Latest methods used to build Alps line

Domenico Ruffa, Giorgio Pesaresi and Eraldo Raffaeli, Snamprogetti, Fano, Italy

Snamprogetti has successfully engineered and constructed a 48-in. gas pipeline through nine pre-existing mountain tunnels in the Swiss Alps, at lengths from 980 ft to nearly 8 mi.

The work was part of a project commissioned by Transitgas to expand an existing pipeline that delivers gas from the North Sea to Italy.

The project entailed several construction challenges. It called for installing 48-in. pipe into tunnels originally built for 34-in. pipe, on severely sloping inclines. In this environment, protecting the integrity of the pipe and its coatings was a concern. Pipe loads, stresses and strains had to be carefully calculated.

The fact that the tunnels had been built for smaller-diameter pipe made it difficult to transport men and vehicles inside the tunnels. Certain sections of the tunnels required restructuring to facilitate installation. In addition, to minimize interruption of gas supply, the project had a tight, six-month construction timeframe.

These challenges were met using pipeline roller supports and an anchoring system that helped crews maintain pipe and coating integrity during installation. A mechanized computer welding system was employed. Girth welds were inspected using nondestructive testing (NDT) technologies, including an automatic ultrasonic orbital system.

Project challenges

The most difficult aspects regarding the project’s design and implementation concerned the line’s route, due to morphological and altimetrical considerations. The route chosen called for installing the pipe through several tunnels consisting of horizontal or inclined sections with planimetric deviations and slopes up to 100%.

Building the line through the Alps presented a series of challenges and forced project planners to consider:

• tunnel sections initially designed for a 34-in. pipeline;
• pipeline supports that would allow the new pipes to be installed inside the tunnels;
• optimization of the roller supports’ size to guarantee the integrity of the HDPE covering during pulling;
• sizing the anchoring system at the tunnel ends and at the various plano-altimetric changes of direction inside, with a view to minimizing construction time; and
• a tight deadline for completing the work.

Within this context, the research on welding technologies and NDT was of strategic importance to limit the time required for construction activity inside the tunnels. This research also helped to ensure quality and productivity requirements, as well as reliable diagnostics of possible welding defects in timeframes compatible with production.

For the above, a mechanized welding process (GMAW) called computerized welding system (CWS) was chosen. Italian company Pipeline Welding Technology set up this system, which provides electronic and computerized control of welding parameters. For girth weld inspection, an ultrasonic automatic orbital system called Rotoscan was used. The Dutch company RTD set it up.

Design and installation

The design activities concerning the pipeline sections inside the tunnels were carried out according to the following successive steps: design and optimization of pipe anchoring and supporting systems; pipe structural assessments (stress-analysis); and pipe installation inside tunnels.

Anchoring, support systems

The first step was to design the pipe-anchoring system. A key consideration was that the 48-in. line, when inside the tunnels, would take up considerable space.

Attention then turned to the study, sizing and execution of support prototypes to meet the following requirements:

• high load-bearing strength due to the considerable pipeline unit weight;
• light enough to be moved by hand;
• small enough to avoid impeding worker movement inside tunnel;
• possibility of assembling some support components before the old line was removed;
• multipurpose to allow for use during pipeline installation through the rollers, as well as for final pipeline support by substituting the rollers with two half-saddles; and
• possibility of using anchoring ties to be fitted on cement blocks, placed at the tunnel ends and at changes of direction inside the tunnels to reduce down time and interference with the various stages of constructing the blocks.

After a preliminary examination, the de­signers decided the pipeline must be fully restrained, longitudinally and transversally, by anchors at the ends of the tunnels and at each change of direction (vertices). Due to the operating conditions, the free thermal expansion would cause considerable pipe elongation. This could jeopardize the stability of the system and make necessary complex works aimed at absorbing the displacements, as well as expensive maintenance during operating.

Yet suitable anchors capable of withstanding high degrees of stress would not be too expensive. This was due to the rock’s good load-bearing capacity, which enables the dimensions of the blocks to be contained within reasonable limits. Also, the other pipeline supports (saddles) along the tunnel are more reliable when the pipe is fully anchored. Saddles are not subjected to fatigue cycles or displacements, which involve several maintenance-related problems.

The anchoring system at the tunnel ends and at the various plano-altimetric changes of direction inside were realized by two kinds of supports (Figures 1 and 2) anchored on concrete anchor blocks.

The supports along the tunnels (Figures 3 and 4) were prefabricated from steel and installed before pipeline laying at different intervals, with a maximum of 41 ft (12.5 m) on the slope section and 57 ft (17.5 m) in subhorizontal sections.

The function of the rollers was to enable the laying of prefabricated pipe strings. Once the laying operation was completed, the rollers would be dismantled and replaced by support saddles, and the pipe fastened to the support base by means of a flexible tie.

The main function of the saddle was to hold the pipe during operating conditions. Heat-shrink sleeves and chloroprene sheets placed between pipe and saddles and along the pipe coating provided cathodic insulation. These supports also had to be capable of holding transversal thrusts, due to imperfect lineup of the pipe. The saddles are firmly anchored to the foundation by anchor bolts calculated to resist all stresses.

A longitudinal thrust also was considered in the design. Such a thrust could occur due to small pipeline displacements ensuing from temperature or pressure changes along a pipeline section, even though such section is firmly anchored at its vertices. The saddle was designed to enable planimetric and altimetric adjustments of the pipe to facilitate erection and guarantee a good load distribution.

Structural analyses

The structural analysis of the gas pipeline sections running through tunnels was carried out using appropriate calculation programs able to take into account:

• mechanical characteristics of the pipe (diameter, thickness, material);
• project data such as temperature, design and hydraulic test pressure variations; and
• tunnel configuration (length, slope, plano-altimetric variations).

Design data

Stress-analysis activities were carried out with the following functional design data:

Di = 1184.3 mm fixed internal diameter

S_p1 = 18.0 mm normal thickness

S_p2 = ­21.2 mm increased thickness

S_p3 = ­26.1 mm reinforced thickness

SMYS = ­448 N/sq mm pipe material yielding (API 5L-X65)

P = 7.5 N/sq mm operating pressure

P_HY =­ ­11.24 to 16.1 N/sq mm hydraulic pressure

T_MAX = ­104°F to 131°F (40°C to 55°C) maximum operating temperature

T_MIN =­ ­­32°F (0°C) minimum operating temperature (vent)

T_HY =­ 50°F (10°C) hydraulic test temperature

T_TIE-IN = ­­­50°F (10°C) installation temperature

The pipeline sections were modeled in terms of length, angles, and anchor and supports position. Coordinates of the tunnel vertices and related geometrical data also were considered.

Design criteria

Para. 833 of ASME B.31.8 Code states how to calculate the stresses on a line free to move. But it does not give a direct method for calculating the equivalent stresses on fully restrained pipeline sections. For this, it is advisable to refer to the criteria suggested by ASME B31.4 at Para. 419.6.4.b.

These stresses (obtained by calculating all forces simultaneously acting) must not exceed the allowable values recommended by the Code for each load conditions: 0.9 x SMYS for operating and vent conditions, and 0.95 x SMYS for hydrostatic test conditions.

Pipeline stress calculation

The pipeline structural analysis has been carried out for the tunnel route. The stresses were calculated considering that the laying method used causes an increasing precompression toward all vertices.

The pipeline stress calculation took into account the loads acting on pipe caused by:

• internal pressure;
• temperature change;
• bending moment due to the span;
• residual torque;
• longitudinal compression due to pipe weight;
• predeformation due to the elastic radius;
• local stresses due to load-bearing on saddles; and
• possible overstresses caused by laying methods other than the one assumed.

The compressive load Fa due to the weight component acting on the upstream anchor block was considered. The resulting compressive stress is:

?_w = -Fa / A

Considering the pipe longitudinally anchored, the longitudinal stresses due to Poisson’s effect and temperature difference are:

?_v = 0.3 P•D / 2t

?_?T = -? • ?T • E

Due to the bending of the continuous beam with equally spaced spans l, the bending moment MS at the support and in the span center M_C are:

M_S = 1/12 qº • l² • cos ?

M_C = 1/24 qº • l² • cos ?

The stresses due to discontinuous load-bearing are:

?_b =­ stress due to the bending moment at saddles

?_bl =­ stress due to the bending moment at the span center

?_s =­ concentrated stress at the saddle edge

? = shear stress due to dead load

The stresses at the saddle and span center are calculated as follows:

?_b,b1 = ± M_s,c / W

where: W = ?[D4-(D-2t)4] / 32D

At the saddle, the stress is tensile type at the upper part of the pipe and compressive in the lower side.

A compressive stress will be present in the upper part of the pipe, whereas in the lower part the stress will be of the tensile type. The stress caused by concentration of stresses at the saddle has been calculated according to the method indicated in Roark’s Formulas for Stress and Strain (6th edition, chapter 12.7, “Pipe on Supports at Intervals”):

?_s = k (P/t²) ln (R/t)

If T= qo• l / 2 is the shear force at each support, the maximum shearing stress acting on the generatrices of the pipe is:

?_max = 2T / A

For effect of the internal pressure P, the circumferential (hoop) stress ?_H is:

?_H = P•D / 2t

The ovalization stress also was considered at saddle. The stresses produced by load convergence on the bends supported by concrete saddle were calculated:

?_A,B,C = K_A,B,C • Q • R_t (6/t²) + (T_A,B,C•Q) / t

The value of the main stress in the axial direction is the sum of the stresses acting at the same time, namely:

?_x = ?_?T + ?_v + ?_W + ?_b + ?_S

The stress value in transversal direction is:

?_y = ?_H + ?_A,B,C

The shear was not considered because the proposed installation method did not cause torque moments (provided no rotation of the pipe string occurs with respect to its axis during laying operations).

The equivalent stress is calculated with the Von Mises criterion:

?_e = (?_x²+ ?_y² - ?_x • ?_y + 3 • ?²)^½

According to the maximum shear theory (Tresca/Guest), as indicated in ASME B.31.4 para. 419.6.(b), the equivalent stress is the higher of the two main stresses ?_x (total longitudinal) and ?_e (total circumferential) if they have the same sign, or the sum of both if they have opposite sign. Once the maximum equivalent stress has been found – as a rule it occurs where the local bending effect causes a compression of the fibers – the equivalent stress value is compared with the yield stress of the steel used. Therefore, the utilization factor k is:

k = ?_e / SMYS

Construction stages

To keep to the deadlines, the construction stages were carried out according to the following schedule:

• civil engineering works to restructure the tunnels and remove rock to make room for the pipeline;
• fitting of anchoring plates that, welded to bolts anchored into the ground, provided the bases for the roller and saddle supports (carried out, in part, before the old pipeline was removed);
• removal of the old pipeline, supports and anchor bolts;
• fitting of the roller supports needed to install the pipeline;
• positioning and fitting of anchoring blocks (placed at plano-altimetric changes of direction inside the tunnels);
• installation of pipeline on roller supports by means of suitably sized pulling winches, proceeding to welding of the individual pipe sections;
• assembly of strongly sloping and subvertical sections, working from top to bottom;
• substitution of roller supports with half-saddle supports for final pipeline installation;
• execution of the various tie-ins inside each tunnel;
• installation of anchoring blocks at each end of the tunnels;
• execution of tie-ins at the ends of each tunnel;
• application of coatings on weld seams using heat-shrink sleeves or alternatively thermosetting material;
• hydraulic inspection of the various pipeline sections; and
• execution of final connection to the buried gas pipeline sections outside the tunnels.

Welding activities

The welding and NDT procedures began in the following chronological order:

• preparation of pipe bevels (narrow grove) for automatic welding using a beveling machine;
• transferring of pipe (one at a time) into the tunnel through side-access openings, using mechanical vehicles, to a widened area of the tunnel in which the operational stages took place;
• lifting and alignment of pipe with the others previously set out on dedicated supports, by means of winches;
• pipe coupling by means of an internal pneumatic device;
• preheating up to 212°F (100°C);
• welding by means of CWS;
• welding inspection using Rotoscan;
• executing necessary repairs;
• manual ultrasonic testing (UT) of repairs;
• laying pipe in the tunnel;
• radiographic testing (RT) of repairs; and
• RT using single-wall exposure inside the tunnel.

Pipe bevel preparation was completed outside the tunnels near the entrance, in a previously designated area. The lines were installed on special carriages, in double joints (6.5 by 49 ft, or 2 by 15 m). Crews then proceeded with column assembly, working from bottom to top.

In sloping sections of the tunnel, shielded manual arc welding techniques were used. The tie-in welds in these sections were tested using manual UT and RT. The work cycle was considered complete after the final radiographic judgement about the repairs.

Mechanized welding system

The welds carried out on the tunnels’ nonsloping sections were mainly executed by means of CWS. This technology consists of two orbital welding heads operating simultaneously on the exterior, one working clockwise and the other counterclockwise.

All the passes are deposited with downhill technique, where the arc is protected by means of a mixture of argon and carbon dioxide. Around the internal lineup clamp, a copper backing supports the root pass.

The welding parameters are calculated and then memorized on dedicated software. The system interacts with a central control unit on a pay welder. An example of general welding-station unit arrangement is shown in Figure 3.

In this instance, the process control boxes were separate from the pay welder, which could not access the tunnel due to its size. The welding heads receive instructions from the process control boxes and so move freely until each pass has been completed under the supervision of two operators. The two operators can intervene by means of peripheral instrumentation should it be necessary to modify the welding parameters within a preset tolerance range.

Normally, when welding with this technology, the same number of welding stations are used as the number of passes to be deposited for each pipe wall thickness. This procedure is not dictated by technical reasons but is aimed at increasing productivity. Due to lack of space inside the tunnels, a maximum of two welding stations were employed.

To qualify the welding procedures, the 48-in. joints were welded in accordance with Snamprogetti specifications. The mechanical tests employed for this purpose (tensile, guided bend and hardness) produced satisfactory results. Also, the results of impact V notch and CTOD testing – carried out at -68°F (-20°C), the minimum design temperature of the aboveground pipes – were deemed satisfactory.

The welding productivity, taking into account the environmental conditions, also was considered satisfactory. To complete the welding operations, daily shifts were organized, covering all 24 hours. Maximum production recorded in one 24-hour day, using two welding stations, was 20 joints.

Automatic ultrasonic system

The Rotoscan system used to check the girth welds offered several advantages. It can be carried out shortly after welding completion, and the results are available virtually in real time. The weld can be checked as soon as the temperature of the joint drops below about 176°F to 212°F (80°C to 100°C). And unlike RT, Rotoscan testing can take place without interfering with the welding operation, since there are no problems of ionized radiation. The system also offers immediate feedback as to the quality of the weld. This allows for immediate corrective action and repair if defects are found.

The Rotoscan system consists of a series of probes with different angles set out in a pattern. The volume of the weld is checked in a single scan. The pattern orbits around the joint and is fixed to a specially designed probe band. The system is electrically powered, and the scan output is connected to a PC that processes the data. The technician in charge then interprets the data and gives his report. At the same time, a diagram is produced showing the indications of the scan. Figure 4 shows the ultrasonic beam view.

The Rotoscan inspection allows volumetric and planar defects to be detected, and is ideal for the inspection of joints made with GMAW processes. Unlike RT, it is able to detect the cold lap and imperfections typical of this type of welding process. However, the same can be applied to the inspection of welds carried out by conventional welding processes with identical performance.

To build the pipeline in the tunnels, 1,600 welds were made with the mechanized CWS system. The percentage of defects found by Rotoscan was about 8%. In addition to Rotoscan inspection, RT also checked about 30% of the welds made using mechanized processes.

The productivity of the Rotoscan inspection is high, since it only takes two minutes to scan a 48-in. weld joint. Further, no problems were found to cover the daily weld production.

Regarding the site organization for the Rotoscan inspection, two or three operators were used on each shift. These operators were adequately trained in the Rotoscan inspection process and qualified according to EN 473 standard on qualification and certification of NDT personnel. Since several sites employed automatic welding, a site coordinator also was available.

Synchronizing work activities

The most critical aspect of working in the tunnels was that of finding the right working synchronism to avoid interfering with other work in progress. Welding and NDT testing play a fundamental role in constructing a pipeline, since they dictate the rhythm of the other tasks necessary to complete the job. It therefore was necessary to program the work timetable carefully, especially regarding the installation of the slants (steeply sloping lengths of pipeline) inside the tunnels.

The technical experience acquired on this project represented another important evolutionary moment in tunneling technology and offers new technical perspectives in the use of routes through new and existing tunnels. The need to create easily maneuverable pipe supports inside the tunnels, together with the need to reduce maintenance interventions, provided the stimulus for an alternative to steel as construction material for the supports.

After careful technological evaluation, re­search and analysis, a prototype support (saddle) was made of aluminum alloy, the same size as the one shown in Figure 1. This support, subjected to stringent load-bearing tests, gave excellent mechanical resistance results.

This new support, although not used on this project, may provide an alternative to traditional steel supports. It provides better maneuverability and is lighter by 60% to 65%, with no need for long-term maintenance.

CWS mechanized welding proved effective. And from a health and safety perspective, this system gives off less fumes than manual welding. On this project, this meant more time could be spent in the tunnels.

Even though a fume-extraction system had been set up, the problem nevertheless impacted work in the tunnels. For future installations inside tunnels, the presence of fumes should not be underestimated.

Regarding NDT’s technical aspects, automatic welding confirmed its potential, especially in terms of mechanical performance, with particular reference to the toughness properties. From an operational point of view, attention should be paid to the placement of process control boxes, since humidity in the tunnels can affect their electronic and computerized instrumentation.

The use of the Rotoscan automatic inspection system proved advantageous in terms of technical performance and reduced time required for NDT inspection of girth welds. Nevertheless, particular care must be given to the arrangement of the equipment in order to avoid the same problems mentioned for the CWS equipment.

Finally, the acceptability criteria of the welds must be established carefully because the simultaneous application of Rotoscan and RT for weld defect diagnosis can give rise to conflicting judgements.

Acknowledgment

This article is based on a paper presented at the International Gas Union’s 21st World Gas Conference in Nice, France.

Authors

Domenico Ruffa is a welding engineer at the Snamprogetti Fano head office, the engineering company of Eni group. In 1983, Ruffa joined Snamprogetti’s piping department, later assuming his present position. He is involved in the company’s field upstream facilities and pipelines. Ruffa holds a degree in mechanical engineering from the Mechanical Institute of Vibo Valentia, Italy, and is certified as European welding inspector technologist in accordance with European Welding Federation guidelines.

Giorgio Pesaresi is a stress analysis engineer at the Snamprogetti Fano office. He joined Snam­progetti’s piping department in 1979, and later moved to the company’s stress analysis department. Pesaresi is involved in efforts to process stress analysis problems for pipeline and unit piping systems. Pesaresi holds a degree in mechanical engineering from the Mechanical Institute of Ancona, Italy.

Eraldo Raffaeli is purchasing department manager for the Snam­progetti Fano office. From 1995 to 2000, Raffaeli served as manager of the piping and layout department of the field upstream facilities and pipelines division. Since 1994 he has been involved in advanced submarine projects. Raffaeli holds a master’s degree in nuclear engineering from the Politecnico of Turin.