Fina

l Pro

gram

PVP 20122012 Pressure Vessels & Piping Conference

New Horizons in Global PressureVessel and Piping Technology

July 15 - 19, 2012Sheraton Centre Toronto Hotel

Toronto, Ontario, Canada

SERRATED METAL GRAPHITE FACED GASKETS FOR RING JOINT FLANGES

José C. Veiga Fabio Castro Carlos F. Cipolatti Braskem UNIB 3 Nelson Kavanagh Santo André, SP, Brasil Teadit Indústria e Comercio Ltda. Valtemir Zadoná Rio de Janeiro, RJ, Brasil Braskem UNIB 2

Triunfo, RS, Brasil

ABSTRACTMetal Ring Joint gaskets are used in piping and

equipment that operate at high pressure and/or temperature. Flange faces for these gaskets are precisely machined and small imperfections on the sealing surface make them unusable, requiring replacement or expensive field machining. During plant shut-downs time constraints may not allow for flange replacement or surface repairs. As a temporary fix gasket manufacturers offer other gasket styles like graphite faced serrated metal or spiral wound. This paper studies the use of graphite faced serrated metal gaskets in ring joint flanges. In addition to laboratory tests, also the results of a successful application of a Double-Rail gasket in a shut-down of a Petrochemical plant are shown.

.INTRODUCTION

Metal Ring Joints are used in piping and equipment that operate at high pressure and/or high temperature. The vast majority of these gaskets are produced following ASME [1] or API [2] standards. There are corresponding flange standards like the ASME B16.5 [3] that specifies the dimensions and sealing surface requirements for grooving where the Ring Joints Gaskets (RTJ) are installed.

During plant shut-downs it is common to find flanges that due to use have their sealing surface damaged. Since the sealing of a Ring Joint Gasket is a metal to metal seal, the mating surfaces must not be damaged. Just replacing the gasket may not assure a good seal at plant start-up. Depending upon the level of the damage it is necessary to machine or replace flanges, a time consuming and in some cases very difficult to do.

As a temporary fix gasket manufactures have been offering Spiral Wound Gaskets (SWG) designed to fit the flat surfaces around the flange groove. The inward buckling of these gaskets is a well known and documented problem [4, 5, 6]. The best alternative to avoid the inward buckling is to incorporate Inner Rings to the SWG design. However, due to the space limitations of the RTJ flanges this solution may not be feasible.

Graphite Faced Serrated Metal gaskets also known as Kammprofile (KAG) have been proved to be very reliable and with a sealability level similar to SWGs [7, 8,] even in very demanding applications.

This paper reports studies performed to evaluate the feasibility of the use of KAG gaskets in RTJ grooved flanges. Gaskets with two serrated rails (known as Double Rail) shown in Figure 1 and Single Rail, as shown in Figure 2, were tested.

In addition to laboratory tests, also the results of a successful application of a Double-Rail gasket in a shut-down of a Petrochemical plant are shown. Due to the lack of time to replace a flange which was damaged in operation a Double-Rail gasket was designed and installed.

FIGURE 1: DOUBLE RAIL GASKET

1 Copyright © 2012 by ASME

Proceedings of the ASME 2012 Pressure Vessels & Piping Conference PVP2012

July 15-19, 2012, Toronto, Ontario, CANADA

PVP2012-78349

FIGURE 2: SINGLE RAIL GASKET

TEST RIGS All tests were performed in ASME B16.5 welding

neck (WN) flanges manufactured in ASTM 105 forged carbon steel [9]. Figure 3 shows the 4 in Class 1500 flange used for this paper.

FIGURE 3: 4 in – CLASS 1500 TEST RIG

Heating elements and thermal insulation were assembled on the outside as shown in Figure 4. Thermocouples were installed near the gasket sealing surfaces to control the temperature.

FIGURE 4: HEATING ELEMENTS

All stud materials were ASTM SA-193-B7 [10] with machined ends to allow a precise bolt elongation measurement. The elongation was used to calculate the bolt load and gasket stress. All dimensions were measured at room temperature. Figure 5 shows the bolt machined end.

FIGURE 5: BOLT MACHINED END

TEST MEDIA Water and steam were the test media. It was chosen to

simulate high pressure service without the expensive cost of other media. At least two pressure tests were performed for each gasket design.

TEST TEMPERATURE AND PRESSUREThe first one, room temperature hydrostatic test at

385 bar (5585 psi), which is the maximum room temperature pressure according ASME B16.5 rating for a Class 1500 flange.

High temperature pressure tests were performed with steam at 170 bar (2465 psi) with daily 350 C (662 F) thermal cycles.

LEAK DETECTION The pressure decay method was used to detect leak. An additional pressure gauge was installed between the two sealing surfaces to detect leak between them as shown in Figure 6.

2 Copyright © 2012 by ASME

FIGURE 6: LEAK DETECTION

TEST GASKETS To establish a baseline for evaluation of the KAG

gaskets both oval and octagonal RTJ shapes were tested. All of them machined from low carbon steel plates with hardness less than 90 HB. No additional machining of the flange surfaces was necessary. it was used the smooth machining normal for these flanges. The gaskets dimension were determined in such way that the sealing area were centered in the flat flange surfaces.

Several combinations of KAG gaskets were tested as follows:

Double Rail (DRKAG) with serrations to seal in both flat surfaces of the flange as shown in Figure 7.

FIGURE 7: DOUBLE RAIL GASKET SEALING SURFACES

Rail (ORKAG) with serrations on the outer sealing surface of the flange as shown in Figure 8.

FIGURE 8: OUTER RAIL GASKET SEALING SURFACE

Single Inner Rail (IRKAG) with serrations on the inner sealing surface of the flange as shown in Figure 9.

FIGURE 9: INNER RAIL GASKET SEALING SURFACE

TEST PROTOCOLThe Test Protocol was designed to reproduce field

conditions of gasket installations. The ends of the studs were prepared to obtain elongation measurements with a micrometer. The stud stretch was used to calculate the bolt load and gasket stress.

A summary of the Test Protocol is as follows:

1 - Install gasket and studs. Finger tighten nuts;2 - Tighten the studs to 50 000 psi stress using 3 cross pattern rounds, followed by two more rotational patterns, as per ASME PCC-1 - 2010 [11];3 - Measure and record stud length;4 – Pressurize with water at 385 bar (5585 psi);5 - Measure and record stud length;6 – Wait 16 hours and check for pressure drop. Record the pressure value;7- If no pressure drop is observed start the temperature test increasing the temperature and pressure to 350 C (662 F) and 170 bar (2465 psi);8 – Keep the temperature and pressure for 16 hours and check for pressure drop. Record the pressure value;9- Turn heating off and wait flange to reach room temperature;10 – Measure and record bolt elongation;11 – Increase the temperature and pressure to 350 C (662 F) and 170 bar (2465 psi);12 – Repeat steps 7 though 11 for a total of 10 thermal cycles;13 – Finish test.

TEST RESULTS The following charts show test results for each representative gasket tested.

TEST RESULTS FOR RTJ GASKETS

The test result for the RTJ gaskets is shown in Figure 10. Test results show that there both gaskets styles had similar relaxation results of 10% for the Octagonal gasket and 13% for the Oval one. The horizontal red line represents the hydrostatic force at 385 bar (5585 psi). It can be noticed that after 10 thermal cycles the there is a “safety margin” of 2.6 times between the bolt load and the hydrostatic force.

3 Copyright © 2012 by ASME

FIGURE 10: TEST RESULTS RTJ OCTAGONAL AND OVAL

TEST RESULTS FOR DOUBLE RAIL GASKET

Two tests were performed with Double Rail KAG gaskets (DRKAG). The first test is shown in Figure 11.

FIGURE 11: DOUBLE-RAIL KAG GASKET

The horizontal red line represents the hydrostatic force at 385 bar (5585 psi) considering the outer rail only and the blue line the inner rail. It can be seen that with 47% gasket relaxation the hydrostatic force is greater than the bolt load, which explains the leak is during the room temperature test. The flange rotates as shown in Figure 12. This rotation increases the hydrostatic area to a point that the pressure forces are greater than the bolt load and, consequently, leakage occurs.

FIGURE 14: FLANGE ROTATION

A second DRKAG test was performed to confirm and the results were similar as the first one, as shown in Figure 13. For this a re-tightening was performed after the gasket started to leak 0.2 ml/hour . However there was not an improvement in performance as the gasket leaked at the tenth hydrostatic test at a rate of 152 ml/hour.

FIGURE 13: DRKAG WITH RE-TIGHTENING

TEST RESULTS FOR OUTER RAIL KAGA gasket with Outer Rail (ORKAG) was tested. As

expected there was leakage at the third thermal cycle confirming the expected result due to the flange rotation, as shown in Figure 14.

FIGURE 14: ORKAG TEST RESULT

TEST RESULTS FOR INNER RAIL KAGTwo gasket with Inner Ring (IRKAG) were tested as

shown in Figure 15. Both gaskets has a similar performance with no leaks and a low bolt load loss.

4 Copyright © 2012 by ASME

FIGURE 15: IRKAG TEST RESULTS COMMENTS ABOUT THE TEST RESULTS It is clear that the DRKAG has a higher bolt load loss. Further studies must be performed to find out the cause of this effect. It was also noted that the flange bending has a major impact in the DRKAG gasket performance. the inner rail looses its contact with the flange sealing surface becoming inoperative. The pressure in the Flange Ring Joint Groove is equal to the media pressure. In order to use this gasket configuration it is necessary to consider the hydrostatic force acting on the outer rail and compare it with the bolt load. If the safety margin is low than this configuration can not be used. For the ORKAG, since it exhibits the same behavior as the DRKAG, the same evaluation must be considered. On the other hand, the IRKAG showed similar results as a RTJ gasket. No leaks and similar Bolt load loss. However, due to the longer moment lever arm between the gasket OD and the bolt circle studies must be performed to assure that the flange bending will be within acceptable limits. Figure 16 shows the bolt load loss comparison between all gaskets tested.

FIGURE 16: COMPARISON OF BOLT LOAD LOSS

A comparison of the initial bolt load after installation and the remaining bolt load at the end of test is shown in Figure 17 comparing it with the Hydrostatic load. By this chart it easy to visualize the results of each test. When the blue bar is

higher or close to yellow one it indicates that the bolt load has reached an unacceptable level.

FIGURE 17: BOLT LOAD LOSS COMPARISON

RESULT OF A FIELD TESTIn April 2008 a major petrochemical plant had to install

an oval Ring Joint in the inlet flange of a super-heated stem turbine. At the installation time it was found that the flange sealing surface was not machined correctly. Re-machining would take several weeks and the plant could not start-up with major production losses. Preliminary laboratories results indicated that a Serrated Metal Gasket with Flexible Graphite facing could be a faster alternative. Since the pressure force was not high as compared with the flange rating a Double-Rail gasket was chosen in order to have a larger sealing area and avoid leaks due to imperfections of the sealing surface and. A picture of the turbine is shown in Figure 18.

FIGURE 18: STEAM TURBINE

The operating conditions were as follows:- Media: Super-Heated Steam - Flange: per ASME B16.5 size 8 in Class 2500;- Flange Material: ASTM A 182 F22 [12];- Design Pressure: 111 bar (1631 psi);- Design Temperature: 510 C (950 F);- Stud Material: ASTM A 193 B16;- Stud Torque: 8800 N-m (6490 lbf-ft);

A drawing of the gasket is shown in Figure 19.

5 Copyright © 2012 by ASME

FIGURE 19: TURBINE GASKET

In September 2009 another Steam Turbine operating under the same conditions started leaking. Upon opening the flanged connection it was found that the sealing surfaces were damaged. Considering that the first gasket was operating successfully the same solution was applied.

Both gaskets have been in operation since they were installed. Both turbines have been shut down for operational reasons without any gasket failure.

Calculating the pressure force x bolt load for this application we have the following:

- With the applied torque of 8800 N-m (6490 lbf-ft) the bolt load is 13858 kN (3115400 lbf).

- Hydrostatic Force at 111 bar: is 444 kN (99815 lbf).As we can see there is a safety margin of 31 times. This

safety margin is enough to assure a good seal. At the time of writing this paper (February 2012) both

gaskets were still running without any leaks even though both turbines had several shutdowns. No re-torque was performed during this period confirming in the field the results expected from the laboratory tests. Other turbines that had RTJ gaskets installed had to be re-torqued at the start-up due to leaks.

CONCLUSIONSIt is feasible to use KAG gaskets in Ring-Joint faced

flanges. However, a careful analysis of the operational conditions and the flange rating must be performed. It can not be considered a “generic solution” but in emergency situations, where the conditions are adequate, it is a fast and reliable alternative to avoid time and costly plant shut downs to machine damaged flange grooves. The IRKAG showed a better overall performance and should be the preferred choice if the Flange bending can be kept within acceptable limits.

REFERENCES

1. ASME B16.20 – 2007 Metallic Gaskets for Pipe Flanges, American Society of Mechanical Engineers, 3 Park Avenue, New York, NY 2008

2. Specification for Wellhead and Christmas Tree Equipment, ANSI/API Specification 6A, American Petroleum Institute, Washington, DC 20005-4070, USA.

3. ASME B16.5 – Pipe Flanges and Flanged Fittings, American Society of Mechanical Engineers, 3 Park Avenue, New York, NY 20084.

4. Understanding and Addressing Radially Inward Buckling (RIB) in Spiral-Wound Gaskets (SWG), J. M. Jenco, JJenco Inc, May 2005.

5. Failure of Spiral Wound Graphite Filled Gaskets– The Health and Safety Executive of Great Britain, March 2008

6. FSA-MG-501-02 – Standard Test Method for Spiral Wound Gaskets, The Fluid Sealing Association, 2002.

7. PVP2010-25966 - Spiral Wound versus Flexible Graphite Faced Serrated Metal Pipe Flange Gaskets in Thermal Cycling and Pressure Comparative Testing.Proceeding of the ASME 2010 Pressure and Piping Conference, Vancouver, Bellevue,Washington, USA.

8. PVP2008-61121 - Heat Exchanger Gaskets Radial Shear Testing, J. C. Veiga, N. Kavanagh, D. Reeves, 2008 ASME Pressure Proc. of the ASME PVP Conference, 2008e Vessel and Piping Conference, Chicago, Illinois, USA.

9. ASTM A-105 Standard Specification for Carbon Steel Forgings for Piping Applications – ASTM International, 100 Harbor Drive, West Conshohocken, PA, USA, 2010.

10. ASTM A-193 Standard Specification for Alloy Steel and Stainless Steel Bolting Materials for High Temperature Pressure Service and Other Special Purpose Applications – ASTM International, 100 Harbor Drive, West Conshohocken, PA, USA – 2010.

11. ASME PCC-1-2010 Guidelines for Pressure Boundary Bolted Flange Joint Assembly, American Society of Mechanical Engineers, 3 Park Avenue, New York, NY 2008.

12. ASTM A182 / A182M - 11a Standard Specification for Forged or Rolled Alloy and Stainless Steel Pipe Flanges, Forged Fittings, and Valves and Parts for High-Temperature Service ASTM International, 100 Harbor Drive, West Conshohocken, PA, USA – 2010.

6 Copyright © 2012 by ASME