A 1.069Mb PDF of this article as it appeared in the magazine—complete with images—is available by clicking HERE
The Trout Brook Storm Sewer Interceptor (TBI) serving St. Paul, Minn., is owned and operated by the Capitol Region Watershed District (CRWD). In 2011, the Minnesota Department of Transportation (MnDOT) informed the district that new bridge construction on I-35E interchanges would inevitably interfere with TBI’s existing alignment.
On top of this, large portions of the interceptor were already failing–understandable, given the TBI’s 100-year service life. This is an absolutely critical component of the region’s stormwater management network, totaling about 6.5 miles of high volume sewers and tunnels serving an 8,000-acre collection area. The section that would be most directly affected by the I-35E interchanges was a mile-long, horseshoe-shaped, cast in place, 13-foot-diameter tunnel. And this section was already failing due to cracking, eroded concrete, exposed steel and, in two sections, subsidence-caused sagging of up to three feet.
The MnDOT notice wasn’t a complete surprise, of course, and CRWD had been looking at sewer realignment and repair possibilities for years. Relocation was also an opportunity to rehabilitate this large piece of vital infrastructure, but initial cost estimates for the work were daunting. A 2006 feasibility study had recommended complete replacement and rerouting of the section, at an estimated cost of $16 million.
In 2008, the district asked Barr Engineering Co., an engineering and environmental consulting firm, to take another look at the sewer, a commission they took rather literally. "We were lowered on harnesses into the tunnel," says Barr Senior Civil Engineer Jim Herbert, PE. "It was dark, wet, and we even saw the occasional angry muskrat. But it was definitely the best way to really find out what was going on."
Barr had previously completed inspection and design work on St. Paul’s Beltline Interceptor System, which was similar to the TBI. The lessons they learned there, together with the visual inspection and extensive hydraulic modeling (which particularly helped designers understand the sagging sections and related sediment accumulation) helped them to put together an entirely different resuscitation plan, one based on rehabilitation rather than replacement. The upside was gigantic, in terms of the bottom line; Barr’s cost estimates came to just over $3 million, a savings of $13 million compared to previous estimates. Of course CRWD and MnDOT were interested.
The plan devised by Barr was logistically challenging, with a lot of moving parts affecting a dozen or more major stakeholders and permit issuers. It involved new footings and piles in cramped spaces, the raising of existing foundations, some open cut trenching, hydraulic redesign, and difficult logistics to keep roadway, rail, and air traffic moving smoothly. Moreover, the design phase and permit issuing had to be aggressively fast tracked so that critical tunnel sections could be completed over winter in order to avoid possible difficulties arising from spring flooding.
Despite the complex construction challenges, short timelines and high stakes in terms of dollars and potential dollars, for today’s urban infrastructure specialists, it’s all in a day’s work. "We did about six months of detailed structural and geotechnical design in just three months, along with a Phase One environmental assessment," says Barr Vice President and Senior Geotechnical Engineer Aaron Grosser, PE.
The project also required a NPDES construction stormwater-discharge permit, which was particularly important to Capitol Region Watershed District given its role as the permitting agency and public protector of water quality. Fortunately, all the agencies involved with the project were able to work together efficiently, and everything came together in time.
One component of the plan emerged early as particularly challenging and troublesome. Work on one section of the TBI would require installation of 162 feet of 12’x9′ box culvert, which in turn would require removal and replacement of 120 feet of rail operated by BNSF, and in the area of track affected, the top of the sewer tunnel was just five feet beneath the top of rail. Understandably, BNSF was concerned about the effect of any work in the area. This is a very busy section of commercial railway–30 to 40 trains daily–and shutdowns are almost unheard of, and stoutly resisted by BNSF. Still, BNSF agreed to an unprecedented 30-hour suspension of rail activity.
To further complicate work in this section, high groundwater and difficult soils made subsidence an issue, and railway settling was a real concern even without construction work. To allay BNSF’s concerns, Barr devised a technically progressive subsidence monitoring system based on 250 track-mounted prisms and two Leica TM30 optical monitoring sensors. With this worked out, the 30-hour "Big Dig" was set to proceed.
Precise Railway Monitoring
The 162 feet of existing tunnel slated for replacement ran directly under the BNSF main railway line, which in turn was under the I-35 bridge. The Big Dig was scheduled for Labor Day 2012, and Barr’s monitoring of the tracks began two weeks before that date to establish movement baselines and for later positional restoration after removal and replacement. Monitoring was also needed to verify track stability during the extensive dewatering that preceded the Big Dig.
"We set up a total of 260 prisms, at 30 foot intervals," Swenson explains. "And we placed them at tighter intervals at critical areas." These were conventional Leica prisms, placed directly on the wooden ties, or sleepers.
The two TM30s were then set up to run continuously, constantly sweeping all 260 prisms. Leica GeoMoS automated deformation monitoring software was used to automate the continuous TM30 shots and note any movement from baseline in near real time. GeoMoS was also used to set the amount of track movement that would trigger alarms. Two tolerance levels were set; movement of a half inch or less would trigger internal review by Barr, and movement over three quarters of an inch would automatically alert BNSF personnel. Vista Data Vision (VDV), a software solution that manages sensor data, was used to publish data and alerts. VDV directly imported data files from the TM30s wirelessly, and distributed status alerts via the internet.
There was no monitoring during the actual big dig, of course, but movement was detected before and after, and the real-time monitoring proved to be very important for safe railway operation. "During the dewatering phase, we did in fact observe settling of about a tenth of an inch," Swenson says. "Fortunately, that amount of settling was within the tolerances for the project, but it was good to be keeping an eye on it, and to verify that dewatering could have an observable effect."
Monitoring continued for several months after big dig completion, and more serious movement was detected. "We watched the tracks from big dig completion in September until winter freeze," Swenson explains. "And substantial settling did occur, about an inch and half. This was well over tolerance, so BNSF resurfaced the bed and adjusted the rails. Then in spring 2013, when the larger bridge project started to affect the site with new piling installation, we saw even more settling–up to two inches, which led to more adjustment. But rail traffic was never actually stopped."
Increasing Importance of Monitoring
Large construction projects are increasingly monitored before a
ny ground is broken, during actual construction, and after for potentially long periods, especially in urban areas. Monitoring during the Alaska Way Viaduct tunneling in Seattle, for example, featured more than 4,000 sensors (including several Leica TM30s) deployed to instantly detect movement in any of the many buildings that Big Bertha (the record-sized tunnel boring machine made for the project) passed under as it created the viaduct tunnel.
The TBI Big Dig is an excellent example of what careful monitoring makes possible. Long term monitoring measuring from existing baselines enabled crews to do a staggering amount of work in a short period of time, while keeping rail traffic safe during the dewatering phase, and speeding up restoration of the rail to its pre-construction position. During the 30 hours of the Big Dig, workers were able to remove 120 feet of railroad track and track ballast, trench, place 162 feet of 12’x9′ box culvert, and replace all track. They were even able to cope when an unsuspected stone arch tunnel was discovered during work; the seven-foot stone-arch tunnel appears to have predated the TBI. The project team and BNSF Railway chose to have the tunnel plugged and capped to allow BNSF to address the remainder of the tunnel at a future date.
Monitoring was also useful after the project; without a monitoring system in place, the unexpected and serious subsidence in August 2012, and early in 2013, could have been an emergency rather than routine maintenance.
For its part, BNSF couldn’t be happier. "BNSF trusted us to do this work, even though the technology was new to them," Grosser says. "Since the Big Dig, they’ve become proponents of the monitoring system we used and have even given presentations on it and how well it worked."
Angus W. Stocking is a licensed land surveyor who has been writing about infrastructure since 2002. For more information about Barr Engineering, visit www.barr. com. To learn more about monitoring solutions, visit www.leicageosystems.us/monitoring.
A 1.069Mb PDF of this article as it appeared in the magazine—complete with images—is available by clicking HERE