May 2009

Special Report: Maintenance/Reliability

Risk-based inspection, panacea for plant failures?

Understand the limitations for an effective implementation

Pullarcot, S. K., International Inspections Centre W.L.L.

Risk-based inspection (RBI) acquired momentum in the late 90s and is being implemented in almost all the industries across the world. RBI is a powerful scientific management tool in optimizing inspection manpower and equipment resources. However, its unscrupulous implementation without identifying the real limitations of the systems will only yield surprising results after a period of time, which will give the impression that RBI is not an effective tool. Therefore, RBI should be implemented with utmost dedication and sincerity, and by duly identifying and acknowledging the inherent deficiencies of the system which shall be given due consideration and remedied accordingly while implementing.

I would like to highlight the various deficiencies associated with RBI. Consider the case of a pressure vessel in hydrocarbon service that operates at a certain temperature and pressure. What is the risk it poses to the surroundings? How can it be quantified? As such, the guidelines provided by the American Petroleum Institute (API) are general and not specific to any industry, leaving out the specifics that are unique to the said equipment for specific service. Moreover, the assessed risk of any operating vessel could be different when assessed by two different individuals. This becomes worse when these two individuals assess the risk with two different methods for arriving at a quantified risk figure. But what is reality? Under certain service conditions, the risk associated with an operating pressure vessel is the same irrespective of who evaluated the risk and what methodology was used. This eventually leads to two different risk figures that may be widely different, if based on different methodologies used for assessment as well as the level of competency of the evaluator. In such instances, the basis on which the RBI is built is bound to collapse at a later date. Though RBI is based on mathematical models that are sound and logical enough, the reliability and dependability of the same is affected by the basic input data. In all probability, the input data will be deficient because of the reasons stated which will question the very credibility of RBI itself, when failures occur contrary to predictions.

In addition, during the equipment service life, deterioration of varying magnitudes takes place based on the severity of service such as the pressure and temperature conditions, cyclic loading, corrosion rate of fluids contained, etc. The guidelines for RBI, API 580 and 581, list various damage mechanisms operating in the oil industry. However, those mechanisms may not be the only ones contributing to deterioration. For example, a stainless steel vessel situated near a seashore experiences external deterioration from a chloride atmosphere prevailing in that area in addition to the usual deterioration mechanisms acting inside the vessel or piping. So a more judicious thinking and application of logic by the evaluator of the failure mechanisms are required. Therefore, any RBI methodology that relies only on API guidelines is not going to provide a realistic picture of the actual deterioration taking place in the vessel. The resulting error that can creep into the system further affects the credibility of RBI which may only be realized 10 or 15 years after introducing RBI.

An overview of the typical RBI implementation strategy adopted by the oil and gas industry is shown in Fig. 1.

 Fig. 1   

Typical RBI implementation strategy.

The principal areas of concern in RBI are shown in Fig. 2.

 Fig. 2   

Principal areas of concern in RBI.

To improve the reliability and dependability of RBI as an effective and rewarding program, serious consideration shall be given to aspects identified as "A" and "B" in Areas 1 and 2 respectively.

Therefore, every industry implementing RBI has to pay attention to the areas of concerns and the following approach is proposed to alleviate the issues to a great extent. As desired in RBI guidelines, this shall be applied to all static equipment and piping to make it comprehensive for the entire plant.

As mentioned earlier, the impediment to a reliable RBI is the subjectivity of the individual assessing the risk (initial/in-process) associated with equipment and piping. The only way to reduce the subjectivity is by increasing the database and to arrive at the initial/in-process risk figure of the equipment and piping and so also the consequences of failure.

The first step to achieve this is to list the various risk elements associated with the equipment/piping under study. Upon listing all the risk elements, severity of each element should be rated in a numerical scale. This numeric figure shall be based on a qualitative quantification made by the evaluator based on guidelines to be developed for the purpose depending on the unique process peculiarities.

As the number of parameters identified increases, the reliability, and thereby the dependability of the figures improves, resulting in more reliable risk values. Since the whole RBI scheme is built on these figures, the reliability of the primary risk values plays a vital role in the reliability of the RBI system itself, which can be termed as a semiquantitative methodology.

Therefore, asset owners should carry out a risk evaluation associated with each vessel in a detailed manner that shall be worked out based on the working parameters, design philosophy and construction of their plants.

For example, the risk associated with a vessel is different when a safety valve is provided compared to another vessel operating at the same operating parameters without it. Similarly, when a trip system from a DCS is enabled, its risk value reduces considerably.

Such a study requires the involvement of a learned engineer who is conversant with the design/operating considerations and parameters. The group has to develop a questionnaire with multiple-choice answers based on these considerations. The question and answers shall be designed in such a fashion that any engineer or operator with a bit of experience (two or more years) shall be able to feed in the required data with reasonable accuracy. The documents they need to answer this questionnaire are the data sheet and drawing pertaining to the equipment and also the material specification and isometric drawings for the piping. However, for answering a few questions, they may need the help of the operations group which can be concluded quickly, provided they consult with the right operating staff.

Since the subject proposed involves many disciplines from engineering and management, the expertise of all these groups is required. However, if the questions are broken down into the simplest possible level, answering them can be made very simple. Moreover, because of the extended database questionnaire generated against each vessel/piping loop, the subjectivity of the individual is reduced considerably in the assessed component risk values, thereby resulting in improved reliability of the assessed risk.

Increasing the database of attributes to risk alone would not improve reliability of the risk figures for equipment and piping. Therefore, the obvious second step shall be to arrive at a realistic weighting factor for these attributes that also have significant bearing on the risk figures arrived at for each vessel and piping loop. This is not easy since it involves concerted efforts of process, mechanical and instrumentation engineers from disciplines like operation and technical services, maintenance, inspection and instrumentation. This has to be accomplished in the study through a sample survey to be carried out by qualified and experienced engineers/technicians from these disciplines.

When steps 1 and 2 are completed, one can arrive at realistic risk figures (initial risk) for all process equipment and piping systems.

As the third step, the risk figures shall be revalidated periodically based on inspection findings from routine and periodic inspection activities. For example, the actual corrosion rate observed during periodic inspection may be less or more than that predicted initially. Therefore, this rate has to be revalidated based on inspection findings.

Lastly, the inspection methodologies proposed for each equipment and piping loop shall be critically evaluated to ensure that they are capable of revealing the deteriorations predicted. Past experience with similar equipment/piping and an awareness of the predominant damage mechanism in such systems would be an added advantage in this regard.

If these RBI aspects are taken care of in a detailed manner, it would definitely improve plant reliability. However, it shall be noted that every requirement has its implications with regard to money and time. Since the cost involved in implementation is only the additional manpower required, the benefit by way of increased plant reliability is expected to be much in excess of the cost. Therefore, this proposal is a worthwhile exercise by which RBI reliability can be substantially increased. I am in the process of developing such a system for a surface production facility in oil and gas. The system thus developed shall be more or less applicable to almost all surface production facilities across the world with minor modifications to customize it to the specific environment of individual producers.  HP


 The author

Sunil Kumar Pullarcot works as a specialized inspection consultant with the inspection and corrosion team of Kuwait Oil Company (KOC). Prior to this, he worked with FACT Engineering and Design Organization (FEDO), India, a premier consultancy organization in South India in various capacities. Mr. Pullarcot has more than 27 years' experience in the manufacture and QA/QC of pressure vessels, heat exchangers, storage tanks, plant and offsite piping and construction activities of fertilizer, chemical, petrochemical and oil/gas projects. He is a Fellow of the Institution of Engineers, and a member of the Nondestructive Testing Society of India and the Indian Institute of Welding. Mr. Pullarcot received a BSc degree in 1981 in mechanical engineering from the University of Kerala, India, and an MTech degree in production engineering in 1990 from the Cochin University of Science and Technology, Kerala, India. He is the author of the book Practical Guide to Pressure Vessel Manufacturing, published by M/s Marcel Dekker Inc., New York, in January 2002 under ISBN 0-8247-0740-0. His second book, Practical Guide to Construction, Inspection and Testing of Above Ground Storage Tanks, is in an advanced stage of publishing. He is a well-known trainer on QA/QC, welding and NDT, and is recognized as a global instructor by the American Society of Mechanical Engineers (ASME).


The Author

Related Articles

From the Archive



{{ error }}
{{ comment.comment.Name }} • {{ comment.timeAgo }}
{{ comment.comment.Text }}