Managing Tunneling Induced Ground Movements: Geocomp

Lucky Nagarajan, Director of Business Development, Consulting & Monitoring, Geocomp; W. Allen Marr, Former CEO and Senior Strategic Advisor, Geocomp

Instrumentation and monitoring are crucial for ensuring safety, stability, and quality in tunneling projects. With tunnels and metro lines being constructed worldwide, it is essential for geotechnical engineers to understand the risks and ensure community safety. The LA Metro Crenshaw/LAX Corridor project demonstrates the value of real-time monitoring in minimizing ground movements and damage from urban tunneling.

The project, part of the Los Angeles County Metropolitan Transportation Authority’s (Metro) expansion, includes 8 stations, cut-and-cover structures, and twin bored tunnels. This paper focuses on Underground Guideway #4 (Figure 1a), spanning 8,430 ft, with three underground stations: Expo, MLK, and Vernon. Ground movements discussed in this paper are along the alignment shown in Figure 1b.

Figure 1. (a) Location of Underground Guideway #4, LAX/Crenshaw Rail Project; (b) Extent of Alignment of Interest and Select MPBX Locations

The twin tunnels, with a centerline separation of 39.13 ft, were excavated using an EPB TBM with a 33-ft shield. The “overcut” dimensions increased from 0.6 inches behind the cutter head to 1 inch at the shield's back. The excavation diameter was 21.49 ft, with precast concrete rings of 18.83 ft I.D. and 5 ft length.

The tunnel axis depth ranged from 51.5–54.5 ft between Expo and MLK Stations and 41.5–61.5 ft between MLK and Vernon Stations. Performance limits for surface movements were 0.5 inches (action) and 0.6 inches (maximum), and for bottom anchor movement, 0.75 inches (action) and 1.5 inches (maximum), measured 5 ft above the tunnel crown (Adib et al., 2018).

Several initial challenges related to optimizing TBM control parameters were overcome. The low advance rate was resolved by adjusting TBM mining parameters, including foam injection and cutter head torque. Ground movements exceeded project performance limits, correlating with malfunctions in the TBM shield bentonite system (Chan et al., 2017). Although the 33-foot shield prevents significant caving, ground movement can occur in the overcut region if a pressurized “envelope” is not maintained (Cording, 2016). The TBM’s bentonite injection system, designed to form a bentonite “cake” within gravelly material, required volumetric control matched to the advance rate and pressure limits (Dias and Bezunjian, 2015). Injection pressure was set close to the calculated face support pressure at the tunnel crown (Chan et al., 2017). Following bentonite injections, surface ground movement was typically 1/4 inch after each tunnel drive.

Project Background

Geological and Geotechnical Information: The project is in the northern Los Angeles Basin, underlain by unconsolidated Quaternary alluvial sediments (Metro, 2012a), subdivided into Holocene-age Young Alluvium and late-Pleistocene Old Alluvium. The tunnels were excavated in the Old Alluvium.

The Young Alluvium, 30 ft thick, consists mainly of fines (silt, lean clay, and fat clay) with frequent organic clay strata. The upper 50 ft of the Old Alluvium (30–80 ft depth) comprises coarse sediment, including pebble-gravel and sand. Groundwater was reported at 45–60 ft depth.

Geotechnical Instrumentation and Monitoring: Geocomp Corporation procured, installed, and monitored the geotechnical instrumentation system to track construction-induced movements and groundwater draw down within the station and tunnel excavation zone. The system included manually and automatically monitored instruments, with data logged into Geocomp’s cloud-based iSiteCentral® and linked to the TBM data management system to correlate ground movement with TBM position.

Subsurface movements were tracked using Multi-Position-Borehole Extensometers (MPBX) with grouted anchors. Each MPBX had three anchors with displacement transducers measuring relative movement between the anchor and MPBX head. Anchors were located 5 ft above the tunnel crown, 5 ft below the surface, and midway between. The MPBX head, housed in a well box and “de-bonded” for independent movement, was surface-protected. Surface movements were monitored using reference points (pk nails) and Feno-type anchors.

Observed Ground Movement

Deformations between Expo and Vernon stations: The ground movements discussed were measured during mining of the SB tunnel between Expo and Vernon stations. The MPBX locations are shown in Figure 1(b).

Figure 2. Measured ground movement between EXPO and MLK stations

Figure 2 summarizes maximum absolute movements at the bottom anchor of each MPBX, corresponding surface movements at nearby settlement points (SMPs), and at the MPBX road box rim, plotted against the distance from the tunnel start at Expo station. Key observations from the data:

• Most relative movement at the bottom MPBX anchors occurred just above the TBM shield after the cutter-head passed.
• Ground movement, both at the surface and depth, decreased following the remedial measures.
• Relative movement at the bottom anchor of MPBX 39 exceeded the uniform maximum overcut thickness, supporting prior findings (Loganathan & Poulos, 1998; Pinto & Whittle, 2014) that the actual gap above the tunnel crown can surpass the overcut thickness.
• The reduction in ground movement at MPBX 39 after the TBM shield tail passed confirms that tail grout pressure can lift the ground, reducing settlement by approximately 0.5 inch.

Deformations at the MLK Station: Settlement monitoring points outside the station show unexpected heave instead of settlement. Figure 3 indicates up to 0.45 inches of heave at the surface, with over half the points consistently showing heave up to 220 ft from the excavation support system. Engineers had predicted up to 1 inch of settlement near the excavation.

Figure 3. Heave measured at points 4615 through 4620 (points located along the northwest exterior face of the MLK station support)

Interpretation

Interpretation of Deformations between Expo and Vernon stations: After implementing the improved bentonite injection and control system, the bottom anchor movement in the southern set was significantly reduced, even less than the surface movement near the MPBX location. In contrast, the northern set showed bottom anchor movement equal to or greater than the surface movement. Literature from the Seattle North Link and University Link projects (Salvati et al., 2016) indicates that in EPB TBM operations, bottom anchor movement can be equal to, greater than, or less than surface movement, independent of TBM performance.

Evidence of ground movement changes observed at the MPBX bottom anchors was supported by transverse surface settlement troughs. Figure 7 shows normalized movements from reference survey points relative to the maximum movement, Smax, at the SMP nearest to the southern set MPBX. The southern set's normalized trough fits a curve with an inflection point at X=0.6HX = 0.6HX=0.6H (H=51.7H = 51.7H=51.7–52.752.752.7 ft), indicating a wider settlement trough. The northern set fits a Gaussian curve (Peck, 1969) with an inflection point at X=0.45HX = 0.45HX=0.45H (H=52.7H = 52.7H=52.7–54.754.754.7 ft) and shows heave away from the centerline.

Figure 7. Normalized transverse settlement at MPBX 19 to 35 arrays

Interpretations of Deformations at the MLK Station: The heaving observed, instead of the expected settlement, is unusual for excavations. Settlement typically occurs due to: 1) inward movement of the ESS, 2) dewatering-induced consolidation, 3) internal heave causing exterior settlement, or 4) ground loss during ESS placement.

At MLK Station, the ESS is extremely stiff, limiting inward movement to ¼–½ inch, as shown by inclinometers—small compared to similar excavations. Minimal dewatering occurred due to deep original groundwater levels, and the high factor of safety in medium to stiff soils kept internal heave minimal. Additionally, ESS construction caused little to no ground loss.

With 65–68 ft of soil and water removed, significant unloading caused heave. Normally, this would be masked by larger settlements from typical mechanisms, which are insignificant here. The heave is expected to stop shortly after excavation completion.

Conclusions

Based on the observed tunnel induced ground movement pattern, it was necessary, for interpretation purpose, to aggregate these 18 MPBXs into two spatially distinct sets: a northern and a southern set.

Monitoring data indicated that following the bentonite injection and control system, the relative movement of the bottom anchor of the southern set was measured to be significantly less than before, even smaller than the surface movement near the MPBX location. This behavior was in contrast to the observed behavior of the northern set, where the relative movement of the bottom anchor was about equal or more than the surface movement.

The mechanisms that typically cause ground settlement around a deep excavation are not significant for the MLK Station. Since 65-68 ft of soil and water have been removed from the excavation large unloading took place contributing to heave. Normally such heave would be masked by the larger settlement from the mechanisms described above but that is not the case here. Such heave would be expected to stop soon after completion of excavation.