Risk-based inspection (RBI) programs applied in an Aristech Chemical Co. polypropylene plant (Fig. 1) and an Altura Energy Ltd. gas-processing plant used plant-specific evidence and historical and theoretical probabilities to develop well-thought-out inspection plans.
The plans are expected to save at least $1 million/plant in inspection costs for future maintenance turnarounds.
The American Society of Mechanical Engineers (ASME), the American Petroleum Institute (API), and others have recognized the RBI approach as a means of determining inspection scopes and of developing a cost-effective inspection and maintenance plan. API 581, a draft1 released in 1995, provides a guide for RBI application.
As RBI methods grow in popularity, it is important that results converge on similar answers regarding risk ranking and guidance for inspection programs. Aptech validated its Risk Directed Mechanical Integrity (RDMIP) system by selecting components from Altura's project that had already been evaluated with API's RBI software. Results from the two applications were similar.
RBI
By risk ranking equipment, an RBI program optimizes its maintenance resources (time and money). The optimization of limited resources is based on risk results in safer, more reliable plant operations.
Applying the appropriate engineering expertise to identify potential damage mechanisms also improves the overall safety of a facility.
Aptech's Technology Application Group (TAG) is providing direction to its RDMIP program, which has been used in chemical companies and refineries.2-4 TAG meets biennially to share experiences and identify areas for future work. Altura and Aristech are both members of TAG.
Strategic, tactical phases
A typical RBI/RDMIP program has a strategic phase and a tactical phase. The strategic phase includes a hazard screen, risk ranking, and development of equipment plans. Implementation of equipment plans occur in the tactical phase.
* Strategic phase. The initial hazard screen identifies all equipment required to be covered under US Occupational Safety & Health Administration regulation, OSHA 1910.119, and the US Environmental Protection Agency's (EPA) risk management program rule. These rules address the quantity of hazardous chemicals in equipment that justify screening.
Some plants elect to include equipment with steam, caustics, or condensates that, while not covered under OSHA 1910.119, may be important from a personnel safety standpoint. In addition, some plants include equipment on the bases of plant reliability and availability.
After gathering equipment data, an operations and technical team performs a risk analysis. The team considers scenarios of what can go wrong, the likelihood of failure (LOF), and the consequences of failure (COF).
The product of LOF and COF provides a measure of risk. The API program tabulates the results of each equipment item on a 5 x 5 matrix (the RDMIP program tabulates the results on a 4 x 4 matrix) to create a risk ranking. The 25 API or 16 Aptech risk ranks contain four risk categories: high, medium-high, medium, and low for purposes of inspection planning.
* Tactical phase. Upon completion of the risk ranking, the team develops equipment plans. These plans consist of inspection schedules, scope, and techniques.
The risk analysis identifies equipment for which the relative risk is judged to be high. That is, in relationship to other equipment at the facility, the relative consequences of the component failing and the likelihood of it failing provide sufficient concern to warrant some immediate measures.
Immediate measures may include gathering additional evidence concerning the integrity of the equipment or additional information on process or metallurgical conditions.
A critical equipment list consists of all high-risk items with a high LOF.
When sufficient knowledge exists on high-risk ranked items to demonstrate suitability for service, the facility can change from a traditional inspection schedule to one based on the risk analysis. A program based on risk analysis is managed via the plant's maintenance management system and inspection result database (Fig. 2).
The data in the management system are continuously improved. Re-evaluation of the risk ranking takes place as a result of additional evidence through the inspection or corrosion-control program or of new knowledge associated with mechanical integrity of the equipment.
API RP 579, recommended practice for fitness for service analysis, is a guide for assessing equipment problems for which suitability for service is questionable.
Aristech Neal plant
Aristech's Neal plant licenses Montell's Spheripol process for manufacturing polypropylene. The facility conducted an initial hazard screen to determine equipment to be covered under the OSHA 1910.119 ruling.
Process and instrument diagrams (P&IDs) identify fixed and rotating equipment as being either covered or excluded. Covered equipment includes those that contain highly hazardous chemicals, are the last or next to last line of defense, or are perceived to be of concern by employees.
Excluded equipment includes those that do not contain highly hazardous chemicals. Failure in excluded equipment would not result in failure of a covered component. Such failure would impact only such items as product quality, capacity, or run time.
Covered and excluded equipment have other identifying characteristics not listed here.
Hazard review
Through the application of the hazard review, 46% of the fixed equipment and 77% of the rotating equipment in the polymerization and material recovery area were identified as excluded. The propylene unloading, storage, and distillation area had fewer exclusions.
As a whole, 32% of the fixed equipment and 49% of rotating equipment in the Neal plant could be considered excluded from the formal mechanical integrity program.
By excluding items not critical to process safety, Aristech was able to focus its resources on inspection, testing, and documentation of the equipment most critical to safety. The hazard review process typically excludes 80-90% of plant instrumentation and controls from the mechanical integrity program. All safe shutdown devices must be covered, however.
The hazard screen will provide about $3 million in cost savings to the Neal plant. This number consists of the cost for restructuring documentation and 20-year life-cycle costs.
Mechanical integrity analysis
Aristech implemented Aptech's RDMIP at the Neal plant. An overview of the methodology has been previously published (OGJ, Aug. 2, 1999, p. 47). The scope of the risk analysis included the covered equipment discussed above with the exception of instruments and controls.
Risk ranking for pumps and compressors was done by considering equipment reliability. The analysis includes failure experience, quality of overhauls, and quality of the predictive or preventive maintenance programs.
For fixed equipment, equipment plans provide recommendations for risk reduction through the mechanical integrity program. The risk analysis team made recommendations for improvements to the maintenance program and design or material changes where warranted.
The risk analysis team analyzed reactors, heat exchangers, storage tanks, pumps, and compressors. The RDMIP process results in a risk ranking of 1 to 16, with 1 being the highest risk items and 16 the lowest. Fig. 3 summarizes the results of the risk analysis of the fixed and rotating equipment.
The plant is constructed mostly of carbon steel (with low-temperature toughness as appropriate) and some stainless steel. Primary damage mechanisms included pitting, fatigue, uniform corrosion, corrosion under insulation, and atmospheric corrosion. The primary damage mechanisms affecting the dynamic equipment included seal or packing failure, erosion, and corrosion.
As a result of the mild process conditions, the LOFs were generally not high. The nature of the plant's business and chemicals, however, resulted in considerable COF for some vessels.
RDMIP cost benefits
The hazard screen and RDMIP analysis optimized the mechanical integrity program of the Aristech Neal Plant. The most readily quantifiable cost benefit of the program is a reduction in the scope and frequency of inspections of items that do not need to be inspected.
Unquantifiable cost savings currently include reducing the duration of turnarounds and minimizing new discoveries during turnarounds.
In some cases, the new program increases costs as a result of more-sophisticated inspections of potential problem areas. The cost of conducting routine internal inspection varies as a function of equipment complexity; it may be as low as $1,000 or as high as $10,000 or more.
According to some industry averages, the cost of fixed-equipment inspections is about $600/inspection. The average industry cost of preparing for the fixed-equipment inspections (blinding, scaffolding) is about $4,000/inspection.
Through the use of the RDMIP program, the Neal plant eliminated more than half of the originally planned thickness-monitoring locations on fixed equipment. Pending verification of no active damage mechanisms, some vessels that were risk-ranked three and four will go beyond the 10-year maximum interval proposed by API 510.
For a design life of 40 years, the elimination of two low-risk ranked vessel inspections equates to a conservative savings of more than $1 million.
The most important cost savings will result, however, from the avoidance of potentially catastrophic failures by applying additional engineering analysis to the inspection program.
Altura Slaughter plant
Altura used API software to perform a Level III (quantitative) analysis of the fixed equipment and the associated piping for the plant.
Fig. 4 shows an overview of the steps to calculate equipment risk. The approach begins with the collection of process and equipment information. Various scenarios show how leaks may occur and how they can progress into undesirable events.
The output from the LOF and COF analysis (described next) is combined in a linear (non-weighted) matrix, which assumes equality between like-ranked elements of LOF and COF.
Results
Generally, 10-20% of vessels in a process facility are ranked as high risk. The RBI program at Altura's Slaughter plant indicated that 28% of the vessels and piping circuits carried the majority of the risk.
This number is higher than the industry average because of the age and condition of the plant. The sour-gas inlet-compression unit and the amine and caustic treating units accounted for most of the plant's risk.
About 48% of Altura's fixed equipment items were identified as medium risk, and about 24% were low risk. Altura used these results to define scope and frequency of inspection for each vessel and associated piping consistent with API 510 and 570.
The plant also used Aptech's software to develop inspection plans for each piece of equipment, which detailed operating conditions, materials, possible damage mechanisms, appropriate nondestructive examination techniques, risk ranking, and the scope and frequency of inspection.
These plans also document the reasons for certain decisions, which form an important part of a defensible inspection program.
The study recommended use of advanced external inspection methods for certain internal damage mechanisms. These methods include ultrasonic mapping, time of flight diffraction, and pulsed eddy-current inspection techniques. Improving the level of confidence in inspection results, prioritizing equipment, and conducting comprehensive on-line inspections allows increased operating times between turnarounds.
The company has estimated that plant turnaround times will be reduced by about 67%.
Before conducting the study, it was estimated that the internal inspection of fixed-equipment items for its 2001 turnaround would be $2-3.5 million. Based on results from the RBI study and application of on-stream condition assessments, the present scope is estimated to be at least $1.5 million less than initially planned.
Some items will depart from their present 10-year inspection cycle; others will be inspected much more frequently, such as every 2 years. Altura believes that the overall result of this work is the creation of a safer workplace at a lower cost.
Validation
The Altura project afforded the opportunity to compare results from API software, version 2.4.1 and Aptech's RDMIP software, version 1.2.4. The RDMIP software imported Altura's data from the API database.
Fig. 5 compares the COF and LOF analysis results.
API's analysis starts complex and uses simplifying assumptions later in the evaluation to reduce the conclusions for purposes of inspection scope and frequency planning. Aptech's software makes simplifying assumptions in the initial analysis and proceeds to more complex analysis as needed.
For most equipment items, the results are consistent. That is, most equipment falls into the same relative risk-rank categories. There are a few items that fall into a different risk-rank category for each software program. For the most part, the minor deviations occur as a result of the various ways each software derives and treats the inputs.
As it is important that different methodologies converge to provide similar results on the same equipment, this validation experiment is a first step in verifying procedures.
Deferred maintenance risk
How might a company use risk concepts to decide which items to work on during a turnaround? And, how should it assess the risk of deferred maintenance so that management may make the best decisions to support corporate goals?
First, the company should establish a technical (no dollar costs yet) risk ranking of the items on the turnaround punch list. For each item on the punch list, the company should determine a risk ranking (LOF x COF) if the item were not done in the upcoming turnaround. The highest risk ranking should be at the top of the list.
Next, for each item, the company should estimate the project cost of the planned maintenance activity.
Calculation of the cumulative cost of items ranked 1 through n, where n is the last item on the list that can be accommodated within the proposed budget, provides the ranked list of activities that can be completed at the next turnaround.
Stakeholders should review this list before it is final.
At this point, the company has only dealt with relative risk to define the hierarchy of work items. To develop a measure of deferred maintenance risk, the LOF and COF analyses must include detailed cost impacts. The evaluation of the risk of deferred maintenance to turnaround planning can be done in the following manner:
- Develop a fault-tree scenario of what can go wrong for each of the major components for which maintenance is planned to be deferred. Identify the likelihood of the events that make up the scenario using the available evidence and document the basis for the evidence. This will provide a best estimate for the probability of an event occurring.
- Develop the outgoing tree for recovery from the identified events and assign dollar values to these. This will give us a measure of the consequences.
- Calculate risk and express it in dollars as a value or cost of deferred maintenance.
- Scrutinize the above trees to determine if mitigating measures are available to minimize the risk. If so adjust the numbers. F
Acknowledgment
The authors thank the contribution to this article by Geoff Egan, Bob Heller, Hung Duc Tran, and many others.
References
- API 581, Base Resource Document on Risk Based Inspection, draft, 1995.
- Merrick, Edwin A., Leonard, C. Ron, Eckhardt, Phil, and Baughman, Harry, "Risked-based methods optimize maintenance work scope," OGJ, Aug. 2, 1999, p. 47.
- Merrick, E., Young, J., Parra, A., and Egan, G., "Risk Based Mechanical Integrity: Its Impact on Environmental, Health, and Safety, and Process Safety Management," Aptech Engineering Services Inc. TP108, 1/98 edition, Presented at Petro-Safe Conference, Houston, winter 1998.
- Anderson, S., Merrick, E., "Risk Directed Mechanical Integrity," Presented at International Chemical and Petroleum Industry Inspection Technology, Post Topical Seminar, Houston, June 1999.
The authors-
Edwin A. Merrick is director of the petroleum and chemical business units for Aptech Engineering Services Inc. He has worked at Aptech for more than 10 years. Formerly, Merrick spent 16 years with the Tennessee Valley Authority working on mechanical integrity issues for the fossil and nuclear power industry. Merrick holds a bachelors degree in mechanical engineering from Vanderbilt University, Nashville, and a masters degree in materials science from the University of Tennessee, Knoxville.
Stephen A. Anderson is a senior materials scientist at Aptech Engineering Services Inc. For the past 5 years, he has specialized in metallurgical and corrosion science in refinery and chemical plant equipment and the development of inspection and maintenance programs. He has more than 15 years of industrial experience. Anderson holds a BS in both physics and chemistry and an MS in materials science, all from the University of Cape Town, South Africa.
Since 1991, Robert M. Newberry has worked in the process safety management area at Altura Energy. He currently works in all areas of regulatory compliance. Newberry has 29 years of experience in the oil and gas industry. He has 3 years of experience in oil production, 8 years of experience as a gas processing plant operator, and 9 years of experience as a gas-processing plant-operations supervisor.
Thomas Blunk is a maintenance engineer at Aristech Chemical Co. Formerly, he was a design engineer for Bechtel Savannah River Inc. and Lockheed Martin Energy Systems. He has also worked at AK Steel as a maintenance engineer. Blunk holds a BS in mechanical engineering from North Carolina A&T State University, Greensboro.
