Narragansett progress report Mar 2005

By mid-February, tunnel excavation by Shank-Balfour Beatty (S/BB) on the Narragansett Bay CSO abatement project in Providence, Rhode Island, was approaching mid-way. Shanks 30ft (9m) o.d. TBM was about 7,400ft (2,280m) into its total 16,215ft (5km) long drive and progressing at between 50-75ft (20-23m) in an 11h production shift/day.

At the same time, excavation of the contract's 430,000ft³ (12,180m³) pumping station cavern was complete. Following final walk-through on 12 February, the works were turned over on schedule to the M&E installation contractor. In addition, drill+blast excavation of the total 4,100ft (1,250m) of connecting adits to the main tunnel was more than 10% complete. The first 695ft (212m) long adit off the north access shaft, into which the TBM will hole through, was 75% complete and the first of five adits to be excavated off the main tunnel line and concurrent with TBM advance, had started.

"Excavation of the cavern went well," said John Kaplin, excavation works Resident Engineer for the projects Gilbane/Jacobs Associates construction management JV. "In addition, TBM advance has improved steadily with the current best advance of 76ft (23m) in an 11hr mining shift achieved in January. Prior to that, progress was being slowed in a reach of about 3,000ft (900m) through wetter rock conditions in which grouting cycles had to be incorporated into the production cycle. Tunneling has since passed into a dryer zone of shale and sandstone and over the last 900ft (275m) or so, production had improved by being able to mine on all five production shifts/week instead of mining for three or four and grouting for one or two. At present TBM excavation is expected to be complete by the end of this year."

Injection grouting station at the back of the TBM

The GBR for the TBM driven tunnel predicted cumulative water ingress of 3,000gal/min for the entire tunnel. At about 40% through the drive, inflow had reached 1,700gal/min, said Kaplin, but the need for grouting is triggered by water ingress through the two mandatory and systematic probe holes.

"Probe drilling is carried out during the two maintenance shifts/day," said Mike Shank for Shank-Balfour Beatty. "Two probes of about 220ft (67m) are drilled and if water inflow from either is more than 15gal/min, this triggers a grouting cycle using a four-hole pattern. The 220ft probes give us about three mining days before having to repeat the process to maintain the required 30ft overlap. We do the same for the grouting cycles - drill 220ft long injection holes, place the packers at about 20ft and grout up the 200ft lengths. On occasions we have injected up to 4-5,000ft into the ground through the four injection holes."

Grouting is a unit price item in the contract, explained Kaplin. "We are not trying to achieve a watertight tunnel. Probes that exhibit an inflow of 10gal/min for example do not trigger a grouting cycle, but in such cases, we can end up with 70-80gal/min into the tunnel through that section. It is the cumulative ingress as measured at the sump in the 290ft (88m) deep TBM access shaft that we are monitoring and controlling."

Ground water inflows to date, said Shank, have been "greatly in excess of what was anticipated, and we have already exceeded the contractor’s planned quantities of grout injection. This was also costing considerable time and we suggested equipment upgrades to increase grout pumping capacity and install the ability to drill two holes concurrently. The owner [the Narragansett Bay Commission] agreed and paid for the new equipment in order to improve progress from three days mining, to two days grouting, to one day grouting, four days mining. So far we have cut off several thousand gal/min of potential ingress before moving into the current drier reach of ground."

Ahead, the ground type remains the same - a combination of sandstone, shales and siltstones, with the same amount of rock cover, at about 100-150ft (30m-45m) and another 50-140ft (15-43m) or so of topsoil. The site investigation data however, as prepared for the contract by project geotechnical engineer, Haley & Aldrich, predicts less water ingress for the second half of the tunnel alignment. "But you never can tell," said Kaplin, "and we must grout when necessary to control inflow and contribute to the performance specification as designed by the owner's project consulting engineer The Louis Berger Group, of no more than 150gal/min of ground water seepage through the secondary in-situ concrete final lining."

Fork lift platform (top) to advance drill+blast of connection adit concurrent with TBM excavation with service trains passing underneath (bottom)

By mid-February TBM tunnel progress was running about 10 months behind schedule. This is due largely to about eight months lost during ground-freezing construction of the TBM working shaft. Through a subsequent schedule revision, Shank has saved about 100 days of that but still the delay is about 10 months. Current forecasted completion of the contract has slipped as a result from March to December 2006.

Initially S/BB had prepared its bid on achieving an average of about 80ft (24m) in two production shifts with one maintenance shift/day. In reality two maintenance shifts after each 11hr production shift are needed to clean out the excavation chamber of the TBM's particular design and to change cutters. Cutter consumption is running at about one disc/5ft in abrasive sandstone and at one disc/13ft as an overall average to date.

To save time, Shank's schedule revision proposed excavating the five on-line drill+blast adits concurrent with TBM advance rather than completing these once the TBM had finished. "This is where a major portion of the 100 days recovery was achieved," said Kaplin, "so excavation of the adits concurrent with main tunnel excavation is a significant time saving."

Shank has just started the first on-line adit and is progressing the 12ft diameter horseshoe heading from a 60ft (20m) long working deck elected at just below spring line in the 28ft (8.5m) i.d. tunnel. "Frorn this deck we can work three drill+blast production shifts/day and maintain TBM service train traffic underneath. The longest of the five online adits is 1,700ft (518m). The next longest is 695ft (211m). Completing these adits, even concurrent with TBM advance, is the schedule path. We plan to have two adits under excavation at the same time. Once all excavation is complete, the largely unreinforced flnal in-situ concrete lining, except in zones of poorer rock conditions and at adit junctions, will be cast."

With concurrent excavation of the adits; more positive TBM progress rates of late; the prediction of drier conditions ahead; and higher production grouting cycles when needed; there is the chance to gain further on the delay as tunneling progresses.