FSU PIPELINE DESIGN CODE COMPARED TO U.S. CODES

Alexander Aynbinder, Yuriy Tabakman, J.T. Powers, Patrick Dalton
Gulf Interstate Engineering Co.
Houston

A major pipeline design code of the former Soviet Union (FSU) has been translated into English and, where possible, compared to relevant parts of pertinent U.S. codes.

"Design of Transmission Pipelines" (SNIP 2.05.06 85) governs stress criteria and calculations of pipeline stress variables. The comparable U.S. codes are "Liquid Transportation Systems for Hydrocarbons, Liquid Petroleum Gas, Anhydrous Ammonia, and Alcohols" (ASME B31.4) and "Gas Transmission and Distribution Piping Systems" (ASME B31.8).

(SNIP is the English pronunciation of the Russian acronym for Construction Standards and Regulations.)

The opening of the oil and gas industries of the countries in the former Soviet Union to foreign investment offers great opportunities and challenges for Western companies. To take advantage of these opportunities and successfully meet the challenges, Western companies must know as much as possible about the pipeline design approach and philosophy of their FSU joint venture partners.

This is particularly true when a project must obtain the approval of the FSU ministries involved, which means the design or study is subject to technical review by FSU engineers. To satisfy these technical reviews and possibly accelerate negotiations, Western companies must know how pipelines have been designed in the FSU to date.

PRESSURE CONCEPTS

SNIP 2.05.06 85 is the FSU standard concerned only with the design of transmission oil and gas pipeline systems. This discussion concentrates on Chapter 8 of SNIP 2.05.06 85, titled "Pipeline Calculations for Strength and Stability."

One of most important concepts in pipeline design is how a code uses the term "internal pressure." The FSU code uses two internal pressure concepts in design:

The size of the load factor, which represents part of the pipeline's overall safety factor, reflects the accuracy with which loads can be predicted.

For gas pipelines, SNIP 2.05.06 85 assigns the internal pressure a load factor of 1.1. For petroleum or petroleum product pipelines, the load factor is 1.1 or 1.15 (depending on the pipe diameter and whether a storage tank is connected).

The load factor for pressure may be less than 1.1 or 1. 15 (but not less than 1) if it can be shown that the pipeline's controls and operations can guarantee that the pipeline's actual operating conditions will not vary from the design criteria.

ASME B31.4, like SNIP '2.05.06 85, uses two pressure terms (excluding test pressure):

The maximum steady state operating pressure cannot, by definition, exceed the internal design pressure. The level of the maximum steady state operating pressure may rise as a result of surges and other variations from normal operation due to upsets but may not exceed the internal design pressure by more than 10%.

ASME B31.8 uses three pressure terms (excluding test pressure): internal design pressure, maximum operating pressure, and maximum allowable operating pressure.

ASME B31.8 states that the maximum allowable operating pressure (MAOP) shall not exceed the internal design pressure. For equipment installed at compressor stations, suitable protective devices must be installed and maintained to ensure that the MAOP exceeds the internal design pressure by no more than 10%.

Comparing these concepts, the SNIP term "normative pressure" is very close to "maximum operating pressure" used in U.S. codes.

"ALLOWABLE STRESS"

A fundamental difference between the codes is that the U.S. transmission pipeline codes are "allowable stress" codes. They do not use load factors, unlike SNIP 2.05.0685.

Allowable stress design uses a factor (less than 1) multiplied by an allowable stress limit (typically the steel's specified minimum yield strength, SMYS).

This factor changes depending on which stress condition is under examination. Load factors are generally used in codes with load resistance criteria in which a predicted load has a load factor applied to it and the limits are based on the steel's SMYS or specified minimum ultimate strength (SMUS).

ASME B31.8, however, allows a strain criterion for predictable, noncyclic loads when designing offshore gas pipelines. Other U.S. design codes, such as ACI 318, "Building Code Requirements for Reinforced Concrete," have concepts similar to the SNIP factored pressure. In ACI 318, the load factor for fluid pressure is 1.4.

SNIP 2.05.06 85 lists probable loads and load effects to which a pipeline system will be subjected during installation, testing, and operation.

This list includes various loads such as dead, sustained, short duration, and special (intermittent). A load factor is assigned to each load.

A load for which the magnitude and distribution can be established with certainty is increased by a smaller load factor than a load whose exact intensity cannot be predicted with precision. For example, the weight of pipe, which can usually be computed accurately, is increased by a load factor of 1.1 or, when for calculation of buoyancy control, reduced by load factor of 0.95.

Conversely, a snow load, which is subject to greater variation, is increased by a load factor of 1.4. When calculations provide for a combination of loads, the total factored load is the sum of the loads multiplied by their load factors, the product of which is then multiplied by a reduction factor of less than 1.

(The reduction factors are found in another referenced SNIP and are not addressed here.)

In SNIP 2.05.06 85, the allowable stress remains constant regardless of whether a single load or a combination of loads is used, unlike ASME B31.4 and B31.8, which have different levels of allowable stress.

To calculate the allowable stress, SNIP 2.05.06 85 uses the SMUS and SMYS. SNIP 2.05.06 85 and ASME B31.4 and B31.8 interpret SMUS the same.

The two code systems interpret SMYS differently, however: SNIP 2.05.06 85 uses the tensile stress required to produce a residual strain of 0.2%, while ASME B31.4 and B31.8 do not specify how SMYS is determined.

The ASME codes reference API SPEC 5L, "Specification for Line Pipe," in which SMYS is determined at a total elongation of 0.5%. For a particular steel, the difference in these two values may be up to 5%.

SNIP 2.05.06 85 and FSU pipe technical specifications require ratios of SMUS to-SMYS for pipe steel different from those in API SPEC 5L. Usually, the FSU requirement for this ratio is higher.

WALL THICKNESS

The design nominal wall thickness is determined with Equation I (see equations box).

When longitudinal axial compressive stresses are present, the wall thickness is determined by Equation 2. The coefficient [see formula], is determined by Equation 3.

Equations 1, 2, and 3 are based on the stress criterion that the factored hoop stress (Equation 4) and stress intensity (Equation 5; Von Mises' combined stress) cannot exceed the code's allowable stress (S,).

The SNIP 2.05.06 85 calculation of hoop stress uses the same formula (Barlow's) as ASME B31.4 and B31.8. The SNIP uses an annulus diameter equal to the pipe's ID, however, instead of the OD used in the U.S. codes.

When considering biaxial stresses in the pipe, SNIP 2.05.06 85 uses the maximum distortional energy (Von Mises) theory, while ASME B31.4 prefers the maximum shear stress (Tresca) theory.

The wall thickness may need to be increased beyond that determined solely by Barlow's formula. In these cases, SNIP 2.05.06 85 requires that the need for this wall thickness increase (and the accompanying material cost increase) be demonstrated based on feasibility studies considering different pipeline system designs and the product temperature.

SNIP 2.05.06 85's additional requirements and limitations for the pipe's nominal wall thickness are unrelated to stress criteria and are, therefore, excluded from this discussion.

The methods for calculating the allowable and actual stresses are described later.

STRESS CRITERIA

SNIP 2.05.06 85 uses two sets of stress analysis calculations. The first set pertains to factored loads; the second, to loads set by standards or specifications.

For the first set, the stress criteria are based on the material's SMUS. The second set, which is usually referred to by U.S. pipeline engineers as "design specifications," is based on the material's SMYS.

SNIP 2.05.06 85 requires an analysis of the factored strength of buried and embankment (soil covered) pipelines in the longitudinal direction, subject to the condition in Equation 6.

For tensile longitudinal axial stresses [SEE FORMULA] is 1. For compression ([SEE FORMULA]), it is determined by Equation 7.

An analysis of the design specification's strength criteria for buried or embanked pipelines is required by the SNIP, subject to the conditions in Equations 8 and 9. For tensile longitudinal axial stresses ([SEE FORMULA]),is 1 For compression ([SEE FORMULA]), it is determined by Equation 10.

An analysis of the design strength criteria for aboveground pipelines is required by the SNIP subject to the condition in Equation 11. For tensile longitudinal axial stresses ([SEE FORMULA] is 1 For compression ([SEE FORMULA]), it is determined by Equation 12.

For a supported system of multiple spans without vibration resonance of the pipeline as a result of wind as well as axial restrained monospan straight overpasses, an analysis of the design strength criteria may be conducted subject to the conditions found in Equations 13 to 15.

Equations 13 and 14 deal with factored loads and load effects and Equation 15 with design specification loads and load effects.

Equations 13 and 14 were obtained from the limit analysis of beams, which have annular cross sections, and where the fiber stress reaches the yield stress.

The stress in the fiber consists of axial and bending stresses. Steel is assumed to be an inelastic material. A biaxial stress condition in the pipe (that is, hoop, longitudinal) is considered a uniaxial stress.

The yield strength is equal to S2 for fibers in tension and [SEE FORMULA] for fibers in compression.

If the axial and hoop stresses are zero, 144 = 1 and the shape factor' is obtained from Equation 14 is 1.27, in agreement with the solution for a plastic section modulus with annular cross section.

COMBINING LOADS

SNIP 2.05.06 85 uses one value for the allowable stress for the combination of all loads and load effects. When calculating this combination, however, it uses a combination factor less than 1 (found in another referenced SNIP).

Soil covered and aboveground pipelines have different allowable stresses. There is a different allowable stress for the analysis of each of the following:

The allowable factored stress for wall thickness calculation and other calculations involving soil covered pipelines is determined by Equation 16.

The allowable factored stress for calculations involving aboveground pipelines is determined by Equation 17.

The allowable design specification stress for calculations involving soil covered and aboveground pipelines is determined by Equation 18.

The various coefficients taken in combination may be thought of as the "design safety factor." Following is a brief description of the meanings of these coefficients and their values.

The coefficient m is very close to the design factor used in ASME B31.8. This coefficient depends on the location category and/or facility.

For SNIP 2.05.06 85 Categories Ill and IV, m = 0.9; Categories I and II, m = 0.75; and Category B, m = 0.6. For example, m would equal 0.6 for compressor station piping and 0.75 for a gas line crossing a river. For an oil line crossing the same river, however, m would equal 0.6.

Coefficient k, depends on the steel pipe manufacturing process including the welding processes, types of seam welds, heat treating process, type and percentage of inspection of parent metal and seam, and under thickness tolerance.

The values of this coefficient are 1. 34, 1.40, 1.47, and 1.55. For example, k1 has a value of 1.34 for welded pipe that meets the following criteria:

ASME B31.8's longitudinal joint factor may be considered as part of these criteria.

For welded pipe, the value of coefficient k2 depends on the ratio of SMYS to SMUS. If this ratio is less than or equal to 0.8, k2 = 1.15; if the SMYS SMUS ratio 0.8, k2 = 1.2.

The value of coefficient k2 for seamless pipe is 1.1 . SNIP 2.05.06 85 does not allow pipe materials with a ratio of SMYS to SMUS of more than 0.9.

The value of coefficient kn depends on the pipeline's diameter and pressure. It may be different for gas and petroleum pipelines.

Table 1 contains the values of this coefficient.

In Equation 1, the thickness is inversely proportional to the allowable stress, which in turn depends on the pipeline's diameter and pressure by way of coefficients. Therefore, there is indirect, as well as direct, correlation between the pipe wall thickness and diameter and pressure, which the U.S. codes do not have.

STRESS COMPUTATION

The method used in SNIP 2.05.06 85 to calculate the hoop stress was described in the previous section.

An important philosophical difference between the two code systems is that SNIP 2.05.06 85 considers the steel's elastic plastic behavior when determining the longitudinal stress for a soil covered pipeline.

The SNIP states that the design model of a pipeline should also consider the pipeline's working conditions during operation and interaction with the soil.

For straight, restrained pipelines without transverse displacement, SNIP 2.05.0685 uses Equation 19 for calculating the longitudinal axial stress.

Equation 19 is similar to the equation in ASME B31.4 (paragraph 419.6.4) for the same case. The only difference is in the fundamental understanding of E and v. According to ASME B31.4, E is the modulus of elasticity; v, Poisson's ratio, both considered independent of the stress level.

In SNIP 2.05.06 85, E is the secant modulus 2 and 1, the effective coefficient of transverse strain (determined by Equations 20 and 21, respectively). The transformed stress strain diagram is calculated from the nominal S E diagram by Equations 22 and 23.

An FSU pipeline design handbook 3 describes the calculation method for developing the nominal stress strain diagram with input data consisting of the SMYS, specific minimum tensile strength, minimum elongation, and all coefficients that determine the safety factor.

If the computed, factored stress intensity reaches the value S,, the secant modulus can be up to 20% less than the constant elastic modulus and the effective coefficient of transverse strain, up to 30% more than the constant Poisson's ratio. This allows an increase in the allowable temperature differential for a restrained pipeline while not requiring an increase in wall thickness.

RESULT COMPARISONS

Gulf Interstate Engineering has developed several programs that perform pipeline calculations meeting SNIP 2.05.06 85 requirements. These programs convert SMYS values and calculate the secant modulus and effective coefficient of transverse strain vs. allowable or actual stress as well as the wall thicknesses of pipe with compressive axial strain.

These programs provided a numerical comparison of the results of conforming to SNIP 2.05.06 85 and ASME B31.4 and B31.8.

The most common initial question asked by Western pipeline engineers designing to the U.S. and FSU codes is: Which code yields a thicker, more conservative wall thickness?

Unfortunately, a simple answer does not exist.

In the FSU code, the wall thickness depends on a variety of environment specific factors, such as those previously described.

Because of the large number of coefficients used, there are instances in which either code system may require a larger wall thickness. Thus, it is necessary to design with both codes and compare the results to determine the more conservative wall thickness.

The following example compares the results of calculated wail thicknesses for a pipeline of 1,066.8 mm (42 in.) OD and an internal working pressure of 6.50 MPa (928 psig). The pipe is manufactured to Russian steel technical standards in the first scenario and the U.S. line pipe code, API SPEC 5L, in the second scenario.

Scenario 1, Russian steel technical standard:

Steel manufactured to an FSU specification can have the mechanical properties of SMYS = 36 kg force/sq mm (51.2 ksi) and SMUS = 52 kg force/sq mm (74.0 ksi).

Then, according to SNIP 2.05.06 85 using typical coefficients (k, = 1.47, k,, = 1, n = 1.1), the required thicknesses are the following:

During design according to ASME B31.4, the SMYS value must first be calculated. For a total strain of 0.5%, the SMYS value for this steel is 36.57 kg force/sq mm or approximately 52 ksi. According to ASME B31.4, the required wall thickness is 13.2mm (0.521 in.).

Scenario 2, U.S. steel pipe standard:

Grade X 52 steel pipe manufactured in accordance with API SPEC 5L can have the mechanical properties of SMYS = 52.0 ksi and SMUS 66.0 ksi.

Then, according to SNIP 2.05.06 85, the wall thicknesses are 13.16 mm (0.518 in.) for Categories III and Iv and 15.7 mm (0.618 in.) for Categories I and II. According to ASME B31.4, the wall thickness is the same as Scenario 1 (that is, 0.521 in,) because the SMYS value is the same.

TEMPERATURE EFFECTS

Another important consideration in wail thickness calculations for liquid pipelines is the effect of the temperature differential, the difference in temperature between line installation and operation.

For example, for the previously described pipeline as a restrained pipeline with Grade X 52 pipe manufactured in accordance with API specification 5L, the thicknesses for a temperature differential of 32.2 C. (90 F.) determined by the two codes are nearly equal (0.521 and 0.518 in.).

For a temperature differential of 48.9 C. (120 F.), however, the pipe's wall thickness must be increased 10% according to ASME B31.4, while SNIP 2.05.06 85 does not require an increase in the wall thickness because of its consideration of the secant modulus and effective Poisson's coefficient.

REFERENCES

  1. Popov, E.P., Introduction to the Mechanics of Solids, Prentice Hall Inc., Englewood Cliffs, N.J., 1968, p. 198.
  2. Lankford, E.T. (ed.), The Making, Shaping, and Treating of Steel, Association of Iron and Steel Engineers, 10th Edition, Pittsburgh, 1985, pp. 1398 99.
  3. Aynbinder, A.B., Calculation of Strength and Stability, Field and Transmission Pipelines, Nedra, Moscow, 1991.

Copyright 1994 Oil & Gas Journal. All Rights Reserved.