream sewers
- The Olympic stadium and surrounding developments
- The Athens city centre and core sewer network, including the main Kappa Alpha Alpha (KAA) trunk sewer
- The Fallirikion sports complex, along the Athens coastline
- The impact of sewage from cruise ships docking at Piraeus harbor
The modelling was set in the context of the ancient city of Athens, which has a residential population of around 3,850,000 within the existing sewerage model and covers a total area of 41,000ha. The city’s drainage catchment covers a variety of geographical areas, from mountainous slopes to coastal beaches. It sits within a ring of mountains that create a geological basin, a feature that has constrained development and resulted in high population densities in the older parts of the city.
The topography of the area around the coastal lowlands is generally steep, with the coastal strip being between 2km to 15km wide and extending up the Kifissos and Ilissos river valleys.
The area has a mix of uses - domestic, commercial and industrial premises with several zones containing heavy industries such as railway engineering, steel works and manufacturing facilities. Most of the largest heavy industrial sites are to the west of Athens and are not included in the digitized sewerage network.
The city’s sewerage developed over the years as a number of discrete systems. Over the years the smaller systems have amalgamated, resulting in a network with very deep sewers and a large number of branches. The coastal sewerage system has also been developed in this way, over many decades, resulting in the construction of a number of pumping stations in series along the sea front. These take flows from the low-lying coastal strip to an interceptor sewer and then on to the Pystallia wastewater treatment works.
The expansion of the city has resulted in an extended sewerage system with pipes of varying diameters, shapes, depths and materials. The catchment is predominantly drained by separate foul and stormwater systems, with just one proper combined area to the east of the KAA trunk sewer. The combined area is one of the oldest, most densely populated regions in the Athens catchment and has suffered badly from flooding in the past.
The catchment’s dense development, combined with the steep gradients, has created a number of problems including high volumes of stormwater runoff over short periods of time during storm events. These, combined with a shortage of capacity in both the storm and foul sewer networks, have given rise to extensive flooding. The benefit of the topography is that, apart from the coastal strip, the sewerage system is a gravity sewer network with low operation and maintenance costs.
Modelling challenges and procedures
To meet the objectives of the study project, the modeling had to overcome a number of significant challenges and to help with this process it was divided into five sections of work. First was the digitization of the network - prior to the study, there were no digital sewer records for the Athens network.
In order to define what data to collect and the extent of the digitization, foul sewers on engineering plans were marked up to cover the primary network - pipes with diameters of 500mm and above. Additional digitization detail was undertaken in the areas where Olympic developments were planned, where known deficiencies existed and where pumping stations, combined sewer overflows (CSOs) or penstocks were located.
The digitization process aimed to identify key nodes - manholes - to a maximum of 2000 nodes. This corresponds to an average of 1500 persons per node, and around 30ha per node. Further discussions between EYDAP and Atkins agreed that a more comprehensive model could be created by digitizing an estimated 5000 nodes.
Available plans were used to provide information on manhole location, referencing and ground levels as well as conduit invert levels, shapes, sizes (both width and height) and connectivity.
EYDAP presented ‘as built’ drawings and schematic diagrams for most of the pumping stations and CSOs. Where such details were not available, it was necessary to make engineering assumptions based on experience.
Connectivity checks were made on the network to establish whether the surface water system drained into the foul system at any points in the catchment. The surface water system was modeled in areas where it received CSO spills, or where it connected to the foul network and there were known interactions.
For most of the trunk sewers all manholes were included, though some non-critical areas of the network were allowed a degree of simplification. Where there were no Olympic developments, flooding, structural or operational issues, the network data plans were checked to decide whether it was necessary to include these areas.
Athens’ sewer network was known to suffer from a number of problems including infiltration, illegal connections and cross connections. Some of the issues associated with stormwater entering the sewerage system include:
- A lack of non-return valves on the CSOs on the KAA trunk sewer
- Illegal connections on private properties of stormwater to foul sewers
- Excessive infiltration of groundwater, especially during rainfall
- Diversion of polluted stormwater low flows from the river system via flow diversion structures
These poorly maintained control facilities allow large volumes of flow to be diverted to the trunk sewers during rainfall events. Within the limitations of time and data availability, but with the help of the know-how of operational staff, these defects were considered during the optioneering stage.
Before and after calibration, the stability of the model was tested using ten-year return summer storms of varying duration. With the calibration completed, the remaining model was checked for CSO spills and areas of flooding during dry conditions.
With the initial calibration complete, the model had to be calibrated against actual historical data from the five core areas to ensure it was robust. Data from 25 rain gauge stations was collected and analyzed to assess Athens’ typical rainfall profile. This was then used as reference data for the validation.
Further testing and investigation were undertaken, and the model adjusted and refined by a process of continuing iteration until it was considered that the pre Olympic Games Athens model had been calibrated in accordance with the recommendations outlined in the initial study.
For the Olympic Games investigation, the model was amended to reflect the future design horizon, incorporating the ongoing sewerage rehabilitation scheme for the coastal region; the ongoing sewerage collection system for Piraeus harbor, the Games developments and the new Olympic Village collector sewer (to assess its downstream impact). All of the developments were assumed to contribute foul flow and there was a 5% allowance for storm flows to take illegal and cross connections into account.
Once the process was complete, locations were identified where the model predicted significant flooding of over 25m3 from a 60 minute, one in ten year design rainfall event. The model predicted hydraulic problems in the central Athens sewerage system, where the majority of the combined sewerage network was located. EYDAP confirmed that problems did exist there, but it was agreed that there were no historical DWF problems in the area.
It was also clear that the pre-Games model predicted flooding during storm conditions in the Athens coastal region, so this could not be attributed to the Olympic developments. However, it was clear that the Olympic venues would exacerbate the existing deficiencies. These identified deficiencies were taken forward to the optioneering stage.
At this stage, solutions were developed for all the deficiencies identified. These included individual or combinations of schemes to form work packages. The schemes proposed included provision of offline storage tanks, or high level overflows; localized, small-scale storage; increasing the capacity of the upstream overflows that discharge excess storm flows to the sea; fitting screens to CSOs; storm relief sewers and full separation of the sewer systems. Also considered was provision of a storm relief sewer, with transfer of flows to the KAA trunk sewers.
During the Olympic Games, water supply side demand exceeded the mean by 5%, rising from 1.1M.m3/day to 1.2M.m3/day, which was less than forecast because of an exodus of the local population from the city. Attendance at the Games was also reduced because of press coverage relating to the threat from terrorism. Flows in the sewer system reached 750,000m3/day, corresponding to 60% of water consumption.
EYDAP organised and prepared for the Games by allocating a large personnel resource for the duration of the event; increasing expenditure to complete construction of the networks; and creating an Operational Centre for special support during the Games as well as providing direct connections to the telephone support centre. They also kept a fleet of vehicles on standby in readiness for any possible problems or damage inside the Olympic installations. Its preparations for the Paralympic Games, which followed after the Olympic Games finished, were also successful.
*This case study is derived from a paper originally presented to the 2004 International WaPUG conference by Costa Ripis of EYDAP Athens and Gwion Kennard of Atkins Water.