Marking the centennial anniversary of their famous trek, the Lewis and Clark Exposition in 1905 attracted people from all over the world to Portland. Visitors crossed wooden bridges to an island in scenic Guild’s Lake to attend.

Old, black and white images show still waters surrounding an elaborate entrance building, which would be taken down after the show. The lake would be filled in, in only 15 years’ time—with anything and everything, it seems.

As the City of Portland wraps up a 20-year program to control combined sewer overflows (CSOs), it’s exploring the geological history of the city’s northwest side while constructing the Balch Consolidation Conduit.

This 8,500ft (2,591m) pipeline will carry combined sewage and storm water runoff from the Balch Drainage Basin via the existing West Side Big Pipe to the Columbia Boulevard Wastewater Treatment Plant. With the completion of the tunnel and the program—they both have a December 1 deadline—CSO volume emptied into the Willamette River, which passes through the city, will be reduced by more than 94 per cent.

Design work began in 2007, when the city put out a consultant services RFP in January, which was awarded to Kennedy/Jenks Consultants, in association with Staheli Trenchless Consultants and Shannon & Wilson.

Guild’s Lake’s demise
Ground conditions along the Balch alignment are less than ideal, partly because of the former lake. “This pretty good sized lake was filled in with manmade materials and our project had to navigate through all these different features associated with the lake,” says Jerry Jacksha, lead geotechnical design engineer, who is a senior associate with Shannon & Wilson. Of all the projects the City of Portland has undertaken in its CSO program, he says, “It’s fair to say we probably have a wider range of subsurface conditions than any of those.”

This varies from very soft lake sediments, which are unique to the whole west side of Portland; red sand materials that were man-placed; wood trestles from railroad beams that ran through the lake and sluiced gravel from the nearby West Hills.

“In order to provide a cheap way of getting fill into this lake area from the West Hills, they used high-pressured water and they turned this native material into a kind of fluid, sluice material that flowed down into the lower areas,” Jacksha explains.

This sluice material made its way to the lake shortly after the 1905 exposition finished when the area began to see new development. Scott Clement, project manager, with the City of Portland’s Bureau of Environmental Services, explains: “The Portland Regional Dock Authority was also doing navigation work on the Willamette River where they would dredge and then dispose into Guild’s Lake. The City of Portland also had an incinerator on site where local trash was burned, and then deposited as landfill, which was also going in the lake.”

As different organizations used Guild’s Lake as a catch-all, it was filled in and was developed over during the 1920s-40s. But that’s not the only challenge for the Balch Consolidation Conduit. In terms of the native materials Alluvial deposits are predominant, Jacksha says. “These were deposited from up to 40, estimated, flood events. They’re called Missoula floods—a lot came from that part of the country [the city in western Montana] when the big ICE dams would break and send huge amounts of water down the Columbia River Gorge into Portland, and then they would move down toward the ocean.”

There are also gravel, cobbles and boulders along the alignment. And there is a material unique to Portland, a cemented gravel called the Troutdale formation. “And that was something we really had to consider with this micro TBM—hitting this cemented gravel and how it would respond. It could be rock material and in other places it could be partially cemented,” Jacksha explains. “And of course all of this area is below the water table, and our shafts had to deal with ground water and other things related to ground water.”

Contracting for risk
When it came to the alignment, of course the City gave heavy consideration to the geology, along with public right-of-way and public impact. “We looked at those on a cost basis and overall just trying to keep our microtunneling drives on the order of about 1,500ft [457m] plus or minus,” says Brad Moore, technical director of the Kennedy/Jenks consultants design team.

That led the design team to the preferred alignment, with six shafts and depths ranging between 40 to 75ft (12 to 23m). There were two existing diversion points where the tunnel needs to pick up flow before delivering it to the west side Nicolai shaft or shaft M (see figure 1).

Originally the tunnel was to be delivered in a standard design-bid-build route. Design plans had progressed to 30 per cent when the City of Portland started looking at the schedule and the risks associated with the project, and decided to look at alternate forms of contract delivery.

“We chose the Portland method, which is similar to CM/GC [Oregon’s Construction Manager/General Contractor], but it doesn’t restrict the prime contractor to perform a limited amount of work,” Moore says. “It also doesn’t have the guaranteed maximum price. The restructure is based on a reimbursable cost, plus a fixed fee. That was the vehicle for completing the design.”

The city issued an RFP for the project’s contractor and selected James W. Fowler Co., which came on board in November 2008 when the design was at about 60 per cent. “We had the contractor, the designer and the city at the table to complete the design of that project,” Moore says.

“Through that process we found many innovations to improve the project in terms of cost, environmental and societal issues. At the end of the PSA phase—when the contractor was on board to participate in the completion of the design—we entered a second contract with the contractor for the construction and that began June 2009.” The contract is worth USD 57.3M, and the project’s total budget is USD 74.4M.

The 84in diameter tunnel will be approximately 6,980ft (2,127.5m) long and is being mined by a Herrenknecht AVN2000D slurry machine. The TBM arrived at the site in November 2009 and was later launched in spring 2010. “During the pre-construction—from 60 per cent to final design—that was a period that allowed the contractor to also look at early procurement of equipment in terms of the micro TBM,” Moore says.

CSM shafts
Five shafts needed to be constructed, GLI, B, D, C and L, and the sixth existed, but needed major work (see side bar). Project plans called for secant or sheet piles to build the shafts, but in the end cutter soil mixing (CSM) was used instead, by suggestion of the contractor. “They proposed that to us initially as a cost saving measure that was anticipated to save us USD 600,000 to USD 1.2M,” Clement says, “And also as a schedule saver because we wouldn’t be subject to the drilling subcontractors providing these services.”

The City of Portland was reluctant to try CSM because it was untested in the project’s unique soil. “We relayed that back to the contractor, and they ended up coming back to the city with a proposal from the manufacturer of this equipment saying, ‘if we’re not successful, the city wouldn’t incur the full cost of it,’” Clement says. “Sort of an incremental purchase—they would prove successful at one site we would pay for so much; prove successful at a another site we would pay more.”

Contractor JWF worked with Jacobs Associates on the CSM, and with the city relieved of its concerns over the risk, a USD 4.5M CSM machine manufactured by German company Bauer was sourced through its Bauer Pileco office in Texas. The machine went through a test program at different sites in Portland to see how CSM would perform, particularly with the area’s gravels.

“The gravel alluvium is an open network of cobbles and I would classify it as boulders,” Clement explains. “It’s basically like a bag of marbles. There is nothing there to hold it. As soon as you release one or release the bag they just run everywhere.”

And that’s what was to happen at shaft L. The contractor needed to go at depths of 35 to 55ft (10.7 to 16.8m) to pre-grout the perimeter of the shaft while the CSM panels were drilled into place.

“We did our research throughout the world of CSM projects, and the kind of material that they ran into in various places around the world,” Jacksha says. “And to our knowledge, this is the only place the machine has ever worked with this amount of gravels or cobbles, and this kind of formation with very little matrix or no matrix.”

In the end the City of Portland is happy with the results. “It’s been very successful, and it came with some added benefits that we hadn’t originally considered,” Clement says. “One of those was, this methodology proved to be watertight. So at shaft B we didn’t have to deal with the contaminated water. It also proved to greatly reduce the amount of waste product that had to be hauled off. So there are additional savings there in just the construction of the support of excavations.”

Construction for shaft M, or the Nicolai shaft, had originally been part of another contract, by another contractor who had multiple failures excavating the shaft before abandoning it. “They had sand boils anytime they had removed enough of the overburden in the bottom of the shaft,” Clement recalls.

Together the contractor, designer and city discussed three options for shaft M, and chose one proposed by JWF to drill down into the existing jet grout fail, trying to supplement that with additional grout to seal the leaks (figure 2). After several attempts the contractor was successfully able to excavate down to grade without a failure on the bottom.

Diverse drives
The MTBM has a 102in (2.6m) outside diameter and is just under 50ft (15.24m) long with 10 disc cutters and 12 carbide bit picks on the face. There is also a 54in (1.37m) diameter drive from shaft B to the Nicolai shaft using a smaller machine through the sluiced West Hills material. Altogether the sequence of drives is: B to GLI; C to B; C to D; B to Nicolai (54in); L to M and then L to D.

The first drive, from B to GLI, was in the project’s softest ground and required CSM panels along the alignment to support the machine’s weight. These panels were placed approx. 15ft (4.57m) on centers for just over 400ft (121.92m). “That assured us that we wouldn’t lose grade or the machine wouldn’t sink, and we finished that drive within specifications for grade,” Clement says.

The project elected to purchase an optional second articulated joint with the MTBM for additional steering in the softer grounds. Along with the first drive, this was also used for the second, the C to B drive. A 9in-thick reinforced concrete jacking pipe, poured locally by Cascade Concrete, is being installed, with neat cement grout injected through pores in the pipe. JWF has been typically running two 10-hour shifts with downtime for maintenance in the second. By shift, production rates have been between 2.5 and 3.5 joints of pipe, or 25 to 35ft (7.62-10.67m).

After slurry lines deliver bentonite to the machine’s face for lubrication, waste lines carry spoil and bentonite back out to a separation plant. “We have a fairly elaborate separation process that allows us to reuse the bentonite,” Clement explains.

The separation plant consists of shakers that segregate the finer materials from the liquids, which go through the centrifuge and clarifiers to produce a cleansed slurry that can be reused for mining. Finer materials, sands and gravels go through shakers and come out to one of three bins: a mud bin that’s still very fine or liquid, a sand bin or gravels bin. The project is reusing more than 7,000t of the sands and gravels for backfill on the shafts.

In the third run, C to D, around 460ft (140.21m) into the drive the MTBM hit what is believed to be a boulder on top of the Troutdale formation, and the whole machine rotated 13 degrees. The operator was able to move the machine past the object, but that caused the it to go into negative pitch.

“The machine was starting to plow, and as the machine continued to slowly progress past this object, the pipe continued to get higher and higher,” Clement says. “Initially when we went over this object, the pipe was online and grade. Once we hit it—the machine went over it—we were 5in (127mm) over grade. By the time the machine got out of plowing, and out of negative pitch, the pipe had risen to over 26in (0.7m) above grade.”

Because this is a gravity system, the 26in (0.7m) hump would reduce conveyance capacity. “From an owner’s perspective, we were looking at a project that was not meeting original design specifications. That was a major issue for us,” he says.

The Portland method contract, with the reimbursable cost, removed the risk for the contractor of not being paid for work, Clement explains. “Instead of going into a mode where the contractor is claiming changed conditions and the owner is trying to refute that, we all sat down—the designer, the owner, the contractor—to discuss a solution.”

Being able to discuss the options together the project could move forward quickly, and that helped avoid a rescue shaft. That’s particularly significant as this section of the alignment is under landfill.

Jacksha explains, “The jacking pressures got to be tremendously high during that run and it really pushed the limits of that machine. But everything performed way better than expected.” Clement adds, had this been a low-bid contract it’s likely the contractor would have claimed changed conditions, and waited for a response. That process would have allowed the jacking forces to increase to the point the only option was to construct the rescue shaft.

“We evaluated the modeling and the hydraulics and found that although it’s not optimal,” he says, “A system with 26in (0.7m) high hump will meet the flow conveyance requirements for CSO events.”

After tunneling from C to D the machine’s face was refurbished and its disc cutters were replaced.

During the drive from L to M the face became gravel bound about 25ft (7.62m) into the drive. An optional airlock had also been purchased with the MTBM, and was then used for an entry under pressure to remove material by hand.

“The most difficult and aggressive soils, gravel alluvium basalts and quartzites from the Missoula floods deposits—we’ve run into them between L and M and L to D,” Clement says. “Quartzite is one of the hardest natural minerals you will find, and coupled with the basalts and the open network makes it a unique tunneling run.”

Even once cleaned, the head still wouldn’t rotate and when the machine was started up it became gravel bound again. The project team needed to find a way to stabilize the material in front of the face without locking up the machine. “You would lock up the machine if you were to do ground improvements with cementitious grout or concrete,” Clements explains. “We ended up using was a chemical grout to try to bind up those open network gravels.”

This grout program was used for 20 to 30ft (6.1 to 9.14m)in front of the face, before the contractor mucked out the chamber and started mining again, successfully finishing the drive. Again, JWF replaced teeth and rotator cutter discs on the TBM before moving on to the final drive, L to D, to be completed in early June.

“One of the things that was so unique about this project was the learning of how to microtunnel in these gravels and cobbles,” says Jacksha. “Throughout the project Herrenknecht provided machine operators. They had never encountered gravels and cobbles like this before.”


The micro TBM is readied Figure 1, the tunnel alignment and location of the six shafts Figure 2, the complex geology at the Nicolai Shaft