EXAMINING COMPRESSOR MAINTENANCE

EVERY MAINTENANCE EVENT MUST BE VIEWED AS AN OPPORTUNITY TO UPGRADE SO AS TO AVOID REPEAT FAILURE

One somewhat dated source cites 13% of all turbocompressor failures as being due to errors or omissions in condition monitoring and maintenance.[ 1] With the advance in monitoring technology and modern operating and maintenance practices one would assume that this general number might not be as high today. What then are good monitoring and maintenance practices?

Compressor condition monitoring has the following components:

1. Proper response to supervisory instrumentation such as alarms and trips.

2. Periodic observation and evaluation of operating parameters such as compressor physical condition and performance efficiency. This would include measuring and judging the rate of mechanical and performance deterioration. Daily compressor operator rounds should be structured following the principles of Operator Driven Reliability.[2]

3. Evaluation of operating trends. This should include auxiliary systems, such as lubrication and seal oil or dry gas seal supply consoles and compressor on-line washing (liquid injection) facilities.

4. Periodic testing of lubrication and seal oils. Six basic analyses are required: Appearance test, testing for dissolved water, flash point test, viscosity test, the determination of the Total Acid Number (TAN), and the determination of the additive content.[3]

5. Periodic testing of emergency safety and shutdown devices (ESD) and other damage- limiting components, such as exercising the compressor’s surge control valve loop and the trip and throttle (T&T) valve on steam turbine driven compressor trains.

6. Data logging and automated record keeping such as the number of unplanned trips per train per year as a basic indication of compressor reliability.

7. Diagnosis of problems, appraising their severity and deciding what action to take.

8. Remedial action & execution planning.

9. Corrective measures should preferably be applied on-stream to reduce the impact on compressor availability. On-line flushing (washing) would be an example.

Maintenance strategies

Generally, turbocompressors have maintenance inspections, overhauls and repairs (MIO&R), often called “IRD”, meaning Inspection and Repair Downtime. These terms are used interchangeably with “turnarounds.” These are scheduled in periodic intervals ranging from two-to-ten years, depending on the type of service.

The extent of MIO&R efforts ranges from simple bearing inspections to opening the compressor and replacing the rotor with a spare. Used rotors are examined for rubs at labyrinth seal locations and for fissures and cracks around impeller eyes on radial compressors. On axial compressors, moving and stationary blades should receive thorough attention. In all cases non-destructive test (NDT) procedures are applied.

As the scheduled compressor turnaround approaches, it is best practice to review the machine’s operating and maintenance history. If there are any defects noted at inspection, it would be well to ask: Are any of these defects repeat occurrences? If so, can they be expected at this turnaround? What steps can be taken to eliminate them? What action should be taken at this time?

A thorough pre-turnaround review should be undertaken to plan the work required. It should consist of an assessment of compressor mechanical condition, a performance check and a diligent review of the machine’s past history.

Maintenance strategies can be either preventive or predictive. Preventive maintenance (PM) is time-based whereas predictive maintenance (PdM) has as its goal compressor operation until detectable defects start to develop. Detection requires high-quality predictive monitoring methods. Application of these methods comes at a cost. Qualified personnel have to be employed and ongoing monitoring efforts are precise. Management expects PdM to determine when a failure will occur and plan an outage accordingly.

It is fair to say that certain predictive routines can be used to minimize the impact of a premature failure, or to understand when a machine drifts into off-design operation. But cost savings always have to be factored in, and none of the various PdM judgments can be made without experience. Real depth of experience will be needed, for example, when several seemingly minor deviations occur and converge.

Preventive maintenance encompasses periodic inspection and the implementation of remedial steps to avoid unanticipated breakdowns, production stoppages, or detrimental machine, component, and control functions. Predictive, and to some extent also preventive maintenance, is the rapid detection and treatment of equipment abnormalities before they cause defects or losses.

This is evident when considering lube oil changes. This routine could be labeled preventive if time-based, and predictive if done only when testing shows an abnormality in the properties of the lubricant. Without strong emphasis and an implemented preventive maintenance program, plant effectiveness and reliable operations are diminished.

In many organizations, the maintenance function does not receive proper attention. Perhaps because it was performed as a mindless routine or has, on occasion, disturbed well-running equipment, the perception can sometimes be that maintenance does not add value to a product. This may lead management to conclude that the best maintenance is the least-cost maintenance.

Armed with this false perception, traditional process and industrial plants have sometimes underemphasized preventive, corrective and routine maintenance. Some have neglected to properly develop maintenance departments, elected not to pursue proper training of maintenance personnel and not to optimize predictive maintenance. Unforeseen compressor failures and safety hazards have resulted.

Correctly executed, maintenance is not an insurance policy or a security blanket. It is a requirement for success. Without effective preventive maintenance, equipment is certain to fail during operation. To be effective, however, maintenance must be selective. A good way to look at it is that selective PM results in damage avoidance whereas effective PdM allows existing or developing damage to be detected in time to plan an orderly shutdown.

Best practices

Four levels of compressor maintenance exist, with some overlap:

1. Reactive, or breakdown maintenance: This includes repair of equipment after failure (run-to-failure). It is unplanned, unsafe, undesirable and expensive.

2. Selective preventive maintenance. Selective preventive maintenance includes lubrication and proactive repair. On-stream lubrication of, say, the admission valve control linkage on certain steam turbines should be done on a regular schedule. In this instance, anything else is unacceptably risky and inappropriate.

3. Corrective maintenance. This includes adjusting or calibrating equipment. Corrective maintenance improves the quality or performance of equipment. The need for corrective maintenance results from preventive or predictive maintenance observations.

4. Predictive maintenance and proactive repair. Predictive maintenance predicts potential problems by sensing operation of equipment. This type of maintenance monitors operations, diagnoses undesirable trends, and pinpoints potential problems. In its simplest form, an operator hearing a change in sound made by the equipment predicts a potential problem. This leads to either corrective or routine maintenance. Proactive repair is an equipment repair based on a higher level of maintenance.

Predictive maintenance instrumentation is available for both positive displacement and dynamic compressors. It exists in many forms and can be used continuously or intermittently. It is available for every conceivable type of machine and instrumentation schemes range from basic, manual and elementary to totally automatic and extremely sophisticated. Recommended instrumentation depends on compressor size and owner’s sparing philosophies.

As an example, a facility may opt to install three 50% machines, two 100% machines or perhaps only one 100% machine in a given service. Moreover, unless the value of downtime avoidance is quantified it will not be possible to make firm recommendations as to the most advantageous level of monitoring instrumentation and shutdown strategies.

There are many competent manufacturers of manual monitoring equipment. Manual monitoring is often used on small air compressors. Advanced predictive maintenance on-stream systems are generally used with large process compressors to continuously monitor vibration behavior. By gathering vibration data and comparing these data with normal operating conditions, both manual and continuous systems can predict and pinpoint the cause of a potential problem. The trouble is that detecting vibration is different from eliminating vibration.

An intelligent but highly selective preventive maintenance program may lead to actions that prevent bearing distress and thus prevent vibration from occurring in the first place. A selective preventive maintenance program may be a more cost-effective program than any program or strategy that waits for defects to manifest. This fact establishes that sweeping management edicts that disallow all manner of preventive maintenance on compressors do not harmonize with the principles of asset preservation and best practices.

Traditionally, industry has focused on breakdown maintenance, and unfortunately, many plants still do. However, to minimize breakdown, maintenance programs should focus on levels 2 through 4 above.

Emergency repairs

Plant systems must be maintained at their maximum level of performance. To assist in achieving this goal, maintenance should include regular inspection, cleaning, adjustment, and repair. And any repairs events must be viewed as opportunities to upgrade.

In other words, the organization must know if upgrading of failed components and subsystems is feasible and cost justified. On the other hand, performing unnecessary maintenance and repair should be avoided. Breakdowns occur because of improper equipment operation or failure to perform basic preventive functions.

Overhauling equipment periodically when it is not required is a costly luxury; upgrading where the economics are favorable is necessary to profitability. Regardless of whether or not PdM routines have determined a deficiency, repairs performed on an emergency basis are three times more costly in labor and parts than those that are preplanned. But bad as the consequences are, much worse is the negative impact from frequent breakdowns on overall performance, including the subtle effect on worker morale, product quality, and unit costs.

Selective preventive maintenance, when used correctly, has shown to produce maintenance savings. Sweeping maintenance including the routine dismantling and reassembling of compressors is wasteful. It has been estimated that one out of every three dollars spent on broad-brush, time-based preventive maintenance is wasted.

A major overhaul facility reported that “60% of the hydraulic pumps sent in for rebuild had nothing wrong with them.” This is an example of the disadvantage of performing maintenance to a schedule as opposed to individual machine condition.

However, when developed and managed correctly, selective preventive maintenance is the most effective type of plan available. It results in improved plant availability, higher equipment reliability, better system performance (or reduced operating and maintenance costs) and heightened safety.

A plant staff’s immediate maintenance concern is to respond to equipment and system functional failures as quickly and safely as possible. Every maintenance event must be viewed as an opportunity to upgrade so as to avoid repeat failure. This sentence is the key to superior maintenance. Systematic upgrading will extend allowable intervals between shutdowns.

The starting point for a successful longterm selective maintenance program is to obtain feedback regarding the effectiveness of the existing maintenance program from personnel directly involved in maintenancerelated tasks. Such information can provide answers to several key questions, and the answers will differ from machine to machine and plant to plant. In-plant data and existing repair records will provide most of the answers to the questions given below. A competent and field-wise consulting engineer will provide the rest: [4]

1. What is effective and what is not?

2. Which time-directed (periodic) tasks and conditional overhauls are conducted too frequently to be economical?

3. Which selective preventive maintenance tasks are justified?

4. What monitoring and diagnostic (predictive maintenance) techniques are successfully used in the plant?

5. What is the root cause of equipment failure?

6. Which equipment can run to failure without significantly affecting plant safety and reliability?

7. Does any component require so much care and attention that it merits modification or redesign to improve its reliability?

It is just as important that changes not be considered in areas where existing procedures are working well, unless some compelling new information indicates otherwise.

It is best to focus on known problem areas. To ascertain focus, continuity of information, and proper maintenance activities, some facilities assign responsibility for welldelineated plant systems to a knowledgeable staff person.

All maintenance-related information, including design and operational activities relating to such a system, are funneled through this expert. He or she refines the maintenance procedures for those systems under their jurisdiction and re-shapes preventive maintenance into selective maintenance.

Maintenance improvement

Problems associated with machine uptime and quality output affect several functional areas. Many people, from plant manager to engineers and operators, make decisions and take actions that directly or indirectly affect machine performance. Production, engineering, purchasing, and maintenance personnel as well as outside vendors and stores use their own internal systems, processes, policies, procedures, and practices to manage their sections of the business. These systems interact, are interdependent, and constrain one another in various ways. Some constraints are appropriate while others can have disastrous consequences.

That said — program objectives need to be clearly defined. An effective maintenance program should ensure that: unplanned maintenance downtime does not occur; equipment condition is known at all times, where justified; preventive maintenance is performed regularly and efficiently; selective preventive maintenance needs are anticipated, delineated and planned; and the maintenance department performs specialized maintenance tasks of the highest quality.

Moreover, all craftsmen are skilled and participate actively in decision-making process; proper tooling and information are readily available and being used; replacement parts requirements are fully anticipated and components are in stock; and maintenance and production personnel work as partners to maintain equipment. Following these general guidelines for centrifugal compressors will give positive results.

One deviation alone might not be enough to bring on a compressor failure, but when several more deviations combine, the failure risk increases exponentially. While it might be possible to avoid more serious failures by implementing an automatic compressor unloading scheme, or adding bells and whistles that annunciate excessive temperatures and vibrations, seal deficiencies and the like, there will never be any substitute for the human mind supplying both logical root cause failure analysis and up-front failure prevention processes.

Achieving these up-front processes before calamities and fingerpointing occur requires both training and accepting accountability. The operator, supervisor or manager accepting a deviation from established practice should somehow be motivated to understand the potential ramifications of bypassing or not following established guidance and should document this understanding in writing.

As the documentation requirements are enforced, fewer deviations would be tolerated. Accept the initial incremental cost outlay needed to do things right. The apparent expenditure of time and money will ultimately bring rewards in safety, reliability, and increased profitability.

References 1. Bloch, Heinz P., and J. J. Hoefner; “Reciprocating Compressors: Operation and Maintenance”, (1998) Gulf Publishing Company; (ISBN 0-88415-525-0). 2. Bloch, Heinz P.; Operator-Driven Reliability,” Maintenance Technology, April 2008 3. Bloch, Heinz P. and Fred Geitner; “Machinery Failure Analysis & Troubleshooting,” Gulf Publishing Co., Houston, TXC (ISBN 0- 88415-662-1) 4. Pipeline and Gas Technology, December 2005

Authors

Heinz P. Bloch, P.E., is an independent Consulting Engineer with decades of machinery reliability improvement experience with Exxon affiliates in the U.S. and overseas.

Fred K. Geitner, P.Eng. is an independent Consulting Engineer with several decades of relevant experience with Cooper-Bessemer and also Essochem affiliates in Canada and overseas. They are the co-authors of “Compressors: How to Achieve High Reliability and Availability”, (2012), McGraw-Hill, New York, NY, ISBN 978-0-07-177287-7. They can be contacted at