Using an ROV to Repair a Leaking Surge Shaft

Excess seepage revealed during routine monitoring of the 1,020 MW Tala powerhouse and dam led to the use of a remotely operated vehicle to both find and repair the source of the leakage.

By Dave Malak and Kencho Dorji

While regular underwater inspection is usually part of the standard maintenance plan for aging hydroelectric facilities, it is less common for such a plan to be in place for newer assets. But conducting inspections on new dams after plant commissioning can help identify premature wear or leaks.

At the 1,020 MW Tala facility on the Wangchu River in Bhutan, Druk Green Power Corp. was monitoring, during routine maintenance activities, a drain that was installed to convey water seeping from the surge shaft. Construction of the 92 meter-high concrete dam was completed by Hindustan Construction Company in 2006 and the powerhouse began operating that same year. During normal operations, the dam would allow seepage within a defined range of flow rates. Through monitoring the drain, Druk Green Power noticed that the rate of flow was higher than intended. The company hired Hibbard Inshore of the state of Michigan in the USA to use a remotely operated vehicle (ROV) to inspect the surge shaft and try to determine the source of the excess leakage.

During that inspection, conducted in 2011, the ROV revealed four holes in the concrete slab of the surge shaft. From the initial video, the diameter of the holes appeared to be between 1.75 in and 2.5 in (3-6 cm). Through the cameras installed on the ROV, it was clear that particulates in the water were moving through three of these holes, indicating flow paths from inside the surge shaft to underneath the structure. This finding was of great concern for the future stability of the structure.

This image, taken using a camera mounted on a remotely operated vehicle, shows a hole in the floor of the surge shaft for the 1,020 MW Tala project in Bhutan, with observed flow.

Potential solutions

Shortly after these holes were discovered, Druk Green Power, Jaiprakash Industries (which built the surge shaft), and the Hibbard Inshore team (which included Dr. Ray Henn of Brierly Associates) discussed solutions to stop the flow permanently. The powerhouse at this dam contains of six 170 MW vertical Pelton turbine-generator sets supplied by Bharat Heavy Electrical Ltd. of India. Because the plant provides a significant source of power and revenue for the region, any solutions that were discussed needed to allow the plant to maintain its generation schedule. This meant that the solution would need to be implemented during daily eight-hour shutdown periods and generation would occur for the other 16 hours.

To reach the holes, equipment would be set up in an access tunnel connecting to the 180 m-high surge shaft, and there would be 100 m of water depth under normal operating conditions. Because flow through this area during operations is about 10 m/sec, all materials needed to be secured prior to the daily operating periods or removed from the surge shaft so that no risk would be introduced that could negatively affect the turbines.

Through discussion, it was determined that the costs and safety risks of a manned entry to make permanent repairs would be prohibitive. Using an ROV provided both safety and financial benefits over attempting to use a commercial dive team, drilling underneath or dewatering the shaft.

The first option for repair discussed was to pump grout from the surface while leaving the shaft flooded. In this scenario, an ROV would have prepared each hole for grouting by clearing debris. Once the hole was cleaned, a customized grout packer would be inserted and the ROV would be used to attach the grout hoses from the surface so the grout could be pumped into each hole.

However, the team decided that pumping grout from the surface would be very risky because the size of the void under the surge shaft slab was unknown. It was also unknown how much grout would be needed to sufficiently fill each hole and whether any of the holes communicated with the others under the shaft. Worries were that the grouting would need to be finished in an eight hour shift at each hole so that the grout could properly set. Also, pressure differentials would need to be closely monitored to avoid a potential disaster due to cracking or heaving.

Plugs were installed in three of the surge shaft holes and then filled with epoxy using a manipulator arm on a remotely operated vehicle.

The team decided that the engineering issues related to grouting were going to be complex. Understandably, the owners were looking for a lower-risk option that could be completed quickly.

Choosing a solution

Grouting was set aside in favor of providing plugs to fill each hole.

The team determined that the slab at the bottom of the surge shaft was M20 concrete. Knowing the type of concrete allowed Hibbard Inshore to determine the forces the slab would need to withstand during various plugging scenarios. The team poured a slab of M20 concrete with holes of similar diameter and separation distance to those at the Tala facility. Information on concrete strength along with testing of a prototype plug in the slab of concrete revealed that specialized plugs could be placed in each hole and inflated to complete a seal without damaging the concrete. It was not known whether the real holes in the surge shaft were lined with pipe and therefore presented a smooth sealing surface or were rough concrete with varying dimensions along the hole depth that would complicate sealing. Because of this, the team tested to ensure that the plugs would work in either scenario.

Each plug was made from stainless steel and coated in vulcanized rubber to provide a sealing surface. The plugs were designed to create a seal while not exerting forces that would exceed the concrete compressive or tensile strengths. The design considered that the plugs would remain in place during the high flow conditions common with normal operations.

Druk Green Power and Jaiprakash Industries decided this was the best method for repair, and the Hibbard Inshore team was hired to design the plugs with Baski Inc. of Colorado, USA, for insertion using an ROV. Under the direction of Hibbard Inshore, the packers were designed, manufactured and tested in the Baski facility, then shipped to Bhutan for installation in April 2012.

Solution implementation

During the initial inspection, the ROV revealed small timber debris lodged in one of the four holes in the surge shaft. To remove the debris, the team brought an ROV-operated saw and drill to the site. During the first day of the 2012 operation, Hibbard Inshore deployed the ROV to examine the holes to determine if there had been any change from the inspection in 2011. The inspection revealed that the timber debris observed in one of the holes in 2011 had been dislodged over the following year. It was also found that since the initial inspection, the flow rates through the holes had increased. The final flow measurement (taken immediately before the plugs were installed) of the leakage from the surge shaft was 1,718 L/min.

Once deployed to the bottom of the surge shaft, the ROV measured each of the four holes, confirming they had not changed in size or nature since the previous inspection. Three holes were determined to have flow, while one was found to be without flow.

After taking the measurements of the holes so their diameters were known, a packer size of 1.75 in diameter was selected, and the ROV prepared each hole to receive a plug. Preparation included inserting a measuring tool into each hole to determine if there was blockage. The plugs had a minimum insertion length that was required to make a proper seal. Differential pressure tests were performed to evaluate communication between the holes and to learn more about the nature of the leaks.

Once the holes were inspected and measured, two packers were inserted using the ROV. The plugs were inflated to seal from the surface while monitored by the ROV. They were epoxied into place and the hoses and interface piece were removed to minimize any material protruding from each hole. Only enough epoxy was inserted into each packer to aid the seal and grouting under the slab was not performed due to the previous concerns. During this operation, a fourth smaller hole was filled with epoxy.

Plugging of the third of the three larger holes occurred in February 2013 once the water cleared. The hole for the third plug had debris present and needed to be cleaned using an ROV-operated drill. During operations to place the third plug, several additional holes were identified in other areas that had not been previously inspected. These holes were not flowing significantly, and flow could not be seen visually. The Hibbard Inshore crew used the ROV to fill six of these additional holes with epoxy. This process proved that the additional holes were not a source of leakage.

How it has worked

Positive results came from the insertion of the initial two plugs. Flow rate through the drain was reduced by about 60%, confirming that these holes were the primary source of leakage and the plug insertions worked properly. After the third plug was inserted and the six additional holes were filled with epoxy in 2013, leakage decreased to 569 L/min, which was determined to be within acceptable levels. The team accomplished significant repairs through the design and implementation of a unique repair method that reduced risks to the structure while maintaining safety without interruption to the regular operating schedule.

Dave Malak is vice president of operations with Hibbard Inshore LLC. Kencho Dorji is chief engineer of the 1,020 MW Tala Hydropower Project for Druk Green Power Corp.

More HRW Current Issue Articles
More HRW Archives Issue Articles

Previous articleBC Hydro seeks engineering of dams, spillways, penstocks
Next articleIncreased study requirements, loss of DOE backing end Admiralty Inlet tidal pilot project

No posts to display