Proceedings of the 2012 9th International Pipeline Conference IPC2012 September 24-28, 2012, Calgary, Alberta, Canada

IPC2012-90299

SEISMIC DESIGN GUIDELINES TO MITIGATE UPHEAVAL BUCKLING OF SMALL DIAMETER PIPES

Hajime SHINKAI Japan Gas Association Tokyo, Japan Yoshito HATSUDA Tokyo Gas Corporation Tokyo, Japan

Nobuhisa SUZUKI JFE Techno-Research Corporation Kawasaki, Japan

ABSTRACT This paper describes The Draft Seismic Design Guidelines to Mitigate Upheaval Buckling of Small Diameter Pipes which were reported by Ministry of Economy, Trade and Industry in 2011. Upheaval buckling of small diameter gas pipes occurred in a high-pressure gas pipeline and three middle-pressure city lines due to the strong ground motion during the 2007 Niigata Chuetsu Offshore earthquake which was a powerful earthquake with moment magnitude of 6.6. A three year research program had been conducted since 2008 to investigate the upheaval buckling behaviors and develop seismic design guidelines to ensure seismic integrity of small diameter high-pressure gas pipelines and middle-pressure main distribution lines.

INTRODUCTION The strong ground shaking during the Niigata Chuetsu Offshore earthquake in 2007 caused upheaval buckling at 15 spots of a high-pressure pipeline and three middle-pressure city lines whose diameters were 4 and smaller [1]. After the 2007 earthquake, The Committee on City Gas Utilities and Facilities Investigating the Effect of the 2007 Niigata Chuetsu Offshore earthquake (the Committee) was held, which was chaired by Prof. Katayama at Tokyo Denki University. The Committee proposed to establish The Seismic Design Guidelines for Small Diameter Pipes Mitigating Upheaval Buckling (the Guidelines) which takes into account strong ground shaking [2]. The three year research program aimed to solve the following three key issues. The first issue aimed to investigate the upheaval buckling behaviors by sand box test and field test using the 4 pipe. Beam buckling test of the 4 pipe subjected to axial compression and bending moment were also conducted

to investigate the effect of the interaction on upheaval buckling and local buckling behaviors. The second issue required to develop a finite element (FE) model which shall be stable to simulate the nonlinear upheaval buckling behaviors. The third issue sighted a simple simulation method for seismic diagnosis of pipelines including pipe bends and/or tee junctions, which method shall also be appropriate to find effective mitigations to prevent upheaval buckling induced by strong seismic waves. Details of the upheaval buckling test and validation of FEA are described in IPC2012-9028 Upheaval Buckling of Small Diameter Pipelines Induced by Strong Ground Shaking [3] and the outcomes of the research are summarized as follows: The upheaval buckling behaviors in the sand box and the test field well captured that observed after the earthquake in 2007. The FE model idealized with four-node shell elements was successfully validated. Pipe diameter, wall thickness, material tensile properties, pipe-soil interaction, and the length of straight pipe are the governing parameter of upheaval buckling. The onset strain of upheaval buckling of the 4 pipe with Round House (RH) type material was 2.29% which imply no possibility of upheaval buckling. The onset strain of upheaval buckling of the 4 pipe with Luders Elongation (LE) type material was 0.31% which tells very high possibility of upheaval buckling. Segmentation of the pipeline and the distribution line is effective to analyze their responses to seismic waves shortly therefore applicable to seismic diagnosis [4]. The length of straight line is taken to be a key parameter which prevents upheaval buckling of the pipeline and the distribution line.

Copyright 2012 by ASME

After the research program, METI drafted Seismic Design Guidelines to Mitigate Upheaval Buckling of Small Diameter Pipes (The Guidelines), in which strong seismic excitation had been taken into account.

UPHEAVAL BUCKLING OF 4 PIPE The upheaval buckling of 4 high-pressure gas pipeline is presented in Figure 1 which occurred during the 2007 Niigata Chuetsu Offshore earthquake [1]. A sand box test set-up for upheaval buckling is presented in Figure 2 and a buckled pipe after test is shown in Figure 3. The general configurations of the buckled portion in Figure 3 present very good agreement with those observed in Figure 1 [1]. And in-field upheaval buckling tests were also conducted however the details are not introduced in this paper for the lack of space. The Guidelines consists of the following five chapters and the details are described using the chapter numbers. 1. General Rules 2. Basic Concept of Seismic Design 3. Preliminary and Specific Seismic Design 4. Preliminary Seismic Design to Mitigate Upheaval Buckling 5. Specific Seismic Design to Mitigate Upheaval Buckling

After test (excavated) Figure 2 Laboratory test of upheaval buckling

1. GENERAL RULES 1.1 Scope of application The guidelines are applicable to high-pressure pipelines and middle-pressure main distribution lines to be constructed whose diameter is 4 and smaller. Remarks: (1) The 2007 Niigata Chuetsu Offshore earthquake caused upheaval buckling of 4 high-pressure pipelines at two spots, and 4 middle-pressure distribution lines at three spots, and 3 at five spots, and 2 at five spots. The upheaval buckling had not been reported before even after the M7.3 Hyogo Prefecture Nambu earthquake in 1995 and the Mm6.6 Niigata Chuetsu earthquake in 2004. (2) The operating pressure of high-pressure pipeline is defined to be higher than and equal to 1.0MPa and that of middlepressure distribution lines is lower than 1.0MPa and higher than 0.3MPa. (3) Furthermore, upheaval buckling of pipes with diameters of 6 and larger was not reported at the same distribution areas as those of 4 and smaller pipes. Hence the Committee concluded that the pipes with diameters of 4 and smaller might undergo upheaval buckling. Hence the guidelines shall be applied to high-pressure pipelines and middle-pressure main lines for city distribution with 4 and smaller diameter. 1.2 Description of technical terms The key technical terms used in the guidelines are defined as follows. Other technical terms have been defined in the JGA Guideline 2006-03. (1) Upheaval buckling Upheaval beam-mode buckling has a wave length defined with pipe bending rigidity and surrounding soil spring properties. A buckled pipe tends to deform upward with increasing axial compressive deformation, then shell-mode local buckling can be observed. Finally the pipe may break

Figure 1 Upheaval buckling of 4 pipe

Pipe setting (before backfilling)

Copyright 2012 by ASME

during post-local buckling behavior and result in leakage of gas. (see Figure 3) (2) Onset strain of upheaval buckling Onset strain of upheaval buckling corresponds to critical longitudinal pipe strain when axial compressive pipe stress reaches its maximum. (see Figure 3) (3) Onset strain of local buckling Onset strain of local buckling is defined as the critical longitudinal pipe strain due to bending when bending moment reaches its maximum. (see Figure 3) (4) Length of straight line In order to conduct seismic diagnosis of the high-pressure pipelines and the middle-pressure main distribution lines, they shall be segmented into the typical patterns, which segments are generally consisted of a straight line and two boundary elements as shown in Figure 4 [1]. A straight pipe is welded to a bend at both ends with an arbitrary bend angle, and the distance between the two weld lines is defined as the length of straight line. (see Examples 1 and 2 of Figure 4). As for tee-junction, a branch is neglected (see Example 3 of Figure 4) and a main pipe is assumed to be composed of a couple of 90 deg pipe bend. (see Example 4 of Figure 4) Remarks: (1) The upheaval buckling deformation process is presented in Figure 3.
Upheaval buckling zone Axial compressive stress Onset of upheaval buckling Local buckling zone Onset of local buckling Crack penetration zone Crack penetration due to ultra-low cycle fatigue (with leakage) Stress Bending moment Axial strain Bending moment

Straight pipe length

<22.5 deg.

Example 1 Pipe bend with a shallow bending angle less than 22.5 deg shall be regarded as straight

Straight pipe length

Example 2 Straight line between two underpasses

Straight pipe length Weld ring

Example 3 Straight line starting from 90 deg bend and the branch of tee junction shall be neglected

Straight pipe length Straight pipe length

Example 4 A part of network consists of two segments Figure 4 Concept of straight pipe length

Figure 3 Upheaval buckling deformation (2) Conceptual examples of straight pipe length are shown in Figure 4.

2. BASIC CONCEPT OF SEISMIC DESIGN (1) Ground Motion Level-2 (GML-2) defined in the JGA Guidelines 2006-03 shall be used in the seismic design.

(2) Pressure integrity of the high-pressure pipelines and main distribution lines shall be ensured, which implies no leakage providing large deformation of the pipes. (3) If upheaval buckling might occur due to the assumed ground motion and if the pipe strain might exceed the allowable strain during the post upheaval deformation, an upper limit for the length of straight line shall be provided to achieve seismic design that satisfies the seismic demand.

Copyright 2012 by ASME

(4) Items that are not defined in this guideline shall comply with the JGA Guidelines 2006-03. Remarks: (1) The design ground motion and the seismic integrity required for gas pipes shall be the same as those of GML-2 defined in the JGA Guidelines 2006-03. (2) According to experiments using 4 pipes to study upheaval buckling, when an axial compression is applied to a pipe, the pipe begins to bend upward (initiation of upheaval buckling) at the point where the axial compression reaches its maximum. Subsequently, pipe deformation due to upheaval buckling progresses, and local buckling then begins (onset of local buckling) at the point where the bending moment reaches its maximum. In the experiments, a pipe was deformed to the extent where axial strain easily exceeded the onset strain of local-buckling; however, any through-wall crack was not found in the pipe. Accordingly, it has been verified that for the longitudinal displacement (only compression), the critical deformation is located beyond the onset of local buckling. Test results show very good agreement with the FEA results obtained using the local FE model [9]. (3) According to the result of FEA with respect to axial displacement (only compression) by using the local finite element model, axial strain at the onset of local-buckling is 2D average compressive strain which is within the range of 6 to 9%. However, when cyclic loading due to seismic waves shall be taken into consideration, it is difficult to determine the critical stage of the pipe at a location beyond the local-buckling start point. Therefore, to be conservative, allowable strain is determined to be 3% in the same manner as the seismic design for GML-2 defined in the JGA Guidelines 2006-03. Thus, seismic integrity equivalent to that of the conventional guideline for seismic design of high-pressure gas pipes can be ensured. (4) While the seismic waves initiate upheaval buckling when the length of straight line is long enough, upheaval buckling does not occur when that becomes insufficient. The basic principle for seismic design to prevent upheaval buckling is to set an upper limit for the straight pipe length.

Chapter 4: Preliminary seismic design to prevent upheaval buckling Item Tensile properties Yield stress Luders elongation Content SMYS defined to be 60% of SMTS Yield plateau terminating at 2% Strain concentration of 0.3% Prevent upheaval buckling

Effect of non-uniform surface layer Checking criteria

Further check required ?

Yes

No

End No

Chapter 5: Specific seismic design to prevent upheaval buckling Materials tensile properties (Chap. 5) Non-uniform thickness surface layer (Sec 4 and 8 of Chap. 5) of Choosing checking criteria: Upheaval or Local buckling (Sec.8 of Chap. 5)

Figure 5 Preliminary and specific seismic design to prevent Upheaval buckling (2) First of all the preliminary seismic design shall be conducted using design parameters presented in Table 3. The parameters are assumed to be conservative in order to obtain the shortest length of straight line. If the shortest length is enough for the pipeline to be constructed, further discussion is not necessary. (3) If the shortest length needs to be updated, precise evaluation shall be required to make the length of straight line longer. The specific seismic design shall be conducted to prevent upheaval buckling actual data in terms of material tensile properties and the non-uniform surface layer and in accordance with Sec. 5.8.

3. PRELIMINARY AND SPECIFIC SEISMIC DESIGN Preliminary seismic design using conservative parameters shall be conducted in accordance with Chapter 4, "Preliminary seismic design to prevent upheaval buckling." In order to ensure seismic integrity taking into account particular parameters for pipeline design, the next step seismic design shall be compliant with Chapter 5 "Specific seismic design to prevent upheaval buckling." referring necessary Appendices of the guidelines. Remarks: (1) The outline of the preliminary and specific seismic design is presented in Figure 5.

4. PRELIMINARY SEISMIC DESIGN TO METIGATE UPHEAVCAL BUCKLING The following preliminary seismic design shall be conducted according to the natural ground period: (1) When a natural ground period is within the range shown in Table 1, the length of straight line shall be equal to or less than that presented in Table 2. (2) When a natural ground period is short or long compared to the range shown in Table 1, the straight pipe can be any length. (3) When the natural ground period is not surveyed, the length of straight line shall be equal to or less than the value shown in Table 2. Remarks: (1) A flowchart for the preliminary seismic design is presented in Figure 6. (2) "Uniform surface layer" means that the subsurface ground has an almost uniform thickness. "Non-uniform surface layer" means that the thickness of the subsurface ground is inclined and the longitudinal ground strain tends to concentrate in the shallow portion.

Copyright 2012 by ASME

Start No

length was slightly shorter when a bent pipe was connected. Therefore, in Table 2, to be conservative, the upper limit value of the straight pipe length in the former case was adopted [12].

Natural period of the ground is known? Yes No Natural period of the ground is within the range shown in Table 4-1 Yes Seismic design is conducted with a value equal to or less than the upper limit value in Table 4-2

Table 1 Natural ground period required for preliminary seismic design

Natural ground period (s) of surface layer Uniform 50 Nominal Outside Diameter (mm) 75 100 100 0.7 1.8 0.8 1.1 0.8 0.9 0.9 1.0 Non-uniform 0.7 or more 0.8 2.0 0.8 1.4 0.9 1.8

End

Figure 6 Preliminary seismic design procedures

(3) When the external force of an earthquake equivalent to the GML-2 assumed by the JGA Guidelines 2006-03 acted on long straight 4 or less gas pipes which were buried in the ground having a natural period within the range shown in Table 1 under the conditions shown in Table 3, upheaval buckling began and the strain generated by subsequent pipe deformation caused the allowable strain of 3% to be exceeded [11]. Therefore, when the natural period of the ground is within the range shown in Table 1, seismic design shall be conducted setting an upper limit for the length of straight line. (4) In Table 1, axial pipe strain is calculated assuming the length of straight line to be infinite; however, as the length of straight line is gradually decreased, strain generated due to deformation after the onset strain of upheaval buckling becomes equal to the allowable strain of 3%. If the straight pipe is not longer than this length, local buckling will not begin. When the length of straight line is further decreased, the generated strain becomes equal to the onset strain of upheaval buckling and the length of straight line becomes equal to the length at which upheaval buckling occurs. If this length is not reached, upheaval buckling will not occur. The upper limit of the length of straight line at which local buckling does not occur is slightly long compared to that where upheaval buckling does not occur. (5) Therefore, to be conservative, we adopted the upper limit value of the length of straight line at which upheaval buckling does not begin [12]. In Table 2, the upper limit values of the length were rounded up. Accordingly, for seismic design, the length of straight line shall be equal to or less than the applicable value [12, 13]. Herein, the length of straight line is the distance between two weld lines where a straight pipe and a bend are welded whose bend angle is 22.5 degrees or more. In this guideline, a main pipe of tee is regarded as a straight pipe, and a branch is regarded as a 90-degree bend. (6) The upper limits of the length were compared between two situations: one in which a bent pipe was connected to the end of the straight pipe and the other in which a loop was connected to the end of the straight pipe. The result was that the straight pipe

Table 2 Upper limit of the length of straight line

Bend angle connected to straight line 90 50 Nominal Outside Diameter (mm) 75 100 100 82 m 92 m 99 m 162 m 45 63 m 71 m 77 m 125 m 22.5 30 m 34 m 37 m 60 m

Table 3 Assumptions for preliminary seismic design

Parameter Pipe type Tensile properties Yield stress Backfilling Critical strain Design seismic motion Trans. pipe-soil interaction Definition SGP, STPG370sch40 Yield plateau terminates at 2% SMYS set to be 60% SMTS 1.5 m FEA with a partial FE model Design seismic motion I Hyperbolic function

(7) When the upper limits of the length of the straight line were compared between the uniform ground and the nonuniform ground, the length of straight line in the non-uniform ground was slightly short compared that in the uniform ground. Hence, in Table 2, to be conservative, the upper limit value of the straight pipe length in the latter case was adopted [12]. (8) A pipe bend with a bending angle less than 22.5 degrees has little influence on the upper limit of the length of the straight line. Therefore, such a pipe is regarded as a straight pipe, and the length of the pipe is added to the length of the subsequent straight pipe and evaluated.

Copyright 2012 by ASME

(9) Calculation of the natural ground period shall be compliant with Section 5.2.1, "Natural ground period" in the JGA Guidelines 2006-03. (10) Table 3, which shows conditions for calculating Table 1 and Table 2, has been established to be conservative. Detailed seismic design can be based on Chapter 5, "Detailed seismic design to prevent upheaval buckling." (11) Compaction of the surrounding soils of pipes is effective for mitigating upheaval buckling because it can ensure an initial tangent of the transversal pipe-soil interaction, which presents a significant effect on the onset strain of upheaval buckling [7].

5.2 Design seismic ground motion Design seismic ground motion shall be set compliant with Section 5.1, "Design seismic ground motion" in the JGA Guidelines 2006-03. 5.3 Subsurface ground displacement and ground strain Surface ground displacement and strain shall be obtained in accordance with Section 5.2, "Surface ground displacement and ground strain" in the JGA Guidelines 2006-03. 5.4 Non-uniform ground Longitudinal strain induced in non-uniform ground tends to become large compared to that generated in uniform ground. Therefore, the non-uniform surface layer shall be taken into account to estimate the strain concentration in accordance with Section 5.3, "Non-uniform ground" in the JGA Guidelines 2006-03. Remarks: (1) As the surface waves defined in the JGA Guidelines 200603 is presented with a sinusoidal function, there are two ways to express the strain concentration in the non-uniform ground. One is to increase the amplitude keeping the wavelength constant and another is to decrease the wavelength keeping the amplitude constant. Pipe strain induced by the former seismic wave pattern will be large compared to that generated by the latter wave function. Therefore the former sinusoidal function shall be adopted to conduct a conservative evaluation [6]. (2) Two-dimensional response analysis to shear waves can be used instead of the above-mentioned sinusoidal function to estimate longitudinal strain induced in the non-uniform ground. 5.5 Pipe-soil interaction Pipe-soil interactions in the longitudinal and transverse directions shall be used to simulate the upheaval and local buckling behaviors and expressed by a hyperbolic function whose maximum values are defined with those values described in Sections 4.4.1 and 4.4.2 of the JGA Guidelines 2006-03. Remarks: (1) The onset strain of upheaval buckling is strongly dependent on the initial tangent of the transversal pipe-soil interaction. Hence the onset strain tends to increase with increasing initial tangent. The longitudinal and transversal pipe-soil interactions described in the JGA Guidelines 2006-03 are expressed with bilinear functions and the initial tangents are small compared to the test data. Hence the pipe-soil interactions in the Guidelines shall be presented with a hyperbolic function determining the initial tangents with the test data and their maximum values with those values calculated in accordance with the JGA Guidelines 2006-03. (2) Test data of pipe-soil interactions can be used instead of the above mentioned description.

5. SPECIFIC SEISMIC DESIGN TO METIGATE UPHEAVAL BUCKLING 5.1 Tensile properties of gas pipes 5.1.1 Yield stress The minimum yield point of the specifications shall be used for the yield stress. Remarks: (1) As a minimum yield point of SGP is not prescribed in the Japanese Industrial Standards (JIS), the yield stress shall be defined to be 60% of the tensile strength prescribed in the JIS in accordance with Appendix 4 of the Announcement of the Gas Business Act. (2) If the yield stress of the material to be used is known, it can be used. 5.1.2 Tensile properties Tensile properties shall be set properly according to the material to be used. Remarks: (1) Materials tensile properties have a significant effect on the onset strain of upheaval buckling. The onset strain of pipes with LE type material, whose plateau terminates at 2% strain for example, tends to be rather small compared to that of RH type materials. The materials tensile properties have not been prescribed in the current pipeline seismic design codes therefore additional tensile tests shall be conducted in order to know the tensile properties. Accordingly, when tensile properties are not known, to be conservative, seismic design shall be conducted assuming the LE type tensile properties with a plateau up to 2%. The onset strain of upheaval buckling of pipes with the RH type material will be large compared to the maximum longitudinal ground strain induced by GML-2 [10]. (2) If the tensile properties of materials to be used are known, they can be used.

Copyright 2012 by ASME

5.6 Onset strain of upheaval buckling Onset strain of upheaval buckling shall be analyzed by finite element analysis using the partial FE model. Remarks: (1) The onset strain of upheaval buckling can be obtained using the partial FE model that has a 15 m center portion. It has been confirmed that the results using the partial FE model presents very good agreement with the test results. The post-upheaval buckling behaviors are unstable to simulate with the global FE model [8, 9, 10]. (2) If the calculations converge, it is possible to use the global model to estimate the onset strain of upheaval buckling. 5.7 Onset of upheaval and local buckling 5.7.1 Onset of upheaval buckling A comparison shall be made between the pipe strain obtained according to Section 5.5, "Seismic design of straight pipes" in the JGA Guidelines 2006-03 and the onset strain of upheaval buckling obtained in accordance with Section 5.6, "Onset strain of upheaval buckling." (1) When the pipe strain is equal to or greater than the onset strain of upheaval buckling, upheaval buckling will occur. (2) When the pipe strain is less than the onset strain of upheaval buckling, further study is not necessary. Remarks: (1) Onset of upheaval buckling is determined by comparing the pipe strain that occurs in a straight pipe having infinite length with the onset strain of upheaval buckling [10]. 5.7.2 Onset of local buckling When it has been determined that upheaval buckling will begin according to Section 5.7.1, "Determination of onset of upheaval buckling", the pipe strain in the buckled portion shall be compared with the allowable strain of 3%. (1) When the pipe strain exceeds the allowable strain, it is determined that local buckling will begin. (2) When the pipe strain does not exceed the allowable strain, further study is not necessary. Remarks: (1) Onset of local buckling is recognized when the maximum pipe strain in the buckled portion exceeds the allowable strain of 3% [11]. (2) Maximum pipe strain during the post-upheaval buckling deformation shall be predicted using an analytical model presented in Ref. [11]. (3) Except the analytical model presented in Ref. [11], proper analytical models can be adopted for predicting the pipe strain in the buckled portion. 5.8 The length of straight line to prevent buckling When it has been determined that local buckling will occur according to Section 5.7.2, "Determination of onset of local

buckling", an upper limit of the length of the straight line shall be calculated which makes the pipe strain in the buckled portion equal to or less than the allowable strain of 3%. (1) When the pipe strain exceeds the allowable strain, it is determined that local buckling will occur. (2) When the pipe strain does not exceed the allowable strain, further study is not necessary. Remarks: (1) The analytical model to calculate the length of straight line is presented in Ref. [12]. (2) The boundary elements at both ends of the straight line shall be set properly according to the number of bend pipes, angle, and other actual conditions [12]. (3) Except the analytical model presented in Ref. [12], proper analytical models can be adopted for predicting the pipe strain in the buckled portion. (4) To be conservative, the length of the straight line at the onset of upheaval buckling can be used instead of that at the onset of local buckling. (5) The upper length of straight line at which upheaval buckling will not occur can be obtained by multiplying the reference length by a reduction coefficient [13].

CONCLUSIONS The Draft Seismic Design Guidelines for Small Diameter Pipes Mitigating Upheaval Buckling issued by METI in 2011 was explained in this paper. The key issues of the Guidelines are summarized as follows; The Guidelines shall be applied to high-pressure pipelines and main distribution lines with diameter of 4 and smaller. The Guidelines shall be applied to small diameter pipes with the LE type materials which have very high possibility of upheaval buckling due to the strong ground shaking defined in the JGA design code. Pipes with the RH type materials will not buckle in upheaval mode due to even extremely strong ground shaking therefore further discussion is not required. The key parameter defined in the Guidelines is the length of straight line whose upper limit to mitigate upheaval buckling is presented in terms of natural ground period, and pipe diameter. Preliminary seismic design shall be performed setting all of the design parameters to be conservative which are the LE type stress-strain curve and SMYS of 60% SMTS. In order to calculate the upper length of straight line for a specific case, Specific seismic design can be performed using the actual data with respect to the pipeline. The Guidelines deals with upheaval buckling and local buckling behaviors induced by strong ground shaking however the results of the buckling test and predicted by FEA will be

Copyright 2012 by ASME

useful to investigate the upheaval and local buckling behaviors of large diameter pipelines induced by other external forces.

ACKNOWLEDMENT A series of investigation works was conducted to establish the seismic design guidelines to mitigate upheaval buckling of small diameter high-pressure pipelines and middle-pressure main distribution lines. The authors would like to thank Ministry of Economy, Trade and Industry of Japan for the permission to publish this paper. Also the authors would like to express sincere thanks to the members of the committee for their invaluable comments and advices.

REFERTENCES [1] METI, Steering Committee Report of the Draft Seismic Design Guidelines to Mitigate Upheaval Buckling, 2011. (in Japanese) [2] Japan Gas Association, JGA Guideline 2006-03 Seismic design standard for high-pressure gas pipelines (in Japanese), 2006. [3] Horikawa, H, Tsunasawa, Y, Koganemaru, H, and Suzuki, N. (2012): Upheaval Buckling of Small Diameter Pipelines Induced by Strong Ground Shaking, IPC2012-90281. [4] Suzuki, N, Koganemaru, K, and Shimizu,Y. (2004): Swift Algorithm to Predict seismic responses of Extensive Buried Distribution Networks, 13th WCEE. [5] METI, Materials Properties of Gas Pipes, Appendix-1 of the Draft Guidelines, 2011. (in Japanese) [6] METI, Effect of Non-Uniform Ground, Appendix-2 of the Draft Guidelines, 2011. (in Japanese) [7] METI, Longitudinal Pipe-Soil Interaction, Appendix-3 of the Draft Guidelines, 2011. (in Japanese) [8] METI, Finite Element Models for Upheaval Buckling Analysis, Global and Local Models, Appendix-4 of the Draft Guidelines, 2011. (in Japanese) [9] METI, Validation of Local Finite Element Model using Test Data, Appendix-5 of the Draft Seismic Design Guidelines, 2011. (in Japanese) [10] METI, Onset of Upheaval Buckling, Appendix-6 of the Draft Guidelines, 2011. (in Japanese) [11] METI, Onset of Local Buckling, Appendix-7 of the Draft Guidelines, 2011. (in Japanese) [12] METI, Upper Limit of Pipe Length to Prevent Upheaval Buckling and Local Buckling Appendix-8 of the Draft Guidelines, 2011. (in Japanese) [13] METI, Calculation of Pipe Length to Prevent Upheaval Buckling and Local Buckling, Appendix-9 of the Draft Guidelines, 2011. (in Japanese)

Copyright 2012 by ASME