8/12/2019 ARMA-06-1071_Ground Deformation and Structure Stability in Highly Stressed Rock Formations
1/16
1. PROJECT SETTING & SITE GEOLOGY
URS Corporation (URS) was retained to undertake
a detailed inspection, investigate existing ground
conditions and evaluate the structural condition ofunderground structures associated with a power
generating hydroelectric facility located on a riverbetween two sets of falls in upper New York state.
The underground structures comprised an intake
tunnel and ancillary structures, including the intakeshaft, the surge tube riser shaft (STRS), the tunnel
transition area (TTA), and the three penstocksincluding the penstock transition zones (PTZs). The
work included a geotechnical investigation and
performance of numerical and empirical analyses to
assess the structural stability of the aforementionedexisting structures. This paper details theassessment of the structural stability of the STRS
only.
The intake tunnel is located beneath a river
extending from the upper falls (higher elevation) to
the lower falls (lower elevation). The ground
surface elevations in the area range from
approximately elevation (EL) 450 feet at the top ofthe river gorge to EL 260 feet below the lower falls.
The elevation of the river varies from EL 392 feetabove the upper falls to EL 252 feet below the
lower falls. Between the two falls, the river is at
approximately EL 350 feet. The intake tunnelconveys water from the impoundment area (approx.
EL 391 feet) located above the upper falls, througha series of three penstocks and turbines, to
discharge points located at the base of the lower
falls. The total elevation head of water between theimpoundment area located above the upper falls and
the base of lower falls is 139 feet.
1.1. Bedrock FormationsThe rock formations exposed at the project siteinclude (from youngest to oldest) the Irondequoit
Limestone, Rockway Dolomite, Williamson Shale
(Upper Maplewood Shale & Lower Maplewood
ARMA/USRMS 06-1071
Ground Deformation and Structure Stability in
Highly Stressed Rock Formations
Paul HeadlandURS Corporation, Gaithersburg, MD, USA
Mohamed Younis
URS Corporation, Gaithersburg, MD, USA
Copyright 2005, ARMA, American Rock Mechanics Association
This paper was prepared for presentation at Golden Rocks 2006, The 41st U.S. Symposium on Rock Mechanics (USRMS): "50 Years of Rock Mechanics - Landmarks and FutureChallenges.", held in Golden, Colorado, June 17-21, 2006.
This paper was selected for presentation by a USRMS Program Committee following review of information contained in an abstract submitted earlier by the author(s). Contents of the paper,as presented, have not been reviewed by ARMA/USRMS and are subject to correction by the author(s). The material, as presented, does not necessarily reflect any position of USRMS,ARMA, their officers, or members. Electronic reproduction, distribution, or storage of any part of this paper for commercial purposes without the written consent of ARMA is prohibited.Permission to reproduce in print is restricted to an abstract of not more than 300 words; illustrations may not be copied. The abstract must contain conspicuous acknowledgement of whereand by whom the paper was presented.
ABSTRACT:The project site is situated in highly stressed rock formations located in upstate New York. The project locationand client name are confidential. The objective of the study was to investigate the stability of an 87-year old surge chamber shaft
embedded in a river gorge side slope with a slope height in excess of 150 feet. The shaft is concrete lined and approximately 22feet in diameter and 75 feet high. Based on a visual inspection, the concrete shaft structure displayed extensive cracking with a
crack pattern that appeared to be the result of high non-uniform stresses imposed by the surrounding rock formationswith possiblerock expansion and movement towards the gorge. The rock formations present at the site consist primarily of limestones,
sandstones, and shales. An engineering geological investigation was designed to study the in situ rock mass characteristics, in situstresses and ground behavior. The study included geological rock mapping (RMR & Q System), rock coring, geophysical
investigation, dilatometer in situ testing, and laboratory rock testing. A numerical model of the rock gorge was built to simulate
the rock stresses and behavior using two-dimensional Fast Lagrangian Analysis of Continua (FLAC) software. This paperpresents the geological study and the analysis. The analysis results obtained were found to concur with initial field observations.
8/12/2019 ARMA-06-1071_Ground Deformation and Structure Stability in Highly Stressed Rock Formations
2/16
Shale), Reynales Limestone (Wallington Limestone,Seneca Park Hematite & Brewer Dock Limestone),
Maplewood Shale, Kodak Sandstone, CambriaShale, Thorold Sandstone, Grimsby Sandstone,
Devils Hole Sandstone, and Queenston Shale.
Different formations exposed on the sidewalls ofthe river gorge exhibited various degrees of
weathering depending upon the nature of theformation.
The general stratigraphic profile (ground surface to
depth) present in the vicinity of the STRS on thesidewalls of the river valley adjacent to the
hydroelectric station is summarized in Table 1.
Table 1: Geologic Profile
Formation Depth from (EL ft) Depth to (EL ft) Thickness (ft)Glacial Till 460.0 443.5 16.5
Irondequoit Limestone 443.5 425.5 18.0
Rockway Dolomite 425.5 413.5 12.0
Williamson Shale(1) 413.5 386.5 27.0
Reynales Limestone(2) 386.5 366.5 20.0
Maplewood Shale 366.5 347.5 19.0
Kodak Sandstone 347.5 343.0 4.5
Cambria Shale 343.0 328.3 14.7
Thorold Sandstone 328.3 320.5 7.8
Grimsby Sandstone 320.5 288.2 32.3
Devils Hole Sandstone 288.2 284.0 4.2
Queenston Shale 284.0 260.0 24.0
(1)Williamson Shale formation comprises the Upper Williamson Shale (thickness = 10 ft) and the Lower Williamson Shale (thickness = 17 ft).(2) Reynales Limestone formation comprises Wallington Limestone (thickness = 16 ft), Seneca Park Hematite (thickness = 1 ft), and Brewer Dock
Limestone (thickness = 3 feet).
1.2. GroundwaterThe groundwater regime in the vicinity of the
hydroelectric facility appears to be primarily
controlled by the presence of the adjacent river.Based on historical water level information, it
appears that the observed levels correspond with the
adjacent water levels in the river.1.3. Tectonic SettingThe bedrock formations in the project area are
known to contain relatively high horizontal in situstresses. These stresses developed as a result of
historic large land mass movements. It is believed
that as the river gorge developed, and rock waseroded away, the horizontal stresses were forced to
concentrate in the strata below the river. It istherefore thought that the intake tunnel was
constructed in, and currently exists in, a relatively
highly stressed rock mass.
2. DETAILED INSPECTION
2.1. Tunnel InspectionURS inspected the underground structuresassociated with the hydroelectric facility between
July 2003 and November 2003. The inspection
culminated in a detailed assessment of the conditionof the tunnel and associated appurtenant structures,
including the Intake Shaft, STRS, Main Tunnel,
TTA, three PTZs, and three Penstocks. The
reported observations for the STRS are summarizedbelow.
Surge Tank Riser Shaft
The STRS is a 22-foot internal diameter concrete-
lined vertical shaft which extends from EL 370 atthe base of the surge tank to the crown of the intake
tunnel at approximately EL 289 (shaft length = 81feet). The center of the surge tank riser shaft is
located at STA -0+43 within the transition zone
between the tunnel and the penstocks, where thetunnel envelope widens immediately north of STA
0+00 of the tunnel alignment. The following keyobservations were made of the STRS shaft during
the inspection.
Shaft walls were generally in poor condition;
Shaft lining appeared to deteriorate considerablywith increasing depth below ground surface;
Structural cracks of varying size, extent, andfrequency were observed throughout the depthof the shaft;
Concrete spalling was observed to varyingdegrees along the entire depth of the shaft;
Minor seepages and limited staining wereobserved on the shaft lining walls;
8/12/2019 ARMA-06-1071_Ground Deformation and Structure Stability in Highly Stressed Rock Formations
3/16
Evidence of buckling (bulging) was observed atseveral locations within the shaft.
Buckling appeared to be more pronounced towardsthe base of the shaft, including one large area of
buckling (4 feet wide by 12 feet high) locatedapproximately 10 feet above the tunnel crown on
the northeast wall of the shaft.
3. FIELD INVESTIGATION & ROCKPROPERTIES
The URS field investigation was conducted in thefall of 2003. The investigation consisted of geologic
mapping, borehole drilling, in situ dilatometer
testing and laboratory testing of selected rock corespecimens. A summary of each of these field
activities is presented below.
3.1. Geologic MappingURS conducted detailed geologic mapping of
exposed sidewalls of the river gorge and collected
orientation and discontinuity condition data for each
of the geological formations observed. URSconducted geologic mapping of the exposurepresent along gorge sidewalls to evaluate rock mass
characteristics and structure in relation to the tunnel,
two shafts, and three penstocks associated with the
hydroelectric facility. The mapping evaluateddiscontinuity (bedding/joint) orientations,
conditions, and characteristics. Figure 1presents aprofile view of the pertinent geologic formations
URS identified during geological mapping
activities.
3.2. Borehole DrillingThe URS field investigation comprised two (2)
borings. Boring B-1 was completed in October 2003
and is located within the TTA below the STRS(approximately EL 268 feet). Boring B-2 was
completed in November 2003 and is located
approximately 70 feet to the northeast of the STRSat ground surface (approximately EL 460 feet).
Both borings are located to the east of a river. Thelocations of the two URS borings are shown on
Figure 2.
Figure 1: Geological Profile
8/12/2019 ARMA-06-1071_Ground Deformation and Structure Stability in Highly Stressed Rock Formations
4/16
Figure 2: Boring Locations
URS also conducted an extensive review of the
geotechnical data associated with these borings in
order to develop an overall rock mass
characterization and classification for the rockformations intersected by the structures associated
with the facility.
The URS field investigation proceeded as follows: B-1 was a vertical boring drilled and
continuously cored from the TTA invert
(approx. EL 268 feet) to a depth of 24 ft
below the invert (approx. EL 244 feet) of thetunnel lining immediately beneath the center
of the STRS using mud rotary drillingtechniques with continuous sampling
methods. Downhole testing consisted of four
rock dilatometer tests undertaken in situ toevaluate rock mass bulk modulus and in situ
stress magnitude. Upon completion, thisborehole was backfilled with grout and sealed
with concrete flush with the surface of the
tunnel with concrete. One anchor bolt 15 feetin length was installed into the boring at this
location and grouted in place per the clientsrequest.
B-2 was a vertical boring drilled from groundsurface (approximately EL 460 feet) and open
holed to 90 feet below ground surface
(approximately EL 370 feet) and then
continuously cored to a depth of 200 feet below
ground surface (approximately EL 260 feet).
Boring URS B-2 was located approximately 70feet to the northeast of the STRS using mud
rotary drilling techniques with continuoussampling. Downhole testing consisted of 11
rock dilatometer tests to evaluate rock mass
bulk modulus and in situ stresses. Uponcompletion, this borehole was backfilled with
grout from EL 260 feet to ground surface (EL460 feet).
Strength descriptions presented on the boring logs
are based on actual unconfined compressive
strength (UCS) laboratory testing results below theKodak Sandstone only. Above the KodakSandstone, the strength descriptions on the boring
logs are based on classification tests undertaken inthe field.
3.3. Dilatometer TestingDilatometer testing was undertaken at 15 locations
within the two borings completed as part of the
URS field investigation. Dilatometer tests wereperformed using the Probex 1 dilatometer system
manufactured by RocTest. The dilatometerapparatus consists of a probe, volume measurement
instrument, and a hydraulic pump with pressure
8/12/2019 ARMA-06-1071_Ground Deformation and Structure Stability in Highly Stressed Rock Formations
5/16
gage. The 70-mm or 2.76-inch (N size) diameterand 15-inch long cylindrical probes contain a high-
pressure expandable rubber membrane that isinflated with water during the test. During testing,
the volume change induced in the probe for each
pressure applied was measured using a linearvolume displacement transducer. Expansion of the
probe was controlled by applying hydraulic pressure
from a hand-operated hydraulic pump and pressuregage. The maximum pressure capacity of the
dilatometer system is 350 tons/ft2 (700 kips per
square foot [ksf]).
The dilatometer test data were processed and
computer generated test curves produced, followedby the calculation of geotechnical parameters
interpreted from the test curves. Each of the
geotechnical parameters derived from thedilatometer test curves are discussed below.
Elastic Deformation Moduli
The initial (E), reload (E+), and unload (E-) elastic
deformation moduli were calculated from the linearportions of the dilatometer test curves. The initial
moduli (E) for all test locations (DT1 through
DT15) ranged from 303 psi (B1-PM3) to 3,183 psi(B2-PM9). The reload moduli (E+) ranged from
2087 psi (B1-PM3) to 71,580 psi (B2-PM14). Theunload moduli (E-) ranged from 862 psi (B1-PM3)
to 56,380 psi (B2-PM7). The elastic moduli
parameters as determined from the dilatometertesting are summarized in Table 2.
Table 2 - Dilatometer Test Results Elastic Moduli Parameters
Boring Formation Tested
Test Depth
from(ft)
Test Depth
to(ft)
Initial
Modulus (E)ksi
Unload
Modulus (E-)ksi
Reload
Modulus (E+)ksi
B1-PM1 Queenston Shale(1) 5.375 6.625 (2) (2) (2)
B1-PM2 Queenston Shale(1) 9.375 10.625 (2) (2) (2)
B1-PM3 Queenston Shale(1) 14.375 15.625 303 2,087 862
B1-PM4 Queenston Shale(1) 19.225 20.475 2,280 56,620 29,860
B2-PM5 Maplewood Shale 95.375 96.625 741 2,067 1,692
B2-PM6 Maplewood Shale 106.375 107.625 740 4,106 3,057
B2-PM7 Kodak Sandstone 114.375 115.625 1,352 (2) 56,380
B2-PM8 Cambria Shale 126.375 127.625 1,579 11,930 7,896
B2-PM9 Thorold Sandstone 133.375 134.625 3,183 (2) (2)
B2-PM10 Grimsby Sandstone 144.375 145.625 1,889 10,960 10,350
B2-PM11 Grimsby Sandstone 152.875 154.125 1,413 9,266 6,313
B2-PM12 Grimsby Sandstone 168.375 169.625 1,529 12,460 6,663B2-PM13 Queenston Shale(1) 176.375 177.625 2,354 (2) 16,230
B2-PM14 Queenston Shale(1) 188.375 189.625 1,697 71,580 18,410
B2-PM15 Queenston Shale(1) 197.375 198.625 1,796 18,980 7,909(1)The Queenston Shale encountered in borings B-1 and B-2 was described as a SILTSTONE becoming shaley in parts.(2)
Data contained errors or produced unacceptable results and therefore was not used to provide geotechnical parameters.
In Situ Stress State
The in situ horizontal total stress (ho) was
determined from the dilatometer test results as the
stress (Po) corresponding to the initiation of the
linear elastic response. The vertical overburdenstress in boring B-2 was calculated using anestimated total unit weight for the rock as
determined from the laboratory test results. It
should be noted that the vertical stresses calculatedfrom the tests completed in boring B-1 have
assumed that the overburden stress at the tunnelsurface (start of boring) is equal to zero. However,
the stress distribution is vertically non-linear andwould be greater than the typical vertical stress
distribution (depth below ground surface multiplied
by the rock unit weight) due to the transfer anddistribution of in situ stresses within the rock mass
arching around the tunnel. The in situ stress
parameter results as determined from thedilatometer testing are summarized in Table 3.
8/12/2019 ARMA-06-1071_Ground Deformation and Structure Stability in Highly Stressed Rock Formations
6/16
Table 3 - Dilatometer Test Results In Situ Stress Parameters
Boring
Test Depth
from
(ft)
Test Depth
to
(ft)
Formation Tested
In Situ
Horizontal
Stress (ho)
psi
In Situ
Horizontal
Effective
Stress
(ho) psf
In Situ
Vertical
Effective
Stress
(v) psf
Coefficient
of Earth
Pressure at
Rest (Ko)
B1-PM1 5.375 6.625 Queenston Shale (1) (1) (1) (1)
B1-PM2 9.375 10.625 Queenston Shale 120 16,656 1,051 15.85
B1-PM3 14.375 15.625 Queenston Shale 130 17,784 1,576 11.28
B1-PM4 19.225 20.475 Queenston Shale 340 47,723 2,084 22.90B2-PM5 95.375 96.625 Maplewood Shale 220 31,680 16,003 1.98
B2-PM6 106.375 107.625 Maplewood Shale 210 30,240 17,837 1.70
B2-PM7 114.375 115.625 Kodak Sandstone 160 23,040 18,147 1.27
B2-PM8 126.375 127.625 Cambria Shale 200 28,800 20,447 1.41
B2-PM9 133.375 134.625 Thorold Sandstone 225 32,400 21,457 1.50
B2-PM10 144.375 145.625 Grimsby Sandstone 260 37,440 23,186 1.61
B2-PM11 152.875 154.125 Grimsby Sandstone 290 41,760 24,222 1.72
B2-PM12 168.375 169.625 Grimsby Sandstone 290 41,760 27,817 1.50
B2-PM13 176.375 177.625 Queenston Shale 260 37,440 29,644 1.26
B2-PM14 188.375 189.625 Queenston Shale 625 38,880 31,654 1.23
B2-PM15 197.375 198.625 Queenston Shale 290 41,760 33,161 1.26(1)
Data contained errors or produced unacceptable results and therefore was not used to provide geotechnical parameters.
The results of these tests are valid for the specific
materials and locations tested and are not to beconstrued to be representative of the entire geologic
unit present at the site. Variations in engineering
properties and differences in conditions are oftenencountered within each geologic unit.
The following observations can be made from the
dilatometer testing results.
The ratio of horizontal (ho) to vertical (v)
effective stress (Ko) increases significantlytowards the center of the river gorge. Ko is
significantly greater immediately beneath theTTZ (B-1) than in the ground mass as
determined from the dilatometer testingcompleted in B-2 located approximately 90 feet
to the northeast of the TTZ/STRS;
The values of Ko in boring B-1 varysignificantly (Ko range 11.28 to 22.90)
throughout the length of the formation tested(EL. 258 feet to EL 248.2 feet);
The values of Ko in boring B-2 are relativelyconstant (Korange 1.23 to 1.98) throughout thelength of the formation tested (EL 364 feet to
EL 262 feet);
The stiffest materials are the sandstone units;the less stiff materials are the shale units.
3.4. Laboratory TestingA total of 19 samples were selected for laboratory
rock testing. Twelve samples were selected forunconfined compressive strength testing (ASTM
D2938), three sample for Brazilian Split testing
(ASTM D3967), two samples for point load testing(ASTM D5731), and three samples for slake
durability testing (ASTM D4644).
Based on the results of the laboratory rock testing
on selected samples from borings B-1 and B-2 thefollowing observations can be made.
The unconfined compressive strength valuesranged from 10,137 psi (Grimsby Sandstone) to23,775 psi (Thorold Sandstone);
The unit weight values of the formations testedwere very consistent and ranged from 157.8 pcf
(Grimsby Sandstone) to 167.76 pcf (QueenstonShale);
The point load strength values ranged from 437psi (Maplewood Shale) to 19,043 psi (BrewerDock Limestone).
4. ROCK MASS STRUCTURE
Based on examination of the rock core collected
from borings B1 and B2 the intact rock mass
structure is generally considered to be sub-
horizontally bedded unweathered, moderately hardto hard, fine (e.g., shale and siltstone) to coarsegrained (e.g., sandstone), laminated (e.g., shale) to
8/12/2019 ARMA-06-1071_Ground Deformation and Structure Stability in Highly Stressed Rock Formations
7/16
thickly bedded (e.g., sandstone), highly (e.g., shale)to moderately fractured (e.g., sandstone)
sedimentary rock with unweathered andsmooth/planar bedding and unweathered
smooth/planar joints becoming slightly rough in
parts. The dip of the bedding ranges between 3 and4 and is classified as flat (dip of bedding = 0 to
20). The above description is based on generalized
observations of the intact rock mass.
4.1. Bedding CharacteristicsThe strike, dip and dip direction data of bedding
plane discontinuities within the rock mass located at
the hydroelectric facility was collected as part of theURS geological mapping completed as part of the
URS field investigation.
Bedding thicknesses are typically from 0.1 to 0.5inch (laminated) and 0.5 to 2.0 inch (very thinly
bedded) in the shale units (Williamson Shale andMaplewood Shale) and from 2 inches to 2 feet
(thinly bedded) and 2 feet to 3 feet (thickly bedded)
in dolomite, limestone, sandstone, and siltstoneunits. Based on examination of the rock core
collected from borings B1 and B2 the followinggeneral observations can be made regarding the
bedding of the formations present at hydroelectric
facility.
Bedding is sub horizontal (classified as flat);
Shale bedding thickness is generally between0.1 inch and 2.0 inches;
Limestone, dolomite, sandstone and siltstonebedding thickness is generally between 2 inches
and 3 feet;
Bedding roughness is generally smooth andplanar becoming slightly rough locally; and
Bedding planes are generally unweathered andtightly healed with no infill material.
Based on the information contained within theabove table, the average dip and dip direction of the
formations exposed adjacent to the hydroelectricfacility is 3 and 166 respectively. Discontinuity
data was not collected for the Williamson Shale and
Maplewood Shale units due to the weathered natureand access issues of these exposed in situ
formations.
4.2. Joint CharacteristicsThe strike, dip, and dip direction data of joint setdiscontinuities within the rock mass located at the
hydroelectric facility were collected as part of the
URS geological mapping. Using the full data setcollected for each formation the average strike, dip
and dip direction have been calculated. Onepredominant conjugate set of joints was noted
during the geological mapping.
Based on examination of the rock core collectedfrom borings B1 and B2, the following general
observations can be made regarding the jointing of
the formations present at the hydroelectric facility.
Jointing is sub-vertical (80 to 90);
Shale joint spacing is generally very closelyspaced (0.07 foot to 0.2 foot) to closely spaced(0.2 foot to 0.7 foot);
Limestone, dolomite, sandstone and siltstonejoint spacing is generally closely spaced
(0.2 foot to 0.7 foot) to widely spaced (2 feet to6.6 feet);
Joint roughness is generally smooth and planar,becoming slightly rough locally; and
Joints are generally unweathered and tightlyhealed with no infill material.
The average joint set discontinuity data(Discontinuity Set A & Discontinuity Set B) for
each formation are summarized in Table 4 below.
8/12/2019 ARMA-06-1071_Ground Deformation and Structure Stability in Highly Stressed Rock Formations
8/16
Table 4 - Average Joint Set Discontinuity Data
Discontinuity Set A (Average
Measurements)
Discontinuity Set B (Average
Measurements)
Formation Strike
Direction
()
Dip () Dip
Direction
()
Strike
Direction
()
Dip () Dip
Direction
()
Irondequoit Limestone 165/345 86 255 55/235 84 325
Rockway Dolomite 176/356 87 265 51/231 86 323
Williamson Shale 161 341
(1)
86
(1)
252
(1)
65/245
(1)
87
(1)
323
(2)
Wallington Limestone 146/326 85 238 80/260 88 170
Seneca Park Hematite 144/324 86 234 82/262 86 172
Brewer Dock Limestone 174/354 86 264 66/246 86 156
Maplewood Shale 172/352 (1) 86 (1) 264 (3) 72/252 (1) 86 (1) 156 (3)
Kodak Sandstone 170/350 87 80 78/258 86 348
Cambria Shale 163/343 87 253 33/213 85 123
Thorold Sandstone 158/338 86 251 29/209 87 210
Grimsby Formation 163/343 87 253 29/209 87 209
Devils Hole Sandstone 139/319 85 229 80/260 86 170
Queenston Shale 41/221 85 131 87/267 87 179(1)
Values estimated based on average overlying and underlying formation measurements. No credible shale measurements were made due to thedegree of weathering and inaccessibility of the formations in situ.(2)
Values estimated based the overlying Rockway Dolomite measurements due to significant disparity between dip direction of overlying andunderlying strata.(3)
Values estimated based the overlying Brewer Dock Limestone measurements due to significant disparity between dip direction of overlyingand underlying strata.
A random set of discontinuities was also noted forthe Brewer Dock Limestone, Rockaway Dolomite,
and the Irondequoit Limestone. These threeformations are located closest to ground surface.
4.3. FaultsNo faults intersect the project site based on the
findings of the field investigation, geological
mapping, and a review of all available geologicalpublications and project information supplied by the
client. The nearest known faults occur within theClarendon-Linden Fault Zone located in the
Allegheny Plateau physiographic province, which is
approximately 30 miles to the south of the projectsite.
5. ROCK MASS CLASSIFICATIONBieniawski (1989) [1] and Barton et al. (1974) [2 &
3] developed rock mass quality indices, namely theRock Mass Rating (RMR) system and the Rock
Mass Quality (Q-System) system, respectively, thatare widely used to classify rock quality and to
estimate tunnel support requirements. The Q-
System was developed primarily for classifyingmetamorphic rock mass quality. Rock mass
classification using the Q-System was undertaken
during the URS geological mapping. However, theQ-System classification results are not presented.
URS evaluated general rock mass quality based on
evaluation of rock mass conditions for eachgeologic formation as observed during URS
geological mapping and to a lesser extent based onevaluation of rock cores collected during the URS
field investigation.
It should be noted that the UCS values used for rockmass classification purposes are based upon
laboratory test data for the Kodak Sandstone and all
underlying formations. The UCS values for allformations above the Kodak Sandstone were based
upon visual observations made during the fieldinvestigation.
The RMR values were estimated predominantly
from weathered exposures and to a lesser extent the
individual core runs (B1 & B2), and therefore onlyapproximate the local site conditions that may be
encountered in situ. Differences may result from:encountering joint swarms, differing seepage
conditions, overbreak during construction (tunnels
and shafts) along bedding, and other conditions.
In general, rock mass classes determined from theRMR were Class II Good Rock to Class III Fair
Rock. The RMR ranged from 39, which equates toClass IV Poor Rock (Williamson Shale and
8/12/2019 ARMA-06-1071_Ground Deformation and Structure Stability in Highly Stressed Rock Formations
9/16
Maplewood Shale), to 82, which equates to Class I Very Good Rock (Devils Holes Sandstone, Kodak
Sandstone, and Seneca Park Hematite).
Comparing the ratings from the two systems (RMR& Q-System), it can be seen that the Williamson
Shale, Maplewood Shale and formations containing
shale layers/partings (Irondequoit Limestone,Rockway Dolomite, and Queenston Shale are
classified as fair rock with all the other rockformations (dolomite, limestone, siltstone, and
sandstone) being classified as good rock or better.
The only anomaly was the Kodak Sandstone, whichwas classified as fair rock according to the Q
System and good rock according to the RMRSystem.
6. METHOD OF STRUCTURAL ANALYSIS
URS conducted a combination of numerical
analyses, using Universal Distinct Element Code(UDEC) [4 & 5] and Structural Analysis and Design
Professional 2002 (STAAD.Pro 2002) [6 & 7]
software programs, and empirical analysis toinvestigate and asses the existing conditions of the
STRS. Below is a brief description of the analysismethods, numerical models, and empirical analyses
performed
6.1. Numerical Analysis Using Universal DistinctElement Code (UDEC)
An analysis was conducted using a two-dimensionaldistinct element program, UDEC, which was
developed specifically for modeling of jointed rockmasses. UDEC is a two-dimensional numerical
program based on the distinct element method for
discontinuum modeling. The discontinuousmedium is represented as an assemblage of discrete
blocks. The discontinuities are treated as boundaryconditions between blocks; large displacements
along discontinuities and rotations of blocks are
allowed. Individual blocks behave as either rigid or
deformable material. Deformable blocks aresubdivided into a mesh of finite-difference
elements, and each element responds according to aprescribed linear or non-linear stress-strain law. The
relative motion of the discontinuities is alsogoverned by linear or non-linear force-displacement
relations for movement in both the normal and shear
directions. UDEC has several built-in materialbehavior models, for both the intact blocks and the
discontinuities, which permit the simulation ofresponse representative of discontinuous geologic
or similar, materials. UDEC is based on aLagrangian calculation scheme that is well suited
to model the large movements and deformations ofa blocky system. A model was created using UDEC
to analyze the existing conditions and behavior of
the rock slope encompassing the STRS. Theanalysis results were input into a STAAD.Pro2002
three dimensional (3-D) model to analyze the
structural conditions and response of the STRS.
6.2. Numerical Analysis Using STAAD.Pro 2002STAAD.Pro 2002 is a structural analysis program
used to model structure response to external loads.
STAAD.Pro2002 uses different structural analysismethods such as finite element and finite difference.
The STAAD.Pro2002 analysis is purely an elasticanalysis. A model was created using STAAD.Pro
2002 to model STRS concrete lining response to
external loading from the surrounding rock
medium. The loading regime surrounding theSTRS was obtained from the UDEC Base Model.
6.3. Empirical MethodsEmpirical analysis methods including typical
geological in-situ stress distribution and rock loadsas determined using Terzaghis method for tunnelswere used to estimate loading regimes on the STRS
and the tunnel lining respectively. In addition,
internal stresses such as bending moment and axial
loads within the STRS lining were estimated basedon arbitrary distortion of the lining. Empirical
analysis was performed to estimate external loadingon the tunnel using typical lateral thrust from the
rock based on the coefficient of lateral earth
pressure obtained during URS field geotechnicalinvestigation. The loading regime was applied to
the tunnel to estimate bending moment and axialloads based on arbitrary distortion.
7. STRS ANALYSIS
The structural analysis undertaken for the STRSincluded a UDEC analysis to calculate themagnitude and distribution of the STRS, a
STAAD.Pro 2002 analysis to evaluate internal
loads in the STRS concrete lining, and an empiricalanalysis to estimate the effects of horizontal loading
on the STRS concrete lining. The purpose of theempirical analysis was to compare the computer-
based STAAD.Pro 2002 analysis with the hand-
calculated empirical analysis.
8/12/2019 ARMA-06-1071_Ground Deformation and Structure Stability in Highly Stressed Rock Formations
10/16
7.1. UDEC Base ModelThis model was developed to investigate themagnitude and distribution of stress regime
surrounding the STRS embedded within the gorgesidewall adjacent to the facility. The horizontal
stress regime data generated by the UDEC analysis
were used as input data for the STAAD.Pro 2002analysis.
The model geometry is 800 feet wide (x direction)by 450 feet high (y direction). Displacement
boundaries were located on the left, right, andbottom of the model. A surcharge load (1,875 lb/ft)
representing the glacial till was applied at the top
boundary of the rock in the model. The modelgeometry for Base Model is shown on Figure 3
(UDEC Base Model Geometry).
Figure 3: UDEC Base Model
For analysis purposes, selected geological
formations were grouped together based on their
similar characteristics and rock mass properties.
The modeled geological profile consisted of sixgeological layers including the Irondequoit
Limestone and Rockaway Dolomite (Layer 1) witha combined thickness of 30 feet, Williamson Shale
(Layer 2) with a thickness of 30 feet, Reynales
Limestone (Layer 3) with a thickness of 20 feet,
Maplewood Shale (Layer 4) with a thickness of 20feet, Kodak Sandstone, Cambria Shale, ThoroldSandstone, Grimsby Sandstone, and Devils Hole
Sandstone (Layer 5) with a combined thickness of
60 feet, and Queenston Shale (Layer 6) with athickness of 185 feet.
The rock mass properties included bulk modulus
(K), shear modulus (G), bulk unit weight (),
friction angle (), dilation angle (), cohesion (c),
and tensile strength (T) for each of the six geologic
layers. These properties are summarized in Table 5
below. In addition, joint properties including joint
normal stiffness (jkn), joint shear stiffness (jks),
joint friction angle (jfric), joint friction angle (jdil),and joint cohesion (jcoh) for each of the geologic
formations are presented in Table 6 below. Theseproperties were derived from the results of the fieldinvestigation, laboratory testing, geological
mapping, and available geological literature.
8/12/2019 ARMA-06-1071_Ground Deformation and Structure Stability in Highly Stressed Rock Formations
11/16
Table 5 - Formation Rock Mass Material Properties
Layer Formation NameK
(psf*)
G
(psf)
(pcf) c(psf) T(psf)1
Irondequoit Limestone,
Rockaway Dolomite1.5x108 6.9x107 170 10o 8.6x105 1.2x105
2Upper WilliamsonShale, Lower
Williamson Shale
9.0x107 4.2x107 165 10o 8.6x105 1.2 x105
3 Reynales Limestone 1.5x10
8
6.9x10
7
170 10
o
8.6x10
5
1.2x10
5
4 Maplewood Shale 8.9x107 6.9x107 165 10o 8.6x105 1.2x105
5
Kodak Sandstone,Cambria Shale, Thorold
Sandstone, GrimsbySandstone, Devils Hole
Sandstone
2.3x108 1.1x108 161 10o 1.2x106 1.9x105
6 Queenston Shale 1.9 x108 8.9 x107 167 10o 7.9x105 1.1x105* psf = pounds per square foot
Table 6 - Formation Joint Properties
Layer Formation Namejkn
(psf)
jks
(psf)jfric jdil
Jcoh
(psf)
1Irondequoit Limestone& Rockaway Dolomite
2.1x108 8.1x107 30o 5o 0
2Upper WilliamsonShale & Lower
Williamson Shale
1.3x108 5.0x107 20o 5o 0
3 Reynales Limestone 2.1x108 8.1x107 30o 5o 0
4 Maplewood Shale 1.3x108 4.8x107 30o 5o 0
5
Kodak Sandstone &Cambria Shale &
Thorold Sandstone &Grimsby Sandstone &
Devils Hole Sandstone
1.1x109 4.0x108 30o 5o 0
6 Queenston Shale 7.0x108 2.7 x10
8 25
o 5
o 0
The modeling approach included two stages. Thefirst stage represented the initial condition prior to
the formation of the river gorge. In the second,
stage, the gorge was excavated to generate theexisting in situ stresses at the project site, prior to
the construction of the tunnel and associatedstructure. The horizontal stresses at the locations
where the shaft would be located were obtained
from the model. A graphical representation
showing the magnitudes of these horizontal stresseswithin the STRS envelope is presented in Figure 4.
7.2. STAAD.Pro 2002 ModelThe 3-D STAAD.Pro2002 model was developed to
evaluate the response of the STRS concrete lining
under the stress regime as determined from theUDEC Base Model. The stress values from the
STAAD.Pro 2002 model at various elevations
within the STRS lining were then compared to thestrength (tensile/compressive) of the concrete.
The STRS lining was modeled as an 80-foot-high
(EL 290 to EL 370 ft) and 24-foot outside diameter(OD) cylinder. Please refer to Figure 4
(STAAD.Pro2002 STRS Model Geometry) for the
model geometry. The internal diameter of theSTRS lining was 22 feet with a concrete lining wall
thickness of 1 foot. The STAAD.Pro 2002 modelassumes fixed support at the bottom of the STRS
sidewalls.
8/12/2019 ARMA-06-1071_Ground Deformation and Structure Stability in Highly Stressed Rock Formations
12/16
Figure 4: Horizontal Stress Magnitudes in the STRS Envelope
The concrete liner properties included:
Compressive strength (fc),
Youngs modulus (E),
Poisons ratio ( ),
Unit weight (), and
Tensile strength (T).
These values were developed based on concrete
strength and thickness measurements made duringthe URS 2003 tunnel inspection.
The STAAD.Pro2002 model was used to calculate
hoop and longitudinal (vertical) stresses throughoutthe STRS concrete lining. The maximum tensile
stress in the longitudinal direction obtained was ashigh as 5,778 psi. It must be noted that this analysisis an elastic analysis and the tensile stress
magnitude obtained from the analysis is muchhigher than the concrete tensile strength. The
concrete tensile strength is typically on the order of
1/10 of the concrete compressive strength. Figure
5 shows the tensile stress contours overlaid onto the
crack distribution map as observed in the fieldduring the URS inspection. In addition, the
compressive stress in the longitudinal direction was
as high as 10,800 psi, which is much higher than the
compressive strength of the concrete measured in
the field (1,500 psi).
The hoop stresses obtained from the analysis ranged
from approximately 10 psi to 6,000 psi. Thesestresses are compressive in nature. The hoopstresses at four elevations of EL 300 feet, EL 315
feet, EL 325 feet, and EL 335 feet are presented inTable 7 to compare with the empirical analysis
results.
Table 7 - Hoop Stresses Acting on STRS Lining
(STAAD.Pro2002)
STAAD.Pro
2002 Scenario
Hoop
Stresses (psi)
Concrete fc
(psi)
EL. 300 6030 1500
EL. 315 3020 1500
EL. 325 1891 1500
EL. 335 762 1500fc compressive strength
It is evident from the results that at EL 300 feet, EL315 feet, and EL 325 feet the hoop stresses are in
excess of the compressive strength of the concrete
lining.
8/12/2019 ARMA-06-1071_Ground Deformation and Structure Stability in Highly Stressed Rock Formations
13/16
Figure 5: Tensile Stress Contours vs. Crack
Distribution
7.3. Empirical AnalysisEmpirical analyses were undertaken to evaluate themagnitude of hoop stresses acting upon the STRS
lining at EL 300 feet, EL 315 feet, EL 325 feet, and
EL 335 feet. The hoop stress values from theEmpirical analysis at various elevations within the
STRS lining were then compared to the strength
(tensile/ compressive) of the concrete. The results ofthe hoop stresses calculated from the empirical
analyses are presented in Table 9 below.Table 9 - Hoop Stresses Acting on STRS Lining
(Empirical)
Empirical
Scenario
Hoop Stresses
(psi)
Concrete fc
(psi)
EL. 300 1350 1500
EL. 315 1015 1500
EL. 325 830 1500
EL. 335 646 1500
It is evident from the results that at EL 300 feet, EL315 feet, EL 325, and EL 335 feet, the hoop stresses
exceed the compressive strength of the concretelining. These stress values do not represent any
distortion that would create bending moments
resulting in higher stresses in the STRS lining.
6.4 STAAD.Pro 2002 v Empirical
A comparison between the STAAD.Pro 2002analysis and empirical analysis is presented in the
Table 8. The hoop stress results obtained from theempirical method are less than those obtained from
STAAD.Pro 2002 analysis. It must be noted that
the empirical analysis has limitations. Theempirical analysis does not take into account rock
structure interaction and the non-uniformdistribution of stresses around the STRS.
Table 8 - Comparison of STAAD.Pro 2002 FOS v
Empirical Hoop Stresses
ScenarioHoop Stresses
STAAD.Pro2002
(psi)
Hoop Stresses
Empirical (psi)
EL 300 6030 1350
EL 315 3020 1015
EL 325 1891 830
EL 335 762 646
8. STRS ANALYTICAL RESULTS
The modeling of the STRS involved a two-stage
analytical process. Stage One (UDEC) comprised
modeling the rock slope and gorge to better estimate
the horizontal stress regime in place around theshaft structure. Stage Two (STAAD.Pro 2002)involved using the rock and field stresses from
Stage One in another model to identify the
distribution of the internal stresses within the STRSstructure. The models (Stage One amd Stage Two)
show a non-uniform stress distribution around theSTRS structure and a non-linear stress distribution
over the extent of the STRS structures height.
The presence of non-uniform horizontal and vertical
stresses around the STRS structure is reasonable
considering the proximity of the cliff face and thedistribution of relatively high horizontal stressaround the river gorge. It would be expected that
the net overall driving stress would be towards the
gorge emanating out of the cliff face (east to westdirection) as observed in the model.
Physical signs of this type of stress distribution
would be the outward displacement and intermittentinstability of the cliff face. Manifestation of this
behavior has been observed during the inspections
8/12/2019 ARMA-06-1071_Ground Deformation and Structure Stability in Highly Stressed Rock Formations
14/16
and remedial work on the PTZ, TTA, and STRS aswell as being evident from the observed debris at
the base of these slopes.
The non-linear increase in horizontal stress is alsoexpected in rock masses where there are
interbedded hard (limestone and sandstone) and soft
(shale) strata. Stronger strata are able to transmithigher stresses than weaker strata. The higher
stresses in these strata result in higher lateral stressvalues and therefore a higher ground load onto
tunnel and shaft linings from these strata as
compared to the softer strata.
The STRS model shows all of the intuitively correctresponses to the aforementioned conditions, and
therefore we feel that our analysis realisticallymodeled the response mechanism of the STRS
structure to the ground loading. The model and
subsequent calculations show that anticipated
current stress conditions are sufficient to crack thelining,and that when these cracks appear they occurin a pattern that closely resembles those observed in
the field.
The cracking observed in the field shows significantflexure and shear displacement, and this type of
cracking appeared in the model as a result of the net
overturning forces associated with gorge slopestability. It is likely that these stresses occurred
over a period of time, but because of the continuallydegrading slope, it is also anticipated that these
loads continue to be applied to the STRS. Thecomparison of modeled tensile stress distributionand actual observed cracks are presented on Figure
5 (Tensile Stress Distribution in STRS Lining).
The results of our analysis show that the majority of
deformation, cracking and loading of the STRS is
caused by high horizontal stresses in the hard strata
and the unbalancing forces of slope movement. Themass slope movement is continuous due toweathering of strata causing unloading of the slope.
The overall mechanism of what will become theprogressive failure of the STRS is that the internal
structural stresses increase over time until the liningcracks and re-distributes this load plastically around
the structure. The stresses then increase again as aresult of the continuing dynamic nature of the slope
instability and the internal forces until buckling of
the structure, loss of integrity and ultimately
collapse of the STRS take place.
The current condition observed in the STRS liningappears to have moved through the cracking phase
with buckling of the structure taking place. Theanalysis has shown clearly that the mechanism and
process of deformation identified above will
continue.
It has been argued with some reason that the STRScracking was caused by old loads perhaps even
loading during construction that have long sincedissipated and pose no threat to the stability of this
structure. The analysis in concert with empirical
calculations and observed evidence of the structureand slopes refutes this argument and makes a
compelling case that the loading that caused thecurrent level of structural distress to the STRS
remains a dynamic force on the structure and will
cause further damage to the existing STRS lining.
Further dynamic loading is added to the STRS by
the rapidly rising and falling internal waterpressures that are part of the normal function of thisstructure. The effects of changing water pressure
on the structure in its current condition are
significant. The turbulent water provides anotherchanging stress environment that can work
particularly at the bulge and crack location to erodethe weaker concrete material from this area. The
water will leak at this location, eroding the rock and
shale surrounding the shaft, and providing furtherbasis to assume a changing stress environment with
more asymmetric load conditions on the structureand a higher degree of slope movement due to
weathering and erosion.
Future work at the facility should carefully considerthe effect of surface work adjacent to the slope on
the underground structures and particularly the
STRS. For example, if the station building itselfwere ever to be demolished, there would be serious
potential consequences for slope stability and forthe STRS that should be considered fully before this
building is removed.
Limiting the current level of slope deterioration and
movement is a key part of our proposed strategy toobtain long-term stability of the structure. This
should be done in concert with rehabilitation of theSTRS structural lining.
Slope stabilization would be one component of a
comprehensive rehabilitation program. Slope
stabilization will prevent establishment of adynamic loading environment on the structure andthe rehabilitation can then be adequately designed
8/12/2019 ARMA-06-1071_Ground Deformation and Structure Stability in Highly Stressed Rock Formations
15/16
for a static load environment and long-termstability.
The evidence gathered to date by URS suggests
very strongly that the STRS lining should bestructurally repaired as a matter of urgency. A new
lining for this structure can be either steel or
concrete, but should be designed in accordance withexisting engineering practices to take no
consideration or contribution from the remainingstrength of the existing lining. It is also our
recommendation that the new lining should be
independent from the existing TTA structure. It isour recommendation that a structural ring beam be
constructed at the base of the STRS that will carrystresses from the TTA, and then a structural hinge
be made between this ring beam and the rest of the
STRS structure.
9. TTA ANALYTICAL RESULTS
It is evident from this observed cracking pattern that
significant stress transfer has taken place between
the STRS and the TTA structures. These cracksprovide further evidence that the proposed
mechanism of loading and displacement of theSTRS structure as modeled is sound.
Our modeling has not analyzed the interaction of
the TTA with the shaft, as this is a highly 3-D
problem and is not feasible to analyze using 2-Dmethods such as those described here. Our
objective in the analysis of the TTA was todetermine if there was any underlying structural
problem under the current loading conditions. We
removed considerations of the shaft structure andused our base model of the ground including the
cliff face to investigate the TTA structure.
The modeling results of the TTA show that thisstructure, when not influenced by additional shear
loading from the shaft structure, may have structural
issues of concern. The concrete strengths assumedin the model (500 and 1,000 psi) were exceeded by
the maximum internal stresses in the TTA structure.Concrete strengths within the structures likely vary,
and degradation of concrete strength is usually
somewhat patchy. However, it can be seen that therequired concrete strength for a 24-inch lining is
around 3,000 psi and it should be noted that weobserved several areas with strength measurements
lower than 3,000 psi.
The concrete strengths selected for our analysiswere considered a worst-case scenario based on our
field measurements and are obviously notrepresentative of the entire structure. This approach
was taken in the analysis as it is standard practice togenerally assess whether there is a possible
problem. Further investigation of the concrete can
quickly quantify the extent of our potential problem.
We recommend that the rehabilitation of the STRSprovide a structural hinge between the shaft and
the TTA so that the TTA can be treated as anindependent structure. This allows more flexibility
to consider the structures separately and, depending
on the results of further investigation, we canconsider non-structural repairs of the TTA as part of
an overall maintenance program.
10.CONCLUSIONS & RECOMMENDATIONS
10.1.STRSThe conclusions and recommendations pertaining tothe STRS structure based upon the results of theanalyses are listed below.
The UDEC base model provided horizontalstress information for further structural andempirical analysis of the STRS structure;
The STRS models indicated a consistentmechanism of ground loading on the STRS
structure;
The numerical analysis and subsequent
empirical calculations showed that theanticipated current load conditions are sufficient
to crack the lining and that the cracks formedclosely resemble those observed in the field
during the URS 2003 Inspection;
The continuous weathering and mass movementof the rock slope provides a significant portionof the overall ground loading and deformation
of the STRS structure;
The continuous weathering process indicates
that loading of the STRS structure will continueto increase over time and impose a load uponthe STRS;
Further dynamic loading on the structure isinduced by rapid changes in internal waterpressure;
Rehabilitation of the STRS structure shouldconsider the need to stabilize the slope and
prevent further weathering in the vicinity of the
facility;
8/12/2019 ARMA-06-1071_Ground Deformation and Structure Stability in Highly Stressed Rock Formations
16/16
Rehabilitation should consider the need tostructurally separate the STRS and TTAstructures and replace the lining with a newly
designed STRS lining in either steel or concrete;
and
The rehabilitation work described above shouldbe given the highest possible priority.
10.2.TTAThe conclusions and recommendations pertaining
to the TTA based on the results of the analysesspecific to the structural stability of the STRS are
listed below.
The crown of the TTA shows structuralcracking consistent with shear forces transferredfrom the observed deformation of the STRS;
The UDEC TTA model indicated that the TTA
has underlying stress-induced structural issuesdue to degraded concrete strength; and
Once the above rehabilitation of the STRS hasbeen carried out the TTA can be considered as aseparate structure.
REFERENCES
1. Bieniawski, Z.T., 1989, Engineering Rock MassClassification A Complete Manual for Engineers and
Geologists in Mining, Civil and Petroleum Engineering,John Wiley & Sons.
2. Barton, N., R. Lien, and J. Lunde, 1974, EngineeringClassification of Rock Masses for the Design of Tunnel
Support,Journal of the International Society for Rock
Mechanics, December 1974, Vol. 6 No. 4, pp. 189-236.
3. Barton, N., 2002, Some New Q-Value Correlations toAssist in Site Characterisation and Tunnel Design.International Journal of Rock Mechanics & Mining
Sciences39, 185216.
4. Itasca Consulting Group, Inc. (1995) Theory andBackground, First edition, January 2000, Minneapolis:
ICG.
5. Itasca Consulting Group, Inc. (1995) Users Guide,First edition, January 2000, Minneapolis: ICG.
6. Research Engineers, Intl, a Division of netGuru Inc.
STAAD.Pro 2002 Technical Reference Manual,April 2002, REI.
7. Research Engineers, Intl, a Division of netGuru, Inc.STAAD.Pro 2002 Software Release Report,
February 2002, REI.