Karstic Limestone Dam Foundation: Grout Curtain Adequacy and Cavity Treatment

Convert the test results first. One Lugeon is roughly 1 L/min per metre at 1 MPa, equivalent to about 1.3×10⁻⁷ m/s; 35–60 LU is therefore of the order of 5–8×10⁻⁶ m/s in the upper 20 m, while the 2–4 LU zone below 45 m is about 3–5×10⁻⁷ m/s.

A rough seepage check shows why 5 LU is too lax. With 40 m head across 20 m thickness and a 100 m wide abutment, k = 5×10⁻⁶ m/s gives q ≈ k·i·A ≈ 5×10⁻⁶ × 2 × 2,000 ≈ 0.02 m³/s, or 20 L/s - comparable to the entire spring. Karst conduits can bypass the grout entirely, so the spring is at real risk.

Depth: 60 m is defensible because it penetrates well below the 45 m low-permeability zone, but depth alone does not fix a single-row curtain in karst. Single rows are readily bypassed around solution features. I would require:

• a minimum three-row curtain in the upper 30 m, collapsing to two rows below;
• staged primary–secondary–tertiary holes with real-time pressure and take monitoring, using GIN or similar controlled grouting;
• acceptance tightened to 1–3 LU, verified by check holes between grout holes;
• grout mixes graded from cement-bentonite to microcement, with a contingency for polyurethane or hot bitumen if fine conduits do not respond;
• an independent review of the left abutment and a dye-tracing test before impoundment.

The cavity is the highest-risk feature because clay-silt filling will not accept normal cement grout and may be erodible under the future 40 m head. Treatment must start with characterisation, not blind grouting: run a borehole camera and cross-hole seismic tomography between BH-7 and neighbouring holes to establish cavity extent and whether it connects to the surface or to other solution features.

If the cavity is limited and accessible, I would flush it with water and air-lift, then fill it with a low-heat, high-slump concrete or cement-stabilised sand placed by tremie through a large-diameter borehole. For a clay-filled void that cannot be cleaned, permeation with a low-viscosity grout or jet grouting from an adjacent hole may be needed. Grouting pressures must be kept below the vertical effective stress at 28 m to avoid hydrofracturing; at a unit weight of about 20 kN/m³, that is roughly 0.5 MPa, so stage pressures should stay well below this.

Verification should include: post-treatment water-pressure tests in check holes reaching 1–3 LU; core recovery through the treated zone; a second camera inspection; and a dye-tracing or salinity test from the reservoir area to the downstream spring during first filling. If connection to the spring is demonstrated, impoundment should be conditional on a successful supplementary grouting campaign.

The plan must measure both pressure and flow, because karst failure often appears as turbidity or a sudden gain in spring discharge before piezometers respond.

Instrumentation:

• vibrating-wire piezometers in the curtain, in the karst zone and downstream of the abutment, at 10–15 m vertical spacing;
• V-notch weirs and continuous conductivity/turbidity loggers at the spring and at two downstream stations;
• surveying targets on the abutment and gallery walls for vertical and horizontal movement;
• pressure cells and joint meters in the grouting gallery;
• periodic dye-tracing during filling.

Action levels should be defined before filling, not after. Typical thresholds for a dam of this height: seepage increase of 5 L/s above baseline triggers daily inspection; 10 L/s or any turbidity above 10 NTU triggers a hold in filling and investigation; piezometric levels above the design phreatic line, or movement exceeding 10 mm, triggers drawdown. Filling should proceed in stages of 5 m head, with a 7–10 day observation period at each stage. Any turbid discharge is treated as a potential internal-erosion signal regardless of flow rate, and the spring must be sampled for limestone fines and dye.