41 Rizos Et Al-2018-Geomechanics and Tunnelling
41 Rizos Et Al-2018-Geomechanics and Tunnelling
41 Rizos Et Al-2018-Geomechanics and Tunnelling
76 © 2018 Ernst & Sohn Verlag für Architektur und technische Wissenschaften GmbH & Co. KG, Berlin · Geomechanics and Tunnelling 11 (2018), No. 1
D. Rizos/D. Williams/A. Fouda/T. Amin/M. A. Dshiesh/A. D. Nicola · TBM tunnelling under the Suez Canal – Port Said tunnels in challenging ground conditions
first 12 m at 2.5 MPa and achieves values of 6 to 7 MPa at was present in all boreholes on both sides of the Suez Ca
sea level –40 m. nal, which could have a significant impact on the safety of
the tunnel works. As the TBMs were manufactured prior
2.1 Tunnel geotechnical profile to the geotechnical campaign that encountered the gas
hazard, several improvements to existing TBM parts and
The geotechnical profile along the tunnel (see Figure 4) is equipment would clearly be required to reduce the fire/
based on a large number of site investigation boreholes explosion risk.
and the holes drilled for gas venting, therefore has a high Additional investigations were carried out (including
degree of confidence. It was not possible to conduct any geophysics to identify potential gas pockets under pres
drilling below the Suez Canal due to restrictions imposed sure), together with studies to determine suitable mitiga
by the Suez Canal freight traffic rules. For the first 0.8 km, tion measures. One of the mitigation measures was to vent
the tunnels are driven with a downhill gradient in the gas present in the ground along the alignment by drilling
CL 1 horizon, followed by 1.15 km mainly in the SL 2, boreholes (three rows at 25 m spacing) to below the tun
then 0.35 km in mixed soil conditions (CL 1/SL 2) and nel invert, ahead of TBM excavation. The boreholes were
finally 0.55 km again in CL 1. The tunnels below the Suez close to the tunnels so they were sealed before TBM ar
Canal were mostly excavated in SL2, with occasional rival.
lenses of CL 3 encountered in the lower tunnel profile As with the site investigation, the 350 m section be
and also sections with the CL1 in the upper tunnel pro low the Suez Canal could not be penetrated by gas vent
file. ing boreholes. The gas hazard in this section was managed
during tunnel excavation by enhanced monitoring and
2.2 Groundwater strict working practices (including continuous manual
flushing of the air bubble in the working chamber). This
Groundwater in the CL1 is at sea level +1 m on the east decision was reinforced by the experience gained in the
ern and slightly lower at sea level 0 m on the western side sections preceding the planned stop at safe haven 4
of the canal. A confined aquifer occurs in SL 2 with a (SH 4) before the canal.
groundwater head sea level +1 to +2 m. Groundwater The potential for methane gas entering into the atmos
chemical analysis reveals very high aggressivity with very phere of the TBM and tunnel is partially mitigated by the
high salinity of 3.8 %, sulphate content of 0.4 % and a choice of TBM itself. Slurry TBMs are more favourable in
chloride content varying from 1 to 5 % with a mean of 2 %. this respect, as any groundwater containing methane is
evacuated through the TBM slurry circuit to the slurry
3 Risks/challenges – mitigation measures treatment plant (STP) on the surface. Modifications to the
3.1 Methane gas occurrence TBM, including enhanced ventilation, extension of the
Samson valve outlet to the rear of the backup and installa
In April and May 2016, during drilling works on both tion of explosion-proof equipment, were required (includ
sides of the Suez Canal, boreholes penetrating into the SL ing the multi service vehicles). The majority of these
2 encountered methane gas, which in one instance ignited works (and the STP upgrade) were carried out during the
at the surface. Gas monitoring confirmed that methane TBM stoppage at the first safe haven (SH 6) in the early
part of the tunnel drives, in the CL 1 where the methane by distorting the tunnel tube in the longitudinal direc
content was lower. tion, causing sliding and an unacceptable gap opening
The adoption of strict working practices and en between the rings of > 10 mm, if no measures were ap
hanced gas monitoring were applied during the entire plied. Such a gap would adversely impact the lining
tunnel drives. In regard to TBM cutterhead interventions, watertightness with severe consequences for the dura
the risk from methane gas was carefully evaluated and the bility and serviceability of the tubes. Draining of the
use of safe havens and free air interventions minimized CL 1 (due to loss of watertightness) could thus also
the risk. For the hyperbaric interventions, intensive gas cause uncontrolled settlement due to clay consolida
monitoring and the venting of the gas through boreholes tion. To mitigate this, a transition zone was construct
controlled the risk to acceptable levels. ed, consisting of a system of plastic concrete diaphragm
walls and barrettes appropriately distributed to provide
3.2 Very soft clay a gradual stiffness gradient of the tunnel-soil. Figure 6
shows the optimized transition zone at the launching
The upper CL 1 clay layer is a very soft clay for the first 12 shafts (similar for the reception shafts). Such a configu
to 15 m and gradually becomes stiffer with depth. ration allows smoother deformation and curvature of
the segmental lining tube in the longitudinal direction,
3.2.1 CL 1 consolidation status which distributes the gap opening in more than one
circumferential joint.
In addition to the very low mechanical properties, a major In addition, to mitigate geotechnical model uncer
concern was the consolidation status, i.e. under-consoli tainties, a thorough sensitivity analysis was performed
dated (UC), normally consolidated (NC) or over-consoli with respect to the soil parameters, the constitutive mod
dated. The consolidation status could impact significantly els and the possible soil creep effect. The results demon
on the design concept in terms of expected ground defor strated that the main design assumptions are sufficiently
mation (creep) and therefore tunnel structure behaviour conservative and resulted in upper limit values of heave.
in the short and long terms. Figure 7 illustrates the recorded vertical deformation of
Relatively recent dredging and dumping of material the south tunnel tube, with only 15 mm heave occurring
on the east bank of the Suez Canal was found to have and no ongoing deformation for the last six months. This
influenced the consolidation status of the clay. In addi short term heave is less than the corresponding calculated
tion, it was known from construction in the nearby Port figure. The longer term behaviour of the soil-tunnel system
Said port area that the clay is in general under- to nor will be monitored until commissioning using the 3D tar
mally consolidated with resulting ongoing creep defor gets, extensometers and vibrating wire piezometers. The
mation. Therefore, a comprehensive investigation pro requirements for long-term monitoring will be determined
gram was carried out to verify the clay status, with labo based on these results.
ratory testing and numerous in-situ investigations includ
ing CPTU 100 % dissipation tests and the installation of a 3.2.3 TBM tilting
large number of Vibrating Wire Piezometers (VWP). This
campaign concluded that the CL 1 clay is under-consoli The soft clay presented a potential risk of TBM tilting,
dated for 200 m from the eastern bank, while for the re especially since the tunnel invert at the launching area
mainder of the drives it is normally consolidated. Fortu was only 22.5 m deep where the clay is still soft. The tilt
nately, in the UC clay section adjacent to the canal, the ing potential was checked for both normal operation and
tunnel tubes are within the SL 2 sand layer and therefore during any unforeseen stoppage. The risk of tilting was
there is no direct impact on the tunnel structure. SH 4, exacerbated given the tunnel drives launched with a
part of which is within the UC CL1, was studied and downhill gradient of 3.3 %. The transition zone ground
measures were taken to compensate the significant nega improvement at the launching shafts extended to 50 m
tive skin friction and avoid development of long term ahead of the shaft and contributed to the distribution
settlement. and support of the weight of the shield. Outside this
zone, it was verified that the submerged weight is very
3.2.2 Differential ground movements small during normal operation, while for unforeseen
stoppages, tilting could be prevented if only 50 % slurry
The response of the tunnel structure to potential differ drawdown is allowed. The latter is also constrained by
ential movements between the „hard points“ (e.g. the the blow-out check verification in the low overburden
launch/reception shaft and the safe havens) and the sections.
free field, where the tunnel is effectively „floating“ in
the soft clay, presents a major engineering risk. The 3.2.4 Surface development exclusions above the tunnels
high stiffness contrast between the launching shaft,
founded in the denser SL 2 sand, and the very soft clay Due to the sensitive nature of the CL 1, any future infra
has a significant impact on the segmental lining. Analy structure or building development above the tunnels
sis indicated that the expected tunnel lining heave (due could cause ground movement and consequent unaccep
to uplift forces and the long term dissipation of the pore table tunnel deformation. To prevent this, an exclusion
pressures generated by the tunnelling) would be about zone has been defined, within which any proposed future
60 mm only a few metres from the shaft. This significant development will be subject to careful studies to verify any
movement would be detrimental to the segmental lining impact on the tunnels.
3.3 TBM cutterhead clogging potential head and excavation chamber is significantly increased.
CL 3 was expected to be encountered in the deepest sec
The clogging potential was investigated for both clay lay tions of the tunnel alignment, where hyperbaric interven
ers CL 1 and CL 3 (Figure 8). The results showed that the tion pressures exceeded 6 bar, making potential manual
CL 1 has, in general, low consistency and despite its high cleaning a very time consuming and riskier procedure.
plasticity index, low clogging potential. A higher clogging Given the clogging potential for the deeper part of
potential was recognised for limited sections of the deeper CL 1 and CL 3 and the likelihood of encountering the lat
part of CL 1. ter below the Suez Canal, modifications were made on
For CL 3, all test results indicated a high to very high site prior to TBM assembly, including opening of the cut
plasticity, which, combined with a stiff to very stiff consist terhead (from 31 to 46 %) and improving face flushing
ency, gave a high clogging potential. Based on previous with additional nozzles and a higher flushing capacity.
experience in tunnel sections where clays with high clog With the above modifications, clay clogging had little
ging potential (such as CL 3) represent more than 20 % of impact on the TBM cutterhead. However initial progress
the tunnel face, the risk of massive clogging of the cutter problems occurred (learning curve at the eastern side) due
to the STP blockage of the screens, which were upgraded To minimize the possibility of TBM stops under the
resulting in improved progress in the clays on the western canal, extensive cutterhead maintenance was carried out
side. at SH 4 (approx. 70 m before the canal) along with all ca
bles being extended in advance to avoid plant stoppages.
3.4 Tunnel drive below the Suez Canal During excavation below the canal, the operation
team encountered inevitable excavation chamber pressure
A major concern was the geological uncertainty given the peaks during pipe extension. To minimize these peaks, the
lack of site investigation data below the canal. This, along slurry level was maintained at the manlock access level,
with the presence of methane gas, the abrasiveness of the while the evacuated slurry was discharged directly to the
SL 2 and implications for tool wear, and the relatively low slurry treatment plant (STP). Figure 10 shows the oscilla
cover to the canal bed, were all considered carefully in the tion of the actual face pressure of the TBM 1 (north) with
design and in the TBM operational parameters. fluctuation around the target value being occasionally
The recent bathymetric survey revealed the canal bed higher than 10 kPa, but in general within the limits. No
is 2 m deeper than anticipated during the finalization of slurry blow-out or face instability was recorded during the
the tunnel alignment – the cover below the canal being passage below the canal.
18 m rather than the assumed 20 m. This resulted in a During the short stoppages for ring building, settle
maximum allowed slurry pressure very close to the mini ment of sand particles caused a blockage of the opening in
mum required for face stability. Figure 9 shows the very the submerged wall, resulting in slight pressure spikes
small bandwidth of the design operational parameters when excavation was resumed since the slurry was sup
with the minimum pressure at the upper pressure cell plied to the chamber but the discharge line was temporar
equal to 472 kPa and the maximum allowed pressure of ily blocked. To prevent this, prior to advancing the TBM,
502 kPa (540 kPa with a reduced safety factor). Clearly the cutterhead was rotated to mix the settled sand parti
careful control of the face pressure was required and spe cles and unblock the opening. Other small pressure spikes
cial training sessions held to prepare the TBM operation occurred during TBM standstill due to small leakages
team. from the feed pipe valves.
Fig. 9. Face pressure along the tunnel alignment – minimum required; maximum allowed; target value for safe operation
Fig. 10. North TBM – face pressure oscillation below the canal lower point
Observations of the discharge at the STP showed The crossings under the latter two were driven with over
that below the canal, both TBMs excavated a face mainly burden of less than one tunnel diameter (approx. 10 m).
in SL 2, but with either a small portion of CL 1 in the Both machines successfully traversed all these critical
crown or CL 3 at the invert. The former condition had points.
potential for soil instability, but this risk was mitigated Figure 11 shows the monitored settlements of the
by maintaining good bentonite quality and being pre railway line which were within the expected values. Dur
pared to inject fresh slurry into the shield annular gap if ing the crossing of the railway line, the operator applied
necessary. precautionary measures such as reducing the speed of
Both TBMs passed under the canal without any ma trains.
jor incident apart from very small overbreaks, subsequent For the crossings below the gas pipeline, the pipeline
ly filled during ring grouting, and some minor unavoidable operator provided staff on standby in case of any adverse
stoppages due to electrical problems. event. Careful control of the TBM parameters (face pres
sure and grout pressure) led to minimal settlement of the
3.5 Critical points along the alignment gas pipeline.
The ground above the tunnels is largely undeveloped, be 3.6 Concrete segment durability
ing desert to the east of the canal and mostly agricultural
land to the west. The west side however has four critical The saline groundwater raised durability concerns for the
points which are: single-pass concrete segmental lining. The durability and
– The Port Said – Ismailia single track railway, concrete mix design were analysed thoroughly to comply
– The main four-lane highway, with both BS 8500 [2] and CS 163-2008 [3]. Cement
– A natural gas pipeline, CEMIII-A+SR (with a high percentage of ground granu
– The Altina Canal. lated blast furnace slag and silica fume) was used for the
SH 6 23+222 Plastic concrete plug 14 m TBM maintenance and modification for
methane gas under atmospheric conditions
SH 4 22+330 Plastic concrete plug + 11.5 m and 24 m TBM maintenance before the canal under
dewatering watertight enclosure atmospheric conditions
SH 2 21+442 Plastic concrete plug + 4.6 m TBM maintenance with pressure 2.9 bars and
dewatering controlled dewatering
chines (and the STP) at this stop and carried out under
atmospheric conditions.
The most critical and demanding maintenance stop
was before the canal at SH 4, where the machines were
inspected thoroughly and all cutting tools replaced. SH 4
is located at the lowest point of the tunnel with a hydro
static pressure up to 6 bar. As the required maintenance
was planned in as short a time as practically possible,
SH 4 was designed to allow an atmospheric intervention.
Hence, the safe haven comprises a soil substitution block
to accommodate the cutterhead in safe face conditions
within a watertight enclosure of diaphragm walls and
deep pumping wells, to lower the groundwater to accom
modate the shield in dry conditions (Figure 12). The tun
nel profiles were in the water-bearing SL 2 and the enclo
sure was plugged by the CL 1 above and by the CL 3 at the
base (the diaphragm panels were embedded in the CL 3).
Both TBM interventions were carried out successfully un
der atmospheric pressure, despite a slower than expected
dewatering process raising concerns about the watertight
ness of the system. The interventions were carried out al
most concurrently as both machines entered the safe hav
en only a few days apart.
The fast track nature of the project meant ongoing
design development during construction. Both SH 4
and SH 6 were initially positions of cross passages (CP).
Therefore, at these locations, the soil substitution
blocks were relatively large as they were also planned
as soil stabilization measures for the proposed CP con
Fig. 12. Layout of SH 4 struction.
SH 2 was designed without consideration of a poten
tial CP, being located at the western side of the canal and
segment concrete (grade C45/55) with cover to reinforce the first planned stop for maintenance after the 0.89 km
ment of 50 mm. High performance criteria were specified drive through mixed soil conditions and below the canal.
with maximum allowed water penetration 10 mm; chlo At this location, the tunnel face is in mixed face condi
ride ion penetration 2,000 coulombs; maximum water ab tions (50-50 CL 1 and SL 2) with a hydrostatic pressure of
sorption 2 %. 5 bar. The length of SH 2 is only 4.6 m to provide a stabi
lizing plug in front of the cutterhead of 2.3 m. As the block
4 TBM maintenance and safe havens is small and the shield lies mostly within the mixed soil
formation, hyperbaric interventions were carried out un
Safe and efficient TBM operation is only ensured if regu der an air pressure of 2.9 bar. This pressure was enabled
lar inspections and maintenance take place. Three stops by a controlled dewatering system comprising deep wells
were planned (excluding the stop in the launch starting and submersible pumps (at sea level –55 m) in the SL2,
plug) at selected locations for maintenance of the cutter reducing the confined water aquifer head by 20 m. Both
head and brushes (Table 1). TBMs successfully completed the maintenance stop at
The first stop at SH 6, 350 m from the launching SH 2. The intervention, which lasted ten days (including
shaft, was planned before reaching the abrasive SL 2 and dewatering, inspection, maintenance and relaunching),
the deeper potentially sticky clay. Modifications due to required cleaning, significant tool replacement and repairs
the occurrence of methane gas were made to the ma to the mixing arms.
6 Conclusion
Derek Williams
Systra Egypt Branch
World Trade Center Complex Buildings Mohamed Aboudshiesh
1191 Corniche El Nile Arab Contractors
Cairo, Egypt 11221 34 Adly Steet, Cairo, Egypt
[email protected] [email protected]