The Foundation You Cannot See
A dam is only as safe as the rock it stands on, and the work that makes that rock safe is the one part of the structure you can never re-inspect once it is buried. This week, why foundations, not concrete, are the leading cause of concrete-dam failure; how the Lugeon test reads the tightness of rock before first fill; and a field record from the Krishna where a six-metre band of bad rock at ninety Lugeon was brought under three.
A.K. Sthapak, Managing Director, PCCI
From the Field
On 2 December 1959, the Malpasset arch dam above Fréjus in southern France failed on first filling and sent a wave down the Reyran valley that killed more than 420 people. The inquiry that followed did not blame the concrete. The thin double-curvature shell was sound. The dam failed because reservoir pressure drove uplift into a foliation plane in the gneiss of the left abutment, the rock moved, and the support the arch was leaning on simply left. Malpasset became the textbook proof of a hard idea: a dam is only as good as the rock it sits on, and that rock is almost never as good as it looks.
This is why the most consequential work on a concrete dam happens before the first lift of structural concrete is placed, in the foundation, where almost no one will ever see it again. The job has two halves. The first is to know how leaky the rock is. The second is to make it tight.
Knowing comes from the Lugeon test, a water-pressure test run in stages inside a sealed section of a borehole. One Lugeon is a water take of one litre per minute, per metre of test length, at an injection pressure of ten bar, which is roughly one megapascal. It corresponds to a rock permeability of about one ten-millionth of a metre per second. The number matters because it sets the target: for the foundation of an important dam, the grout curtain is commonly required to bring the rock down to between one and three Lugeon, tight enough that the reservoir cannot find a path through or under the structure.
Making it tight comes from grouting, and the two operations are not the same. Consolidation grouting is shallow and spread across the whole footprint of the dam, a blanket of holes that stiffens the upper, jointed rock and reduces near-surface seepage so the structure has a sound base to bear on. Curtain grouting is the deep, narrow line of holes driven near the heel, on the upstream side, to build a vertical barrier against seepage and, together with the downstream drainage, to hold down the uplift pressure that tries to float the dam off its base. The Indian standard is explicit about what the curtain is for.
Grout curtains are established under the heel of concrete and masonry dams to prevent erosion and loss of water from the reservoir, and, in conjunction with the drainage, to reduce uplift pressure.
IS 11293 (Part 2):1993, Guidelines for the Design of Grout Curtains, Foreword (now consolidated into the current IS 11293:2018).
A curtain is not drilled all at once. It is built by split spacing: a line of widely spaced primary holes, then secondary holes between them, then tertiary holes between those, each round closing the gaps the last one missed, until a verification hole confirms the rock is below the target. The principle is that you never know in advance where the bad rock is. You find it by grouting.
The Krishna river gives a clean illustration. The Lal Bahadur Shastri Dam at Almatti, the masonry gravity dam completed in North Karnataka in 2005, was grouted from a foundation gallery running more than a kilometre, Blocks 1 through 52. For most of that length the rock was fresh, hard gneiss and the curtain was routine. Then the foundation changed character. In one stretch the holes hit artesian water during drilling, the pressure in the rock so high that, in the words of the contractor's own record, the jets came out "practically touching the roof of the gallery." Beyond it lay the worst of it: a weak band up to six metres thick where quartzite met the granite beneath, filled with weathered rock and broken fragments, core recovery as low as one part in ten, permeability running as high as ninety Lugeon. Drill rods dropped suddenly into voids. One stage in one hole took 3,200 kilograms of cement before it would hold pressure.
The treatment was to grout the band down through primary, secondary and tertiary lines at pressures stepped up to ten kilograms per square centimetre, and then to drive a second curtain line from the upstream heel where the worst blocks were. The result, as the grouting contractor reported it in a case history presented to the Fifth International Conference on Case Histories in Geotechnical Engineering, was unambiguous.
Single line grout curtain of permeability less than 3 Lugeon has been effectively formed below the foundation, despite very poor strata having maximum pre-grout permeability of 90 Lugeon.
Bidasaria, M., "Treatment of Unconformity Zone and Curtain Grouting in Foundation of Almatti Dam," 5th ICCHGE, 2004.
Post-grout test holes through the worst blocks read between 1.62 and 3.00 Lugeon. Ninety brought under three. These are the contractor's reported figures, presented to a peer conference rather than audited by the owner, and they should be read as such; but they are internally consistent and they describe exactly the job the curtain is meant to do.
The reason all of this has to be right the first time is that none of it can be done again. Once the concrete covers the foundation and the reservoir fills behind it, the grout curtain is buried under millions of tonnes and head of water. You cannot excavate it, you cannot re-inspect it, and fixing it is a different order of problem. Wolf Creek Dam in Kentucky, built on cavernous limestone, spent decades chasing seepage that found its way around a first round of grouting; a concrete cut-off wall slowed but did not stop it as the water found new paths, and the final remedy ran to a deeper wall and a double line of grout to more than three hundred feet. The lesson of the buried curtain is that it costs a fraction to build correctly and a fortune to correct.
The lesson: The foundation is the one part of a dam you treat blind and bury permanently, and it is where dams most often fail. Test the rock with water before you trust it, in Lugeon units against a stated target. Grout to that target and verify it with holes drilled for the purpose, not assumed from the takes. Treat the curtain as unrepairable, because for all practical purposes it is. The concrete you can see is rarely the concrete that decides whether the dam stands.
Read more: Dam Foundation Grouting: Curtain, Consolidation, Contact →
Did You Know?
53%
Foundation deficiencies, not weak concrete, are the single largest cause of concrete-dam failure.
In the body of failures compiled after the International Commission on Large Dams (ICOLD) analysis, the foundation was the culprit in 29 of 55 documented concrete-dam failures, about 53 percent, with uplift accounting for several more. Malpasset is the case everyone remembers, but it is not an outlier. It is the rule. The most consequential concrete on a dam is the part you bury in the rock, which is exactly why the water-pressure test before first fill is the last honest look anyone gets at the foundation the curtain is meant to seal.
Sources: U.S. Bureau of Reclamation, Dam Safety, "Concrete Dam Foundation Risks" (citing ICOLD failure statistics); ICOLD analyses of dam failures by cause.
Worth Knowing
Dam Foundation Grouting: Curtain, Consolidation, Contact. 3 Methods, Pressures, and QC Criteria
The PCCI master guide to the three grouting operations under a dam, how they differ in depth and purpose, the pressures they run at, and the QC criteria that tell you when the rock is tight enough to stop.
Foundation Contact Grouting for Concrete Dams: Timing, Method, and Verification
The PCCI brief on the grouting that seals the contact between the concrete and the rock it sits on, when in the construction sequence it is done, and how the seal is verified.
Water Testing for Grouting (Reclamation Geology Field Manual, Chapter 16)
The USBR field reference on the Lugeon test itself: how the stepped pressure stages are run, how the value is computed, and how to read the flow pattern that tells you whether the rock is taking water by laminar flow, turbulence, dilation, or wash-out.
Newsletter
Concrete Pulse
Stay ahead on concrete technology. Subscribe to our weekly newsletter. Field-tested insights on mass concrete, dam engineering, and QA/QC, delivered straight to your inbox.
Free. No spam. Unsubscribe anytime.