Strutted Excavation Exceeding the Action Level Next to a 1930s Masonry Building
A 7.2 m deep, 40 m by 25 m basement is being excavated in firm to soft marine clay using a contiguous bored pile wall (600 mm diameter at 750 mm centres) with three levels of steel struts. A four-storey load-bearing masonry building from the 1930s sits 5.5 m from the wall face on shallow footings. Instrumentation comprises inclinometers cast into the piles, vibrating-wire piezometers, surface settlement points, crack gauges and strain gauges on the top strut. Trigger levels are 20 mm (alert), 25 mm (action) and 35 mm (emergency). At the third excavation stage the inclinometer in a mid-wall panel reads 34 mm at 6.5 m depth, with a rate of 2.5 mm per day sustained over four days. Building settlement points show 11 mm maximum with 1:900 angular distortion and a 0.4 mm crack in a brick pier. The top strut load reads 780 kN against a design value of 950 kN. The contractor has 1.2 m of clay left to excavate and proposes omitting the lowest strut, which has not yet been installed, to bring the formation pour forward.
1. Interpret the monitoring data. What is the significance of deflection exceeding the action level while the strut load remains below design?
Read the worked answer
Deflection first. 34 mm is above the 25 mm action level and below the 35 mm emergency level, but the governing number is the rate: 2.5 mm per day over four days is roughly five times a typical serviceability threshold of 0.5 to 1 mm per day. A wall that has stabilised at 34 mm is a different problem from a wall moving at 2.5 mm per day, which is not stabilised and points to a mechanism that is still active.
Shape of the profile. Maximum movement at 6.5 m depth, just above formation level, is the classic bulge of basal heave in soft clay rather than a simple prop-supported cantilever. The passive resistance below formation is being mobilised and the toe is yielding.
Why the low strut load is not reassuring.
- Struts are preloaded to a proportion of design load; a strut that has not been re-tensioned, or has lost preload through temperature change or seating, reads low while the wall still moves.
- Load migrates downwards. With only two of the three struts installed, a large share of the lateral support is being provided by the soil below formation, which is exactly the mechanism the deflection profile suggests and exactly what basal heave destroys.
- Check the strain gauge installation and calibration before believing the number; a strut that is not bearing reads low.
What to check immediately. Piezometer response behind the wall (a rising water level or perched water indicates a leaking joint or a lost seal), the excavation sequence and elapsed time of the current stage, ground surface cracking, and whether the soffit of the masonry building has been surveyed recently.
2. Evaluate the contractor's proposal to omit the lowest strut. Support your answer with the mechanics, not just the monitoring thresholds.
Read the worked answer
Reject the proposal.
Mechanics. With the lowest strut omitted, the wall spans as a cantilever from the second strut to 7.2 m below ground, roughly 3 m of unsupported height plus the fixity developed in the clay below formation. Cantilever deflection scales approximately with the fourth power of the span and inversely with the wall stiffness, while the applied moment scales with the square of the span. Extending the unpropped length by even a metre at this location increases deflection and bending moment disproportionately, and the wall is already at 34 mm with 1.2 m of clay left to remove.
Consequence for the building. The building is 5.5 m away on shallow footings with 11 mm settlement and 1:900 angular distortion already recorded. A 1:900 distortion is within typical crack-control limits for framed buildings but marginal for unreinforced brickwork, and a 0.4 mm crack has already opened. A further 10 mm of wall movement would drag the settlement trough outwards and could take distortion past 1:500, which is the range where masonry damage becomes visible and costly to repair.
What to do instead. Install and preload the lowest strut as designed before removing the last 1.2 m; excavate the remaining clay in short panels with a central berm left in place; restrict plant loading on the berm; and, if the rate does not fall below 1 mm per day after strut installation, add cross-lot bracing or ground improvement rather than accelerate.
Decision rule. The contractor's programme gain is a few days. The downside is a damaged building, a停工 notice and a remedial grouting bill many times the saving. There is no engineering case for the change.
3. Set out the immediate actions, the revised monitoring regime and the criteria for resuming excavation.
Read the worked answer
Immediate actions (within the shift).
- Stop excavation in the affected panel and leave the remaining 1.2 m of clay in place as a berm.
- Install and preload the lowest strut to the design preload, then re-read the strain gauges.
- Notify the building owner and the supervising engineer, inspect the building jointly and photograph and gauge the existing crack.
- Check the piezometers and the wall joints for leakage; any seepage path must be sealed before further excavation, because water is the usual trigger for a sudden loss of passive resistance.
- Keep heavy plant off the crest within the active wedge.
Revised monitoring.
- Inclinometers twice daily until the rate falls below 0.5 mm per day, then daily for a week, then weekly.
- Settlement points and crack gauges daily, with a level survey of the building by an independent surveyor.
- Strut loads and piezometers daily.
- Set a new interim action level at 30 mm with a rate trigger of 1 mm per day, so that the next response is not delayed until 35 mm.
Criteria to resume.
- Rate below 0.5 mm per day for three consecutive days.
- Deflection trend asymptotic, not increasing with each excavation step.
- Strut loads consistent with the design apparent-pressure diagram after preload.
- Building distortion still better than 1:1000, no new cracks.
Back-analysis. Use the measured profile to back-figure the mobilised soil modulus and the apparent pressure diagram, and revise the design before the next excavation stage. If the back-analysed modulus is far below the design value, the wall section or the strut spacing needs review, not just tighter monitoring.