From Master Data to World-class Maintenance: a case study at a case company

(1)

From Master Data to World-class Maintenance: a

case study at a case company

Msc. Technology & Operations Management

University of Groningen

Faculty of Economics and Business

Groningen

,

August 2014

First Supervisor:

Dr. ir. W.H.M. Alsem

Co-assessor:

Prof. dr. R.H. Teunter

Student:

Remco Hebben

Student number:

1987070

(2)

Abstract

In the face of the current global competition and increasing demands from stakeholders, there is a basic business demand to improve manufacturing performance and this need to improve has brought the maintenance function into limelight. In order to perform maintenance, plant registers or maintenance databases should exist. As there is only limited research on Master Data Management and practical research has been widely understudied, this research focusses on the research question: “What Master Data is needed to perform World-class

Maintenance?” Relevant literature and a case study at the case company are used for

(3)

List of Abbreviations

BOM Bill of Materials

CBM Condition-based Maintenance

CM Corrective Maintenance

FFA Functional Failure Analysis

FMEA Failure Mode Effect Analysis

FMECA Failure Mode, Effect and Criticality Analysis

FTA Fault Tree Analysis

IER Improving Equipment Reliability

KPI Key Performance Indicator

MD Master Data

MDM Master Data Management

MPI Maintenance Performance Indicator

MPM Maintenance Performance Measurement

PI Performance Indicator

PM Preventive Maintenance

RCAM Reliability-centred Asset Maintenance RCM Reliability-centred Maintenance

SHE Safety, Health and Environment

WCM World-class Maintenance

(6)

1. Introduction

In the face of the current global competition and increasing demands from stakeholders, there is a basic business demand to improve assets manufacturing performance and this need to improve has brought the maintenance function into limelight (Muchiri, Pintelon, Gelders, & Martin, 2011). The use of maintenance systems for predicting and monitoring the health of machines can provide a valuable support to maintenance operators to isolate and fix the asset components that possibly will fail (Espíndola et al., 2013). Therefore maintenance systems could increase the manufacturing performance of assets.

In order to perform maintenance, a plant register or maintenance database should exist to keep track of asset information and make identification of assets possible. As organizations are facing information technology landscapes that are littered with legacy, packaged and developed applications, coupled with multiple data warehouse, and uncontrolled, unstructured data across the enterprise (White & Genovese, 2006) the emerge for high quality data arises. High-quality Master Data is a prerequisite for organizations to accomplish strategic business goals and provides organisations with the ability to integrate, analyse and exploit the value of their key data assets (Ofner, Straub, Otto, & Oesterle, 2013; Tuck, 2008).

Master Data (Management) is one of the most recent topics in the information system discipline and the need for a set of fully integrated, high quality data has for a long time been recognized, but there is only limited research on Master Data (Silvola, Jaaskelainen, Kropsu-Vehkapera, & Haapasalo, 2011). The practical research has been widely understudied, and this practical research is needed in order to understand current phenomena and develop new solutions.

(7)

The remainder of this paper is organized as follows: In the next section the research questions will be addressed. In chapter 2 the theoretical framework will be provided. In the third chapter the methodology will be addressed. In chapter 4 a case description will be provided. In chapter 5 the initial analysis conducted will be summarized. In chapter 6 enhancing equipment reliability will be addressed. In chapter 7 the results will be provided. In chapter 8 this research will be discussed and in the last chapter a brief conclusion will be given and recommendations for further research will we provided.

1.1. Research Question

The focus of this thesis is to gain insight in the data needed in order to perform World-class Maintenance. For this purpose a case study is provided to gain additional insight. Indeed, it is clear to the case company that in order to continue to add value to their services in the future, they need to have the necessary maintenance data available. As the demand for Master Data emerged, they are willing to know which Master Data should be in place to be able to perform World-class Maintenance. In the face of the current global competition the case company has, as well as other manufacturing and maintenance firms, a basic business demand to improve manufacturing performance. The general research question assessed in this research will be:

“What Master Data is needed to perform World-class Maintenance?”

This research adds knowledge to the existing knowledge base of Master Data and is relevant to academics and practitioners active in the field of Master Data. Therefore, the main purpose of this research will be concentrated at the development of Master Data for World-class Maintenance.

In order to answer the general research question, other sub research questions will be used to address subsections for this paper. These sub research questions include:

- How should Master Data, maintenance performance, and World-class Maintenance

be defined?

- Which challenges and preconditions related to Master Data exist? - What data is needed to perform World-class Maintenance?

- What Master Data and Master Data priorities exist?

(8)

2. Theoretical framework

The theoretical framework gives an overview of literature on Master Data (Management), manufacturing performance and World-class Maintenance. Section 2.1 provides an introduction to the main theory. Section 2.2 discusses the general definitions of Master Data. Section 2.3 discusses maintenance performance. Section 2.4 relates to how to define World-class Maintenance. Section 2.5 considers challenges related to Master Data and discusses preconditions for Master Data. At last, in section 2.6 a conceptual model will be provided.

2.1. Introduction

In recent years, rapid and uncoordinated innovations in the information technology sector have left businesses and institutions with heterogeneous computing applications and systems (Mashiku, Stephen, & Kleiner, 2011). Increasing storage capabilities, layers of “enterprise” solutions, multiple groups managing data, ownership issues, and short-term workarounds are factors that increased the data management challenge (H. A. Smith, 2008). As a result, managing and using information from multiple vendors and platforms could be expensive and time consuming. Systems integration is about bringing together the components of computing subsystems into one system that functions as a single unit, to increase performance, reduce complexity and optimize the information technology infrastructure, and to enable effective information sharing and collaboration between employees (Mashiku et al., 2011). When organizations have multiple different copies of information which should be the same, then the business stakeholders could encounter problems. Master Data (MD) play a key role in the core operations of a business, and are shared and used by several applications that make up the whole system (Mashiku et al., 2011).

2.2. Master Data

(9)

and business operations should revolve around Master Data (Mashiku et al., 2011). Not all data is Master Data, but Master Data has the ability to integrate, analyse and exploit the value of key data assets (Ofner et al., 2013; Tuck, 2008). As said, it is used across the entire company (Priglinger & Friedrich, 2008), and is often referred to as a “single version of the truth,” and according to Loshin (2009), Mertens (2000) and Ofner et al. (2013) Master Data has a clear distinction from other types of data, as Master Data always describes the basic characteristics of objects from the real world, usually remain largely unaltered, are quite constant with regard to volume, and constitutes a reference for transaction data. The important aspect to note is that Master Data only involves the essential/key business elements. Where, Master Data Management (MDM) comprises all activities for creating, modifying, or deleting that Master Data (H. A. Smith, 2008).

Figure 2.1 illustrates where Master Data and its management fits into the “data ecosystem,” where a data ecosystem describes data cycles consisting of a network of interactions among data, and between data and its environment. Therein, it clarifies activities and technologies that are related to MDM (H. A. Smith, 2008). It is important to understand the relationships between the different entities in the ecosystem in order to recognize the work that must be done to integrate any MD initiative into other concurrent information management activities within the organization and a make a MD initiative successful (H. A. Smith, 2008).

(10)

Master Data is all about delivering a single view of key business entities and harmonizing and managing the system, which has often been referred to by other researchers (Media, 2006; H. A. Smith, 2008; White, Radcliffe, & Newman, 2007).

According to Smith (2008) Master Data has certain obvious benefits: - Better information, as there is only “one version of the truth.”

- Cost savings, as the data quality increases, which leads to better decision making and less data verification.

- Improved business capabilities, as quality information is critical to the ability to improve agility and performance.

- Improved technical capabilities, as data management and access will be simplified and streamlined by eliminating data redundancies.

2.3. Maintenance Performance

Maintenance management became relevant for companies to stay productive and proﬁtable. Well-defined performance indicators can potentially support identification of performance gaps between current and desired performance and provide indication of progress towards closing the gaps. Within the maintenance management function, maintenance performance measurement (MPM) is perceived as an important function to achieve sustainable performance of any manufacturing plant (Muchiri et al., 2011; Pintelon & Van Wassenhove, 1990). In order to achieve this, maintenance managers need to set objectives and need to be able to measure maintenance performance. Based on literature and their experience, Van Horenbeek and Pintelon (2014) summarized a generic list of maintenance objectives on the strategic and tactical levels of the organization:

- Maintenance budget: maintenance costs, maintenance value.

- Functional and technical aspects: availability, reliability, maintainability, Overall Equipment Effectiveness, productivity, output quality and maintenance quality.

- Plant design life: capital replacement decisions and life-cycle optimization. - Support: inventory of spare parts and logistics.

- People and environment: environmental impact, safety/risk/health and personnel management.

(11)

tactical maintenance objectives as well as operational maintenance objectives. However, Muchiri et al. (2011) found that performance indicators should not be defined in isolation, but should be the result of a careful analysis of the interaction of the maintenance function with other organisational functions.

2.4. World-class Maintenance

The term ‘world-class’ began to emerge in the eighties when numerous organizations decided to develop strategies for improvement of productivity. Many authors have defined World-class Maintenance (WCM) in different ways (Mishra, Anand, & Kodali, 2006). A World-World-class Maintenance System (WMS) refers to the best practices in maintenance that are followed and adopted by various organizations to transform themselves to be a ‘world-class manufacturer’ (Mishra et al., 2006). Todd (1995) defined World-class Maintenance as being the best in one’s particular sector of industry in the world. Kodali, Mishra, and Anand (2009) attempted in their paper to understand what World-class Maintenance is, what its characteristics are and how it differs from other maintenance systems. They said that “best practices” will leads to the development of World-class Maintenance Systems. Norman (2001) defined WMS as “a holistic system which is created when organizations combine coherent, visionary leadership with robust processes and a supportive culture to ensure that the vision and ownership of appropriate maintenance methods permeate the organization”. WMS can also be defined as the collection of best practices in maintenance that are followed and adopted by various organizations to transform themselves to be a “world-class manufacturer” (Kodali et al., 2009).

World-class can be deﬁned as “a tool used to search for and allow a company to

perform at a best-on-class level (Labib, 1998).” Extending this definition of world-class to

World-class Maintenance results in the definition of (Ingalls, 2000), who defined World-class Maintenance as the consistent demonstration of industry best practices and producing bottom-line results as well and it can be described as “maintenance without waste” (Mishra et al., 2006).

2.5. Challenges and Preconditions Related to Master Data

(12)

quality, having a data friendly company culture, having clear definitions for roles and responsibilities, building an organizational structure to support the data processes, having clear definitions for processes, having managerial support, and having a unified data model for the information system. They also found data-related, process-related and information system-related challenges experienced by companies concerning Master Data implementation. Companies had problems with defining Master Data and Master Data models. Besides, data are in different formats and are unreliable. Concerning process-related challenges they found that companies had no clear definitions for data ownership, difference in importance, underpowered employees, laborious protocols, and incomplete data processes. Concerning the information system-related challenges, most of the companies had problems with integration but also with different data storages.

These findings are supported by Smith (2008) as he found that an information management strategy and principles should be important contributors to an organization’s enterprise architecture, data management, data quality, and data integration. He stated that an enterprise information policy should be developed, business ownership should be provided, governance should be created, and the role of IT should be specified.

(A. J. J. Braaksma & Meesters, 2012) also provided a number of problems for the information management in a maintenance environment. He found problems as, the uncertainty of future information needs, maintenance knowledge is insufficiently accessible, information cannot be used without additional knowledge, maintaining high quality asset data is complex and costly, heterogeneity of storage applications, data hand-over problems, and the lack of information standards.

These challenges regarding Master Data should be resolved. Before starting any Master Data management program, a business case with costs, timings and benefits should be created (Silvola et al., 2011). The next steps according to Silvola et al. (2011) include creating a solid picture of the current status of the data, minimizing the number of different data storages and applications, and at last ensuring data quality.

2.6. Conceptual Model

(13)

the result of implementing Master Data and meeting the preconditions. The performance can be evaluated on different strategic and tactical performance indicators, such as maintenance budget, functional and technical aspects, plant design life, support, and people and

environment. As maintenance performance is perceived as an important function to achieve performance for any manufacturing plant, increasing maintenance performance will

eventually lead to World-class Maintenance as World-class Maintenance is the consistent demonstration of industry best practices. Data management is not all that matters to increase maintenance performance, as for example maintenance strategies and policies play key roles, however, other factors are left outside the scope of this thesis.

Preconditions:

- Data ownership - Defined roles and

responsibilities

- Well defined protocols - Complete data processes - Unified data model - Data quality

- Data friendly company culture and organizational structure to support the data processes - Managerial support World-class maintenance Maintenance performance - maintenance budget - functional and technical aspects - plant design life - support

- people and environment Master Data

(14)

3. Methodology

In this chapter, the methodology of this thesis will be discussed, with the primary aim of

theory building from a practical perspective. The thesis will be executed by studying relevant literature in Master Data (Management) and equipment reliability. The literature in this study is identified using multiple academic databases, using "forward and backward" searching.

At first, a literature review was conducted to gain an accurate understanding about Master Data Management and other core concepts, such as equipment reliability, maintenance policies and maintenance data. The next step was to analyse this literature to gain insight in the process of Master Data. The scope of this research is limited to the Master Data needed for the process of (continuously) improving maintenance (enhancing equipment reliability) due to the time restrictions within this research. Therefore the data needed to perform this process were identified first using existing literature. Then, the case was analysed to find additional data elements that are important for enhancing equipment reliability. These both resulted in data requirements needed to perform the process of enhancing equipment reliability. Finally, based on the characteristics of Master Data, the data the Master Data was identified. This identification resulted in an overview of the Master Data needed for enhancing equipment reliability in order to perform World-class Maintenance.

(15)

4. Case Description

The theoretical basis proposed in this research will be developed in cooperation with a case company. This chapter will provide a general description of this case company. This company is a relatively young company providing services focussed on supporting the process industry. Their current goal is to become world class in the field of maintenance, which means that they aim at being one of the best and leading companies in maintenance.

The process industry is a challenging work environment, which is focussed on safe and optimal production. The case company wants to contribute to the growth and development of the process industry in Europe, by supporting companies in the process industry with their services.

(16)

5. Initial Analysis

As other organizations, the case company wants high-quality data, but they lack a roadmap to get there. Smith (2008) said that MDM is being promoted as a means of developing such a roadmap, to focus companies on the job of creating a single view of their most important pieces of information in order to improve the accessibility of their most critical data (e.g., customers, products, employees).

Master Data Management is not a relatively new concept within the case company. The benefits of MD are recognized but work is still ongoing. Interviews revealed that MD is a challenging concept and hard to separate from common data practices: although the subject of the interviews where initially based on Master Data, soon other data practices were discussed. Even as the case company has provided a clear (working) definition of Master Data and its management, the differences between master, reference and transactional data are still quite vague. Additionally, after the business case has been created, necessary for achieving support from higher managerial levels and for a successful MDM project (Silvola et al., 2011), it is still unclear how to improve the Master Data.

The case company has defined their MDM vision as: “To acquire appropriate Master Data

Management.” The case company is using the seven building blocks of MDM, by Gartner

Inc. for achieving this vision. They cleared that appropriate MDM is the key to the development of world class which should be accomplished by:

- Effective and efficient (business) processes. - Discipline related methods.

- Reliable management information, which should be complete, correct timely,

consistent, and available and accessible.

(17)

creating a data friendly company culture and organizational structure to support the data processes, and in total ensuring data quality.

Proceeding with the initial analysis, it can be seen that the pressure on MD is increasing within the case company as there is a trend towards Reliability-centred Maintenance, decreasing inventories, and a need for an increase in efficiency. Due to this increased pressure, the case company (besides the difficulties encountered with the preconditions) is also struggling with the identification of its current data. As said by Silvola et al. (2011): it is necessary to acknowledge the relevant data needed to do business and to create a solid picture of the current status of the data. This conception is shared by the case company, as they are willing to gain insight into their current state of data and identify business relevant data before creating a bigger picture of Master Data.

5.1. The Scope

(18)

Figure 5.1: Structured approach to achieve WCM (Campbell, 1995).

(19)

Improving Equipment Reliability Reliability-centred Maintenance Im p le m en t P P M p ro gr am Define optimum PPM plan (FMEA) Criticality raking of equipment S tu d y p re p ar at io n C ri ti ca l i te m se le ct io n Sy st em s el ec ti o n an d d ef in it io n FM EA /F M EC A Se le ct io n o f m ai n te n an ce ac ti o n s Tr ea tm en t o f N o n -cr it ic al It em s D at a co lle ct io n P re ve n ti ve M ai n te n an ce C o m p ar is o n A n al ys is D et er m in at io n o f m ai n te n an ce in te rv al s Im p le m en ta ti o n M o n it o r re lia b ili ty p er fo rm an ce a n d im p ro ve P P M p la n s A n al ys e an d im p ro ve P P M P ro gr am Monitor, Analyse and Improve Ex ec u te m ai n te n an ce

Figure 5.2: Reliability-centred Maintenance and Improving Equipment Reliability.

(20)

6. Enhancing Equipment Reliability

Maintenance plays an important role for achieving World-class Maintenance, where many programs have been raised for the purpose of improving maintenance, such as Total Productive Maintenance (TPM), Reliability-centred Maintenance (RCM), Business-centred Maintenance (BCM) and Life Cycle Costing (LCC) (Waeyenbergh & Pintelon, 2002). In pursuit of continuous improvement, only two of these available methodologies enhance the reliability of physical assets (A. K. Jardine & Tsang, 2013):

- Total Productive Maintenance: a people-centred methodology. - Reliability-centred Maintenance: an asset centred methodology.

However, a significant limitation of RCM is its lack of capability of determining which maintenance strategies are the most cost effective (Bertling, Allan, Eriksson, & Member, 2005). The Reliability-Centred Asset Maintenance approach (RCAM) is a quantitative approach of RCM relating preventive maintenance of equipment to system reliability and total cost (Bertling et al., 2005). RCAM will not be evaluated in this thesis, as quantitative data is not Master Data of nature as it does not remain largely unaltered. Besides quantitative data does not describe basic characteristics of objects and is not a reference for transaction data. However, as the equipment reliability improvement process is an asset centred top-down methodology, TPM is briefly described in section 6.1 and section 6.2 will focus in more detail on RCM.

6.1. Total Productive Maintenance

TPM can be regarded as an improvement programme to promote productive maintenance by establishing a comprehensive productive-maintenance system throughout the entire life of the equipment, spanning all equipment-related fields, with the participation of all employees from top management down to shop-floor workers, to promote productive maintenance through motivation or voluntary team-based activities (Chan, Lau, Ip, Chan, & Kong, 2005; Konecny & Thun, 2011). Therefore, TPM is an approach which involves and empowers production related employees to establish maintenance improvement. Moreover, TPM contains the following five elements:

- It aims at getting the most efficient use of equipment.

- It establishes a thorough system for the equipment’s entire life span.

(21)

- It promotes productive maintenance through autonomous, small-group activities.

Since TPM aims to maximise equipment effectiveness, the overall equipment effectiveness (OEE) is the core metric of the TPM processes (Chan et al., 2005; Waeyenbergh & Pintelon, 2002). The OEE, which is the product of performance efficiency, availability and quality rate, can be considered as a combination of the operation maintenance, equipment management, and available resources designed to determine the reliability of assets (Chan et al., 2005; Chen, 2013). For the equation see Appendix 3.

Although, the key component in TPM is maintenance, TPM is not really a maintenance concept. TPM goes much further than maintenance only, but is incomplete as a maintenance concept because it does not provide clear rules to decide which basic maintenance policy will be used (Waeyenbergh & Pintelon, 2002). The need for data for TPM will not be discussed in this thesis, as TPM is a people-centred incomplete maintenance concept, whereas the case company uses RCM, an asset centred methodology.

6.2. Reliability-centred Maintenance

An alternative approach for enhancing asset reliability is RCM, which is design focussed. Reliability-centred Maintenance is a system engineering method at present that is

used to determine and optimize maintenance tasks. Its primary objective is to preserve system functions and therefore enhance system reliability. The various steps of the RCM analysis require a variety of data, like design data, operational data, and reliability data.

There are three basic approaches for estimating equipment reliability (A. K. S. Jardine & Buzacott, 1985). First, empirical data on the equipment or of similar design could be analysed. Data on equipment lifetimes from the field or the laboratory should be acquired and analysed using statistical techniques, so that a lifetime distribution could be fitted to the equipment life data (A. K. S. Jardine & Buzacott, 1985). The lifetime distribution can therefore be defined as the probability that the unit breaks down before or at a certain point in time. Second, assuming that the equipment consists of components with independent failure processes, equipment reliability could be predicted from component reliability (using FMECA) (A. K. S. Jardine & Buzacott, 1985). Third, developing models based on the

(22)

physical, chemical or other mechanisms of failure could help for estimating equipment reliability (A. K. S. Jardine & Buzacott, 1985). The best know example of the mechanism of failure is the strength interference model, where a failure will occur if the realized value of applied load/stress (random variable) is greater than the realized value of strength (random value with a given probability distribution) (A. K. S. Jardine & Buzacott, 1985).

The RCM analysis may be executed as a sequence of activities. The RCM method can be divided into three main activities: the identification of maintenance significant items, the assignment of suitable PM tasks for the significant items, and the implementation and update of the PM tasks (Selvik & Aven, 2011). When, evaluating the twelve steps for RCM provided by Rausand (1998) and combining it with the eight steps for RCM provided Yuan et al. (2012), this resulted in a sequence of eleven activities for RCM described below. However, activities could be overlapping in time due to their cohesive nature.

Table 6.1: Reliability-centred Maintenance

Step Activity

1 Study preparation

2 System selection and definition

3 Critical item selection

4 Data collection

5 FMEA/FMECA

6 Selection of maintenance actions

7 Determination of maintenance intervals

8 Preventive maintenance comparison analysis

9 Treatment of non-critical items

10 Implementation

11 Monitor, analyse and improve

In the next sections, the activities of this RCM analysis will be described and, if necessary, be supplemented with a description of the case analysis. Additional the need for data will be described briefly.

6.2.1. Step 1: Study Preparation

(23)

with preventing unacceptable equipment reliability and integrity risk. It does so by defining, implementing, executing and continuous improving the maintenance actions.

However, the establishment of a project group is more a managerial and organisational one time off establishment, and lies more in the subject of establishing a group which reduces for example groupthink, increase group diversity, etc. As this establishment of a project group is business specific, not maintenance directed, and in most cases is a one time off operation, the identification of data needs for this step will be left out of scope. Besides, a project group or department responsible for the RCM project is already established within the case company.

6.2.2. Step 2: System Selection and Definition

All systems may in principle benefit from RCM. However, with limited resources, priorities between systems should be made. Therefore, before performing an RCM analysis at a plant, two methods for selection should be addressed. First the system for analysis should be selected by evaluating to which systems an RCM analysis will be beneficial. Second the system needs to be defined by evaluating at what level of the assembly (plant, system, component or part) the analysis should be conducted, where most of the operating plants already have developed some sort of assembly hierarchy.

Case analysis

Based on defined relevant Key Performance Indicators (KPI), regarding Safety, Health and Environment (SHE), integrity, and economic value, additional with the manufacturing plan and the maintenance strategy, the manufacturing goals and thresholds will be identified. This, together with the defined systems, based on their hierarchy in the structure list (from plant via functional location and equipment, to components), will lead to the selection of systems on which to perform the analysis. The manufacturing plan and the maintenance strategy are two important aspects, as they both provide insight into the objectives of a plant (e.g. reducing (maintenance) cost, or increasing availability, etc.).

Data need

(24)

6.2.3. Step 3: Critical Item Selection

This third step as functional failure analysis (FFA) has three different objectives. First, the system’s required functions and performance criteria should be identified and described. Usually a system will have more different functions and for the RCM analysis it is essential that all the important system functions are identified. Second, the input interfaces required for the system to operate needs to be described. The third step of the FFA is to identify ways in which the system might fail to function, so to identify and describe the potential system failure modes. The criticality must be judged on the plant level, and should be evaluated over four important performance classes (Rausand, 1998; Yuan et al., 2012):

- Safety.

- Environmental.

- Economic (availability). - Maintenance cost/task.

This step will end with the identification of potential critical items with respect to the functional failures. It may be beneficial to focus particularly on critical items, to not waste time and money, however in some cases all items should be analysed.

Case analysis

Instead of using a functional failure analysis, the case company is using a criticality matrix to determine critical equipment. In determining the criticality matrix, plant data, equipment data, process information, asset utilization data, maintenance data, maintenance costs, and SHE data is used to determine the criticality of sections. Then sections are prioritized, on where to start first, using the plant breakdown structure. Then the criticality will be determined by the effect of a failure on first safety, second health, third environment, fourth production (losses, or contrary availability), fifth maintenance (costs), and the sixth and last one quality (problems), where the probability of occurrence will be used as well. This process of criticality ranking will result in A (high critical), B (medium critical), and C (low critical) items.

Data need

(25)

equipment data, (6) process information, (7) asset utilization data, (8) maintenance data, (9) maintenance costs, and (10) SHE data.

6.2.4. Step 4: Data Collection

As the various steps of the RCM analysis require a variety of input data, in this section the necessary reliability data will be described briefly. Among others, the following reliability data should be collected (Bertling et al., 2005; A. K. S. Jardine & Buzacott, 1985; Martorell et al., 1999; Rausand, 1998; Subcommittee et al., 2001; Yuan et al., 2012):

- Mean Time To Failure (MTTF). - Mean Time Between Failures

(MTBF).

- Mean Time to Repair (MTTR). - Overall equipment effectiveness

(OEE).

- Relative frequency histogram. - Survival/reliability function. - Failure rate function.

- Performance indices. - Inspection records. - Equipment data. - Equipment manuals. - Failure data. - Maintenance data. - Process data. - FMEA data/reports. - Engineering data. - Test reports.

- Failure data and statistics. - Maintenance protocols. - Maintenance history.

- Maintenance labour and material costs.

The operational and reliability data are collected from available operating experience and from external files where reliability information from operating conditions (empirical data) and systems with similar design may be found (e.g., data banks, data handbooks, field data from own data storage, manufacturer‘s recommendations) (A. K. S. Jardine & Buzacott, 1985; Rausand, 1998). Plant specific information is required, which must be used closely related to operational practices, programs and work management of individual plants.

Case analysis

(26)

Data need

The data needed to be collected is: (1) reliability data, (2) the provided format, (3) previously used data, (4) information from other plants, (5) data for redefinition, and (6) information on maintenance strategies should be collected.

6.2.5. Step 5: FMEA/FMECA

RCM is a system engineering method, which uses a decision making method of reliability analysis based on failure mode effect analysis (FMEA) to determine the failure mode, risk rank and criticality of equipment (Yuan et al., 2012). However, as risk rank and criticality are already added to the FMEA, the notation of failure mode, effect and criticality analysis (FMECA) by Rausand (1998), will be used from here on. The objective of this step is to identify the dominant failure modes of the selected item. As noted earlier and also supported by Braaksma, Klingenberg, and Veldman (2013) FMECA is applied on a limited selection of “critical” assets. However, identifying the failure modes for conducting an FMECA is one of the main challenges and those failure modes and effects are not always identified with sufficient accuracy. This could be the result of absence of standards and not always applying the FMECA according to clearly defined paper- or software-based procedures. As in most cases FMECA is executed as a one-off exercise, and after which changes were usually made to the preventive maintenance plan without reference to the original FMCEA assumptions and outcome (rendering them out-of-date), the FMECA method does generally not ensure consistency in maintenance decision making ( a. J. J. Braaksma et al., 2013). If there is, like in most cases, an absence of empirical failure data or other measured data the use of this data is generally not possible and therefore, the FMECA then relies predominantly on expert judgment ( a. J. J. Braaksma et al., 2013). In Appendix 4 an example of a FMECA is given.

Case analysis

(27)

Data need

Most of the data needed to perform a FMEA is given in the above section of data collection. This data will help to identify system functions and its failure modes. However, this can be supplemented with the data format provided in the FMECA, which results in additional data: (1) functional location (tag/code), (2) operation mode (e.g. running, standby, etc.), (3) function, (4) failure mode, (5) effect of failure, (6) MTTF, (7) criticality, (8) failure cause, (9) failure mechanism, (10) %MTTF, (11) failure characteristics or detection methods (e.g. measurable, age-dependent or random), (12) maintenance action, (13) failure characterises measure, (14) recommended maintenance interval, (15) preferred data requirements, and (16) additional data collection actions.

6.2.6. Step 6: Selection of Maintenance Actions

(28)

Mainenance Strategy Preventive Maintenance Corrective/reactive Maintenance Time-based Maintenance Condition-Based Mainenance Continuous Monitoring Opportunistic Mainenance Calendar/fixed time-Based Mainenance Time/running-Based Mainenance Predictive Mainenance

Figure 6.1: Maintenance Strategies.

Rausand (1998) has identified a structured approach (decision logic) also applied by Selvik & Aven (2011) for selecting the right maintenance action. However, as that RCM logic tree does not take preventive maintenance, but only overhaul and replacement, into account, an updated logic/decision tree provided by Afefy (2010) is used and can be seen in Figure 6.2.

Will the failure have direct and adverse effect on safety,

healt and environment and mission (quantity or quality)?

Will the failure result in other economic loss (high cost

damage to machines or system)?

Is there an effective CBM technology or approach?

Run-to-Fail?

Develop and schedule CBM task to monitor condition and

perform CBM task.

Is there an effective Interval-Based task?

Develop and schedule Interval-Based task.

Redesign system, accept the failure risk, or instal

redundancy. Yes Yes No No Yes No No

(29)

First it will be identified if the failure will have an effect on health, safety, security and environment. If not, it will be identified if the failure will have an adverse economic effect. If not, the maintenance action will result in corrective maintenance and the system will run to failure. If one of both is so, it will be evaluated whether condition-based maintenance is possible. This means that condition indicators should be measurable and detectable, an effective CBM technology or approach should exist. Then the maintenance item could be monitored continuous or be predicted by inspections. Otherwise time-based maintenance should be applied if an effective interval-based task exists. At last, if a time-based maintenance task is not feasible, the system should be redesigned, the failure risk should be accepted, or redundancies should be installed.

(a) Corrective Maintenance

Corrective maintenance has a main feature that actions are only performed when a machine breaks down. No interventions are undertaken until a failure has occurred. After the failure has occurred actions can be taken directly (unplanned corrective) or can be planned to be performed later (planned corrective). Actions to be undertaken can be divided into (minimal) repair and replacement.

(b) Preventive Maintenance

Preventive maintenance (PM) is a proactive maintenance strategy based on equipment reliability characteristics. This data makes it possible to analyse the behaviour of the equipment and allows for a definition of a periodic maintenance program for the machine. The preventive maintenance policy is implied to determine a series of inspections, replacements and/or component overhaul and repair with a frequency related to the failure rate of the equipment/component. In other words, preventive maintenance is effective in overcoming the problems associated with the wearing of components. Preventive maintenance can be performed calendar based or running based. This means that it can be performed for instance once a month hour or after on month of running the equipment.

(c) Opportunistic maintenance

(30)

perform all relevant maintenance interventions at the same time. Opportunistic maintenance often yield plant shutdown. However, considering the scope of this research this type of maintenance will not further be evaluated.

(d) Condition-base Maintenance

Condition-based maintenance (CBM) is another proactive maintenance strategy. A prerequisite for the application of condition-based maintenance is the availability of a set of measurements and data acquisition systems to monitor the machine performance in real time. The continuous monitoring of conditions can point out abnormal, allowing for performing the necessary controls and, if necessary, stop the machine before a failure can occur. Then, the actions to be undertaken can be divided into overhaul and repair or replacement.

Condition-monitoring techniques can be classified according to the type of symptoms they are designed to detect, which are as follows (Moubray, 1990):

- Dynamic effects, such as vibration and noise levels. - Particles released into the environment.

- Chemicals released into the environment.

- Physical effects, such as cracks, fractures, wear and deformation. - Temperature rise in the equipment.

- Electrical effects, such as resistance, conductivity, dielectric strength.

(e) Predictive Maintenance

A third proactive maintenance strategy is predictive maintenance. Unlike the condition-based maintenance policy, the predictive maintenance policy analyses acquired control parameters data periodical to ﬁnd possible temporal trends. This makes it possible to predict when the controlled value will reach or exceed the threshold values, and therefore be abnormal. Then, depending on the operating conditions, unavoidable component substitution or revision can be planned for preventive maintenance.

Case analysis

(31)

used to select maintenance actions is the result of the FMEA and its identified failure mechanisms, which could lead to control mechanisms or redesign. The strategies are evaluated on the estimated impact on life cycle costs, quality and plant integrity. However, to evaluate the control mechanisms, the shutdown-time and frequency, repair cost and costs of spare parts warehousing should be known.

Data need

The data needed to perform the logic tree is based on: (1) previous identified data, (2) the FMEA and its failure mechanisms, and (3) the available control mechanisms, (4) shutdown-time and frequency, (5) repair cost and (6) costs of spare parts warehousing.

6.2.7. Step 7: Selection of Maintenance interval

Determining an optimal interval for performing some maintenance policy is a very difficult task, as it depends on a large amount of reliability data like the failure rate function, the likely consequences, costs and risk of the task, the MTTF, the OEE, etc. A general problem is that the necessary input data is rarely available or not in the format required. Even if the necessary input data is known, the selection intervals is still submission to the fact that maintenance task could be grouped into maintenance packages carried out at the same time (opportunistic maintenance), or should be carried out in a specific sequence (Rausand, 1998). These problems are sometimes found to be so overwhelming that practitioners do not use these maintenance optimization models to optimize the maintenance intervals, where they do solely rely on the manufacturers’ recommendations and past experience (Rausand, 1998). In most cases this will lead to a situation ending up with too frequent maintenance activities.

Case analysis

(32)

decreasing it is becoming more attractive and perhaps costs effective. The use of terminology, for example, is currently expanded to find risks of failures by increasing hotspots.

The determination for the use of duplications, are made preferably within the design, using FMEA with design for maintenance and design for reliability. However, this use of duplications depends on the strategy of the plant or installation. If duplications are not included in the initial design, however the failure of a plant or installation results in critical failures, the case company will look at possible solutions in design, CBM, preventive maintenance and corrective maintenance. Design can result in the use of a different type of equipment, building redundancy, or using spare parts. Based on the RCM methodology certain strategies will be selected. However, certain strategies can be supplemented like the combination of corrective maintenance with spare parts inventory. However, as said, it depends on the criticality of equipment.

Data need

All previous collected and identified data will be used for the evaluation of maintenance intervals.

6.2.8. Step 8: Preventive Maintenance Comparison Analysis

In RCM two criteria for selecting maintenance tasks are used. Plants must satisfy quantitative reliability requirements while at the same time minimizing their costs (Bertling et al., 2005). Under the principle of keeping the reliability and safety of equipment at minimal cost, each task must be applicable and each task must be cost-effective (Deshpande & Modak, 2002; Rausand, 1998; Yuan et al., 2012).

Applicability - A maintenance task is applicable if it will prevent, mitigate or eliminate a

failure, or at least reduce the probability of occurrence. A maintenance task will also be applicable if it will detect onset of a failure or discover a hidden failure.

Cost-effectiveness - A maintenance task does not cost more than the failure(s) it is going to

(33)

performing it, this means that it is possible that the maintenance task with the lowest costs is not the most cost-effective option. Cost-effectiveness is based on finding a balance between costs and benefits.

Case analysis

At this moment maintenance tasks are selected based on risk, where risks are weighted against costs. When a certain failure exposes high risk, so A critical items, the task must be applicable, it must preferably eliminate a failure, but in a lesser extent do costs matter. Contrary, for low critical items maintenance task, the task should be applicable, however with costs as low as possible. As said, risk plays an important factor. For example for critical equipment especially the cost of downtime can be very high, therefore preventive maintenance actions will be more beneficial than the cost of downtimes.

Data need

Costs and the risk of a failure and its maintenance prevention program are important for comparing preventive maintenance analysis. However, all previous collected and identified data will be used for comparing preventive maintenance analysis.

6.2.9. Step 9: Treatment of Non-critical Items

When step 1-8 is performed, the maintenance task for the critical items selected in step 4 will be identified; however what remains is what to do with the items which are not critical? For plants already having a maintenance program, a brief cost evaluation should be carried out and if the existing maintenance cost related to the non-critical items is insignificant, it is reasonable to continue this program (Rausand, 1998). However, as one of the biggest disadvantages of RCM is its complexity and its price, the RCM method will not be applied to non-critical maintenance items (Waeyenbergh & Pintelon, 2002).

Case analysis

(34)

critical A items. Where A “high” critical items should be less than 20% of equipment due to the amount of work required to perform FMEA’s.

Data need

The data needed to perform the treatment of critical items could be used to perform the treatment of non-critical items, however the analysis of these data will be more superficial and therefore less data could be needed.

6.2.10. Step 10: Implementation

Necessary for implementing the result of the RCM maintenance optimization analysis is that the organizational and technical maintenance support functions are available (Rausand, 1998). This means that the organization must be able to support the maintenance tasks as well as that the technical maintenance support functions should be available to be able to improve the reliability, availability, and profitability of plants. As experience has shown that many accidents occur during maintenance or as a result of inadequate maintenance, it is therefore of vital importance to consider the maintenance tasks associated risk when implementing a maintenance program (Rausand, 1998).

Case description

When implementing the maintenance strategy, the case company translates these plans into tasks what has to be done. Separate strategies can be combined into one maintenance plan taking into account the different frequencies. These plans will be entered into the maintenance management system and will eventually (automatically) generate work orders that have to be performed. However, a phenomenon that could be seen within the case company is, there are less interferences during the holidays, as fewer people are present, such as operators that change the process settings. It appears that decommissioning and maintaining a system often leads to a period with the occurrence of more failures (e.g. leaking gaskets after a stop and start).

Data need

(35)

6.2.11. Step 11: Monitor, Analyse and Improve

As more operating data becomes available, as RCM systematically analyse and documents data, the operation experience can be used to adjust and update initial decisions. For each significant failure that occurs in the system, failure characteristics should be compared with the initial FMECA (Rausand, 1998). This step will be described in more detail in section 6.4.

6.3. Execute maintenance

Reliability-centred Maintenance is the basis of improving the reliability of equipment by selecting appropriate maintenance actions. The need of data for performing maintenance will be

identified in this section, as this provides an important input for maintenance improvement. The data resulting from the execution of maintenance will be used in performing the RCM methodology. Therefore, the maintenance data identified in this section could enlarge the view of Master Data needed in order to perform World-class Maintenance. According to ISO/DIN Standard 14224 (2004), and NEN-EN Standard 13460 (2002) the following reliability and maintenance data should be collected: 6.3.1 Equipment data, 6.3.2 Failure data and 6.3.3 Maintenance data. Collection of this data should make it possible to exchange reliability and maintenance data within a company and between companies. It is recommended that each piece of information is coded where possible, as it is advantageous versus free text because the facilitation of queries and analysis of data, the ease of data input, the consistency check undertaken at input, by having predefined code lists, and by minimization of database size and response time of queries.

Case description

The purpose of doing maintenance is to ensure asset/equipment reliability. Related to the process of doing maintenance the case company has identified several data that has to be maintained: (1) functional location, (2) equipment data, (3) functional location Bill of Materials (BOM), (4) equipment BOM, (5) material BOM, (6) functional location task list, (7) equipment task list, (8) general maintenance task list, and (9) proactive maintenance plans. This need of data together with the previously identified (10) failure data and (11) maintenance data is needed in order to perform maintenance and should be available for analysis.

(36)

6.3.1. Equipment data

This section provides insight into equipment data needed in order to perform World-class Maintenance. Equipment data is categorized into classification data (e.g. industry, plant, location, system), equipment attributes (e.g. manufacturer’s data, design characteristics), and operation data (e.g. operating mode, operating power, environment). Additionally, some equipment specific data may be required. It should be clear that for example an equipment identification tag should be available in order to select equipment. Besides, operational information should be available for analysis, like the initial equipment commissioning date and the operational time. The equipment data is used as input for both the RCM process and the process of IER and is therefore data needed in order to perform both RCM and IER.

Case analysis

The format below (table 6.2) is used in the case company as well. This equipment data is captured within SAP. However, some interviewee identified that they were missing equipment data among the age of equipment. Sometimes the initial equipment commissioning was even missing. Besides there was a need to know the life expectancy, economic life, theoretical life and the technical life, which was entirely not available. They said: “These are available to a lesser extent, but so essential that we do not even know them. Besides, to the fact of missing equipment data, a lot of data is data unstructured. There is much written in open text, which make analysing the data much more difficult and time consuming. Additional, data is in different applications.”

Table 6.2: Equipment data

Data category Data Business example

Use/Location attributes

Industry Natural gas

Business category Midstream

Installation category Pipeline

Installation code or name Beta gas line

Owner code or name Johnsen Inc.

Geographic location Europe

Plant/Unit category Compressor station

Plant/Unit code or name CS 3

Section/System Compression

Operation Category Manned

Equipment attributes

Equipment class Compressor

(37)

Equipment identification/Location C1001

Equipment description Main compressor

Unique equipment identification number (tag code)

10101

Manufacturer’s name Wiley

Manufacturer’s model designation CO2

Design data relevant for each equipment class and subunit/component as applicable

Equipment specific

Operation

Normal operating state/Mode Active standby

Initial equipment commissioning date 2003.01.01

Start date of current service 2003.02.01

Surveillance time, h 8000

Operating time, h 100

Number of demands during the surveillance period as applicable

2 Operating parameters as relevant for each

equipment class

Equipment specific

Additional information

Additional information in free text as applicable

Specify as needed

Source of data Specify as needed

6.3.2. Failure Data

This section provides insight into failure data needed in order to perform World-class Maintenance. Failure data is categorized into identification data (e.g. failure record number and related equipment that has failed) and failure data for characterizing a failure (e.g. failure date, items failed, failure impact, failure mode, failure cause, failure detection method). Additionally, it could be necessary to collect specific types of data. In order to analyse failures and improving equipment reliability, for example the data when the failure occurred, what its impact has been, or what its cause has been should be available. If this data is not available, the analysis will become impossible. The failure data is also used as input for both the RCM process and the IER process and is therefore identified data needed to perform RCM and IER, which will be evaluated in chapter seven, based on the characteristics of Master Data.

Table 6.3: Failure data

Data category Data to be recorded Description

Identification Failure record Unique failure record identification

Equipment identification/Location E.g. tag number (see Table 6.2)

(38)

(year/month/day)

Failure mode Usually at equipment-unit level

Failure impact on plant safety Usually zero, partial or total Failure impact on plant operations Usually zero, partial or total Failure impact on equipment

function

Effect on equipment-unit function: critical, degraded, or incipient failure

Failure mechanism The physical, chemical or other

processes which have led to a failure Failure cause

The circumstances during design, manufacture or use which have led to a failure

Subunit failed Name of subunit that failed

Component/Maintainable item(s) failed

Name of the failed maintainable item(s)

Detection method How the failure was detected

Operating condition at failure Running, start-up, testing, idle, standby

Remarks Additional information

Give more details, if available, on the circumstances leading to the failure: failure of redundant units, failure cause(s) etc.

6.3.3. Maintenance Data

(39)

Table 6.4: Maintenance data

Data category Data to be recorded Description

Identification

Maintenance record Unique maintenance identification

Equipment Identification identification/location

E.g. tag number

Failure record Corresponding failure identification

record

Maintenance data

Date of maintenance Date when maintenance action was

undertaken or planned

Maintenance category Main category

Maintenance priority High, medium or low priority

Interval Calendar or operating interval

Maintenance activity Description of maintenance activity

Maintenance impact on plant operations

Zero, partial or total

Subunit maintained Name of subunit maintained

Component/maintainable item(s) maintained

Specify the component/maintainable item(s) that were maintained

Spare part location Availability of spares

Maintenance resources

Maintenance man-hours, per discipline

Maintenance man-hours per discipline Maintenance man-hours, total Total maintenance man-hours

Maintenance equipment resources E.g. intervention vessel, crane

Maintenance times

Active maintenance time Time duration for active maintenance work being done on the equipment

Down time Time duration during which an item is

in a down state Maintenance delays/problems

Prolonged down time causes, e.g. logistics, weather, scaffolding, lack of spares, delay of repair crew

Remarks Additional information

Give more details, if available, on the maintenance action and resources used

6.4. Monitor, Analyse and Improve

As more operating data becomes available, when the RCM method systematically analyse and documents data, the operation experience can be used to adjust and update initial

decisions. For each significant failure that occurs in the system, failure characteristics should be compared with the initial FMECA (Rausand, 1998). This process of analysing and

(40)

updating should be concentrated on three major perspectives: short term interval adjustments, medium term task evaluation, and long term revision of the initial strategy (Rausand, 1998). The short term update may be considered as a revision of the previous analysis, based on updated failure and reliability data. Significant failure causes identified by maintenance experience not considered in the initial analysis may update the medium term strategy and will review the selection of maintenance actions. At last, the long term revision will consider the whole analysis. RCM is a proactive reliability enhancing methodology; however what remains is what to do after a failure has occurred. Two well-recognised popular reactive techniques for identifying failure characteristics, (a) Root Cause Analysis and (b) Fault Tree Analysis, will be briefly described next. These techniques are used after a failure has occurred, to analyse the failure and improve the implemented maintenance action(s). Some authors have placed the RCA around the FMEA, however like other authors, RCA is in this thesis captured within monitor, analyse and improve, as it is a reactive approach which should be used to redirect the initial FMEA.

(a) Root Cause Analysis

Besides failure modes and effects analysis (FMEA), root-cause analysis (RCA) is one of the most popular failure analytical methods used by practitioners around the globe. RCA is known as a process of identification based on a structured approach of the root cause of a problem and to take the actions necessary to eliminate it (Andersen & Fagerhaug, 2006). Andersen & Fagerhaug (2006) have identified seven activities, bifurcated into two broad categories (Sarkar, Mukhopadhyay, & Ghosh, 2013), which should be performed for conducting the root cause analysis: (1) identification of the potential causes; problem understanding, problem cause brainstorming, problem cause data collection and problem cause data analysis, and (2) validation to root cause; root cause identification, root cause elimination, and at last solution implementation. Where RCA classifies the problem into associated categories, such as people, procedures or hardware, and tries to prevent problem recurrence (Chen, 2013).

(b) Fault Tree Analysis

(41)

analysis that works from the top down – assuming the system has failed and then trying to work out why it failed (Chen, 2013).

6.4.1. Performance Measurement

Besides comparing failure characteristics with the initial FMECA, after the occurrence of a significant failure, performance measurement are important in managing the maintenance function (Muchiri et al., 2011). Performance indicators could potentially support the identification of performance gaps between current and desired performance and provide indication of progress towards closing the gaps. However, the need of data for performance indicators will be left out of the analysis, as they monitor performance and are not Master Data, as they do not describe the basic characteristics of object, and do not usually remain largely unaltered. In Appendix 5 a definition format for Performance Indicators (PI) is presented to gain insight in the data needed for a performance indicator. Additional to this format, Appendix 6 provides an overview of commonly used performance indicators, maintenance performance indicators, and maintenance performance measurements.

Case analysis

Performing a FMEA will result in the identification of certain failure modes. It will be attempted to prevent those failure modes from failing with preventive maintenance. However, once a failure has occurred, subsequently a RCA will be performed. The RCA could form the basis to alter the initial proactive maintenance plan and the initial FMEA is updated as well. Unfortunately, this does not always happen at the case company. However, this is not always needed, as in some cases the basic assumptions of the FMEA before and after the occurrence of a failure are comparable. Then, there is no need to alter the initial FMEA.

(42)

(43)

7. Results

In this chapter the data identified in previous chapters will be evaluated, and screened against the characteristics of Master Data. As said, not all data is Master Data, but Master Data has the ability to integrate, analyse and exploit the value of key data assets (Ofner et al., 2013; Tuck, 2008). Master Data involves first of all key business elements. Besides, Master Data always describes the basic characteristics of object(s), usually remains largely unaltered, is quite constant with regard to volume, and constitutes a reference for transaction data. The data used in the process of enhancing equipment reliability will be evaluated in this chapter. Using the five characteristics of Master Data above, it will be identified which data is Master Data and which is not. This will create insight into what Master Data is needed to improve equipment reliability and eventually will lead to World-class Maintenance.

In Table 7.1 all identified data needed in order to perform RCM and IER is screened against the five characteristics of Master Data. In the first column of Table 7.1 the data attributes are described. In the second column of Table 7.1 it is evaluated if this data attribute is a key business element, in the third column it is evaluated if it describes the basic characteristic of (an) object(s), in the fourth column it is evaluated if the data attribute remains largely unaltered, in the fifth column it is evaluated if it is quite constant with regard to volume, and in the sixth column it is evaluated if the data attribute is a reference for transaction data. If a data attribute meets a characteristic of Master Data this will be identified with an “X” in the row of the data attribute and in the column of the characteristics it satisfies. Master Data will then be identified with an “X” in the seventh and last column in Table 7.1, if all five characteristics are met.

From Master Data to World-class Maintenance: a case study at a case company