Geotechnical Instrumentation Monitoring System for Shallow

Geotechnical Instrumentation Monitoring System for Shallow
Freeway Tunnel Crossings with EPBMs
Sam Swartz, PE
Jacobs Associates, Seattle, Washington
Rick Smith, PE
Jacobs Associates, Seattle, Washington
John Sleavin, PE
Central Puget Sound Regional Transit, Seattle, Washington
Pierre Gouvin
GEO-Instruments, Narragansett, Rhode Island
Mike Switzer
GEO-Instruments, Lynnwood, Washington
ABSTRACT: The U230 Contract for Sound Transit’s University Link Light Rail Project in Seattle,
Washington, required two EPBM tunnel drives beneath a major arterial freeway. Each tunnel had roughly
one diameter of cover or less at the crossing, with tight WSDOT-imposed limits on allowable ground
movements. A redundant geotechnical instrumentation monitoring system was developed during design, and
modified slightly and carefully monitored during construction. Precision excavation and close collaboration
of all parties resulted in successful crossings. Resultant movements were well below allowable limits. This
paper discusses this monitoring system and insights gained into monitoring and managing settlements for
shallow tunnel crossings.
INTRODUCTION
The University Link (U-Link) Project is a critical extension to Central Puget Sound Regional Transit’s
(Sound Transit) light rail system in Seattle, WA. The U230 Contract extends the existing system from
downtown Seattle up to the heavily populated Capitol Hill neighborhood to the northeast. As part of the
alignment, the twin tunnels both cross beneath the Interstate 5 (I-5) freeway, the major arterial through the
Downtown Seattle area. Cover above these tunnels is less than a diameter beneath the high-occupancy
vehicle (HOV) lanes, and just over a diameter beneath the main travel lanes in the southbound direction. See
Figure 1 for a picture of the I-5 at the undercrossings, and Figure 2 for a profile of the alignment.
Working closely throughout design and construction with the Washington State Department of
Transportation (WSDOT), the project team implemented requirements to limit impacts to the freeway. As
part of these requirements, a robust and redundant geotechnical instrumentation system was set up for
monitoring movements throughout construction. After the first successful crossing of the freeway,
modifications to the system were made for the second crossing, which also proceeded without incident. This
paper discusses the background and requirements for the project, the geotechnical program developed during
the design phase, modifications to the program during construction, results of settlement monitoring during
the crossings, and recommendations for future crossings with similar conditions.
Figure 1. I-5 Freeway at Undercrossing (Looking North-Northeast)
Figure 2. Cross Section Below I-5 (Source: JCM)
PROJECT BACKGROUND
The U-Link Project provides an extension to the current system that runs in the Downtown Seattle
Transit Tunnel (DSTT). A previous contract created the Pine Street Stub Tunnel (PSST) beneath heavily
traveled Pine Street, both to provide a turnaround point for trains in the DSTT and to provide for future
extensions to the system. The U-Link Contract U230 extended the system from the PSST up to the Capitol
Hill Station, with twin tunnels running approximately 1,158 m (3,800 ft) to a new underground cut-and-cover
station.
As part of the alignment, the twin tunnels pass beneath the I-5 freeway. As part of the previous U215
Contract, large-diameter secant piles had to be removed from the path of the tunnel boring machines (TBMs).
During both design and construction, Sound Transit worked closely with the WSDOT, as well as designers
and the contractors, to provide a successful construction project. WSDOT set critical movement levels for the
retaining walls adjacent to the freeway, overpass structures in proximity to the tunnels, and the pavement
itself. A Yellow, or Threshold, Level was set at approximately 12.7 mm (0.5 in.), and a Red, or Maximum,
Level was set at approximately 25.4 mm (1 in.). Additional requirements for observations and repairs were
also set, including regular monitoring of both settlement and TBM performance, real-time data presentation
via a web-site, and emergency plans in case levels above were exceeded.
DESIGN DEVELOPMENT
A number of issues were carefully considered during the design phase, including requirements for tunneling
beneath the I-5, estimating of settlements and setting of action levels, and development of geotechnical
instrumentation requirements.
Tunneling requirements
During the design phase, requirements were set for TBM capabilities, operations, and maintenance prior
to and during the crossings. Primary requirements for the TBM to limit settlements during mining below the
I-5 freeway included the capability to provide positive face pressure to resist hydrostatic and earth pressures
(EPBM); and the ability to inject bentonite around the shield through ports to limit shield-related ground
losses. Operational provisions included 24-hour mining during the crossing to limit the number of prolonged
stoppages, and careful coordination between all parties and contingency plans during the crossings to prepare
for any foreseen and unforeseen circumstances.
As part of the previous U215 Contract, controlled density fill (CDF) blocks were created on either end
of the two TBM drives to replace removed material and stabilize the excavations. The blocks on the east side
of the crossings were required to be used to observe TBM and cutterhead conditions under free air prior to
the crossings, and to perform maintenance if needed. These inspection and maintenance stops were intended
to limit potential TBM issues and maintenance during the crossings themselves.
Estimated settlements and action levels
As part of the approval process from WSDOT for the crossings, estimates on the intended settlements
were performed during the design phase. The two crossings were far enough apart to limit overlap of
settlement profiles. However, the shallow alignments beneath the I-5 presented challenges for limiting
settlements to levels acceptable to WSDOT. Further complicating the crossing was a 1,065 m diameter (42
in.) storm drain beneath the main southbound travel lanes with only limited access for monitoring and
observing during construction.
A number of analyses were performed to estimate settlements, using a range of assumptions for
anticipated ground loss and settlement trough width. Results from these analyses showed estimated
settlements of between approximately 6.3 and 12.7 mm (0.25 and 0.5 in.). Two action levels were then set for
settlements based upon allowable limits set by WSDOT, as discussed above. A Trigger Level of 12.7 mm
was set, above which additional monitoring and consideration of additional contingency measures were
required. A Maximum Level of 25.4 mm (1 in.) was also set, above which TBM operations could be halted,
and the freeway itself shut down. Some post-construction verification of impacts to the freeway paving was
also required if either or both levels were exceeded, including repairs to the pavement surfaces.
Geotechnical instrumentation
Monitoring of movements during the crossings was critical to successful construction. The primary
goals of the instrumentation system were real-time monitoring (data available hourly) of movements to allow
contingency measures to be implemented; dissemination of the information to the team via a web-based
portal; redundancy of instrumentation to provide a check on information, and in case one system was not
working properly; and an accurate system as action levels were set at low values.
The previous U215 Contract required significant instrumentation to monitor movements during work
adjacent to the freeway. Most of the instrumentation was associated with the retaining walls on either side of
the freeway, as impacts of this contract were more focused on the modifications to the existing retaining wall
structures. Instrumentation consisted of linked beam sensors and tiltmeters installed on subject walls,
inclinometers, glass reflector targets installed on the retaining walls read by an Automated Total Station
(AMTS), and pulsed laser scanning of the walls with reflective laser targets.
While access to the I-5 limited the installation and use of extensometers within the actual freeway limits,
numerous wireless extensometers were installed along the alignments prior to the crossings. The intent was to
carefully monitor these extensometers to determine how well the TBM operations were controlling
movements. Wireless extensometers provide very quick feedback on movements and allow for adjustments to
TBM operations to be made in real time.
Similar to the requirements of the U215 Contract, laser scanning of the freeway pavement was part of
design requirements to monitor U230 pavement movements of I-5. This method was intended to provide
detailed movements across the area of concern, and could be compared to previous readings to determine
how movements were developing. As discussed later in this paper, issues with processing of data made it
very difficult to have real-time information using this method. Concerns about accuracy had also been also
raised during the U215 Contract.
As a redundant form of instrumentation for the crossing, horizontal in-place inclinometers (HIPI) were
installed between the tunnel alignment and the overlying pavements. These devices were installed during the
previous U215 Contract, and wiring was routed to locations that were accessible after completion of that
contract. A total of eight HIPIs were installed—two HIPIs above each tunnel installed from each side of the
freeway crossing.
Installed instrumentation
During U215 construction, extensive monitoring of instrumentation was implemented, and some of the
instruments installed during the U215 Contract were intended to monitor the U230 undercrossing as well.
The discussion here will focus on the systems used, and important lessons learned. During the U215 Contract
the eight arrays of HIPI’s were installed from within shaft excavation as construction proceeded. Each array
included six to eight 3 meter (10 foot) gage lengths, installed under the North HOV and South bound lanes of
I-5. HIPI’s were chosen as they could be installed in horizontal holes drilled form within the excavations,
thus minimizing disruption of traffic on I-5. These installations precluded installation of other systems, such
as vertical multiple position extensometers, as these would have required several days of closures on I-5 to
drill and install the vertical instruments. The HIPIs were intended to provide the primary measurement of
road movement. The instruments were installed and cables run through the CDF backfill in the U215
excavations, up to the data logger locations on the I-5 retaining walls. Running and protecting cables in CDF
blocks during shaft construction was a challenge. Access to the HIPI’s would be impossible, as they were to
be buried in the CDF backfill. This unique application required careful planning, reading verification and
system checking of instruments as no access to troubleshoot and/or check installations would be possible
after the CDF was placed. See Figures 3 and 4 for a photograph of HIPI installation.
Figure 3. HIPI Installation During U215
Figure 4. HIPI Casing Installation During U215 Contract, note temporary internal push rod and
counterweight for casing grouting operation.
As a redundant monitoring method, the U230 Contract intended to use pulsed laser scanning to provide
real time measurements of the road surfaces above the tunnel alignment to provide real time readings in road
surface changes. Laser scanning is a powerful tool, and provided meaningful results on the U215 Contract
with the use of scan targets for wall movements, but it became clear during use on U215 that this technology
would be unable to provide the real time measurement needs (less than 30 minutes) and measurement
accuracy of the project. The laser scanning option was replaced with the use of road prisms, consisting of
reflective glass prisms installed in a protective plastic housing and affixed to the road surface with hot melt
tar or epoxy. Road prisms were installed at the concrete pavement panel joints, and read with the AMTS
system provided for use in the U215 Contract. See Figures 5 and 6 for photographs of a typical AMTS
installation, and installed roam prisms, respectively.
For the U230 Contract, the first TBM crossing relied upon the HIPI, road prisms installed within the
HOV lanes, and reflectorless EDM pavement monitoring for the I-5 main southbound lanes. The second
TBM crossing relied upon the HIPI, and road prisms installed within the HOV lanes and within the I-5 main
southbound lanes.
Figure 5. Typical AMTS Installation
Figure 6. Road Prisms at Pavement Panel Corners
CONSTRUCTION AND MONITORING
Overall, construction of the tunnels below the freeway proceeded with no significant issues. However,
significant efforts went into planning and monitoring the crossings.
Planning for crossing
Prior to both crossings, multiple meetings were held by all parties involved, and responsibilities were
carefully assigned. Instrumentation and remote reporting of the instrumentation system was checked and
double checked, and personnel were assigned for remote monitoring during tunnel passage. A large part of
the success of the crossings was due to this coordination effort. Provisions that were developed during these
meetings included a calling tree with multiple individuals at each level; coordinating the start of excavation
with the nightly HOV lane closures to limit traffic on the roadway during the initial excavation; having a
clean-up team in close proximity in case of any pressure blow-outs of materials onto the roadway; having
individuals within the HOV lanes during the road closure to directly observe conditions; having traffic
control teams on stand-by if needed; and round-the-clock tracking of settlement data and TBM performance.
First crossing
For the first crossing (the Northbound track tunnel), continuous excavation took about 60 hours from
November 15th to November 18th, 2011, with mining of approximately 60 m (200 ft) proceeding from east to
west. TBM excavation began at approximately 9 p.m. in order to coincide with an early closure of the HOV
lanes, which provided an additional precaution for initial excavation. Excavation parameters from the TBM
were carefully monitored, and no unusual excavation volumes or excessive tail void grout volumes were
noted. At the median area between the HOV lanes and the main southbound travel lanes, a small open
conduit between the tunnel elevation and the ground surface allowed a limited amount of pressurized material
from the cutterhead to escape to the surface. This material was quickly cleaned up, and no subsequent issues
were observed.
Both the pavement and HIPI were carefully monitored throughout the crossing. Overall, the HOV road
prisms provided the most accurate and timely information, and movements appeared to track the TBM
progress well. See Figure 7 for a typical time-history plot of a road prism. The HIPI information for the HOV
lanes also appeared to provide good correlation with the pavement markers. See Figure 8 for typical
movements observed from one of these instruments. For the southbound main lanes, the HIPI showed results
within the normal operating range of the instruments, indicating no significant movement during TBM
passage. However, after years of proper function, and hours after the TBM completed, the system started
reporting suspicious movements along every sensor in the HIPI arrays, indicating linear movement/rotation.
While other available information seemed to indicate that no movements should be occurring, the post TBM
passage data created significant consternation amongst all parties. The next morning, a systematic analysis of
the measurement components related to the HIPI was conducted, and manual and electronic readings were
taken of the HIPI sensors, isolating the multiplexer and other measurement peripherals. This process
determined that a multiplexer had malfunctioned. Multiplexers are used with datalogging systems to allow
measurement expansion of multiple sensors to one logger resource. Normally these components work
flawlessly. Since the system was installed and measuring for several years, with readings that were within the
normal operating range if the system, this was the last place a problem was expected. With the new manual
HIPI readings, road prisms and manual surveys, movements were confirmed to be less than WSDOT
imposed levels. As a contingency measure, most of the rings below the southbound lanes were cored and
proof grouted. Cores indicated solid grout backfill between the segmental lining and the excavated ground,
and grout takes were low throughout. This coring and grouting further confirmed the TBM and manual
survey information.
Ground movements were compiled and plotted, with the most complete information coming from the
road prisms on the HOV lanes. Resultant movements were on the order of up to 7.6 mm (0.3 in.) throughout
the crossing, with movements dissipating from the centerline of the tunnel. Movements were less than the
Yellow Level limit set by WSDOT, and thus no additional actions were required.
Figure 7. First Crossing, Typical Road Prism Plot
Figure 8. First Crossing, Typical HIPI Deformation Plot
Second crossing
For the second crossing (the Southbound track tunnel), continuous excavation took about 67 hours from
April 23rd to April 26th, 2012, with mining of approximately 60 m (200 ft) proceeding from east to west.
TBM excavation began at approximately 11 p.m. in order to coincide with the normal closure of the HOV
lanes, which provided an additional precaution for initial excavation. Excavation parameters from the TBM
were carefully monitored, and no unusual excavation volumes or excessive tail void grout volumes were
noted. Data logger systems components were manually checked prior to the tunnel crossing to obtain manual
sensors measurements, and compared to automated readings. This minimized the potential for a system
component problem such as determined with the multiplexer on the first crossing. No similar issues with
multiplexers were detected.
Both the pavement and HIPI were carefully monitored throughout the second crossing, in a manner
similar to the first crossing. Overall, the HOV road prisms again provided the most accurate and timely
information, and movements appeared to track the TBM progress well. See Figure 9 for a typical time-history
plot of one of the road prisms. The HIPI information also appeared to provide good correlation with the road
prisms. See Figure 10 for a typical movement plot observed from one of these instruments. No issues
associated with the multiplexers occurred for the HIPI for the second crossing, mainly because of careful
examination of the units just prior to the crossing.
Ground movements were compiled and plotted, with the most complete information coming from the
road prisms. Resultant movements were on the order of up to 7.6 mm (0.3 in.) throughout the crossing,
similar to the first crossing, with movements dissipating from the centerline of the tunnel. Movements were
less than the Yellow Level limit set by WSDOT, and thus no additional actions were required.
Figure 9. Second Crossing, Typical Road Prism Deformation Plot
Figure 10. Second Crossing, Typical HIPI Deformation Plot
CONCLUSIONS AND RECOMMENDATIONS
Both tunnels were excavated with no significant impacts on the overlying I-5 Freeway. Movements did
not exceed any of the limits set by WSDOT for the U230 Contract. For data analysis and reduction, plotting
the data from both crossings on a cross section and using a best-fit Gaussian curve, a maximum resultant
ground loss of approximately 0.1 to 0.2% was estimated, and a settlement trough width factor of 0.35 to 0.4.
See Figure 11 for a plot for the HOV lanes for the first crossing, and Figures 12 and 13 for a plot of the HOV
lanes and the main southbound lanes for the second crossing, respectively. The data for the southbound lanes
for the first crossing were not considered to have the required accuracy for the development of reasonable
correlations.
Figure 11. First Crossing HOV Lanes Road Prism Data, Actual versus Predicted
Figure 12. Second Crossing HOV Lanes Road Prism Data, Actual versus Predicted
Figure 13. Second Crossing Main Southbound Lanes Road Prism Data, Actual versus Predicted
Based upon experience from this project, a number of recommendations are provided for future projects
with similar undercrossings of critical infrastructure:
• Throughout planning, design, and construction, closely work with impacted third parties to develop
realistic expectations and goals, including setting points of contact, critical settlement limits for
structures of concern, and contingency measures. Meetings at key times are also considered vital,
including during early development of the project alignment, at key stages during design, during
preconstruction meetings, and prior to the actual crossings.
• Develop detailed requirements for the construction team members, including assigning
responsibilities, developing calling trees (include back-up individuals), developing contingency plans
for potential occurrences, and making sure key information is being carefully observed and shared
within the team.
• For geotechnical instrumentation systems, the following items are recommended:
– Redundant monitoring systems;
– Innovative approaches, such as the road prisms used on the U230 Contract;
– Real-time monitoring using AMTS with reading schedules of every 30 to 60 minutes, and
download time to a web-based system of less than one hour;
– Multiple teams or individuals concurrently watching information and communicating observations.
– For long term (more than 6 month) applications requiring systems installed long before they are
needed, a full check of all components including manual redundant measurement capability should
be planned for, and executed within a time frame to execute changes if needed.
– Do not become complacent with automated readings that are within the normal operating range of
the sensor.
– Question readings that are valid, but make no sense, for example linear readings of movement
hours after events should have been detected.
• Provide for tie-in communications and data sharing between the TBM heading and the geotechnical
instrumentation.
ACKNOWLEDGMENTS
The success of this project depended upon numerous individuals throughout all the organizations
involved, and it would be difficult to list them all. Parties involved in this successful project include Sound
Transit staff on both the design and construction management side; the U230 Contractor JCM (Jay-Dee,
Michels Coluccio Joint Venture), including help with the undercrossing cross section; Geo-Instruments
(subcontractor responsible for instrumentation); WSDOT key personnel; the START CM team (Jacobs
Engineering and CH2M Hill Joint Venture); and members of Northlink Transit Partners (AECOM, HNTB,
and Jacobs Associates Joint Venture), who were closely involved in project design and monitoring.