Sei sulla pagina 1di 169

Technical Report Documentation Page

1. Report No.

A450, A464

2. Government Accession No.

3. Recipient's Catalog No.

4. Title and Subtitle

Analysis of Effects of Deep Braced Excavations on Adjacent Buried Utilities Richard J. Finno, Kristin M. Molnar, Edwin C. Rossow

5. Report Date

December, 2003

6 Performing Organization Code 8. Performing Organization Report No. 10. Work Unit No. (TRAIS) 11. Contract or Grant No.

7. Author/s

9. Performing Organization Name and Address

Department of Civil and Environmental Engineering Northwestern University 2145 Sheridan Road Evanston, IL 60208
12. Sponsoring Organization Name and Address

DTRS98-G-0016
13. Type of Report and Period Covered

U.S. Department of Transportation Research and Special Programs Administration 400 7th Street, SW Washington, DC 20590-0001
15. Supplementary Notes

Final Report, April 1, 2002 September 30, 2003


14. Sponsoring Agency Code

16. Abstract Ground movements resulting from deep braced excavations impose the risk of damage to adjacent buried pipelines. Accurate assessment of the effects these movements have on pipelines allows potential damage to be avoided or mitigated. A predictive process for determining the stresses occurring in a pipeline adjacent to deep braced excavations is presented. The method can be used to establish rational criteria for determining allowable maximum values for excavation-induced ground movements. The ground movement distribution around the excavated area is predicted using a complimentary error function, an assessment of the maximum ground deformation, and knowledge of the geometry of the excavation. The pipeline is assumed to move with the ground enabling the behavior of the pipeline to be represented by the ground surface movements at its location. Conservative analyses for determining the bending stresses and joint rotations along a pipeline caused by its deformation are established. Allowable values for both the tensile bending stress and joint rotation resulting from the excavation-induced movements are presented for comparison with the computed maximum values. The predictive methodology is applied to three gas mains surrounding a deep braced excavation in downtown Chicago and four cast iron mains from various excavations in Chicago. For these cases, the calculated bending stresses in the pipelines were significantly smaller than allowable values, but the joint rotations were observed to be the more critical case. 17. Key Words Excavation, Pipelines, Deformation, Seismicity, Infrastructure 18. Distribution Statement

No Restrictions

9. Security Classification (of this report)

Unclassified

20. Security Classification (of this page)

Unclassified

21. No. Of Pages 168

22. Price

ANALYSIS OF EFFECTS OF DEEP BRACED EXCAVATIONS ON ADJACENT BURIED UTILITIES

By Kristin M. Molnar Richard J. Finno Edwin C. Rossow

By School of Civil and Environmental Engineering Northwestern University Evanston, IL December, 2003

ABSTRACT Ground movements resulting from deep braced excavations impose the risk of damage to adjacent buried pipelines. Accurate assessment of the effects these movements have on pipelines allows potential damage to be avoided or mitigated. A predictive process for determining the stresses occurring in a pipeline adjacent to deep braced excavations is presented. The method can be used to establish rational criteria for determining allowable maximum values for excavation-induced ground movements. The ground movement distribution around the excavated area is predicted using a complimentary error function, an assessment of the maximum ground deformation, and knowledge of the geometry of the excavation. The pipeline is assumed to move with the ground enabling the behavior of the pipeline to be represented by the ground surface movements at its location. Conservative analyses for determining the bending stresses and joint rotations along a pipeline caused by its deformation are established. Allowable values for both the tensile bending stress and joint rotation resulting from the excavation-induced movements are presented for comparison with the computed maximum values. The predictive methodology is applied to three gas mains surrounding a deep braced excavation in downtown Chicago and four cast iron mains from various excavations in Chicago. For these cases, the calculated bending stresses in the pipelines were significantly smaller than allowable values, but the joint rotations were observed to be the more critical case. Excessive leakage was observed at a rotation of 6 x 10-3 radians for a cast iron pipeline with lead caulked joints.

ACKNOWLEDGEMENTS A number of people and organizations were instrumental in providing the data from the Lurie Research Center project that form the basis of this work. Inclinometer data were obtained by Construction Testing & Instruments, Inc., and Professionals Associated obtained the vertical and lateral survey data. Turner Construction Company was the general contractor and Case Foundation Company was the excavation support subcontractor. The help and interest of Dr. Jerry Parola and Ms. Dhooli Raj of Case and Mr. Ron McAllister of Turner made this work possible. Mr. John Brzezinski and Ms. Jo LeMieux-Murphy of the Facility Management group at Northwestern provided access to the project and were very helpful throughout its duration. Northwestern University students who generously gave their time to assist in the field monitoring effort included Frank Voss, Tanner Blackburn, and Terry Holman. Jill Roboski of Northwestern University developed the method to estimate surface settlement profiles along lines parallel to an excavation. The authors also thank Professor Thomas ORourke from Cornell University who generously provided us with a number of reports he authored related to pipelines.

ii

TABLE OF CONTENTS Abstract Acknowledgements Table of Contents List of Tables List of Figures List of Symbols i ii iii iv v vi

INTRODUCTION BACKGROUND Pipe Material Joints Initial Stresses ANALYSIS OF PIPES Approach Bending Stresses Joint Rotations Allowable Stresses and Joint Rotations PREDICTION OF EXCAVATION-INDUCED MOVEMENTS Summary of Procedure CASE STUDIES Lurie Center Chicago Excavation Case Studies CONCLUSIONS REFERENCES TABLES FIGURES APPENDIX A: LURIE CENTER DATA

1 2 2 4 5 6 6 8 11 12 15 15 16 16 19 21 23 26 32 40

iii

LIST OF TABLES Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 Engineering Properties for Piping Materials Failure Rotations for Selected Cast Iron and Ductile Iron Joints Allowable Bending Stresses from Excavation-Induced Movements Allowable Joint Rotations for Cast Iron and Ductile Iron Joints Strength Reductions at Location of Line Pipe Welded Joints Description of Cast Iron Pipelines Parallel to Chicago Excavations 26 27 28 29 30 31

iv

LIST OF FIGURES Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Schematic of Joint Rotation for Joint j of Rigid Pipeline General Layout of Lurie Center Site Instrumentation and Adjacent Gas Mains Comparison of Ground and Pipeline Movement during Excavation at Lurie Center Ground Displacements at Location of Gas Mains along North, South and West Walls after Completion of Excavation Maximum Tensile Stress in Pipelines Adjacent to North, South and West Wall During Excavation Maximum Relative Rotation Encountered Along North, South, and West Pipelines During Excavation for 3.6m Pipe Sections Observed and Predicted Maximum Bending Stresses and Joint Rotations for Final Stages of Construction Rotations in Cast Iron Pipelines Adjacent to Excavations in Chicago 32 33 34 35 36 37 38 39

LIST OF SYMBOLS A0, A2, B2 coefficients for no slippage between pipe and soil equations

A0*, A2*, B2* coefficients for slippage between pipe and soil equations B C dc dc/ds di dl do ds dw E EA EI Fu Fy h H HDB I j dimensionless constant related to lateral stress ratio dimensionless constant related to lateral stress ratio thickness of caulking in joint rotation to cause metal binding failure in lead caulked cast iron joints inner diameter of pipe cross section depth of lead outer diameter of pipe cross section depth of bell depth of packing material distance between rubber gasket and end of pipe modulus of elasticity circumferential extensional stiffness per unit length circumferential bending stiffness per unit length ultimate stress yield stress relative distance from data point to node of grid for radial basis interpolation depth of pipe hydrostatic design basis value moment of inertia of cross section stress exponent

vi

K L Lji m M Ms Mt Mx Mz n N

lateral stress ratio length of diagonal of extent of data range for radial basis interpolation original length distance along pipeline between points i and j (i,j = 1, 2, 3, n) modulus number moment Newmark coefficient constrained modulus of soil tangent modulus of soil moment around x-axis moment around z-axis number of survey points along pipeline number of data points in radial basis interpolation thrust Boussinesq coefficient internal pressure interaction load at interface of pipe and soil uniform pressure mean radius of pipe horizontal distance from top of pipe to application of load smoothing factor for radial basis interpolation wall thickness interaction shear at interface of pipe and soil extensional flexibility ratio bending flexibility ratio

p Pr q r R R2 t Tr UF VF

vii

w W xi x(Yi) Xi Yi z zi z(Yi) Zi i L T ji ALLOW i imax L max

curvature of deflection curve w concentrated wheel load distance to extreme fiber at point i under Mz (i = 1, 2, 3, n) curvature of lateral movement profile at point i (i = 1, 2, 3, n) total lateral movement at survey point i (i = 1, 2, 3, n) distance from origin to survey point i along pipeline (i = 1, 2, 3, n) distance from centroidal axis of deflection curve distance to extreme fiber at point i under Mx (i = 1, 2, 3, n) curvature of vertical movement profile at point i (i = 1, 2, 3, n) total vertical movement at survey point i (i = 1, 2, 3, n) coefficient of thermal expansion relative rotation between two adjacent pipe sections at joint i change in length change in temperature strain differential lateral movement between points i and j (i,j = 1, 2, 3, n) Poissons ratio angle from horizontal of pipe cross section allowable joint rotation from excavation-induced ground movements angle from horizontal for cross section of pipe at point i (i = 1, 2, 3, n) angle of principal plane of pipe cross section at point i (i = 1, 2, 3, n) rotation to cause excessive leakage in lead caulked joint for analysis rotation to cause metal binding failure in rubber gasket iron joints

viii

M ji ALLOW B H i INITIAL imax r v

rotation to cause metal binding failure in rubber gasket joints for analysis radius of curvature of centroidal axis of deflection curve differential vertical movement between points i and j (i,j = 1, 2, 3, n) effective stress of soil allowable stress from excavation-induced ground movements design bending stress of pipe material hoop stress (circumferential stress) bending stress at point i (i = 1, 2, 3, n) initial stress in pipeline prior to stress analysis maximum longitudinal bending stress in pipe cross section at point I reference stress (atmospheric pressure = 100kPa = 1atm) vertical soil stress

ix

INTRODUCTION Buried pipelines within urban environments are exposed to large ground movements as the result of adjacent excavations. These excavations may include tunneling, trenching, or deep open cuts. The ground movements resulting from these conditions impose stresses on the pipeline that could lead to its failure. Deep braced excavations are known to cause significant ground movements at great distances away. Most commonly buried utilities are exposed to ground movements from trench construction for the installation or repair of an adjacent pipeline. Much work has been done in this area to determine the patterns of movement caused by trench construction and the stresses encountered in the pipeline. Field experiments (Carder, et al. 1982, Carder and Taylor 1983 and ORourke and Kumbhojkar 1984), beam on elastic foundation analysis (Crofts, et al. 1977 and Tarzi, et al. 1979), and finite element simulation (Nath 1983 and Ahmed, et al. 1985) have been performed to determine a pattern of behavior for ground movements from trench construction and their effects on adjacent pipelines. Information regarding the effects of pipelines adjacent to deep braced excavations is very limited. Methods for predicting the magnitude and location of maximum ground movements around an excavation have been determined from field observations (Clough and ORourke 1990 and Hsieh and Ou 1998). Maynard and ORourke (1977) report field observations for four cast iron gas mains in Chicago exposed to ground movements from neighboring deep excavations. They reported excessive leakage from one of the mains making it necessary for it to be taken out of operation. If excavation-induced ground movements are sufficiently large, imposed stresses on a pipeline can cause failure. One therefore needs to determine the effects of an excavation on a

pipeline prior to construction so that rational criteria for preventing damage can be included in the design of the excavation support system. A method for predicting the longitudinal bending stresses and joint rotations incurred by a pipeline from ground movements resulting from an adjacent excavation is presented. The complimentary error function (Roboski and Finno, 2004) is presented for producing a distribution of movements parallel to an excavation wall based on the excavation geometry. The ground distributions are imposed on the surrounding pipelines. Equations are derived for bending stress and joint rotation analyses to determine the distribution of stresses along the pipeline. Distributions of the bending stresses and joint rotations are determined by the analysis. The maximum values can be determined and compared to allowable values determined from previous experimental and empirical observations. This approach provides a rational method of establishing excavation-induced ground movement limits when surrounding utilities are the critical structure impacted by the excavation.

BACKGROUND Pipe Materials The most common metallic materials found in pipelines in urban areas are cast iron, ductile iron and steel. The more modern installations use plastics because of their flexibility. Polyethylene is a popular material for buried utility installations for gas and water transportation. General engineering properties for these materials are presented in Table 1. Cast iron is the oldest metal found in pipelines, and is relatively common since many pipelines installed over 100 years ago are still in operation. Cast iron pipe was formed by pit casting, the predominant process until the 1930s when the centrifugal cast process was developed. Of the two manufacturing processes, centrifugal cast pipe has a greater strength due

to a better distribution of graphite flakes within the iron matrix. The tensile stress-strain behavior of both types of cast iron exhibits irrecoverable deformations at low strains. There is no apparent yield point, and brittle failure occurs at relatively low strain values. Cast iron pipe was the dominant pipe material until the 1950s when ductile iron was introduced. The metallurgy of the materials is very similar, but ductile iron has increased strength and ductility from carbon that exists as small spheroids. The presence of the carbon in this form introduces fewer discontinuities into the matrix. Ductile iron also does not exhibit a yield point. Angus (1976) observed a true elastic range of stress values in which irrecoverable deformation did not occur. When the material begins to plastically deform, the graphite nodules do not deform with the matrix and the useful cross sectional area is reduced which in turn reduces the apparent pipe stiffness. At the end of the 19th century, steel, or line, pipe was implemented into the transport of gas and oil. Pipe sections are manufactured as welded or seamless. Welded pipe has two halves of pipe longitudinally welded together. Seamless pipe contains no welds and is produced by either hot piercing or cupping and drawing. Steel differs from cast iron and ductile iron in that it exhibits a yield point at the end of a region of linear elastic behavior, with a well-defined modulus elasticity. The stress-strain behavior is dependent on the carbon content within the alloy. The yield point becomes less sharp, the yield plateau becomes less prominent, and the ultimate stress increases as the carbon content increases. In the more recent installations of pipelines, plastics have been used due to the great lengths of pipe available and its flexibility and durability. Plastics are solid materials with one or more polymeric substances that can be formed by flow (Plastics Pipe Institute, 1993). Common

plastics used in piping are polyvinyl chloride (PVC) and polyethylene (PE). Polyethylene pipe is commonly used for the transportation of water and gas. The behavior of polyethylene under stress is complex due to its viscoelastic properties and its dependency on the load duration, environment, and temperature. The shape of the stress-strain curve is highly dependent on the load duration due to its material rate-dependent behavior. The Plastics Pipe Institute (1993) define a flexural, short-term, and long-term modulus for polyethylene dependent on the nature of the load and the rate at which the load is applied.

Joints The methods of joining pipe sections also have steadily improved over time allowing greater rotations without attendant loss of service. Joining mechanisms may be used to allow rotations or the pipe sections may be joined to one another directly. Table 2 show typical failure rotations for cast iron and ductile iron joints. Early installations of cast iron pipe were joined with rigid connections and were constructed with metal-to-metal contact within a bell and spigot connection to prevent leaks, while allowing very little rotation. Semi-rigid joints were the more predominant mode of joining cast iron pipe and were constructed by adding a packing and caulking material within a bell and spigot connection. The packing material was soft, usually jute or yarn, to permit a certain amount of flexibility within the joint. The caulking held the packing in the joint and, for most installations, was lead. These joints were able to withstand small rotations before failure was observed. Joints allowing exhibiting flexible behavior within a cast iron pipeline are constructed by adding a rubber gasket to eliminate leaks; these joints allow greater rotations without loss of service. Rubber gaskets are used in push-on and mechanical joints.

Ductile iron pipe is predominantly joined with rubber gasket joints similar to that of cast iron pipes. There are rubber gasket push-on joints and bolted-gland mechanical joints. For more adverse conditions there is a ball and socket joint, which is free to rotate up to 0.27 radians (15 degrees). Rubber gasket joints are available for steel pipe as well, but the more common method of joining steel pipe sections is by welding. Lap, single-butt, and double-butt welds are the more common types of joining welds and produce a joint with strength very similar to that of the strength of steel. Of the three types of welded joints, the single-welded lap joint introduces the greatest reduction in the strength of the steel of approximately 25 percent. Similar to that of welding steel pipe, polyethylene is joined by fusion, wherein two ends of adjacent pipes are melted to a fluid-like state and then forced together to join a continuous section after cooling. Melting of the ends of the pipe can be from direct heat from a hot surface or a coiled wire. The fusion joint is equal in strength to that of the rest of the polyethylene pipe.

Initial Stresses The installation of a buried pipeline introduces stresses within the pipe upon which are added additional stresses caused by movements associated with excavations. These stresses can be a result of any combination of an operating internal pressure, the soil cover load, cyclic or static surface loads, the installation procedure, previous ground movements, or environmental effects. Taki and ORourke (1983) analyzed the effects internal pressure, thermal fluctuations, repeated loading, and installation procedures on cast iron mains. They determined typical amounts of tensile or bending strain induced by these conditions for low pressure pipelines.

From these calculations, they suggest assuming that a buried pipeline has an initial bending strain value between 0.02 to 0.04 percent. Internal pressures can cause a circumferential tensile stress due to the imbalance of interior and exterior pressures (Timeshenko 1951). The stresses induced by the soil cover (e.g. Carder et al. 1982; Carder and Taylor 1983) and static or cyclic loads (e.g. Pocock et al. 1980) can cause ovalling of the pipe cross section with attendant stresses that vary around the pipe. The installation of pipelines in the ground can create a stress or joint rotation from uneven bedding or a curved laying pattern (e.g. Pocock et al. 1980). Prior to an adjacent construction, the previous construction history may have resulted in movements of the pipeline causing bending stresses and joint rotations (e.g. Maynard and ORourke 1977). The environment in which the pipeline is buried can cause different stresses to be imposed. A pipeline installed in a location with fluctuations in temperature can cause strains in the pipeline (Attewell 1986). Moisture changes in the soil surrounding the pipeline can cause corrosion to occur which could weaken the strength of the pipe walls (e.g. Sears 1986).

ANALYSIS OF PIPES Approach Pipelines parallel to deep excavations undergo deformations due to the displacement of the surrounding soil. For most utilities that parallel a large excavation in an urban environment, the pipeline can be assumed to move with the soil. Carder, et al. (1982) and Carder and Taylor (1983) conducted field experiments with 100 mm diameter cast iron pipelines buried 0.75 m deep parallel to trench excavations in different types of soil. They determined that the movements of the pipelines calculated from the data obtained from strain gauges along the pipeline were similar to that of the ground displacements observed. Nath (1983) conducted a

three-dimensional finite element analysis on cast iron pipelines ranging from 75 to 450 mm in diameter buried 1 m deep parallel to open trench excavations of differing dimensions. He concluded that pipelines with diameters of 150 mm or less move with the ground providing little or no restraint to movements, whereas larger pipes provided some restraint against the movement of the surrounding soil. For pipelines whose movements are consistent with the displacements of the surrounding soil, one can make assumptions concerning joint flexibility to analyze the effects of ground movements on the pipe. By assuming a pipe is either flexible, wherein it is assumed a pipe connection does not affect the mechanical behavior of the pipe, or rigid, where all the movement is assumed to occur at a joint, one can bound the response of the pipe to the imposed deformations. The question becomes defining the critical condition, either excessive bending stresses for the flexible condition or large rotation at a joint for the rigid condition, possibly leading to excessive leakage or fracture at a joint. Special care should be taken when applying this approach to larger diameter metal pipelines because they tend to restrain movements more than that of small diameter pipelines, and the restraint provided by the pipe will result in higher stresses. Flexible pipe is assumed herein to move along with the ground causing bending within the pipe sections and no rotation at the joints. With the introduction of more ductile materials, bending stresses have become less of a concern; however, this is still of great concern for the old cast iron mains that can fail as a brittle fracture at a low strain, or in cases of pipes subjected to high pressures. The displacements of the pipe sections are assumed to be small allowing for axial displacements to be neglected. The deflection within a pipe is assumed to follow the Bernoulli-Navier theory of bending wherein the plane normal cross sections remain plane and

perpendicular to the deflected centroidal axis and the transverse normal stresses are negligible (Baant and Cedolin, 1991). Rigid pipe is assumed herein to deform along with the ground displacement profiles as rigid links connected by points that are free to rotate. The effects of the ground movements on the pipe are concentrated in the joints as relative rotations between adjacent pipe sections. The pipe sections are assumed to have a large flexural rigidity thus preventing any curvature to develop. The joints are assumed to have no rotational rigidity allowing free rotation. The rotation at the joints is assumed to be longitudinal due to bending of the pipeline, therefore torsional behavior is neglected.

Bending Stresses Flexible pipes exposed to ground movements develop bending stresses within the pipe sections. The bending stresses obtained from this analysis should be compared to established allowable values to determine the structural adequacy of the pipeline. If there are high internal pressures, a Mohr circle analysis should be made to find the maximum of the combined stresses. A displacement profile of the pipeline should be established within a convenient local coordinate system, as shown in Figure 1a. The origin is situated adjacent to the corner of the excavation, the positive x-axis represents the lateral movement towards the excavation, the positive y-axis is the longitudinal axis of the pipeline, and the positive z-axis is directed upwards. Following the assumption that the pipeline moves along with the ground, the lateral and vertical ground surface movements at the location of the pipeline determine the lateral and vertical displacement profiles of the pipeline. Bending is of greatest concern at the location of the displacement profiles where the curvature is the largest. The curvature is calculated at a point j from the lateral and vertical

displacement profiles causing a differential movement of point j with respect to two adjacent points along the pipeline by the following equations:

(Z Z j ) (Z j Zi ) 2 k ( ) (Yj - Yi ) Y Y k j = z" (Yj ) = (Yk Yi )

(X X j ) (X j X i ) 2 k (Yk - Yj ) (Yj - Yi ) x" (Yj ) = = (Yk Yi )

2 kj ji L kj L ji L ki

(1)

2 kj ji L kj L ji L ki

(2)

where ji = Xj-Xi is the differential lateral movement between points i and j (i,j = 1,2,3,..n), Lji = Yj-Yi is a characteristic length defined as the distance along pipe between points i and j, and ji = Zj-Zi is the differential settlement between points i and j. The displacements are defined in the local coordinate system as Xi and Zi representing the total lateral and vertical movements at point i, respectively, and Yi is defined as the distance from the origin along the pipeline to point i. In defining a characteristic length, Lji, for the curvature calculations, the distance should be large enough to reasonably reflect the curvature in the pipeline. A characteristic length of approximately 6.1 m or greater has shown to give an adequate representation of the curvature values for a pipeline undergoing bending. Treating the lateral and vertical profiles separately allows a simple three-dimensional analysis for determining the principal planes of the pipe cross sections. This provides a more accurate calculation of the maximum stresses along the pipeline than calculations from the resultant movements due to the continual change of the angle of the principal planes along the pipeline. For a pipe in bending there is a distribution of tensile and compressive stresses within the cross section. The maximum tensile stress existing in the pipe is critical due to the greater

strength of many pipe materials in compression. The distribution of normal stresses, i, within the pipe cross section exposed to both lateral movements in the x-direction and vertical movements in the z-direction, assuming the pipeline behaves as an elastic beam, is found by:

i =

M z x i M x zi I I

(3)

in which Mx and Mz are the moments around the x- and z-axis, respectively, with the x-axis horizontal and the z-axis is vertical, and I is the moment of inertia about the neutral axes of the cross section of the pipe. The terms xi and zi represent the distance to the most extreme fiber with respect to the moment. Substituting the expression for bending moment with small deflections, M=EIw", the equation for stress at a point i along the pipeline becomes:

i = Ex i z" (Yi ) Ez i x" (Yi )

(4)

where x"(Yi) and z"(Yi) are the curvatures of the lateral and vertical displacement profiles at point i, respectively. To express equation (4) in terms of the pipe radius, r, and the angle from the positive x-axis, one can write:

i = Er[z" (Yi )sin i + x" (Yi )cos i ]


To calculate the maximum tensile stress within the cross section, this expression has to be

(5)

maximized. By taking the derivative of the stress with respect , the expression for the angle of the principal plane can be found as:
z" (Yi ) x" (Yi )

imax = tan 1

(6)

The value of the maximum tensile stress is determined by substituting the results of (6) into (5).

10

Joint Rotations

It is assumed in this limiting case that rigid pipelines conform to the ground movements through rotations at the joints due to the infinite stiffness of the pipe sections. To determine if there is failure at a joint, the relative rotation between the two adjacent pipe sections needs to be calculated. Since the pipeline is assumed to follow the ground movements, the joints are located along the displacement profile at distances equal to the length of the pipe sections. The differential movements, as defined earlier, need to be determined to determine the change in slope along the pipeline at the joint. The rotation at a joint can be calculated using vector mechanics. The pipe sections can be represented as vectors that intersect at the joint. The angle between them can be calculated if the differential movements are known. Figure 1b shows the vector representation of two adjacent pipe sections along a pipeline that have undergone both lateral and vertical movements relative to joint j. The differential lateral and vertical displacements, ji and ji, are as previously defined for the bending stress analysis. The characteristic length, Lji, is defined by the pipe section length of the pipeline. The majority of cast iron mains consist of pies 3.6 m in length. Ductile iron pipe sections range in length of 5.5 to 6.1 m. If the pipe section length is unknown, a conservative assumption of 6.1 m should be used. This will yield the largest rotations because of the greater differential displacements. Using this established convention, the rotation at joint j, j, can be calculated using the following expression:

j = cos 1

ji kj + Lji Lkj + ji kj
2 ji

+ Lji + ji
2

) (

2 kj

+ Lkj + kj
2

(7)

For buried pipelines, the locations of the joints may be unknown. In this case, in order to determine the maximum potential joint rotation along a pipeline, a joint should be assumed at

11

one end of the displacement profile. Assuming small displacements, the next joint should be located on the displacement profile a distance of a pipe section length along the pipeline. The line connecting these two points represents the pipe section between these two points. This should be continued to the other end of the profile. Once the rotations are calculated, the same procedure should be repeated after offsetting the location of the first joint. Multiple analyses should be completed with the offset increasing until it is as long as the length of a pipe section. From these analyses, the maximum potential joint rotation can be determined.

Allowable Stresses and Joint Rotations

The failure of the pipeline from excavation-induced ground movements can occur from excessive bending stresses or large rotations at the joints. Stresses and rotations from preexisting conditions must be considered when allowable values for imposed stresses, ALLOW, and rotations, ALLOW, are established, for bending stresses and rotations, respectively. These allowable values are presented in Tables 3 and 4. For cast iron pipe, the failure in a pipe due to bending occurs as an abrupt brittle fracture. Attewell et al. (1986) recommend a maximum design stress for cast iron under direct tensile load equal to one-quarter of the ultimate tensile strength of the material. For cast iron exposed to bending, a rupture factor of 1.6 needs to be applied which results in a maximum design stress equal to 40 percent of the ultimate tensile strength of the material or a factor of safety of 2.5. For ductile iron in bending, which is a much more flexible material than cast iron, the Ductile Iron Pipe Research Association (2001) recommends using a factor of safety of 2.0 for calculations involving bending. Attewell, et al. (1986) recommend using a value of 85 percent of the minimum yield strength of the material since the failure of the material would not occur at

12

the yield point, but when plastic yielding exists through the section. This definition corresponds to a safety factor of 1.2. Steel line pipe is usually the material used in the transportation of oil and gas under extremely high internal pressures. Under these conditions, a combined stress analysis should be used to determine an allowable design stress. For safety considerations, pipelines installed in urban environments generally are maintained at low internal pressures and a combined stress analysis is unnecessary. Steel pipe for these installations can be assumed to follow the ground movements behaving as a continuously supported beam. For laterally supported beams subjected to bending moments, the allowable stress can be calculated as 60 percent of the yield stress which is equivalent to a factor of safety of 1.67. Polyethylene pipe design strengths are established through an internal pressure analysis. The Plastics Pipe Institute (2000) recommends using a hydrostatic design basis value to determine a limiting strength for the pipe material. They recommend using a factor of safety of 2 resulting in allowable hydrostatic design values for PE80 and PE100 grades as 4.3 MPa and 5.5 MPa. Failure at a pipe joint can occur in the form of an excessive amount of leakage or a fracture of the pipe joint itself. These are critical behaviors for only cast iron and ductile iron pipelines since steel and plastic pipe joining procedures ensure joints that are of equivalent strength to that of the material. Experimental data have been obtained to determine rotations at which failure of a joint could occur. The majority of cast iron pipe installations are joined with lead-caulked joints. These semi-rigid joints allow some rotation because of the soft packing material, but failure can occur when the packing is forced from the joint resulting in excessive leakage. Fracture of the joint

13

can occur when the rotation is large enough to cause metal-to-metal contact between the bell and the spigot inducing large bending stresses. Maynard and ORourke (1977) observed excessive leakage from a cast iron main exposed to ground movements with a joint rotation of 0.006 radians (0.34 degrees). With the application of a safety factor of 1.25 to be conservative, this can define an allowable ground movement-induced rotation of approximately 0.0048 radians (0.275 degrees). For rubber gasket joints for both cast iron and ductile iron the failure at the joint occurs as metal-to-metal contact. This rotation is dependent on the size of the pipe leading to different dimensions within the joint. Leakage is no longer a primary concern due to the flexibility of the rubber material to seal holes that could possibly form from movement at the joint. From observed conditions of cast iron and ductile iron mains with flexible joints following installation, Attewell, et al. (1986) suggest assuming an initial rotation within the joint up to 0.026 radians (1.5 degrees). This value is reflected in the allowable ground movement-induced rotations shown in Table 4. The majority of steel pipe is joined by welding with a minimal loss in strength along the pipeline from this method. Table 5 shows typical percentages for the reduction in strength for a steel pipe for different welds presented by Watkins and Anderson (2000). The greatest strength reduction of 25 percent of the strength of the steel is for single welded lap joints. This strength reduction should be applied to the design bending stress for the pipe material at the location of the joint for the bending stress analysis. The thermoplastic behavior of polyethylene allows for the material to be heated and reformed to another shape without losing strength. The fusing of two pipes by heat or electrofusion produces a joint with equal or greater strength of that of the material.

14

PREDICTION OF EXCAVATION-INDUCED GROUND MOVEMENTS

Semi-empirical methods have been developed for determining the values of maximum movement resulting from a deep braced excavation. For application of the stress analysis presented earlier, the distribution of the movements along the excavation are necessary for determining the displacement profile of the pipeline. A method for predicting the distributions of the lateral and vertical movements by the application of the complementary error function is presented by Roboski and Finno (2004). They determined that the ground movement distribution might be adequately represented with the following formula: 1 x A (8) B 2 where (x) is the settlement or lateral movement at distance x from the corner of the wall, max is

( x ) = max 1 erfc

the maximum movement, A is the distance to the inflection point of the function to the corner of the wall, and B is an empirical shape factor. Positive values of settlement should be used and lateral movement should be considered positive towards the excavation. The value of A is determined from a relationship with the ratio of the depth of the excavation to the length of the wall, He/L. The value of B can be calculated from the value of A from the following expression: L A 2 B= 2.8 A thorough explanation of the derivation and application of this procedure is presented by Roboski and Finno (2004).

(9)

Summary of Procedure

To determine the magnitudes of stresses and rotations in a pipeline caused by excavationinduced movements:

15

1.

Determine the maximum lateral ground movements from semi-empirical or detailed methods. Estimate the maximum vertical movement from the value of the lateral movement, and develop a proposed ground surface settlement profile at the pipeline location based on procedures summarized in the previous section.

2.

The stresses and rotations in the pipeline can be determined from the two limiting conditions. The pipeline is assumed flexible and a bending analysis is conducted to determine the maximum tensile stress occurring in the pipe. The pipeline is then assumed to behave rigidly, and a joint rotation analysis made to determine the largest possible rotation along the pipeline.

3.

The maximum tensile stress and joint rotation values should then be compared to the allowable values given in Tables 4 and 5, respectively. If the values predicted fall below the minimum allowable values, the pipeline should be safe under the imposed ground deformations.

CASE STUDIES Lurie Center

Construction of the Lurie Medical Research Center included a deep braced excavation in downtown Chicago. The dimensions of the excavation were approximately 82 m by 69 m by 12.8 m deep. The support system for the excavation was a sheet pile wall with three levels of tiebacks. A more complete description of the project is reported in Finno and Roboski (2004). The area surrounding the Lurie Center site is heavily populated with underground utilities transmitting water, waste, gas, electric lines, and telecommunication cables. The analyses

16

presented herein focuses on gas mains along the north, west, and south walls of the excavation. The locations of the instrumentation and gas mains with respect to the excavation are shown in Figure 2. The instrumentation around the site consisted of eight inclinometers, 150 surface points, and 18 embedded soil anchors, and 30 points on utilities. Detailed ground movement measurements were collected throughout the excavation process and ground movement distributions were computed from the data by a radial basis interpolation. The gas mains along the west and north walls of the excavation are old ductile iron mains with mechanical joints. The main along the west wall is a 500 mm diameter pipeline located 5.5 m from the excavation wall. The main along the north wall is a 150 mm diameter pipeline located approximately 15.5 m from the north wall of the excavation. Along the south wall, there is a 300 mm diameter main approximately 8.1 m from the southern edge. The pipelines were assumed to move with the ground and provide no restraint against the soil. Therefore, it was assumed that the ground displacement profiles at the locations of the gas mains defined their movement. To support this assumption, comparisons were made between settlement readings taken from the survey points located directly on the gas mains and the approximated settlement profile at the pipeline location. Only the values for vertical movements were compared because lateral readings from the gas mains were unable to be obtained. Figure 3 shows a comparison between the pattern of the pipeline determined from a survey point located directly on the pipeline and that of the observed ground surface movement above for the pipeline located along the west wall of the Lurie Center excavation. Excluding an initial offset of the observed ground movement, the displacement patterns are the same within the accuracy of the optical survey measurements. Because the pipe along the west wall had the largest diameter, the

17

ground displacement profiles at the locations of the three gas mains can be then assumed to be the movements of the pipelines. Figure 4 shows the ground movements observed at the locations of the gas mains at the end of the excavation, where the maximum vertical and lateral movements were the greatest. This should correspond to the maximum stresses in the pipeline. Only settlement values only for the gas main to the north of the excavation are reported because the pipeline was located further from the excavation than the furthest row of lateral survey points. Assuming that the pipelines behave flexibly, a bending stress analysis was completed for the ground displacement profiles in Figure 4. Figure 5 shows the variation in time of the maximum tensile stress for the three gas mains throughout the excavation process. There is a gradual increase in the magnitude of the maximum tensile stress for all three gas mains. The magnitude of the maximum tensile stress for all mains at the completion of the excavation is representative of the maximum stress incurred for the duration of the construction. The maximum tensile stresses in the north, south, and west gas mains were 2.5, 10 and 25 MPa, respectively. For cast iron and ductile iron pipelines, these values are well below the allowable stress values of the pipe for design (Table 3). From the bending stress analysis, it can be concluded that no failure due to bending would occur in these gas mains due to the excavation. Assuming the gas mains adjacent to the Lurie Center excavation behave as rigid pipelines, an analysis on the relative rotation at the joints must be completed to determine if the rotations are within the allowable limits. The joints for the north and west gas mains are known to be mechanically bolted joints. The type of joint for the cast iron pipeline along the south wall is unknown.

18

Figure 6 shows the maximum joint rotations for the north, south, and west gas mains for the time length of the construction for pipe section lengths of 3.6 m. As shown in Table 4, lead caulked joints began to show large amounts of leakage at rotations equaling 0.006 radians (0.34 degrees). From Figure 6, it is apparent that the rotations in the joints from the ground movements for the north and west gas mains remained below the critical rotations for rubber gasket joints. The joint rotations for the south gas main, which was cast iron, reached critical magnitudes for lead caulked joints. A comparison of the maximum stress and joint rotations obtained from the ground movements computed as a radial basis interpolation and of pipe displacement on the error function model showed very similar results. Figure 7 shows the comparison of the predicted, calculated, and allowable maximum bending stresses and joint rotations values for the gas main along the west wall of the excavation, a 500 mm diameter ductile iron gas main with mechanical joints. The maximum calculated stress from the complementary error function-based ground movements for the final stages of the construction were within 8 MPa of those computed from the radial basis approximation using the observed ground movements. The joint rotation analysis yielded similar results. The joint rotations from the complementary error function-based ground movements differed from those computed as a radial basis function by approximately 2.5 x 10-3 radians (0.14 degrees). Similar conclusions concerning the effects of the excavation on the pipelines at this site can be drawn when computing stresses and rotations with both methods.

Chicago Excavation Case Studies

Maynard and ORourke (1977) present ground surface movement data for different braced excavations in Chicago where cast iron pipelines were impacted by the effects of excavations. Table 6 presents the information for the four cast iron mains and their approximate

19

maximum movements. The ground movements were observed within a distance of approximately 3 m behind the pipeline to eliminate the effects of the edge of the trench on the data. This produced ground movement data representative of the movement of the pipelines. The displacement profiles for the four pipelines show the maximum movements occurring near the center of the excavation with large curvatures at the edge. From a bending stress analysis, assuming a modulus of elasticity of 100 MPa and negligible change in rotation of the principal planes, the largest stresses occurred at the edge of the excavation where the largest curvatures were located. The characteristic length for determining the curvatures for each pipeline was taken as the distance between the data points along that pipeline. The distances for the four mains ranged from 7.6 to 15.2 m, which is a reasonable characteristic length for the calculation of curvature being that they are greater than 5.5 m, which was determined to be the lower bound. For two of the pipelines the maximum calculated bending stresses proved to be greater than the minimum value of allowable bending stress from excavation-induced ground movements for cast iron presented in Table 3. However, there was no evidence of fracture of the pipes due to excessive bending showing the conservativeness of the bending stress analysis. For an analysis of the rotations at the joints along the pipelines, it was assumed that the joints were located at the data points. Since only the total vector movement data was available, the relative rotations could be calculated from the changes in slope of the displacement profiles. Figure 8 shows the rotations calculated at the data points. The open symbols represent mechanical joints and the filled symbols denote lead caulked joints. For the two mains joined by lead caulked joints, the 300 mm diameter pipeline showed the largest relative rotation at a joint of 6 x 10-3 radians (0.34 degrees) at which leakage at the joints was observed and was taken out of service. The mains equipped with mechanical joints

20

experienced a maximum rotation of 8 x 10-3 radians (0.46 degrees). This rotation is greater than that observed to cause failure at a lead caulked joint, however, for mechanical joints it is within the allowable limits of 0.044 radians (2.5 degrees). Both pipelines joined with mechanical joints experienced larger relative rotations yet from observations they did not show excessive leakage, illustrating the benefits of the more flexible joints.

CONCLUSIONS

Based on the analyses presented herein and the data obtained at several deep excavations in Chicago, the following conclusions can be made. 1. From comparison of ground displacements interpolated from collected data and field observed movements of buried utilities, it was shown that the pipeline tracked the movement of the surrounding soil within the accuracy of the optical survey data. For computation of bending stresses and joint rotations induced in pipelines from ground movements related to a deep braced excavation, the pipeline displacement profile may be assumed to be that of the displacement of the surrounding soil when the displacement in the pipe is small in relation to the length of the pipeline, i.e. sin . When utilizing this assumption, special consideration of construction activity, differential soil behavior, and local effects must be taken into account. 2. A methodology was presented for computing the longitudinal bending stresses and joint rotations induced in a pipeline from an adjacent deep braced excavation. The validity of the method for calculating the bending stresses and joint rotations is illustrated by comparisons of calculated ground movement

21

values and direct observations made in the field for the Lurie Center excavation and the various Chicago excavations presented by Maynard and ORourke (1977). The method proved to be conservative for both bending stress and joint rotation analysis. 3. The more critical condition for a cast iron or ductile iron main considered herein is excessive rotation at a joint. The bending stress analysis on the ground movements presented by Maynard and ORourke (1977) showed large longitudinal tensile stresses with no observed cracking or rupture. The small calculated joint rotation of 6 x 10-3 radians (0.34 degrees) proved to cause excessive leakage in a lead caulked joint.

22

REFERENCES

Ahmed, I. (1990). Pipeline Response to Excavation-Induced Ground Movements. PhD thesis, Department of Civil and Environmental Engineering, Cornell University, Ithaca, NY. American Petroleum Institute. (1991). Specification for Line Pipe, 39th Ed. American Petroleum Institute, Washington, DC. Attewell, P.B., Yeates J., and Selby, A. R. (1986). Soil Movements Induced by Tunneling and
Their Effects on Pipelines and Structures. Blackie and Son, Ltd., London.

Bonds, R. W. (2003). Ductile Iron Pipe Joint and Their Uses. Ductile Iron Pipe Research Association, Birmingham, AL. Carder, D. R., Taylor, M. E., and Pocock, R. G. (1982). Response of a Pipeline to Ground Movements Caused by Trenching in Compressible Alluvium. Department of the
Environment Department of Transport, TRRL Report LR 1047, Transport and Road

Research Laboratory, Crowthorne. Carder, D. R. and Taylor, M. E. (1983). Response of a Pipeline to Nearby Trenching in Boulder Clay. Department of the Environment Department of Transport, TRRL Report LR 1099, Transport and Road Research Laboratory, Crowthorne. Clough, G. W. and ORourke, T. D. (1990). Construction Induced Movements of In-Situ Walls. Design and Performance of Earth Retaining Structures, Proceedings of a
Specialty Conference at Cornell University, ASCE, New York, 439-470.

Croft, J. E., Menzies, B. K., and Tarzi, A. I. (1977). Lateral Displacement of Shallow Buried Pipelines due to Adjacent Deep Trench Excavations. Geotechnique, 27(2), 161-179. Finno, R. J., and Roboski, J. F. (2004). Three-Dimensional Responses to a Tied-back Excavation Through Clay.

23

Hsieh, P. G. and Ou, C. Y. (1998). Shape of Ground Surface Settlement Profiles Caused by Excavation. Canadian Geotechnical Journal, 35, 1004-1017. Maynard, T. R. and ORourke, T. D. (1977). Soil Movement Effect on Adjacent Public Facilities. Preprint No. 3111, ASCE Annual Meeting, San Francisco, CA. Nath, P. (1983). Trench Excavation Effects on Adjacent Buried Pipes: Finite Element Study.
Journal of Geotechnical Engineering, ASCE, New York, NY, 109(11), 1399-1415.

Plastics Pipe Institute. (1993). Engineering Properties of Polyethylene. PPI Handbook of


Polyethylene Piping, Plastics Pipe Institute, Washington, DC.

Plastics Pipe Institute. (2003). Specifications, Test Methods and Codes for Polyethylene Piping Systems. PPI Handbook of Polyethylene Piping, Plastics Pipe Institute, Washingtion, DC. Plastics Pipe Institute. (2000). Model Specification for Polyethylene Plastic Pipe, Tubing and
Fittings for Water Mains and Distribution. Plastics Pipe Institute, Washington, DC.

Roboski, J. F., and Finno, R. J. (2004). Distributions of Ground Movements Parallel to a Deep Excavation. Salmon, C. G., and Johnson, J. E. (1996). Steel Structures: Design and Behavior, Emphasizing
Load and Resistance Factor Design, 4th Ed. HarperCollins College Publishers, New

York, NY. Sears, E.C. (1968). Comparison of the Soil Corrosion Resistance of Ductile Iron Pipe and Gray Cast Iron Pipe. Materials Protection, 7(10), 33-36. Tarzi, A. I., Menzies, B. K., and Crofts, J. E. (1979). Bending of Jointed Pipelines in Laterally Deforming Soils. Geotechnique, 29(2), 203-206.

24

Timoshenko, S. and Goodier, J. N. (1951). Theory of Elasticity. McGraw-Hill Book Company, Inc., New York, NY. Untrauer, R. E., Lee, T. T., Sanders, Jr., W. W., and Jawad, M. H. (1970). Design Requirements for Cast Iron Soil Pipe. Bulletin 199, Engineering Research Institute, Iowa State University, Ames, IA. Watkins, R. K. and Anderson, L. R. (2000). Structural Mechanics of Buried Pipes, CRC Press, New York, NY.

25

Tables

Pipe Material
Vertically Pit Cast Iron Centrifugally Cast Iron Ductile Iron Steel Grade A Steel Grade B Steel Grade 414 Polyethylene PE80 Polyethylene PE100

Coeff. of Ultimate Modulus of Yield Stress, Fy Stress, Fu Elasticity Poisson's Thermal Exp. Ratio (per C) (MPa) (MPa) (GPa)
145 20 207 20 420 331 413 517 31 31 83 14 114 14 166-180 200 200 200 552-758 758-1103 0.26 0.26 0.28 0.29 0.29 0.29 0.42 0.42 11 x 10-6 11 x 10-6 11 x 10-6 12 x 10-6 12 x 10-6 12 x 10-6 2 x 10 -6 2 x 10 -6

Reference
Ahmed (1990) Ahmed (1990) DIPRA (2001) API (1991) API (1991) API (1991) PPI (2003) PPI (2003)

300 207 241 414 15-18 21-24

Table 1 Engineering Properties for Piping Materials

26

Material

Joint

Limiting Rotation
Leakage, rad. (deg.) Failure, rad. (deg.)
0.09-0.1 (5-6) 0.07-0.09 (4-5) 0.07 (4) 0.05-0.09 (3-5) 0.03-0.14 (2-8) 0.26 (15)

Reference
See Note 1 Attewell, et al. (1986) Attewell, et al. (1986) Bonds (2003) Bonds (2003) Bonds (2003)

Lead-Caulked 0.0094-0.017 (0.54-1.0) Cast Iron Rubber-Gasket Mechanical Rubber-Gasket Ductile Iron Mechanical Ball and Socket

Note 1: Adapted from Untreaur, et al. (1970), O'Rourke and Trautmann (1980), Harris and O'Rourke (1983), and Attewell, et al. (1986)

Table 2 Failure Rotations for Selected Cast Iron and Ductile Iron Joints

27

Yield Ultimate Initial Stress 1 Strength Strength ( ), (Fy), MPa (Fu), MPa MPa 13.8 - 38.6 145 --Pit Cast Iron Spun Cast Iron 20.7 - 52.4 207 --33.1 - 71.7 300 420 Ductile Iron 41.4 - 82.8 207 331 Grade A Steel 41.4 - 82.8 241 414 Grade B Steel 41.4 - 82.8 517 414 Grade 414 Steel 3 0.13 - 0.26 8.6 --PE80 0.28 - 0.56 --11 PE100 Pipe Material
INITIAL

Factor of Safety
2.5 2.5 1.2 1.67 1.67 1.67 2.0 2.0

Design Bending Allowable Stress 2 Stress ( ), MPa ( ), MPa


B ALLOW

0.4Fu 0.4Fu 0.8Fu 0.6Fy 0.6Fy 0.6Fy 0.5HDB 0.5HDB

58 82.8 336 124.2 144.6 248.4 4.3 5.5

19.3 - 44.1 30.3 - 62.1 264.3 - 302.9 41.4 - 82.8 62.1 - 103.4 165.4 - 206.8 4.04 - 4.17 4.94 - 5.22

1 Adapted from Taki and O'Rourke (1984) assumed initial longitudinal bending strain of 0.02 to 0.04% 2 Allowable bending stress from excavation-induced ground movement = ALLOW = B - INITIAL 3 Polyethylene designed for internal pressure. Allowable values expressed as Hydrostatic Design Basis

(HDB).

Table 3 Allowable Bending Stresses from Excavation-Induced Movements

28

Mode of Joint Type Failure Leakage Lead-Caulked Lead Caulked Cast Iron Metal Rubber-Gasket Push-on Binding Mechanical (metal-tometal Rubber-Gasket Push-on Ductile Iron contact) Mechanical Ball and Socket Material
1 ALLOW

Failure Rotations
Radians
0.0094 - 0.016 0.09 - 1.0 0.07 - 0.09 0.07 0.05 - 0.09 0.035 - 0.14 0.22 - 0.26
2

Allowable Rotations, ALLOW 1


Radians
0.0048
3

Degrees
0.54 - 0.92 5-6 4-5 4 3-5 2-8 12.5 - 15

Degrees
0.275 3.5 - 4.5 2.5 - 3.5 2.5 1.5 - 3.5 0.5 - 6.5 11 - 13.5

0.06 - 0.08 0.044 - 0.06 0.044 0.026 - 0.06 0.009 - 0.11 0.19 - 0.24

represents allowable excavation-induced rotation with assumed 0.026 rad. (1.5 deg.) initial rotation (Attewell, et al., 1986) for flexible joints, ALLOW = METAL BINDING - INITIAL. ALLOW = LEAKAGE/F.S. for lead caulked joints where F.S. = 1.25. 2 Observed from laboratory tests to cause excessive leakage. 3 Observed from field data to cause excessive leakage (initial rotation already occurred).

Table 4 Allowable Joint Rotations for Cast Iron and Ductile Iron Joints

29

Welded Joint
Butt Welded Joint 1 Single Welded Lap Joint 2 Double-Welded Lap Joint

Percent Strength Reduction


0 25 20

1 For a full-penetration weld through thickness of pipe. 2 For a gap smaller than 3.2 mm.

Table 5 Strength Reductions at Location of Line Pipe Welded Joints (Watkins and Anderson, 2000)

30

Main Type

Year

Joints

Internal Diameter Depth Pressure (mm) (m) (kPa)


300 1200 900 150 2.1 2.3 1.7 1.4 207 310.5 207-276 1.75

Symbol
WL-1 WL-2 WM-1 GM-1

1938 Lead-Caulked Water Main pre-1900 Lead-Caulked 1960 Mechanical Gas Main 1935 Mechanical

Table 6 Description of Cast Iron Pipelines Parallel to Chicago Excavations (Maynard and ORourke, 1977)

31

Figures

kj
i

ji
j L ji

L kj

ji
a. Bending

kj

kj
i

ji
L ji j j

L kj

ji
b. Rotation

kj

Figure 1 Coordinates for Bending and Joint Rotation Analyses

32

E. Superior St.

Existing Pedestrian Tunnel

LURIE MEDICAL RESEARCH CENTER EXCAVATION


N. Fairbanks Ct.

Prentice Women's Hospital

LEGEND
Gas Main Surface Point Utility Point Soil Anchor Inclinometer

0 2 4 8 Scale in meters

E. Huron St.

Figure 2 General Layout of Lurie Center Site Instrumentation and Adjacent Gas Mains

33

Days from Completion of Sheet Pile Wall Installation


0 0 10 20 30 40 50 60 50 100 150 200 250 300

Settlement (mm)

Ground Surface Settlement from Contours

Survey Data from Pipe

Figure 3 Comparison of Ground and Pipeline Movement during Excavation at Lurie Center

34

80 70

Settlement (mm)

60 50 40 30 20 10 0 -20 0 20 40 60 80 100

Distance from Corner of Excavation (m)

North

South

West

90 80 70 60 50 40 30 20 10 0 -20 0 20 40 60 80 100

Lateral Movement (mm)

Distance from Corner of Excavation (m)


South
West

Figure 4 Ground Displacements at Location of Gas Mains along North, South, and West Walls after Completion of Excavation

35

Max. Tensile Stress (Mpa)

30 25 20 15 10 5 0

50

100

150

200

250

300

Days from Completion of Sheet Pile Wall Installation

North Pipeline

South Pipeline

West Pipeline

Figure 5 Maximum Tensile Stress in Pipelines Adjacent to North, South, and West Wall During Excavation

36

Rotation (x10-3 rad.)

6 5 4 3 2 1 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation


North Pipeline Allowable Rotation
South Pipeline Failure Rotation
West Pipeline

Figure 6 Maximum Relative Rotation Encountered Along North, South, and West Pipelines During Excavation for 3.6 m Pipe Sections

37

300

Tensile Stress (MPa)

250 200 150 100 50 0 100

150

200

250

300

Days of Construction of Sheet Pile Wall Installation


Radial Basis

Error Function

Maximum Allowable Stress

a) Maximum Tensile Bending Stresses


30

Rotation (x10-3 rad.)

25 20 15 10 5 0 100

150

200

250

300

Days from Completion of Sheet Pile Wall Installation


Radial Basis

Error Function

Maximum Allowable Rotation

b) Maximum Joint Rotations

Figure 7 Observed and Predicted Maximum Bending Stresses and Joint Rotations for Final Stages of Construction for Gas Main Along West Wall

38

9 8 7 6 5 4 3 2 1 0 -5 0 5 10 15 20 25 30 35 40 45 50

Rotation (10-3 rad.)

Distance from Corner of Excavation (m)


WL-1

WL-2

WM-1

GM-1

Note: Leakage Observed in WL-1.

Figure 8 Rotations in Cast Iron Pipelines Adjacent to Excavations in Chicago

39

APPENDIX A: LURIE CENTER DATA

TABLE OF CONTENTS

Table of Contents List of Tables List of Figures

40 42 44

A.1 A.2

Introduction Pipeline and Material Properties A.2.1 Cast Iron A.2.2 Ductile Iron A.2.3 Steel A.2.4 Polyethylene Case Study: Lurie Medical Research Center A.3.1 Ground Movements due to Excavation A.3.2 Ground Movements at Locations of Gas Mains A.3.3 Ground Movement and Pipeline Movement Comparison Stress Conditions on Buried Pipelines A.4.1 Initial Stresses A.4.1.1 Hoop Stress A.4.1.2 Ring Stresses from Soil Cover A.4.1.3 Traffic Loads A.4.1.4 Stresses from the Installation Procedure and Adjacent Construction History A.4.1.5 Environmental Effects A.4.2 Analysis of Effects of Ground Movements from Adjacent Excavations on Pipeline A.4.2.1 Soil Pipeline Interaction A.4.2.2 Sign Convention and Definition of Terms A.4.2.3 Calculation of Bending Stress for Flexible Pipeline A.4.2.4 Calculation of Relative Rotation at Joint for Rigid Pipeline Summary Conclusions

48 49 49 55 57 60 63 64 66 69 73 73 74 75 81

A.3

A.4

83 84
87 90 91 92

98
104 108

A.5 A.6

40

References Tables Figures

111 115 126

41

LIST OF TABLES

Table A-1 Table A-2 Table A-3 Table A-4 Table A-5 Table A-6 Table A-7 Table A-8 Table A-9 Table A-10 Table A-11 Table A-12 Table A-13 Table A-14 Table A-15 Table A-16

Tensile Strength and Stress-Strain Properties for Cast Iron Typical Dimensions for Lead Caulked Joints Typical Dimensions for Flexible Joints for Cast Iron Pipe Experimental Results for Rotation at Leakage for Lead Caulked Cast Iron Pipe Joints Tensile Strength and Stress-Strain Properties for Ductile Iron Typical Dimensions for Flexible Joints for Ductile Iron Pipe Strength Reductions at Location of Welded Line Pipe Joints

115 115 116 116 117 117 118

Typical Mechanical Properties for Polyethylene Gas Distribution Pipe 118 Locations of Underground Utilities with Respect to Lurie Center Excavation Definitions of Stages of Construction 119 119

Magnitudes of Maximum Ground Movements Surrounding Lurie Center Excavation 120 Magnitudes of Maximum Ground Movements at Locations of Gas Mains Adjacent to Lurie Center Excavation 120

Locations of Utility Survey Points with Respect to Nearest Corner of Excavation 121 Sample Hoop Stress Calculations for Gas Mains Adjacent to Lurie Center Excavation Overpressure Stresses for Gas Mains Adjacent to Lurie Center Excavation Allowable Bending Stresses from Excavation-Induced Movements 121 122 123

42

Table A-17 Table A-18 Table A-19 Table A-20

Allowable Excavation-Induced Joint Rotations for Semi-Rigid and Flexible Cast Iron and Ductile Iron Joints Typical Engineering Properties for Piping Materials Dimensions and Maximum Tensile Stress Values in Gas Mains Adjacent to Lurie Center at End of Excavation

123 124 124

Dimensions and Maximum Joint Rotation in Gas Mains Adjacent to Lurie Center at End of Excavation 125

43

LIST OF FIGURES

Figure A-1 Figure A-2 Figure A-3 Figure A-4 Figure A-5 Figure A-6 Figure A-7 Figure A-8 Figure A-9 Figure A-10 Figure A-11 Figure A-12 Figure A-13 Figure A-14 Figure A-15 Figure A-16 Figure A-17 Figure A-18 Figure A-19

Typical Stress-Strain Curves for Cast Iron Typical Lead Caulked Cast Iron Joint Typical Flexible Iron Joints Stress-Strain for Cast Iron and Ductile Iron Rotational Stiffness of Ductile Iron Rubber Gasket Joint Typical Joint Welds for Line Pipe Stress-Strain Curve for Polyethylene Under Controlled Conditions Typical Joining Methods for Polyethylene Pipe General Layout of Lurie Center Site Instrumentation and Adjacent Underground Utilities Vertical Ground Movements Along North Wall on Day 146 Lateral Ground Movements Along North Wall on Day 146 Vertical Ground Movements Along North Wall on Day 192 Lateral Ground Movements Along North Wall on Day 192 Vertical Ground Movements Along North Wall at End of Excavation Lateral Ground Movements Along North Wall at End of Excavation Vertical Ground Movements Along South Wall on Day 157 Lateral Ground Movements Along South Wall on Day 157 Vertical Ground Movements Along South Wall on Day 203 Lateral Ground Movements Along South Wall on Day 203

126 126 127 127 128 128 129 129 130 131 131 132 132 133 133 134 134 135 135

44

Figure A-20 Figure A-21 Figure A-22 Figure A-23 Figure A-24 Figure A-25 Figure A-26 Figure A-27 Figure A-28 Figure A-29 Figure A-30 Figure A-31 Figure A-32 Figure A-33 Figure A-34 Figure A-35 Figure A-36

Vertical Ground Movements Along South Wall at End of Excavation Lateral Ground Movements Along South Wall at End of Excavation Vertical Ground Movements Along West Wall on Day 146 Lateral Ground Movements Along West Wall on Day 146 Vertical Ground Movements Along West Wall on Day 185 Lateral Ground Movements Along West Wall on Day 185 Vertical Ground Movements Along West Wall at End of Excavation Lateral Ground Movements Along West Wall at End of Excavation Maximum Magnitude of Vertical Movement During Excavation at Location of Gas Main Adjacent to North Wall Maximum Magnitudes of Movements During Excavation at Location of Gas Main Adjacent to South Wall Maximum Magnitudes of Movements During Excavation at Location of Gas Main Adjacent to West Wall Re-zeroed Settlement Values Along North Gas Main at Completion of Excavation Re-zeroed Settlement Values Along South Gas Main at Completion of Excavation Re-zeroed Settlement Values Along West Gas Main at Completion of Excavation Comparison of Utility Survey Point U-1 Movement with Approximated Ground Movement Values Comparison of Utility Survey Point U-2 Movement with Approximated Ground Movement Values Comparison of Utility Survey Point U-3 Movement with Approximated Ground Movement Values

136 136 137 137 138 138 139 139

140 140 141 141 142 142 143 143 144

45

Figure A-37 Figure A-38 Figure A-39 Figure A-40 Figure A-41 Figure A-42 Figure A-43 Figure A-44 Figure A-45 Figure A-46

Comparison of Utility Survey Point U-4 Movement with Approximated Ground Movement Values Comparison of Utility Survey Point U-5 Movement with Approximated Ground Movement Values Comparison of Utility Survey Point U-6 Movement with Approximated Ground Movement Values Free-Body Diagram of Forces Resulting from Internal Pressure Sign Convention for Thrust, Moment, Displacement, and Stress Equations Local Coordinate System Convention for Pipeline Analysis Definitions of Dimensions and Differential Ground Movement Designations Pipeline Profiles with Established Local Coordinate System for Analysis of Bending Stress for Flexible Pipeline Pipe Cross Section and Sign Convention Comparison of Characteristic Lengths of 6.1, 9.2, and 15.3 m for Stress Analysis Along Gas Main Adjacent to North Wall During Excavation Comparison of Characteristic Lengths of 6.1, 9.2, and 15.3 m for Stress Analysis Along Gas Main Adjacent to South Wall During Excavation Comparison of Characteristic Lengths of 6.1, 9.2, and 15.3 m for Stress Analysis Along Gas Main Adjacent to West Wall During Excavation Maximum Tensile Stress in Pipelines Adjacent to North, South, and West Walls During Excavation Maximum Tensile Stress Along North Gas Main at Completion of the Excavation Maximum Tensile Stress Along South Gas Main at Completion of the Excavation

144 145 145 146 146 147 147 148 148

149

Figure A-47

149

Figure A-48

150 150 151 151

Figure A-49 Figure A-50 Figure A-51

46

Figure A-52 Figure A-53 Figure A-54 Figure A-55 Figure A-56 Figure A-57 Figure A-58 Figure A-59 Figure A-60 Figure A-61

Maximum Tensile Stress Along West Gas Main at the Completion of the Excavation 152 Pipeline Profiles with Established Local Coordinate System for Analysis of Joint Rotations for Rigid Pipeline Schematic of Joint Rotation at Joint j of Rigid Pipeline Maximum Relative Rotation Encountered Along North, South, and West Pipelines During Excavation for 3.6 m Pipe Sections Comparison of Maximum Relative Rotation in North Gas Main for 3.6 and 6.1 m Pipe Sections Comparison of Maximum Relative Rotation in South Gas Main for 3.6 and 6.1 m Pipe Sections Comparison of Maximum Relative Rotation in West Gas Main for 3.6 and 6.1 m Pipe Sections Joint Rotations and Pipeline Settlement Along Pipeline Adjacent to North Wall at Completion of Excavation Joint Rotations and Pipeline Settlement Along Pipeline Adjacent to South Wall at Completion of Excavation Joint Rotations and Pipeline Settlement Along Pipeline Adjacent to West Wall at Completion of Excavation 152 153 153 154 154 155 156 157 158

47

A.1 Introduction

Within urban environments, buried pipelines may be exposed to large ground movements either by tunneling, mining, or open cut constructions, such as trenching or deep braced excavations. These large ground movements can induce deformations in pipelines resulting in stresses within the pipeline. These stresses, if excessive, could result in damage or complete failure of the pipeline. Failure in a pipeline due to excavation-induced ground movements could be caused by large bending stresses in the pipe or relative rotations between two adjacent pipe sections at a joint. Large bending strains in a pipe and rotations in a joint are the result of significant differential movements along the pipeline. This mainly occurs near the edge of an excavation due to the transition of the pipeline from being restrained to it being free to move with the ground. A conservative analysis of the effects of the ground movements from deep braced excavations is presented. The analysis considers two different methods of deformation separately, either curvature in the pipe or rotation at the joints. The displacements along the pipeline are used to calculate the maximum stresses and rotation imposed on the pipeline. These values can be compared to established allowable values to determine whether the pipeline could be damaged or remain safe and in operation. This analysis is presented and applied to two case studies where data was obtained from excavations in downtown Chicago. Maximum calculated values are computed and compared to the allowable values established from previous experimental and empirical studies. Conclusions are drawn for the stresses that were imposed on the pipeline due solely to the ground movements resulting from the deep braced excavations.

48

A.2 Pipeline Material and Properties

Four different materials used in pipeline engineering will be discussed; cast iron, ductile iron, steel, and polyethylene. An overview of the manufacturing methods, engineering properties, stress-strain behavior, standard dimensions, and joining methods for all four pipeline materials will be presented and compared.
A.2.1 Cast Iron

Many cast iron gas pipelines in use today have been in operation for over 100 years. The initial growth of the use of gray cast iron pipe in the pipeline industry in the United States occurred around 1816. Foundries for production of cast iron specifically in the form of pipe began in the eastern states and spread quickly westward. The metallurgic composition of cast iron is an alloy of iron and carbon with a percentage of silicon and manganese. The carbon exists in the form of graphite flakes and gives the material much of its strength. Two main manufacturing methods were used in the production of cast iron pipe: pit casting and centrifugally, or spin, casting. The majority of cast iron pipes installed during the duration for which cast iron was the main piping material were pit cast. This was due to the late introduction of centrifugal casting in the 1920s. In the pit casting process, the molten mixture of metals was poured into either a horizontal or vertical mold and allowed to set in place as it cooled. Vertical pit casting was the preferred method of manufacturing due to the longer sections of pipe that could be cast. Horizontal casting was limited by the flexural rigidity of the mold core where bending of the core could cause inconsistent wall thickness along the length of the pipe. Vertical pit casting increased the length of pipe sections available from 1.2 to 1.5 m to

49

lengths of up to 3.6 m, in turn decreasing the amount of joints necessary. Both horizontal and vertical pit cast pipe was available in a range of diameters from 75 to 1500 mm. In the 1920s, the process of centrifugally casting gray cast iron pipe was introduced and became the primary manufacturing method of cast iron pipe by the early 1930s. This procedure involved the pouring of the molten material into a horizontal spinning mold. The rate of rotation of the mold was controlled to obtain the desired thickness of the pipe. The centrifugal forces generated by the spinning produced a material with a much more consistent cross section and even distribution of impurities within the pipe section. The graphite flakes characterizing the strength of the pipe were more evenly distributed to produce a much stronger and more consistent material. Cast iron pipe was available in pipe sections 3.6 to 6.1 m in length and in sizes ranging from 75 to 1200 mm diameter. The stress-strain behavior of cast iron exhibits a brittle behavior with no yield point and an abrupt fracture at failure. Under an applied stress there exists no completely elastic behavior for any stress value. Cast iron undergoes an amount of plastic strain under the application of an increment of load. The total strain at all stress values is composed of both elastic and plastic components. The elastic strain behavior is characterized by a curve of recoverable strain, which does not agree with the traditional straight-line path of linear elastic materials. Therefore, the definition of a modulus of elasticity of material is more challenging to calculate and is usually defined by an initial tangent modulus. A typical stress-strain curve for both pit cast and centrifugally cast iron presented by Attewell, et al. (1986) is shown in Figure A-1. The lack of elastic behavior by cast iron is clearly visible from the non-linearity of the elastic strain curves. The brittle behavior of both pit cast and centrifugally cast iron is apparent with rupture failure occurring at a relatively low value of

50

axial strain. Centrifugally cast iron shows approximately a 50 percent increase in strength over that of pit cast. This can be attributed to its quick solidification, uniformity of its cross section, and the isolation of impurities accomplished through the improved method of production. Cast iron has been shown to behave differently in tension and compression. This can have a large effect on the flexural behavior of the pipe. Schlick and Moore (1936) conducted 12 tests on specially cast plates of four different grades of strength with 13, 23, and 32 mm thickness in direct tension and compression. From the 12 tests, it was shown that the compressive strength of cast iron to be on average 3.6 times greater the tensile strength with a standard deviation of 0.31. The difference in the behavior of cast iron in tension and compression is most notable for strains above 0.1 percent, when the slope of the tensile curve decreases at a faster rate than that of the compression curve. Due to the complex behavior of cast iron, extensive testing has been done to determine appropriate engineering properties to represent its response to loading conditions. Ahmed (1990) compiled the test results and established recommended ranges for values for the ultimate stress, the initial tangent modulus, and the failure strain for pit cast and centrifugal cast iron. Table A-1 shows his suggested values for both pit cast and centrifugally cast iron. The variation of the values is the result of the improvements in the production process from pit to centrifugal casting. Cast iron pipe failure due to excessive bending occurs as an abrupt brittle fracture at low strains. Attewell, et al. (1986) recommend a maximum design stress for cast iron under direct tensile load equal to one-quarter of the ultimate tensile strength of the material. For cast iron in bending, a rupture factor of 1.6 needs to be applied which results in a maximum design stress equaling 40 percent of the ultimate tensile strength of the material. This is equivalent to using a factor of safety of 2.5.

51

Cast iron pipe joints are divided into three classifications depending on their behavior when subjected to a bending moment: rigid, semi-rigid, and flexible. Rigid joints are constructed with surface-to-surface contact of the bell and spigot of the joined pipe sections not allowing for any rotation when a bending moment is applied without large stresses in the joint. A pipeline with rigid joints behaves as a continuous beam with no relative rotations at the joints. Flanged and turned and bored joints are classified as rigid joints. These types of joints were used in the earliest installations of cast iron pipelines. Semi-rigid joints provide flexibility at the joint allowing an amount of rotation with the application of a bending moment. Prevention of leakage is provided by a packing and caulking materials within a bell and spigot joint. For the packing, materials such as a jute yarn, hemp, and oakum were used. This material prevents the caulking from entering the pipeline, centers the spigot in the bell, and provides the rotational flexibility in the joint. The different caulking materials available for these types of joints were Portland cement, lead, leadite, and hydrotite. Figure A-2 shows the general construction of a lead caulked cast iron joint with jute packing, the overall most common of these types of joints, and Table A-2 lists its general dimensions for different sizes of pipe. Lead caulked joints were the main joining method for cast iron pipe up until 1934 (Attewell, et al., 1986). The behavior of these joints is highly dependent on the type and condition of the packing and caulking materials. Differences have been noted in the behavior of these joints in gas and water mains. Before natural gas, the manufactured gas that was transported through the cast iron pipelines contained heavy hydrocarbons that were absorbed into the packing around the spigot. This low moisture gas dried out the packing leaving it hard and resistive to movements (Harris and ORourke, 1983). In water mains, the packing remained pliable and deformable allowing for

52

rotation at the joint. The packing also swelled with the absorption of water that helps in sealing any points of leakage. Harris and ORourke (1983) conducted bending tests on cast iron joints to determine the rotational stiffness of lead caulked joints with hard jute packing for 100, 150, and 200 mm diameter pipes. From their experimental results, they determined 100 and 200 mm diameter pipes to have a rotational stiffness range of 0.8 to 2.8 m-MN/rad. The 150 mm diameter joints exhibited rotational stiffness values of 1.7 to 6.8 m-MN/rad. Flexible joints are able to accommodate substantial amounts of rotation. Typical cast iron pipe flexible joints are shown in Figure A-3 with typical dimensions listed in Table A-3. The installation of these joints for cast iron was dominant from 1935 until 1970 (Attewell, et al., 1986). A soft, pliable rubber gasket is used for the cushioning of the spigot in the bell and provides rotational flexibility. The rubber gasket is initially compressed in between the bell and spigot to allow for rotation at the joint without allowing leakage. Flexible joints for cast iron pipelines are the rubber gasket push-on and the bolted-gland mechanical joint. Failure at a joint can be partial failure or complete failure. Partial failure occurs at lead caulked joints when the rotation is enough to cause an opening to form in the joint where the packing and the pipe are in contact. This allows leakage to occur at the joint. Complete failure of both semi-rigid and flexible joints occurs when the rotation at the joint is great enough to cause the spigot to force the packing or rubber gasket out of place and for the spigot and bell to come into contact. This metal-to-metal contact between the two pipe sections, called metal binding, is the point where the joint is considered to have completely failed because with the addition of a small rotation will cause a large increase in the stress within the joint. For cast iron

53

pipes 75 to 600 mm in diameter, Attewell, et al., (1986) determined the rotation necessary for metal binding to occur by the ratio dc/ds, which ranges from 0.09 to 0.1 radians (5 to 6 degrees). Table A-4 contains the suggested rotational limitations from experimental and finite element analyses. Harris and ORourke (1983) ran several bending tests on lead caulked joints with soft jute packing for 75, 150, and 200 mm. diameter cast iron pipe. From their tests, they observed an average rotation of 0.0094 radians (0.54 degrees) with a standard deviation of 0.015 radians (0.89 degrees) corresponding to maximum leakage from the joint. A self-healing phenomenon of the joint was observed after the maximum leakage was reached where there was a reduction in the flow from the joint following the maximum. This can be attributed to a plastic deformation of the caulking that begins to seal the path of flow. Attewell, et al. (1986) looked at the rotation necessary to force the lead out of the joint socket. They observed this occurring at a rotation of about 0.0697 radians (4 degrees). They also suggest that the installation and construction of the pipeline can cause an initial rotation at the joints. They suggest assuming an initial rotation of 0.026 radians (1.5 degrees), leaving an allowable rotation for lead caulked joints of 0.017 radians (1.0 degree) for gas mains and 0.026 radians (1.5 degrees) for water mains. The smaller allowable rotation for the gas mains is due to the observed hardening of the packing from the hydrocarbons. Untrauer, et al. (1970) conducted nine tests on 100 mm diameter cast iron pipe lead caulked joints under bending to determine the rotation where leakage began. They determined an average rotation of 0.016 radians (0.9 degrees) at first leakage. Maynard and ORourke (1977) observed cast iron pipelines with both lead caulked and mechanical joints subjected to ground movements adjacent to various deep braced excavations in Chicago. They observed

54

leakage to occur at the lead caulked joints when a rotation of 0.006 radians (0.34 degrees) was present.
A.2.2 Ductile Iron

In 1955, ductile iron was introduced into the pipeline market as a more favorable material for pipes due to its increased strength and greater resistance to impact and corrosion. Ductile iron pipe is produced by the centrifugal casting process. It is available in sizes ranging from 75 to 1600 mm in diameter and lengths of 5.5 to 6 m. The metallurgy of ductile iron is very similar to that of gray cast iron except for the form that carbon takes within the matrix of iron. Instead of carbon existing as graphite flakes, an inoculant is added which results in the carbon forming as graphite nodules, or spheroids. Carbon existing in this form within the iron matrix has a great impact on the strength of the material. Ductile iron was a favorable replacement for gray cast iron because of its stress-strain behavior. Unlike gray cast iron, ductile iron behaves elastically for a certain stress range (Angus, 1976). However, similar to that of gray cast iron, a definitive yield point does not exist. Figure A-4 shows a stress-strain behavior comparison of pit cast, centrifugal cast, and ductile iron. The information pertaining to the ductile iron curve was reported by Reese (1949) and for the cast iron by tests performed by Harris and ORourke (1983). Figure A-4 shows the increased strength of ductile iron to that of the two types of gray cast iron. The contrast is apparent between the two materials, cast iron behaving as a very brittle material with failure occurring at a low stress where as the ductile iron is able to undergo significant plastic deformation before failure. Failure in ductile iron occurs at around an elongation of approximately 15 percent. The Ductile Iron Pipe Research Association (2001) reports standard minimum strengths for ductile

55

iron pipe with a yield tensile strength of 300 MPa and an ultimate tensile strength of 420 MPa and a minimum elongation of 10 percent. The different behavior between ductile iron and cast iron is a result of the different formations of graphite within a very similar matrix. Graphite existing in the form of nodules, or spheroids, in ductile iron reduces the useful cross sectional area, in turn reducing the modulus of elasticity. The modulus of elasticity for ductile iron ranges from values of 166 to 180 GPa. Table A-5 lists the mechanical properties representative of most ductile iron pipe and suggested by the standards. Ductile iron obviously exhibits a much more flexible behavior in bending than that of cast iron. For the design of ductile iron pipe exposed to bending, the Ductile Iron Pipe Research Association (2001) recommends an allowable flexural stress of 331 MPa, or 80 percent of the minimum ultimate strength of the material. This is equivalent to applying a factor of safety of 1.25. The joining procedures for ductile iron pipe are very similar to those of the cast iron pipe joints. There are similar constructions to those of the rigid and flexible joints. Rigid joints are available for aboveground applications. They come in the form of a flanged configuration and a grooved joint. Flexible joints for ductile iron pipe are a rubber gasket push-on, bolted gland mechanical, or ball and socket joints. The construction of these joints is similar to that for cast iron pipe shown in Figure A-3 and maximum rotations are listed in Table A-6. The rubber gasket push-on joint was developed in 1956. It is a bell and spigot configuration with a rubber gasket installed around the spigot to block leakage. From Table A-6, the range of maximum rotation that the push-on joint can withstand is from 0.052 to 0.087 radians (3 to 5 degrees). The mechanical

56

joint is a bolted-gland configuration with a rubber gasket to prevent leakage. The mechanical joint, for pipe diameters from 75 to 600 mm, can withstand maximum rotations in the range from 0.035 to 0.14 radians (2 to 8 degrees). The ball and socket joint is mostly used for subaqueous installations. The design of the joint allows for the rubber gasket to remain compressed throughout the range of deflection without leakage occurring. The ball and socket joint can accommodate a rotation as large as 0.27 radians (15 degrees). Singhal (1984) presented the rotational stiffness for ductile iron rubber-gasket push-on joints for 100, 150, 200, and 250 mm diameter pipes. The moment-rotation relationships for these four sizes of pipe are shown in Figure A-5. The moment-rotation relationships are shown up to 0.7 radians (4 degrees) when metal-to-metal contact occurs creating high stresses in the joint within the joint. Maximum rotations presented by Burns (2003) are the ultimate rotations when failure was the result of the metal of the bell coming into contact with that of the spigot. For the pushon and mechanical joints, this rotation varied from 0.035 to 0.14 radians (2 to 8 degrees), depending on the diameter of the pipe. Attewell, et al. (1986) suggest assuming an initial rotation of 0.026 radians (1.5 degrees) for rubber gasket joints, which they suggest an allowable rotation of 0.044 radians (2.5 degrees).
A.2.3 Steel

Towards the end of the 19th century, the use of steel, or line, pipe was introduced into the pipeline industry primarily used in gas and oil transmission applications. Line pipe is manufactured in two different styles: welded and seamless. Welded pipe is produced with longitudinal welds to join two halves of pipe section. These welds can either be butt-welded or

57

lap-welded joints. Figure A-6 shows typical welds for steel pipe. Seamless pipe has no weld along the length and is produced by hot piercing or cupping and drawing of the material into its desired shape. Line pipe for gas and oil transmission can be manufactured in nominal diameter sizes of 50 to 1600 mm. The stress-strain behavior of steel differs from cast iron and ductile iron because of its ability to behave perfectly elastic and the existence of a definitive yield point where plastic deformation begins to occur. At the initial onset of incremental stress to both cast iron and ductile iron, there is a plastic deformation in the material that is an unrecoverable deformation. With steel, for a certain range of stress, deformation occurs that is recoverable upon removal of the stress. The range of these stresses and strains makes up the elastic region of the stress-strain curve. The boundary at which elastic deformation ceases and the strains begin to become plastic, or unrecoverable, is the yield strength of the material. The slope of the stress-strain curve prior to the material reaching the yield stress is the modulus of elasticity of the material. For structural steels, this value is approximately 200 GPa. The classification of the steel is important in determining its yield strength. For carbon steels and high strength low-alloy steels, a yield point in which there is an obvious change in slope in the stress-strain relationship represents the yield strength. With high strength steels, there is not an apparent yield point. Therefore, the yield strength is defined as the stress at a 0.2% strain offset (Salmon and Johnson, 1996). Multiple grades of steel pipe ranging in minimum yield strength, designated as the specified minimum yield strength (SMYS), are available for different applications. Steel Grade A and Grade B have minimum yield strengths of 207 MPa and 241 MPa, respectively. Other steel grades available are classified by their specified minimum yield strengths. For example,

58

Grade 290 is steel with specified minimum yield strength of 290 MPa, grade 414 is steel with specified minimum yield strength of 414 MPa. Line pipe is available in grades up to 550 with specified minimum yield strength of 550 MPa. Steel pipe is mostly used for the transmission of gas and oil at high pressures. In most cases, for determining an allowable stress for a steel pipe exposed to excavation-induced movements, a combined stress analysis, such as that developed by von Mises, needs to be considered. Within urban environments, most pipelines are maintained at a low to medium pressure. Therefore, the stresses due to internal pressure are small in comparison to those induced in the pipeline due to ground movements and a combined stress analysis is not necessary. With the assumption that the pipeline follows the ground movements, the pipeline behaves as a continuously supported beams under bending. For a continuously supported steel beam, the allowable stress value is 60 percent of the yield stress of the material. For steel pipe, this would be the specified minimum yield stress specified for the pipe. The corresponding factor of safety for calculations for the allowable bending stress for steel pipe is 1.67. Steel pipe may be joined by some of the previously mentioned methods, such as mechanical, rubber-gasket push-on, and ball and socket joints. The majority of line pipe have been joined by welding in order to create a leak-tight joint and a pipeline that behaves as a continuous beam. This method of joining line pipe can be accomplished in the field during installation. A few types of welds that can be used are the single- and double-butt weld joint, and the lap weld joint. The quality of construction of these joints defines the strength of the pipe at their location. Table A-7 lists the typical percentage of reduction of steel strength at the location of welded joints. The advantage of joining line pipe by welded joints is that there is a

59

very small reduction in the strength of the pipeline at the location of the joint. Also, the potential for leakage from the location of the joints is eliminated. The percentage of loss of strength for the three types of joints mentioned range from only 0 to 25 percent.
A.2.4 Plastic

In more recent years, plastic pipe has been implemented for the majority of pipeline installation and replacement. Plastics are solid materials with one or more polymeric substances that can be formed by flow (Plastics Pipe Institute, 1993). A common plastic used in pipelines for the transmission of gas, water, waste, electric cable, and telecommunication cable is polyethylene. Polyethylene is a compound composed of a polyethylene resin polymer with additives for stability, color, and corrosion protection. It is classified in plastics as a thermoplastic, which are plastics that have the ability to be melted down and reformed. This thermoplastic property is an important advantage for the joining of polyethylene pipe. Polyethylene was first used in pipeline applications in 1939 for the insulation of telephone cables and coaxial cables. The use of polyethylene pipe increased in the 1950s when the method of production became much safer and more economical. Today polyethylene pipe has become one of the dominant materials in pipeline construction. The behavior of polyethylene under stress is very complex. Figure A-7 shows an example of a stress-strain relationship of polyethylene under controlled conditions. The stressstrain behavior of polyethylene is greatly influenced by the duration of loading, temperature, and the environment. It does not behave elastically and the stress-strain relationship is curvilinear. There are no true elastic constants for polyethylene due to the lack of a clearly defined elastic behavior region because there is no apparent proportionality between the stress and strain values. The modulus of elasticity cannot be easily defined and there exists no sharp yield point. The

60

modulus of elasticity of polyethylene is typically determined from the slope of a line connecting the origin of the stress strain curve to the point on the curve at 2% strain. The value is very sensitive to the rate of loading due to the molecular structure of polyethylene. The slower the stress is applied to the material, the lower the value of the modulus is determined. The complex behavior of polyethylene can be attributed to its viscoelastic property. Polyethylene is made up of crystalline and amorphous regions. These two regions behave differently when under stress. The crystalline regions behave elastically to the load being applied. The amorphous areas react in a fluid-like fashion. The overall mechanical response is a combination of these two behaviors. In the design of polyethylene pipelines, the creep behavior of the material must be considered. Polyethylene is a material that has a very large plastic region and can deform plastically to great strains. When a load is initially applied to the material, an initial resulting strain quickly forms. Following this, the plastic continues to deform at a much slower rate for an indefinite amount or until failure is reached by fracture. This gradual increase in strain due to a constant load is important to consider with pipelines that are constructed with the intention of a long lifespan. Therefore, another modulus must be defined to characterize this behavior. The apparent, creep, or long-term modulus is determined from the ratio of stress to strain from the creep curve. This should be used as the modulus in design calculations due to the long life expectancies of pipelines for direct loading. The apparent modulus for polyethylene pipe varies from 140 to 210 MPa where as the tensile modulus from short-term testing varies from 700 to 900 MPa. Where bending is concerned, the outermost fibers are put in both extreme tension and compression. The modulus pertaining to bending behavior is the flexural modulus. This is the

61

ratio of the applied stress in flexure to the corresponding strain in the outermost fiber within the elastic range. This is usually the tangent modulus because of the absence of perfectly elastic behavior. For gas distribution systems, there are two different common grades of polyethylene pipe commercially available. They are designated as PE80, a medium density polyethylene, and PE100, a high density. They both are available in diameter sizes ranging from 13 to 610 mm. Table A-8 lists typical mechanical properties for both grades of polyethylene pipe. For the design of polyethylene pipe, the design standard is determined by the long-term tensile strength of the material. This is represented by the hydrostatic design basis (HDB), which are appropriate to use for an elastic analysis for design. The allowable value is determined by the application of a factor of safety of 2. The value designating the long-term allowable tensile strength of the material with application of the factor of safety is the pressure class of the polyethylene. For PE80 and PE100 the pressure ratings are half the value of the hydrostatic design basis, or 4.3 and 5.5 Mpa, respectively. There are a number of joining techniques for polyethylene pipe. There is a mechanical joint where a component called the body of the joint is placed over the ends of the pipe and held in place by a gasket and a gripping ring. There is also a bolted mechanical joint which holds the pipe ends together with a bolt ring. The more common procedure for joining polyethylene pipe is by fusion, which takes advantage of its thermoplastic property of being able to be re-melted and remolded into new shapes. There are two procedures for fusing together pipe, heat fusion and electrofusion, as illustrated in Figure A-8. The process of producing a butt joint with heat fusion is carried out by aligning the pipe ends to each other, heating the ends of the pipe to the required temperature to

62

change the plastic to a molten consistency, then forcing the ends of the pipe together and holding until the plastic has reformed into a continuous pipe. This process can be completed in the field with special fusion joint machinery. Butt joints can be formed in a much similar fashion with electrofusion, the main difference being the application of heat to the ends of the pipes. In electrofusion, the pipe is heated internally by wire coils at the interface of the joint or by conductive polymers. In this procedure, the pipes are clamped together, a current is sent through the set up and the coils are heated. The ends of the pipe become molten and join together. Once the joint is allowed to cool, the pipe sections are joined and behave as a continuous beam.

A.3 Case Study: Lurie Medical Research Center

The site at which the data for this study were acquired was the deep braced excavation for the Lurie Medical Research Center located in downtown Chicago, IL. The dimensions of the excavation were approximately 82 by 69 m. The depth of the excavation reached approximately 12.8 m. The support system consisted of a sheet pile wall with three levels of tiebacks. To the north, west, and southern directions of the site, there are two-lane city streets with multi-story buildings beyond. Located adjacent to the east wall of the site is The Prentice Womens Hospital. Underground and aboveground pedestrian tunnels connect many of the surrounding buildings. This report focuses on the underground utilities surrounding the site. Within a 20 m radius of the site, there exist a steamline, water mains, sewer mains, gas mains, and electric cables. Figure A-9 illustrates the locations of these with respect to the excavation site. Table A-9 lists the distances of the utilities from the edge of the excavation.

63

The site was heavily instrumented to study the ground movements resulting from the excavation. Surrounding the site were 150 surface points, 18 embedded soil anchors, and 30 points located on local buried utilities. The site was also equipped with 8 inclinometers to document the subsurface movements. The layout of the instrumentation surrounding the construction site is shown in Figure A-9. The survey points surrounding the site were surveyed weekly for both vertical and lateral movements. The inclinometers were also read on a weekly basis for lateral movements.
A.3.1 Ground Movements due to Excavation

The survey data collected from the surface points, utility points, and soil anchors surrounding the site was entered into Surfer 7 (Golden Software Inc., 1999), a three-dimensional surface mapping software program, to produce contour maps representative of the ground movements around the site. A multiquadric radial basis function was used for interpolating the data to create a grid map of the surface ground movements. The grid map represents the interpolated values at a user-defined spacing. From these values, the software can be used to generate a contour map to display the pattern representative of the entered data. Radial functions are functions whose values are dependent on a distance from a center point. Linear combinations of these radial functions centered at specific points throughout the range of data determine the unknown functions. The multiquadric radial basis function used by Surfer 7 is:
B(h ) = h 2 + R 2

(A-1)

where: h = relative distance from data point to node of grid R2 = smoothing factor The default value for R2 calculated by Surfer 7 is determined by the equation:

64

R2 =

L2 25N

(A-2)

where: L = the length of the diagonal of the extent of the data range N = the number of data points The data presented here is representative of the movements resulting from the excavation following the installation of the sheet pile support wall. The survey data along each wall was zeroed to the date for which the installation of the respective wall section was completely installed. Figures A-10 through A-27 present the vertical and lateral ground surface movements along the north, south, and west walls of the excavation, respectively, following the completion of three definitive stages of the excavation process that are defined in Table A-10 along with the dates pertaining to the completion for each side of the excavation. The data for the north and west walls are zeroed to the movements observed on May 29, 2002. The data for the south wall are zeroed to the data reported on May 18, 2002. The vertical movement contours show negative values for settlement. The lateral movement contours show positive values for movement towards the excavation. There is an assumed 13 mm movement both laterally and vertically at the sheet pile wall. The movements compared from the inclinometer data support this assumption. The lateral movement contours are generated solely on the information gathered from survey points located directly on the ground surface due to inaccuracy of readings from utility points and soil anchors. From the data obtained around the site, the maximum movements at the completion of excavation in both the lateral and vertical directions occurred at an average of 5 to 7 m from the edge of the excavation. The maximum vertical movements ranged from 50 to 70 mm settlement. The maximum lateral movement towards the excavation observed from the data collected ranged from 55 to 90 mm. The survey data showed the maximum lateral movements to occur at the

65

survey points furthest from the excavation, leaving undefined the behavior beyond this distance. Therefore, the contours for the lateral movement display an ongoing increase in movements beyond the survey data that may be inaccurate for the behavior that actually occurred. Table A11 shows approximate values for the maximum magnitudes of lateral and vertical movements observed along the north, south, and west walls of the excavation along with the length of the wall. Along the northern wall of the excavation there exists an underground concrete pedestrian tunnel close to the western corner. The location of this tunnel is shown in Figure A-9. The section of the tunnel running parallel to the sheet pile wall is supported on timber piles. The portion of the tunnel angled away from the excavation is supported solely by the underlying soil. From the contour plots along the north wall, the influence of this large tunnel is obvious. The movements in the northwest side of the excavation are significantly smaller than those experienced in other areas. The restraint of the ground movements by the tunnel is apparent in Figures A-10 through A-15. Along the south wall, the ground movements were heavily influenced by the construction activity throughout the excavation process. The access ramp for the equipment was positioned for access from E. Huron St., located to the south of the excavation. Therefore, much of the heavy equipment for removal of the soil traversed this street, influencing the behavior of the soil below.
A.3.2 Ground Movements at Location of Gas Mains

When looking at the effects of excavation induced ground movements on adjacent pipelines, the main concern is for the prediction and prevention of failure to local gas mains. Damage or fracture to one of these gas lines could have serious and harmful effects. Therefore,

66

the focus of this paper will be the movements imposed on the gas mains local to the site and the resulting stresses. The gas mains located around the Lurie Center site are constructed out of cast iron or ductile iron. They are shallow buried pipelines with a depth of cover of about 1 meter. In the installation process of a pipeline, the pipeline is usually bedded into the bottom of a trench and then backfilling is done to fill the trench and enclose the pipe in soil. Due to the length of the pipeline adjacent to this size of excavation and the lack of restraints to movement along the pipeline, it is assumed that the pipeline will move with the ground. Since there is a shallow depth of cover, it is assumed that the movements at the location of the pipeline below are similar to that of the surface. Therefore, the vertical and lateral ground surface movements are assumed to be the same as those of the pipeline below and are analyzed to predict what occurred in the pipeline. The maximum lateral and vertical ground movements at the gas pipelines were observed throughout the entire excavation. It is expected that there should be a gradual increase in the magnitude of maximum movement in both the lateral and vertical directions for the duration of the excavation schedule. Figure A-28 shows the maximum magnitude of the vertical movement of the ground surface at the location of the gas main adjacent to the north wall of the excavation. The gas main is located approximately 15.5 m from the north wall of the excavation and is a 150 mm diameter ductile iron pipe with mechanical joints. The positive values shown for vertical movement are the magnitudes of settlement observed during the excavation. The lateral movements at this pipeline could not be adequately determined due to insufficient data at such a great distance away from the excavation edge. The maximum settlement that was observed was approximately 35 mm. This occurs at a location approximately 50 m east of the northwest

67

corner of the excavation. From the ground surface profile, the influence of the underground pedestrian tunnel is apparent and produces a large curvature in the ground movement profile. Figure A-29 shows the maximum magnitudes of vertical, lateral and total ground surface movements at the location of the gas main adjacent to the southern wall of the excavation. The gas main is a 300 mm diameter cast iron pipe located approximately 8.1 m from the southern edge of the sheet pile support wall. Again, the positive values shown for vertical movement are the magnitudes of settlement observed during the excavation. The lateral movements are shown as positive values for movement to the north towards the excavation. The total movement is the resultant of the vertical and lateral movement. The maximum settlement and lateral movement at the location of the gas main are both approximately 55 mm. The maximum settlement occurred at 60 m west of the southeast corner. The maximum lateral movement occurred at a distance of 40 m west of the southeast corner of the southern wall of the excavation. The maximum total movement of the ground at the gas main location was 75 mm. This movement is downwards and towards the north in the direction of the excavation at a location 60 m west of the southeastern corner of the excavation. The maximum magnitudes for the vertical, lateral and total ground surface movements at the location of the gas main adjacent to the western wall of the excavation are shown in Figure A-30. The gas main is a 500 mm diameter ductile iron pipe with mechanical joints located approximately 5.5 m from the edge of the excavation. The values for vertical movement are positive to show the magnitude of the settlement observed. The lateral movements are positive for movement in the easterly direction towards the excavation. The total movement is the resultant of the vertical and lateral movement.

68

The maximum settlement at the location of the gas main adjacent to the west wall of the excavation neared 70 mm. The location of this movement occurred 30 m north of the southwest corner of the excavation. The magnitude and location of the maximum lateral movement was 83 mm at 25 m north of the southwest corner of the excavation. The maximum resultant movement of the ground surface was 110 mm, which occurred at 30 m north of the southwest corner. Overall, the gas main adjacent to the west wall of the excavation was subjected to the greatest ground movements. Along the pipeline, the movements reached a settlement of 70 mm, a lateral movement of 83 mm, and a total movement of 110 mm. Table A-12 shows the maximum movements for each wall at the location of the gas main along with its distance from the edge of the excavation. For this data, as the distance from the edge of the excavation increased, the maximum lateral and vertical movements decreased. After the completion of the excavation and the site was brought down to final grade, the movements along each wall had reached their maximum. Figures A-31 through A-33 show the ground movements along the pipeline for the data collected on this date for the north, south and west walls, respectively. The vertical movements are positive for the magnitude of the settlement and the lateral movements are positive towards the excavation. The north wall lateral movement is not shown due to lack of survey data at that distance away from the excavation. All survey data along a wall of excavation was zeroed after the completion of the installation of the sheet pile wall. This allowed for the ground movements caused by the excavation to be isolated for the analysis. The excavation is plotted to reference the location of the corners of excavation.
A.3.3 Ground Movement and Pipeline Movement Comparison

Without detailed instrumentation on the gas mains around the site, assumptions about their behavior must be made to calculate the resulting stresses. As mentioned earlier, the

69

assumption made was that the pipeline moves with the ground due to its lack of restraint and shallow depth of burial. Finite element studies and full-scale experiments have been conducted to determine the legitimacy of this assumption. Nath (1983) conducted a three-dimensional finite element study for pit cast iron pipelines adjacent to an open, unsupported trench construction. He examined 75, 150, 300, and 450 mm diameter pipes buried at a depth of 1 m in a linearly elastic medium adjacent to an unsupported trench construction to explore the relative movements between the ground and the pipeline as a function of pipe diameter. The results showed that small diameter pipes, less than 150 mm, provided little or no restraint to the movement of the ground and were actually pushed by the ground to the same shape of deformation. Therefore, the ground movement at any point could be assumed to be the movement of the pipe at that location as well. He showed that the bending moment in the pipe was shown to be directly proportional to the flexural rigidity of the pipe, EI. For large diameter pipes, 150 mm or greater, this proportional relationship was shown not to agree. Also, there was evidence of the larger diameter pipes restraining the movement of the ground. Carder, et al. (1982) conducted full-scale tests of cast iron pipelines exposed to ground movements from deep trench excavations in different types of soils. In a compressible alluvium, they instrumented and installed a 100 mm diameter cast iron pipeline with 0.75 m cover and exposed it to ground movements from an adjacent trench excavation. In their analysis, they compared the monitored ground movements with that of the movements calculated from the measured strains recorded by strain gauges along the pipeline and found them to be very similar. In a similar testing fashion, Carder and Taylor (1983) instrumented and installed another 100 mm diameter cast iron pipeline with 0.75 m cover in Boulder clay. Again a deep braced

70

trench was constructed and the ground movements and pipe strains were monitored. Calculating the movements from the strains in the pipe, they again proved to be similar to those that were observed from the ground surface. While this existing data suggest small diameter mains move with the ground for shallow amounts of embedment, the data obtained at the Lurie Center site can be used to evaluate this assumed behavior for larger diameter pipes encountered at this site (i.e. from 150 to 500 mm). The buried utilities adjacent to the Lurie Center site were instrumented with a total of 30 survey points, 7 of which were located on the adjacent gas mains. Table A-13 defines the location of the utility points with respect to the corner of the excavation. Accurate readings of only the vertical movement of the pipeline could be obtained. To determine the behavior of the pipelines and evaluate the assumption of equal pipe and ground movements, comparisons were made between the survey values obtained in the field from the pipe data only and the values returned by the contouring software based on the surface settlement points. Plots of the magnitude of vertical movement surveyed at the gas main survey point and that generated by Surfer 7 are presented in Figures A-34 through A-39 for the gas mains along the north, south, and west walls, respectively. The 150 mm ductile iron gas main along the north wall was equipped with three monitoring points to determine the settlement of the pipe, U-1, U-2, and U-3. The values from these three data sets, and the approximated behavior of the ground surface mostly agree to within 3 mm, or within the accuracy of the optical survey data. Significant differences are noted for days when large ground surface settlements occurred, but following the incident, the pipeline responds and follows the movement pattern of the ground. As seen in Figures A-34 through A-

71

36, the days where there is a large amount of ground movement, there is a similar movement in the pipe at that location. The 300 mm cast iron gas main closest to the southern wall of the excavation was monitored at two points along its distance, U-4 and U-5. The ground surface and pipeline at the same point behave in a very similar manner. The data from utility point U-4 are very similar to that of the approximated values of the ground surface settlement, with agreement again within the accuracy of the optical survey data. The data reported for the utility point U-5 show a consistent offset of about 10 mm from the beginning of the excavation. If that offset is subtracted, then agreement is again noted with the accuracy of the optical survey data. Along the west wall, there were two survey points established for monitoring the movement of the adjacent 500 mm ductile iron gas main. Unfortunately, throughout the excavation, U-7 was unable to produce any comparative data because vehicles covered this point when the optical surveys were made. Also, for approximately the first 100 days of the excavation, the utility point U-6 was unable to yield any data for the same reason. For the data that could be obtained from U-6, there is adequate agreement between the behavior of the pipeline and the ground surface movements. From the data collected for these pipes ranging from 150 to 500 mm in diameter, the settlement pattern shows to be consistent with the ground surface movement at that point. The pipelines displacement patterns for the duration of the excavation follow that of the approximated ground movements from the survey data. Therefore, the assumption that the pipes move with the ground is adequate for the stress analysis of the pipes.

72

A.4 Stress Conditions on Buried Pipelines

Pipelines are designed to remain in service for a long life span. Cast iron, ductile iron, and steel pipelines have shown to withstand decades of adverse conditions while remaining in service. During their lifetime, they are exposed to many different conditions that can result in the accumulation of different stresses and rotations. These and the addition of ground movements are what need to be analyzed when predicting the failure of a pipeline.
A.4.1 Initial Stresses

In addition to ground movement-induced stresses cause by adjacent excavation, there exist stresses as a result of a combination of hoop stresses from internal pressures, ring stresses from soil cover, traffic loads, residual stresses from the installation or adjacent construction history, and environmental effects. Due to the preexisting stress condition of buried pipes, Taki and ORourke (1984) suggest assuming an initial longitudinal bending strain in the range of 0.02 to 0.04 percent. This determined an allowable amount of incremental bending strain of 0.05 to 0.075 percent resulting from an adjacent excavation. They note that incremental strains of this magnitude would lead to stresses of approximately 45 percent the tensile strength of pit and spun cast iron. Harris and ORourke (1983) performed tests on 21 caulked joints. From their experimental results, they determined an initial rotation in the joints ranging from 0.0017 to 0.033 radians (0.1 to 1.9 degrees) with an average deviation of 0.34 radians (1.0 degree). From these results, Attewell, et al. (1986) recommend assuming an initial rotation in both lead caulked and rubber gasket joints up to 0.026 radians (1.5 degrees).

73

A.4.1.1 Hoop Stress Pipelines in urban environments transmit pressurized materials, such as gas and water. This pressure subjects the pipe material to loading in multiple directions resulting in a tensile force within the pipe wall. This tensile stress is called the circumferential, or hoop, stress, H, which is illustrated in the free-body diagram of a pipe cross section in Figure A-40 with inner radius r, wall thickness t, and internal gauge pressure p. The gauge pressure, the pressure measured above atmospheric pressure, is used in the calculation because this represents the imbalance between internal and external forces on the walls of the pipe. For calculating the hoop stress within a pipe section, the simplified method of thin wall analysis may be used with minimal error for pipe sections with an inner-radius-to-wall-thickness ratio greater than or equal to 10. For pipe materials such as cast iron, ductile iron, and steel, this is most likely the case. For plastic pipes, the ratio of inner-radius-to-wall-thickness can fall out of the range for a thin walled cylinder and a thick-walled analysis should be used. From the free-body diagram of the thin-walled pipe section, the hoop stress can be calculated by force equilibrium. From basic mechanics, the equation for the circumferential stress throughout the cross section is:

H =

pr t

(A-3)

For thin walled pipes, this stress will not vary much throughout the wall thickness and can be assumed to be constant, or uniform. For thick walled pipes, the stress may vary throughout the section. For these pipes with an internal pressure, the maximum stress occurs at the inner surface of the wall. The equation to calculate the magnitude of this tensile stress is:

74

H =

p (d o ) + (d i )
2

(d )
o

(d i )

(A-4)

where: p = internal gauge pressure do = outside diameter di = inside diameter H = maximum hoop stress Within Chicago, buried gas pipelines for supplying the population range from low- to medium-pressure. The gauge pressure for low-pressure mains is approximately 1.75 kPa and for medium-pressure it reaches about 138 kPa. Table A-14 contains results of example calculations for the adjacent gas mains at the Lurie Center for both low- and medium-pressure. A.4.1.2 Ring Stresses from Soil Cover For pipes buried at shallow depths below the ground surface, the weight of the soil above the pipe produces stresses within the cross section of the pipe. This overpressure could cause the pipe to begin to deform into an elliptical cross section. For those loads exceeding the rigidity of the pipe cross section, excessive ovalling can occur causing longitudinal cracking, or even more critically, buckling of the pipe. Field and laboratory testing have been completed to determine the magnitudes of longitudinal stresses in a pipeline following its installation due to the backfilling of the trench. Carder, et al. (1982) and Carder and Taylor (1983) conducted two similar experiments in different types of soil to determine pipeline behavior for the different construction activities for a pipeline installation. During experiments in boulder clay and compressible alluvium, they monitored 100 mm diameter cast iron pipelines buried to a depth of 0.75 m throughout their installation process. The pipelines consisted of ten pipe sections rigidly jointed together. Each pipe section was instrumented with eleven strain gauges, five located at equal spacing on the bottom and side of the pipe and the other gauge centrally located on the opposite side of the pipe. 75

The pipes were installed into trenches with depths of 0.85 m and widths of 0.5 m. The bottoms of the trenches were hand-trimmed to ensure a flat laying surface for the pipes allowing for support along the entire length of the pipeline. The individual sections of pipe were laid into the trenches and joined while in the trench with bolted sleeves. The backfilling procedures were completed in 100 mm thick layers of soil that were compacted. For the pipeline in the boulder clay, the strain gauges along the pipeline varied due to differences in the bedding of the pipe and were independent of the location of the joints. The gauges located on the side of the pipeline reflected only small strain values, with a maximum reaching 20 x 10-6. This is equivalent to a stress of 2 MPa for cast iron with a modulus of elasticity of 100 GPa. The maximum strains recorded on the bottom of the pipe were a tensile bending strain of 60 x 10-6 and a compressive bending strain of 130 x 10-6. These are equivalent to a tensile stress of 6 MPa and a compressive stress of 13 MPa for cast iron with a modulus of 100 GPa. The instrumented pipeline installed at the compressible alluvium site was also monitored during the installation procedure by strain gauges. A maximum bending strain of 40 x 10-6 was observed along the side of the pipeline equaling a stress of 4 MPa. The maximum strain recorded on the underside of the pipeline was 80 x 10-6, equivalent to a stress of 8 MPa. Similar to the other field tests, the joint locations proved to not affect the strains along the pipeline. Burns and Richard (1964) performed an analysis for an elastic circular cylindrical shell embedded in an elastic medium under a surface overpressure to develop equations for the stresses within the circular cylinder and within the medium at any point. They looked at the interaction loads between the buried cylinder and the surrounding medium, the circumferential thrust and moment in the cylinder, the displacement of the cylinder, and the stresses within the

76

surrounding medium. The analysis pertains to both rigid and flexible pipes that either have a full bond with the surrounding soil in which they move together and there is no slip or have a friction free behavior with full slippage between the pipe and the soil. For determining the stresses induced in the pipeline cross section, the properties of the soil and the material properties of the pipe must be defined. With the model suggested by Burns and Richards, the soil is assumed to behave elastically, so the elasticity constants of the soil may be used; the modulus of elasticity, Es, and Poissons ratio, , commonly taken to be 0.33. Relationships between these two parameters can determine the lateral stress ratio, K, and the constrained modulus of the soil, Ms. The relationship to determine the lateral stress ratio of the soil is: K=

(A-5)

This relationship yields a value for the lateral stress ratio of 0.5 with a Poissons ratio of 0.33. The modulus of elasticity of the soil can be determined by the Janbu tangent modulus method derived from the definition of tangent modulus, Mt (Fellenius, 1999). The definition for the tangent modulus is: ' Mt = = m r r where: = effective stress = strain induced from effective stress j = stress exponent m = modulus number r = reference stress = 100kPa = 1 atm This approach follows the basic principle of non-linear stress-strain behavior for all types of soils. The relationship between the stress and strain behavior of the soil is a function of two non1 j

(A-6)

77

dimensional parameters, the modulus number, m, and the stress exponent, j. These two parameters are unique to the soil type. For determining the values of the non-dimensional parameters m and j, the types of soil in which the pipe is embedded must be known. At the Lurie Center site, the pipes were shallowly buried at a depth of approximately 1 m. From soil samples around the site, it is noted that the upper layers of the soil consist of a medium-dense to dense sand. For sandy or silty soils, Fellenius (1999) suggests a value of 0.5 for the stress exponent, j, and a modulus number, m, in the range of 250-1000 for medium-dense to dense sands. The value of the effective stress at the top of the pipe can be calculated by computing the weight of the soil column located above the pipe. The average depth of cover of the gas mains surrounding the Lurie Center site is approximately 1 m. A conservative value of 19 kN/m3 can be taken for the unit weight of soil near the surface. Therefore, the value of the stress at the top of the gas mains is found to be approximately 19 kPa. For these values, assuming a value of 800 for the modulus number, the modulus of elasticity of the soil can be approximated as 35 MPa. The relationship for obtaining the constrained modulus, Ms, of the soil given by Burns and Richard (1964) is: Ms = M t (1 ) (1 + )(1 2 ) (A-7)

For the calculations used with the data from the Lurie Center, a Poissons ratio of 0.33 and a modulus of elasticity of the soil equal to 35 MPa, the constrained modulus equals 52 MPa. To determine equations for the thrust, moment, displacements and interaction forces for the pipe cross section due to the overpressure load, Burns and Richard (1964) use the extensional shell theory. Figure A-41 shows the sign convention used for these equations. For convenience

78

when calculating the thrust and moments, they established equations for dimensionless constants, B and C, related to the lateral stress ratio: B= 1 (1 + K ) 2 1 (1 K ) 2 (A-8)

C=

(A-9)

Constants defining the interaction between the surrounding soil and the surface of the pipe with a mean radius, R, and thickness, t, they define as the extensional flexibility ratio, UF:
UF = 2B MtR EA

(A-10)

where: EA = circumferential extensional stiffness per unit length = Et/(1-2) and the bending flexibility ratio, VF:

MtR3 VF = 2C 6EI

(A-11)

where: EI = circumferential bending stiffness per unit length = Et3/12(1-2) After the application of boundary conditions and mathematical manipulation of the Michells stress function presented by Timoshenko and Goodier (1951), equations for the thrust, moment, displacements, and interaction forces were determined for both cases of a full bond and friction free behavior between the soil and the pipe surface. These equations are presented here for explanation purposes. The derivation of these equations is thoroughly explained by Burns and Richard (1964). Assuming a full bond between the pipe and the soil, and the displacements both radially and tangentially of the pipe agree with those of the surrounding soil, the equations for the coefficients are:

79

A0 =

UF 1 UF + (B C ) C(1 UF)VF UF(C B) + 2B (1 + B)VF + C[VF + (1 B)]UF + 2(1 + C)

(A-12)

A2 =

(A-13)

B2 =

(B + CUF)VF 2B (1 + B)VF + C[VF + (1 B)]UF + 2(1 + C)

(A-14)

The equations for the interaction loads at the interface of the pipe and soil, the thrust, and the moment equations for the case where no slippage occurs at the boundary are presented, respectively: Pr = P[B(1 A 0 ) C(1 3A 2 4B2 )cos 2 ]
Tr = PC 1 + 3A 2 + 2B 2 sin 2

(A-15) (A-16) (A-17) (A-18)

N = PR [B(1 A 0 ) + C(1 + A 2 )cos 2 ] CUF C M = PR 2 (1 A 0 ) + (1 A 2 2B2 )cos 2 2 6VF

The equations for the constants for the case where there is slippage at the soil-pipe interface are: A0 = UF 1 UF + (B C ) 2VF 1 + (1 B) 2VF 1 + (3 B) 2VF 1 2VF 1 + (3 B) (A-19)

A2* =

(A-20)

B2 * =

(A-21)

Due to the friction free movement between the soil and pipe surface, there is no tangential interaction load. The equations for the radial interaction load, thrust and moment are: Pr = P[B(1 A 0 ) C(1 + 3A 2 * 4B2 *)cos 2 ] (A-22)

80

C N = PR B(1 A 0 ) + (1 + 3A 2 * 4B2 *)cos 2 3 CUF C M = PR 2 (1 A 0 ) + (1 + 3A 2 * 4B2 *)cos 2 3 6VF

(A-23)

(A-24)

Table A-15 shows the values obtained from these equations for the three adjacent gas mains to the Lurie Center excavation with respect to the angle from the horizontal. A.4.1.3 Traffic Loads Pipelines within an urban environment mostly run below roadways and sidewalks exposing them to additional stresses due to the traffic above. Pocock, et al. (1980) subjected a 100 mm diameter cast iron pipe to static and rolling wheel loads of varying sizes and weights. For the static wheel load, they obtained maximum tensile and compressive strains of 20 x 10-6 and 10 x 10-6 respectively for a well-bedded pipeline. This corresponds to tensile and compressive stresses of 2 MPa and 10 MPa, respectively, for a cast iron pipe with modulus of elasticity of 100 GPa. For a poorly bedded pipeline, the strains were about double that of the well-bedded pipeline. A similar comparison was made for the strains resulting from the rolling wheel tests. Watkins and Anderson (2000) suggest either using the Boussinesq or Newmark method for determining the maximum vertical soil stress at the location of a shallow buried pipe. The Boussinesq method equation that defines the vertical soil stress below a concentrated load at the top of a buried pipe is:

NW H2

(A-25)

where: W = concentrated wheel load H = depth of pipe N = Boussinesq coefficient

81

The load may be considered a concentrated load if the depth is greater than three times the diameter or length of the area of the surface being loaded. The value of the coefficient is dependent on the ratio of the depth of the top of the pipe to the horizontal distance of the top of the pipe to the point of application of the wheel load. The value of the coefficient can be found by:
N= 3H 5 2R 5

(A-26)

where: N = Boussinesq coefficient H = depth of pipe R = horizontal distance from top of pipe to application of load Clearly, from this equation, the vertical soil stress at the top of the pipe decreases as its distance from the concentrated load increases. The maximum stress, when the wheel load is directly over the pipe, is approximately 47.7 percent of the load divided by the square of the depth of the pipe The Newmark method was developed by an integration of the Boussinesq method of infinitesimal areas over an area to apply this concept for distributed loads. The resulting equation determines the vertical soil stress at a point below the surface under the corner of an applied rectangular load greater in length than width. The equation resulting from the integration is:

v = Mq
where: v = vertical soil stress q = uniform pressure M = Newmark coefficient - dependent on load dimensions and depth of pipe

(A-27)

82

For calculating the stresses in the gas mains adjacent to the Lurie Center site due to traffic loads, a worst case of heavy traffic can be assumed in order to be conservative. Watkins and Anderson (2000) give standard dimensions and weights for a HS-20 truck. They give a total load of the vehicle to be approximately 142 kN with each wheel transmitting 71 kN downwards. For a tire with a pressure of 720 kPa, the dimensions of tire contact are approximately 560 by 180 mm. The results for the analyses by both the Boussinesq method and the Newmark method for a wheel located directly over the top of the pipe are 37 kPa and 43 kPa, respectively. A.4.1.4 Stresses from the Installation Procedure and Adjacent Construction History Installation procedures for buried pipelines have proven to cause initial stresses within the pipeline. Pocock, et al. (1980) conducted field experiment on a 100 mm diameter cast iron pipeline composed of eight pipe sections joined by lead caulked joints at a depth of approximately 0.6 m in compacted London clay. The bedding of the pipeline varied to impose variation in stresses along the pipeline. The backfilling of the trench was completed in 100 mm deep sections. Where there was poor bedding underneath the pipeline, the maximum tensile strain reached a value of 120 x 10-6, or a stress of 12 MPa, in the center of the pipe section. Another consideration for cast iron pipelines with semi-rigid caulked joints is the stresses within the joint upon completion of the construction of a joint. The caulking in the joints is meant to avoid leakage, however, this may put the bell and spigot under a tensile stress state. Prior (1935) performed laboratory tests on semi-rigid cast iron joints with four different caulking materials; lead, cement, leadite, and hydrotite. Strains were measured upon completion of the caulking process and after the joint underwent a bending test. The results from the tests showed that the caulking process alone contributed approximately 0.33 to 0.5 of the maximum bell

83

strains achieved in the lead caulked joints from the later applied bending test. Conversely, the cement joints showed a very small amount of strain following the caulking procedure. In addition to the construction of the pipe joint itself, an initial rotation at the joint could be a result of the laying procedure. Harris and ORourke (1983) observed 21 caulked joints after installation was completed. They noted an initial rotation at the joint ranging between 0.0017 and 0.033 rad. (0.1 to 1.9 deg.) with an average deviation from straight of 0.017 rad. (1.0 deg.). Attewell, et al. (1986) suggests assuming an initial rotation up to 0.026 rad. (1.5 deg.) for both lead caulked and rubber gasket joints. One of the noted advantages of piping materials is their durability and longevity. Cast iron and ductile iron pipes have been praised for their ability to remain in service for well over 50 years. In turn, their long lifespan exposes them to many different ground movements caused by natural effects and construction procedures. As we focus on the effects of excavation induced stresses on pipelines, these pipelines most likely have already been exposed to ground movements due to previous construction activity. These effects are not easy to evaluate and leave an element of unknown in the analysis. A.4.1.5 Environmental Effects Pipelines being buried in direct contact with the soil are exposed to environmental factors that could be limiting to the existing strength of the pipeline. Fluctuations in ground temperature can cause stresses in the pipeline due to axial deformations and the exposure of the pipe surface to corrosive soil conditions, which induce chemical reactions on the external surface of the pipe can cause weakening of the wall of the pipe. These factors differ from geographical area. Seasonal temperature changes in the weather can produce an expansive and contractive behavior within a pipe sections. For those pipelines equipped with semi-rigid or flexible joints

84

this behavior is in the form of small axial displacements at the joint. Limiting values of axial displacement for lead-caulked joints can be determined from the dimension dw of the joint shown in Table A-2. For cast iron pipes ranging in diameter of 75 to 600 mm, the maximum pull-out distance ranges from 27 to 38 mm. To avoid severe leakage from lead caulked joints, Attewell, et al. (1986) suggest a value of 75 percent of the maximum pull-out distance for lead caulked joints. Similarly, the typical values for the maximum pull-out distance, dw, for rubber gasket push-on and bolted-gland mechanical joints for pipes ranging in size of 75 to 600 mm in diameter are 38 mm and 35 to 65 mm, respectively. These values are shown in Table A-4. Due to the pipelines exposure to other stress conditions, they suggest values of 10 to 15 mm axial displacement for lead caulked joints with no initial leakage noted and 25 mm for rubber gasket joints. Carder and Taylor (1983) and Carder, et al. (1982) observed strain changes in 100 mm cast iron pipelines due to temperature changes for two different types of soil. Carder and Taylor (1983) noted for a temperature change of 6C and 10C in boulder clay fluctuations of axial strains of 15 to 125 microstrain and 50 to 20 microstrain, respectively. Carder, et al. (1982) observed a temperature change of 9C in compressible alluvium resulting in axial strains ranging from 40 microstrain. The Plastics Pipe Institute (1993) gives a comparison of the difference in axial displacement of a steel and polyethylene pipe. They give typical examples of coefficients of thermal expansion for steel and polyethylene as 1.8 x 10-5/C and 1.8 to 2.16 x 10-6/C, respectively. The equation for calculating thermal expansion is:

85

L = LT

(A-28)

where: L = change in length L = original length = coefficient of thermal expansion T = change in temperature From this equation and the typical values for the thermal expansion coefficients for steel and polyethylene, a 30.48 m unconstrained polyethylene pipeline exposed to a 12.2C temperature difference will displace axially 25.4 to 30.48 mm. A steel pipeline of the same length will only displace approximately 2.54 mm axially. Corrosion of metal pipe from contact with adverse soil conditions can cause local weaknesses in the pipe surface resulting in fissure cracking or rupture. Some methods that have been incorporated into metal pipeline installations to protect them in extreme conditions are bituminous coatings, plastic encasement, sacrificial metals, and cathodic protection. Analyses have been completed to compare the resistance of corrosion between cast iron and ductile iron. Through studies discussed by Sears (1986), ductile iron has proven to be equal or greater in resisting corrosion than cast iron. This similar behavior is mostly due to the similarities in chemical composition of the two materials. The process of corrosion for iron pipe is characterized as graphitization, where the initial contact of the pipe surface to moisture produces a graphitic layer upon the external surface of the pipe. This layer adheres to the underlying surface and provides protection for the material against the effects of the soil. Therefore, the design of the pipe must have an adequate wall thickness for graphitization to occur while leaving enough wall thickness to support other stress conditions. As the strength of iron pipe increased, the effects of corrosion grew in concern. The increase in strength of the material allowed for a decrease in the necessary pipe wall thickness.

86

For thinner pipe walls, the pipe material left intact following the graphitization process decreased leaving a weaker pipe. This also increases the chances of failure of the pipeline to pitting, a concentration of corrosion in one area making protective measures more applicable. The chemical composition of steel makes line pipe much more susceptible to corrosive behavior than cast and ductile iron. The process of graphitization in cast irons is due to the existence of carbon in the form of graphite. Carbon in steel exists in a combined form, mainly as pearlite, which does not form the protective graphitic layer observed in iron graphitization. The protective procedures mentioned earlier should be applied to steel pipe installations in highly corrosive areas. Polyethylene deteriorates in a different fashion than that of the graphitization and corrosion of the previous metals. The environmental damage to polyethylene is due to exposure to oxidizing agents or organic chemicals. Oxidizing agents can attack the surface of the plastic and cause a gradual deterioration of the material. Organic materials are absorbed into the pipe and over time can cause swelling and softening of the material. The addition of strain to these phenomena can cause environmental stress cracking failure of the pipeline.
A.4.2 Analysis of Effects of Ground Movements from Adjacent Excavations on Pipeline

Pipelines adjacent to excavations are exposed to lateral and vertical ground displacements. These ground movements can cause excessive longitudinal bending stresses in the pipe sections or large relative rotations at the joints. Table A-16 lists values for allowable amounts of bending stresses resulting from excavation-induced ground movements. Table A-17 lists allowable joint rotations from excavation-induced ground movements after an assumed initial rotation is considered.

87

There are many miles of pipeline underlying the ground surface. Many of these pipelines are within feet of each other. Installation, replacement, or repair of any of these pipelines will ordinarily require a shallow or deep trench excavation. Much work in the way of field experiments and finite element analysis has been completed to determine the effect trench excavations have on adjacent utilities to establish a method for predicting the resulting stresses and strains in pipelines. Carder, et al. (1982) performed a field experiment with a 100 mm diameter cast iron pipeline adjacent to a trench excavation in compressible alluvium. The pipeline was instrumented with strain gauges and settlement and lateral ground movements were monitored. They concluded that there were small increases in bending strain due to the adjacent trench excavation. Larger values of bending strain were observed during the backfilling with a large percentage of these being permanent. From the strain gauge measurements and the ground movements, they established a correlation between the strain and deflection in the pipeline. Carder and Taylor (1983) reported similar findings for a 100 mm cast iron pipe adjacent to a trench excavation in boulder clay. The strain gauges were located only on the top and side of the pipeline adjacent to the trench construction. The strains recorded from the strain gauges do not reflect the maximum strain values within the pipe cross section because the pipe underwent both lateral and vertical movements. The principal plane at which the maximum values occur is necessary to be determined. Ahmed, et al. (1984) conducted two- and three-dimensional finite element analyses for a deep trench excavation adjacent to a cast iron pipeline. Under the assumption that the pipeline remains in complete contact with the soil, the strains were less than the allowable strength limits

88

for pit cast iron pipe. They concluded that pipelines located adjacent to trench excavation need not be replaced if they remain in contact with the surrounding soil. Nath (1983) analyzed the effect of a shallow trench excavation on adjacent 75, 150, 300, and 450 mm diameter cast iron pipes buried at a depth of 1 m. From his results he categorized small pipes as those having diameters smaller than 150 mm, which follow the movement of the surrounding soil. Pipes with diameters of 150 mm or greater tended to not move with the ground and were shown to restrain the movement of the surrounding soil. Therefore, the assumption of equal pipe and soil movements with large diameter pipes should be used to get a conservative estimate of the stresses in the pipeline due to the adjacent excavation. Maynard and ORourke (1977) looked at the effect of various deep braced excavations in Chicago on adjacent pipelines, both parallel and perpendicular to the excavation. The point of maximum curvature in the pipeline is the most critical. They determined this to be located near the area of restraint, which in pipelines parallel to excavations is usually located near the edge of the excavation where the pipeline intersects with others. The curvature, if large enough, could cause leakage at the joint or cracking of the pipe. From the excavation that they observed in Chicago, they concluded that excessive stress in pipelines parallel to excavations would not occur for displacements of the support wall less than 100 mm, which can be related to the subsurface movements. They determined that 230 mm of movement of the retention system would most assuredly be cause of failure of the pipeline. They examined four different cast iron pipelines adjacent to the excavations in Chicago. Two of the pipelines were equipped with lead caulked joints and the other two with mechanical joints. There was an observed leakage from one of the pipelines with lead caulked joints that

89

showed a maximum joint rotation of 0.006 radians (0.34 degrees). Even though the pipelines with mechanical joints experienced larger joint rotations, leakage was not observed. Ahmed (1990) looked at the response of pipelines adjacent to deep braced excavations with a finite element model. He validated the use of a sine curve for representing the ground movements near the edge of an excavation to evaluate the stresses in a pipeline running parallel. He recommends for cast iron pipelines, that an increment of 0.05 percent strain is allowable from the ground movements. For ductile iron, he suggests a value of tolerable strain due to ground movements of 0.15 percent. Ahmed (1990) studied cast iron and ductile iron pipe joints to determine limiting values of rotation and pull-out distance. For cast iron pipe he looked at the lead-caulked joint and determined maximum allowable values for the rotation of 0.009 rad. (0.5 deg.) and an allowable pull-out distance of 19 mm. The maximum allowable values for mechanical and push-on flexible joints for ductile iron determined were 0.044 rad. (2.5 deg.) rotation and 25 mm pull-out distance. A.4.2.1 Soil Pipeline Interaction The behavior of a pipeline with relation to the ground movements can be categorized as either a flexible system or a rigid system. A perfectly flexible pipeline is one that will deform along with the ground and the pipe sections of the pipeline will be subjected to stresses in the form of longitudinal bending. The joints are considered to be rigid connections and allow for no relative rotation of the pipe sections. The deformation of the pipeline will exist as flexural strain due to bending of the pipe sections. The possible forms of failure in a flexible pipeline are longitudinal cracking, rupture, or buckling.

90

Rigid pipelines behave in the opposite manner. They continue to follow the movement of the ground, but rigid pipelines deform by relative rotations between adjacent rigid pipe sections at the joints. The joints are considered to be flexible allowing for the displacements along the pipeline. The stresses in the pipeline due to the ground movements are concentrated in the joints. Failure of rigid pipelines will occur at the joint as an excessive leakage or complete failure of the joint with metal binding. The question then becomes which is the more critical mode of failure for the pipeline: excessive bending stress or large rotations at a joint. The behavior described for the flexible and rigid pipelines is an ideal behavioral pattern. More than likely the pipeline will deform as both bending strain in the pipe section and relative rotation at the joints. This is due to the flexural and rotational stiffness of the pipe and joint materials. For a more conservative analysis, calculations for bending stress will assume the pipe to be perfectly flexible and for joint rotation calculations will assume the pipe to be perfectly rigid. A.4.2.2 Sign Convention and Definitions of Terms The local coordinate system used for the pipeline stress and joint rotation calculations presented is shown in Figure A-42. The local axes are established so that the origin is parallel to the corner of the excavation where the positive x-axis is directed towards the excavation, the positive y-axis is along the length of the pipeline towards the other corner, and the positive z-axis is directed upwards. This should set up the coordinate system for the excavation to be on the right side when looking along the positive y-axis. The lateral movements towards the excavation will be positive and settlement will be in the negative z-direction according to this local coordinate system.

91

Figure A-43 shows the designations for the differential ground movements from the ground movement profiles. The survey data points are numbered starting from the furthest negative y value in the positive y-direction. The excavation is displayed to show location of the corner to the pipeline. The variables are defined as follows: n = number of survey points along pipeline Xi = total lateral movement at survey point i (i = 1,2,3,n) Yi = distance from origin to survey point i along pipeline (i = 1,2,3,n) Zi = total vertical movement at survey point i (i = 1,2,3,n)

ji = Xj - Xi = differential lateral movement between points i and j (i,j = 1,2,3,n)


Lji = Yj - Yi = distance along pipe between points i and j (i,j = 1,2,3,n)

ji = Zj - Zi = differential settlement between points i and j (i,j = 1,2,3,n)


A.4.2.3 Calculation of Bending Stress for Flexible Pipelines Flexible pipelines follow the ground displacements with deformation in the lengths of pipe as curvature. The curvature in the pipe sections produce tensile and compressive stresses along the pipeline that, if in excess, can cause failure of the pipeline. There is important to consider for old cast iron gas mains because of their brittle failure at low strain values. In determining the maximum tensile stress along the pipeline, certain assumptions are made. For a flexible pipeline, the assumption is that the pipe deformation due to the ground movements is in the form of curvatures in the pipe sections and there is no relative rotation at the joints. The flexural behavior of the pipe sections follows the Bernoulli-Navier theory of bending. This theory assumes that for a slender beam undergoing bending, the plane normal cross sections remain plane, normal to the deflected centroidal axis, and the transverse normal

92

stresses are negligible (Baant and Cedolin, 1991). This will become the basis for the calculations of the stresses in the pipe sections. To determine the maximum value of stress along a flexible pipeline from the ground displacements, the ground displacement profiles for both lateral and vertical movement at the location of the pipeline should be established. From the assumption that a pipeline moves with the soil, the ground movement profiles represent the pipeline movement profiles. The local coordinate system should be established for the pipeline with respect to the corner of the excavation. An example of a flexible pipeline layout for the bending stress analysis is shown in Figure A-44. In determining the maximum stresses in the pipeline, the location of the maximum stress in the cross section should be considered first. Due to the combination of lateral and vertical movement along the pipeline, the maximum stress plane at a point does not necessarily exist in the plane of movement at that point and the angle of this plane continually changes along the pipeline. Figure A-45 shows the coordinate system and defines the measurement of the angle . To calculate the magnitude of the maximum tensile stress and the angle of the plane at which it exists in the pipe cross section, the Bernoulli-Navier theory of bending assumptions should be considered. From their hypothesis, they imply that the axial normal strain is:

(A-28)

where: = axial normal strain z = distance from centroidal axis of deflection curve = radius of curvature of centroidal axis of deflection curve Assuming the pipeline behaves as an elastic beam, the axial normal stress can be expressed by the expression:

93

= E =

Ez

(A-29)

where: = axial normal strain E = Young's elastic modulus z = distance from centroidal axis of deflection curve = radius of curvature of centroidal axis of deflection curve Substituting this into the expression for bending moment, a relationship between the bending moment and the radius of curvature can be established. M= EI

(A-30)

where: M = bending moment I = moment of inertia of cross section To determine a relationship between the bending moment and the deflection profile of the pipeline, assuming the slopes of the deflection curves to be small, the linearized approximation of the reciprocal of the radius of curvature yields: 1

= w"

(A-31)

where: w = curvature = second derivative of profile function with respect to length coordinate of pipeline Substituting this expression into the bending moment equation gives the equation for bending with small deflections:
M = EIw"

(A-32)

To determine the maximum principal stress due to flexure in the cross section, the equation for stress due to an applied moment accounting for both vertical and lateral movement should be used:

94

i =

M z x i M x zi I I

(A-33)

where: Mz = moment around z-axis Mx = moment around x-axis xi = distance to extreme fiber of cross section at point i under Mz zi = distance to extreme fiber of cross section at point i under Mx Substituting the equation for bending moment for the vertical and lateral movement profiles gives:

i = Ex i z"(Yi ) Ez i x" (Yi )


where: x(Yi) = curvature of lateral movement profile at point i z(Yi) = curvature of vertical movement profile at point i

(A-34)

The following equations may be used for calculating the curvature at point j, located between points i and k, of the lateral and vertical movement profiles, respectively when they are unable to be calculated directly: (X X j ) (X j X i ) 2 k ( ) (Yj - Yi ) Y Y k j = x" (Yj ) = (Yk Yi ) (Z Z j ) (Z j Zi ) 2 k ( ) (Yj - Yi ) Y Y k j = z" (Yj ) = (Yk Yi ) 2 kj ji L kj L ji L ki

(A-35)

2 kj ji L kj L ji L ki

(A-36)

Expressing Equation A-34 in terms of pipe radius, r, and the angle from the horizontal, , the expression for the stress at any point around the circumference of the pipe cross section is:

i = Er[z"(Yi )sin i + x" (Yi )cos i ]


The location of the maximum value of stress in the cross section of the pipe can be determined by finding the value of i in which the stress equation, Equation A-37, is a

(A-37)

95

maximum. The value of i for which the stress equation is a maximum is when the slope of the stress equation is equal to zero. From this, we can define an expression for determining imax:

imax = tan 1

z" (Yi ) x" (Yi )

(A-38)

This equation gives the angle at which the principal plane exists for the pipe at point i. The principal plane is the plane at which there exist the maximum tensile and compressive stresses in the pipe cross-section. Substituting this in for in the stress equation yields an expression for determining the value of maximum stress within the cross section of the pipe at point i:

imax = Er[z" (Yi )sin imax + x" (Yi ) cos imax ]


From Equations A-38 and A-39, it is obvious that for simplicity of calculations the

(A-39)

settlement and lateral movement profiles should be considered separately. The total movement, calculated as the resultant of the lateral and vertical movement, would not necessarily exist at the angle imax. The angle imax is the angle between the magnitudes of the curvature, not the angle found by the resultant of the lateral and vertical displacements of the pipeline at the point i. This would be true for movement in one direction, vertical or lateral only, but for combined movement, the differential movement of the pipeline becomes more complex. From Equation A-39, the distribution of maximum stresses along the pipeline can be calculated. Applying Equation A-39 to survey points along the length of the pipeline will give a representation of the stresses within the pipeline due to the displacement of the pipeline. When calculating the stresses along the pipeline, the distance between the points along the displacement profiles is important for smoothing out rough data. Therefore, a characteristic length, Lji, should be established to reduce the effects of highly erratic data. From the data obtained from the Lurie Center site, the stresses along the pipeline from the ground movement

96

profiles calculated by Surfer 7 and its least squares fourth order polynomial fit were compared. Figures A-46 through A-48 show the comparison of the maximum tensile stresses calculated along the pipeline for characteristic lengths of 6.1, 9.2, and 15.3 m and those calculated from the polynomial for the north, south, and west gas mains, respectively. The least squares fit represents an ideal smooth ground movement profile. Reasonably conservative values for the stresses within the pipeline were found for characteristic lengths of 9.2 m or greater. All of the stress data presented will be in conjunction with a characteristic length of 9.2 m. As the excavation proceeded at the Lurie Center site, the ground movements were shown to gradually increase along the three sides of the excavation that were monitored. Figure A-49 shows the values of the maximum tensile stress for the pipelines adjacent to the north, south, and west walls of the excavation for the duration of the excavation. As the ground movements increase the pipeline is subjected to larger curvatures to accommodate the greater movement. It follows that the tensile stresses in the pipeline will increase, essentially reaching the maximum tensile stress at the completion of the excavation. Figures A-50 through A-52 show the maximum tensile stress values along the north, south, and west pipelines at the completion of the excavation, respectively. The 150 mm diameter ductile iron pipeline located 15.5 m from the north wall of the excavation experienced tensile stresses of much lesser value than those pipelines that were adjacent to the west and south walls. This pipeline had a maximum settlement of approximately 35 mm. The resulting critical curvature in the pipeline was approximately 1.6 x 10-7 mm-1, equivalent to a radius of curvature of 6.5 x 103 m. The maximum tensile stress in the pipeline was 2.5 MPa located approximately 14 m from the northwest corner of the excavation. The allowable excavation-induced bending stress for a 150 mm ductile iron pipe is the range of 259.2 to 298.0 MPa.

97

The 300 mm diameter cast iron pipeline 8.1 m from the south wall of the excavation underwent a maximum settlement of 55 mm and a maximum lateral movement of 55 mm towards the excavation. The maximum tensile stress in the pipeline during the excavation was approximately 10 MPa at 16 m from the southeast corner of the excavation. This corresponds to a curvature at this location of 5.1 x 10-7 mm-1 and radius of curvature equal to 1.9 x 103 m. The allowable range of bending stress induced by ground movements resulting from an adjacent excavation for cast iron pipe is 19.3 to 62.1 MPa. The maximum tensile stress in the 500 mm diameter ductile iron pipeline located 5.5 m from the west wall of the excavation was 26 MPa at approximately 55 m from the northwest corner of the excavation. The curvature in the pipeline at this location was equal to 4.9 x 10-7 mm-1. The radius of curvature in the gas main was 2.1 x 103 m. Again, the range of allowable bending stresses able to be tolerated by a ductile pipeline from excavation-induced ground movements, taking into account the initial conditions, is 259.2 to 298.0 MPa. Note that the maximum stresses in the north and west gas mains, 2.5 MPa and 26 MPa, are well below the range of allowable excavation-induced bending stress values for ductile iron of 259.2 to 298.0 MPa. Also, the tensile stress in the gas main to the south of the excavation, 10 MPa, is well below that of the established ranges of allowable excavation-induced bending stresses for both pit and spun cast iron of 19.3 to 44.1 MPa and 30.3 to 62.1 MPa, respectively. A.4.2.4 Calculation of Relative Rotation at Joints for Rigid Pipeline A rigid pipeline exposed to ground movements conforms to these deformations with rotations at the joints due to the high stiffness of the pipe sections. Two modes of failure exist at this type of joint, partial failure and complete failure. Table A-3 shows the rotations where partial failure of lead caulked joints was observed. Tables A-2 and A-4 display the ultimate

98

rotational capacity of lead caulked joints, as the ratio of dc/ds, and rubber gasket joints, as max. Table A-17 show the allowable excavation-induced rotations for semi-rigid and flexible types of joints for cast iron and ductile iron. Most cast iron pipes were installed with the semi-rigid lead caulked joints. Partial failure of the joint is the existence of substantial leakage from the joint. This occurs when there is enough rotation at the joint to force the caulking from the joint causing excessive leakage. There could also be failure by fracture of the bell due to the rotation at the joint. The spigot could rotate relative enough to the bell to produce a large stress on the bell from metal binding, enough to cause a fracture of the bell. For flexible joints equipped with rubber gaskets, leakage is avoided by the flexibility of the rubber gasket within the bell. Complete failure of these joints can occur when the rupture strength of the joint is achieved and fracture of the bell occurs. Certain assumptions are made for the analysis of a pipeline that is considered rigid. It is assumed that all of the pipeline deformation is concentrated in the joints in the form of relative rotation between two pipe sections at a flexible joint. The joints behave as links between a series of very rigid segments. The pipe sections are assumed to have a very high flexural stiffness to not allow any deflection of the pipe section, imposing no curvature. Therefore, the bending stress analysis is not necessary. The joints are assumed to have no rotational stiffness allowing free rotation. The relative rotation between the pipe sections that the joint must accommodate is dependent on the differential ground movements between the ends of the pipe sections joined at the joint. To determine the relative rotation of a joint, again the pipeline should be established in its local coordinate system. As shown in Figure A-53, the pipeline will now follow the ground displacement profiles at evenly spaced intervals equal to the length of the pipe sections of the

99

pipeline. The numbered locations are the locations of the joints connecting the pipe sections. The differential displacements shown are the same as defined earlier. When calculating the relative rotation at the joint it is more convenient to calculate the three-dimensional deformation of the pipeline from separate vertical and lateral movement profiles. When considering movement separately, vector mathematics can be used to determine the rotation between two vectors representing the pipe sections. Axial strain in the pipe section is neglected, thereby assuming small axial displacement within the joint. A schematic for the calculation of the relative rotation at joint j is shown in Figure A-54. The definitions of the variables are as follows:

ji = Xj - Xi = differential lateral movement between points i and j (i,j = 1,2,3,n)


Lji = Yj - Yi = distance along pipe between points i and j (i,j = 1,2,3,n)

ji = Zj - Zi = differential settlement between points i and j (i,j = 1,2,3,n)


Xi = total lateral movement at survey point i (i = 1,2,3,n) Yi = distance from origin to survey point i along pipeline (i = 1,2,3,n) Zi = total vertical movement at survey point i (i = 1,2,3,n)

i = relative rotation at joint i (i = 1,2,3,n)


The following relationship may be used to determine the relative rotation of two pipe sections joined at joint j, j:

j = cos 1

ji kj + Lji Lkj + ji kj
2 ji

+ Lji + ji
2

) (

2 kj

+ Lkj + kj
2

(A-40)

As the ground movements gradually increase during the excavation, the joints will experience larger rotations. Figure A-55 shows the gradual increase of the maximum joint rotation for the pipelines along the north, south and west walls of the excavation for the duration 100

of the excavation. These values were calculated for the most common cast iron pipe section length of 3.6 m. Comparisons between these rotations and those for 6.1 m sections yielded very similar results, as shown in Figures A-56 through A-58 for the north, south, and west gas mains, respectively. The most critical of the three gas mains is obviously the south gas main, which has a maximum rotation of approximately 6.3 x 10-3 radians (0.343 degrees). Assuming that this cast iron pipeline was installed with lead caulked joints, for the most limiting case, the maximum rotation exceeds that observed by Maynard and ORourke (1977) of 6 x 10-3 radians (0.34 degrees) to cause excessive leakage. The mode of failure of the gas main would have been excessive leakage from one or more of the joints. To determine the maximum relative rotation along the pipeline, the values for the settlement and lateral movement for points located at 3.6 m intervals were taken and the differential movements were calculated. These values were then entered into Equation A-40 for calculating the angle between the two pipe segments. This was done along the pipeline at intervals of one-third of a meter. Figures A-59 through A-61 show the relative rotations along the north, south, and west pipeline with a joint located at the maximum at the end of the excavation along with the corresponding rigid pipeline movement profiles for reference. In Figure A-59, the relationship between the rotations at the joints and the pipeline settlement is shown along the north wall. The settlement profile is clearly influenced by the existence of the underground pedestrian tunnel. There is a clear boundary between the ground restrained by the tunnel and that not influenced which is a transition from small to large movements creating a substantial change in slope of the pipeline. This is noticed in the pipeline profile approximately 15 m east of the northwest corner of the excavation. This would be the most critical location for a joint along the pipeline. This is shown in Figure A-59(a) where the

101

maximum joint rotation along the pipeline corresponds to this location of maximum differential movements between two pipe sections from Figure A-59(b). The maximum joint rotation along the pipeline, from Figure A-59(a) is approximately 1.6 x 10-3 rad. (0.092 deg.). Comparing this rotation to the allowable rotational limits for lead caulked joints established from prior experimentation and shown in Table A-17, the relative rotation is considerably less. The allowable rotation from excavation-induced ground movements, when initial stresses are considered, for a 150 mm diameter ductile iron mechanical joint is approximately 0.095 radians (5.5 degrees). This gas main was located the furthest from the excavation, 15.5 m away, and therefore was subjected to substantially less ground movements and smaller curvature which could potentially produce large rotations in the joints. In Figure A-60, the relative rotations at the joints along the pipeline located south of the excavation with pipe sections 3.6 m long and the pipeline displacement profiles are shown. Since the movement both in the vertical and lateral directions were available, locating the point where maximum joint rotation occurs is more difficult to determine. From comparing the pipeline displacements, approximately 15 m and 60 m from the southeast corner there are large magnitudes of differential settlement. About 75 m from the southeast corner, there was a large differential lateral movement. These three locations along the pipeline show large relative rotations for a joint at this location, the maximum of these occurring at a distance 15 m from the corner. The maximum relative rotation is approximately 6.3 x 10-3 rad. (0.343 deg.), comparatively larger than that occurring in the gas main adjacent to the north wall. This value is near enough to that which has bee observed to cause failure for a lead caulked joint. As determined before, earlier in the excavation process, the maximum rotation experienced by a joint in the south main was greater than that observed to cause leakage in a lead

102

caulked joint. The larger maximum rotation of the joints is agreeable with the larger displacements incurred by the pipeline due to its closer proximity to the edge of the excavation. Along the south wall of the excavation, there was much construction activity due to the access ramp being situated to enter and exit from the street to the south. The influence of the construction traffic is apparent in the ground movement profiles and could be the cause for the large rotations. Excluding this rotation, assuming an error in the data, the rotations along the pipeline remain relatively minor, reaching a maximum of 3.5 x 10-3 rad. (0.20 deg.). This is a safe value for lead caulked joint rotations. Figure A-61 shows the relative rotation at joints along the pipeline adjacent to the west wall of the excavation and the corresponding displacements. Similar to that of the pipeline to the south, the locations of the greatest differential settlement and lateral movement are of greatest concern. This gas main was that of most concern because of it being located within 5.5 m of the sheet pile support wall. From the displacement profiles, there is a large difference in differential settlement 55 m from the southwest corner. This could be attributed to the local effect of the location of a corner in the bracing wall system. At this location there exists the maximum relative rotation of 4.1 x 10-3 rad. (0.23 deg.). For a 500 mm diameter ductile iron gas main with mechanical joints, the allowable rotation for the pipe joints to undergo from ground movements caused by an excavation is 26 x 10-3 radians (1.5 degrees). Even though the gas main underwent great amounts of displacement, the joints were within the tolerable limit. All three gas mains were exposed to substantial ground movements, however, for the ductile iron mains with flexible joints, the rotations have been shown to be less than that necessary to cause failure. For the more limiting case of lead caulked joints, the south main was the only pipeline that experienced displacements great enough to potentially cause a leakage

103

failure. The maximum relative rotation that occurred in a joint along the north, south, or west gas mains at the end of the excavation with pipe sections 3.6 m in length are 1.6 x 10-3, 5.8 x 10-3, and 4.1 x 10-3 radians (0.092, 0.33, and 0.23 degrees), respectively. These rotations are less than the minimum rotation observed to cause leakage by Maynard and ORourke (1977) of 0.006 radians (0.34 degrees), however, the gas main to the south of the excavation was observed to reach greater rotations during the excavation which could have lead to failure at the joint by excessive leakage.

A.5 Summary

The focus of this paper was the effect that ground movements due to a deep braced excavation has on adjacent buried utilities. The main objective was to determine a way for establishing reasonable limitations for maximum ground movements from deep braced excavation construction that would not cause damage to adjacent buried utilities. The data presented was from the excavation in downtown Chicago for the Lurie Medical Research Center. The site is surrounded by underground utilities carrying water, waste, gas, electric lines, and telecommunication cables. The pipelines of main concern for this case study were the gas mains due to their close proximity to the edges of the excavation and the detrimental effects if failure were to occur. Different properties of four piping materials were given for their behavior under stress: cast iron, ductile iron, steel, and polyethylene. Common engineering properties for these materials are summarized in Table A-18. Cast iron is no longer used in pipeline installation; however, there still exist many cast iron pipelines still in use today. Ductile iron, steel, and polyethylene are commonly used in pipeline replacement and installation.

104

Initially, information was gathered on the existing conditions of the gas mains surrounding the site. The information compiled included the pipe material, dimensions, depth of burial, and location with respect to the excavation limits. Further study involved the collection of information on the available dimensions, engineering properties, and joining methods of the existing piping materials and other select materials. The properties of the pipe materials are presented in Figures A-1 through A-8 and Tables A-1 through A-8. The site was monitored and the ground movements around the site that resulted from the excavation process were documented. The site was heavily instrumented on the north, south, and west sides of the excavation with over 150 survey points, shown in Figure A-9. To the east of the site was an existing building. The locations of the underground utilities within 20 m of the site are noted in Table A-9. The survey data from the monitor points was collected and input into a software program to generate contours that reflected the movements adjacent to the excavation. From the contours, ground movement profiles were obtained for movement in both the vertical and horizontal directions at the locations of the pipelines. Comparisons were done between these movements and the optical survey data obtained to study the agreement between the soil and pipeline movement. The data showed the pipes to behave similarly to that of the soil supporting the assumption that the pipelines move with the soil causing little to no restraint. The error between the observed values and those calculated by the contouring software was within the range of error for the optical survey. The ground movement profiles at the location of the pipelines were collected throughout the excavation process and the maximum magnitude of the vertical and lateral movements was observed for all three gas mains. The maximum movements at the end of the excavation for each

105

pipe are shown in Table A-12. Lateral movements at the location of the gas main to the north of the site were not measured. Calculations were performed to determine the possible initial state of the pipeline prior to its exposure to additional ground movements. Initial stresses in the pipeline may be due to internal pressure, soil cover, traffic loads, installation procedures, and environmental effects. Tables A-14 and A-15 list the results for the stresses induced by internal pressure and soil cover. The location of the concentration of the stresses is defined by the soil-pipeline interaction. The soil-pipeline interaction can be divided into two different behaviors: flexible or rigid. For the flexible model, the pipeline is assumed to consist of flexible pipe sections connected with rigid joints. The displacement of the pipe is concentrated in the pipeline as curvatures in the pipe sections. The joints allow no relative rotation between adjacent pipe sections. The pipe sections are assumed to follow the Bernoulli-Navier theory of bending, where plane sections remain plane and perpendicular to the centroidal axis of the deformed shape. The pipeline behaves as a continuous beam on a flexible foundation. Axial deformations are neglected and assumed to be small enough to exist within the joints. An equation was derived from the theory of bending to determine the stress throughout the cross section of a pipe that has undergone vertical and lateral displacements. This equation was maximized to determine the principal stress and the angle of the plane on which it acts within the pipe cross section. The equation for the angle of the principal plane and the maximum tensile stress are:

imax = tan 1

z" (Yi ) x" (Yi )

(A-38) (A-39)

imax = Er (z"sin imax + x" cos imax )

106

From Equations A-38 and A-39, the magnitude and location of the maximum tensile stress along the pipelines was obtained. This value was compared to that of the strength limitations of the pipe material to conclude whether the pipeline would fail due to the ground movements. For the pipelines adjacent to the Lurie Center site subjected to the ground movements documented during the excavation, the longitudinal bending stresses were small in comparison to the pipe strength. Table A-19 summarizes the dimensions, locations, and maximum tensile stresses in the north, south and west gas main at the completion of the excavation. The rigid model for the soil-pipeline assumes that the pipeline behaves as a set of rigid sections linked together with perfectly flexible joints. The pipe sections are assumed to be completely rigid, with a very high flexural rigidity to resist any curvature due to bending moment. The joints are assumed to be perfectly free to rotate and the displacements of the pipeline are concentrated in the joints as relative rotation between the adjacent pipe lengths. For the analysis of the joint rotations within the pipeline, the vertical and lateral movement profiles were kept separate to analyze the pipeline behavior as a series of vectors connected at the joints. The equation for calculating the angle created between the pipe sections assumes that the pipe is initially perfectly straight. The angle of rotation of one pipe section in relation to an adjacent pipe section is determined by the values of the differential movements between the pipe section ends. The equation derived for calculating the angle of rotation at joint j between two pipe sections is:

j = cos 1

ji kj + Lji Lkj + ji kj
2 ji

+ Lji + ji
2

) (

2 kj

+ Lkj + kj
2

(A-40)

This equation may be used with the pipeline displacement profiles to determine the location along the pipeline that is most critical for a joint location. For the pipelines adjacent to the Lurie Center site, the rotations along the pipeline were calculated, and the maximum rotations were 107

less than those previously observed to cause leakage at a joint. Table A-20 lists the possible maximum relative rotations at joints along the north, south and west gas mains.

A.6 Conclusions

A method established for the determination of the effects a deep braced excavation has on adjacent buried pipelines to establish reasonable limitations for maximum allowable ground movements set to prevent damage. Damage to a pipeline from large displacements can occur as either excessive bending stresses in the pipe or large rotations at the joints. The analysis was divided to consider both of these cases separately. Allowable values for the stress and joint rotations induced by the ground movements caused by the excavation were established from experimental and empirical data for comparison. The method was applied to ground movement observations collected from a deep braced excavation in downtown Chicago. The site was surrounded with utilities transporting water, waste, gas, electric lines, and telecommunication cables. Failure of any of these lines would be costly, but the danger imposed by the pressurized gas mains was the main focus. The site was surrounded on the north, south, and west side by gas mains of either cast iron or ductile iron. The sizes of the pipelines ranged from 150 to 500 mm in diameter and from distances of 5.5 m to 15.5 m. From data collected around the site, the ground surface movement distribution was interpolated in both the lateral and vertical directions. In previous studies, it has been assumed that the pipeline will move with the surrounding soil. Data from survey points located on the utilities were collected for the vertical movement of the pipe. These values and the interpolated ground surface settlements at that point were compared. The patterns of movement between the

108

ground surface and the pipeline at the same point showed very similar displacement patterns. Differences in magnitude were noted from initial offsets, but when the data was adjusted, the settlement behavior was very similar differing within the accuracy of the optical survey data. Therefore, the assumption that the pipelines move with the ground is valid for this study. Since the pipeline movement was shown to agree with the ground surface movements, the ground displacement profiles at the location of the pipelines were taken to be the displacement profiles of the pipelines. From these lateral and vertical movement profiles, the bending stress and joint rotation analyses could be completed for all three gas mains. The bending stress analysis is a conservative approach because the joints are assumed to tolerate no rotations. From the bending stress analysis for the north, south, and west gas mains, the maximum tensile bending stress in the pipeline could be calculated. The values for the maximum stresses in the north, south, and west gas mains are 2.5, 10, and 26 MPa, respectively. The north and west gas mains were constructed with ductile iron, with a limiting allowable bending stress of approximately 259.2 to 298.0 MPa, much greater than that observed to have been caused by the ground movements. The south gas mains was made of cast iron pipe, which has an allowable bending stress of approximately 19.3 to 62.1 MPa, dependent on the manufacturing method. As shown for the previous gas mains, the bending stress was substantially lower than that which is allowed for the material from excavation-induced movements. For the joint rotation analysis, the pipelines were assumed to deform by rotation at the joints, with a conservative assumption that the pipes were rigid and did not undergo bending. The north and west gas mains were installed with mechanical joints. The joining method of the south gas main was unknown, but can be assumed to be the most limiting case of lead caulked

109

joints. From the joint rotation analysis, maximum joint rotations along the north, south, and west gas mains of 0.0016, 0.0058, and 0.0041 radians (0.092, 0.33, and 0.23 degrees), respectively. For the north and west ductile iron mains with mechanical joints, these rotations were less than those allowable for excavation-induced ground movements. At the end of the excavation, the south gas main had undergone rotations at which cast iron lead caulked joints had been observed to fail. From the results of the bending stress and joint rotation analysis for the Lurie Medical Research Center, it is apparent that the more critical case of the two limiting values is that of the rotation at the joints. Cast iron pipelines with lead caulked joints are the most limiting of the pipeline applications presented. Therefore, analysis needs to be completed prior to an excavation to determine the magnitudes of stresses and joint rotations in these pipelines as a result of the ground movements. Even though the north and west pipelines, constructed out of better materials and joints, showed little effect from the ground movements, analyses should be completed for all pipelines that may be exposed to excavation-induced ground movements. This method of analysis has been applied to observed ground movements from a deep braced excavation to determine the stresses following the construction. From the established analysis, a rational method for predicting the behavior of pipelines may be developed. Given a predicted ground movement distribution, the displacements of the pipeline may be assumed from which the stresses and joint rotations may be calculated. Compared with the values for the allowable stresses and rotations induced by ground movements from adjacent excavations, failure or damage to a pipeline may be predicted. Also, more reasonable allowances for maximum ground movements for flexible excavations may be determined.

110

REFERENCES

Ahmed, I. (1990). Pipeline Response to Excavation-Induced Ground Movements. PhD thesis, Department of Civil and Environmental Engineering, Cornell University, Ithaca, NY. American Petroleum Institute. (1991). Specification for Line Pipe, 39th Ed. American Petroleum Institute, Washington, DC. Attewell, P.B., Yeates, J., and Selby, A. R. (1986). Soil Movements Induced by Tunneling and
Their Effects on Pipelines and Structures. Blackie and Son, Ltd., London.

Baant, Z. P. and Cedolin, L. (1991). Stability of Structures: Elastic, Inelastic, Fracture, and
Damage Theories. Oxford University Press, New York, NY.

Bonds, R. W. (2003). Ductile Iron Pipe Joint and Their Uses. Ductile Iron Pipe Research Association, Birmingham, AL. Burns, J. Q. and Richard, R. M. (1964). Attenuation of Stresses for Buried Cylinders.
Proceedings of the Symposium on Soil-Structure Interaction, University of Arizona,

Tucson, AZ, 378-392. Carder, D. R., Taylor, M. E., and Pocock, R. G. (1982). Response of a Pipeline to Ground Movements Caused by Trenching in Compressible Alluvium. Department of the
Environment Department of Transport, TRRL Report LR 1047, Transport and Road

Research Laboratory, Crowthorne. Carder, D. R. and Taylor, M. E. (1983). Response of a Pipeline to Nearby Trenching in Boulder Clay. Department of the Environment Department of Transport, TRRL Report LR 1099, Transport and Road Research Laboratory, Crowthorne.

111

Cast Iron Pipe Research Association. (1952). Handbook of Cast Iron Pipe, 2nd Ed. Cast Iron Pipe Research Association, Chicago, IL. Clough, G. W., and ORourke, T. D. (1990). Construction-Induced Movements of In Situ Walls. Proceedings of Earth Retaining Structures, American Society of Civil Engineers Special Conference, Ithaca, NY, 439-470. Croft, J. E., Menzies, B. K., and Tarzi, A. I. (1977). Lateral Displacement of Shallow Buried Pipelines due to Adjacent Deep Trench Excavations. Geotechnique, 27(2), 161-179. Ductile Iron Pipe Research Association. (2001). Ductile Iron Pipe. Ductile Iron Pipe Research Association, Birmingham, AL. Fellenius, B. H. (1999). Basics of Foundation Design, 2rd Ed. BiTech Publishers Ltd., Richmond, B.C. Golden Software, Inc. (1999). Surfer 7 Users Guide: Contouring and 3D Surface Mapping for
Scientist and Engineers. Golden Software, Inc., Golden, CO.

Harris, C. W. and ORourke, T. D. (1983). Response of Jointed Cast Iron Pipelines to Parallel Trench Construction. Geotechnical Engineering Report 83-5, School of Civil and Environmental Engineering, Cornell University, Ithaca, NY. Hetenyi, M. (1949). Beams on an Elastic Foundation. University of Michigan Press, Ann Arbor, MI. Hseih, P. G. and Ou, C. Y. (1998). Shape of Ground Surface Settlement Profiles Caused by Excavation. Canadian Geotechnical Journal, 35(6), 1004-1017. Janbu, N. (1963) Soil Compressibility as Determined by Oedomoeter and Triaxial Test.
Proceedings 4th European Conference on Soil Mechanics and Foundation Engineering,

1, Wiesbaden, 19-25.

112

Kyrou, K. (1980). The Effect of Trench Excavation-Induced Ground Movements on Adjacent Buried Pipelines. PhD Thesis, Department of Civil Engineering, University of Surrey, UK. Maynard, T. R. and ORourke, T. D. (1977). Soil Movement Effect on Adjacent Public Facilities. Preprint No. 3111, ASCE Annual Meeting, San Francisco, CA. Nath, P. (1983). Trench Excavation Effects on Adjacent Buried Pipes: Finite Element Study.
Journal of Geotechnical Engineering, ASCE, New York, NY, 109(11), 1399-1415.

Ou, C. Y., Chiou, D. C., and Wu, T. S. (1996). Three-Dimensional Finite Element Analysis of Deep Excavations. Journal of Geotechnical Engineering, ASCE, New York, NY, 122(5), 337-345. Plastics Pipe Institute. (1993). Engineering Properties of Polyethylene. PPI Handbook of
Polyethylene Piping, Plastics Pipe Institute, Washington, DC.

Plastics Pipe Institute. (2003). Specifications, Test Methods and Codes for Polyethylene Piping Systems. PPI Handbook of Polyethylene Piping, Plastics Pipe Institute, Washingtion, DC. Plastics Pipe Institute. (2000). Model Specification for Polyethylene Plastic Pipe, Tubing and
Fittings for Water Mains and Distribution, Plastics Pipe Institute, Washington, DC.

Prior, J. C. (1935). Deflection Test of Cast-Iron Water Pipe Bell and Spigot Joints with Special Reference to Bell Strength. Bulletin 87, College of Engineering, Ohio State University, Columbus, OH. Reese, D. J. (1949). Symposium on Nodular Graphite Cast Iron. Transactions, American Foundrymens Society, 57, 576-587.

113

Salmon, C. G., and Johnson, J. E. (1996). Steel Structures: Design and Behavior, Emphasizing
Load and Resistance Factor Design, 4th Ed., HarperCollins College Publishers, New

York, NY. Schlick, W. J. and Moore, B. A. (1936). Strength and Elastic Properties of Cast Iron in Tension, Compression, Flexure, and Combined Tension and Flexure. Bulletin 127, Iowa Engineering Experiment Station, Ames, IA, 1-95. Sears, E.C. (1968). Comparison of the Soil Corrosion Resistance of Ductile Iron Pipe and Gray Cast Iron Pipe. Materials Protection, 7(10), 33-36. Singhal, A. C. (1984). Behavior of Jointed Ductile Iron Pipelines. Journal of Transportation
Engineering, ASCE, New York, NY, 110(2), 235-250.

Taki, H. and ORourke, T. D. (1984). Factors Affecting the Performance of Cast Iron Pipe.
Geotechnical Engineering Report 84-1, School of Civil and Environmental Engineering,

Cornell University, Ithaca, NY. Tarzi, A. I., Menzies, B. K., and Crofts, J. E. (1979). Bending of Jointed Pipelines in Laterally Deforming Soils. Geotechnique, 29(2), 203-206. Timoshenko, S. and Goodier, J. N. (1951). Theory of Elasticity, McGraw-Hill Book Company, Inc., New York, NY. Untrauer, R. E., Lee, T. T., Sanders, Jr., W. W., and Jawad, M. W. (1970). Design Requirements for Cast Iron Soil Pipe. Bulletin 199, Engineering Research Institute, Iowa State University, Ames, IA. Watkins, R. K. and Anderson, L. R. (2000). Structural Mechanics of Buried Pipes, CRC Press, New York, NY.

114

TABLES

Engineering Property
Ultimate Stress, MPa Initial Tangent Modulus, GPa Strain at Failure, % Poisson's Ratio

Vertically Centrifugally Cast Pit Cast 145 20 207 20 83 14 114 14 0.45 0.2 0.55 0.2 0.26 0.26

Table A-1 Tensile Strength and Stress-Strain Properties for Cast Iron

Lead Caulked Joint Dimensions Nominal dC /dS dS dW dL dC Diameter mm mm mm mm mm rad. (deg.) 75 100 150 200 250 300 350 400 450 500 600 89 89 89 102 102 102 114 114 114 114 114 27 27 27 32 32 32 35 35 35 35 38 44 44 44 51 51 51 57 57 57 57 63 10 10 10 10 10 10 10 12 12 12 12 0.1 (6) 0.1 (6) 0.1 (6) 0.09 (5) 0.09 (5) 0.09 (5) 0.09 (5) 0.09 (5) 0.09 (5) 0.09 (5) 0.09 (5)

Table A-2 Typical Dimensions for Lead Caulked Joints (adapted from Cast Iron Research Association, 1952; Attewell, et al., 1986)

115

Nominal Diameter mm 75 100 150 200 250 300 450 600

Rubber-Gasket Push-on Joint dS dW max rad. (deg.) mm mm 82 82 82 89 96 104 107 125 38 38 38 38 38 38 38 38 0.09 (5) 0.09 (5) 0.09 (5) 0.09 (5) 0.09 (5) 0.09 (5) 0.07 (4) 0.07 (4)

Bolted-Gland Mechanical Joint dS dW max rad. (deg.) mm mm 76 82 89 93 93 100 108 35 45 50 50 50 55 65 0.07 (4) 0.07 (4) 0.07 (4) 0.07 (4) 0.07 (4) 0.07 (4) 0.07 (4)

Table A-3 Typical Dimensions for Flexible Joints for Cast Iron Pipe (adapted from Attewell, et al., 1986)

Source Untrauer, et al. (1970) Maynard and O'Rourke (1977) O'Rourke and Trautmann (1980) Harris and O'Rourke (1983) Attewell, et al. (1986)

rad.
0.016 0.006 0.009 0.0094 0.017

deg.
0.9 0.34 0.5 0.54 1.0

Table A-4 Experimental Results for Rotation at Leakage for Lead Caulked Cast Iron Pipe Joints

116

Mechanical Properties of Ductile Iron


MinimumYield Strength, MPa Minimum Tensile Strength, MPa Minimum Elongation, % Modulus of Elasticity, GPa Poisson's Ratio 300 420 10 166-180 0.28

Table A-5 Tensile Strength and Stress-Strain Properties for Ductile Iron

Nominal Diameter (mm) 75 100 150 200 250 300 350 400 450 500 600

Push-on Joint max deg. rad. 0.09 0.09 0.09 0.09 0.09 0.09 0.05 0.05 0.05 0.05 0.05 5 5 5 5 5 5 3 3 3 3 3

Mechanical Joint max deg. rad. 0.14 0.14 0.12 0.09 0.09 0.09 0.06 0.06 0.05 0.05 0.03 8 8 7 5 5 5 3.5 3.5 3 3 2

Table A-6 Typical Dimensions for Flexible Joints for Ductile Iron Pipe (after Burns, 2003)

117

Welded Joint
Butt Welded Joint 1 Single Welded Lap Joint 2 Double Welded Lap Joint

Percent Reduction
0 25 20

1 For a full-penetration weld through thickness of pipe. 2 For a gap smaller than 3.2 mm.

Table A-7 Strength Reductions at Location of Welded Line Pipe Joints (Watkins and Anderson, 2000)

Property
Yield Tensile Strength, MPa Ultimate Tensile Strength, MPa Hydrostatic Design Basis, MPa Elongation at Failure, % Flexural Modulus, MPa Modulus of Elasticity (Short-term), MPa Modulus of Elasticity (Long-term), MPa Poisson's Ratio

PE80
15-18 28-31 8.6 > 600% 552-758 800 150 0.42

PE100
21-24 31-38 11 > 800% 758-1103 1000 150 0.42

Table A-8 Typical Mechanical Properties for Polyethylene Gas Distribution Pipe

118

Direction from Excavation


North

Utility
Pedestrian Tunnel 200 mm Water Main 600 mm Sewer Main 150 mm Ductile Iron Gas Main 200 mm Water Main Steamline 300 mm Cast Iron Gas Main 600 mm Sewer Main 300 mm Cast Iron Gas Main 300 mm Water Main 500 mm Ductile Iron Gas Main 1500 mm Sewer Main 200 mm Water Main

Distance from Excavation


1.5 m 3m 8.7 m 15.3 m 17.1 m 0.6 m 8.1 m 10.2 m 12-17.7 m* 18 m 5.5 m 7.8 m 15 m

South

West

* Gas main transitions from 12 m from excavation edge to 17.7 m from excavation edge at gradual angle.

Table A-9 Locations of Underground Utilities with Respect to Lurie Center Excavation

Stage of Construction

Work Completed
Excavation to depth of 6.71 m, installation of tiebacks at 5.49 m Excavation to depth of 10.06 m, installation of tiebacks at 9.45 m Excavation to depth of 12.8 m, subsurface structure

Stage 1 Stage 2 Stage 3

Table A-10 Definitions of Stages of Construction

119

Side of Excavation North South West

Length of Wall (m) 80 80 66

Magnitude of Max. Movement Vertical (mm) Lateral (mm) 50 70 70 60 55 90

Table A-11 Magnitudes of Maximum Ground Movements Surrounding Lurie Center Excavation

Side of Distance from Magnitude of Maximum Movement Excavation Excavation Vertical (mm) Lateral (mm) Total (mm)
North South West
15.5 m 8.1 m 5.5 m 35 55 70 -55 83 -75 110

Table A-12 Magnitudes of Maximum Ground Movements at Locations of Gas Mains Adjacent to Lurie Center Excavation

120

Survey Side of Excavation Point


North South West U-1 U-2 U-3 U-4 U-5 U-6 U-7

Reference Corner
Northwest Northwest Northwest Southeast Southeast Southwest Southwest

Distance from Corner


14.6 m 38.8 m 68.0 m 26.2 m 58.5 m 17.0 m 54.0 m

Table A-13 Locations of Utility Survey Points with Respect to Nearest Corner of Excavation

Nominal Inner Side of Radius Diameter Excavation (mm) (mm)


North South West
1 2

Thickness
(mm)

Hoop Stress, H 2 Low, kPa Medium, kPa


1

150 300 500

81 153 265

6.4 14.2 9.2

21.8 18.6 50.0

1750 1500 4000

Low pressure = 1.75 kPa Medium pressure = 138 kPa

Table A-14 Sample Hoop Stress Calculations for Gas Mains Adjacent to Lurie Center Excavation

121

Direction from Excavation


North South
Cast Iron 300 167.64 153.42 160.53 14.22 100 0.26 1.53 25.70 35 52 0.33 0.50 19 0.75 0.25 0.0082 0.692 -0.330 0.449 -0.264 0.392 0.088

West
Ductile Iron 500 274.32 265.18 269.75 9.14 180 0.28 1.79 12.43 35 52 0.33 0.50 19 0.75 0.25 0.0118 6.84 -0.328 0.220 0.251 0.840 0.759

Pipe Properties
Material Nominal Diameter, mm Outer Radius, mm Inner Radius, mm Mean Radius, mm Thickness, mm Modulus of Elasticity, GPa Poisson's Ratio Extensional Stiffness/Length, GN/m Bending Stiffness/Length, kN-m Ductile Iron 150 87.63 81.28 84.46 6.35 180 0.28 1.24 4.17 35 52 0.33 0.50 19 0.75 0.25 0.0053 0.624 -0.331 0.460 -0.287 0.372 -0.059

Soil Properties
Modulus of Elasticity, MPa Constrained Modulus, MPa Poisson's Ratio Lateral Stress Ratio Effective Stress, kN/m

Equation Constants
B C UF VF A0 A2 B2 * A 2 B*2

No Slippage
Pr , kN/m Tr, kN/m N, kN/m M, kN-m/m 18.97-3.64cos2 18.95-3.36cos2 18.93-3.16cos2 8.58sin2 8.65sin2 10.27sin2 1.60+0.59cos2 3.04+1.11cos2 5.11+1.56cos2 6.4x10 -5+0.019cos2 0.00032+0.066cos2 0.00013+0.048cos2 18.97-8.94cos2 18.95-8.67cos2 18.93-2.29cos2 1.60+0.25cos2 3.04+0.46cos2 5.10+0.21cos2 6.4x10 -5+0.021cos2 0.00032+0.074cos2 0.0013+0.055cos2

Full Slippage
Pr , kN/m N, kN/m M, kN-m/m

Table A-15 Overpressure Stresses for Gas Mains Adjacent to Lurie Center Excavation

122

Yield Ultimate Initial Stress 1 Pipe Material Strength Strength ( ), (Fy), MPa (Fu), MPa MPa 145 13.8 - 38.6 --Pit Cast Iron Spun Cast Iron 207 20.7 - 52.4 --33.1 - 71.7 420 300 Ductile Iron 331 207 41.4 - 82.8 Grade A Steel 41.4 - 82.8 414 241 Grade B Steel 41.4 - 82.8 517 414 Grade 414 Steel 0.13 - 0.26 8.6 --PE80 0.28 - 0.56 --11 PE100
INITIAL

Factor of Safety
2.5 2.5 1.25 1.67 1.67 1.67 2.0 2.0

Design Bending Allowable Stress 2 Stress ( ), MPa ( ), MPa


B ALLOW

0.4Fu 0.4Fu 0.8Fu 0.6Fy 0.6Fy 0.6Fy 3 0.5HDB 0.5HDB

58 82.8 331 124.2 144.6 248.4 4.3 5.5

19.3 - 44.1 30.3 - 62.1 259.2 - 298.0 41.4 - 82.8 62.1 - 103.4 165.4 - 206.8 4.04 - 4.17 4.94 - 5.22 to 0.04%

1 Adapted from Taki and O'Rourke (1984) assumed initial longitudinal bending strain of 0.02 2 Allowable bending stress from excavation-induced ground movement = ALLOW = B - INITIAL 3 HDB = Hydrostatic Design Basis

Table A-16 Allowable Bending Stresses from Excavation-Induced Movements

Mode of Joint Type Failure Leakage Lead-Caulked Lead Caulked Cast Iron Metal Rubber-Gasket Push-on Binding Mechanical (metal-tometal Rubber-Gasket Push-on Ductile Iron contact) Mechanical Ball and Socket Material
1 ALLOW

Failure Rotations
Radians
0.0094 - 0.016 0.09 - 1.0 0.07 - 0.09 0.07 0.05 - 0.09 0.035 - 0.14 0.22 - 0.26
2

Allowable Rotations, ALLOW 1


Radians
0.0048
3

Degrees
0.54 - 0.92 5-6 4-5 4 3-5 2-8 12.5 - 15

Degrees
0.275 3.5 - 4.5 2.5 - 3.5 2.5 1.5 - 3.5 0.5 - 6.5 11 - 13.5

0.06 - 0.08 0.044 - 0.06 0.044 0.026 - 0.06 0.009 - 0.11 0.19 - 0.24

represents allowable excavation-induced rotation with assumed 0.026 rad. (1.5 deg.) initial rotation (Attewell, et al., 1986) for flexible joints, ALLOW = METAL BINDING - INITIAL. ALLOW = LEAKAGE/F.S. for lead caulked joints where F.S. = 1.25. 2 Observed from laboratory tests to cause excessive leakage. 3 Observed from field data to cause excessive leakage (initial rotation already occurred).

Table A-17 Allowable Excavation-Induced Joint Rotations for Semi-Rigid and Flexible Cast Iron and Ductile Iron Joints

123

Pipe Material
Vertically Pit Cast Iron Centrifugally Cast Iron Ductile Iron Steel Grade A Steel Grade B Steel Grade 414 Polyethylene PE80 Polyethylene PE100

Coeff. of Ultimate Modulus of Yield Stress, Fy Stress, Fu Elasticity Poisson's Thermal Exp. Ratio (per C) (GPa) (MPa) (MPa)
145 20 207 20 420 331 413 517 31 31 83 14 114 14 166-180 200 200 200 552-758 758-1103 0.26 0.26 0.28 0.29 0.29 0.29 0.42 0.42 11 x 10-6 11 x 10-6 11 x 10-6 12 x 10-6 12 x 10-6 12 x 10-6 2 x 10 -6 2 x 10 -6

Reference
Ahmed (1990) Ahmed (1990) DIPRA (2001) API (1991) API (1991) API (1991) PPI (2003) PPI (2003)

300 207 241 414 15-18 21-24

Table A-18 Typical Engineering Properties for Piping Materials

Pipe Size (mm)


150 500 300

Material
Ductile Iron Cast Iron

Maximum Dir. from Dist. from ALLOW Excavation Excavation Tensile Stress MPA 15.5 m 259.2 - 298.0 2.5 MPa North West 5.5 m 259.2 - 298.0 26 MPa 8.1 m 19.3 - 62.1 10 MPa South

Table A-19 Dimensions and Maximum Tensile Stress Values in Gas Mains Adjacent to Lurie Center at End of Excavation

124

Pipe Size (mm)


150 500 300

Material
Ductile Iron Cast Iron

Dir. from Dist. from Excavation Excavation 15.5 m North West 5.5 m 8.1 m South

Allowable Rotation1 Maximum Joint Rotation, rad. (deg.) rad. deg. 0.095 5.5 0.0016 (0.092) 0.026 1.5 0.0041 (0.23) 0.006 2 0.34 0.0063 (0.343)
rotation already occurred)

1 Allowable amount of rotation from excavation-induced ground movements 2 Observed rotation in field to cause excessive leakage in lead caulked joint (initial

Table A-20 Dimensions and Maximum Joint Rotation in Gas Mains Adjacent to Lurie Center at End of Excavation

125

FIGURES

300 FRACTURE 250 SPUN CAST

Tensile Stress, MPa

Elastic Strain

200 150 100 50 0 0 0.1 0.2 0.3 0.4 0.5 Total Strain PIT CAST

Strain (%)

Figure A-1 Typical Stress- Strain Curves for Cast Iron (after Attewell, et al., 1986)

Run lead or lead wool compressed by caulking tools Jute yarn packed tight
dC

dL

dW dS

Figure A-2 Typical Lead Caulked Cast Iron Joint

126

Gland Ring Rubber Gasket Rubber Gasket

dW dS dS

dW

Rubber-Gasket Push-on Joint

Bolted-Gland Mechanical Joint

Figure A-3 Typical Flexible Iron Joints

483
DUCTILE IRON

414

Tensile Stress, MPa

345 276 207 138 69 0


PIT CAST FRACTURE SPUN CAST

0.1

0.2

0.3

0.4

0.5

0.6

12

13

14

15

Axial Strain (%)

Figure A-4 Stress-Strain for Cast Iron and Ductile Iron (after Ahmed , 1990)

127

250

250 mm

200

Moment, kN-mm

200 mm 150

100

50

150 mm 100 mm 1 2 3 4 5

Rotation, deg.

Figure A-5 Rotational Stiffness of Ductile Iron Rubber Gasket Joints (Singhal, 1984)

Single-Butt Weld Joint

Double-Butt Weld Joint


Outside weld

Inside weld may be substituted for outside weld

Lap-Welded Slip Joint

Figure A-6 Typical Joint Welds for Line Pipe

128

Fracture Ultimate Strength

Yield Stress

Stress

Strain
Figure A-7 Stress-Strain Curve for Polyethylene Under Controlled Conditions (Plastics Pipe Institute, 1993)

Typical Heat Fusion Butt Joint


Fitting

Wire coils

Typical Electrofusion Joint

Figure A-8 Typical Joining Methods for Polyethylene Pipe (Plastic Pipe Institute, 1998)

129

E. Superior St.

Existing Pedestrian Tunnel

LURIE MEDICAL RESEARCH CENTER EXCAVATION


Prentice Women's Hospital

LEGEND
Water Main Gas Main Electric Line Sewer Main Steamline Surface Point Utility Point Soil Anchor Inclinometer N. Fairbanks Ct.

0 2 4 8 Scale in meters

E. Huron St.

Figure A-9 General Layout of Lurie Center Site Instrumentation and Adjacent Underground Utilities

130

0
Existing Underground Pedestrian Tunnel

-25

Excavation
Denotes Survey Point Contour Values Shown in Millimeters Contour Interval = 5 mm 0 5 10 15 20m

Figure A-10 Vertical Ground Movements Along North Wall on Day 146
Data Re-zeroed for Completion of Wall Installation

Existing Underground Pedestrian Tunnel

25

Excavation
Denotes Survey Point Contour Values Shown in Millimeters Contour Interval = 5 mm 0 5 10 15 20 m

Figure A-11 Lateral Ground Movements Along North Wall on Day 146
Data Re-zeroed for Completion of Wall Installation

131

-25
Existing Underground Pedestrian Tunnel

-25

Excavation
Denotes Survey Point Contour Values Shown in Millimeters Contour Interval = 5 mm 0 5 10 15 20m

Figure A-12 Vertical Ground Movements Along North Wall on Day 192
Data Re-zeroed for Completion of Wall Installation

Existing Underground Pedestrian Tunnel

25

Excavation
Denotes Survey Point Contour Values Shown in Millimeters Contour Interval = 5 mm 0 5 10 15 20 m

Figure A-13 Lateral Ground Movements Along North Wall on Day 192
Data Re-zeroed for Completion of Wall Installation

132

-25
Existing Underground Pedestrian Tunnel

-50 -25

Excavation
Denotes Survey Point Contour Values Shown in Millimeters Contour Interval = 5 mm 0 5 10 15 20m

Figure A-14 Vertical Ground Movements Along North Wall at End of Excavation
Data Re-zeroed for Completion of Wall Installation

Existing Underground Pedestrian Tunnel

50 25

Excavation
Denotes Survey Point Contour Values Shown in Millimeters Contour Interval = 5 mm 0 5 10 15 20 m

Figure A-15 Lateral Ground Movements Along North Wall at End of Excavation
Data Re-zeroed for Completion of Wall Installation

133

Excavation

-25 0

-25

Denotes Survey Point

Contour Values Shown in Millimeters Contour Interval = 5 mm 0 5 10 15 20 m

Figure A-16 Vertical Ground Movements Along South Wall on Day 157
Data Re-zeroed for Completion of Wall Installation

Excavation

25 0

Denotes Survey Point

Contour Values Shown in Millimeters Contour Interval = 5 mm 0 5 10 15 20 m

Figure A-17 Lateral Ground Movements Along South Wall on Day 157
Data Re-zeroed for Completion of Wall Installation

134

Excavation
-25

-25 Denotes Survey Point Contour Values Shown in Millimeters Contour Interval = 5 mm 0 5 10 15 20 m

Figure A-18 Vertical Ground Movements Along South Wall on Day 203
Data Re-zeroed for Completion of Wall Installation

Excavation
25

25 0

Denotes Survey Point

Contour Values Shown in Millimeters Contour Interval = 5 mm 0 5 10 15 20 m

Figure A-19 Lateral Ground Movements Along South Wall on Day 203
Data Re-zeroed for Completion of Wall Installation

135

Excavation
-25

-25 -50 -25

Denotes Survey Point

Contour Values Shown in Millimeters Contour Interval = 5 mm 0 5 10 15 20 m

Figure A-20 Vertical Ground Movements Along South Wall at End of Excavation
Data Re-zeroed for Completion of Wall Installation

Excavation
25
50

Denotes Survey Point

Contour Values Shown in Millimeters Contour Interval = 5 mm 0 5 10 15 20 m

Figure A-21 Lateral Ground Movements Along South Wall at End of Excavation
Data Re-zeroed for Completion of Wall Installation

136

-25

25

Excavation

Denotes Survey Point Contour Values in Millimeters Contour Interval = 5 mm 0 5 10 15 20 m

Denotes Survey Point Contour Values in Millimeters Contour Interval = 5 mm 0 5 10 15 20m

Figure A-22 Vertical Ground Movements Along West Wall on Day 146
Data Re-zeroed for Completion of Wall Installation

Figure A-23 Lateral Ground Movements Along West Wall on Day 146
Data Re-zeroed for Completion of Wall Installation

137

25

Excavation

-25

0 25 0 75 50

-25

Excavation

Denotes Survey Point Contour Values in Millimeters Contour Interval = 5 mm 0 5 10 15 20 m

Denotes Survey Point Contour Values in Millimeters Contour Interval = 5 mm 0 5 10 15 20m

Figure A-24 Vertical Ground Movements Along West Wall on Day 185
Data Re-zeroed for Completion of Wall Installation

Figure A-25 Lateral Ground Movements Along West Wall on Day 185
Data Re-zeroed for Completion of Wall Installation

138

Excavation

-50

-25

0 25 0 75 50

-50

-25

Excavation

Denotes Survey Point Contour Values in Millimeters Contour Interval = 5 mm 0 5 10 15 20 m

Denotes Survey Point Contour Values in Millimeters Contour Interval = 5 mm 0 5 10 15 20m

Figure A-26 Vertical Ground Movements Along West Wall at End of Excavation
Data Re-zeroed for Completion of Wall Installation

Figure A-27 Lateral Ground Movements Along West Wall at End of Excavation
Data Re-zeroed for Completion of Wall Installation

139

Excavation

40

Maximum Vertical Movement (mm)

35 30 25 20 15 10 5 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation

Figure A-28 Maximum Magnitude of Vertical Movement During Excavation at Location of Gas Main Adjacent to North Wall
80

Maximum Ground Movement (mm)

70 60 50 40 30 20 10 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation


Settlement
Lateral Movement
Total Movement

Figure A-29 Maximum Magnitudes of Movements During Excavation at Location of Gas Main Adjacent to South Wall

140

Maximum Ground Movement (mm)

120 100 80 60 40 20 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation


Settlement

Lateral Movement

Total Movement

Figure A-30 Maximum Magnitudes of Movements During Excavation at Location of Gas Main Adjacent to West Wall

35

Ground Surface Settlement (mm)

30 25 20 15 10 5 0 -20 0 20

Excavation
40 60 80 100

Distance East from Northwest Corner (m)

Figure A-31 Re-zeroed Settlement Values Along North Gas Main at Completion of Excavation

141

80

Ground Surface Movement (mm)

70 60 50 40 30 20 10 0 -10 -20 0 20

Excavation
40 60 80 100

Distance West from Southeast Corner (m)


Settlement

Lateral Movement

Total Movement

Figure A-32 Re-zeroed Ground Movement Values Along South Gas Main at Completion of Excavation

Ground Surface Movement (mm)

120 100 80 60 40 20 0 -20 -20 -10 0 10

Excavation
20 30 40 50 60 70 80

Distance North from Southwest Corner (m)


Settlement

Lateral Movement

Total Movement

Figure A-33 Re-zeroed Ground Movement Values Along West Gas Main at Completion of Excavation

142

12 10

Settlement (mm)

8 6 4 2 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation


Ground Settlement Values from Contours

Survey Data from Pipe

Figure A-34 Comparison of Utility Survey Point U-1 Movement with Approximated Ground Movement Values

30 25

Settlement (mm)

20 15 10 5 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation


Ground Settlement Values from Contours

Survey Data from Pipe

Figure A-35 Comparison of Utility Survey Point U-2 Movement with Approximated Ground Movement Values

143

25 20

Settlement (mm)

15 10 5 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation


Ground Settlement Values from Contours

Survey Data from Pipe

Figure A-36 Comparison of Utility Survey Point U-3 Movement with Approximated Ground Movement Values

45 40 35

Settlement (mm)

30 25 20 15 10 5 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation


Ground Settlement Values from Contours

Survey Values from Pipe

Figure A-37 Comparison of Utility Survey Point U-4 Movement with Approximated Ground Movement Values

144

80 70

Settlement (mm)

60 50 40 30 20 10 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation


Ground Settlement Values from Contours

Survey Data from Pipe

Figure A-38 Comparison of Utility Survey Point U-5 Movement with Approximated Ground Movement Values

60 50

Settlement (mm)

40 30 20 10 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation


Ground Settlement Values from Contours

Survey Data from Pipe

Figure A-39 Comparison of Utility Survey Point U-6 Movement with Approximated Ground Movement Values

145

H r p t

Figure A-40 Free-Body Diagram of Forces Resulting from Internal Pressure

M w

Pr Tr v

Figure A-41 Sign Convention for Thrust, Moment, Displacement, and Stress Equations (after Burns and Richard, 1964)

146

Excavation
y Buried pipe y x z

Figure A-42 Local Coordinate System Convention for Pipeline Analysis

Y4

4
X4

5 6

Zi Xi

32
1

2
Lateral Movement Profile z Settlement Profile

8
y

32

3 4 L43

Z4

6 5

Figure A-43 Definitions of Dimensions and Differential Ground Movement Designations

147

Y4

4
X4

5 6

Zi Xi

32
1

2
Pipeline Lateral Movement Profile F = x(y) z Pipeline Settlement Profile F = z(y)

8
y

32

3 4 L43

Z4

6 5

Figure A-44 Pipeline Profiles with Established Local Coordinate System for Analysis of Bending Stress for Flexible Pipeline

r y

Figure A-45 Pipe Cross Section and Sign Convention

148

Maximum Tensile Stress (MPa)

4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation

6.1 m

9.2 m

15.3 m

Least Squares Polynomial

Figure A-46 Comparison of Characteristic Lengths of 6.1, 9.2, and 15.3 m for Stress Analysis Along Gas Main Adjacent to North Wall During Excavation

25

Maximum Tensile Stress (MPa)

20 15 10 5 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation


6.1 m

9.2 m

15.3 m

Least Squares Polynomial

Figure A-47 Comparison of Characteristic Lengths of 6.1, 9.2, and 15.3 m for Stress Analysis Along Gas Main Adjacent to South Wall During Excavation

149

50

Maximum Tensile Stress (MPa)

45 40 35 30 25 20 15 10 5 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation


6.1 m

9.2 m

15.3 m

Least Squares Polynomial

Figure A-48 Comparison of Characteristic Lengths of 6.1, 9.2, and 15.3 m for Stress Analysis Along Gas Main Adjacent to West Wall During Excavation

30

Maximum Tensile Stress (Mpa)

25 20 15 10 5 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation


North Pipeline

South Pipeline

West Pipeline

Figure A-49 Maximum Tensile Stress in Pipelines Adjacent to North, South, and West Walls During Excavation

150

3 Maximum Tensile Stress (MPa) 2.5 2 1.5 1 0.5 0 -0.5 -10 0 10 20 30 40 50 60 70 80 90 Distance East from Northwest Corner (m)

Figure A-50 Maximum Tensile Stress Along the North Gas Main at Completion of the Excavation

10

Maximum Tensile Stress (MPa)

9 8 7 6 5 4 3 2 1 0 -10 0 10 20 30 40 50 60 70 80 90

Distance West from Southeast Corner (m)

Figure A-51 Maximum Tensile Stress Along the South Gas Main at Completion of the Excavation

151

30

Maximum Tensile Stress (MPa)

25 20 15 10 5 0 -10 0 10 20 30 40 50 60 70 80

Distance North from Southwest Corner (m)

Figure A-52 Maximum Tensile Stress Along the West Gas Main at Completion of the Excavation

x(y)

y Zi

Y4

4
X4

5 6

Xi

32
1

2
Pipeline Lateral Movement Profile z Pipeline Settlement Profile

8
y

32

3 4 L43

Z4

6 5
z(y)

Figure A-53 Pipeline Profiles with Established Local Coordinate System for Analysis of Joint Rotations for Rigid Pipeline

152

kj
i

ji
L ji j j

L kj

ji

kj

Figure A-54 Schematic of Joint Rotation at Joint j of Rigid Pipeline

Maximum Rotation (x10-3 rad.)

6 5 4 3 2 1 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation


North Pipeline

South Pipeline

West Pipeline

Figure A-55 Maximum Relative Rotation Encountered Along North, South, and West Pipelines During Excavation for 3.6 m Pipe Sections

153

Maximum Rotation (x10-3 rad.)

1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation


3.6 m

6.1 m

Figure A-56 Comparison of Maximum Relative Rotation in North Gas Main for 3.6 and 6.1m Pipe Sections

Maximum Rotation (x10-3 rad.)

7 6 5 4 3 2 1 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation


3.6 m

6.1 m

Figure A-57 Comparison of Maximum Relative Rotation in South Gas Main for 3.6 and 6.1m Pipe Sections

154

Maximum Rotation (x10-3 rad.)

5 4 3 2 1 0 0 50 100 150 200 250 300

Days from Completion of Sheet Pile Wall Installation


3.6 m

6.1 m

Figure A-58 Comparison of Maximum Relative Rotation in West Gas Main for 3.6 and 6.1m Pipe Sections

155

1.8 1.6

Relative Rotation (x10-3 rad.)

1.4 1.2 1 0.8 0.6 0.4 0.2 0 -0.2 -20 0 20 40 60 80 100

Distance East from Northwest Corner (m)

a) Joint Rotations

35 30 25 20 15 10 5 0 -20 0 20 40 60 80 100

Pipeline Settlement (mm)

Distance East from Northwest Corner (m)

b) Pipeline Settlement Figure A-59 Joint Rotations and Pipeline Settlement Along Pipeline Adjacent to North Wall at Completion of Excavation

156

Relative Rotation (x10-3 rad.)

6 5 4 3 2 1 0 -10 0 10 20 30 40 50 60 70 80 90

Distance West from Southwest Corner (m)

a) Joint Rotations
60

Pipeline Displacement (mm)

50 40 30 20 10 0 -10 0 10 20 30 40 50 60 70 80 90

Distance West from Southeast Corner (m)


Settlement

Lateral Movement

b) Pipeline Movements Figure A-60 Joint Rotations and Pipeline Movements Along Pipeline Adjacent to South Wall at Completion of Excavation

157

4.5 4

Relative Rotation (x10 -3 rad.)

3.5 3 2.5 2 1.5 1 0.5 0 -10 0 10 20 30 40 50 60 70 80

Distance North from Southwest Corner (m)

a) Joint Rotations

90

Pipeline Displacement (mm)

80 70 60 50 40 30 20 10 0 -10 0 10 20 30 40 50 60 70 80

Distance North from Southwest Corner (m)


Settlement

Lateral Movement

b) Pipeline Movements Figure A-61 Joint Rotations and Pipeline Movements Along Pipeline Adjacent to West Wall at Completion of Excavation

158

Potrebbero piacerti anche