Development of Maintenance Management Strategy Based on Reliability Centered Maintenance for Marginal Oilfield Production Facilities ()
1. Introduction
Oil and gas exploration and development venture is highly capital intensive and involves a lot of uncertainties in terms of project development costs and revenue. A major deal breaker for exploration decisions is the reserves’ size. In recent times, there has been increasing interest in oilfields that are initially considered unattractive for development due to technical, economic, or strategic reasons. These fields are referred to as Marginal Oilfields [1] . [2] highlighted the economic factors that classify an oilfield as marginal. These factors include high capital expenditure (CAPEX) and operating expenditure (OPEX) costs, unattractive revenue dependent on recovery factor, low production rates, technological constraints, Government regulations and policies, etc. However, advancements in petroleum engineering technologies such as 3-D seismic and opportunities for low CAPEX—phased development, among others, have favored interest in marginal oilfield development. While the low CAPEX requirement promotes the start-up of marginal field operations, many operators struggle to keep-up with operations due to the high OPEX, especially arising from the maintenance of the facilities.
Maintenance is critical to the healthy operations of any plant or facility. In the process industry where production operations run continuously round the clock, it is very essential to ensure that maintenance is properly planned to achieve a high level of equipment availability, because accidental stoppages result in substantial financial losses [3] . More so in the oil and gas industry, downtime resulting from improperly planned maintenance is shown to have a significant negative impact on the OPEX [4] . Downtime in the oil and gas industry is estimated to range between 5% to 10%, which is higher than other industries’ average of 3% to 5%. This is because 90% of oil and gas companies are said to practice time based preventive maintenance, while about 5% to 20% adopt reactive maintenance [5] . According to [6] 40% of net operating expenses in the oil and gas industry is accountable to unplanned (reactive) and scheduled (time-based) maintenance, while unplanned plant shutdown accounts for nearly half of the overall losses of an oil facility. The impact of downtime is even more detrimental to the operations of a marginal oilfield, due to the compact size of field’s production capacity. Therefore, it is incumbent to develop an efficient maintenance management strategy to ensure that the oilfield equipment are reliable, available and optimally operated. In this regard, this paper presents a Reliability Centered Maintenance (RCM) framework to support the maintenance management of a typical marginal oilfield production facility in Nigeria for a reduced OPEX and enhanced profit.
The concept of RCM was initially presented in theory as far back as 1969 by Nowlan and Heap [6] , with the notion that failure distribution is not related to age and the frequency of performing maintenance. Thus, RCM is viewed as a technique that provides a bespoke approach to maintenance, bearing that facility equipment does not have the same level of importance to the operation and safety of the facility, therefore, such facility maintenance should not be generalized. Thus, RCM involves a systematic analysis of the functions and failures of a system to determine the appropriate maintenance to implement for such a system [7] . The outcome is a mix of specific-based maintenance techniques, which identifies equipment or components that should be run-to-fail, i.e., corrective maintenance, those that require time-based preventive or scheduled maintenance, and more substantially, promotes the practice of condition-based (CBM) maintenance and predictive maintenance (PdM) [8] .
RCM considers the functions of a system in normal or desired operating conditions and ways in which the system can fail to meet its desired or normal operating condition, i.e., functional failure. Thereafter, the causes of the functional failures (the failure mode) are identified, together with the immediate effects and consequences of the failure. Opportunities to predict the failure are then explored, if not predictable, default actions are considered to prevent the failure [6] . Thus, in achieving a successful RCM, the following tools are essential: Failure Mode Effect (FMEA), Criticality Analysis, Fault Tree Analysis (FTA), Event Tree Analysis (ETA), Logical Tree Analysis (LTA), and other risk-based decision-making tools [9] .
[10] highlighted the attribute of RCM as an integrated approach that capitalizes on the collective strengths of several maintenance techniques applied optimally together, rather than independently, thereby maximizing facility and equipment reliability while simultaneously minimizing life-cycle cost. The author presented a methodology for RCM using a process steam plant as a case study; a significant reduction in OPEX, including spare parts and labor costs, were estimated, as well as a reduction of downtime by 80%. The limitation from this study is that the procedure only considered a single unit. Similarly, [11] presented a study on RCM procedures considering radical maintenance using an ethylene plant as a case study. The findings from the study facilitated a more efficient resource utilization and an improved maintenance program for the facility. While these techniques are adoptable for this current study, the key limitation is that the case study was not in the oilfield industry. [12] presented a comprehensive review of maintenance practices in the oil and gas industry, particularly in marginal oilfields. The major gap identified was the lack of study on the application of RCM to marginal oilfield maintenance. The study recommended that implementing RCM will potentially provide an efficient maintenance strategy for the marginal oilfield production facilities by reducing downtime and maintenance related OPEX [12] .
This study hereby explores the techniques of RCM in developing an efficient maintenance strategy for marginal oilfield production facilities. A brief overview of a typical marginal oilfield production facility is presented in section 2, followed by the methodology used in developing the RCM-based maintenance strategy. The results were presented and discussed in section 4, and lastly, a conclusion section with key findings from the study and recommendations for further works.
2. Overview of the Production Facility in the Marginal Field Case Study
This study used a typical 10,000-barrels/day Early Production Facility (EPF) within the Niger Delta region of Nigeria as case study. The selection of the production facility was based on the data that EPF is the most common oil and gas production facility utilized by marginal oilfield operators in the country [2] . The case study EPF is designed to process up to 10,000 barrels of oil per day (bopd) and 20 million standard cubic feet (MMSCF) of gas (2000 GOR), produced from onshore oil wells within 500 - 3000 meters of the production facility. Stabilized crude oil from the EPF is transferred to temporary storage tanks onsite, after which the product is evacuated through the export facilities. Currently, the associated gas from the process is primarily disposed of by flaring. However, there are plans to process the gas for domestic use and export in the context of Nigeria’s gas utilization policy.
The scope of this study was limited to five (5) units of the EPF based on the data available for the study, which include well control, gathering system, separation and stabilization, and process utilities, as shown in Figure 1.
The maintenance strategy currently adopted at the case study facility is predominantly time-based preventive across the entire facility. However, maintenance records from the facility also showed a high reliance on corrective maintenance in response to equipment failure or damage during normal operation.
3. Methodology
In this study, the RCM methodology adopted involved evaluating the EPF system to properly understand its functions and functional failures. This is followed by a systematic risk-based criticality analysis for the selected systems and an
![]()
Figure 1. Functional block diagram of the EPF.
FMEA for maintenance task selection in that order. Figure 2 presents the flow diagram of the RCM methodology adopted.
3.1. Information and Data Collection
Relevant data required for the RCM were obtained from the case study facility by administering a technical questionnaire to key operations and maintenance personnel working at the EPF. This included the operations superintendent, maintenance lead and the health safety and environment (HSE) superintendents. Field visits were also conducted to verify the information/data provided. The technical questionnaire captured details such as facility overview, equipment list, equipment functions and functional failures (complemented with theories from literature), and equipment failures and maintenance records/history.
3.2. System Description
In addition to the overview of the case study facility presented in Section 2, further inputs including the equipment list, the process flow diagram (PFD) and the system “units” functions were used to obtain a comprehensive description of the facility. Based on the selected unit systems for RCM, clear boundaries were defined across the systems, as shown in the functional block diagram in Figure 1. The diagram shows the interactions of equipment within the same boundary and across different boundaries. This is important for the equipment criticality analysis to ascertain how an equipment failure can impact the overarching system.
3.3. Equipment Criticality Analysis
Equipment failure criticality analysis (FCA) is performed to evaluate the impact of equipment failure on the overall system. To achieve the FCA, equipment maintenance history and failure records were obtained from the case study. Criticality analysis was used to evaluate the risk of a failure occurring against its consequences and the impact on the entire system or the business at large. The criteria of evaluation referred to as “risk factors” that were considered in this study included Production Loss (PL)—any failure event that can lead to production deferment or downtime of oil production, safety (S)—any failure event that could lead to injuries or fatalities, Environment (E)—any failure event that could negatively impact the environment either by pollution or damage, and Maintenance Cost (M)—direct cost associated with an equipment failure ranging from minor repairs to complete replacement.
A five-by-five (5 × 5) risk factor matrix evaluation procedure adapted from [11] was used to evaluate the risk factors based on five (5) levels of potential consequences. The failure criticality of an equipment with respect to a specific risk factor is given by the probability of failure (failure frequency) multiplied by the corresponding consequence, as shown in Equation (1). This yields a Risk Priority Number (RPN), also called Risk Rating, which determines the risk level of the failure. Table 1 illustrates the evaluation of the RPN using a 5 × 5 risk matrix. The resulting RPN risk level is either low, medium, or high, as shown in Table 2.
Probability of failure P × Consequences C = RPN (1)
The description of the consequence level used in evaluating the 5 × 5 risk matrix is shown in Table 3, while the description of failure probability/frequency is shown in Table 4. The risk-based equipment criticality analysis was conducted with a team of experts from the case study facility. Inputs to the assessment included equipment list, maintenance cycle and equipment history/failure records.
![]()
Table 1. Illustration for the evaluation of RPNs using 5 × 5 risk matrix.
![]()
Table 2. Description of RPN risk levels for the risk factors.
![]()
Table 3. Description of risk matrix consequnce levels.
![]()
Table 4. Description of risk matrix consequnce levels.
The analysis included the maintenance engineering supervisor, the maintenance lead, and the Health Safety and Environment (HSE) superintendent. The resulting 5 × 5 matrixes are presented in Table 5, within the results and discussion section.
An “Initial Composite” risk priority number denoted by RPN’ was obtained by multiplying all four RPNs (PL, S, E, M) for each piece of equipment as shown in Equation (2). The availability of redundancy or standby was also considered in evaluating the criticality of a piece of equipment, primarily in the area of production loss risk factor. A piece of equipment with a standby or redundant unit is expected to reduce the impact of failure only in production recovery. This is because the tendency of the failed unit to impact safety, the environment, and the maintenance cost remains the same. Thus, a “Residual Composite” risk priority number denoted by RPNR was evaluated considering the availability of standby equipment where applicable. Equipment with redundancy is expected to contribute to the production loss by an operational rule of thumb of 10%, which accounts for the time taken to switch over from the failed equipment to its backup. The calculation for the residual composite RPN is expressed in Equation (3).
(2)
where:
RPN' = Initial composite risk priority number;
PL' = Production loss risk priority number;
S' = Safety risk priority number;
E' = Environmental risk priority number.
Thus,
(3)
where:
RPNR = Residual composite risk priority number;
R = Availability of equipment standby.
And:
where:
: when there is an availability of equipment standby, and;
: when there is no availability of equipment standby.
After the residual composite risk priority number RPNR is obtained, the next step is to determine the risk level or category for the RPNR.
Recall the risk level description from Table 2 where the risk levels were defined as Low, Medium, and High for each risk factors obtained from the individual 5 × 5 risk matrix. The risk level for the RPNR simply considers the upper limits of the individual risk factors’ RPN. Multiplying these upper limit yields the range for the composite RPN for each risk level, this we termed Max Composite RPN for the respective risk levels.
Thus, the max composite RPN for each risk level is given by the expressions below:
(4)
where:
RPN− = Max Composite RPN;
PL− = upper limit of Production loss RPN;
S− = upper limit of safety RPN;
E− = upper limit of safety RPN;
M− = upper limit of maintenance cost RPN.
The max composite RPN risk levels are therefore defined as follows:
Low: 1 to Low level max composite RPN
Medium: low level RPN− +1 to Medium level RPN−;
High: Medium level RPN− +1 to High level RPN−.
The Max composite RPN for the three risk levels are presented in Table 6 within the results and discussion section.
3.4. Failure Mode Effect Analysis
The FMEA is adopted to identify how the equipment at the production facility might fail and the relative impact of the identified failures. The main objectives of the FMEA are to the identification of the possible ways in which failure can occur (failure mode), their causes and the magnitude of the effects on the equipment or the system (failure effects) [13] . The FMEA, in the context of this study, was employed to analyze the equipment identified with medium and high failure criticality (residual risk priority number, RPNR) to the system, thereby recommending appropriate maintenance tasks. The inputs to conducting the FMEA included the equipment functional failures that have occurred in the past, or those with the tendency to occur. Functional failures that have occurred were obtained from the facility equipment failure log.
The FMEA considered the equipment on a component basis. Each component of the equipment is analyzed to identify the failure modes, causes and the effect of the failure on the three dimensions described as follows
1) Local effect: component level;
2) System effect: equipment level, and;
3) Plant effect: effect of the failure on the overall EPF.
The outputs of the FMEA are contained in Table 7 and Table 8, in the result and discussion section.
3.5. Maintenance Task Selection
Maintenance task selection for critical equipment is proposed to promote reliability-based maintenance. Thus, the aim is to identify equipment or components that can be maintained in the category of Corrective maintenance, Preventive Maintenance, and Condition-based maintenance that an artificial intelligence program can support. To achieve the maintenance selection, the outcome of the FMEA is further analyzed as shown in the flowchart in Figure 3.
Failures with low or no effect on component and system level are recommended for corrective maintenance. This is because such failures are usually associated with non-critical parts, which do not affect related parts or the system upon failure. In addition, the failure can be easily corrected with readily available spares at lower cost, compared to carrying out routine preventive maintenance, which according to [14] , can be imperfect, thereby accelerating the failure mode.
Failures with medium system level impact were further analyzed using the 5 × 5 criticality analysis risk matrix. A low RPNR indicates that the component can be placed under the corrective maintenance scheme, as with the previous case. If the RPNR falls within the medium or high-risk rating, then the component can be considered for routine or time-based preventive maintenance. Similarly, if the failure has a high system effect and up to medium plant level impact, such should be considered for routine or preventive maintenance. Lastly, if failure poses a high risk to the plant, a Root Cause Failure Analysis (RCFA) is performed through a Fault Tree Analysis (FTA), to identify the type of failure exhibited by the component.
![]()
Figure 3. Maintenance task selection flowchart.
Failures are broadly categorized into three types in relation to the bathtub curve: early life failures, random (or constant failure), and wear-out failures [15] [16] . Early life failures are failures that occur at the early stage of equipment utilization, resulting from the faulty assembly, transportation or installation damage, or design error. Random failures are those that occur within the useful life of the equipment. They tend to have a random frequency and may be due to external events such as human error, improper operating procedures, overloads, etc. Reliability predictions and evaluation play a significant role in this type of failure. Lastly are wear-out failures, which increase towards the end-of-life of equipment or component.
Early life failures usually occur regardless of maintenance intervention; such failures fall in the category of reactive or run-to-fail. Engineering best practices in design, installation and commissioning are considered the best way to prevent such failures. Random failure is considered for CBM, particularly AI-based, to enable reliability monitoring and identification of potential failure before manifesting into functional failure. Wear-out failure, on the other hand, especially for non-repairable components, could have a pre-determined Mean Time to Failure (MTTF), either from the experience of operating the equipment or from industry standards and guidelines. If the MTTF is known, time-based preventive maintenance is recommended; otherwise, such can also be considered for condition-based maintenance (CBM).
4. Results and Discussions
The risk matrix generated for the criticality analysis is shown in Tables 5(a)-(d), which was used to obtain the equipment failure risk priority numbers as described in section 3. The residual composite risk priority numbers (RPNR) were categorized into respective risk level using the max composite RPN risk level shown in Table 6. A plot summarizing the equipment criticality analysis conducted is shown in Figure 4. The well control fixed choke assembly and the steam boiler unit were identified as equipment with the most failure criticality to the EPF. Other equipment with low RPNs were not considered for further analysis in this study; as such they were recommended for routine inspection and maintenance as per industry best practice and or OEM recommendations. Nonetheless, further system level-based RCM can be performed to address such equipment.
Further analysis was performed on the identified critical equipment using FMEA, CA and RCFA/FTA. As shown in Table 7, the steam boiler components fell mostly within the category of CM and PM. The most critical component was narrowed to the Pressure Safety Valve (PSV), which could lead to a catastrophe
![]()
Figure 4. Plot of equipment failure criticality analysis.
![]()
Table 8. Wellhead choke FMEA and maintenance task selection.
in the EPF facility in the scenario of failure or unavailability when required. There are industrial recommendations and statutory requirements on periodic inspections and recertifications of the PSV based on best practices. Thus, the PSV was recommended for PM. The result for the wellhead choke assembly, however, as presented in Table 8, showed that the equipment’s main component, the choke nozzle, exhibits random failure tendencies, and when it occurs, it is difficult to identify by physical inspection because the failure is mostly hidden. Such occurs within the internals of the equipment [17] . This necessitates the need for close monitoring of the performance conditions. As a result, the wellhead choke was recommended for condition-based monitoring maintenance, while failures associated with the choke body can be addressed by appropriate pre-commissioning pressure tests and periodic integrity test post-commissioning.
5. Conclusion
The outcome of the RCM conducted for the case study EPF within the Niger Delta zone of Nigeria provided an indication of equipment whose failure can significantly affect operations at the production facility. The steam generation unit and the wellhead choke assembly. The result of the component level FMEA conducted on the equipment aided the development of a robust maintenance management strategy, which is based on an optimized mix of corrective, preventive and condition-based monitoring maintenance the EPF. The proposed maintenance management strategy has the potential to reduce OPEX because it reduces routine preventive maintenance, which subsequently reduces costs from spare parts, labor and risk of failure from imperfect preventive maintenance. Furthermore, it enables the maintenance team to identify non-critical equipment parts that can be run to failure and thereafter replaced or corrected, which saves costs on routine parts replacement and prevents imperfect preventive maintenance that could result in unprecedented damage to parts or equipment. Such parts are common within the steam generation unit. In addition, the wellhead choke’s main component was identified to require condition-based monitoring maintenance because of the failure mode it exhibits, which is hidden in nature. This has the potential to cause a major loss to the plant’s operation, specifically causing damage to the oil reservoir if failure is not immediately addressed. Therefore, the future research direction would be to integrate the CBM with Artificial Intelligence capabilities such that it can trend the performance data of the equipment and flag any case of deviation from the expected outcome.