Deficiencies Observed in Design, Engineering, Procurement, Construction & Operating Phase of Reciprocating Compressors and Resolution of the Problems.

Reciprocating compressors are used for high pressure and critical application in all industries. These equipment need through review of design of various components for a failsafe design. API-618 specifies various details about construction of various components. But due to OEM design by compressor vendor, proper review is not done or due to cost economics of equipment these are not given importance. API RP 686 is a recommended practice for machinery installation and installation design. However due to the constraints in timelines and project deliveries they may be some lacuna in erection and commissioning of these machines. Also even though approved SOPs and operating manuals specify the safe operation of the machinery, excursions or aspects beyond the machinery boundary impact the safe operations of the machinery as well.

There are many examples when accident took place when these issues were ignored or correct assembly procedure for assembly of components is not followed and there were mistakes in design of piping system or a hurry to commission the machinery at the cost of reliable or sustained safe operations.

In this paper some such critical deficiencies observed in various phases in the asset life cycle of the reciprocating compressors i.e, Design / Engineering & procurement phase, Construction and Commissioning phase and finally the sustained operation phase are discussed in detail. The issues were observed in the lifecycle of a refinery’s CCR Platformer units Net Gas Compressor and HNU Unit’s make up hydrogen compressors.

In the design, engineering & procurement phase, The issues described are designs of unloaders and venting system, unloader and vent piping, cylinder tap off connections, end clearance pocket unloaders, distance piece and crank case venting system. In the construction and commissioning phase the issues described are related to civil foundation aspects,? purge system for motors, requirement of startup fine mesh strainer & pulsation studies for piping supports.

In sustained operation phase we described about supporting of SBCs or small bore connections which play a crucial role in reliability as well as safety aspects and also about issues in lubricators. It may be noted that some issues like motor purge and SBC supporting fall under multiple phases i.e, engineering & procurement, construction & commissioning as well as sustained operations.

?All issues have been discussed in paper and various solutions and fail safe design have been suggested to improve safety and reliability. The studies also indicates that proper start up procedure and follow up of correct checklists during overhauling and start up can eliminate/ identify such mistakes.

A. DESIGN / ENGINEERING & PROCUREMENT PHASE ISSUES:

A1. VALVE UNLOADERS DESIGN ISSUES:

Valve unloaders are installed on suction valves; it consists of valve cage, fork assembly, valve cover, unloader stem and sealing arrangements. Top portion of unloader assembly is called air actuator and consist of spring, diaphragm/piston assembly, air intake points and position indicator.

Air actuator may be of direct acting, reverse action. In directing acting unloader assembly air supply directly presses piston assembly for unloading of valves. In case of reverse acting spring force keeps valve unloaded condition. Whenever valve is to be loaded, air supply pressure acts in reverse direction and releases spring force. Now these days direct acting unloaders are preferred.

?Unloader stem assembly sealing is done by O Rings, O Rings with back up rings and throttle bush on high pressure side. The sealing arrangements are designed in such a way that intermixing of process gas with instrument air do not take place. For this venting systems are designed, it may be single, double, triple venting system.

These are following areas where leak and problem have been observed:

1. Stem to valve cover sealing: This is the most problematic area in this design and there are always issues of leakage from this joint. Some vendor uses lead gasket/soft iron gaskets at this location. Once this joint is leaking, efforts are made to attend leak by tightening the lock nut for giving compression to gasket.

2. It is essential to know how stem sealing is designed, Leakage through single o ring may lead to leak from both side (air & hydrocarbon side) of stem. In case of excessive leakage from hydrocarbon side, it can lead to loss of hydrocarbon to vent and can create unsafe condition.

3. Chances of air contact from vent off line (in case there is leak from air actuator side, stem o ring is not holding)

4. Chances of screw becoming out of adjustment during leak attending process due to tightening of locknut. This may result in non functioning of valves wherever there is appreciable change in overall length of unloader assembly. This change in assembly length may lead to partial unloading of valve or ineffective unloading of valve as unloader fork assembly is not touching valve plates.

5. Airs vent point and leak of point if open to atmosphere, there are chances of buds blocking these passage which can result in non functioning of unloader.

6. Some time due to unloader design it is difficult to know from outside correct position of unloader stem position.

?

IMPROVEMENTS DONE:

1. Valve unloaders should be pneumatic type where connection is provided from side on top area and top most portion is left for valve unloader position indication.

2. Stem sealing arrangement was reviewed for leakage path and arrangement for sealing. Single O Ring design for sealing should not be considered for hydrocarbon service. Instead bush combined with double o ring/back up ring design was considered for replacement. Unloader side sealing can be of single o ring sealing.

3. API 618 also addresses this issue in section 2.7.13 (Pneumatic unloaders shall be designed so that air used for unloading cannot mix with the gases being compressed, even in the event of failure of diaphragm or another sealing component. A threaded gas vent connection shall be provided at the stem packing.)

4. Breathers were installed in Air vent connection on unloader to avoid chocking from insects/bugs. A choked vent connection will not release air and it will lead to non-functioning of unloaders.

5. Unloader sliding pushrod exposed to atmospheric condition was replaced with corrosion -resistant material. (Please note When trace quantity of H2S are present in process gas, or the amount of H2S is uncertain, Components to which NACE requirement apply shall also include clearance pockets, valve covers all component with in cylinder)

6. Tubing for valve unloader leak off and air supply connection (shall be) were modified to be of different size for distinguishing two different services and avoiding mistake during maintenance.

7. Each individual unloading device was provided with a visual indication of its position and its load condition (loaded or unloaded).

?

A2. ISSUES WITH UNLOADER VENT CONNECTIONS:

Following issues were encountered with unloader piping/tubing:

?

1. It is seen that little care is taken for various vent /air connections tubing while deciding layout. There are various issued observed with supply made by compressor suppliers. Some of the issues are improper tubing lay out, supporting. As in most of the cases unloader combined vent connections are left to the owner, little effort is given to proper layout /design.

If these issues are not corrected it may lead to snapping off tube or piping due to vibration. Improper attention during maintenance may end up in intermixing of air supply / vent tube after maintenance work and if it goes unnoticed it may result in problem during start up or can be serious problem due to mixing of vent and air supply.

?

2. Sometime unloaders Vent lines are combined together to a common header. While laying piping routing and supporting is left to contractor’s team as no any drawing available for the same.

?

A CASE STUDY INTERMIXING OF UNLOADER TUBINGS AND NON DETECTION LEADING TO SAFETY INCIDENT:

One risk, considered to be non-credible by the designer, is the contamination of the instrument air with flammable gas. The instrument air may get contaminated with hydrocarbon in unloader leaking from stem. The vent tube from unloader that should have been connected to flare may get interchanged with the instrument air connection. In this event, if stem sealing arrangement is leaking and service is high pressure service, instrument air will get contaminated with hydrocarbon.

?

The same mistake if not noticed & compressor started it may lead to the incident where compressor can show showed a strange vibration pattern, instrument air header get contaminated. If no precaution was taken to ensure that the tubing was correctly reinstalled after the overhaul there are 75% chance to make the fatal mistake.

?

As can be seen on below figures, the tubing to the clearance pocket unloader is installed in parallel on the top of the cylinders. The tubes have the same dimension, 6 mm and there is no marking which tube is for the instrument air and which for the leak gas from the space between the 2 seals.

Several plants reported safety related incident when intermixing of tube goes unnoticed. The sequence of event leading to incident are given below

?

Equal size of the air and vent tubing --> Connecting couplings are next to each other --> Tubings are not properly clamped at different locations --> Low awareness of the impact of intermixing of these 2 services --> maintenance procedures clearly do not differentiate the identification and work required on both service tubings --> SOP for checking unloading Pattern prior to start up was absent --> Contamination of instrument air with Hydrogen --> Incident.

?

In order to prevent such incident, it is essential that there should be some differentiation between two connections so that chances of mistakes can be avoided. It was be achieved by –

1.?????? Keeping tubing size different for air or vent connection

2.?????? Different colour of air or vent tubing. Normally blue colour for air and red colour for vent tubing.

It is necessary to ensure that after each maintenance on compressor valves/overhauling, unloading pattern should be checked and verified for sequence of unloading steps. If any valve unloading pattern is not matching, it should be corrected before compressor is taken in service.

?

?

A3. ISSUES WITH CYLINDER TAP-OFF CONNECTIONS:

?

Cylinder tap off connections are provided for taking compressor pressure profile to enable condition monitoring of reciprocating compressor by taking pressure/temperature profile for drawing PV diagram. Tapped holes are provided on cylinders and normally these are kept plugged. For taking profile reading, Kiene valves are installed on these tap off connection after removal of plugs. Normally for tap off connection a drilled stud with gasket is mounted on cylinder.

Some of the cylinder design uses cover plates cover for jackets. Cover plates are kept in position by bolts mounted on cylinder body tapped holes. Cylinder tap off connection bolt is also used for tightening cover plate. These designs have issues related with –

1.?????? High pressure gas causing failure of cover gasket or

2.?????? Gas entering the jacket water system.

3.?????? Failure of tap off bolts

Although relief is provide between tap off bolt and cover plate, but adaptor nut prevents passage of leaking gasket (in event of leak) and in this case pressurization of jacket may take place as jacket cover gaskets are designed for jacket side water pressure. There is one more problem associated with this design that length of adaptor bolt is higher and during tightening/ fixing of Kiene valves there are chance of breaking this bolt.

?

Better design for tap off connection –

API-618, Figure below indicates fool proof arrangement and will not lead to mixing of cylinder gases to water jacket in event of leak from gasket. Indicator connection is through solid boss in cylinder casting and is not exposed to water jacket.

For cylinder with plate cover typical arrangement is given in sketch which indicates free space for gas release in event of leak from tap off connection bolt as given in picture below:

B. CONSTRUCTION AND COMMISSIONING PHASE

?

B1. CIVIL FOUNDATION – POCKET ISSUES.

Reciprocating compressors come either in completely assembled skid mounted or in separate components which are to be erected at site and assembled. In either case, a civil foundation with pockets for foundation bolts is a requirement. In one such machine, the civil foundation was made with pockets. But the inside of the pockets were lined with MS plates. The foundation was casted with these plates in position.

?

Post discussion with compressor OEM, it was analysed that once the machine was erected and levelled and the foundation bolts put in place, the grouting to be done with EP2 / epoxy grout. However with the plates in position there will be no adhesiveness and the foundation will not behave as a single unit but 2 distinct units i.e the civil foundation base separate and the epoxy grouted part separate and this will have implications for sustaining the operations of the compressor.

?

Hence, after confirmation from OEM, the plates were gas cut from the pockets and removed and the machine erected and master levelling done. Once the same was done, foundation bolts were put in the pocket and pocket grouting with epoxy done to made the complete foundation whole as a single entity withstanding the sustained operation of the make up gas compressor.

?

?

B2. FINE MESH IN START UP STRAINER – PISTON & RIDER RING FAILURE.

In one such net gas compressor, OEM mandated a startup conical strainer with 40 mesh for initial trial and commissioning of the compressor. However on start up frequent failure of the piston rings and rider rings were observed. As contamination from process side was suspected, the startup conical strainer was wrapped with a fine mesh – 200 mesh based on joint analysis with OEM and compressor restarted with Delta P across the mesh being monitored on DCS. After few rounds of cleaning the strainer with fine mesh based on high Delta P indication in DCS, the wearing and failure of the piston and rider rings drastically reduced. Once the operation was streamlined with the same, later on 200 mesh was removed from the strainer.

?

B3. MOTOR PURGE CONNECTION RELATED ISSUES:

Motor purge connections are provided on large motors to ensure that rotor and stator air gap is purged for removing contaminant specially when motors are located in hydrocarbon area. In the UK, during the period 1984 to 1991, six incidents were recorded, including five explosions, three offshore and two onshore. These incidents involved motors operating at 6.6kV and above and were confined to type Ex N motors. In the incidents involving explosions, flammable gas had entered the motor enclosure either by an external gas release being drawn into the enclosure following motor cool down, or via other routes such as common lube oil systems.

Several incidents after 2003 again highlighted the problem. This incident resulted from a substantial leakage of gas centrifugal compressor seal oil, allowing gas to migrate from the compressor through the seal oil system into the common compressor and motor lube oil system, and then into the motor enclosure. On starting the motor an explosion occurred within its enclosure. This motor was not fitted with the pressurisation or pre-start purging system required as special conditions of its certification (indicated by an X on the certificate).

Ignition of any flammable atmosphere within the machine has the potential to cause an internal explosion which may damage the machine. This ignition also has the much more serious potential to ignite any flammable atmosphere surrounding the machine.

Several sources of ignition need to be considered for cage motors:-

1. high internal or external surface temperatures

2. sparking across the air gap

3. sparking at the surface of the stator

4. sparking at the surface of the rotor

5. static generated on any parts not equipotentially bonded (e.g. the integral cooling fan)

6. sparking where circulating current is interrupted

7. Mechanical rubbing.

?

GAS INGRESS AND ITS CONTROL:

Ex N, Ex n and Ex e motors which have cooling provided by an external heat exchanger have no means of dispersing any gas which does enter the motor enclosure, as they circulate a largely trapped volume of air. Therefore, any flammable mixture that occurs within the motor enclosure is likely to remain there for a considerable time.

Another possible source of gas ingress occurs where the motor drives a load which handles flammable gases in compressors. During start up of a compressor when venting is carried out, there is a potential path for gas to enter the motor enclosure is through mistakenly connected vent headers of compressor and motor purge vents.

Although Ex N, Ex n and Ex e motors should not cause sparking the recent standards acknowledge the potential rotor and stator winding sparking on motors >1kV and require the user to consider additional control measures e.g. Ex p protection.

Hence it is necessary to ensure that the motor should not get contaminated with gas. Gas detection within the machine enclosure shall be provided.

Pre-start purging

Pre-start purge (one of the special measures that is applied to HV Ex e, N or n motors) to purge clean air (Instrument air or nitrogen) through the motor enclosure to remove any residual potential flammable gas from within. Air flow sensors and timers may monitor this process before allowing the starting of the motor. Once the motor is started no further air flow monitoring is provided.

Pre-start purging may not be sufficient to prevent ignition risks owing to stator sparking which might also be present during running conditions. This has only recently been recognized and users should assess their ignition risk control measures taking this into account.

?

C. SUSTAINED OPERATIONS PHASE:

?

C1. ISSUES RELATED WITH VOLUME BOTTLE DRAIN PIPING:

Normally all drains from compressor cylinder and volume bottles are routed through oily water sewer system (OWS). Typical configuration is shown in given sketch

This arrangement has one critical issue related with high pressure gas passing to cylinder distance pieces or coming to outside from point near OWS system. Normally all volume bottle have pulsation/vibration when compressors are in operation. Due to pulsation/vibration there are chances of loosening of drain valve of volume bottle over a period of time. In case this valve gets opened up and blind shown in above sketch is in position, leaking gas from valve will find way towards distance pieces as normally all drain valve of distance pieces are kept in open condition when compressor is in operation. In case above referred blind is not in position, gases will come out side to atmosphere from end point of drain header. If it remains unnoticed, it can result in unsafe condition

In order to prevent passing of high pressure gas to distance pieces, it is essential to provide additional blind after isolation valve of volume bottle. The volume bottle can be drained after taking out the blind downstream of this valve.

For hydrocarbon/hydrogen service detector shall be installed inside compressor house at truss location so that early warning signs can be noticed. These detectors have detected many leak incidents in hydrogen compressor house and timely action averted major incident.

C2. PRESSURISATION OF CYLINDER COMPARTMENT AND GAS TRAVELLING TO INSTRUMENT TERMINAL BOXES :

?

API-618, Para 3.6.6.8: specifies that flexible metallic conduit shall have a liquid tight thermosetting or thermoplastic outer jacket and approved fittings. For Division 1 locations, an NFPA –approved connector shall be provided.

It is observed that in event of rod packing leaks or improper venting system there is possibility of cylinder distance piece gets pressurized and connected instrument (Rod drop monitor, Temperature elements) in that compartment & may become possible source of gas reaching up to terminal box outside compressor through conduit pipe. This issue needs to be tackled at design and fabrication stage by using proper grommets/sealing devices to seal escaping gas and oil through capillary action

?

?

C3. ISSUES WITH LUBRICATOR:

In the net gas compressor a NP machine with dual stage, double acting, 6 cylinder machine with 7.2 Mw drive, there is a lubricator mechanism to lubricate the liner of all the 6 cylinders. The lubricator motor was an independent drive and the cam shaft mechanism pushed the lubricating oil through the tubing from the lubricator via a NRV into the respective cylinders.

?

In operation, bubbling was seen in the lubricator oil tank and the H2 detectors detected hydrogen in operation. Upon shutdown and analysis, an offspec product was observed through the compressor eventhough the compressor was designed for the same, the sticky material had deposited in the ball of the NRV making into to close improperly. This causes flow of H2 from the cylinder back to the lubricator.

?

Subsequent to the same, apart from ensuring that offspec product was not routed to the compressors, two additional modifications were made in the lubricator system.

1.?????? Single NRV changed to double NRV design.

2.?????? And a piping U loop liquid oil seal was made between the 2 NRVs to prevent ingress of the sticky offspec product into the 2nd NRV in case first was compromised.

?

?

C4. ISSUES WITH SMALL BORE CONNECTIONS:

?

API-618 Standard requires pulsation and vibration analysis of reciprocating compressors systems to prevent the system from damages. This analysis is carried out on major piping but often small bore connections are missed out in analysis and mistakes in site fabrication add up to this problem. These small bore connections on reciprocating compressor causes serious problems due to vibrations around reciprocating compressor.

?

issues of small-bore connection on Refinery CCR Platformer Net Gas Compressors were observed where series of failures occurred on small bore connection and associated tubing. Vibration/pulsation was the major contributing factor to various failures. Contributory Reasons for failure were analysed and reasons was found due to inadequate design of piping & tubing supports, Transmittal of vibration to platforms. It was found that small bore connections for PT, volume bottle drains, TI/TW were vibrating with high vibration velocity and amplitude. Max Vibration readings were found 705 micron. The dominant peak of vibration frequency is located at 15/17 X RPM. All vibrations were low frequency vibrations (< 300 Hz).

?

?The problem was overcome by reducing SBC length and provision of better supports, reduced span for tube and pipe supports, replacement of tubes with higher schedule carbon steel pipes, shifting of instruments to less vibration areas such as basement of compressor house etc. Various modifications explained for reduction of vibration levels. Study compares condition prior to modification and new modification proposals. Modification improved vibration levels and no further failure reported.

?

Study indicates that more attention is warranted for small-bore attachments both in design stage and during operation. It is recommended to remove SBCs wherever possible, redesigning them so they are less susceptible to vibrations and relocating them to areas of less base motion.

?

Small-bore connections are used on reciprocating compressor system for pressure and temperature measurement in each stage. These connections cannot be eliminated as these are required for critical process condition measurements. Small bore connection on reciprocating compressor causes serious problems due to vibrations around reciprocating compressor. These vibrations are generated due to pulsations, cylinder stretch, cantilever mass and poor supporting standards. Hence more attention is warranted for small-bore attachments both in design stage and during sustained operations as well.

?

A SAMPLE INCIDENT:

Platformer Net Gas Compressor - C, 1/2" SS tube for PT-037 on 2nd stage suction sheared off at ferrule joint, hydrogen at 14-kg/cm2 pressure started releasing from failed location as shown in Figures below. This was immediately noticed through hydrogen detectors in compressor house and compressor stopped and system was isolated to prevent further release.

COMPONENT FAILURE ANALYSIS OF TUBE:

The tube was found sheared at the location where it comes out from fitting. Tube fitting ferrule was found holding the broken tube piece. Thickness of tube was measured and no loss of thickness observed. There was no material defect and during DP test no abnormality noticed. No corrosion was observed in the failed portion of tube.

?

HISTORY

History of compressor indicated that during project stage, small bore piping were field routed as per site condition and no isometrics were generated as piping size above 2” were only considered for making isometrics. Since no drawings were available, it is left to site engineers and fitters to complete small bore piping work. Vibrations were observed on entire piping network of net gas compressor during commissioning. For reduction in vibration several new structural members were added to net gas compressor up to fin fan coolers and in other area. Similarly the bolted structure of NGC compressor house platforms were found loose due to piping transmitted vibrations. Later on all nuts were tack welded to prevent loosening. With this modification, there was reduction in piping vibration.

However high vibrations were observed on suction lines having tapping for PG/PT/TI , PDT, Valve unloader air and vent tubing were also found vibrating and several failures reported from various SBC and tubing. Typical failures were on pressure gauge adopters, valves, tubing. Piping vibrations were found on higher side when all three compressors were in operations.

?

?

SOURCES OF VIBRATION IN RECIPROCATING COMPRESSORS:

Vibrations related problems are generally observed in attached components on reciprocating compressor frame/ piping. Following are typical locations where vibrations are observed –

1. Valve actuator

2. Tubing

3. Inspection opening and instrument connections (thermocouple, pressure transmitters,

4. Flow measurement instruments,

5. Suction separator level control instrumentation,

6. Small branch connections for instrument connections and drain and vents

7. Instrument panel mounted on compressor frames

?

Various statistics indicates that most of the failures of SBC on reciprocating compressor occur due to Fatigue. This is mainly caused by excessive vibrations due to following dynamic excitation forces:

1. Pulsation-induced forces.

2. Stretching of cylinders

3. Mechanical unbalance and misalignment forces and moments from compressors

?

Pulsation induced vibration:

Pulsation that results in high shaking forces cause excessive vibration when the piping system is unable to transmit or absorb the shaking forces without damage. Excessive vibrations can occur even in cases where the dynamic forces are low. This occurs when the excitation frequency is close to or coincides with a mechanical natural frequency. When this happens, the vibration can be magnified by 5-50 times the off-resonant condition, which is mainly determined by the damping.

Pulsation itself will not produce vibration of the piping system; points of acoustical-mechanical coupling are necessary to develop a dynamic force, which in turn produces the vibration.

Pulsation induced vibrations are forced, repetitive, and occurs over a relatively long period of time. Excessive pulsation induced vibration can cause a fatigue failure in the pipe due to a large number of high stress cycles. This failure normally occurs at stress concentration point (e.g., branch connection, threaded connection, fillet weld, elbow, etc.). Geometric discontinuities such as elbows, reducers, tees and capped ends are common force-coupling points in compressor piping systems. Elbows or bends have differential areas due to outside and inside radii differences, which produce forces in the plane of the elbow. Force-coupling points also occur at reducers, tees and other points in the piping having unbalanced areas. .

?

Cylinder stretch

Cylinder stretch is the motion in the direction of piston movement which is generated by the pressures inside the compressor cylinder clearance volume. During each stroke, gas is first compressed on the head end, and then on the crank end of the piston. These pressures result in alternating force acting on the compressor cylinder. This force causes the cylinder assembly to lengthen and shorten during each stroke.

Cylinder stretch is a significant source of dynamic excitation for the attached piping and vessels on cylinder. The vibration amplitude generated by cylinder stretch can be minor, or severe, depending on the cylinder gas force and the MNF of the piping system. These forces are typically strongest at 1x, 2x running speed and are primarily a concern only in the immediate vicinity of the compressor.

?

Mechanical unbalance and misalignment forces:

Pulsation in cylinder nozzle can create significant unbalance forces in the vertical direction that can excite mechanically resonant component in vertical direction. The pulsation induced forces in the bottles and piping can be significant at high frequency as well. Unbalance forces due to pulsation resonances in the cylinder gas passages (between the head end and crank end) have been implicated on occasion as cause of high vibration and failures of appendages. The so called gas passage resonance force will often be in the range of 250 – 350 Hz on high speed compressor. A coincidence of gas passage resonance frequency and MNF of the cylinder in the stretch direction can result in high vibration.

Steady-state vibration can also cause failures at small diameter connections and tubing or cause flange leakage due to loosening of the studs. Steady-state vibration can be either low frequency (< 300 Hz) or high frequency (> to 300 Hz). Low frequency vibration will typically cause lateral displacement of the pipe, while high frequency vibration can cause flexural vibration of the pipe wall itself in addition to lateral pipe movement.

?

EXCITATION OF VIBRATION IN SMALL BORE CONNECTIONS:

Failure of small bore connections generally happens due to alternating tresses caused by vibration of these components at or near mechanical frequency of attachment resonance. The orientation of SBC connected to a system has a significant impact on how much vibration an connection will exhibit. Orientating an appendage in the direction of compressor cylinder motion is typically better for appendage close to compressor cylinder.

There are several sources of high frequency forces that can excite the mechanical natural frequency of small bore connections. Generally these MNF are above fourth order of run speed and often as high as 150-200HZ. Gas forces within cylinder contain harmonics of run speed and act in the direction of piston motion (stretch direction).Generally these forces decreases in amplitude as frequency increases. Any mechanical natural frequency of the cylinder assembly (typically between 150-250 hz) amplify the effect of these harmonic forces, causing more base motion of cylinder (stretch vibration)

Excitation is more severe when following condition exists -

1. Cantilever pipe with masses at the end (e.g., flanges, valves)

2. Long unsupported small-bore piping which vibrates due to pulsation and cylinder movement

?

?

COMPARISON WITH SAFE VIBRATION LIMITS:

?

a) Recommendation of API-618:

7.7.1.8: Connections DN 40 (11/2 NPS) and smaller shall be designed to minimize overhung weight. Connections shall be forged fittings or shall be braced back to the main pipe in at least two planes to avoid breakage due to pulsation-induced vibration.

?

b) Manufacturer’s recommendations:

Various other manufacturers’ standards are conservative on vibration velocity limits and they indicate following:

1. Vibration velocity of 7mm/sec is considered as alarm limit.

2. Vibration velocity beyond 7mm/sec to 21 mm/sec considered as danger zone and immediate corrective actions are required.

?

c) Recommendation of ASME-76 –PET-18

ASME 76-PET-18 (Ref figure below) has also set limit for displacement as per following conclusion:

1. 0.2 mm vibration amplitude for 300 Hz frequency level

2. 2.5 mm vibration amplitude for 2 HZ frequency level

?

On comparing, it was evident that vibrations levels on SBC related with Pressure transmitter (PTs) and drain points were beyond danger limit and need to be attended on urgent basis.

?

CHECKING FOR SOURCES OF HIGH VIBRATION:

?

1. Checking for higher pulsation in compression system:

One of reasons of high vibration in reciprocating compressor piping system is inadequate suppression of pulsation in compressor piping system. API -618 also recommends for checking and limiting these pulsations. Compressor supplier has responsibility for limiting this by carrying out pulsation study and an interactive acoustical simulation study should be carried out when the compressor unit is to be installed in parallel with other compressor units. This is normally achieved by:

?1. Proper design of volume bottles, varying suction piping diameter

?2. Installation of orifices in piping system

?3. Adding new support or modifying existing supports

?4. Installation of vibration/pulsating suppression devices.

?

?This was checked and found in order.

?

2. Inadequate design of piping and tubing supports:

Compressor piping does not have independent piping supports taken from the floor. The rigidity of supports for vent piping, tubing was found inadequate due to following reasons –

1. Long cantilever supports were used which have low stiffness (fabricated from angle).

2. In adequate size of Angle were used for fabrication of supports

3. Supports were taken from members which were vibrating in different plane.

4. Supporting span of piping and tubing was found very high

The supporting structure for the PG/PT tubes/ pipes on suction line was having more overhung length. As seen clearly from Figure above that pipe supports have a cantilever arrangement and poor stiffness of support due to weak structure, which contributed for excessive vibration. Supporting arrangement for small bore connection and main piping were inadequate. This was felt by vibration transmitted to structures and cracked fireproofing material around main supports.

?

3. Transmittal of vibration to platforms:

Net gas compressor piping supports were taken from platforms; these were transmitting piping vibrations to platforms. Since pressure gauges and PT frames are installed on platforms, vibration of platform is causing vibration of instrument and connected tubing.

As seen in Figure above, at some locations cylinder supports were found touching platform gratings resulting into transmittal of vibration to platform and associated supports of SBC.

?

4. Two vibrating at different frequencies and in different plane are connected together:

Compressor cylinder vibrations are in horizontal direction. However the vent piping is vibrating in vertical plane due to non- rigid supporting arrangement. Tubing support taken from vibrating piping was resulting into additional stresses on the tube. Due to vibration frequency and non-stiff supports the vibration amplitudes are getting amplified at mid span.

?

5. Vibration dampeners not functioning/not installed:

Restraints that are added to reduce vibration must not increase the pipe thermal expansion stresses or end-point reaction loads to unacceptable levels. It may sometimes be necessary to use hydraulic snubbers to stop vibration rather than fixed restraints. Such snubbers permit pipe thermal movement while still dampening vibration.

It is essential that all vibrating member need to be supported properly and for cushioning/dampening good quality supports were not used. Tubes were supported individually with GI U clamps without any dampening inserts. As number of tubes passing in a parallel layout from different unloaders to vent piping header. Each tube is vibrating in a plane and providing support at vibrating plane was resulting in reduction of vibration level. Pairing tubes in group was resulting in overall reduction in vibration. In order to reduce vibrations vibration isolators need to be installed in strict accordance with the design and the manufacturer's instructions.

Particular attention is required to isolator spring adjustment, anchor bolt adjustment, and correct installation of rubber pads to minimize the transmission of vibration and noise into the building structures. Equipment mounted on vibration isolators need to be free to oscillate and not be restrained by piping, conduit, etc

?

RESOLUTION OF VIBRATION PROBLEMS:

A pipe will not vibrate if it is prevented from moving. However, this concept does not necessarily help due to movement of cylinder and associated piping. Preventing piping from movement that could make situation worse. Therefore, for addressing this vibration problem, flexibility design of the piping system was considered.

Following 3R philosophy was considered for resolution of vibration problem-

1. Removing SBC connections which are not needed

2. Redesigning SBC so that they have less cantilevered and unsupported mass

3. Relocating SBC to locations of less base motion

?

1. REMOVAL OF SBC:

As most of the problems were due to matching of vibration frequency with MNF. SBC mounted on compressor frame, volume bottles were also important consideration due to base motion. One of the first measure which was considered was removal of SBC from compressor volume bottles, piping which are not accessible and can pose threat in case of any leak. This was very important especially on volume bottles which always have slight movement due to pulsation from compressor motion and compression of gases. Cantilevered valve, instruments at these locations have high chances of failure.

In case where instruments cannot be removed, it was decided to relocate them to rigid area (near foundation base).For this tubes with longer run was supported by using stiffening/ insulation rings for taking support. Isolation of SBC from the base motion was another option that was considered as an extra measure for reducing base motion of compressor. This was done by remote mounting the SBC.

?

2. REDISIGNING SBC:

2.1. Reduction in SBC length and provision of better supports:

Improvement in Design and construction of SBC was considered by changing orientation, reducing cantilever mass, improving supporting spans so that MNF of SBC could be reduced and excitation avoided. The structure natural frequency of a vibrating piping can be changed by attaching additional supports of sufficient rigidity. Deflection stress factor in SBC connection can be simplified as:

Ya = L2/D

Where

L = span length

D= Outer diameter of SBC piping/tube

Ya=deflection stress factor

?

The above equation indicates that the deflection stress factor will reduce with lower support spans and higher diameter of piping. This was considered in relocation supports for reduction in span. Moving support closer together will shift the natural frequencies and increase the stiffness and may significantly reduce the vibration for those systems with low frequencies of less than 300 Hz.

It was also considered to orient SBC close to a compressor cylinder and frame. For this stiffness of support increased by using higher cross section of steel members and replacing long angle supports due to their poor stiffness.

Number of elbows in small bore piping & tubing system minimized to reduce the number of force coupling points. Clamps were located near elbows on these SBC to prevent dynamic forces at these locations. For SBC fittings overall unsupported length was kept as short as possible. Mass of unsupported valves/instrument minimized (PT037, PT036) by replacing cantilever gate valves with similar size ball valves of similar pressure rating. Mass at the free end of cantilever was supported in both directions perpendicular to axis of small bore.

Smaller dial pressure gauges (4”) provided on reciprocating compressors. Possibility of replacing PT with pressure transmitter having digital indicators was also explored. This helped in eliminating the extra PG tapping provided on volume bottle.

?

?

2.2 Eliminating cantilever arrangement

In order to eliminate cantilever arrangement, ball valves of suitable pressure ratings were used in place of normal gate valve. Piping, tubing lay out corrected to eliminate cantilever. Wherever cantilever arrangement could not be eliminated support at these location were provided from compressor frame/cylinder.

?

2.3 Reduction of supporting span of tube and pipe supports:

Integrity of all small bore/ tube supporting arrangements on net gas compressor piping was checked. Tube supports provided as per standard specification (figure below) to eliminate sagging and vibration.

A reduced tube supporting span of 700 mm was considered for the tubes to further increase the stiffness (inversely proportional to cube root of span length between supports). In addition to this, the tubes were laid on a tray and a support for this tray was taken from the compressor body. Existing tapped hole Compressor body and frame were utilised for making new supports.

?

2.4 Provision of better quality supports for holding SBC/ tubes:

The support stiffness must be sufficient to restrain the piping and it should be at least three times the stiffness of the intermediate pipe section. Compressor tubes were supported on structure/frame by galvanised U bolts. These U bolts were found corroding over a period of time and in case of vibrating tubes it was acting as pointed support. Many tubes were found cracking at U clamp location. In order to improve this and to increase clamping area, bolted polypropylene clamp support as per details given in Figure below were provided. These clamps were also found useful for absorbing shock and vibration.

3. RELOCATING SBC WHICH ARE VIBRATING WITH HIGH AMPLITUDE:

?

3.1 Shifting of instrument to less vibration area:

1st stage suction pressure transmitter PT036/001 & discharge pressure transmitter PT003/005 stand shifted from the main platform. This stand relocated to ground floor. This arrangement reduced tubing length, vibrations and for compressor maintenance better access will be available. 2nd stage suction pressure transmitter PT037/006 & discharge pressure transmitter PT009/010 stand shifted from the main platform. This stand also relocated to ground floor. 1st & 2nd stage discharges pressure gauges PG004/008 relocated to ground floor (Ref figure below.) Tapping for these PGs was taken by providing a Tee after the hard piping of PT005 for 1st stage and PT010 for 2nd stage.

3.2 Control/Eliminate vibration:

It was necessary to inspect complete platform carefully & all the contacts between compressor / compressor components and the platform to be removed. This helped in two ways -

? Rubbing between the platform & the compressor components/piping/tubing can be avoided.

? Eliminating contact between compressor frame and platforms will have direct effect on reduction of platform vibrations. With reduction in platform, vibration all SBC for which support are taken from platform will also experience favourable condition for reduced vibration.

?

OTHER CONSIDERATIONS:

Adequacy of net gas compressor piping supports was checked as various cracks were observed on concrete pedestals compressor common suction / discharge piping. Spring restraints were found loose in some location due to absence of lock nuts. These were corrected for entire piping network.

?

Hence, It is seen that although reciprocating compressor main piping and system may exhibit normal behaviour but due to excitation mechanism, small bore connection have tendency to vibrate. This excitation may result from cylinder stretch, inadequate stiffness of supports, long supporting spans and stiffness of piping and tube itself. Many failures are reported from SBC connections on reciprocating compressors and most of the failures have occurred due to vibration. More attention is warranted for small bore attachments not only in design stage but also during operation. As a result of this field work in line with 3R philosophy, vibration levels reduced considerably. Hence 3R approach on resolution of reciprocating compressor SBC vibration problem is very useful.

?

1. Removal of SBCs wherever possible specially on compressor volume bottles, frames which have tendency to vibrate due to cylinder stretch, pulsation generated by compressor.

?

2. Redesigning them so they are less susceptible to vibrations by reducing support span, increasing stiffness of supports, better tube supports, elimination of cantilever weight supporting from compressor frame itself to avoid relative motion.

?

3. Relocating them to areas of less base motion near base of foundation, rigid structure

Regular LLF (look, listen & feel) rounds by operation and maintenance people always helps & can indicate early warning signs.

References & Acknowledgements:

1. API 618.

2. Literature on valve unloader by M/S Dresser-Rand

3. Literature on Reciprocating compressor by M/S GE

4. Literature on reciprocating compressor valves by M/S Hoerbigher

5. High voltage electric motors for use in hazardous atmospheres, Offshore Information Sheet No. 3/2010, (Issued April 2010 by HSE, UK

6. Brian C. Howes, Chris B. Harper, Vibration Related Failures of Small-Bore Attachments- Beta Machinery Analysis Ltd., 2003, Calgary, AB, Canada, T3C 0J7

7. Everett Houdyshell, Bill Eckert, Wally Bratek. Recommended Approach to Control Vibration from Cylinder Gas Forces, Proceedings of GMC2009: Gas Machinery Conference, October 5-72009 Atlanta, Georgia

8. Brian C. Howes, Vibrations in reciprocating Compressors, Beta Machinery Analysis Ltd., Calgary, Alberta, Canada

9. N. Sackne, B. Fofonoff, Pulsations & vibration control for small reciprocating compressors, Beta Machinery Analysis Ltd. Calgary AB, Canada,

10. M. HAMBLIN, Fatigue of cantilevered pipe fittings subjected to vibration, Woodside Energy, GPO Box D188, Perth, Western Australia 6840 Australia.

11. By J C Wachel, JD Tison, Vibration in reciprocating machine and piping systems, Engineering Dynamics Inc, published at 23rd Machinery symposium 1994.

12. SAFETY ISSUES WITH RECIPROCATING COMPRESSOR DESIGN, OPERATING PRACTICES & MAINTENANCE, K.C. Upreti, Amar Dev, T V Prasad, B Pundarikaksha.

13. ?PREVENTING SMALL BORE PIPING/ TUBING FAILURES IN RECIPROCATING COMPRESSORS, K C Upreti, Dwaipayan Banerjee.

?