An embankment dam can be damaged by internal erosion. During this process, the soil material erodes and is transported out of the dam structure. Some common types of damage due to internal erosion are piping, less dense soil zones, and zones where the fine material has been washed out.

To prevent ongoing internal erosion from developing into a damage or a breach, injection grouting may be performed to replace the eroded soil material. In an embankment dam, injection grouting is usually performed vertically. A pipeline is drilled from the crest of the dam to the damaged zone, to which the grout material is delivered via a pump. The injection grouting methods suitable for embankment dams are compaction and permeation grouting. In compaction grouting, a lowmobility grout material is injected at high pressure, compacting the surrounding soil. In permeation grouting, the injection pressure is lower and the grout material flows more easily, allowing it to permeate the porous structure of the soil into which it is injected.

A grout material for embankment dams should have properties, i.e., particle size distribution, water content, shear strength, and bulk density, similar to those of the original core soil after injection. A grout material with these properties will however be very stiff and difficult to pump, and permeation will be difficult to achieve. Therefore, a new type of non-hardening grout material has been developed and tested in the laboratory. The grout material is a low-mobility grout, but its viscosity and yield strength can be temporarily lowered by replacing the fine aggregates with a limestone filler and by adding a superplasticizer. After injection, the effect of the superplasticizer wears off, leaving a grouted zone with geotechnical characteristics similar to those of the original core soil. The grout material consists of 0–4 mm aggregates, limestone filler, dry bentonite powder, water, superplasticizer, and an air-release agent. The grout material properties and the influence of injection method were tested in three laboratory investigations and the results were presented in four papers.

Development, and fresh and hardened properties of the grout material were investigated in Paper I. The key findings are: (1) The grout material attracted air when homogenized. When homogenized longer than 15 minutes, it was difficult to pump. Air content up to 16.5 % was observed. (2) After 34 days of storage, the water content was ~10 % and the bulk density ~2250 kg/m3, which are very similar to those of the core soil. The undrained shear strength was ~13 kPa, which was initially lower than that of the core soil but it slowly increased with time.

The factors affecting a grout material’s ability to permeate a core soil damaged by internal erosion were investigated in a pilot-scale permeation test series and the results are presented in Paper II. Three different coarse-grained materials with d15of 35, 75, and 110 mm were grouted. The key findings are: (1) The ratio between limestone filler and aggregates in the grout material greatly influenced the permeation. A grout material with a ratio of 1.7 performed far better than a grout material with a ratio of 1.4. (2) A higher consistency measurement of the groutmaterial (150 mm vs. 100 mm) improved the permeation if low injection pressure was used. At higher pressure, the role of consistency was minor. (3) A higher maximum particle size, Dmax, of the grout material (4 mm compared to 2 mm) improved the permeation. The difference was most probably caused by higher viscosity and higher yield strength of the grout material with Dmax = 2 mm compared to that with Dmax = 4 mm. The lowest ratio between the minimum particle size of the coarse-grained material and the maximum particle size of the grout material was 4, and using higher pressure, the grouting was successful. Ratios below 4 were not tested.

From Paper II, the most suitable materials for permeation grouting were chosen to be investigated further for their resistance to flow. The results of this investigation were presented in Paper III. Resistance to flow in the pilot-scale permeation test was found to occur within the pipeline and at its exit, where the grout material downflow was redirected 180° to an upward flow. Total frictional losses could be estimated by regarding the grout material flow as Newtonian laminar. Total frictional losses in the 1.3 m length and 0.075 m diameter pipeline during all testswere measured to be 1–67 kPa/m at grout material velocity of 0.01–1.03 m/s. Frictional losses due to the grout material’s permeation of the coarse-grained materials could be estimated with the hydraulic conductivity. The mean hydraulic conductivities in the d15 = 75 mm coarse-grained material, when permeated by the Dmax = 2 mm and 4 mm grout materials, were measured to be 1.7x10-4 and 1.4x10-4m/s, respectively; where as in d15 = 110 mm, the values were 2.1x10-4 and 3.3x10-4m/s. These observed values of the hydraulic conductivity were very close to the expected values. With the Newtonian approach, pressure losses may be easily estimated. This will facilitate the estimation of how much of the grouting pressure at the pump is transferred into the core soil during a grouting operation. The possibility to quantify pressure losses during the permeation of the coarse-grained material with hydraulic conductivities can be used when estimating permeation depths vis-à-vis applied injection pressure.

The method, “Identification – Localization – Characterization – Remediation”, was tested at the abutment of a large-scale test embankment dam with the newly developed grout material and presented in Paper IV. The seepage rate was successfully reduced to 40 % directly after the injection grouting, and up to 70 % after one year. Most of the seepage reduction was caused by the rotary percussion drilling. Remedial grouting should not be regarded as a last resort, but as a part of a maintenance program.