Maretta Tubb, Managing Editor
Under a contract designed and managed by Brown & Root Inc., Houston, Ceanic Corporation recently carried out underwater repair work at two inverted siphon locations in Baytown for the San Jacinto River Authority (SJRA). At one site, the divers relied on underwater sliplining to repair two 42-inch diameter siphon pipes, while at the second site specially constructed cofferdams provided a solution to allow repairs to be made in the dry to a 60-diameter siphon pipe.
The involvement of the SJRA in this and other raw water transmission projects dates back to 1937, when the agency was formed by special acts of the Texas State Legislature.
Today, the SJRA is one of 15 major fiver authorities in the State of Texas. With general offices in Conroe TX, the authority supplies water and collects and treats wastewater in The Woodlands, manages Lake Conroe and the Highlands Canal System. The agency's boundaries include the entire watershed of the San Jacinto River and its tributaries excluding Harris County. Specifically, all of Montgomery County and parts of Walker, Waller, San Jacinto, Grimes and Liberty Counties are within the boundaries of the agency. By virtue of its statutory order, the SJRA operates in East Harris County.
The Highlands Canal System, where work was recently completed at two separate inverted siphon locations, is made up of approximately 900 acres of land, 38 miles of canals, a 500 acre reservoir and a 100 million gallon per day (mgd) pump station on Lake Houston. The raw water transmission system, which supplies the water needs of industrial clients and agricultural irrigation customers in southeast Harris County, includes some 40 inverted siphons that carry the water under roads, highways, railways, bayous, streams and drainage ways.
When the Highlands Canal System was constructed in the mid 1940s, pipe joint sealing materials such as jute and oakum offered the best sealing solutions at the time. Over the years, this early sealing agent is thought to have dissolved, causing some leakage problems. Problems with leakage at two separate inverted siphon locations, Nos. 29 and 39, became of concern to the SJRA in 1996.
According to SJRA, at the No. 29 siphon, water was observed bubbling up in the ditches on both sides of the road. Water also regularly seeped up through the roadway, a condition credited with having contributed to pavement failures on several previous occasions.
Further downstream of the Highlands Canal, the ground condition at an industrial railroad spur above the No. 39 siphon had become a problem. Even if it hadn't rained in weeks, the ground remained soft and unstable. As a result, excessive rail movement was being experienced when industrial railcars traveled the tracks.
At that point, Brown & Root Inc.'s Water Resource/Environmental Division was retained by the SJRA as the consulting engineer on the project to evaluate the problems and come up with a solution.
Michael Spero, P.E., engineering manager of Brown & Root's Water & Wastewater Rehabilitation, indicated that shortly after the contract award, the sites were carefully evaluated. At both locations the large diameter siphon pipes were found to be situated in about 12 to 17 feet of water and located between concrete inlet and outlet structures.
The problem faced by Brown & Root was how to access the siphons. Spero said, 'It was obvious from the outset that we would have to rely on divers to conduct the inspections and carry out any needed repairs. In addition, since the two siphons delivered approximately 60 mgd of raw water to area industrial users, by-pass systems would be needed at both sites to handle deliveries while the siphons were being inspected and repaired.'
Spero recalled that when the project was put out for bid, most area utility contractors had little if any experience with underwater construction. Fortunately, Houston-based Ceanic Corporation (formerly American Oilfield Divers Inc.), that had gained worldwide experience in underwater construction, was low bidder on the project.
The siphon at the No. 29 location runs under Ellis School road. It has two 42-inch lines, in parallel, that were installed in 1945, and a third 60-inch line slightly off to the side built in 1979.
To better evaluate the problems, divers from S&J Diving, Houston, made an initial inspection. Spero pointed out that plans called for the underwater inspection to be carried out by one diver, with a tender on the surface. A second, back-up diver and tender were on the surface as a relief crew and for safety.
The water coming out of Lake Houston is somewhat turbid. Any underwater inspection and repair work conducted at the two sites would have to be done primarily by feel, as the divers would be working in basically zero visibility.
According to Spero, during the initial inspection dive at the No. 29 siphon, the divers were unable to detect any structural problems with the three pipes making up the siphon, such as offset joints, excessive gaps, etc. 'However, as the divers passed through the two 42-inch concrete lines during the inspection, the air bubbles produced by the divers were observed coming up in the ditches on either side of the road. 'This indicated the pipe joints of the two 42-inch lines were not properly sealed,' he said.
Following the inspection, several methods to repair the siphon were considered. The solution ultimately devised was to slipline the two 42-inch lines with a 36-inch Hobas Low Profile fiberglass pipe with rubber gasketed joints. However, since the size of the inlet and outlet structures limited the length of pipe that could be used, instead of standard 20 foot joints, five foot joints of pipe would be required for the sliplining operation.
Also in conjunction with the sliplining, diving repair crews were pressed into service to divert the flow of water entering the siphon while the underwater sliplining was taking place. For this, the SJRA supplied two steel slide gates that were modified by Ceanic and then set in existing stop log grooves in the inlet structures.
Spero is quick to point out that the slide gates did not allow dewatering of the siphon, only diverted the water flow to the other two siphon pipes and prevented the current from passing through the siphon.
The equipment spread mobilized to the site to conduct the sliplining basically included a traditional diving crew made up of a diver and tender, plus a backup diver and tender, a 10 ton crane to facilitate pipe handling and an airtugger or winch.
As noted by Spero, the winch was an essential component of the equipment spread. It was rigged up on the upstream headwall of the siphon and had a 3/4-inch diameter wireline leading to a snatchblock at the bottom of the upstream slide gate.
During the sliplining operation, the diver's work would consist of aligning each joint of Hobas pipe. As each joint of pipe was lowered into the water, the diver would 'swim' the pipe inside the existing 42-inch line, check the alignment and then signal for the operator of the winch to begin pulling the pipe in tight to make the final joint connection. The diver would follow the pipe in, swim inside and confirm that the joint was 'homed' properly. As the diver exited the pipe he would bring the line with him, hook up the next joint of pipe and start the procedure again. All this work was done strictly by feel.
The divers wore wet suits, and a dive helmet. They were connected to the surface with an air hose, safety line and communication cable and a hose that pumped heated water (100 degrees F) into their wet suits. This was to protect them against the 50 degree water coming out of Lake Houston during the dives which were expected to last two to four hours.
In order for the underwater sliplining to work, it would have to be closely coordinated both above and below water.
Ceanic Project Manager Russ Fogel, charged with overseeing both land based and diving operations, said that prior to the sliplining commencing, land and undersea communications were carefully coordinated to ensure that the divers and land crew were well versed in both verbal and hand signals that would be heavily relied on during this phase. 'We had a superintendent at a communication console that was in constant contact with the diver in the water. There was also a worker stationed at the siphon outlet structure that used hand signals to direct crane operations and there was constant radio communication with the winch operator positioned the other end of the siphon,' he said.
Continuing, Fogel indicated the underwater sliplining was a text book case. 'We sliplined 14 Hobas joints in each respective siphon. It took about two and a half days to slipline 67-feet of pipe in the first 42-inch line, and less than a day to complete the second,' he said.
After the final section of pipe was installed, PVC grout and vent tubes and a grout filled cloth sock bulkhead were installed and used to fill the annular space between the two pipes with cellular cement grout.
As noted by Spero, it was expected that about nine cubic yards of grout would be required to fill the annual space at each respective location. However, 18 cubic yards of grout was ultimately used. 'The voids above the pipe were filled, solving the structural problems of the pavement above. In fact, these lines should now have a very long life.'
The second part of the project, at the railroad spur location, was a similar type set up. Siphon No. 39 also has three RCPs in parallel. Two are 42-inch diameter lines, measuring 130 feet in length constructed in 1945, while the third is a 60-inch diameter line, 115 feet long, installed at the site in 1979.
Located in Baytown, the siphon is situated several miles downstream of Siphon 29.
Spero noted that there was no way to completely block off or divert the flow to conduct the inspection of the lines. Some 60 mgd of water pass through the three siphon pipes at this location and the water velocity inside the 60-inch diameter line was running about four feet per second. 'When divers tried to enter the 42-inch lines to conduct the inspections, debris, coupled with the high water velocity, was such that there was concern the diver's air hoses would become fouled, putting the divers at risk.'
Based on the above-ground inspection, there was no indication the 42-inch lines were leaking. There was, however, significant evidence of leakage from the 60-inch line.
While concern for divers prompted SJRA to decide to rely on water jets and grappling hooks to clean the 42-inch lines, it was felt an interior inspection of the 60-inch line was required to fully assess the problems.
According to Spero, inspection of the 60-inch line was accomplished by lowering a diver carefully into the pipe from the upstream side. A tether was also used to provide the diver some stability against the strong current. Although bouncing around during the inspection, the diver was able to determine that each 7 1/2 foot joint of the 60-inch diameter pipe had a lifting hole at the top which had either never been plugged properly or the plugs had worked loose. Holes found at the top of all 16 pipe joints that make up the line were allowing water to flow into the soil behind the pipe, causing unstable ground conditions under the railroad spur.
Complicating things further was the pipe configuration. This particular line had two 30 degree vertical bends at each end extending up toward the surface.
In looking at viable solutions to repair the large diameter line, several methods were evaluated. Investigations into underwater sliplining revealed that, even with short sections, 48-inch O.D. would be the largest pipe diameter that could be used with the 30 degree bends in the host pipe. 'This created too much hydraulic loss and was found unacceptable,' Spero said.
Also evaluated was the use of a cured-in-place pipe (CIPP). This method, which seemed to offer a viable solution had never been attempted on a 60-inch underwater line. For this reason, Brown & Root felt the method would prove too costly and push the envelope of the technology too far, and decided against recommending it as well.
'What we ended up doing was constructing two cofferdams to shut off the flow of water in the 60-inch line to allow dry repair of the line,' Spero said.
Unlike Siphon 29, the No. 39 location is situated near industrial plants that relied heavily on the siphon for raw water to conduct daily operations and the railroad spur to both receive and ship products. For this reason, one of the first concerns was a bypass system to handle raw water deliveries.
Ceanic's Fogel said the by-pass system was specified to handle 28 mgd of raw water and was made up of five 12-inch pumps. 'In this way, we could be assured raw water supplies to area industrial users would not be interrupted.'
There was, however, no way to allow the railroad spur to be used once the water had been diverted. The hoses diverting the raw water would need to be placed on mud mats atop the tracks. Once the water was diverted, all incoming and outgoing rail shipments to the plant had to be halted until the work was completed.
Prior to the project start, plant officials had agreed to allow the railroad spur to be out of service for 14 days. Since time was of the essence, as soon as the pumps were turned on and the water diverted, divers prepared to set the cofferdams.
The cofferdams, constructed off-site by Ceanic workers, consist of 54-inch diameter corrugated metal pipe. At one end, a 30 degree bend was fabricated to accommodate the angle of the concrete pipe. Also, a specially constructed expansion clamp with a rubber gasket was devised that could be inserted into the pipe and expanded out to form a seal from the inside.
Fogel noted that like most construction projects there are always conditions you can't find until you actually do the work. The cofferdam installation was no exception. At both sides of the siphon, apparently when the headwalls were constructed either form leakage or a form blowout had occurred. As a result, concrete was in the pipe itself, which had to be chipped away with an air chisel before a secure seal of the cofferdams could be achieved.
'Once we got a good seal, the pumps were turned on to dewater the line so the work could be conducted in the dry,' he said.
Although some seal leakage around the cofferdams occurred, it amounted only to several inches of water in the bottom of the pipe, or a few gallons per minute, which was easily handled with a small submersible pump.
As quickly as the cofferdams were in place, lights and a ladder were placed inside the pipe to facilitate handling of the repairs. A short time later work commenced to clean the pipe and seal the lifting holes with Master Builders EMACO cement mortar and Splashzone, an epoxy repair mortar.
Although plans initially called for the all the pipe joints in the line to be air pressure tested, the concrete pipe proved to be too porous. Instead, using hand trowels, workers resealed all the pipe joints with the epoxy grout.
Once the grout was sufficiently dry the siphon was flooded and divers removed the cofferdams. The diversion pumping was then stopped.
Spero said, 'The plan worked to perfection. Crews had 14 days to complete this phase but actually completed it in only 13 days.'
It is anticipated that at some future date, if leakage problems develop, the 42-inch siphon pipes will have to be sliplined. Since it isn't creating a problem at this time, SJRA's maintenance staff plans to continue closely monitoring conditions at the site.
Not surprisingly, Spero indicated he was happy with the $184,000 project's success. 'It was completed on time, with only a few minor change orders for unexpected conditions that were discovered and without any major problems.'
|Printer friendly Cite/link Email Feedback|
|Comment:||Carried out repair work at two inverted siphon locations in Baytown for San Jacinto River Authority|
|Date:||Jul 1, 1998|
|Previous Article:||Used HDD Market Emerging|
|Next Article:||heavy load|