Your browser is no longer supported

For the best possible experience using our website we recommend you upgrade to a newer version or another browser.

Your browser appears to have cookies disabled. For the best experience of this website, please enable cookies in your browser

We'll assume we have your consent to use cookies, for example so you won't need to log in each time you visit our site.
Learn more

Large scale geotechnics on the Bosphorus Railway, Istanbul

Vasilis Karakitsios, Parsons Brinckerhoff, Hideki Sakaeda and Robin Miles, Pacific Consultants International.

The Bosphorus Project is a 76km long, combined metro and freight railway providing the first rail connection between Europe and Asia.

Click here for map

The main structures and systems include a 1.4km immersed tube tunnel, 9.8km twin tube tunnels excavated by tunnel boring machines, 2.4km cut and cover tunnels, four underground stations, 37 surface stations, an operations control centre, stabling yards, workshops, maintenance facilities, 250km of new track, electrical and mechanical systems, rolling stock and a 25kV electrification system.

The construction programme began in 2004 and is due to complete in 2012. The cost is $3bn (£1.5bn) with financing by the Japan Bank for International Cooperation, the European Investment Bank and the Council of Europe Development Bank.

Project owner is the Ministry of Transport Turkey, represented by the General Directorate of Railways, Harbors and Airports. Contract BC1 (design and build) covering all tunnels and the four underground stations is being constructed by the Japanese/Turkish joint venture of Taisei, Gama and Nurol.

Avrasyaconsult is the employer's representative, led by Pacific Consultants International (Japan), with Parsons Brinckerhoff (UK/US), Yüksel Proje (Turkey), Japan Railway Technical Service, and Oriental Consultants (Japan).

Parsons Brinckerhoff (PB) leads the immersed tube tunnel and electrical and mechanical teams, and contributes to station architecture, hydraulics, marine environmental, seismics, geotechnics, and railway design issues.

The benefits are significant:
 Istanbul train journeys will be increased from 3% to 27% of all passenger journeys
 Motor vehicle pollution and severe traffic congestion will be significantly reduced
 Crossing the Bosphorus Strait will only take four minutes
 New modern stations and fully air-conditioned rolling stock
 Seamless interchanges with the planned new tram and existing metro systems
 No permanent impact on the Bosphorus' rich aquatic habitat
 No negative visual impacts – the spectacular views over Istanbul's historic skyline will be preserved.

Project challenges include extensive archaeological excavations at the historic port of Constantinople (including a Byzantine naval vessel), world heritage monuments and old buildings on shallow foundations. Construction will also take place around Istanbul's dense urban population, designing to the highest international earthquake standards, and all life safety systems for a 100MW freight train fire.

This paper discusses design and construction of the two largest cut and cover stations: Yenikapi Station on the European side, and Üsküdar Station on the Asian. Geotechnical and seismic issues of the immersed tube tunnel are also discussed.

Geotechnical design issues

Istanbul's geological setting at the interface between two continents presents unique challenges for design and construction of the immersed tunnel and station boxes.
Notable among these are:
 A highly seismic area due to its proximity to the active North Anatolian fault
 Highly fractured and variably\weathered rock with numerous shear zones
 Deep historic fill depths at Üsküdar Station
 Highly variable soft marine soils below the immersed tube, with soils susceptible to liquefaction.

Site investigations

Site investigations for the project were carried out between 1985 and 1987 for preliminary design, then from 2002 to 2003 during the contract tender period.

Site investigation during the 1980s included core drilling in rock and soil with systematic standard penetration testing (SPT) and sampling. A total 772m borehole length at 20 marine locations and 1394m at 37 land boreholes was carried out.

Marine investigation was done from a barge using conventional drilling equipment. Geophysical and bathymetric surveys were conducted across the Bosphorus Strait to assess condition of marine soils. Physical properties of the rock and soil were determined by laboratory tests.

The 2002 to 2003 site investigation between Yedikule to Sögütlüçesme Stations consisted of:
 A total length of 1.07km core drilling in 26 land boreholes
 Undisturbed and SPT sampling in soil layers
 A geophysical survey (up/down hole and cross hole) at deep stations
 Laboratory testing to define geomechanical properties of the rock and soil
 Seven boreholes (350m length) and a maximum 85m depth from seabed level
 Cone penetration testing (CPT) and undisturbed sampling every 3m in marine soils
 Continuous CPT tests down to depth of 40m below seabed at three locations
 P-S logging in two holes, from seabed to 10m into bedrock, to obtain shear data for seismic analyses
 Environmental sampling and testing from seabed to 9m depth in marine soils.

The investigation confirmed findings about the soft soils and provided information that helped take timely contingency measures. For example, a local surficial zone of fill was stabilised, and contaminated material on the European side of the crossing was removed and disposed of.

Earthquake design

The primary tectonic feature affecting the project site is the Marmara fault system. PB in collaboration with the Department of Earthquake Engineering at Bogaziçi University assessed the distribution of earthquakes that have occurred in this region since 1500, and proposed designs that would allow all structures to withstand earthquakes measuring up to 7.5 on the Richter scale.

Both probabilistic and deterministic hazard models were developed for seismic design. The initial probabilistic model employed was based on the homogeneous Poisson model. This, however, was not fully applicable to the tunnel because of the unique time-dependent behaviour of the fault system, so a time-dependent conditional probability model was used to characterise the segmentation behaviour of the main Marmara Fault.

Consideration was given to the time elapsed since the segment last ruptured. The results of the analysis indicated an about 50% probability of an earthquake with a moment magnitude greater than 7 during the next 50 years.

For the deterministic hazard analysis, an earthquake of 7.5 moment magnitude was selected and assumed to take place 16km south of the immersed tunnel. The attenuation relationships used in deriving ground motion parameters (such as peak ground acceleration, velocity and displacement) were based on the median-plus-one standard deviation values. This was because of the importance of the project and the relatively high uncertainties associated with large magnitude earthquakes.

For the longitudinal tunnel response analysis, the tunnel was modelled as beam elements taking into account the non-linear characteristics of the tunnel member rigidity. The interaction between soil and ground was modelled as non-linear spring elements with limiting resistance, and the structural characteristics of the joints in the immersed tunnel section taken into account.

To determine the sectional forces and displacements in the structural elements under the transverse racking deformation effect, the analysis consisted of the following steps:
1. Determine the free-field deformation profile from site response analysis
2. Develop a two-dimensional free-field finite element plane-strain model
3. Apply boundary displacements or loads on the model boundaries (or a pseudo-static seismic coefficient) to achieve the free-field soil deformation profile
4. Add tunnel structural elements in the model and remove the soil elements to be excavated
5. Apply the same boundary displacements or loads (or the pseudo-static seismic coefficient) used in step 3.

Immersed tube tunnel geotechnics

The immersed tunnel is being constructed in a trench dredged and backfilled to a minimum 2m cover with material not vulnerable to liquefaction (to protect against falling anchors and sunken ships).

The soils below the tunnel's invert towards the European side are composed primarily of silt clay, so liquefaction is not a serious concern. However, borehole and geophysical data indicated deposits of liquefiable soils (loose to medium dense sands) towards the Asian side of the tunnel alignment.

Loose to medium sandy soils are present to a depth of about 10m below the immersed tube invert. Seismic analysis indicated that these sandy deposits are prone to liquefaction in the event of an earthquake, with post-liquefaction settlements of up to 350mm.

Ground stabilisation on a massive scale was conducted, and took one and a half years to complete. The operation involved injecting 2778 grout columns into marine sediments on a 1.7m grid to a depth of 8m.

The depth of the Bosphorus seabed (a minimum of 30m), the strong and unpredictable sea currents, and the heavy shipping traffic made compaction grouting a challenging operation.
Other geotechnical works on the sea bed included removal of 140,000m3 of contaminated soil from the upper 3m near the Golden Horn Estuary on the European side and ground stabilisation work involving over-dredging and replacing a 3m thickness of marine soil also at the European end.

Station geology
The geology at Yenikapi Station comprises historic fills, marine deposits, Süleymaniye Formation and Trakya Formation. The geology at Üsküdar Station is similar, except that Süleymaniye Formation was not encountered.

The principal rock formation along the European and Asian land tunnel alignments is the Trakya Formation. This is composed of sandstone, siltstone and mudstone sequences, and is generally moderately weathered to fresh, with close to moderately spaced discontinuities.

Typically, the intact rock has moderate strength, with average unconfined compressive strength of 70MPa. However, at fault and shear zones, up to several metres thick, the rock can be highly fractured and weathered and have strengths in the range of 2MPa to 20MPa. Intrusive dykes of Diabase, having strengths higher than 180MPa, were also encountered.

Süleymaniye Formation consists of silty clay with sandy lenses. The clayey layers are highly overconsolidated, with stiff to very stiff consistency. The sandy lenses often contain perched groundwater.

According to the Unified Soil Classification System, the cohesive layer of the Süleymaniye Formation is usually classified as CH and MH-OH.

Üsküdar Station construction (Asian side)

Üsküdar Station will be one of the biggest underground stations in Istanbul. Extensive soft fills to significant depths were identified during the ground investigation. The station box is being constructed with 1.5m thick diaphragm walls, laterally supported by reinforced concrete struts. Table 1 presents the geotechnical design parameters for Üsküdar Station.

The soil is removed by excavators and cranes with clamshells to load dump trucks at the surface. When excavation reached the first strut level, reinforced concrete struts and whaler beams were installed to provide temporary support. The soil is then excavated to the next strut level where another tier of temporary support is put in place.

Ground movements and groundwater levels are monitored continuously by geotechnical instruments because of the many sensitive and historic structures in the area, including a mosque. Dewatering is carried out by submersible pumps operating in sump pits in the excavation box and will remain active until no risk of uplift action remains.

Yenikapi Station construction (European Side)

Yenikapi Station is on the site of what was once Istanbul's busiest harbour between the 3rd and 8th centuries AD. The harbour gradually filled with river sediments and waste dumping until it was abandoned.

The station layout has been revised several times because of the archaeological findings and changes in Istanbul's transport planning. As a result, the station and ventilation buildings are now being built separately.

Because of the historical importance of the site, archaeological excavations are being performed from ground level down to -6.5m. This has involved roughly 130,000m3 of hand excavation, which has delayed the project by over a year. Excavation will continue to its final depth of -18m by mechanical means. Groundwater level is close to the surface.

Borehole data revealed marine deposits to a depth of -6.6m and Süleymaniye Formation to the excavation's final level.

This enabled construction of the station box with secant piles. The pile diameter ranges from 800mm to 1.5m, and reaches a toe level at 23m to 25.5m depth.

Temporary horizontal support during soil removal is provided by reinforced concrete whaler beams and tie-back ground anchors.

Steel pipe struts and king-post piles are used to support the excavation for the ventilation building.


Legend has it that when the huge Persian army of the 483BC invasion of Greece met the Marmara Sea, King Xerxes ordered a bridge made of flax and papyrus to be built.

But the unpredictable sea currents destroyed the bridge causing the death of thousands. To punish the sea, Xerxes ordered the waters to be whipped 300 times with iron chains.

Through innovative techniques and design and construction management, project participants have proved there are better ways of bringing Europe and Asia closer (without angering the sea spirits).

The economic and environmental benefits generated by the Bosphorus Project will be significant, including direct projected benefits to Istanbul of £300M per year.

Have your say

You must sign in to make a comment

Please remember that the submission of any material is governed by our Terms and Conditions and by submitting material you confirm your agreement to these Terms and Conditions. Please note comments made online may also be published in the print edition of New Civil Engineer. Links may be included in your comments but HTML is not permitted.