Swelling Claystone in a Headrace Tunnel: Support Redesign Under Squeezing Ground
A 3.5 km long, 6.0 m diameter headrace tunnel is being driven through Eocene flysch comprising alternating siltstone and claystone. X-ray diffraction on claystone samples returns 15–20% smectite, and laboratory swelling tests give a swelling pressure of 1.2 MPa. Overburden reaches 180 m. The contractor is using drill-and-blast with 3.5 m rounds and immediate 150 mm steel-fibre-reinforced shotcrete. Three weeks after passing chainage 2+150, the invert has heaved 400 mm, the shotcrete has developed parallel cracks with 10–20 mm opening, and steel arches are visibly deformed. Water inflow is modest (2–3 L/s), but the claystone is softening and slaking in the invert. The construction schedule allows a 6-week pause for redesign, and the client is concerned about long-term lining loads over the 50-year operating life.
1. Diagnose the ground behaviour. Is this swelling, squeezing, or both, and what evidence supports your diagnosis?
Read the worked answer
The evidence points to a combination, with swelling as the dominant driver in the invert.
The 15–20% smectite and measured swelling pressure of 1.2 MPa are direct mineralogical and laboratory evidence of expansive behaviour. The invert is where water accumulates and where confinement is lowest during construction, so it is the first place to heave: 400 mm of floor rise, slaking and softening are classic swelling responses. Shotcrete cracking parallel to the tunnel axis and deformation of steel arches indicate that the support has been loaded beyond its capacity, not merely that the ground is squeezing under overburden.
Squeezing is also present. At 180 m overburden, the vertical stress is roughly 4.5 MPa, and the tunnel is in weak, foliated flysch; significant time-dependent convergence is expected. However, pure squeezing would produce more uniform radial convergence and less preferential invert heave. The modest inflow of 2–3 L/s is enough to trigger smectite hydration but not enough to indicate a major water-bearing zone.
The practical distinction matters: squeezing is managed by yielding support and rapid ring closure, while swelling requires water control and accommodation space. Both must be addressed, or the permanent lining will be overstressed in service.
2. Redesign the excavation and support. Give dimensions, sequence and rationale.
Read the worked answer
The redesign should combine a yielding support system with a waterproofing and drainage strategy.
Excavation: reduce rounds to 1.5–2.0 m and use smooth blasting to limit overbreak and disturbance. Over-excavate by 300–500 mm to provide a compressible annulus. Use a systematic forepoling umbrella in the claystone intervals.
Support sequence:
- immediately after excavation, apply 75 mm of steel-fibre shotcrete, not 150 mm; a stiff thick shell attracts load and cracks before it can yield;
- install TH-section steel arches at 0.8–1.0 m spacing with sliding joints or compressible block elements designed to yield at 0.5–0.8 MPa;
- close the invert with a curved steel set and concrete invert strut within one round; leaving the invert open is unacceptable in swelling ground;
- place a 100 mm compressible layer (compressible board or geotextile wrap) between the ground and the final concrete lining;
- install a drainage layer and weep holes around the invert and use a sprayable waterproofing membrane, because water is the swelling trigger.
The final lining should be designed for a reduced, time-averaged swelling pressure, not the full 1.2 MPa, but only if drainage and the compressible layer are maintained for the design life. That requires a maintenance strategy for the drains, since blocked drainage would restore full swelling pressure.
3. Propose a monitoring and decision framework for the 6-week pause and for long-term operation.
Read the worked answer
During the pause, install a monitoring section at chainage 2+150 and at least two more in the claystone intervals.
Instruments: convergence arrays (five points) every 20 m; inverted floor heave points; piezometers in the invert; pressure cells behind the shotcrete; strain gauges on the steel arches; and load cells on the sliding joints. Read convergence daily for the first two weeks, then twice weekly.
The decision framework should be based on convergence rate, not absolute movement. A commonly used criterion is that the lining may be installed when the rate falls below 1–2 mm/month, or when the rate is decaying according to a fitted exponential curve. If the rate does not decay, the support is too stiff or drainage is inadequate, and the design must be revised again.
Long-term: monitor invert heave every six months for the first five years and annually thereafter; inspect drains and weep holes annually; set an action level of 25 mm of invert rise or 10 mm of additional convergence. A rise above 50 mm should trigger a drainage rehabilitation and a structural review. The contract should include a geotechnical baseline that distinguishes swelling ground from ordinary rock, so the contractor is not penalised for a condition that was always present.