There are two separate contracts for the underground construction on WTP4 – the raw water system and the Jollyville Transmission Main. The raw water system mainly comprises an intake on Lake Travis, a raw water tunnel, a pump station and the treatment plant.
For WTP4, the client, the Austin Water Utility, is working with the City of Austin, which is managing the design and construction. Aecom, with subcontractor Brierley Associates, is the designer for the raw water system. The contractors on the project include Manson Construction, for the raw water intake, and Obayashi USA, for the raw water access shaft, raw water tunnel and raw water transmission main tunnel.
"The City of Austin owned a piece of property in Lake Travis, so the place where we could withdraw water was restricted to property that they owned. We also needed vertical access to the water level," says Shelby Eckols, project manager with Aecom.
"The lake tap is now complete and the raw water tunnel is nearly complete. We’ve also completed the access shaft and the transmission main tunnel, and they will be _ ooding the raw water tunnel and the access shaft imminently."
Encountered Geology
Excavations were through three expected different formations: the Edwards, a limestone formation with solution features; Walnut, which is composed of hard limestone and nodular marls; and Glen Rose, which consists of hard and softer limestone. Most of the tunnel excavations were expected to be through the GlenRose.
Above ground the area is mostly green fields, and no issues were expected with subsidence. However, the tunnels go under critical environmental features that can’t be disturbed. The entire project, but particularly pertinent to the raw water system, was smack in the middle of an environmentally sensitive area.
"Environmental preserves were surrounding the pump station and we could not gain access," says Eckols. "Normally we would have soil borings along the tunnel route to the lake and from the raw water pump station to the WTP4 site. Both of our tunnels are roughly one-mile-long [1.6km] and we could not get access to those soil borings. "For the raw water tunnel, we only did about two soil borings at the raw water pump station site, two along the route to the lake and two on the way to the water treatment plant site."
With limited soil borings to identify the rock, other borings in the general area were investigated. Through investigations it was found that the Walnut and Glen Rose formations were generally consistent, Eckols explains: "Based on that information we prepared geotechnical base line reports for both the raw water intake, raw water tunnel and raw water transmission main.
"With that investigation, the geology that we encountered was generally consistent with what we anticipated." However, Eckols notes that on the raw water transmission main – the tunnel that begins at a tunnel portal on the raw water pump station site – the excavation began in the Glen Rose formation, transitioned into the Walnut formation, then rose vertically through the Edwards formation.
"We faced a challenge with the ground," says Eckols. "The chosen route minimised excavation in the Edwards formation and the potential for encountering karst formations. In the tunnel section that transitioned from the Glen Rose to the Walnut formation, there was a short section that required a change in rock bolts. The change was from an epoxy anchored rock bolt to an expansion type rock bolt. Otherwise, excavation was routine.
"After excavation, a 7ft [2.1m] diameter steel pipe, with welded joints, was installed in the tunnel. The steel pipe was cement lined in place and the tunnel annulus was grouted."
Eckols explains that while the rock is actually very good in the area, the rock was not self-supporting. The design called for tunnel excavation to be supported at a minimum with two rock bolts at the crown of the tunnel – 5ft (1.5m) centres. In areas of weak rock, an additional two rock bolts, wire mesh and shotcrete was required.
"That’s the way the tunnel was designed and that’s the way we wanted it built. The reality is that it may have supported itself in many areas but we required them to install those rock bolts," Eckols says.
Excavation
The contractor used one Antraquip AMQ 100 using pattern rock bolts for support for excavation. A TBM was considered, but in the end a roadheader was the preferred method. TBMs, which dig faster than road headers, are an attractive choice but Eckols says other considerations, such as dismantling times would have been time consuming. "The project was designed so the route and the turning radius of the TBM would fit within the easement, but it was a tight fit," notes Eckols. "The use of the roadheader resulted in flexibility of excavation to stay within the underground easements and resulted in a tunnel that met the project requirements."
Two major challenges to the project are making the connection from the raw water tunnel to the intake, and the 450ft (137m) deep access shaft to the raw water tunnel. The 25.5ft (7.8m) diameter access shaft is located at the pump station site and was excavated by roadheader; construction of the shaft was completed roughly a year ago.
Originally, the shaft’s design called for a shotcrete lining, but at MWH and Obayashi’s request the design was adjusted to accommodate a concrete lining. "We designed the access shaft to have excavation supported by rock bolts, welded wire mesh and shotcrete for the shaft excavation," explains Eckols.
"The contractor proposed to use a cast-in-place concrete for the initial and the permanent support for the access shaft. We realised the design to accommodate the contractors preferred method and the shaft was excavated with roadheader and the concrete liner was installed as the excavation progressed. After the work completed, the contractor had to go back and perform some patching up; finish-up on the joints in the liner and drilled weep holes in the liner, as originally required."
Obayashi chose to excavate both tunnels by roadheader. The raw water tunnel excavation was some 12ft (3.6m) diameter in a horseshoe-shape. The tunnel had a very slight slope and the route in plan view was serpentine in shape for a length of 4,306ft (1,312m) to stay within the acquired underground easements. When the contractor completed excavation it was on line and grade, and only a couple of inches off the centreline of the lake tap pipe.
"One of my concerns was finding that pipe buried in the lake bottom, but the contractor came within a couple of inches of being right dead on the centre – phenomenal control on part of the contractor. The tunnel was lined with a steel pipe at the lake tap, then transitioned to an unreinforced concrete liner for the remainder of the tunnel. The finished diameter was 9ft [2.7m]."
The raw water tunnel concrete lining was a cast-in-place concrete liner. The liner was placed from the lake tap back to the access shaft. "The challenge was getting the concrete down the access shaft and then a mile up the tunnel to begin the concrete placement; that was a something that the contractor overcame," notes Eckols.
In addition, the specifications required contact grouting be performed after placement of the concrete liner. The contract stated that the grouting be done 30 days after the concrete was placed. "We debated not lining the tunnel at all," adds Eckols. "This was because the rock was so good, but we couldn’t get the hydraulic requirements for the conveyance of water."
Finishing the job
While Obayashi USA excavated the access shaft, Manson Construction, responsible for marine work, was constructing raw water intake in the lake. The raw water intake includes a lake tap – 9ft (2.7m) diameter pipe – that is embedded about 85ft (25.9m) into the lake bottom. This 9ft (2.7m) diameter pipe was installed by excavating a 12.5ft (3.8m) diameter hole into the lake bottom, installing the 9ft (2.7m) diameter pipe and grouting the pipe in place. The pipe has an internal blind flange about 40ft (12.1m) from the bottom and an external blind flange on top of the pipe at the bottom of the lake.
The raw water intake construction was completed months before the raw water tunnel. Eckols describes how the connection from the raw water tunnel to the intake tunnel was managed: "It was intended that the intake in the lake be completed well before the tunnel got there, so we had no work in the lake that could endanger people working in the tunnel.
"Obayashi USA excavated the tunnel and found the blind flanges right on target, this is to stop the water rushing in. Obayashi then completed all the work on the tunnel; they finished doing the concrete liner on the tunnel all the way back to the access shaft and all the work on the access shaft. After completion of the tunnel work, the lower blind flange of the lake tap was checked to confirm the upper blind flange was not leaking, then the lower blind flange was removed. Some minor work was then required within the lake tap pipe and that work is being performed presently. Upon completion of this work, a valve on the upper blind flange will be opened and the tunnel will be flooded. After the tunnel is flooded, the contractor will then remove the upper blind flange by using divers and cutting devices. That will complete the raw water intake and tunnel"
