The Musaimeer pump station and outfall project is located directly south of Hamad International Airport in Doha, on the eastern side of Qatar. The project is owned by public works authority Ashghal and is designed to receive both ground and storm water from 270km2 of southern Doha.
Although predominately desert, Qatar can experience torrential rainfall in the rainy season which occurs from September to March. The quantity of rainfall received in a short period can cause severe disruption to the transport systems for days at a time. Groundwater level is high in the Doha area and is rising generally in the Middle East; this presents problems for urbanization and increases the cost of construction.
Ashghal has a policy of reducing groundwater level to minus 3m (-3m) and maintains this by a process of dewatering; the Musaimeer project will collect and process this excess ground water. The outfall tunnel will extend 10.2km offshore from the pumping station, and will connect via a riser shaft to a diffuser field. This will allow the safe and environmentally-compliant discharging of storm and groundwater flows directly into the Gulf.
TBM SELECTION
The contractor for this project was a joint venture between HBK (Qatar) and PORR (Austria). The CREG/ Wirth TBM selected was an earth pressure balance (EPB) with a cutterhead designed to support a working pressure of 4.5bar (full hydrostatic). The machine also includes an excavation chamber, main bearing, screw conveyor, front, middle and tail shields, personnel locks, thrust system, articulation, erector and brushes as part of the tail-shield seal system. In addition, there are 18 gantries totalling 160m in length and accommodating all electromechanical parts, additive tanks and rescue chambers. The machine was designed to install a universal segmental ring of 1.3m width. The main components of the CREG/Wirth EPB TBM were broadly as follows:
? Optimized cutting-tool combination for both hard and soft rock;
? Open- or closed-mode operation;
? Materials and equipment to support a maximum face working pressure of 4.5bar;
? Eliminating the risk of uncontrolled water inflow into the tunnel during construction;
? The screw conveyor had two sets of gates, for emergency use under high pressure;
? A long screw conveyor to dissipate the maximum 4.5bar pressure;
? An extra row of wire-brush seals;
? A removable probe-drilling machine;
? Real-time probing system using BEAM (Bore-tunneling Electrical Ahead Monitoring).
CHALLENGES
This article will discuss three of the main challenges and the management employed to overcome and successfully complete this demanding tunnel. The three main challenges were:
1. Geotechnical conditions
2. Single access shaft
3. Groundwater extremes.
CHALLENGE 1 – GEOTECHNICAL CONDITIONS
Prior to tender stage, Ashghal undertook a significant geotechnical investigation program which consisted of drilling 22 offshore boreholes at 500m centers and which provided the following geotechnical data:
? Cores for inspection
? In-situ packer tests
? Pressure meter tests at each location
? Geophysical surveys including sonar bathymetry, magnetometer, seismic reflection and seismic refraction
? Laboratory mechanical and chemical tests on soil, rock, and water samples.
This investigation data was combined with a long geotechnical profile and formed the basis for selecting the TBM and tunnel segmental lining type. .
Although there was significant experience of TBM tunneling in the Doha area, there was no previous experience of sub-sea tunneling. The investigation data had indicated the rock quality along the tunnel alignment to be competent. However, there were two major tunneling risks which had to be considered by the contractor and their designer.
Karstic features
The first challenge was the possibility of the TBM encountering karstic features. Although the tender geotechnical investigation showed no significant evidence of these features, the TBM was nevertheless equipped with grouting ports located both ahead and around the TBM shield, as well as all the associated drilling facilities required to deal with such features should they arise.
Hydraulic connection to seabed
The second risk was associated with the connection between the seabed and the TBM-bored tunnel. In this case, the full hydrostatic water pressure would be directed to the advancing TBM cutterhead, requiring increased face pressure and thrust to advance the tunnel. The machine was designed to withstand the 4.5bar pressure. There was significant evidence from the existing geotechnical investigation that in 85% of the tunnel alignment there was competent and/or lowpermeability rock at least one tunnel diameter above the tunnel crown
Anticipated geological formations
Referring to borehole information and additional geophysical investigations, the TBM would drive through three general geological formations: Rus Formation, Midra Shale and Simsima Limestone (Figure 1). As the TBM transitioned from one to another, the excavation face included both strata. Both low pressure (LP) and high pressure (HP) water was expected but the extent was not known. The TBM excavated through seven zones, as follows:
? Rus Formation (RF) – full face
? Rus Formation with Midra Shale (MS)
? Midra Shale – full face
? Midra Shale with Simsima Limestone (SL)
? Simsima Limestone – full face LP and HP
? Simsima Limestone with Midra Shale
? Simsima Limestone – full face LP and HP.
Post-tender investigations
To provide more complete information, the contractor implemented a detailed geophysical investigation survey to help mitigate the major geotechnical risks.
Two different types of offshore geophysical surveys were conducted along the outfall tunnel alignment. The Electrical Resistivity Tomography (ERT) and the Seismic Reflection Geophysical Tomography (SRG).
Electrical resistivity tomography survey
ERT determines the subsurface distribution in two or three dimensions (2D and 3D) of resistivity, using geological parameters such as soil/rock type and mineral content, as well as the porosity and degree of water saturation in the rock.
Correlation between resistivity value variations and the three different geologic units was then identified. A number of ‘conductive anomalies’ and ‘resistive anomalies’ were detected along ERT profiles.
Seismic reflection geophysical survey
SRG records the acoustic waves that are generated at the surface of the sea which have been reflected back by any sub-surface structures or interfaces. A reflection will usually occur when density and/or velocity changes occur at the boundary between two different materials. The acoustic waves are recorded from numerous hydrophones, installed at fixed distances along a seismic cable (‘streamer’).
The SRG survey was used to correlate the existing borehole-defined stratigraphy; confirm the extent of topsoil formations, and identify possible vertical weak zones which could result in excessive water ingress along the tunnel alignment.
Based on the interpretative analysis, certain structural vertical zones/faults were identified indicating possibilities of increased water ingress. These features are points where abrupt changes in the seismic wave profiles occur. This assessment identified 20 locations which were coded F1 to F20.
Real-time non-destructive radar
The TBM was provided with a non-intrusive, electrically-induced polarization prediction system, able to predict ground conditions and provide interpretations up to three tunnel diameters ahead and two thirds of the tunnel diameter in real time. The Bore-tunneling Electrical Ahead Monitoring (BEAM) system is based on the geotechnical principle that rock mass has different resistivities at varying frequencies, expressed with the Percentage Frequency Effect (PFE). The information is displayed in a matrix which combines both resistivity and PFE. The interpretation of the matrix, shown in Figure 2, is based on the columns and rows referring to the karst interpretation (P1:P4) and the possibility of water inflow(R3:R1). This data was available in the TBM operation cabin.
In Figure 3, the TBM is shown tunneling in a rock mass with a higher potential risk of karsts and aquifers; this particular location refers to a fault that had been identified previously (F8).
The BEAM system gave indication of 10m–15m before entering those zones. It was observed that 12 of the 20 F zones were correctly identified and a good correlation with predicted condition. The correlation greatly increased over the tunnel alignment chainage from 2,600m to 10,200m, identifying 11 out of 20 F features, so allowing TBM advance rates to be maximized.
CHALLENGE TWO – SINGLE ACCESS SHAFT
Long tunnels with only one access shaft pose many logistical program problems, but combined with the requirement to drive directly out under the seabed for 10.2km served to only increase the challenge. Once TBM selection was complete, the next task was to ensure efficient removal of excavated material, and the option of locomotive and material skip was chosen. The expected total excavated volume of material required to be removed for the erection of one 1.3m segment ring comprises three factors:
? Excavated volume per 1.3m advance
? Bulking factor applied to above
? Soil conditioning.
Using the above criteria, the excavated volume would amount to 18.9m3 and a bulking factor of 50% was then applied to give 28.6m3 of excavated material in the loose, while the average soil conditioning (foam, water and bentonite) would be 2m3. Therefore, the total volume per segmental ring would be 30.6m3. To accommodate this volume, four eight-cubic metre-capacity material skips were provided. In addition, two segment cars and a general-purpose transport car were also included. The final addition was the diesel locomotive.
ACCESS SHAFT ARRANGEMENTS
Logistic shaft
The most effective method to construct the outfall tunnel was to use a temporary logistics shaft. This was located directly behind the drop shaft (48m deep) and along the same alignment as the outfall tunnel. This structure consisted of a 26m-diameter shotcrete-lined shaft and an extension tunnel of 26m, capable of housing three separate rail tracks side by side.
TUNNEL LOGISTICS OPTIONS
Single and multiple California crossing Considerations
As a tunnel construction progresses, locomotive journey times increase until, at some point, they result in the TBM having to stop and wait for the locomotive to bring back empty material skips and additional supplies to it before it can start mining again. To avoid this, one or more rail crossings can be introduced in long tunnels to keep the tunneling continuous. This allows more than one locomotive to operate on the railway line at once and allows locomotives to wait closer to the TBM, ready to supply it from a short distance.
The ‘California’ crossing is a complex structure which must take into consideration many variables and so will take months to fabricate. Variables to be considered in its design for the MPSO tunnel are:
? Safety of TBM personnel;
? Number and duration of working shifts per day;
? Number of working days per week;
? Length of combined material/segment and personnel trains (normal conditions);
? Additional excavated material skips for adverse conditions;
? Ease of installation;
? Ease of relocating within the tunnel, two days per relocation;
? Impact of TBM services within the tunnel;
? Minimal impact on ventilation system;
? Efficient access and exit of the structure for rolling stock.
There are then three key interrelations to consider:
1. Maximum speed: the higher the speed that rolling stock can travel on the single rail, the better, but this requires a higher standard of rail-track installation and maintenance;
2. Accessing, traversing and exiting the California crossing;
3. Acceleration from the California crossing and travelling to the TBM, and positioning for the next excavation cycle. This time should ideally be less than the time it takes to build one tunnel ring.
After the simulation, the final decision was that a maximum of two California crossings would be required, but that this would be a marginal decision, with the possibility that only one would be required. The key finding was the number of idling hours per week which should not exceed 20. Therefore, the contractor’s initial plan was to install two California crossings at chainages 3,300m and 7,900m. Only one would be fabricated, with the second put on hold pending assessment of actual performance for the first 4km of tunnel construction.
SINGLE RAIL CROSSING ACTUAL LOCATIONS
As tunneling progressed beyond ring 4,000, and based on site observations using 3–4 production locomotives travelling at speeds of 15km/h, the TBM operation did not incur excessive idling time for the supply and delivery of segments and transportation. After continuous examination of the idling time, the analysis showed that one California crossing would be sufficient to complete the entire tunnel. The travelling speed of the transport system was increased to 20 km/ hour since the high standard of rail maintenance would justify this.
Figure 4 shows the recorded idling time for locomotive transport (production trains) over the full length of the tunnel of 7,813 rings or 10.2km. Figure 4 also shows the seven positions of the single California crossing used during tunnel construction.
CHALLENGE THREE – GROUND WATER EXTREMES
Two extremes of ground water were encountered during tunnel excavation:
1) Midra Shale with less than 5% moisture content;
2) Simsima Limestone fractured to the seabed with hydrostatic pressure of 3.5bar to 4.5bar;
During excavation, ground conditioning was carried out at the tunnel face within the TBM excavation chamber and screw conveyor. Soil conditioning provides many benefits, including increased face stability, reduction in friction and cutterhead torque, enhanced control of the pressure inside the cutting chamber, better control of groundwater, improvements in the flow of excavated material through the screw conveyor, less material adhesion during handling, and reduction of wear and tear of cutters, cutterhead and other wear parts. The quantity of soil conditioning agent will depend on the properties of geological formations or the ground condition, which will directly impact the TBM progress.
Midra Shale
The TBM excavated through the Midra Shale from ring 250 to ring 1,150 (900 rings = 1,170 m); this was very slow, with average production of eight rings/day. After ring 571, the addition of water up to 25m3 per ring was required along with an increment of soil conditioning agent up to 75 liters of additives (concentrate) per ring when the normal was 12 liters/ring. This resulted in an increase in excavated volume from 4–9 skips per ring. This meant two trains were required per tunnel ring.
Notwithstanding the excavation time being long (around 90 minutes), the bottleneck was in the logistics shaft where the gantry crane had not accomplished the lifting activities on time, forcing the TBM to wait for a locomotive for between 10–20 minutes per ring. To mitigate the idle time, a third train with the same configuration as previous, was placed on the track and a California crossing was installed at ring 550, then moved to ring 1,000 just behind the TBM. The machine was 90 days ahead of program entering into the Midra Shale, however upon exiting this zone, it was just 51 days ahead of the program.
Simsima Limestone – hydrostatic pressure 3.5bar–4.5bar
TBM excavation in Simsima Limestone from ring 2,830 to ring 4,100 saw slower progress due to high water pressures (3.5bar–4.5bar). The same operational issue was also seen between ring 2,830 to ring 3,497, again due to high water pressures (3.5bar). During this section, two cutterhead inspections were required which caused further delay in the program due to hyperbaric conditions of 3.5bar. To maintain TBM progress under high water pressure, it became challenging to condition the soil despite the addition of conditioners. Because of this, the quantity of skips required increased and was consistent at five material skips per ring, however, this required two trains per ring. The California crossing was relocated three times to mitigate the idling time. TBM productivity was impacted severely, leading to nine days behind baseline program on exiting this zone.
Average soil conditioning and material skips
In the Midra Shale, 25.4m3 of conditioning material was required per ring, while in the Simsima Limestone 4.3m3 of conditioning material was required per ring.
TBM PROGRESS – BASELINE VERSUS ACTUAL PROGRESS
The contractor started the tunnel excavation 90 days early. The tunnel then progressed and, as various operational issues were encountered, the variance with baseline deteriorated, until in May 2020 it reached 14 days behind program. From this date onwards, the variance with baseline improved until at tunnel completion it was 58 days ahead of program. This achievement can be considered even more impressive since only one California crossing was used throughout this period.
CONCLUSIONS
At the early stage of the project, very professional and comprehensive planning and simulations were undertaken in terms of logistics (California crossing and rolling stocks requirements). Also considered and evaluated were all the possible issues that might affect the tunnel transportation system. All assumptions were clearly identified and recorded to allow verification during tunnel construction. The prime objective being to complete the tunnel construction with the minimum number of California crossings.
Precise daily, weekly, and monthly time monitoring not only related to the excavation cycle time of the TBM, but also the tunnel logistics, which allowed operatives to adjust the tunnel transportation system elements at the surface, shaft and California crossings. The analysis of times determined the right moment to relocate the California crossing.
An excellent preventive maintenance schedule – together with a strong rail-traffic management system, including risk assessments for all activities – was established and implemented to increase the speed of the trains up to 20km/h, while minimizing the risk of derailments.
The BEAM had good correlation with the identified F features between 2,600m and 10,200m along the tunnel alignment. Simsima Limestone is favorable for TBM mining, except when associated with high groundwater pressure.
The author would like to thank all parties involved in the completion of this challenging tunnel.
