A Step Forward To Safe and Cost Effective Maintenance Optimization in Offshore Drilling
The Competitive environment in Oil & Gas motivates drilling operations to continuously
focus on increasing safety and cost effectiveness. To ensure this, optimal reliability and appropriate
maintenance of rig equipment is a must. Optimized maintenance plays a key role in balancing the
maintenance parameters (legislative, economic and technical) and the available resources to carry
out the maintenance programme. Effective implementation of maintenance optimization leads to
improved system availability, reduces overall maintenance cost, further improves equipment
reliability and elevates system safety.
Downtime cost, long lead times and spare part costs due to equipment unavailability are prime major cost elements whereas various kinds of planned maintenance lead to a significant increase in planned maintenance costs and undesired disturbances during operations. As a result, operational and maintenance personnel must increasingly be able to identify which maintenance activities are truly necessary, and how to schedule them without significant disturbances. Making decisions on optimum volume of planned maintenance hours on various systems and equipment is considered to be one of the toughest challenges.
Maersk Drilling has been reporting failure and maintenance data in its CMMS for more than 10 years. The main purpose of the collection of maintenance history is to analyze failure causes and lessons learned from the past. To coincide with this, the Maintenance Analysis & Optimization Group is continuously focusing on expanding current best maintenance practices by shifting from “Qualitative” to “Quantitative” decision- making. To achieve this, Maersk Drilling has started to build up competencies for utilizing condition based maintenance, reliability based maintenance, maintenance optimization etc. As a result, various types of Decision Support System (DSS) and tools have been developed to support this change.
Figure 1. Overall framework for maintenance optimization methodology.
Decision Support System
A decision support system (DSS) is a computer- based information system that supports business or organizational decision-making activities. DSS serves the management, operations, and planning levels of an organization and helps to make decisions that may be rapidly changing and not easily specified in advance . To tackle the challenge of optimizing planned maintenance hours, Maersk Drilling recently developed two spreadsheet based DSS’s  for maintenance optimization. The first (DSS-1) identifies over and under maintained equipment from a pool of thousands of pieces of equipment originating from Maersk Drilling’s current fleet of 17 drilling rigs. The second (DSS-2) uses maintenance history, failures, downtime hours and cost data as inputs, and delivers optimized planned maintenance hours plus an optimized planned vs unplanned maintenance ratio at lowest total cost as outputs.
Safety is always a prime concern at Maersk Drilling, therefore the proposed methodology considers cost effective maintenance optimization under safe operations. Inputs from various internal competent engineers and legislative requirements are taken into consideration in the process of optimization.
Every model is only an approximation and is only shadows of reality . The outcomes of the proposed models are therefore considered to move forward in the right direction for significant improvement. Based on practical constraints and reality, the following limitations are considered:
- The quality of input data is important for the outcomes from DSSs.
- The maintenance plans should be finalized with the involvement of relevant competent technical superintendents.
- Maintenance hours and cost have been modified in order to maintain confidentiality as per the company’s regulations. However, the hours and cost ratios have been kept the same as they were in the original real data, in order to obtain real trends in the outputs.
The mathematical model and methodology have been developed and customized based on available data in the company’s maintenance system. An optimization approach [4, 5, 6] of minimizing total cost (planned and unplanned maintenance costs including downtime cost) was adopted as a basis for the optimization model. The methodology starts with grouping the equipment in to high and low critical categories and further calculates relative weights of four critical categories (Safety, Environment, Production and Cost as per Maersk Drilling’s criticality definition) by using the pair wise comparison method . The identified relative weight of each criticality category is used to combine all four criticalities into one number, this is given the name Criticality Index. The Criticality Index is later used to identify the equipment targeted first for maintenance optimization. The proposed mathematical model identifies optimum planned vs. unplanned maintenance ratio and planned maintenance hours per year for the targeted equipment at lowest total cost. Figure 1 briefly describes the overall methodology and interrelations between inputs and outputs of the developed DSSs.
The main challenge facing optimization was to identify equipment to be targeted for optimization, and how to set the criteria on maintenance hours for identifying equipment and organising it in over and under maintenance categories.
A mathematical model for total cost (planned, unplanned and downtime cost) has also been developed. The cost model optimizes planned maintenance hours at lowest possible cost of the system/equipment. The mathematical model will be submitted in a peer-reviewed international journal.
A case study has been carried out at one of the rigs in Maersk’s Drilling fleet. Maintenance activities and cost were collected from 2009 to 2012. Model inputs, outputs, analyses and methodologies are briefly described in the following sections.
Calculation of relative weightings
A pair-wise comparison between criticality parameters has been carried out to calculate the relative weights of the four criticality parameters (Table 1).
Figure 2. Decision Support System (DSS-1) for identifying over and under maintained equipment.
Figure 3. Decision Support System (DSS-2) for optimizing planned maintenance.
Identifying over and under maintained equipment
The DSS-1 automatically categorizes equipment into over and under maintained groups based on pre-defined maintenance threshold values, as well as the value of the criticality index. The output window is displayed in Table 2 showing equipment of Pipe Handling System and Top Drive System. In the output from DSS-1, equipment in the green zone, shows over maintained equipment whereas equipment in the red zone illustrates under maintained categories. Tubular feeding machine (Table 2) has inconsistent behaviour concerning maintenance volume as it was under maintained during the years 2009 and 2010 whereas it was over maintained during the years 2011 and 2012. Therefore no generic conclusion regarding over and under maintained categories can be made and hence no colour is assigned to this equipment. The DSS-1 takes maintenance, criticality and cost data from approximately 4500 pieces of equipment as input and provides 300–400 pieces of equipment in both over and under maintained categories as outputs. The spreadsheet subsequently filters the parameters such as criticality index, maintenance ratio, planned and unplanned maintenance hours etc. The DSS-1 performs searches to identify the targeted equipment on a yearly basis so that trends of over and under maintenance can be appropriately justified. This yearly analysis provides a better picture/view for the maintenance review. The coloured zones during the investigation period (2009–2012) act to verify the status of maintenance volume on individual equipment on a yearly basis, which again is a visual analysis of the trend of over and under maintenance on specific equipment over a period of investigation.
Optimizing planned maintenance
After identifying the targeted equipment for maintenance review, the next step is to optimize the ratio between planned and unplanned maintenance. To demonstrate the outcomes of the optimization model, two systems (Draw work system and Drilling control system) from one rig in Maersk’s drilling fleet has been chosen as an example (Table 3).
Table 2. Output from DSS-1 grouped as over and under maintained equipment
Figure 4. A) An acceptable correlation between planned and unplanned maintenance for Draw Work System. B) The correlation for the Draw Work System leads to an optimized value of planned maintenance.
Figure 5. a) Correlation between planned & unplanned maintenance hours b) Cost curves under maintenance optimization
An acceptable correlation between planned and unplanned maintenance for Draw Work System is shown in Figure 4a indicating that planned maintenance is effective in terms of reducing unplanned maintenance. Figure 5a shows no correlation between planned and unplanned maintenance for the Drilling Control System indicating that planned maintenance does not reduce unplanned maintenance. This information provides a useful input to perform comparative analysis of maintenance for this equipment within the fleet of rigs. This type of trend obtained from the correlation study emphasizes the need to analyze maintenance procedures, human factors, supplier information and detailed failure analysis.
Optimal Ratio of Planned vs. Unplanned Maintenance
As described earlier, the correlation for the Draw Work System leads to an optimized value of planned maintenance (Figure 4b and Table 3). An increase in planned maintenance hours (Figure 4b) leads to an increase in planned maintenance costs (marked blue) and decrease in unplanned maintenance costs (marked green). The sum of both planned and unplanned maintenance costs (total costs) is marked red, which is initially higher due to higher breakdown costs. The total cost decreases to a certain level but subsequently increases due to high planned maintenance costs. The optimum value of planned maintenance has been indentified when total costs reach its minimum level. This means an increase in annual planned maintenance hours may further reduce the cost and significant savings can be estimated.
In the other example, due to no correlation, the model output illustrates that planned maintenance should be performed at its minimum level. The no correlation phenomenon refers to random failures justifying “run to failure” maintenance policy, further justified by the model outputs. However, reasons behind randomness in failure should be analyzed before accepting the recommendations from model output.
The methodology has been identified as a potential tool for maintenance management to determine the most relevant equipment section for improvement and provides a scientific procedure and guideline for moving towards safe and cost effective maintenance optimization. Significant monetary benefits are expected by reducing breakdowns, as well as unnecessary planned maintenance with higher safety.
Based on the outcomes from the case study on one of the Maersk’s Drilling rigs, optimized maintenance on specific systems shows an estimated reduction of 15–20 % in maintenance costs and 10–15 % in downtime.
»»References ››  https://en.wikipedia.org/wiki/Decision_ support_system. ››  Holsapple, C.W., and A. B. Whinston. (1996). Decision Support Systems: A Knowledge-Based Approach. St. Paul: West Publishing. ISBN 0-324-03578-0. ››  Box, G. E. P, (1979). Robustness in Statistics. Academy Press, London. ››  Jardine, A.K.S. and Tsang, A.H., (2005), Maintenance Replacement and Reliability. ››  Dandotiya, R. (2011), PhD thesis, Decision Support Models for the Maintenance and Design of Mill Liners. ››  Kumar, S., Dandotiya, R., Kumar, R. and Kumar, U. (2008), Inspection frequency optimization model for degrading flowlines on an offshore platform, Vol. 15, No. 02, pp. 167-180. ››  Saaty and Thomas, L. (2001). Fundamentals of Decision Making and Priority Theory. Pittsburgh, Pennsylvania: RWS Publications. ISBN 0-9620317-6-3.
The European Agency for Safety and Health at Work (EU-OSHA) welcomed European occupational safety and health (OSH) stakeholders to Bilbao, Spain, for the 2019 Healthy Workplaces Summit on 12.11.2019. After a successful 2-year campaign, the summit provided the perfect opportunity to reflect on lessons learned and share best practice for managing dangerous substances in the workplace.
Neste and Ravago start collaboration to enable chemical recycling of over 200,000 tons of plastic waste annually
At present, no one in the world can recycle fibres from blended fabrics on a large scale. Södra now has a solution that will enable circular flows in the fashion and textile industry. Production will commence at a low rate of 30 tonnes this year, but the long-term target is to add 25,000 tonnes of textiles to the company’s pulp production.