Lucian P. Spiteri, P.E., Paul C. Schmall, Ph.D., P.E., Mark Gentile, P.E., and Michael A. Nuhfer, P.E.
The OSIS Augmentation and Relief Sewer (OARS) project is a 23,300-ft long, 20-ft finished diameter combined storm and sanitary overflow relief sewer mined through dolomite and limestone rock approximately 170 ft below the surface. Designed by DLZ Corp. in association with JEC Consulting and CH2M Hill to augment the existing Olentangy Scioto Interceptor Sewer (OSIS), the OARS project is one of the largest capital investment elements in the City of Columbus, Ohio’s Wet Weather Management Plan. Prior to OARS, overflows from OSIS have emptied into the City’s Olentangy and Scioto rivers in times of excessive rainfall. When completed, the OARS tunnel will collect OSIS overflow and channel it through a flood diversion structure to the Jackson Pike or Southerly wastewater treatment plants.
In addition to the tunnel, primary components of the overall project include construction of six deep shafts, four tangential inlet drop structures, three de-aeration chambers, screening structure, pump station and river outfall structure. The project is under the construction management of Black & Veatch and HR Gray and was bid in two phases. Tunnel mining and shafts 1, 2 and 6 were bid under Phase 1. Phase 2 includes Shafts 3, 4 and 5, the bypass relief tangential inlets and de-aeration chambers for those shafts, the pump electrical control building, and a new river outfall structure. Trumbull Corp. was awarded the contract for Phase 2 construction.
The ground conditions at shafts 3, 4, and 5 were situated in the more permeable sections of rock along the alignment, which raised significant concern for pre-excavation grouting for groundwater control. Phase 2 Shafts 3, 4, and 5 are nominally 17, 30 and 26 ft in excavated diameter, respectively, and extend approximately 170 ft below the surface. They each connect to the sewer tunnel via a de-aeration chamber and an adit. The shafts were constructed by conventional excavation within secant pile earth support installed through the overburden and extending 5 to 7 ft into rock. Drill-and-blast techniques were used to sink the shafts to design depth.
The Phase 2 design and construction teams were faced with several challenges. The drill-and-blast portions of the shafts extended through highly variable karst conditions ranging from small fissures to large solution voids. Groundwater was approximately 20 ft below the surface which, given the shaft depths, meant that the work would be accomplished under significant hydrostatic water pressure and a potentially high groundwater inflow. Addressing this potential was therefore a critical aspect of the overall design and construction program.
The subsurface profile was similar for the three Phase 2 shafts. The overburden consisted of up to 70 ft of mostly high permeability sand and gravel alluvium deposited by the Scioto River. The rock consisted of approximately 40 ft of shale over three different limestone formations that contained karst solution features. No shale was present at the Phase 1 shafts. The presence of the shale at the Phase 2 shafts corresponded to less infilling in the rock and cleaner, more permeable solution features.
The single most telling piece of hydrogeological data was an aquifer pumping test conducted at Shaft 5. An open borehole in rock, drilled to the shaft depth, was pumped at 490 gallons per minute with only 4 ft of groundwater drawdown in the pumping well.
One of the primary parameters for the viability of conventional drill-and-blast techniques for shaft construction was the hydraulic conductivity of the rock. During shaft excavation, estimated inflows of thousands of gallons per minute would not be manageable. Provision for adequate pre-excavation grouting was therefore critical to the success of the overall program and was included in the bid documents.
Alternate Proposed to Original Grouting Scheme
In order to ensure that excessive inflows did not occur during shaft sinking, the original shaft grouting scheme suggested down-stage sequenced grouting from within the shaft as the excavation proceeded, with a circular array of vertical and battered grout holes installed outward and downward in a radial fan pattern extending from a temporary bench. The contractor was required to drill probe holes to assess the adequacy of the grouting program and proceed with excavation only if minimal groundwater inflow occurred through the probe holes.
The general contractor, Trumbull, accepted an alternative approach proposed at the bid stage by specialty geotechnical contractor Moretrench whereby all drilling and grouting for the shafts would be performed from ground surface prior to excavation. In addition to vertical grout holes, a second row of battered grout holes would be installed to intersect to the numerous near vertical fractures anticipated. The major advantage to this option was that the grouting would be performed from the surface, rather than from inside of the shaft.
Grouting Program Challenges
Karst geology is, by nature, highly variable. The grouting program therefore needed to accommodate a very wide range of conditions, from small features to larger voids. With significant hydrostatic head on the order of 150 ft anticipated at each shaft location, even small features under this pressure could result in significant water inflow. Vertical and battered drilling at significant depth would be inherently difficult. Encountering a large feature while drilling through this highly variable ground could also lead to complete loss of drilling fluid as well as hole instability. The proposed drilling and grouting plan had to overcome this challenge.
And last but by no means least, the four discrete rock units needed to be addressed individually in terms of the grouting. The grout had to be able to penetrate small joints and fissures as well as fill large voids effectively and efficiently. This required tailoring the grout formulation to the range of subsurface conditions.
Grouting Program Design & Application
Given these challenges, a deep-hole grouting program was developed that would achieve the program goals. This would be accomplished by creating a two-line grout curtain around the perimeter of the shafts within the bedrock, consisting of vertical holes drilled through reservation sleeves installed in the secant piles, and an outer circle of holes drilled through permanent surface casings installed in the overburden. The maximum hole spacing was approximately 7 ft. Since the potential for significant groundwater inflows below the bottom of the shaft was likely, an array of holes would also be drilled in the center of the shaft to enable grouting of the rock at the shaft bottom.
To address the variable conditions anticipated, a grouting program was established utilizing a series of primary, secondary and tertiary grout holes. Successive stages within each hole were isolated with inflatable packers in order to treat the rock in discrete zones and provide the required degree of monitoring and control over the operation. Whenever significant voids or other unstable conditions were encountered during drilling, the tooling was withdrawn and the unstable zone grouted and allowed to set. Drilling then proceeded through the grouted zone and the hole advanced.
A suite of four balanced-stable grout mixes were developed that exhibited minimal bleed so that the grout would not separate in the fractures. A low-viscosity grout was injected first. Two progressively thicker mixes were utilized as warranted to address more open features. A low-flow, paste-like consistency grout containing sand was designed to address wide open karst features and minimize grout travel outside the target area. At each stage, the thinnest grout mix was injected first, followed by incrementally thicker grout until refusal criteria was reached.
Drilling was accomplished using a modified Commachio MC1200 track-mounted rig equipped for deep rock hole drilling. A Wassara water-fired down-hole hammer system was utilized. This style of down-hole hammer utilizes high pressure water to fire the hammer as well as flush cuttings from the hole. The use of water as the flushing medium produces a cleaner borehole, which maximizes the ability to access finer features in the rock with grout. To track the features encountered during drilling, the drill rig was equipped with an automated data recording system that tracked the drilling parameters. When large features were encountered within a particular stage during the drilling, or grout takes indicated runaway conditions, an allowance was made in the grouting program to rapidly thicken the grout to minimize runaway takes and keep the grout within the target zone.
In order to maintain grout quality during production work, a computer-controlled batch plant was utilized, with Marsh Funnel grout viscosity checks made daily and periodic laboratory testing conducted by an outside agency. The grouting contractor also utilized an automated grouting data acquisition system as a tool to actively monitor and record the progress of the grouting as the work continued. In addition to monitoring and recording the location, duration, flow rates and pressures utilized during injection, the system also plotted the apparent Lugeon value of the stage during grouting. This feature allowed the crew to observe the behavior of each stage over time, make informed decisions on when to change to a thicker grout mix, and assisted the operator in determining when refusal had been reached. This allowed for day-to-day management of the grouting program, even when considering the complex grout hole geometry required to perform the grouting. The initial apparent Lugeon value plotted over time for each shaft displayed a rapidly decreasing trend in the permeability of the rock during the initial phases of the grouting program (primary holes). The rate of decline in apparent Lugeon then tapered quickly as the grouting proceeded through the secondary and tertiary holes toward the 1 Lugeon mark, which is approximately 10-5 cm/sec permeability.
Ground conditions encountered at each shaft were slightly different and challenging in their own respects. Approximately 10,000 lf of surface casing and 19,000 lf of drilling was required for this project. Drilling through karst bedrock resulted in some instability and water loss. Many holes therefore had to be grouted in successive down-stage fashion. Significant communication occurred between open grout holes, sometimes between holes more than 50 ft apart. This typically meant that only one grout hole could be open at a time during the initial phases of the grouting program in order to minimize the risk of inadvertently grouting shut other open grout holes. Adjustments were made to the initial grouting sequencing as well as to the crew size to handle the modified work flow.
Throughout the program, the grout takes were erratic, primarily governed by the features encountered during drilling. As a general rule, grout takes increased with the increased solutioning of the rock with depth. In general, conditions worsened (as measured by karst solution features encountered and increased grout takes) as the grouting operation proceeded from Shaft 3 to 4 and then to 5, with Shaft 5 taking 30 percent more grout than Shafts 3 and 4 combined.
Shaft 5 De-aeration Chamber and Adit Pre-grouting
At Shafts 3 and 4, pre-grouting for the de-aeration chamber and adit tunnels was performed horizontally from inside the shaft, in accordance with the original design. However, given the karstic conditions and high grout takes realized at Shaft 5, pre-treatment grouting of the rock in the adit horizon was performed from ground surface to reduce the risks associated with encountering significant inflows during drilling from the chamber and adit. A three-row grout curtain was designed and drilled from the surface.
Planning Pays Off
A proactive approach to the challenges that presented themselves to the project team prior to and during this project was instrumental in delivering the grouting solution needed for successful completion of the Phase 2 shafts and adits. In particular, the Shaft 5 conditions required numerous day-to-day sequencing changes, but given that the pre-bid aquifer pumping test indicated extremely permeable karst conditions, this end result was exceptional. With grouting completed, excavation was able to be accomplished with minimal conventional sumping needed to handle the low level residual water infiltration.
Lucian Spiteri is a Senior Project Engineer and Paul Schmall is a Vice President and Chief Engineer with specialty geotechnical Contractor Moretrench. Mark Gentile is General Contractor Trumbull’s Project Manager, and Michael Nuhfer, who was at the time of the work a Senior Engineer with Black & Veatch, is Midwest Manager with Aldea Services.