The Hodges Village dam is part of an existing Corps of Engineers flood control project located on the French river, near Oxford, Massachusetts in the US. Constructed in 1959, this embankment dam is 652m long and reaches a maximum height of 16.5m above the inlet channel invert. The embankment crest at an elevation of 158.55m provides 1.5m of freeboard above the designed maximum surcharge pool (el 156.9 m).
The foundation deposits beneath the dam embankment consist mainly of three types of materials. The overall geological profile is divided into two separate reaches by a buried bedrock ridge near the middle of the dam. To the right of the ridge (looking downstream) across the old river channel, the foundation materials consist mainly of well-graded sandy gravel with traces of silt overlying bedrock and numerous boulders. On the left side a deep pre-glacial valley is present below a low dam embankment and left abutment peninsula. The foundation materials consist of clean well-graded stratified gravelly sands and sandy gravels with numerous thin, highly pervious open work gravel strata overlying a layer of relatively uniform graded stratified sands, with silt and silty sands and occasional silt and gravel strata. The embankment of the main dam consists of homogeneous pervious fill with a downstream rockfill shell.
Although the dam was founded on a highly pervious foundation, no provision for an impervious clay cut-off through the foundation, nor an impervious core through the embankment, was made in the original design. The purpose of these would have been to prevent underseepage through the foundation of the dam due to the proposed limited flood control storage pools.
Over the past few decades, seepage and internal erosion have became recurring problems at Hodges Village dam, even during flood events of a frequency as low as 10 years. High exit gradients during flood events have caused sand boils, internal erosion and loss of foundation materials due to piping and damage to the access roads, following excessive seepage through the downstream slopes.
First evidence of this occurrence was observed during the 1968 flood event (a 20-year flood) when the reservoir pool reached an elevation of 148.8m (3.8m below the spillway crest). Seepage emerged from the slopes of the left abutment embankment causing considerable damage to the access road. The second flood event, exceeding a 10-year frequency, occurred in 1987 with the reservoir pool reaching an elevation of 150.5m (2.4m below the spillway crest). Seepage emerged in several areas at the toe of the dam and on the slopes of the left abutment. Several sand boils developed along a 180m area around the tailwater pond downstream of the left abutment.
The worst situation developed during the 1993 flood event; a 10-year event. The reservoir pool reached an elevation of 148.4m (4m below the spillway crest). A big depression measuring about 3.6m in diameter appeared on the surface of the access road located above the foundation drains. In only a few hours, this depression developed into a large sink hole about 15m long — threatening the integrity of the entire left abutment.
Emergency action was undertaken and steel sheet piles were driven into the area surrounding the sink hole to cut off the seepage water and to provide stability to the walls. This action prevented complete failure of the dam.
In view of the recurring problems developed dur-ing the mod-erate flood events, the Corps of Engineers was very concerned about the ability of the dam to impound safely the floods for which the project was designed. An extensive study evaluated the reliability of the dam and its existing condition.
A concrete cut-off wall was selected as the most practical and reliable solution to the problem, based on the specific conditions at the dam site. The cut-off wall was modelled at both spillway and surcharge pool conditions and at various wall depths to determine the most effective depth.
The retained design required the installation of approximately 23,080m2 of vertical concrete cut-off wall through the main dam and the adjacent dike 1 with a thickness of 800mm and a maximum depth of 45m. The cut-off wall is to be keyed a minimum of 1.5m into the underlying sound rock, which consists of fine grain quartz mica schist with strength of up to 127N/m2.
Cut-off wall construction
Considering the alluvial materials to be excavated, the possibility of encountering boulders and the hard rock excavation, bauer of America selected the following equipment:
•One Bauer BC30 trench cutter (35T) mounted on a Sennebogen BS6100 (125T) crawler crane.
•One Bauer BC30 trench cutter (45T) equipped with specially designed and fabricated roller bit cutter wheels to excavate boulders and bedrock which will also be mounted on a Sennebogen BS6100 crawler crane.
•A Bauer GB50 hydraulic clamshell will assist the two cutters for pre-excavation, chiselling and removal of boulders.
Each piece of excavation equipment is supplied with a built-in recording system for continuous, real-time monitoring of excavation parameters, including depth and verticality of both axes. Information data is transmitted to a monitor in the cab as well as computers in the project offices.
Due to the high possibility of slurry losses during excavation a large volume of bentonite slurry, about 2150m3, is stored in three tanks. The bentonite is then supplied to each cutter through a 20.3cm pipe, over a maximum distance of about 800m, using 150HP pumps. Also, due to the long distance, the return slurry line to the desanding plant is equipped with a booster pump.
Two Bauer BE500 desanders separate the excavated materials from the bentonite slurry. A total of 5.5km of bentonite slurry lines have been installed. To meet a contract requirement, Bauer designed and installed, in co-operation with US-based Geomation, a fully automated system to monitor bentonite slurry level in open panels during non-working hours.
In January 1998, tests were performed on a section located in the most challenging area of the main dam, and the selected equipment proved to be adequate in achieving the required excavation — meeting the contract quality requirements within the expected time frame. Core drilling performed on the middle of the concreted panel (Nx size) and the 10.16cm diameter core through the panel’s joint, confirmed the good quality of concrete and joint tightness.
Following the test results, review and the completion of equipment mobilisation, excavation of the main concrete cut-off wall was resumed in mid April 1998. As had been carried out during the test section, the excavation is currently performed following the primary and secondary panel sequence. The dam’s 16,500m2 cut-off wall consists of 137 panels; the cut-off wall on the adjacent dike 1 has an area of 6580m2 consisting of 83 panels. Most of the primary panels, excavated first, have a length of 7m. Two ‘bites’, 2.8m each (corresponding to the width of the equipment), completed one after the other for the full depth of the panel, are excavated first, followed by excavation of the middle wedge to complete the excavation of a primary panel.
Secondary panels, consisting of a single 2.8m long bite, are excavated with the cutter when the construction of the primary panels are completed. During the secondary panel excavation, an average 15.24cm of concrete is cut from the two adjacent primary panels leaving a high quality concrete construction joint for the full depth of the panel.
So far, about 25% of the cut-off wall has been completed and the work is expected to be finished in early 1999. All auxiliary work, including construction of access roads, removal of the top of the embankment dam, construction of working platforms and the guide wall, reinstallation of random fill and a drainage blanket, along with restoration works, have been subcontracted out. The reservoir will be kept in service throughout the contract period, and is expected to be finished by September 1999.
A cut above the rest
Michel Gavillet* explains how a cut-off wall will solve the problems of seepage and internal erosion being experienced by a flood dam in the US