Custom sleeve design stops tight-quarters crude header leak

Dec. 2, 2013
Custom sleeve design provides a cost-effective means for stopping leaks on pipeline supply headers delivering to refineries. A sealing clamp initially used to stop a leak on an oil pipeline began to show leaks around its edges.

Subrata Saha
Mahendrapal Rathore

Reliance Engineering Group
Jamnagar, India

Based on presentation to PetroWorld India, Mumbai, Aug. 22-24, 2013

Custom sleeve design provides a cost-effective means for stopping leaks on pipeline supply headers delivering to refineries. A sealing clamp initially used to stop a leak on an oil pipeline began to show leaks around its edges. A custom-designed sleeve stopped the clamp leak. Fig. 1 shows the system detail of the leaky pipeline.

Site conditions made conventional sleeve design unfeasible (Fig. 2). Custom sleeve design considered space constraints in terms of both on site fabrication and installation (Fig. 3). The sleeve's complex geometry required finite element analysis (FEA) for mechanical design.

The clamp applied to stop a leak on a crude transfer header running between marine storage and a refinery in India began to leak. Space, time, and financial constraints required a customized approach to stopping the leaky clamp (Fig. 2).

Description

Corrosion led to pinhole leaks on about 400 mm of a pipeline's surface. An online sealing clamp immediately stopped the leaks. After 6 months, however, the sealing clamp's edge began to leak. Sealant was used to arrest the leakage, but sweating along its edges raised serious concerns.

The operator considered three possible solutions:

    • Line shutdown, pipe replacement.

    • Hot tapping.

    • Sleeve fabrication.

The line was transporting crude to the refinery and shutdown was not commercially viable. Hot taps, by contrast, would provide a bypass connection, allowing flow to continue while necessary modifications were made. This operation would have been very expensive.

Designing a sleeve to encase the leaky portion of the crude transfer header carried less risk than a hot tap and would also be substantially cheaper. Combined with the lack of any need to interrupt flow, these factors made the sleeve the preferred option.

Solution

The space between crude transfer header (CTH)-1 and CTH-2 measured 85 mm, and that between CTH-1 and the single point mooring (SPM) header 127 mm (Fig. 3). Available space between the adjacent pipes, combined with the lug length sealant clamp, made 38-in. OD pipe a feasible solution. But 38-in. pipe was not available, causing rolled carbon steel plate to be considered.

When rolled plate was checked for installation, however, space between CTH-1 and CTH-2 proved limited, making straight steel plate the best option. Using conical transitions at the ends instead of flat plates minimized the end closures' required thickness. Conical transitions also provided more flexibility than flat plates for absorbing differential thermal expansion. Figs. 3-4 show the best possible sleeve design considering fabrication and installation.

Practical limitations

Sleeve design faced the following practical limitations:

    • Hydrotesting of the oversize sleeve was not feasible.

    • Inspection of all field welds as per original line specification could not be done.

    • Space constraints limited plate thickness.

    • Tight clearance between the grade and the bottom of the pipe prevented drain installation.

Complex geometry and site constraints made hydrotesting either at the shop or in situ unfeasible. No draining could be provided at the site, making the only viable option a service test. A pressure gauge placed on the vent monitored pressure and avoided build up of excess pressure in the cavity.

This testing approach would typically call for 100% nondestructive testing of all weld joints. But since this was not possible, in-situ fabrication was also ruled out. Instead a shop fabricated the sleeve in two halves, reducing the number of field weld joints to four: two longitudinal and two circumferential.

Field engineers conducted die-penetrant tests and radiography or ultrasonic flaw tests on all shop weld joints, and die-penetrant and ultrasonic flaw detection tests on longitudinal field joints. Only die-penetrant tests were feasible on the circumferential joints. Coating the inner sleeve surface with anti-corrosion material prevented corrosion due to stagnant fluid.

Design

ASME 31.4-2006 guided design of the CTHs.1 The same standard also guided pressure thickness calculations for the circular portion of the sleeve, with ASME Section VIII Division 1-2010 covering design of the end closures (Table 2).2 ASME 31.4-2006 Para 402.3.1 guided allowable stress calculations, with material properties considered under ASME BPVC Section II Part D 2001.3 Engineers satisfied ASME Section VIII Division 2, Part 5 "Design by Analysis" requirement for elastic stress,4 while not considering fatigue.

The pipeline consisted of A671 CB6013, 30-in. OD pipe with 0.375-in. WT. Sleeve design parameters set pressure at 170.68 psi and temperature at 40° C.

FEA

Table 1 gives the numerically calculated thickness required for each component to sustain design pressure. But the stress at the junction of sleeve components and sleeve-to-header pipe proved difficult to calculate. The final geometry of the sleeve was very complex. No simplified formulas or algorithms are available for evaluating the flexibility of such complex geometries (Figs. 3-4), requiring development of an FEA model for sleeve design.

Model inputs

ANSYS V.14.5 was the project's software, helping carry out a static linear elastic study using small displacement theory. Material properties used in this report came from ASME IID.3 Stress classification limits are set in accordance with ASME VIII-2.4

Engineers used a sectioned solid model for finite element analysis, removing pipe tolerances (Fig. 5).

A sectioned solid model with pipe tolerances removed provided the basis for analysis. Fig. 5 shows the selected Solid 186 3D 20-Node Structural solid element. Application of tetrahedral mesh allowed the contact elements to be treated as bonded. Reported error measured less than 5% in general regions, as per ABSA FEA Guidelines AB-520 (Fig. 6), validating the mesh selected and model used for analysis.

Analysis required application of a tetrahedral mesh and treatment of contact elements as bonded (Fig. 6).

The sleeve and CTH pipe end had symmetry restraints applied, with internal surfaces subjected to the 170.68 psi and 40° C. conditions defined by the design cases. The model is in balance and restrained from rigid body motion.

The absence of proper modeling software caused weld details and other detailed fillet radii to be left out of the FEA model. Assumptions included a homogeneous isotropic material and a 21° C. installation temperature.

Design cases

The first design case considered a scenario in which an encased header pipe ruptures and the sleeve contains the fluid. Both the sleeve and the encased pipe were assumed to be at fluid operating condition.

The second design case considered a scenario in which the encased header pipe is holding the fluid and there is no leak. In this case, the sleeve is at ambient temperature and pressure while the encased pipe is at fluid operating temperature and pressure. Evaluation focused on stresses generated due to differential thermal expansion at the sleeve and header pipe junction.

Design Case 1

Case 1 analyzed the sleeve under fluid operating pressure and temperature. The preliminary sleeve model used a uniform thickness of 10 mm minus tolerances. Stresses were very high at straight-plate center and at the junction of the straight plate and the sleeve's circular section (Fig. 7).

The stress plot on the 10-mm thick sleeve registers very high at the straight plate center and the junction of the straight plate and the sleeve's circular section (Fig. 7).

Considering this preliminary result, determination of the required plate thickness used Roark's formulas for stress and strain (Table 1).5 High stresses at the junction prompted its reinforcement, with Fig. 8 showing the sleeve's stress plot after reinforcement at the high-stress portion.

Membrane stress, membrane plus bending stress, and total stress all measured within allowable limits (Table 2).

The high stresses shown in Fig. 7 required reinforcing the plate's straight portion, with the new stress plot shown here (Fig. 8).

Design Case 2

Case 2 examined the sleeve at ambient temperature, estimating stresses due to differential thermal expansion of the sleeve and encased header pipe. Localized high stresses registered at the thick straight plate and header pipe junction (Fig. 9). Reducing stresses requires smooth blending of all the fillet welds as shown in Fig 328.5.2A ASME B31.3 -2010,6 addressing concave fillet welds.

The second design case estimated stresses due to thermal expansion of the sleeve and encased header pipe, with the observed localized stresses due to differential expansion shown (Fig. 9).

Pressure thickness calculations for each sleeve component used numerical formulas from various references. Thickness calculations and selections considered available materials and FEA results from Table 1. Numerical formula stresses calculated for selected plate thicknesses also validated FEA results (Table 1). Numerically calculated stresses roughly matched stresses calculated by FEA for sleeve components other than the junction (Fig. 8), validating the FEA model and results.

A valve on top of the installed sleeve prevents over-pressurization (Fig. 10).

Implementation

Fig. 10 shows the installed sleeve at site. The valve provides venting to prevent over-pressurization. The sleeve has been operating successfully for more than 18 months.

References

1. ASME 31.4, "Pipeline Transportation Systems for Liquid Hydrocarbons and Other Liquids," 2006.

2. ASME Section VIII, Division 1, "Rules for Construction of Pressure Vessels," 2010.

3. ASME Section II, Part D, "ASME Boiler and Pressure Vessel Code–Material Properties," 2001.

4. ASME Section VIII, Division 2, "Rules for Construction of Pressure Vessels," 2010.

5. Young, W.C. and Budyna, R.G., "Roark's Formulas for Stress and Strain," 7th Ed., McGraw-Hill, New York, 2001.

6. ASME 31.3, "Process Piping," 2010.

The Authors

Subrata Saha([email protected]) is head of the piping department with Reliance Engineering Group at Reliance refinery, Jamnagar, India. He has a wide experience in detailed engineering for projects in the hydrocarbon and thermal, conventional, and nuclear power segments, including special expertise in trouble shooting plant problems. His areas of interest include FEA analysis and software development. Saha holds a bachelor of technology degree with honors in mechanical engineering from Indian Institute of Technology, Kharagpur, India, and a PhD from Indian Institute of Technology, Kanpur, India.

Mahendrapal Rathore([email protected]) is a pipe stress analyst at Reliance Engineering Group, RCIPTL, a Reliance Group company, handling various grassroots and greenfield projects and troubleshooting various piping and pipeline related problems. His areas of expertise are pipe stress analysis and finite element analysis. Mahendrapal holds bachelor of engineering degree in mechanical engineering from Visvesvaraya Technological University, Belgaum, India.