|Oil & Gas Journal, 92:28, 7/11/1994
Patrick Dalton, D. Bevil, J.T. Powers, Alexander Aynbinder
Gulf Interstate Engineering
Writing acceptable pipe specifications for design and construction of joint-venture pipeline projects in the former Soviet Union (FSU) requires a thorough knowledge of FSU pipeline design codes.
In this occasional series comparing FSU code requirements contained in SNIP 2.05.06-85 with U.S. codes, the initial article (OGJ, Mar. 7, p. 67) provided a general introduction to both systems. The article also examined in detail their stress-analysis provisions.
(SNIP is the English pronunciation of the Russian acronym for Construction Standards and Regulations.)
This second article compares the U.S. and FSU technical requirements for line pipe in pressure piping service.
As is common in Western codes, SNIP is based on officially approved technical specifications (that is, GOSTs and TYs in the FSU system) for steels which form the starting point in the generation of specifications. ("GOST" and "TY" are abbreviations of Russian language titles that translate as State Standards/Specifications and Technical Requirements, respectively.)
For the most part, this article compares SNIP 2.05.0685 with API Specification 5L-92 (Specification for Line Pipe). When applicable, however, it also compares the FSU code with U.S. oil and gas industry codes ASME B31.4 (Liquid Transportation Systems for Hydrocarbons, Liquid Petroleum Gas, Anhydrous Ammonia, and Alcohols) and ASME B31.8 (Gas Transmission and Distribution Piping Systems).
API Specification 5L is the most widely used line pipe specification referred to by ASME B31.4 and B31.8. Pipe conforming to the standards referred to by these codes can be used for the appropriate applications (as prescribed and limited by ASME B31.4 and B31.8) without further qualification.
The importance of knowing the FSU code is demonstrated by the fact that the code permits imported pipe only if that pipe conforms to SNIP requirements.
SNIP requirements concerning pipe dimensions must also be considered.
SNIP 2.05.06-85 provides for a maximum joint length of 11.6 m (38.1 ft). If a Western company orders pipe according API Specification 5L, some joints can be as long as 45 ft (13.7 m). Although a variance on length might be obtained, the process can be lengthy and cumbersome.
Orders in such lengths may create problems in transportation on standard FSU railcars.
Also, Charpy test requirements, especially as described in ASME B31.8, are examined in more detail here because of their importance in the design of future pipelines in the Russian Arctic.
And specifying requirements for a pipe's design wall thickness according to the FSU code uses specified minimum ultimate tensile strength (SMUS) instead of specified minimum yield strength (SMYS), as in the ASME codes.
SNIP provides for the use of seamless and welded pipe for transmission pipelines. Welded pipe can be fabricated with a straight longitudinal seam, a helical seam, or other special designs approved by the appropriate ministry.
An example of a special design is the development and testing program for pipe manufactured by coiling the skelp into a multilayered pipe. Introduced approximately 15 years ago, this pipe never entered full-scale production.
Pipe manufactured from fully killed or semikilled carbon and low-alloy steels is acceptable to 500 mm (19.7 in.) OD. Pipe fabricated from fully killed or semikilled low-alloy steel is acceptable to 1,020 mm (40 in.) OD.
Pipe fabricated from low-alloy steel that is strengthened thermally (by heat treating) or thermomechanically (by heat treating and working) is acceptable to 1,420 mm (56 in.) OD. Cold expansion of the steel during fabrication cannot exceed 1.2% of the pipe circumference before expansion.
API Specification 5L pipe can be seamless or welded, the same as SNIP 2.05.06-85.
The API code has an additional requirement for helical seam pipe not part of SNIP. This requirement specifies that the width of the steel plate or skelp used in pipe fabrication can be no less than 0.8 nor more than 3.0 times the pipe OD.
API also limits helical (spiral) seam pipe to 4.5 in. (114.3 mm) OD and larger.
API Specification 5L has no requirement for the maximum amount of deformation for cold-expanded pipe. ASME B31.4, however, requires the reduction of allowable stress when pipe that has been cold-worked to meet the SMYS requirement (and subsequently heated to 300 C. [600 F.] or higher) is used in a design.
SNIP limits the maximum carbon equivalent of carbon or low-carbon, low-alloy steels to 0.46%. This requirement is independent of how the supplied pipe was fabricated (that is, hot rolled, normalized, or thermally strengthened).
Equation 1 in the accompanying equations box determines the value of the carbon equivalent for low-carbon, low-alloy steels.
For Russian carbon steels (for example, CT-3, CT.10, and CT.20 [Russian abbreviations]) and low-alloy steels that use only a silico-manganese alloying system (for example, 17TC, 17TC1 0.9T2C [Russian abbreviations]; copper, nickel, and chromium included in the pipe steel as impurities are not considered), the carbon equivalent is calculated with Equation 2.
SNIP requires that the actual test values of the carbon equivalents be shown in certificates and marked on each pipe.
For pipe manufactured in accordance with API Specification 5L, the composition must conform to the requirements in Table 3.1 of the code or be a makeup agreed upon by the supplier and the purchaser.
Table 3.1 of API Specification 5L requires that the maximum carbon content be 0.18-0.31%. The actual carbon content depends on the grade of steel and type of pipe (that is, seamless or welded, nonexpanded or cold-expanded).
There are also maximum limits on the content of phosphorus and sulfur as well as how the manganese limits vary with carbon content.
API Specification 5L-92 for the first time has included a supplementary requirement (Number 18) concerning the carbon equivalent (Equation 3). The value of the carbon equivalent cannot exceed 0.43%.
ASME B31.8 requires that carbon steels with carbon equivalents in excess of 0.65% be preheated to the temperature indicated by the welding procedure.
Additionally, it requires welds with carbon equivalents in excess of 0.65% to be stress-relieved as prescribed in Section VIII of the ASME Boiler and Pressure Vessel Code.
In these cases, the carbon equivalent is calculated with Equation 4. ASME B31.4 has similar preheating and postweld heat treatment requirements, unless these procedures can be demonstrated to be unnecessary.
SNIP 2.05.06-85 and API Specification 5L interpret specified minimum ultimate tensile strength (SMUS) the same.
The two code systems interpret SMYS differently, however. SNIP uses the tensile stress required to produce a residual strain of 0.2%, while API Specification 5L interprets SMYS as the stress that produces a total elongation of 0.5% of the gauged length as determined by an extensometer.
For a particular steel, the difference in these two values can be up to 5% and, however slight, must be considered because it affects stress-related calculations.
SNIP contains requirements for the maximum ratio of SMYS-to-SMUS and the pipe's minimum elongation. The SMYS-SMUS ratio cannot exceed the limits in Table 1.
Pipe 1,020 mm (40 in.) OD or larger must be manufactured from sheet and coiled steel that has been 100% inspected by nondestructive examination.
In API Specification 5L, each steel grade has its own mechanical requirements (that is, minimum yield strength and minimum ultimate tensile strength for that grade). The exception is Grade X-80, which also has a requirement for the maximum ultimate tensile strength.
While the SNIP has a maximum SMYS-SMUS ratio requirement for all steels, API Specification 5L has this requirement only for cold-expanded pipe: the ratio of the body steel's yield strength to the body's ultimate strength cannot exceed 0.93.
Fig. 1 shows the ratios of SMYS-to-SMUS plotted against SMYS and SMUS, respectively, for all steel grades in table 4.1 of API Specification 5L. The straight line plots show this ratio for the FSU's different types of steel (Table 1).
It is interesting to note that the Russian steels stop at a yield strength of approximately 70 ksi (480 MPa). FSU gas pipeline designs typically limit their steels to yield strengths closer to 65 ksi (450 MPa).
FSU practice also usually limits the major gas-transmission lines to pressures of 1,088 psig (7.5 MPa).
These practical considerations must be taken into account when examining Fig. 1 and interpolating SNIP 2.05.06-85 requirements to Grade X-80 steel. When the FSU code was written and approved, it was based on the Soviet Union's experiences to that point.
These experiences did not include the use of high-strength steel such as Grade X-80.
Comparing the mechanical properties of actual Russian steels used for pipe with their analogous Western steels is interesting. Russian pipe between 1,020 and 1,220 mm (40 and 48 in.) OD is typically manufactured from the low-alloy, thermally strengthened steel with the Russian abbreviation 17T1CT-(K60/42).
This steel has the following mechanical characteristics: SMYS = 60.56 ksi (after conversion to the API definition of yield stress), SMUS = 85.38 ksi, and SMYS-SMUS ratio = 0.71.
According to API Specification 5L, this steel grade is approximately equal to Grade X-60.
For X-60, the SMYS-SMUS ratio is 0.8 and the SMUS 75 ksi (517 MPa). This comparison demonstrates that for pipes with nearly equal SMYS, the minimum SMUS value for this Russian pipe is 11% more than API pipe.
In SNIP 2.05.06-85, the wall thickness depends on the SMUS, while ASME B31.4 and B31.8 determine wall thickness based on SMYS.
According to SNIP, the elongation of a fivefold standard sample (defined presently) cannot be less than the following:
A fivefold sample is described in GOST 1497-61 (Tensile Testing of Metal) as a sample in which the gauged length is related to the cross-sectional area by Equation 5. This GOST was reissued in 1984 and will be in effect through 1996.
According to API Specification 5L, the minimum elongation of a specimen with a gauged length of 2 in. (50.8 mm) must be that determined by Equation 6. A is expressed in square inches and SMUS in pounds per square inch.
For comparison, the elongations for specimens of different API Specification 5L grades are calculated according to Equation 6. The specimens' gauged lengths are 2 in. and the cross-sectional areas assumed to be (2/5.65)2; the results can therefore be compared directly to the FSU results.
Fig. 2 illustrates the results of this comparison. For this case, A is 0.125 sq in. (80.6 sq mm).
Fig. 2 shows that API Specification 5L requires larger minimum elongations for pipe fabricated from steel grades to X-46 (SMUS = 63 ksi [434 MPa]), while SNIP 2.05.06-85 requires larger minimum elongations for steels with higher yield strengths (and thus higher tensile strengths).
According to SNIP 2.05.06-85, the Charpy energy value for the pipe metal of pipe 6 mm (0.236 in.) W.T. or larger must meet or exceed the requirements in Table 2.
The testing temperature must be equal to the minimum temperature of the pipe during operation.
Table 2 illustrates that, according to SNIP, the Charpy V-notch absorbed energy (CVN) requirement depends on the pipe diameter and the operating pressure but not the wall thickness.
API Specification 5L has a minimum absorbed energy requirement only for X-80 pipe. For this pipe grade, the all-heat average of the order must be 50 ft-lb (68 J[Joules]) when tested at 32 F. (0 C.).
For other grades, this value is determined and specified by the pipe purchaser.
ASME B31.8 requires that pipe be tested in accordance with the procedures of supplementary requirement SR5 of API Specification 5L.
ASME B31.8 requires that the all-heat average of the Charpy energy values meet or exceed the energy value calculated with one of the approved equations developed in various pipeline research programs.
These approved equations are functions of hoop stress and pipe radius. Two equations are also functions of wall thickness. For comparison purposes, the equations have been expressed in terms of pressure, diameter, and SMYS.
For a design factor of 0.72, allowable stress is equal to 0.72SMYS, and the wall thickness as a function of pressure can be expressed as PD/[(2)0.72SMYS)].
Because the pipe diameter is equal to 2R, the equations in paragraph 841.11(2) of ASME B31.8 can be written in the following forms:
In Equations 7-10, SMYS is expressed in thousand pounds per square inch (ksi), D in inches, and P in pounds per square inch gauge (psig).
Fig. 3 graphs calculated CVN energies vs. gas-pipeline diameter values for different constant design pressures: 798 psig (5.5 MPa), 1,088 psig (7.5 MPa), and 1,450 psig (10 MPa). The pipe is assumed to have been manufactured from API Specification 5L Grade X-60 steel (SMYS = 60 ksi [413 MPa]).
Fig. 4 shows the calculated CVN for the four methods vs. the SMYS for pipe 40 in. (1,020 mm) OD and a design pressure of 1,088 psig.
Fig. 5 shows the calculated CVN for the four methods vs. the operating pressure for 40-in. pipe with an SMYS of 60 ksi. This figure demonstrates that the results of calculations using the various methods can differ by more than two times for large-diameter pipe intended to operate at high pressures.
For comparison with the SNIP requirements, the values obtained with Equation 7 are used. The BCL equation is internationally known and respected, its solution often used in engineering practice.
Fig. 6 shows the calculated CVNs for the BCL method (Equation 7) for two different steels with SMYS of 60 and 80 ksi (413 and 551 MPa) and the value from SNIP 2.05.06-85 (calculated from the value in Table 2 for the area under the notch of a full-size, 10 mm x 10 mm sample) vs. the pipe diameter for different design pressures.
Fig. 6a is for 798 psig, Fig. 6b for 1,088 psig, and Fig. 6c for 1,450 psig. In these figures, API Specification 5L's requirement that Grade X-80 pipe have 50 ft-lb (68 J) at 32 F. (O C.) should receive attention.
In general, SNIP's absorbed energy requirements are lower than BCL's. For a 56-in. (1,420-mm) high-pressure gas line constructed from X-60 steel, however, the SNIP requires a higher absorbed energy. In the FSU, the 56-in. trunk line is the major method of cross country gas transportation.
According to SNIP, requirements for the percentage of fibers in fractured specimens from drop weight tear tests (DWTT) apply only to gas lines of 700 mm (19.7 in.) OD and larger. The test temperature must be equal to the pipe wall's minimum temperature during operation.
The fiber percentage in the fractured surface must be determined with full-thickness samples with a height of 75 mm (3 in.) for pipes 10 mm (0.4 in.) W.T. and larger or 50 mm (2 in.) for pipes less than 10 mm W.T.
The required SNIP values for the fiber percentage for different nominal diameters and pressures are presented in the last column of Table 2.
API Specification 5L's supplementary requirement SR5 specifies that the average shear value of the fracture appearance of the three required specimens cannot be less than 60% and the all-heat average for each order (per diameter size and grade) not less than 80%.
According to SNIP 2.05.06-85, the tolerance from the specified outside diameter for the pipe ends (for a distance at least 200 mm [7.9 in.]) cannot be more than 2 mm (0.079 in.) for pipe 800 mm (31.5 in.) OD or larger.
For smaller pipe, the SNIP tolerance is that found in the approved GOST for the particular pipe in question. Table 3 contains the corresponding API Specification 5L tolerances for outside diameter.
SNIP requirements for pipe out-of-roundness (that is, the ratio of the difference between the maximum and minimum diameters to the specified diameter) are that the pipe at the ends cannot be more than 1%.
The pipe out-of-roundness for pipe 20 mm (0.787 in.) W.T. and larger cannot exceed 0.8%.
API Specification 5L does not list out-of-roundness requirements for pipe smaller than 20 in. (508 mm) OD. For pipe larger than 20 in. OD for a distance of 4 in. (101.6 mm) from the pipe ends, the maximum outside diameter cannot be more than 1% larger than specified and the minimum outside diameter no more than 1% smaller than specified.
The different criteria that SNIP 2.05.06-85 uses must be noted. For example, a pipe rejected under API Specification 5L criterion with a maximum diameter 3% larger than specified and a minimum diameter 2% larger than specified is compared with SNIP 2.05.06-85: (1.03Dspec-0.102Dspec)/Dspec = 1%.
This example demonstrates that a pipe rejected under the API code for more than 1% out-of-roundness may be acceptable under SNIP because the out-of-roundness criterion is based on the difference between the maximum and minimum diameters as well as the ratio.
According to SNIP, pipes cannot deviate from straightness by more than 1.5 mm (0.059 in.) over a 1-m (39.4-in.) length. The total deviation cannot exceed 0.2% of the pipe length.
According to API Specification 5L, pipe larger than 4.5 in. (114.3 mm) OD must be randomly checked for straightness. Deviation from a straight tine cannot exceed 0.2% of the length. Thus, SNIP has a local straightness criterion not found in API Specification 5L.
SNIP does not specify a minimum or maximum tolerance on the pipe wall thickness but does require that the allowable stress be reduced if the underthickness tolerance is more than 3%.
The GOST under which the pipe was made has the actual tolerances used in manufacturing (to determine, for example, when a pipe should be rejected).
For Grades X-42 to X-80, the API tolerances for wall thickness at any place along the pipe must be within the following ranges:
The exception to these tolerances is the weld area, which is not limited by the plus tolerance.
The most pronounced effect of API's large acceptable tolerances is that pipe manufactured to API Specification 5L to be used in the FSU must have the allowable stress reduced unless the purchaser's specification reduces the API Specification 5L underthickness tolerance below 5%.
SNIP 2.05.06-85 requires that the length of the pipe joints furnished by the manufacturer be 10.5-11.6 m (34.5-38.1 ft).
According to API Specification 5L, for a nominal double random length pipe of 40 ft (12 m), the minimum length for plain-end pipe is 14 ft (4.27 m), the minimum average length for each order item 35 ft (10.67 m), and the maximum length 45.0 ft (13.72 m). (By agreement between the purchaser and the manufacturer this tolerance applies to each carload.)
SNIP requires each pipe to be hydrostatically tested at the mill for at least 20 sec. The minimum test pressure must meet or exceed the value calculated by Equation 11.
If the hydrostatic test pressure is less than that calculated in Equation 11, the mill manufacturer must guarantee that the pipeline system, when constructed, can be tested to a pressure that induces a hoop stress equal to 95% of the SMYS.
Specification 5L requires each length of pipe to withstand an inspection hydrostatic test without leakage. The test pressure must be held for at least 5 sec for all sizes of seamless and welded pipe 18 in. (457.2 mm) OD and smaller and at least 10 sec for welded pipe 20 in. (508 mm) OD and larger.
The minimum test pressure for a particular steel grade from X-42 to X-80, outside diameter, and wall thickness is computed with Equation 12 (U.S. Customary). API Specification 5L notes that the test pressures for Grades X-42 to X-80 were limited to 3,000 psi (20.7 MPa) to accommodate hydrostatic tester limitations.
The percentage of the SMYS depends on the steel grade and the outside diameter. For Grades X-42 to X-80, the following are standard test pressures for the respective nominal pipe diameters: 5 in. OD and smaller, 60%; 6 and 8 in. OD, 75%; 10-18 in. OD, 85%; and 20 in. OD and larger, 90%.
The mill hydrostatic test pressures using the API and SNIP approaches are determined in the following manner for a 36-in. (914.4-mm) OD oil line, 0.454-in. (11.5-mm) W.T. (design factor = 0.72) manufactured from X-60 pipe according to API Specification 5L. This setup is typical for a design pressure of 1,088 psig (7.5 MPa). According to API Specification 5L (Equation 12), the hydrostatic test pressure should be:
[2(0.9)(60000)0.454]/36 = 1,360 psi.
To determine the hydrostatic test pressure requirements under the SNIP (Equation 11), the API SMYS value must be converted to the SNIP equivalent (that is, the stress required to produce a 0.2% residual strain); this value is 58,730 psi.
The minimum W.T. (using API Specification 5L tolerance of 8%) is (0-92)0.454 = 0.418 in. The pipe's ID is 36 - [(2)0.454] = 35.092 in.
Therefore, the hydrostatic test pressure is:
2(0.95)(58,730)(0.418)/(35.092) = 1,330 psi
It is interesting to compare the hydrostatic test pressure as a ratio to the design pressure (Pdesign) according to both codes because the design pressure calculated by the two code systems is different for a given wall thickness.
For a 36-in. OD pipe manufactured according to API Specification 5L for Grades X-42 to X-80, the design factor is assumed to be 0.72 for calculations involving U.S. codes.
Then, because the design pressure and the hydrostatic test pressure depend on the SMYS, the ratio of Ptest-to-Pdesign is constant (for this example at 1.25) up to the 3,000-psi limitation on Ptest.
Considering this ratio under the SNIP code, additional coefficients not found in ASME or API codes are introduced:
Under a SNIP design, the design pressure for a given wall thickness depends on the steel's ultimate strength, but the hydrostatic test pressure depends on the steel's yield strength.
After performing the algebra, it is possible to express the ratio of Ptest-to-Pdesign as a function of the SMYS-SMUS ratio. For the example shown, the ratio of SMYS-to-SMUS is taken from the values in API Specification 5L's Table 4.1. Fig. 7 compares the two different codes.
As illustrated in Fig. 7, according to API Specification 5L, the minimum required test pressure gives a constant ratio to the design pressure of 1.25.
According to the SNIP, however, this ratio changes with changes in steel grade. For steel grades up to X-60, the Ptest-Pdesign ratio according to API Specification 5L is higher.
For the more technologically developed steel grades above X-60 (that is, X-65 and stronger), SNIP 2.05.06-85 requires that, for the same operating pressure, the mill hydrostatically test the pipe to a higher pressure. Table 4 presents a numerical example considering Grades X-60 and X-80 steel for a 36-in. (914.4-mm) pipeline designed to transport oil at 1,088 psig (7.5 MPa).
Copyright 1994 Oil & Gas Journal. All Rights Reserved.