Maryland-based contractor Bradshaw Construction Corp. had a complex recipe on its hands: a series of four tunnels up to 1,200 ft long in exceedingly hard Baltimore gneiss. The City’s Lower Gwynns Run Interceptor Phase II, a gravity sewer, required good control of line and grade in highly variable hard rock that tested up to 38,000 psi UCS.
“The City of Baltimore has learned that they have extremely hard and abrasive rock – 45,000 to 50,000 psi rock is not uncommon. The rock often tests in the 4.5 to 5 range on the Cerchar Abrasivity Index. Those conditions are exceptionally challenging for a pipe jacking or conventional TBM tunneling operation. The biggest challenge is the cutting tools: disc cutter wear can be excessive and bearing failure can happen if the rock gets too hard,” said Lester Bradshaw, president of Bradshaw Construction.
The struggles and triumphs during tunneling proved that the contractor’s chosen method of excavation was indeed the right one, though tunneling itself was not easy.
A Project Years in the Making
The planned 36-in. diameter gravity sewer interceptor tunnels for the Lower Gwynns Run Interceptor were part of a consent decree to modernize a section of sewer line running below some of the city’s oldest and busiest streets, including Baltimore Street and Franklin Street. The first phase of the interceptor project encountered rock strength substantially higher than anticipated, eventually leading to the decision to remove and redesign the lower reach of the project.
Phase II consisted of four segments, totaling 2,500 ft in length, and individually measuring 395 ft, 260 ft, 1,200 ft and 650 ft. The longest 1,200-ft section traveled directly below active Amtrak and Norfolk Southern Railway lines. In addition, six shafts also needed to be built in various locations in order to launch and re-launch the chosen tunneling machine.
Choosing the Excavation Method
The contractor turned to Robbins for a solution in tunneling through the predicted hard rock. “When it comes to cutting rock, bigger is better. We knew we wanted as big of a bearing capacity on our disc cutters as possible but still needed to go as small as we could for the tunnel size. The 72-in. Rockhead came with a Heavy Duty (HD) cutterhead with 14-in. discs; the biggest cutter and bearing capacity we could get given the size of the tunnel.”
The initial tunnel support design was then upsized from a 60-in. diameter to 72 in. to accommodate the larger machine size. “If we went any smaller, we couldn’t use the 14-in. cutters. At 60 in., it wouldn’t have been possible,” continued Bradshaw. “Microtunneling machines are limited in cutter size; they use 11-in. back-loading cutters that wouldn’t have been suitable for this application.”
The variability of the rock was another reason for choosing the Rockhead: “We were expecting the rock to consist of highly variable formations — it could be incredibly hard in one spot and incredibly weathered in another. That’s why we went with a Double Shield model of the Rockhead. We knew we would be tunneling through mixed face and mixed reach rock conditions, and a conventional, un-shielded TBM with grippers couldn’t be used in those sections,” said Bradshaw Construction project manager Todd Brown. “We were looking for a different, less risky solution. Using a smaller diameter TBM with smaller cutters wouldn’t work, and we weren’t confident that our 83-in. Mark-6 Jarva TBM would be up to the task in loose soils and seamy rock.”
The machine was shipped to the jobsite for launch from the northernmost end of the project in November 2014. Due to utility concerns, the northern end of the project was sunk deeper to a depth of 55 ft, also allowing for an unsupported bore in solid bedrock. The entire tunnel project ranged in depth from 18 ft at the southern end, to 55 ft, at its deepest.
Urban Tunneling: Complex Site Preparations
Preparing the sites for TBM launch and recovery were also challenging. At the southernmost end the sixth and last shaft, which would serve as the tie-in shaft to the existing sewer system, required negotiating a maze of mapped and unmapped utilities. The shaft sat on Baltimore Street, one of the busiest and oldest streets in downtown Baltimore.
“In that main thoroughfare alone, we had a 12-in. water line, a 4-in. gas main, and multiple duct banks for fire safety and communications to contend with. The Verizon duct bank served as the main communication line for the city, so it could not go out. We were faced with building a 32-ft diameter shaft to recover a 60,000-lb machine below existing utilities, as well as a cast-in-place manhole structure, all while maintaining service of the existing sewer system,” said Brown.
“We had to rethink our plan. From the beginning, we had a window to recover the machine, and we had to set our shaft up based on that plan,” he said. “However when we started excavating we found that not all of the utilities were located according to that plan. To recover the machine, we wound up needing to push the TBM out past the utilities in segments, turn them sideways 90 degrees, and slide each of them through different windows as we raised it to surface level.”
Due to the location of the shaft and a mix of soft ground, as well as hard rock, crews hammered the rock out with small excavators rather than using drill and blast. Excavation began in March 2015 and was completed in mid-August 2015, although it did not proceed full-time as tunneling was also going on concurrently. The other five shaft locations were in rock and were excavated by the drill-and-blast method. While most shafts were in industrial areas, vibration was still a concern and all shafts were closely monitored during construction.
Tough Tunneling Below Charm City
Once the machine was launched on its first tunnel, the difficult nature of the rock became apparent. While most of the initial tunneling was able to be driven without adding ground support as bald tunnels, the rock strength made it slow going. As transition zones were encountered in subsequent tunnels, cutter wear went up, even with the HD cutterhead and 14-in. disc cutters, but this was expected. “When we were transitioning from weathered rock into hard and massive rock, we would catch a shock load on the cutters that could chip them. There was higher wear on the gage due to abrasivity,” said Brown.
Steering was also a challenge in the weathered zones and production was significantly slower. “Most of the tunnel was bald, but in the transition zones we had to go through and install ribs and lagging as tunnel support to assist in propulsion of the machine through those sections, and to make the tunnel safe to work in. The front stabilizers could not grip the rock and the TBM wanted to drift up and to the right, so we counteracted that by requiring rear thrust off of the temporary lining support. It was an interesting problem,” said Brown. Several different sections of ribs and lagging, ranging from 15 to 50 ft in length, were required in these transitional zones.
Liner plate was installed below the active railway line as part of the contractual requirement. The longest 1,200-ft crossing went smoothly despite the logistics of the tunnel length at small diameter. “Our ventilation was designed for a longer 1,200-ft run. We used a single muck car train with a tail tunnel in the shaft. The cycle time at that distance went well with one car, and so we did one stroke per car. It took about the same amount of time to cycle muck cars through as it took it to re-grip, so there was not much downtime,” said Brown.
Advance rates reflected the range of obstacles encountered. In very competent rock of moderate strength, crews were able to achieve up to 37 ft in one 10-hour shift. However, significantly lower advance rates in decomposed rock zones, downtime for maintenance and installation of tunnel supports produced an overall average rate for the project of 10 ft per shift.
The final tunnel was completed on Sept. 18, 2015. As it was a 36-in. fiberglass gravity sewer line, the pipe was set to proper grade and then back-filled in place using multi-stage, no aggregate grout. Due to a fly ash shortage on the East Coast at that time, Bradshaw was required to engineer its own cement bentonite grout for the last 20 percent of the tunnels, making the whole process a bit more involved.
Lessons Learned in Hard Rock
Despite the difficulties, the project was ultimately very successful, and the challenges made the job all the more rewarding. “There are highly variable rock formations throughout Baltimore. We knew this from many previous projects in the area. We wanted to maximize the good conditions and minimize the impacts from bad ground. There’s no perfect machine out there that can bore in both very hard and very soft conditions. We felt the Double Shield Rockhead was the best fit. We figured maximizing our performance in the 80 to 90 percent of the tunnel that was in hard rock formations while going slowly through the weathered zones was the best option. The true variability of the rock was the most interesting thing — we really didn’t know what the next section of rock would be like,” said Brown.