Obtaining Leak-Free Bolted Joint Operation by Returning to Basics
By Warren Brown1, Wayne McKenzie2, Shane Ryan3 Abstract
Leakage of pressure vessel and piping bolted joints in refineries is an unnecessary hazard, with high associated cost, that can be easily rectified using currently available technology. There have been advances in gasket testing technology in recent years that have allowed great improvements to be made in the specification of gaskets for refinery applications. This minimizes the likelihood of joint leakage and results in reduce operating cost. In addition, there have also been advances in joint assembly procedures that have enabled significant reduction in joint assembly times, while resulting in a better final gasket stress distribution and therefore lower likelihood of leakage. This paper outlines a basis for justification of the implementation of improved joint and bolting technology in the refinery. It also details the critical calculation methods, assembly procedures and actions required to ensure leak free operation and ways of providing on-going quality assurance to a refinery leak-free program. An example of the recent implementation of a leak-free program in a refinery unit is followed in order to provide clarity on the required steps to achieve leak free operation.
A Brief Operating History of Pressure Vessel and Piping Bolted Joints
Pressure vessel and piping bolted joints have existed, in one form or another, since the introduction of pressure vessels during the industrial age. A design method for bolted joints was introduced into the 1934 ASME pressure vessel code, was updated in the 1943 code and has remained essentially unchanged since that time. The present design method, by self admission in Appendix S of ASME VIII, Div. 1 (ASME [1]), does not analyze most of the actual operational loads on the joint and, until the introduction of ASME PCC-1 (ASME [2]) in 2000, there was no official guidance on the assembly or operation of bolted joints. This is in spite of the fact that throughout history there have been major incidents involving the failure of bolted joints. The ICI risk management manual (ICI [3]) lists an estimate of the “average” number of joint leaks as 5 per year per plant. In most refineries, this is probably a low-ball estimate and probably actually only covers major leakage incidents. In a study of offshore platforms for the UK HSE (HSE [4]), bolted joint leakage was the leading cause of loss-of-containment reports with a total of 174 incidents versus the next important contributor (corrosion and erosion) at 171 incidents. Given that the process conditions offshore are generally kinder on bolted joints (lower pressure and temperature conditions), this is a fairly decent indication that estimates for bolted joint leakage on onshore equipment is an accepted occurrence that largely goes unrecorded. A search of the Chemical Safety Board’s website (www.csb.gov) reveals only a total of 22 documents containing the keywords “gasket” or “flange”. None of the accident investigations among these cases list the leakage of bolted joints as causing the incident. A similar query on the EPA website (www.epa.gov) for accident investigation reports, with the same keywords, also results in 22 documents and, once again, none of the incidents list leaking bolted joints as causing the incident. So, from that one might consider that the apparent disregard for bolted joint leakage might be justified and assume that, in spite of 1
Warren Brown, Ph.D., P.Eng, Principal Engineer, Team Leader – Fitness For Service, The Equity Engineering Group, Shaker Heights, Ohio, USA 2 Wayne McKenzie, Upgrading Turnaround Leader for SSB units, Syncrude Canada Ltd., Fort McMurray, Alberta, Canada 3 Shane Ryan, P.Eng., Plant Integrity Mechanical Engineering Lead , Syncrude Canada Ltd., Fort McMurray, Alberta, Canada
the fact we know joint leakage occurs, all joint leakage is benign in nature. But is this really the case?
Modern-Day Paradigms
If you scratch a little further below the surface what is revealed is not the low risk associated with bolted joint leakage, but in fact an i n d u s t r y p a r a d i g m w h e r e b o l t e d j o i n t le ak ag e i s c o n s i d e r e d an ac c e p t ab le p ar t o f p la n t o p e r at i o n . This is highlighted in the 2006 BP US refinery safety review panel report (Baker [5]), where it is stated at one point that a fire caused by flange leakage was a “ …minimal impact, non-event type of fire… one that we have seen many times over the years on numerous flanges and connections on our coker.”. One can only imagine what the difference in attitude would have been if this
fire had resulted from loss of containment on a weld, for example. The perception that bolted joint leakage is part of an “every-day operating fact of life” seems particularly prevalent in the refining industry. Nowhere is this more evident than in the EPA report [6] on the Tosco Avon Refinery 1997 incident in their hydrocracker unit that led to one operator death, 46 injuries and extensive damage to the refinery (Fig. 1). The cause of the explosion and fire was a run-away temperature spike in a reactor and subsequent rupture of effluent piping due to the high process temperatures (1400°F). The contributing causes identified in the report include inadequate plant maintenance (including flange leakage), however the recommendations for mitigation of similar incidents focus on management philosophy, operator training and instrumentation system design and maintenance. This is somewhat understandable, given the major contribution that failure of those systems played. However, if the paradigm of acceptance of flange leakage had not been as prevalent, then it is likely that the report would have been stronger on recommending elimination of bolted joint leakage. p r e c i p i t a t i n g In this tragic incident the pr c a u s e w a s , in fact, f la n g e le a k ag e . The report makes it clear that leakage of the feed-effluent exchangers on the unit was an accepted part of plant operations. In the days preceding the explosion, the plant operators struggled with leakage of the exchanger joints and employed steam rings and on-line leak sealing clamps in an attempt to “mitigate” the problem. This was normal operating practice at the refinery (in spite of the likelihood of flange leakage being listed as “infrequent” in the plant HAZOP). However, the leakage on one exchanger became so bad that the operators cut back the feed to one of the hydrocracker reactors in an attempt to stop
Figure 1 – Tosco Avon Refinery Incident (Figure 9 from EPA Report [6])
the leak and allow application of an on-line leak sealing clamp to the leaking joint. In cutting back the flow to the reactor they unbalanced the process, which led to the reaction runaway that caused the pipe failure. One could surmise that if the exchanger joints had not leaked, then the reactor flow would not have been diverted, reaction run-away would not have occurred and the operator would still be alive today. Therefore, at the least, f la n g e l e a k a g e s h o u l d h a v e b e e n i d e n t i f i e d a s o n e o f t h e m a j o r c o n t r i b u t o r s s to the incident. Much of the evidence from this one example highlights the industry acceptance of joint leakage: Disconnect between reality and hazard management process (HAZOP identified joint leakage as infrequent, even though it occurred regularly) Standard plant operating procedures built around a reactive approach to joint leakage (steam rings and on-line leak sealing), rather than a proactive approach (preventing the leakage in the first place). Management endorsed operating procedures included continuing to operate, in spite of a breach in equipment pressure boundary (joint leakage). The problem of joint leakage received only cursory focus in the investigation report. The report failed to elevate the importance of joint leakage on the incident and made limited, soft recommendations that joint leakage should be addressed. In the entire 105 page document, outside of the incident description, leaking joints have less than one half page dedicated to them (5 paragraphs total). •
•
•
•
If the exchanger pressure boundary breach had been from a shell or channel weld, t hen this incident probably would never have occurred. Standard operating procedures in most refineries would have called for a unit shutdown and repair (off-line) of the failure. Even if this protocol had not been followed and an incident such as this occurred, then considerable focus would have been dedicated to determining WHY the welded joint leaked. In fact, nowhere is the paradigm more evident than by comparison between welded joints and bolted joints on pressure vessels and piping. In the case of a welded joint, the components for the joint are pre-qualified as acceptable by test, the “assembly” technique is qualified and tested, the “assembler” must be qualified and tested and inspection is conducted afterwards to ensure joint integrity. In the case of a bolted joint, the gasket may not be suitable for the service and is largely untested, the assembler is generally neither qualified nor tested and there are no requirements for post-assembly inspection. Is a bolted joint really that much less important than a welded joint?
Cost to Industry
Certainly, from a cost perspective, this is not supported by recent history. The Tosco Avon Refinery incident (estimated $80MM in property damage) appears in a 2003 list of the top 100 Largest Property Damage Incidents in the Hydrocarbon-Chemical Industry (Marsh [7]). In the same list there are three other incidents (Italy - $65MM, Mexico - $136MM and Saudi Arabia - $60MM) where flanged joint leakage is identified as the cause of the property damage. This is a relatively high proportion of incidents attributable to joint leakage, considering that most of the 100 incidents do not have the cause listed. In addition, since significant unit downtime would be required to repair this damage it is likely that the actual overall cost of these incidents to the companies was even more significant. While it is easy to focus on the large incidents and associate high cost, flange leakage most likely contributes to a significant and poorly identified portion of every refineries operating budget. An on-line leak sealing clamp to seal one joint can cost in excess of $150M and $20M to fit each time leakage occurs. These costs are generally rolled into the maintenance
budget. The unfortunate aspect of this is that the clamp is usually fitted because of leakage from a $200 gasket. Evidently, there is ample reason to consider joint leakage important, but little industry focus. This is made even more remarkable due to the fact that with currently available modern technology, it is possible to eliminate joint leakage from occurring in a refinery with relatively little effort .
Why Joints Leak
The causes of joint leakage are not as complex as they might seem at first glance. In the vast majority of leakage cases, the correct selection of gasket and assembly bolt load will turn even the worst “bad-actor” joint into a leak-free “best-actor”. Once the two basic causes of joint leakage are understood, it is apparent why this approach works. These causes can be summarized as follows: 1. Joint Component Damage, for example: a. Damage to gasket during assembly due to poor technique b. Poor gasket selection for service leading to gasket failure c. Excessive flange rotation caused by high assembly bolt loads d. Gasket damage or excessive relaxation caused by flange rotation e. Gasket damage caused by differential thermal expansion of flanges 2. Insufficient Load on the Gasket, for example: a. Incorrect assembly bolt load selection b. Incorrect application of specified bolt load c. Load loss due to thermal transients d. Gasket load loss due to pressure and/or piping loads e. Excessive gasket relaxation It can be seen that correct specification of gasket, assembly bolt load and correct joint assembly will eliminate most of the above causes. This obviously begs the question “ if it is so simple, why is there still a high incidence of joint leakage”.
Be Wary of Snake-Oil Salesmen
Unfortunately, due to a lack of knowledge surrounding bolted joints in the industry, poor industry standards and a “this is the way we’ve always done it” attitude, it is very rare that an incident analysis for joint leakage will point to basic issues with the plant assembly or gasket selection procedures. There is little chance, therefore, of improving the overall reliability of the plant until the focus is taken from the reactive approach (after joint leakage) to a proactive (identify and fix the root causes of leakage in the plant). More often than not, a reactive solution will be sought that incorporates several standard “fixes” which are thought to be the general panacea for joint leakage. One such fix is the use of Belleville washers, which are usually coupled with an increased focus on joint assembly and perhaps even a gasket change. Many people have had good success with this approach and now swear by Belleville washers. However, it is most likely that their success is, in the most part, due to the other aspects that they changed and not the Belleville washers. If joint elastic interaction calculations are performed often it is found that Belleville washers will have little or no effect on the joint performance. In addition, introducing Belleville washers adds a maintenance item, as in hot services they must be periodically replaced and in marine environments, they are susceptible to Stress Corrosion Cracking. Even worse than Belleville washers are special gasket designs that are touted as miraculous cure-alls that are capable of resolving all leakage issues. A small amount of investigation into these claims will invariably lead to the discovery that there is little to substantiate them
in terms of hard data. There are several gasket designs that have been available for years and tout miraculous performance, however in each case, those same gaskets have been the known cause of joint leakage resulting in fire and plant shutdown. However, due to the lack of industry standards and methods for communicating these failures, the gaskets remain on sale with indemnity even after significant failures. Even “improved” standard gasket claims are difficult to substantiate prior to implementation in the field, as there are very few standard tests available that truly measure a gasket’s field performance. There is little that can be gained from tests such as those for the proposed leakage-based design method (ASME [8]), where a gasket is tested between flat platens in a laboratory and helium leak rates of 10-7 mg/mm/sec are measured. These tests are so far removed from the field operation of the joint, that to imply the results will have any bearing on actual joint leakage in the field demonstrates a lack of reality. Even when a “real-life” test is available, such as the RA.S.T. test (Brown [9]), the miracle gaskets are never tested, despite being invited. One basic fact must be considered when seeking resolution of leakage issues; in the plant environment, the simpler the solution (or gasket construction) the less likely it is that it will result in failure. If the leak can be solved with a change to a simpler gasket style (Kamprofile or Corrugated) and a better assembly procedure, then the chances of long-term success will be much better. The crux of a successful leak-free program is to change existing paradigms, make leakage unacceptable and create success by focusing on basic fundamental solutions.
A Top-Down Approach to Changing Paradigms
The most important asset in achieving leak-free joint operation is the people who will be involved in the effort. It is necessary to have support for the program at ALL levels of the organization in order to expect success. Unfortunately the paradigms that prevent the adoption of a successful leak-free operation are not always simple to break. In keeping with human behavior, the easiest way to break the paradigms and get such support is to demonstrate the benefit from the program at each individual level of the organization. Luckily, this relatively easy to do with a bolted joint “leak-free” program, as it is possible to put the benefits into terms that all levels of the organization will appreciate (Fig. 2). Management & Owners
Engineers & Maintenance Staff
Maintenance Technicians & Contract Joint Assemblers
Process Operators & General Refinery Staff
Cost Benefit & (Depending on Risk Tolerance Level) Safety Reduced Work-Load, Less Emergency Work & Safety
Reduced Work-Load & Safety
Safety & Reduced Work-Load
Figure 2 – Program Benefit vs. Organization Level Diagram
Not surprisingly, it is not the process operator or maintenance technician (who routinely are forced to risk their lives when joint leakage occurs), but the upper echelons of engineering and management who take more convincing that addressing joint leakage should be a high priority. The problem is that these levels of the organization are often not made aware of
the frequency of joint leakage and, if they do become aware when a suitably large incident occurs, the focus is usually on short-term measures to address the specific incident. In fact, in most plants, even the lower levels of engineering support are not advised of minor joint leakage incidents. This is because the last thing that is wanted when a joint is leaking is engineering input, which tends to be on the conservative side and will, in most cases, prevent successful sealing of a leaking joint. It is this sort of “business as usual” attitude that must be broken before a leak-free program can be justified, put into action and joint leakage eliminated. Process operators and maintenance personnel generally need no further justification for the implementation of a leak-free program than the incentive of improving the safety of their work environment and reducing the incidence of outages. For contract joint assembly personnel, who are not generally present when leakage occurs, it is necessary to demonstrate that such a program will reduce their base work-load and, most importantly, reduce the incidence of re-work that is required. As incentive to adopt restrictions on previously acceptable bolting practices (such as the carte-blanche use of air impact wrenches), the introduction of faster joint assembly procedures and better, more assembler-friendly, gasket styles can reduce the basic work-load by more than 30%. This improvement is generally enough to convince even the most skeptical individual to give the program a try. For engineering personnel the increase in safety is generally sufficiently important. However, the reduction of emergency work and the switch from reactive engineering work (trying to seal a leaking joint) to proactive engineering work (selecting the best gasket and assembly load to prevent that leakage from occurring in the first place) is generally one of the best incentives. Senior engineering and management are used to routinely seeing the cost of joint leakage in the plant operating budget. In fact, the cost of joint leakage it is rarely highlighted as an anomaly in maintenance budgets. The author once ran a query for the cost of joint leakage in a refinery cost tracking system across several units and the result was a paltry $15K spent on joint leak sealing. This was an interesting result, considering that two months previously the company had spent $50K having an on-line leak sealing company recondition and re-fit a joint clamp on a leaking joint. If specific fore-thought is not put in to tracking the cost of joint leakage, often major portions of the cost will go unnoticed. So, in order to present a real case to management for the need for a leak-free program it is often necessary to separately track to true costs of joint leakage. This sort of tracking should include the following aspects: 1. Lost Profit Opportunity (LPO) due to plant outage caused by joint leakage 2. Cost of on-line sealing services and leak-sealing clamps 3. Cost of (potential) equipment replacement due to fire 4. Cost of engineering and maintenance hours spent addressing leakage issues 5. Number of incidents occurring (even small leakage incidents, as often the difference between “small” and “large” is luck or lack of an ignition source). 6. Cost of disassembly, repair and machining of joints that have leaked during turnaround These costs can then be compared to the cost of a leak-free program and the justification usually becomes self-evident, due to the comparatively low cost of a leak-free program. The following figures (Fig. 3 and Fig. 4) show the outcome of one such cost benefit analysis conducted over the period of three years for the heat exchangers in three refineries. It can be seen in Fig. 3 that the number of incidents of joint leakage resulting in LPO recorded per refinery was not particularly high (6 in total) and was the third highest level of incidence, behind tube leakage and tube fouling. However, it was the second highest cost in terms of LPO that was recorded, with nearly $2MM in LPO over the three year period for each
Figure 3 – Heat Exchanger LPO Analysis Example
Figure 4 – Cost Benefit Analysis Example
refinery. This alone ought to be enough to justify the introduction of a leak free program. However, the justification can be made even stronger by looking at the Return on Investment (ROI) that is possible. For the highest LPO cost item (tube leaks) the solution is not always simple and can involve costly upgrading of bundle materials (>$200M per bundle), whereas a joint leak-free program can be introduced site-wide for as little as $50M to $150M the first year and $30M to $70M the following years, depending largely on the size of the plant, condition of existing equipment information and the complexity of the system.
The Plan – Return to Basics
By returning to the basic issues of bolted joints (gasket selection, assembly stress, assembly technique and joint history) it is possible to immediately eliminate a very large portion of leakage in the refinery environment. There are several distinct phases, based on the life-cycle of a joint, to such a plan and at each of the phases there is a requirement to provide personnel with the tools that they need in order to succeed. Of course, the most important part of any leak-free program, is having personnel that are open to change and supportive of the program for it to work. The phases can be divided as per the following: 1. Engineering (Joint Specification) 1.1. Correct gasket selection 1.2. Proper selection of the required assembly bolt stress 1.3. Correct location, constraint and width of the gasket sealing element 1.4. Consideration of the effects of piping loads l oads / misalignment 1.5. Quantifying the effects of gasket creep/relaxation 1.6. Quantifying the effects of temperature 1.7. Quantifying the effects of pressure 1.8. Maximum permissible assembly bolt load 2. Field Application 2.1. Improved joint assembly techniques 2.2. Ensuring assembly tool calibration 2.3. Selection of the correct nut factor 2.4. Training and qualification of workers 2.5. Ensuring correct components & loads 3. Field Improvement 3.1. Maintaining an assembly history 3.2. Learning from leakage history 4. Field Reaction to Leakage 4.1. Safe & rapid response 4.2. Reporting and tracking leakage
Engineering
For the engineering phase, it is necessary to have personnel who are appropriately trained and knowledgeable in order to make the correct decisions. They also need to be equipped with (or create) some basic tools for the calculation of the correct assembly stress, which will also aid in gasket selection. Correct Gasket Selection
The correct gasket is one that is capable of withstanding the service conditions (chemical compatibility and temperature) and is capable of sealing with the available bolt load. Gasket selection should also consider how easy it is to assemble a joint, as not all gaskets are equal in that respect and this will impact cost and potential for leakage. In addition, the gasket should ideally protect the flange facing and prevent
the need for periodic machining. For refinery services, graphite has emerged as the sealing material of choice, with kamprofile or corrugated gasket designs in exchangers and kamprofile gaskets in piping joints. These gasket types give you the largest stress range (between minimum required and maximum permissible), offer superior performance to double jacketed gaskets in resisting flange loads (Brown [10]) and are more forgiving on poor assembly techniques by comparison to spiral wound gaskets. They are, however, limited to around 850°F operating temperature (before graphite oxidation becomes an issue). For temperatures in excess of this, spiral wound gaskets will offer a longer service life (although will still ultimately be limited by oxidation) and unfilled solid steel gaskets will be required for long-term operation above 900°F. Spiral wound gaskets are suitable for both piping and exchanger services and may offer lower relaxation levels, however a more complex assembly pattern is required and the potential for damage during assembly is higher. Spiral Wound gaskets ideally should always be contained within the flange recess or groove or by inner and outer rings, in order to minimize the chances of damage during assembly. Corrugated gaskets are not recommended for piping flanges, due to the risk of a small leak rapidly escalating to a larger leak due to failure of the central metal portion of the gasket when leakage occurs (particularly in gas or two phase service). Proper Selection of the Assembly Bolt Load
The assembly bolt load is critical in determining if the joint will leak during service. In general, industry tends to focus on the minimum assembly bolt load that is required to seal a joint. More important, however, is the maximum assembly bolt load that can be applied without damaging any of the joint components. Given the unknown or ill-defined aspects of joint operation (exact levels of gasket relaxation, actual achieved assembly load, etc…) there is really no point in trying to calculate the required minimum assembly bolt load with any great accuracy. It is better to default to a maximum bolt load every time and give the joint the highest margin of relaxation between the assembly bolt load and the minimum required bolt load to seal the gasket during operation. In fact, knowledge of the minimum required assembly bolt load is really only useful for determining when additional proactive steps, such as hot-torque during start-up or more accurate assembly techniques, such as hydraulic tensioning, must be employed. The most effective method for performing this type of calculation is by keeping it simple. In a refinery environment, the full gasket width should be used to calculate the gasket area and rule-of-thumb stress levels (Table 1, determined from field experience and RA.S.T. testing, Brown [10]) for minimum and maximum gasket stresses are generally considered sufficient. These gasket stress levels are then applied to determining the required assembly bolt load (using the bolt tensile area) per the philosophy outlined in Fig. 5. This figure demonstrates the bias towards aiming for the higher end of the bolt stress. Although the other aspects that must be considered (maximum permissible bolt load for the flange, gasket creep/relaxation level, effect of temperature and effect of pressure) can be calculated (as outlined in following sections) it is also possible to perform a high-level analysis by assuming default values. These defaults can be summarized as following: 1. Maximum permissible bolt load = 100% of bolt yield for raised face ASME designed flanges, 60% of bolt yield for raised face ASME B16.5 flanges and 40% of yield for raised face ASME B16.47 flanges. For ring joint type flanges, the maximum permissible bolt load is generally limited to 40% of yield, to prevent ring joint groove cracking.
2. Gasket relaxation = 25% (joints with a single gasket) to 30% (joints with two gaskets, such as a tubesheet joint) of assembly bolt load for spiral wound, kamprofile and corrugated gaskets. 3. The effect of temperature is generally less than 10% of assembly bolt load, however it can be higher for high temperature tubesheet and large diameter joints. 4. The effect of pressure is generally thought to be conservatively modeled by assuming that the load on the gasket is equal to the assembly bolt load minus the hydrostatic pressure load. However, for harder gaskets (solid metal or kamprofile) this may not be conservative, as the bolt load can actually decrease due to pressure and therefore the gasket load is the assembly bolt load, minus both the pressure driven bolt load decrease and the hydrostatic end load. However, in most cases it will be sufficient to assume that the reduction in gasket load is equal to the hydrostatic end force.
Table 1 – Example Gasket Operating Stress & Relaxation Values Gasket Specification
Min. Req’d Operational Gasket Stress
Min. Recom. Operational Gasket Stress
Min. Seating Gasket Stress
Maximum Gasket Stress
% Relaxation 1 gasket / 2 gaskets
% Relaxation after Hot Torque
4 ksi
6 ksi
12 ksi
60 ksi
28 / 35
25 / 31
Kamprofile
4 ksi
6 ksi
12 ksi
60 ksi
32 / 40
26 / 33
Corrugated
4 ksi
6 ksi
10 ksi
60 ksi
32 / 40
26 / 33
Spiral Wound
1 2
1
2
= These values may not be applicable to all applications and caution should be used in their general application = Values are only applicable for a fully constrained Spiral Wound gasket
Thermal Thermal - Calculated Calculated from Op. Conditions 0% Sy Bolt (No Load)
X% Sy Bolt (Min. Req’d Load)
Min. Gasket Stress Required to Seal
Min. recom. Op. Load
Buffer against leakage
% Sy Bolt remaining during operation
Calculate Effect of Press. & Ext. Loading
Effect of Creep/Relax.
% Sy Bolt lost during operation W% Sy Bolt (Minimum Seating Load)
Z% Sy Bolt (Max. Permiss. Load) Y% Sy Bolt 100% Sy Bolt (Assembly Load) (Max. Possible Load) - + joint damage buffer ve ve Bolt Load +/- Assembly Assembly Technique and Procedure Accuracy GOAL = Maximize Green (%Sy Bolt to Spare)
Figure 5 – Joint Assembly A ssembly Load Selection Philosophy
Correct location, constraint & width of the gasket seal ing element
The location of the gasket sealing element can affect whether the joint will remain leak-free. As a general rule, the gasket OD should be maximized, as this has the effect of reducing the moment arm between the bolts and gasket pivot point, which in turn reduces the flange rotation and loss of gasket stress due to pressure and thermal effects. In addition, maximizing the gasket OD reduces the risk of the flange faces coming into contact outside of the gasket OD and riding off the gasket sealing element as the flange rotates. This is particularly important with spiral wound gaskets, where the flange rotation can lead to significant reduction in effective
gasket stress as more load is transferred into the outer centering ring. In the case of spiral wound gaskets in a recessed type flange, typical of many heat exchanger joints, the sealing element must never have an outer ring and therefore the clearance between the male and female of the recess should be minimized. If the clearance (on diameter) exceeds 0.125 in, then additional windings without filler should be added to the OD of the sealing element in order to prevent extrusion of the windings through the gap. As a general rule, the wider the gasket sealing element, the better. This achieves two things; it minimizes the likelihood of a radial defect in the flange face causing leakage and, by increasing the area of the gasket relative to the bolt area, it also reduces the amount of gasket stress lost due to pressure and thermal effects. However, obviously there is a limit as to how wide the gasket may be increased before the lower limits for gasket seating stress or minimum gasket stress during operation are approached. Selection of the assembly bolt load and optimal gasket dimensions can be an iterative problem. One final aspect of gasket width, is that for gaskets on exchanger tubesheet joints where a differential pressure occurs, it is good practice to increase the width of the gasket on the low pressure side as this has the effect of increasing the stiffness of that gasket, which reduces the overall gasket deflection and therefore reduces the loss of gasket stress on the high pressure side of the joint, often preventing leakage.
Quantifying the effects of gasket creep/relaxation
Unfortunately there are no standard tests performed by manufacturers for gasket creep/relaxation from which the end user can estimate gasket performance. In general, the test must be performed on a joint with similar flexibility as the intended joint in the field. The RA.S.T. testing was performed performed on a heat exchanger style flange and the radial movement that was imposed on the gasket is similar to actual joint operation in the field (even piping joints see this sort of movement). Therefore, the data available from that testing (Brown [10] and Fig. 6), provides the best estimate for gasket relaxation in a refinery environment that is available today.
Quantifying the effects of temperature
The effects of temperature are often the last straw that causes a joint to leak. In many cases, a higher initial assembly bolt load may still seal a joint that suffers severe thermal transients. However, it is useful to be able to determine what the effects of temperature on bolt load will be, prior to placing a new gasket into service. In addition, modifications to the flange design may be able to minimize the temperature effect. Calculation of the level of effect is best done by determining the joint component temperatures and the effect on bolt stress using the method outlined in WRC bulletin 510 (Brown [11]).
50 45
Corruga t ed ed / Gra phit e #5
Corruga t ed ed / Graphit e #7
Kampro f ile / Graphit e #1
Spira l W o und / Graphit e #1
Spiral Wound / Graphite #3
40 35
o i t 30 a x a25 l e R20 % 15 10
Start of Radial Shear Cycles
5
Hot Torque 0 0
20
40
60
80
T ime (hours)
Figure 6 – RA.S.T. Test results for Relaxation
Quantifying the effects of pressure
The effects of pressure are generally not significant in a well designed joint (where the gasket area is maximized and the bolt load moment arm is minimized). However, it is not always possible to assess, at first glance, if a joint is well designed without performing at least some preliminary calculations. It is also useful to be able to address the effect of possible proposed new gaskets or Belleville washers on the reduction in gasket stress that will occur. A Kamprofile gasket will, for example, be stiffer than other gasket types and therefore will lose more gasket load due to pressure. The basic design equations for the flange can be modified to include elastic interaction in accordance with the original work by Wesstrom [12] and the effect of pressure on gasket load is relatively easily calculated. A presentation of this effect with relation to the ASME [2] code equations is included in Brown [13]. Maximum permissible bolt load
The issue of maximum permissible bolt load has been debated by engineers for decades. Rodabaugh [14] demonstrated the inadequacies in the code design equations for predicting the maximum permissible bolt load and suggested the use of elastic-plastic FEA for future work. Two papers by Brown [15, 16] used elastic-plastic FEA to illustrate why, for the majority of flanges in a refining type environment, failure of the flange due to bolt overload will not be an issue. The work was expanded in Brown[17] to include an analysis of most ASME B16.5 and ASME B16.47 Series A flanges and a series of equations and checks were developed to enable the end-user to determine an appropriate maximum bolt load that a flange can take prior to significant permanent rotation occurring. These equations can be used in
conjunction with the upper limit of bolt yield and the limit on maximum permissible gasket stress detailed in Table 1 to determine an appropriate upper limit for assembly bolt load. An additional limit, based on flange rotation may be required for some gasket types. As mentioned previously, this upper limit on assembly bolt load is the key to a successful leak-free program.
Field Application
Although the previous engineering section may seem quite extensive, it would be pointless to do much of that work if the recommendations and improvements are not applied to the field. It is essential that individuals assembling the bolted joints be included in the plan of action and given training so that they understand why bolting procedures and gasket specifications are changing and how they will be an important piece of the leak-free program. Generally speaking, since the program will save them effort and result in a better success in their work, very little convincing must be done and field personnel are quick to adopt the improved practices. Improved joint assembly techniques
Prior to the release of ASME PCC-1 (ASME [2]), many companies had internal documents with similar guidance on the use of a gradual step increase in bolt load and a star pattern to ensure uniform compression of the gaskets. However, many bolting specialists in the field have worked out, through experience or trial and error, that much of the standard PCC-1 procedure is not required and, in fact, is not optimal for the assembly of the majority of bolted joints. Research into bolting patterns (Brown [18]) has shown that the key to efficient and successful joint assembly is to increase the bolt load more rapidly than previously thought. The research also demonstrated that simpler bolting patterns, other than the star pattern, are just as effective in achieving uniform compression of the gasket. In fact, for harder gasket types, such as kammprofile and corrugated gaskets, the use of a tightening pattern is really not required and the joint can be assembled with initial tightening of four opposing bolts (to ensure even initial compression) and then only circular tightening patterns at 100% of final bolt load. The Japanese Pressure Vessel Research council has recently released a standard bolting procedure that suggests using only that pattern (Sawa [19]). However, that sort of pattern may cause damage to softer gasket styles (spiral wound and PTFE in particular) and so therefore may not be the best procedure to use if a single, site-wide procedure is required. The main advantage of the newer tightening procedures is that they increase the bolt stress more rapidly and so will reduce the number of passes and effort required to assemble any given joint. They are, therefore, a key weapon in the leak-free program, as more accurate assembly methods (torque control for example) can be introduced into the plant without resulting in an increase in assembly cost or turn-around duration. Ensuring assembly tool calibration
It is one thing to calculate the optimum assembly bolt load, but if the assembly equipment in the field is not calibrated, then it is left to chance as to whether this load will be achieved or not. Calibration of the equipment serves two purposes, it ensures that individuals a sufficiently familiar with the equipment that they can achieve the correct load and it also ensures that damage to the equipment, which might greatly reduce the obtained load, is found and corrected as early as possible. Even the most accurate of methods for determining bolt load will not achieve leakfree operation if they are not calibrated for the application. Ultrasonic elongation measurement has been found by the author in one case to be reading 30% too low
by comparison to actual measured load. The studs being assembled using that item of equipment (supposedly accurate to within a few percent) would be consistently under loaded to only 70% of the specified value. The potential impact of lack of calibration can be severe, as the bolts tightened with poorly calibrated equipment will generally be under loaded. Selection of the correct nut factor
There is little industry guidance as to the appropriate nut factor to use for torque tightening. Manufacturers publish nut factors for their products, but often these nut factors are obtained using techniques suitable for aero-space or automotive applications and the resulting nut factor has little semblance to the nut factors found for the product in a refining application. One study (Brown [20]) found that the nut factors listed for anti-seize products commonly used in refinery applications were often up to 50% different from test results. This study, and a follow-up study conducted (Brown [21]) have highlighted the effect of factors such as assembly temperature, bolt material and bolt diameter on the nut factor. The effect of these factors depends heavily on the individual anti-seize product being used, and so with the appropriate test results, it is possible to select an anti-seize that shows little variation due to these factors, and therefore simplifies calculation of the required torque. An industry standard test, based on the test method used in the previously referenced papers, is presently being proposed in the ASTM F16 committee on bolting, and so therefore may become standard within the next few years. An additional factor that improves joint reliability and can often be easily justified is the use of ASTM F-463 hardened washers and the replacement of bolts at each disassembly. In order to accurately set bolt load with torque, used bolts must be wire-brush cleaned prior to re-use, this takes time and is often more expensive than replacement (particularly of smaller diameter bolts). Even with thorough cleaning of the bolts, tests have shown that the spread in achieved load is almost doubled, by comparison to new bolts (Brown [18]). This means that there is greater likelihood of joint leakage due to poorly loaded bolts. bolts. Training and qualification of workers
As mentioned previously, by comparison to the level of training and certification required by welders, the bolted joint assembler has no formal requirements to be either trained or qualified. The ASME PCC-1 committee has formed a special working group that is presently re-writing Appendix A of PCC-1 to make it more prescriptive as to bolted joint assembler qualification and certification requirements. The intent is to bring an industry standard into effect that will allow bolted joint assemblers to be certified in much the same way welders are today. This may, in the short term, increase cost to industry. However, successful adoption of these requirements by industry will undoubtedly reduce the incidence of leakage. The certification and training guidelines are based around a prescriptive training program and experience requirements that will result in three different levels of assembler qualification, with endorsements for hydraulic equipment, heat exchangers and special (non-standard) joints. As a minimum, a company that wants to eliminate joint leakage must introduce some basic training for the bolted joint assemblers in order to transfer knowledge as to why improved gaskets, assembly techniques and quality assurance techniques are required. Such training should also strive to reinforce that the leakfree program success is dependant on the bolted joint assemblers.
Ensuring correct components & loads
One potential cause of leakage, even in the best of leak-free programs, can be the use of the wrong components or assembly loads due to misidentification of the joint. It is relatively simple to introduce a joint assembly sheet or tag for each joint that identifies the required components (gasket, bolts, nuts, anti-seize and washers) and assembly techniques (assembly load, assembly pattern and assembly equipment). Using this type of approach all but eliminates the likelihood that an incorrect component or load will be used in a joint. This step is, in many ways, similar to the welding procedure sheet that is commonly used throughout industry.
Field Improvement
While implementing the engineering and field work outlined above would result in a significant improvement in leakage performance, the only way to ensure leak-free operation is to also implement ways of tracking the joint history, during both the assembly and operational phases. Maintaining an assembly history
The joint assembly sheet or tag discussed earlier can be easily modified to include a sign-off section for the people assembling and inspecting the joint. Ideally there should be a sign-off at each significant point during assembly (check correct components, hand-assemble joint, snug-tight joint and tighten joint fully). The signoff should be completed by the individual responsible for the activity. This achieves two things; it clearly signifies to the individual that they have responsibility for the activity (as they are signing it off as complete) and it enables tracking of less frequent causes of joint leakage, such as faulty assembly equipment. These record sheets should be retained during operation of the joint, as they are essential in determining possible causes of leakage and also in recommending improvement to the assembly procedure should leakage occur. Without knowledge of the last assembly parameters used, there is a big part of any leakage investigation that will be missing. Learning from leakage history
If the previous steps have been taken, and the joint still leaks during operation, then it is important to track this leakage in order to make modifications to the gasket selection or joint assembly procedure the next time that the joint is assembled. This is the only way to achieve complete leak-free operation. It is, in essence, learning from history and will often highlight site-specific requirements that may exist due to peculiarities with the process or original equipment specification. To correctly implement leakage history tracking, any level of leakage should be flagged and actions taken at the next opportunity to disassemble the joint to determine the root cause of the joint leakage. One important aspect of this is to examine the gasket upon removal from the joint to determine if there are any flaws to the gasket construction or if the gasket has been damaged during assembly or by the service in any way. The joint disassembly procedure should include a gradual release of the bolt load in a star pattern for these gaskets, to ensure that damage to the gasket does not occur during disassembly.
Field Reaction to Leakage
In the early years of implementation of a leak-free program, leakage will most probably still occur, although at a reduced frequency. It is important that reaction to such leakage is rapid and avoids escalation of the incident. There are two reasons for this; one is that a
rapid response will give the joint the best chance of being able to be sealed by increasing the bolt load during operation and the second reason is that a rapid response will minimize the chances of environmental or safety impact, which can undermine the overall success of the leak-free program. Safe & rapid response
Ideally, leakage response procedures should be determined prior to joint operation. The most common means of initial response to leakage is to attempt to tighten the bolts further. This makes sense as, unless the flange face or gasket is permanently damaged, the most likely cause of leakage will be low gasket stress due to creep/relaxation. Even if damage to the gasket has occurred, it is often possible to seal the joint by increasing the gasket stress. However, the problem with increasing the gasket stress is always in determining the limits as to how much additional bolt load can be applied. In most cases, a torque based on bolt yield and a normal assembly nut factor is used as an upper limit. The problem with this is that operation at temperatures above 350°F for an hour or operation for several days at ambient temperature will result in the anti-seize product becoming ineffective, due to the loss of the lubricating properties of the carrying oil. To overcome this, on some sites the procedure is to remove one bolt at a time, re-lubricate it and then tighten it to the higher assembly bolt load. The problem with that approach is that by removing the bolt you are decreasing the gasket stress local to that bolt and therefore increasing the likelihood of leakage. Since someone has to be at the bolt location to perform the operation, such a procedure is placing that individual at risk of injury and therefore the method is generally limited to non-hazardous services. A better method is to use a “turn-of- nut” approach for increasing the bolt load. This provides a safer approach to eliminating leakage. Since the gasket stress is always increasing there is minimal likelihood that leakage will increase during tightening activities. The “turn-of-nut” approach assumes that, because the gasket is leaking, the gasket stress is lower than a minimal limit (Table 1) and therefore must be increased to a more suitable stress level above that limit. The approach must include calculation of the flange rotation, gasket compression and bolt elongation, versus the amount of reduction in bolt length that is achieved by turning the nut a set amount (based on the threads per inch). In actual fact, particularly at lower bolt loads, there is flexibility in the threads of the nut that must also be accounted for in order to more accurately estimate the increase in bolt load. However, neglecting this effect will be conservative, in that the actual increase in bolt load will be less than that calculated, and so therefore this complexity may be neglected. The “turn-of- nut” approach allows tightening of a leaking joint by specifying the number of flats that the nut must be turned in order to raise the gasket stress the desired amount. Since the number of flats that the nut must be rotated is unique to the joint configuration, it may be calculated in advance for each joint and, if leakage occurs, someone can be deployed into the field to perform the tightening with minimal preparation. Rapid response to leakage is required due to the fact that most leaking mediums will damage the gasket or flange and therefore prevent re-sealing of the joint without the use of external leak-sealing clamps, which are an expensive and cumbersome solution. High pressure steam services or gas services that contain small amounts of liquids are the worst services to re-seal once leakage has commenced, due to cutting of the flange face. Once the gasket or flange face is cut, an external leak-sealing clamp or wire-peen will be required to stop the leak. Repair to the flange face and damage caused by the on-line leak sealing will also be required at the next turnaround. This process can be quite expensive and often will damage the gasket,
preventing post-incident inspection of the gasket for the cause of leakage. If the leak is caught when it is only a wisp, then it is usually possible to re-seal it by only tightening the bolts, thus avoiding the additional cost. Reporting and tracking leakage
As mentioned previously, it is essential for improvement to the leak-free program that leakage is tracked. However, since the tools will be available to rapidly deal with a leak, there is a possibility that reporting of the leaks will not occur. This will not allow the improvement of assembly techniques or gasket selection to eliminate leakage and will effectively undermine the program. A system of tracking leakage and the steps taken to eliminate leakage must be introduced to ensure continual improvement may be implemented. Recording the steps taken to stop the leak will enable easy assessment of possible improvements that may proactively eliminate leakage. For example, if a given nut rotation successfully sealed the joint then this may point to the fact that a higher initial bolt stress or, more likely, the use of a hottorque procedure on the joint during start-up, will enable it to operate without leakage in future. The most effective way of tracking leakage is to have a system that requires the user to enter in details about the joint leakage prior to obtaining the instructions on the use the “turn-of- nut” approach. In this way, they can then be forced to return to the system to update the joint history on the effectiveness of the tightening attempt. This reduces the likelihood that leakage will not be effectively tracked.
Putting the Plan into Action
The preceding items may seem somewhat daunting at first, however once the program is put into action, they tend to fall into place quite well and are easier to understand as they are applied in the field. If suitable tools are available (software, for example) to assist the team, then the full plan may be put into place from day one. However, in most cases a phased approach to the introduction of the leak-free program will assist in understanding and also serves to avoid requiring too much time from key individuals during the initial phase. The key to a phased introduction is to keep the initial phase as simple as possible and to focus on 90% of joint leakage that is easily resolved. This is best achieved by the introduction of gasket assembly stress calculations, improvement in joint assembler training and assembly techniques and the introduction of a joint assembly and leakage tracking system. These three items will often eliminate the majority of nuisance leakage in the refinery and leave only the true “worst-actors” continuing to leak. This then reduces the number of joints that must be further assessed to determine why they are leaking and what may be done to prevent them leaking. In summary, the initial phase should focus on determining the best (maximum) assembly bolt load to be applied (see Fig. 7), transfer this into the field by training and the use of joint assembly sheets or tags and then continue to track the assembly and operational leakage using the joint assembly sheet or tag. If software is not available, then a spreadsheet may be constructed that has dimensional inputs (flange, bolts and gasket) and gasket stress limits (similar to those in Table 1) in order to determine the maximum assembly bolt stress for each joint. The same spreadsheet may also be used to print a joint assembly sheet or tag and, if a comments section is added to the bottom of the sheet, the spreadsheet can be used to track the assembly history and any operational leakage. This can be somewhat cumbersome, given that each item of equipment will need a workbook and each joint on the equipment will have a spreadsheet, but it is an effective way to manage a leak-free program when joint assembly software is not available.
Once the initial phase is successfully put into operation, changes to gaskets and other higher level calculations may then be performed to eliminate the worst-actor joints that have not been resolved by the initial phase. Increase Bolt Load, Reduce Gasket Area or Change Gasket Type N 1
Assign Initial Assembly Bolt Load
Stress < Maximum Allow. ?
Calculate Achieved Gasket Stress
2
Stress > Minimum Seating ?
Y
Y
N
Increase Gasket Area
Subtract Relaxation (25-30%) Thermal Effects (10%) & Hydrostatic End Force
Decrease Bolt Load or Increase Gasket Area
N
3
4
Assembly Bolt Load Selected
Y
Bolt Stress > 40% Sy
Y
Stress > Minimum Operat.?
1 = 2 = 3 = 4 =
N
Gasket stress should be less than the maximum allowed for that gasket (Table 1) Gasket stress should be greater than the minimum seating stress allowed for that gasket (Table 1) Gasket stress should be greater than the minimum allowed operating stress for that gasket (Table 1) Bolt stress should be greater than 40% of bolt yield at ambient, less than that indicates a gasket that is too narrow.
Figure 7 – Assembly Bolt load Selection Procedure
Case Histories 1 – Plant-Wide Introduction of the Initial Phase
The initial phase of a leak-free program was recently introduced into a small chemical plant in Canada. After one day of training for the engineering and maintenance department they proceeded to implement change to their methods of calculating assembly bolt load. Prior to implementing the new program, they actually had an advanced system of calculating the required assembly bolt load using a combination of code equations and PVRC leakage based calculations (ASME [8]). This already required the use of a spreadsheet for each joint and these spreadsheets had been written to provide a joint assembly sheet for the field. So, on the surface their existing systems for bolted joint maintenance were relatively advanced by industry standards. However, they were continually getting nuisance leakage and fires on their exchangers and piping. The key aspect that caused this leakage was the fact that the calculations they were using to determine assembly bolt load were not giving correct results and they were, for most joints, j oints, grossly under loading the gaskets. Upon incorporating the recommendations outlined in this paper, a review of the specified assembly bolt load that they had been using, versus the recommended new assembly bolt load, revealed an average increase in bolt load of 154% over 91 joints. Some bolt loads were increased by 3 times the previous assembly bolt load. The chance of joint leakage was therefore significantly reduced by obtaining a significant additional buffer (54% of load on average) against joint leakage. Since they also checked the minimum operating stress, as outlined in the paper, it was possible for them to determine which joints were more at risk of leakage than others and to flag those for hot-torque during start-up. In addition, they were able to determine which gaskets should be modified (gasket width increased) by looking at those joints where assembly bolt load was less than 40% of the bolt yield. In effect, by switching to a simpler system of assembly bolt load calculation and focusing on the important markers for joint leakage (assembly and operating gasket stress levels) it was possible to significantly increase their chances of successful operation without leakage. In addition, they were able to use the “turn-of-nut” technique to retro-actively increase the operating bolt load of exchanger joints which were flagged as having the potential for leakage due to the low previous joint assembly load. Having part of the initial phase in operation already (joint assembly sheets) made the location of these potential issues possible and highlighted the usefulness of recording joint history. Introduction of the later stages of the program (optimization of gasket selection and continued joint history tracking) will now be phased in over the coming years, with a focus on any joints that continue to have leakage issues. 2 – Resolution of Problem Exchangers
Once the initial phase is introduced, the joints that still leak required closer inspection to determine the root cause of the failure. The following outlines analysis of two exchangers and highlights how simple changes to the gasket and assembly bolt load result in elimination of joint leakage. The first example is a type BEM (fixed-fixed tubesheet) exchanger that leaked and caught fire (Fig. 9) during an upset in operation where cold fluid was spiked into the tubeside. The root cause of the leakage was determined to be the low assembly bolt load, with a secondary contributing factor being the narrow gasket width. Finite Element Analysis using the ABAQUS FEA program (Abaqus [22]) was performed on the exchanger to demonstrate the concepts explained in this paper regarding operating gasket stress and the influence of assembly bolt load on leakage. The advantage of using the Abaqus program is that the actual elastic properties of the gasket can be accurately modeled. However in many cases this added complexity is not required and hand calculations generally will also reveal the basic issues. The exchanger gasket stress versus time in operation is shown in Fig. 10. The
effect of gasket creep/relaxation and off-loading of the gasket due to pressure can be seen in the initial steps of the graph. The present case shown on the graph was with the gasket dimensions that leaked, and it can be seen that from an initial assembly gasket stress of 17ksi, the operating gasket stress (neglecting thermal transients) is only around 9ksi. When the thermal transient hits, within 5minutes (and again after about an hour) the gasket stress is reduced a further 4ksi and this leaves the operating gasket stress below the recommended minimum gasket stress of 6ksi for the kamprofile gasket used in this joint. j oint. By increasing the gasket outer diameter and the width from 0.5 in. to 1.0 in., the effect of both the pressure and thermal loading can be seen to be dramatically reduced. If this modification is coupled with an increase in assembly bolt load, the actual operating gasket stress is now kept at around 12ksi, even during the thermal upset condition. Since this gasket stress is well above the recommended minimum gasket stress limit of 6ksi it can be seen that a simple change in gasket dimensions and optimization of the assembly bolt load (which combined probably added no additional operating cost to the plant) resulted in elimination of leakage from this joint. Previous attempts at rectifying the problem included changing the gasket type and using external leak sealing clamps, which were expensive and did not work. It can be seen that identification of the root cause and implementation of a proactive solution was much more effective in this case. Figure 9 – Failed Exchanger Joint 25
Initial Assembly Bolt Load 20
Stress after Gasket Creep/Relaxation Stress after Hydrostatic End Force (Pressure Applied)
) i s k ( s s 15 e r t S t e k s a G e g 10 a r e v A
Start of Upset Cycle
Loss due to Thermal Transient
5
Present C ase
Recommended C ase
40
60
0 0
10
5min
20
30
50
70
Time (minutes) (minutes)
Figure 10 – Exchanger Gasket Stress versus Time
80
90
The second case study involves an exchanger that was classified as a worst-case actor and required continual attention, including on-line leak sealing to resolve leakage issues. The exchanger had continual leakage issues on the high pressure shell side of the tubesheet joint. The exchanger shell contains hydrogen and so leakage from the joint was considered serious. The joint originally had solid steel gaskets installed, although the operating temperature of the joint was less than 800°F and so graphite was an option. The joint details were run through the VCE IntelliJoint program and a comparison between the original gasket, bolt load selection and dimensions (Fig. 11) and a revised gasket selections (Fig. 12) revealed that, with a change of gasket and increased assembly bolt load, the joint was far less likely to leak. The original gasket dimension (0.656in. wide) limited the maximum allowable bolt assembly stress to 22ksi, due to the maximum allowable gasket stress (taken as twice yield for the solid steel gasket). Even though the joint was normally assembled to 50% of yield, the additional bolt load was not effective in eliminating leakage, since it exceeded twice yield. It can be seen, that changing the gasket to a Kammprofile type gasket with the same gasket dimensions would probably have resolved the leakage issues, even though the assembly bolt load would have needed to be reduced to 22% of yield. Since this sort of bolt load is not desirable, the gasket width must be increased, which has a number of other advantages, as mentioned earlier. The results of increasing the gasket width can be seen in Fig. 12. Two things are evident from this graphic, one is that increasing the width of the solid metal gasket would have dramatically improved the performance and most likely would have resolved the leakage issues. The second is that, by using a Kammprofile gasket and increasing the gasket width to 1.25in., there is now a 15% buffer between the assembly bolt load and the minimum required operating gasket stress. This is obviously a significant advantage over the original gasket selection (which had no buffer between assembly and operation). In fact, the graphic for the revised gasket width Kammprofile indicates that the joint assembly could be reduced to hydraulic bolt torque and leakage would still not be expected. If successful operation is obtained for the next run with the revised gasket type and width, the assembly technique may be revised at the next turnaround.
Figure 11 – Second Example, Original Gasket Width
Figure 12 – Second Example, Revised Gasket Width
Conclusions
This paper has outlined how focusing on the basic issues that cause joint leakage can result in the elimination of joint leakage in refineries. The most important aspect that must be overcome prior to industry-wide improvement in joint leakage is the paradigm that joint leakage can not be eliminated. Modern techniques in gasket testing, joint analysis and joint assembly have given industry the tools that are required to implement leak-free programs and eliminate joint leakage. Using the techniques outlined in this paper, it is possible to justify a joint leak-free program program and eliminate joint leakage in a methodical methodical manner.
References
[1] ASME. 2004, ASME VIII, Div 1, Boiler and Pressure Pressure Vessel Code, Appendix Appendix 2, American Society of Mechanical Engineers, NY, USA
[2] ASME PCC-1. 2000, “Guidelines for Pressure Boundary Bolted Flange Joint Assembly”, American Society of Mechanical Engineers, NY, USA [3] ICI Engineering Department, “Process Safety Guide 14 – Reliability Guide” , ICI UK PLC
[4] Health & Safety Executive, UK, 1999, “ Corrosion Risk Assessment and Management for Offshore Processing Facilities ”, Offshore Technology Report 1999/064, UK HSE [5] Baker Report, 2007, “ The Report of the BP U.S. Refineries Independent Safety Panel Review ”, ”, BP, USA
[6] EPA Report, 1998, “ EPA Chemical Investigation Report – Tosco Avon Refinery, Martinez, California”, EPA Report 550-R-98-009, Environmental Protection Agency, USA [7] Marsh, 2003, “ The 100 Largest Losses, 1972-2001, Large Property Damage Losses in the Hydrocarbon-Chemical Industry ”, ”, Report by Marsh Risk Consulting Practice, 20 th Edition, February 2003 [8] ASME-BFJ draft 2000, 2000, “Background and Commentary on Appendix BFJ - New Rules for bolted Flanged Connections with Ring Type Gaskets”, Internal correspondence of ASME Special Working Group BFJ. [9] Brown, W., 2003, “Recent North American Research into Several Pressure Vessel Bolted Joint Integrity Issues”, Proceedings of the ICPVT 10 , Vienna, Austria [10] Brown, W., 2002, “The Suitability of Various Gasket Types for Heat Exchanger Service”, Proceedings of the ASME PVP 2002 , ASME, Vancouver, Canada, 433, pp. 45-51
[11] Brown, W., 2006, “Analysis of the Effects of Temperature on Bolted Joints”, Welding Research Council Bulletin 510 [12] Wesstrom, D.B., Bergh, S.E., 1951, “Effect of Internal Pressure on Stresses and Strains in Bolted-Flange Connections”, Transactions of ASME, 73, n.5, pp 508-568, ASME, NY, USA [13] Brown, W., 1993, “Design and Behaviour of Bolted Joints” Proceedings of the 3 rd International Conference on Fluid Sealing, CETIM, Nantes, France, pp. 111-113 [14] Rodabaugh, E.C., Moore, S.E., 1976, “Evaluation of the Bolting and Flanges of ANSI B16.5 Flanged Joints – ASME Part A Design Rules”, ORNL Report W-7405-26, Oak Ridge, Tennessee, USA [15] Brown, W., 2006, “Considerations for Selecting the Optimum Bolt Assembly Stress For Piping Flanges”, Proceedings of the ASME PVP 2006 , ASME, Vancouver, Canada, PVP2006ICPVT11-93094 [16] Brown, W., 2006, “Flange Assembly Bolt Load Selection Based on Leak before Break Analysis”, PVP2006-ICPVT11-93075, PVP2006-ICPVT11-93075, Proceedings of the ASME PVPV conference, Vancouver, Canada [17] Brown, W., 2007, “An Update on Selecting the Optimum Bolt Assembly Stress For Piping Flanges”, Proceedings of the ASME PVP 2007 , ASME, San Antonio, TX, USA, PVP2007-26649 [18] Brown, W., 2004, “Efficient Assembly of Pressure Vessel Bolted Joints”, Proceedings of the ASME PVP 2004, ASME, La Jolla, USA [19] Sawa, Sawa, T., Kobyashi, T. Presentation during during the ASME ASME PVP 2006 in Vancouver Vancouver during during the Panel Session on Flange Design Methods. [20] Brown, W., Marchand, L., Lafrance, T., 2006, “ Bolt Anti-Seize Performance in a Process Plant Environment ”, ”, Proceedings of the ASME PVP 2006, Vancouver, Canada, PVP2006PVP2006ICPVT11-93072 [21] Brown, W., Marchand, L., Evrard, A., Reeves, D., 2007, “ The Effect of Bolt Size on the Assembly Nut Factor ”, ”, Proceedings of the ASME PVP 2007, San Antonio, TX, USA, PVP200726644 [22] Abaqus v6.5-4, General Finite Element Analysis Software, Dassault Systèmes, Rhode Island, NY, USA [23] VCEIntelliJoint, Pressure Vessel and Piping Bolted Joint Analysis Software, The Equity Engineering Group, Shaker Heights, OH, USA