Case study

Braced Excavation Basal Heave Case Study

Civil engineering case study on braced excavation basal heave, hydraulic uplift, soft-clay base stability, piezometric head, stop-work decisions, sequence change, and validation evidence.

A braced excavation can look controlled at the wall line while the base is becoming unsafe. Struts may keep wall deflection within trigger values, yet the excavation bottom can heave because the soft clay below the base has insufficient undrained resistance or because upward groundwater pressure reduces the effective confinement.

This case study follows a hypothetical deep excavation where the monitoring team observes upward movement at the base after rainfall and dewatering changes. The case is written for engineering education. It shows how a civil engineer should connect basal heave, hydraulic uplift, piezometric head, construction sequence, stop-work authority, and validation evidence.

The central question is:

Can excavation continue because wall movement and strut loads remain below red triggers, or must the next stage be stopped because base stability is no longer credible?

The correct decision is to stop the next excavation stage until the base-stability mechanism is understood and corrected. Wall movement is only one part of excavation safety.

Case Context

A two-level braced excavation is being constructed beside utilities and an existing masonry building. The excavation is in soft clay with a confined water-bearing layer below the planned base. The support system is a soldier-pile and lagging wall with walers and internal struts.

ItemValue or observation
Current excavation depth10.5\ \text{m}
Unit weight of removed soil18.5\ \text{kN/m}^3
Temporary surcharge near edge18\ \text{kPa}
Verified undrained shear strength below base34\ \text{kPa}
Screening bearing factor for base heaveN_c=5.7
Minimum required basal-heave factor of safety1.5
Confined piezometric head above excavation base6.0\ \text{m}
Clay plug thickness above confined layer2.8\ \text{m}
Saturated unit weight of clay plug19.0\ \text{kN/m}^3
Water unit weight9.81\ \text{kN/m}^3

The values are simplified. A real basal-heave review must use the approved ground model, construction sequence, wall embedment, excavation width, soil anisotropy, strain-softening behavior, groundwater observations, dewatering influence, surcharge control, and the governing design standard.

Field Evidence

The warning signs are not dramatic at first:

EvidenceEngineering interpretation
survey points on the exposed base rise by 35\ \text{mm} over two shiftspossible basal heave or rebound beyond predicted elastic recovery
piezometers below the base show a confined head increase after rainfallupward water pressure has increased
wall inclinometers remain below red movement triggerwall stability alone does not prove base stability
lower strut load drops slightly while base survey risessupport force may be redistributing as the base moves
small cracks appear in the working platformdeformation is reaching the construction surface

The observation is a coupled geotechnical problem. A monitoring table that checks only wall deflection could miss the governing failure mode.

Basal-Heave Screening

Use a simplified undrained basal-heave screen:

\displaystyle FS_{heave}=\frac{N_c s_u}{\gamma H+q}

where:

  • N_c is a screening bearing factor for the base mechanism;
  • s_u is undrained shear strength below the excavation base;
  • \gamma H is the overburden stress removed by excavation depth;
  • q is surcharge near the excavation.

Compute the driving stress:

\gamma H+q=18.5(10.5)+18
\gamma H+q=212.25\ \text{kPa}

Compute the undrained resistance term:

N_c s_u=5.7(34)=193.8\ \text{kPa}

The screening factor of safety is:

\displaystyle FS_{heave}=\frac{193.8}{212.25}=0.91

This is below the required value:

0.91<1.5

The calculation is simplified, but it is not a small shortfall. The current excavation state should be treated as outside the accepted design basis.

Hydraulic Uplift Check

The confined water pressure below the clay plug also needs checking. The upward water pressure is:

u=\gamma_w h
u=9.81(6.0)=58.9\ \text{kPa}

The downward total stress from the clay plug is:

\sigma_v=\gamma_{sat}t
\sigma_v=19.0(2.8)=53.2\ \text{kPa}

The simplified uplift factor of safety is:

\displaystyle FS_{uplift}=\frac{\sigma_v}{u}=\frac{53.2}{58.9}=0.90

Again, the result is below unity. The base is not only weak in undrained bearing; it is also vulnerable to upward hydraulic pressure.

Engineering Decision

The next excavation stage should be held. The decision basis is:

  1. observed base heave is trending upward;
  2. the basal-heave screen gives FS_{heave}=0.91;
  3. the hydraulic uplift screen gives FS_{uplift}=0.90;
  4. the confined head increased after rainfall and dewatering changes;
  5. continuing excavation would reduce confinement further;
  6. wall movement below a red trigger does not remove the base failure mode.

The field instruction should be explicit:

Stop further excavation in the affected bay, remove nonessential surcharge, restrict personnel exposure at the base, stabilize drainage and piezometric control, and require engineering review before any deeper cut.

This is a construction-stage safety decision, not only a calculation update.

Corrective Sequence

The reviewed correction combines operational and physical controls:

  • remove temporary surcharge from the excavation edge;
  • stop excavation at the current level until the base response stabilizes;
  • install relief wells to reduce the confined head;
  • place a controlled working platform and localized treated plug in the critical bay;
  • verify the treated plug strength before release;
  • define a staged excavation sequence with smaller depth increments;
  • increase base survey and piezometer reading frequency after rainfall.

After correction, the reviewed values for the release bay are:

ItemCorrected value
effective reviewed excavation depth for release stage10.5\ \text{m}
controlled surcharge near edge5\ \text{kPa}
verified equivalent undrained resistance in treated plug55\ \text{kPa}
confined head after relief wells3.2\ \text{m}
clay plug thickness retained2.8\ \text{m}
added working-platform equivalent stress16\ \text{kPa}

Recalculate basal heave:

\gamma H+q=18.5(10.5)+5=199.25\ \text{kPa}
N_c s_u=5.7(55)=313.5\ \text{kPa}
\displaystyle FS_{heave,new}=\frac{313.5}{199.25}=1.57

This clears the screening requirement:

1.57>1.5

Recalculate uplift. Upward pressure after relief wells:

u_{new}=9.81(3.2)=31.4\ \text{kPa}

Downward stress from clay plug and working platform:

\sigma_{v,new}=19.0(2.8)+16=69.2\ \text{kPa}
\displaystyle FS_{uplift,new}=\frac{69.2}{31.4}=2.20

The corrected condition is no longer controlled by the same immediate uplift screen, but the excavation should still proceed by hold points because soil improvement, piezometric control, and construction tolerance are variable field controls.

Monitoring Interpretation

The monitoring review should combine base movement, piezometric head, wall movement, and strut load:

QuantityBefore correctionAfter correction target
base heave trend35\ \text{mm} in two shiftsstable or reducing rate
confined head above base6.0\ \text{m}\le 3.2\ \text{m} for release
surcharge near edge18\ \text{kPa}\le 5\ \text{kPa}
basal-heave screen0.911.57
uplift screen0.902.20

A single green wall-deflection reading cannot override adverse base and groundwater indicators. The excavation should be released only when the governing mechanism is controlled and the data trend supports the revised model.

RPN Screen

A simple risk-priority-number screen can document why the response escalated:

RPN=S \times O \times D

Before correction:

FactorValueRationale
Severity S10Basal instability can expose workers, utilities, and adjacent structures to sudden ground movement.
Occurrence O4Soft clay, surcharge, deep excavation, and elevated head make the mechanism credible.
Detection D5Wall monitoring may not detect base failure early unless base and piezometer data are reviewed together.
RPN_{initial}=10(4)(5)=200

After relief wells, surcharge removal, treated plug verification, and staged release:

FactorValueRationale
Severity S10Consequence remains high if the mechanism returns.
Occurrence O2Physical controls reduce likelihood.
Detection D2Base survey and piezometer hold points improve detectability.
RPN_{controlled}=10(2)(2)=40

The RPN does not approve excavation by itself. It records why the failure mode became better controlled after the engineered intervention and monitoring update.

Validation Evidence

The release package should include:

Evidence itemWhy it matters
piezometer trend before and after relief wellsproves uplift pressure was reduced
base survey trendconfirms heave has stabilized or slowed within the release criterion
treated-plug strength recordssupports the revised s_u value used in the calculation
surcharge exclusion inspectionconfirms the lower q value is real on site
excavation stage hold-point recordprevents the sequence from drifting back to the failed condition
wall and strut monitoringverifies that correcting base stability did not create another support problem
rainfall and pumping logexplains groundwater changes and validates the response
independent engineering reviewconfirms the calculation model and field evidence are consistent

The closeout should state what excavation depth is released, which bay or chainage is covered, which groundwater head limit controls the next stage, and who can stop work if the trend reverses.

Engineering Lessons

The first lesson is that a braced excavation is not safe just because the wall has not reached a red deflection trigger. Base stability can govern before wall strength or serviceability appears critical.

The second lesson is that surcharge, excavation depth, undrained strength, and groundwater head interact. A modest field change in any one of them can remove the margin in soft ground.

The third lesson is that upward movement at the base should be treated as diagnostic evidence, not as harmless rebound, until the mechanism is checked. Trend, rate, groundwater, and sequence matter more than a single isolated survey reading.

The final lesson is that release after a basal-heave warning requires evidence. Relief wells, surcharge control, plug verification, base survey, piezometer trend, and hold-point records must all support the decision to continue.

REF

See also