Treatment of high pressure seepage27 October 2004
Murray Gillon, Peter Amos and Tom Newson report on a grouting operation to fill the high pressure seepage path under the foundations of New Zealand's Arapuni dam
In September 2000, pressures being monitored in a geological fracture beneath New Zealand’s Arapuni dam were found to be rising significantly, indicating that a deteriorating condition was developing in the foundation. Two boreholes drilled in 1995 had intersected high water pressures within the fracture in an area close to the downstream face of the dam. The increasing pressures had the potential for major seepage to develop downstream of the dam, where the fracture was exposed. Investigations since September 2000 confirmed the extent of high pressure in the dam foundation and the nature of the deterioration, providing a baseline to assess the feasibility and performance of treatment options. The foundation fracture bearing the high water pressure was successfully grouted in December 2001 without lowering the reservoir or damaging the dam’s porous concrete underdrain network. The key to the success of the operation involved three important elements in the approach to grouting; controlling foundation pressures within safe limits during grouting, designing a grout mix suitable for application in high flowing seepage water and implementing measures to prevent grout entering and blocking the dam underdrain network.
Arapuni dam, located on the Waikato river 55km upstream from Hamilton, is a 64m high curved concrete gravity dam completed in 1927. The dam across the Waikato river bed forms the reservoir for a 186MW hydroelectric power station, sited 1km downstream at the end of a headrace channel that follows the left abutment. The Arapuni dam and power station are owned and operated by Mighty River Power Ltd.
The dam is founded on a 40m thick sheet of Ongatiti Ignmibrite, a point-welded tuff. A feature of this ignimbrite sheet is the lack of regular orthogonal vertical jointing often seen in ignimbrites. Three major subvertical cracks or fractures were identified during dam construction crossing diagonally across the dam footprint in an east-west orientation as shown in Figure 2. These fractures extend for the full depth of Ongatiti. Beneath Ongatiti, about 40m below the base of the concrete dam, are older ignimbrite deposits. The ‘no-fines concrete’ porous drain network (Figure 2) is the main uplift control at the dam/foundation interface.
Recent seepage history
In 1995, eight holes were drilled from the downstream toe into the dam foundation to investigate foundation uplift pressures. Two of these holes, known as OP05 and OP06, intersected high water pressures and each flowed at several hundred litres per minute after drilling. Closing the holes to measure pressure indicated water pressure around 13m below lake level. The high pressure flow was encountered at depths down the holes that coincided with one of the sub-vertical fractures in the dam foundation recorded during dam construction. Between 9 and 12 days after the drilling, the underdrain flows increased from 15litres/min up to 200litres/min.
During an annual dam performance evaluation in 2000 it was noted that pressures and flows from the fracture were increasing. An extensive investigation programme was initiated to locate and characterise the flow path in the fracture system as a pre-requisite for remediation. By September 2000, solid material (clay and bitumen particles up to 20mm in diameter) was observed to be exiting from the relief drain. Drill hole OP06 was operated as a relief well to control the water pressure in the fracture so as to minimise the potential for seepage to propagate to the downstream toe of the dam.
While an immediate action involving grouting through the two holes OP05 and OP06 could have been carried out in September 2000, this was not considered the best long-term solution. Clearly, an eroded path through fractures in the dam foundation had developed, but little was known about the seepage path, including the size of the void, the nature of the eroding material, the source of seepage water or the path of the leak. The excess pressures above reservoir pressures necessary to effect grouting were also considered to have the potential to initiate blowout of the remaining fracture infill between OP06 and the exposed fracture downstream of the dam with the possibility of surface leakage. Furthermore, the high pressure water was entering the dam’s underdrain in at least one location, indicating that grouting OP05 and OP06 could lead to grout entering the underdrain and blocking it.
On this basis the decision was made to investigate the high pressure area and determine the seepage path prior to developing the grout treatment method. The investigations would be undertaken with the full reservoir in place, but in the knowledge that a deteriorating seepage condition existed in the dam foundation. The option to grout OP05 and OP06 was retained during investigations as a contingency should seepage conditions become unacceptable.
A wide range of investigations (Gillon and Bruce, 2002) showed that:
• The high pressures were confined to the fracture system.
• The flow path within the clay infilled fracture varied in width up to 80mm wide.
• There were strong hydraulic connections to the lake and to the diversion tunnel located in the right abutment.
By October 2001 sufficient information had been gathered to identify the seepage path and to confidently plan a grouting operation. Grouting was programmed for early December 2001. The primary objective of the grouting operation was to fill the high pressure seepage path with a secondary objective of preserving the existing foundation under-drain network. It was acknowledged that erodible joint infill materials would still remain in the fractures.
The design for the grouting had three key elements; control of pressures and flow areas in the foundation and seepage path by using relief wells, designing a grout mix suitable for application in high flowing seepage water and preserving the dam’s porous concrete underdrain network
A total of 10 drill holes (Figure 4) were located in the seepage flow path prior to the grouting operation. Eight of these were available for use as relief wells. The highest pressure recorded in the seepage path was used as the grouting pressure on the basis that this pressure had not caused seepage to emerge at the toe of the dam. By using the most upstream wells for pressure relief, low velocity and low pressure conditions were created in the downstream part of the flow path. This was conducive to grouting in this critical area for the protection of the toe of the dam. The low flow velocity minimised the problem of washout of the grout. The low water pressure in the fracture enabled a grouting pressure 2bar above this pressure without exceeding the maximum limits for the grouting operation in the toe area.
The grout mixes to be used were cement-based grouts, with varying water/cement ratios, designed to be placed in either static or flowing water conditions. The mixes incorporated bentonite to reduce bleed, an anti-washout agent and a superplasticiser. The proposed mix designs developed by the designers were tried and tested in a mobile laboratory set up next to the mixing station. The mixes were modified as indicated by the tests to ensure they would be stable, durable and possessed the appropriate rheological and hydration properties.
Grout mixes had water/cement ratios in the range of 0.8:1 to 1:1 by weight and three mixes were sand/grout combinations to be used if a runaway condition developed during grouting. During grouting only two water/cement mixes were used, and no sand mixes were required.
Protection of the dam under-drain system was achieved by flushing of the drains and using wood chips to seal the underdrain connection with the high pressure seepage. The wood chips were injected into the flow path through a relief well and the flow to the drain drew the wood chips to the underdrain connection sealing it off. Monitoring during grouting showed that grout did not enter the underdrain system.
Before grouting commenced, the seepage properties of the foundation were confirmed by testing so that a baseline could be determined, suitable for control of pressures during grouting and to determine if grouting had led to improvement in the seepage condition.
Conventional grouting equipment was used. Important items of equipment included a high shear (or colloidal) grout mixer, a Mono (helical screw type) grout pump and an electronic flowmeter grout control panel. Each drill hole had been fitted with branched pipework that allowed for grout injection line, pressure gauge and bleed pipe to test for grout arrival at the hole.
Grouting was planned as a 24 hour operation and commenced on 12 December 2001. The grout and QA teams worked on a shift basis. Grout teams were dedicated to the mixing station at the dam crest, the grout injection station at the toe of the dam or the flushing water supply station.
Three hours before commencement of grouting, the most upstream pressure relief well was opened to depressurise the foundation by 20m head. Two hours before grouting, the underdrain flushing operation was started.
Grout injection commenced from the most downstream hole and was intended to work upstream towards the upstream pressure relief holes. Grout flowed upstream at a fast rate. The order of grout appearance at other holes indicated that grout first filled lower flow paths and then built up to appear at higher elevation holes.
Grouting took 12.5hr; a total of 11.5m3 of grout was placed. Of this total volume, 4.4m3 was placed in the upstream holes, leading to the conclusion that grout placement had extended upstream of the dam.
Engineering design and detailed specification for grouting the feature was carried out by DamWatch Services, together with its specialist consultants Geosystems LP (USA). DamWatch was also responsible for on-site management of dam safety throughout the grouting preparations and during the grouting operation including the responsibility for operating the relief wells to control pressures under the dam.
Dam safety during grouting
Once it began, grouting was a continuous round-the-clock operation. To ensure continuity throughout the grouting operation, dam safety teams worked in shifts with shift changes timed to avoid shift changes of grouting team.
Dam safety teams monitored key indicator instruments. A data logger recorded readings from transducer instruments every 5min. Recorded maximum precedent foundation pressures formed the upper limit for grouting pressures.
The dam safety team leader had the authority to suspend grouting at any stage if the grouting operation compromised dam safety.
Following completion of the grouting, foundation pressures have been closely monitored to validate the continued performance of the grouting work and to ensure that the former high pressure seepage does not develop an alternate path.
Verification of grouting effectiveness
As a direct result of the grouting, seepage flows from the dam foundation decreased from their former 800litres/min to 20litres/min. Foundation uplift pressures dropped by up to 17m immediately beneath the dam. There were no indications of transfer of pressures to adjacent parallel fissure systems following grouting.
Four boreholes (PR10-13) were drilled from the dam toe into the grouted fissure immediately after grouting. Verification drilling did not detect significant flow or pressure in the fissure. One of the verification holes returned a core that contained a well grouted fracture with no signs of voids remaining. The other holes returned core with intact infill material present, but no grout. Coupled with the observation that the total grout quantity injected was approximately one third the void volume assuming the fracture infill was totally removed, it appears that the void beneath the dam was a network of interconnecting discrete flow paths through the fissure infill that extended back to the reservoir. Based on grouting and verification data, a view of the extent of grout in the fissure is shown in Figure 4.
The underdrain capacity was tested and confirmed that grouting had not affected drain capacity.
The high pressures and seepage flows in a geological fracture under the foundations of Arapuni dam were successfully grouted.
Careful planning and the application of precautionary measures during grouting, like pressure relief and sealing off the underdrain, allowed a comprehensive and well-executed grouting operation to be run, without compromising dam safety. A large void allowing high pressure, high flow seepage from the reservoir to be present under the dam has been successfully filled using modern grouting practices, while the reservoir remained fully operational. The work was completed without damaging the dam underdrains.
Murray Gillon and Peter Amos are with DamWatch Services Ltd, Wellington, New Zealand. Tom Newson is with Mighty River Power Ltd, Hamilton, New Zealand.
For more information contact: Murray.Gillon@DamWatch.co.nz.
Many people from a variety of organisations worked together to successfully implement this unique grouting operation. The authors would like to acknowledge the contributions of all these people. The authors also acknowledge the permission of Mighty River Power and the New Zealand Society of Large Dams to utilise information presented in a previous paper by Messrs Amos, Newson, Gillon and Stewart.