BALANCED FORCES AND FOUNDATIONS KEYS TO RECIPROCATING-COMPRESSOR LIFE

Sept. 3, 1990
Phil Applegate Warren Petroleum Co. Division of Chevron U.S.A. Monument, N.M. Reducing the effects of unbalanced forces and vibrations on reciprocating compressor foundations will reduce crankshaft and engine maintenance problems. There are guidelines and strategies for balancing, conducting vibration studies, and designing and maintaining foundations. There is a very large amount of reciprocating horsepower installed throughout the industry.
Phil Applegate
Warren Petroleum Co.
Division of Chevron U.S.A.
Monument, N.M.

Reducing the effects of unbalanced forces and vibrations on reciprocating compressor foundations will reduce crankshaft and engine maintenance problems. There are guidelines and strategies for balancing, conducting vibration studies, and designing and maintaining foundations.

HIGH-SPEED UNITS

There is a very large amount of reciprocating horsepower installed throughout the industry.

A large percentage of the reciprocating horsepower is big, slow speed (300 rpm) integral engines/compressors. The slow speeds were built primarily in the 1930's through the 1960's.

There are still quite a few being built today. Since that time, the industry has moved towards higher rpm machines that are separate from the reciprocating compressors.

These high-speed units (900 rpm) are usually skid mounted and bolted onto a concrete pad. These skidmounted units can be moved quickly and inexpensively from one location to another as gas availability dictates.

The slow-speed units are mounted on concrete pedestals usually connected to a common concrete mat on which multiple engine pedestals and engines are mounted.

The industry will continue to struggle with finding new and better ways to maintain this reciprocating horsepower. Innovations within the last 1 0 years have included better electronic engine-shutdown panels, better solid-state ignition systems, and improved lubrication systems.

One of the most expensive maintenance problems derives from frame distortion. This results in poor alignment, causes premature main journal bearing failure, and eventually breaks crankshafts. Visit any crankshaft repair shop to witness the industry's problems with broken and bent crankshafts.

Another problem has been excessive piping vibration. This can result in catastrophic results if the piping happens to break in the wrong place.

These are just some of the more costly repairs that are being seen. By using a little common sense, one can see that there are probably other mechanical repairs that are caused from excessive vibration and poor foundations.

There are two ways to reduce these costly problems: Reduce the unbalanced or shaking forces or revise the foundations to transmit the forces into the earth while still maintaining proper alignment.

UNBALANCED FORCES

Most people in the industry are also aware of how critical balancing is to centrifugal equipment such as turbines, compressors, and pumps. The higher the speed the more critical the balance. Reciprocating equipment is no different.

Unbalanced forces have for several years been a "black box" to many in the compressor-maintenance industry. High speed, separable, horizontally opposed compressors have always been closely watched for unbalanced forces.

The horizontally opposed pistons, crossheads, etc. must not vary by more than 2 lb from each other. Maintenance personnel have learned that if the weights differ by more than 2 lb, the compressor sometimes shakes excessively.

Any reciprocating machinery that is located offshore must also be balanced as close to perfection as possible. Otherwise, the platform cannot handle the excessive vibration. In fact, original-equipment manufacturers (OEMS) have certain reciprocating engines that usually go offshore because of unbalanced forces and moments inherent to some engines but not to others resulting from their crankshaft geometries.

The term "unbalanced force" simply means that there are inertial forces within a compressor created by the piston weights that are unbalanced. The "unbalanced force" term is used sometimes generally to describe specific forces and moments; or in other words a catch-all phrase for the entire topic.

All reciprocating compressors and engines have unbalanced forces and moments. Slow speed, high-speed integral, and separable, all have certain inherent forces depending upon their design.

There are components within the reciprocating compressor and engines which operate in a rotating motion.

An example would be counter weights.

Roughly two thirds of the weight of the connecting rods operate in the rotating direction. The remaining one third operates in the reciprocating direction.

Counter weights also have a center of gravity. There are two ways to find the center of gravity. The first method is to make a paper template of each counter weight. Then find its center of gravity by dividing the counter weight area into various form-fitting shapes such as squares, rectangles, triangles, parabolas, and circles.

The center of gravity is found by summing the centroids for each area.

The other method is by balancing the counter weight on a fulcrum and marking the spot where the counter weight balances. The angle of the counter weight's center of gravity in relation to No. 1 piston top dead center (TDC) must also be determined no matter which method is chosen.

For a single-throw machine, the primary and secondary inertial forces are unbalanced in the horizontal direction unless a counter weight is used to oppose the weight.

When the crankshaft rotates 90, the counter weight is in the vertical direction and then is unopposed by the piston weights. This means that as the crankshaft is rotated 360, the machine will tend to jump up and down vertically.

The foundation and anchor bolts are the only opposition to the unbalanced force in the vertical direction. This is usually preferred, however, because the anchor bolts and foundations are better able to handle this movement.

In the horizontal direction, only the piping and the friction between the chocks and the anchor bolts restrict the movement. Therefore, excessive horizontal movement causes frame distortion, broken anchor bolts, fretted chocks, and cracked piping.

To help restrict excessive horizontal movement, the preload or torque of the anchor bolts must be adequate. The coefficient of friction between the frame and chocks and the chocks and pedestal must also be kept as high as possible.

Therefore, any contaminant such as oil and grease should be strictly avoided. This means some prevention measures should be taken to keep the oil away from the chocks. For engines installed on a full-bed grout, this is probably not very feasible.

TRADE OFFS

There are, however, certain trade offs in decisions of which force or moment to reduce. Vertical forces and horizontal and vertical moments can be slightly increased to reduce the primary horizontal force. This is because the foundation can usually better withstand vertical forces, and the couples are usually smaller in magnitude than the forces when spread out over the length of the frame. Every attempt should be made to reduce the primary horizontal force whenever possible.

Secondary forces are caused by the compressor connecting-rod motion. Secondary forces and couples cannot be balanced with counter weights attached to the crankshaft because they do not operate at the same speed.

Secondaries can only be balanced by making all pistons weigh the same or by using geared counter weights that operate at twice the running speed.

For integral engines that have multistage pistons, it is impractical or uneconomical to have all the pistons weigh the same. For two-stroke piston scavenged engines, the problems are more complex. For instance, an RA and BA Clark has two scavenging pistons which can weigh less than one half of the larger compressor pistons.

Other factors include the number of throws, crankshaft geometry, distance between throws, different crosshead weights, and the size and shape of counter weights that can be used.

Through normal maintenance procedures, the industry has indiscriminantly changed the compressor pistons without checking the weights of replacement parts on slow-speed integral engines. It was thought that the slower-speed units with good foundations did not need to be balanced.

Other mistakes have included the mixing of heavy and light crossheads, counter weights in the wrong position, counter weights missing, and dummy weights for crossheads removed.

Machine repair shops throughout the industry have provided identically sized pistons that might differ in casting weight by as much as 100-150 lb.

The inertial forces covered so far are created by the weight of the pistons as they accelerate and decelerate. The inertial forces are the same when the compressor is idling unloaded as when the compressor is compressing gas; the weight of the pistons, crossheads, etc. is still the same.

Therefore the gas load does not affect the unbalanced forces or moments. Vibration studies confirm this fact.

It follows that the rod load does not affect unbalanced forces and moments for the compressor frame when treated as one rigid body. Inertial forces can significantly affect the "combined rod load," however.

"Combined rod load" is the term used to describe the combined effect of gas and inertial loading on the frame, crosshead, rod, etc. One of the side benefits of weighing the compressor components is that the "combined rod load" can now be calculated more accurately.

Quite frequently several compressors are identically configured with the same size pistons, crossheads, etc. If this is the case, then some thought should be given to weighing all the pistons and components for each unit and averaging the weights for the same size pistons. Then as each unit is balanced, all the same size pistons will weigh the same.

If a piston needs to be replaced, there will be only one spare needed. This will simplify the required inventory of spare parts. Otherwise, when a replacement is needed, there will likely not be a spare that weighs within tolerance.

A personal-computer program to analyze the unbalanced forces and moments can be used to help the equipment operator optimize the inertia balance.' The OEMs as well as consulting engineers and applied research groups will also furnish detailed inertial balance studies.

Either way, it is very important to optimize the balance to help eliminate some of the pounding and shaking that the foundation must absorb and transfer. The industry for the most part has ignored the inertial balance on the assumption that a good foundation and soil conditions compensate for the excessive equipment movement.

If a balance study has not been performed in several years, however, chances are high that things have changed, and a study should be performed. A lot of money is being spent every year to "regrout" slow-speed units, and it is worth the time and expense to protect the investment with an inertial balance study.

VIBRATION STUDY

Figs. 1 and 2 represent a vibration study that was performed on a slow-speed integral engine.

The figures are renderings of the peak-to-peak displacement or movement measured in thousandths of an inch (mils). The readings were taken at one-time engine speed with a special accelerometer which outputs in velocity and remains accurate down to 1 hz.

Fig. la represents data both in the horizontal and vertical directions taken before the engine was shut down to rebalance. Fig. 1b represents data taken after rebalance. Nothing else was done to the unit except to rebalance it.

Fig. 2 represents a slowspeed integral engine configured identically as the unit in Fig. 1. Fig. 2a represents the vibration data taken before the unit had an almost complete foundation overhaul. The unit was supported with full-bed cementitious grout about 12 years old.

Fig. 2b represents the data taken after installation of a new concrete pedestal anchored with post-tensioned rebar along with a new epoxy grout cap with extra large epoxy chocks. The unit was also balanced similarly to the unit in Fig. 1 b.

Fig. lb shows the benefit of rebalancing an existing engine to reduce the vibrations and extend the foundation life. Fig. 2b illustrates the further benefits of repairing the foundation in addition to rebalancing.

The vibration studies conducted so far have proved to be very enlightening and allowed the user to quantify some movement observations. As more experience is gained, the vibration studies will prove to be very useful.

FOUNDATIONS

Structural engineers tell us the purpose of the foundation is to transmit the forces into the soil below while still maintaining proper alignment. The following will focus on five major areas, which will achieve this goal: foundation design, soil quality, concrete quality, anchor bolts, and epoxy grouting technology.

FOUNDATION DESIGN

The primary purpose of the foundation is to absorb the pounding and shaking of the engine and compressors and transfer these forces into the soil below. This is not an easy job if the concrete in the foundation is weak and oil soaked.

The concrete must perform its job properly or it will become the weak link in the chain. Weak concrete and poor soil conditions will cause increased costs from excessive equipment movement.

The engine typically rests on a concrete pedestal which varies in height from 3 to 1 0 ft for most engines. The pedestal is joined to the wide thin concrete mat with steel rebar. The concrete mat is usually between 12 and 48 in. thick and usually at least as wide as the building, depending on the design. The concrete mat rests on soil which may be rock, sand, or something in between.

The old rules-of-thumb have been replaced with newer theories precisely to predict the required foundation design.

Design criteria using the elastic-half-space theory are currently being used by structural engineers to determine the adequacy of a dynamically loaded block-type foundation.

A personal-computer program is available to determine various parameters of the pedestal and mat .2 Input includes unbalanced forces, block shape, mat size, and soil properties.

The properties calculated include natural and resonance frequencies, displacement, velocity, and acceleration vibrations. The results are compared with design criteria to determine if the block will vibrate at tolerable levels; if not, dimensions are redesigned to meet proper design criteria.

In order for the pounding and shaking forces to be transferred down the block and into the soil, the total concrete foundation must act or move together. The concrete pedestal should not be separate from the concrete mat.

This is not the case for most of the blocks, however, because typically the mat was poured first with only stubs of rebar to stick up into the pedestal. The concrete pedestal was then poured to the desired height.

As many know, old concrete does not bond to new concrete or one pour to the adjacent pour. Therefore, there is virtually no bond between the pedestal and the mat except for what rebar ties the block and pedestal together.

Basically, these separate pours create a horizontal crack which covers the entire pedestal. This crack is often referred to as the cold joint.

Watson states that the concrete pedestal is then supported by a series of rebar which act as compressed springs .3

His spring theory states that when the engine is running, the pounding and shaking are reflected into the engine itself, thereby increasing maintenance and failure of the engine, compressor, and piping. In time, the cold joint or horizontal crack will begin to wear.

If problems are suspected, core samples will help determine the quality of the concrete, condition of the cold joint, and the soil properties beneath the mat.

The solution for the cold joint problem of a new installation is to pour the new concrete mat and pedestal in one pour.

Watson recommends that, for an existing mat, the solution is to use post-tensioned bolts. This is accomplished by vertical holes being drilled through the pedestal and down into the mat.

The bolts are then post-tensioned or tightened to tie the pedestal to the mat. The void between the drill hole is then filled with epoxy grout.

These post-tensioned bolts will be the solution to most existing pedestals and mats unless it is determined that the mat should be replaced (Fig. 3).'

The usual practice in the past has been for several engines to be set on a common mat. Watson feels the problem with a common mat is that the excessive vibrations are transferred from one engine to another. When the conditions are just right, vibrations resonate throughout the mat and damage foundation pedestals up and down the length of the mat.

The solution is to pour the foundations with separate foundation mats. This keeps the vibrations from damaging other foundations.3 For continuous common mats that have already been installed, the cost of separating or removing the mats must be weighed against the problem each particular installation is experiencing with this type of problem.

SOIL QUALITY

Soil conditions under the concrete blocks and mats can create additional problems. Foundations resting on rock do not absorb the forces, while sand might move and create additional alignment problems.

The recommended solution is to use a properly compacted soil which will transfer these shaking and pounding forces into the soil below.3 This soil will be different in different parts of the country. It is common to see different types of soil within the same plant site.

If the proper type of soil is not available, then it will have to be found and brought to the job site.

CONCRETE QUALITY

The next step in proper concrete foundations is the quality of concrete. Any new concrete poured for compressor foundations should have minimum compressive strength and tensile strength established before the foundations are poured.

Each company should have its own guidelines for minimum concrete specifications. Over-designing the cement content or water/cement ratio, however, might cause excessive internal temperatures and damage the concrete during curing.

The aggregate should be crushed rock, instead of river rock, with at least 95% fractured faces.

This is a very important consideration. The crushed rock provides more rough surface for cement bonding. Thus, it will greatly enhance the tensile strength of the concrete. It also will increase the compressive strength of the concrete compared with concrete with smooth river rock.

The smooth river rock pops loose from the concrete too easily, which is why crushed rock is preferred. The crushed rock and sand should be washed to remove any residue. If there is any doubt as to meeting specifications, a batch test should be conducted by a reputable engineering firm.

It is important to know what kind of concrete will be poured before it is poured.

The water-cement content of the concrete is probably the most important factor affecting the strength of the concrete mixture. There should be only enough water to provide workability to the mix and to ensure the forms are filled without voids.

In order to keep the concrete as dry as possible but still flowable enough to be pumped and placed in the forms, a chemical is added to the concrete called "superplasticizer."

This chemical creates a temporary increase in slump or workability for about 60-90 min. The effect of the chemical then disappears and the concrete reverts to the consistency of the original mixture. This means a dry, high-strength concrete which can be poured without placement and workability problems.

There should be two areas of caution in use of a superplasticizer. When used on a very warm day, the superplasticizer will wear off more quickly, and extra dosages might be needed. Extra chemicals should therefore be on hand during the pour. Extra superplasticizer will not hurt the concrete and will temporarily increase the slump again.

Also, the superplasticizer does not seem to mix as well into a very dry mixture so that a minimum slump should be specified as well as the maximum allowable slump.

The block should have adequate rebar in the form to reinforce the concrete foundation.

Sharp corners within the concrete form should be strictly avoided. This will help eliminate the stress-concentration point for the concrete to start cracking during curing.

All corners should be radiused or chamfered at least 11/2 in. Inside corners require 2 in. If the oil-pan area of the block is formed in concrete after the form is removed, the oil-pan area should be painted with an epoxy paint or other material to protect the concrete from oil penetration.

Any cracks in this area will allow the oil to penetrate into the concrete and drastically weaken its strength.

ANCHOR BOLTS

Anchor bolts have two very simple purposes: holding the engine and compressor down and controlling and transferring unbalanced forces into the concrete block and foundation.

The bolt material is probably the most important factor. There are usually two types of materials used: high strength 4140 steel and mild cold-rolled steel. Each type has different characteristics.

The normalized 4140 steel is a high-strength steel with a yield strength of approximately 100,000 psi. Mild cold-rolled steel has a yield strength of 36,000 psi.

In the past few years,high-strength 4140 steel has been specified because the bolt can be loaded more than mild steel. This extra loading is used to carry out the purpose of the anchor bolt which is to hold the engine down.

The high-strength bolt allows the bolt to have more preload or torque applied to it. This extra torque can be as much as four to six times higher than mild steel.

The amount of stretch is directly proportional to the load or force being applied to stretch the bolt. The bolt is anchored in the concrete on one end and tightened with a nut at the other end. The length in between is known as the "free length" of the bolt.

This allows the bolt to stretch while reducing the chance of cracking due to the bolt being too short. This is one reason that care must be taken when existing anchor bolts of mild steel are being tightened.

In older bolting systems, it is likely there was very little free length in the bolt. This creates an additional risk of cracking when retorquing the anchor bolts.

When anchor bolts are installed into the concrete form, every effort is made to align the bolts as precisely as possible. However, some small variations frequently occur.

The increased flexibility of a bolt that is able to move a little helps solve this alignment problem.

The anchor bolt will have adequate free length because it was protected with a grout sleeve. This will allow the bolt to be moved slightly when the engine is lowered.

The bolt must be properly secured to the form so that as the concrete is poured the bolt will not move. The top and bottom of the bolt should be secured to the steel rebar in the form. The bolt should never be welded on to fasten the bolt to the rebar. This might ruin the integrity of the high-strength steel bolt.

The grout sleeve should be sealed with an oil-resistant compound to prevent oil penetration into the block.

In the past, horizontal cracks have developed in the concrete at the exact depth of the anchor bolt hook or Jbend. This was a result of all the anchor bolts being installed at the same depth. For this reason, the depth of engagement of the anchor bolts in new blocks should be varied in order to reduce the chance of these types of cracks.

Standard-cut threads are a series of stress concentration points that create cracks in the bolts, usually in the first thread below the nut. The chance of this type of crack can be reduced if cold-rolled threads are used instead of standard-cut threads.

Another effective idea for anchor bolts is the use of matching convex/concave washers and nuts. These center the bolt load and provide a better matched surface over flat washers and nuts. These should be constructed of the same materials as the bolt.

When an anchor bolt breaks and a replacement is needed, the bolt should be divided into two pieces and connected with a collar or coupling nut.

Below the nut and collar the bolt is embedded into the concrete. Above the collar or nut the grout sleeve protects the bolt and provides the free length needed. If the upper portion of the bolt breaks again, it is simply unscrewed and replaced with a new section.

The nut or collar should be constructed of the same material as the bolt to ensure the integrity of the anchor-bolt system.

"Preload" is the term used to describe the load or force put into a bolt as it stretches while being tightened. Cyclic fatigue on the anchor bolts is one of the major causes of anchor-bolt breakage.

Cyclic fatigue is the stretching and relaxing of steel over and over again. If the bolt is preloaded with enough force that the cyclic fatigue load does not exceed the preload, then theoretically the bolt will not fail. This preloading of each bolt has two purposes: to restrict the engine movement and to keep the cyclic fatigue problems to a minimum.

There are several methods of preloading the anchor bolts: stretch method, torque wrenches, strain gauges, crushable washers precalibrated for a set load, and turn-of-the-nut method. Whatever the method, the preload should be checked on a regular basis in addition to checking crankshaft web deflections.

The initial preload always relaxes due to a number of factors, including embedment relaxation, vibration loosening, and stress relaxation .5 Because of these factors, the bolts should be retorqued when new anchor bolts are installed to ensure that the proper preload is maintained.

These torque values should be recorded also for future reference and trending. Crankshaft web deflections should be checked in conjunction with this retorquing.

EPOXY GROUT

Fig. 4 shows a typical diagram of a concrete foundation, epoxy grout and chocks, and engine frame. The epoxy grout cap is poured on top of the concrete foundation. The epoxy chocks are used to support the engine after it has been aligned.

The primary functions of epoxy grout are to maintain the proper original alignment and to resist chemical and oil penetration into the concrete foundation. Epoxy chocks are designed to prevent thermal humping of engine frames and crankshafts by allowing air flow underneath the engine thereby reducing "thermal humping."

Epoxy grout is the material that bonds to the top of the foundation and provides a tough, stable, oil and chemical-resistant surface for the engine base.

The epoxy grout bonds tightly to clean, dry concrete. It will, however, not bond to oil-soaked concrete. Epoxy grouts are being used today over other cement grouts because of their oil resistance and strong physical properties. Although lubricating oil will ruin concrete, it usually cannot penetrate epoxy materials unless the epoxy grout is cracked.

Epoxy grouts are stronger in compressive strength and shear and tensile strength than cement grouts. Epoxy grouts are comprised of epoxy resin, epoxy hardener, and aggregate. The aggregate is used to increase the overall strength of the grout similar to the function of aggregate in concrete.

Epoxy chocks, similar to epoxy grouts, are specially formulated materials that contain an epoxy resin, hardener, and possibly a fine aggregate. The chocks must provide a tough, uniform surface that provides almost complete contact with the engine-bed plate. The chocks must resist thermal creeping.

The concrete foundation must be inspected before grouting to be sure it is good quality concrete, structurally sound, and properly cured with no visible cracks. If the block is in good condition, the entire top of the foundation must be chipped off to assure that a good clean surface is exposed.

Typically when concrete is poured into a block, a layer of cement and sand rises to the surface. This layer must be chipped down to expose the aggregate. Jack hammers should not be used on new concrete. A smaller chipping gun should be used for this task.

The top of the exposed concrete must be cleaned to remove any dust or other loose material. This can be done with a vacuum or clean oil-free air. All outside foundation edges must be rounded or chamfered. The block and equipment should be shaded from direct sunlight to prevent excessive shrinkage in the grout.

The temperatures of the concrete block, outside air, and grout materials are very important to a successful pour. The effect of temperature cannot be over-emphasized. Epoxy grout and chocks should not be poured when it is either too hot or too cold. The job schedule should be planned so that the ambient temperature surrounding the block meets the grout manufacturer's tolerances.

The forms should be chamfered on all corners and edges. Inside right angles or inside corners should be strictly prohibited because of the stress points corners produce. Expansion joints perform a very important function; they place the crack that might develop during curing where it does the least damage.

The expansion joints also serve to divide up the grout pour into smaller sections thus reducing the probability of a crack. Expansion joints are placed perpendicularly to the crankshaft.

The equipment configuration should determine the exact location of the expansion joint. Expansion joints should not be placed directly under chocks, but instead moved over. Proper curing time of the chocks must be allowed before bolts are torqued. This is very important. A hardness tester may be used to check the hardness of the chock when there is a possibility that the chock did not cure properly. This usually happens during cold temperatures.

When the anchor bolts are being torqued, measure the "pull down" with a dial indicator. Pull down is a measure of how much the chocks shrink or deflect when the preload is being applied. If the pull down is considered excessive, then the chocks will have to be repoured.

In the past, cracks have developed 0.25-0.50 in. below the concrete grout bond or interface. These cracks, known as "edge-lifting cracks," usually occur in recently poured concrete because it has not developed enough tensile strength to resist these lifting forces.

These edge-lifting cracks are caused by the grout shrinking during curing. This lifts the concrete up causing cracks on the outside edges of the foundation block.

These types of cracks can be prevented. The edge of the concrete-block foundation should be rounded off down to floor level (Fig. 5). This removes the weak shear plane which is where the cracks start.

ACKNOWLEDGMENT

Special thanks are due to Randy Millier of Warren Petroleum Co. for his assistance with the vibration CAD programming.

REFERENCES

  1. Enpro, Belle Chasse, La.

  2. Structural & Vibration Engineers, Houston.

  3. Watson, W.D., "Vibration analysis key to compressor-foundation maintenance," OGJ, Aug. 8, 1988, p. 40.

  4. Watson, W. D., Structural Foundation Seminar, Oct. 18, 1988.

  5. Bickford, J. H., "That Initial Preload - What Happens to It?" Mechanical Engineering, October 1983.

Copyright 1990 Oil & Gas Journal. All Rights Reserved.