MANAGING AND LEVERAGING ACTION KNOWLEDGE: THE CASE OF FRONT-LINE OPERATORS IN THE PETROCHEMICAL INDUSTRY

ABSTRACT

In the petrochemical industry, improvements in the technology and in production processes over the last few decades have led to a reduction in accidents. However, the resulting processes have also created the need for large numbers of highly skilled operators. The anticipated wave of retirement of operators in the next few years means that the industry will experience a significant loss of senior expert operators. Establishing ways to effectively leverage and manage knowledge about refinery operations is, therefore, a critical concern. This research conceptualizes knowledge that operators in the petrochemical industry possess as “action knowledge,” that is, as a basis for action—both as tacitly stored in operators’ heads and explicitly written as codified procedures; and examines the problem as well as proposes solutions following this conceptualization. The research reported in this dissertation is organized in three essays. Essay 1 develops and describes an innovative approach and a software tool for analyzing operator procedures into chunks of explicit action knowledge. This approach uses design science as the research method. Essay 2 evaluates the feasibility and effectiveness of the proposed approach based on an empirical evaluation of the software tool. Essay 2 uses empirical analyses based on use of authentic procedures obtained from multiple refineries, and feedback from expert operators. Essay 3 develops a new framework to understand tacit action knowledge in the context of operators working in the petrochemical industry. Essay 3 uses a modified iterative grounded research methodology to analyze work practice descriptions gathered from operators following the critical incident technique. The dissertation concludes with a discussion of findings across the three essays for managing action knowledge in the petrochemical industry, and contributions to the stream of research on knowledge management along with directions for future work.

 

Chapter 1

Introduction and Motivation

1.1 Managing Operational Knowledge in Process Industries

Process industries are characterized by operations that are run either as continuous or batch processes involving material transformations. Examples of process industries include food, beverages, chemicals, petrochemicals, coal, and paper products, among others. Firms in process industries are concerned with using formulas and recipes to processing resources in order to produce other products (Makins, 1991). In process industries, the relevant factors are ingredients, not parts; formulas, not bills of materials; and overall operation processes, not individual units.^[1] An output produced through such a process has produced cannot be returned to its original components.

In process industries, human operators play a critical role in monitoring operations, responding to emergencies and failures, optimizing processes and recovery, and coordinating maintenance and repair tasks (Schragenheim, Cox, & Ronen, 1994). Operators perform these tasks around the clock every day and coordinate their work across multiple shifts (Isermann, 2006).  During their tenure in the process industry, they acquire and cultivate expertise that is difficult that newcomers simply do not bring to the industry (Strahan, 2005). Without this expertise, operations can suffer, and processes may be derailed. The result: accidents that can cause significant damage to assets, and worse, to injuries to operators and loss of life. In recent years, a number of technological advances have been made and automation systems introduced that have significantly changed the role of the operator (Ernst & Lundvall, 2004). These high-tech production processes have not replaced the operators. Instead, they have changed the operator roles significantly. They have forced the operators to acquire more sophisticated knowledge and skills. And they have prompted firms in process industries to document and codify practices in an effort to render the operators less error-prone and more effective (Grote, 2008; Rowe, 2009). This “operational,” knowledge thus, refers to the knowledge and skills that these operators (should) possess and draw on in order to ensure that the plants in a process industry are run effectively and without any serious problems.

1.2 Research Setting: Petrochemical Industry

The petrochemical industry is the focal industry for the present research. In this industry, operations are focused on the continuous processing, including extraction and refinement, of crude oil. The scale of the problem – how to manage operational knowledge – can be appreciated by considering the number of large refineries. Within the continental U.S., there are about 140 refineries, each with about 250 operators.

I selected the petrochemical industry as our research setting for two reasons. First, failures in this industry can result in catastrophic consequences with loss to property and life, which make the problem important. Over the past few decades, although both technology and government regulations have been developed to improve safety in this industry, major industrial accidents in this context may be as likely today as they were 10 years ago. For example, since 1998, neither the U.S. nor the European industry has shown a reduction in either the fatality rate or the major accident rate (Wolf, 2001) (see Figure 1-1). Beyond the fatality and non-fatality rates and the direct financial loss, the most dangerous potential consequences of petrochemical industry accidents are environment pollution and harm to members of the general public, such as people living close to a petrochemical facility. As the industry grows, the number and size of the facilities including the toxicity of the materials and the number of people living or traveling nearby may also increase (Perrow, 1999).

Second, human resource management is a big concern in the petrochemical industry. On one side, operator error is still high in the petrochemical industry. Over the last thirty-years, operator error has been the second largest cause of loss in the industry (Nivolianitou, Konstandinidou, & Michalis, 2006) (see Figure 1-2). On the other side,, the high rate of operator turnover becomes a critical issue for petrochemical refineries at the present time. Even worse, the number of operators in the petrochemical industry is predicted to decrease with a significant wave of retirements in the next few years, resulting in the loss of 20% of the operators, effectively clearing the ranks of expert operators.

 Establishing an effective way to manage operational knowledge in the petrochemical industry is, therefore, an extremely urgent concern for two reasons: (1) the industry’s accident rate is high and accidents have serious consequences, and (2) the number of experienced operators capable of avoiding or reducing the impact of accidents is decreasing. It is for these reasons that my selected the petrochemical industry as the research setting of this study.

1.3 Research Conceptualizations

Given the characteristics of the research setting, I adopted three key conceptualizations to drive my investigation. The first relies on the fundamental conceptualization of knowledge as related to operator actions. In prior work, knowledge is considered to be a justified true belief that answers question of “what” (Lehrer & Paxson, 1969). Instead of this traditional conceptualization, our conceptualization emphasizes that operational knowledge is important insofar as it supports action, namely, the knowledge to answer the “how” question. I follow the recent work of Bera and Wand to define action knowledge as “knowledge that supports effective action” (Bera & Wand, 2009). The conceptualization ‘action knowledge’ corresponds to the idea of operational knowledge, and may be a collection of rules, prescriptions, and propositions that give operators the ability to choose and execute a course of actions to achieve their goals.

Therefore action knowledge constitutes the first research conceptualization of this study.

The second conceptualization differentiates between tacit and explicit action knowledge. This conceptualization concerns how action knowledge is stored and expressed. For example, action knowledge in the petrochemical industry could be stored in the operator’s mind. As both technologies and government regulations have developed, more and more action knowledge has been written down in the form of codified procedures. These two forms are recognized by the second conceptualization and, in turn, map onto the classic distinction between explicit and tacit knowledge proposed by Nonaka and Konno (1999).

The third conceptualization focuses on the possessor of action knowledge. In the petrochemical industry, action knowledge could be possessed by individual operators in the form of their own experience and expertise. It could also be codified and managed by individual refineries or by the industry as a whole. This difference is consistent with the classification of knowledge as either individual and organizational in nature (Barney, 1986; Bhatt, 2002), and provides the possibility for analyzing how such knowledge can be managed at different levels of analyses.

1.4 Research Questions

Research concerns pertaining to action knowledge in process industries as well as the selection of the petrochemical industry as the research setting suggests the research question of this study: How should action knowledge be leveraged and managed in process industries?  In this study I select the research domain as the petrochemical industry, and investigate three related sub-questions in order to answer our overall research question: (1) How can action knowledge, codified in operational procedures, be extracted and chunked so that it may be managed as an organizational asset? (2) Is the chunking process designed for explicit action knowledge feasible and effective in the petrochemical industry? And (3) What strategies do operators use to apply action knowledge to perform tasks in the petrochemical industry? Together, these three questions are aimed at addressing the concerns outlined above about managing action knowledge in the process industry; and to contribute to the research stream related to knowledge management.

Each of these sub-questions is investigated via a research essay.

1.5 Structure of this Thesis

Essay 1 (in Chapter 5) deals with explicit action knowledge, which is action knowledge codified as standard operating procedures. The essay describes the design and implementation of a tool called the SPA (Semantic Procedure Analyzer), an approach to converting action knowledge codified in procedures into knowledge chunks that can be made available to operators at the point of action and managed by the organization as a knowledge asset. Essay 2 (in Chapter 6) evaluates the implemented tool with a set of authentic procedures obtained from multiple companies in the petrochemical industry. The evaluation is supplemented with expert feedback and assessment. Both the evaluation data pertaining to parsing accuracy and to user acceptance are analyzed. Together, these two essays follow the design science research methodology. Essay 3 (in Chapter 7) explores application of action knowledge, which including both tacit and explicit one. The essay offers a framework for understanding strategies to apply action knowledge with a particular emphasis on several categories of resources of action knowledge. It also compares the use of strategies by novice and expert operators, in order to understand the relationship between operator experience and use of strategies to apply action knowledge. Together, the three essays address the overall research question: How should action knowledge be leveraged and managed in the petrochemical industry?

This dissertation comprises eight chapters. Chapter 2 introduces a review of related research and Chapter 3 generates the research scope of this study. Chapter 4 describes the research methodologies adopted in this study. Chapters 5, 6, and 7 describe the three essays respectively. The final chapter offers a conclusion, describes the contributions of the present study, and suggests directions for future research.

 

[1] http://en.wikipedia.org/wiki/Process_manufacturing

MANAGING AND LEVERAGING ACTION KNOWLEDGE: THE CASE OF FRONT-LINE OPERATORS IN THE PETROCHEMICAL INDUSTRY

Sharing is caring!