Abstract Number: 002-0243

Business Modelling in Support of Innovative Process Development in the Speciality Chemical Industry

Second World Conference on POM and 15th Annual POM Conference,

Cancun, Mexico

April 30 - May 3, 2004

T. F. Burgess1, and N.E.Shaw2

1Author to whom correspondence should be addressed,

Leeds University Business School,

University of Leeds,

LS2 9JT,

UK

,

Phone: +44 (0)113 343 2615

2 Leeds University Business School

University of Leeds,

LS2 9JT,

UK.
Abstract

We report on research collaboration between industry and academia to improve manufacturing process development in the speciality chemical industry. Faculty drawn from Operations Management personnel in a Business School and from personnel in Chemical Engineering and Mechanical Engineering Departments have combined with industry practitioners to develop frameworks and tools to support an innovatory approach to designing and developing manufacturing processes within what can be characterised as a traditional, batch production environment. The background to the research is described and the major action research style initiatives outlined that culminated in industry-based case studies. The business aspects of the developed framework and tools will be presented from an Operations Management perspective.

Keywords: Innovation, Process Design, Business Modelling, Speciality Chemicals

1. Introduction

Porter and Ketels (2003), in advice to the UK government, have recently advocated that UK companies should pursue business strategies that rely on innovation to accentuate high-value products, presumably relinquishing commodity manufacture to other companies such as those in emerging economies. Such strategies are found in, and are advocated for, many manufacturing companies in the pharmaceutical, agrochemical and specialty chemical industries (Chapman and Edmond, 2000; Burgess et al., 2002; Walsh and Lodorfos, 2002). Emphasising innovation as a key organisational success factor means that agility becomes necessary as a key organisational capability (Goldman et al., 1995). Agility needs to be present in all areas of company operation - emphasising business or technical domains in isolation will not deliver the necessary overall capability to innovate (Daft, 1978). The ability to speedily introduce new products is a function of not just agile production plant but is also constrained by the company’s ability to develop and introduce new processes into the manufacturing plant. Clearly the business routines for introducing a new product, and new process, are powerful determinants of whether an organisation achieves agility, just as the technical and manufacturing capabilities are. However, where companies have relied for many years on familiar development routines that identify traditional manufacturing processes to produce products that are not so high-valued then upgrading the company’s agility level poses challenges. This is the situation that many, both academics and practitioners, see applies in the speciality chemical industry. In this paper we outline the results of collaborative research carried out to develop a framework for agile process and plant design within the speciality chemicals industry; a framework that integrates business and technical aspects and provides tools to support the design process. First some theoretical background is provided, then the research methodology (action research) involving collaboration between academics and industry practitioners is described. Next the results of the research are covered and some conclusions drawn. In particular, this paper emphasises that part of the research that involved Operations Management academics.

2. Background

As indicated in the introduction, companies in the speciality chemicals industry are pursuing, or being exhorted to pursue, strategies that focus on higher value added products than hitherto (Chapman and Edmond, 2000; Burgess et al., 2002; Walsh and Lodorfos, 2002; Porter and Ketels, 2003). Chemicals are seen as comprising such an important industry (Arora, Landau and Rosenberg, 1998; Howitt, 2000) that many argue that such a strategic response is necessary in countries like the UK – the industry cannot simply be allowed to whither away in the face of overseas competition. The advocated strategies require companies to respond better to customer requirements for new products to be delivered speedily in small volumes. Such requirements place demands for companies to become more agile (Goldman et al., 1995), and call for radical changes in the business and manufacturing routines to produce new products and process of what is often characterised as a traditional, slow to change industry. Some features of the industry are:

  • disconnected business and technical criteria in the new product/process activity (Olson et al., 2001)
  • inadequate knowledge of the chemical processes (Charpentier and Trambouze, 1998)
  • over reliance on batch production methods and equipment that had been used in the past (Stephanoupolos et al., 1999)

However, this is not to deny that innovation that has contributed to the growth of the chemicals industry in the past (Freeman, 1990; Landau, 1994), but such innovation has tended to be more to do with improving manufacture of commodities (Barnett and Clark, 1997).

3. Methodology

A major three-year programme funded by industry and the UK Engineering and Physical Science Research Council (EPSRC) forms the basis for the research described here, but has been extended subsequently. The programme aimed to significantly improve agile process and plant design in the speciality chemicals industry. In particular, the aim was to provide methods which: halve the total project time from the start of process development to manufacture; lead to significant reductions in total manufacturing time; lead to 30 to 40% reductions in the capital cost of new investments compared to conventional plant; and lead to plant which is inherently more versatile. The aim would be achieved by creating a framework that promoted a radical change in approach to chemical process development, linked to the design of readily reconfigurable plant and an integrative business model that facilitates assessment of technical and financial options. This paper focuses on the work of one of the four UK university groups involved in the research activities, namely the group based at Leeds University whose remit was to concentrate on business and technical integration aspects of the research project.

The main approach to the research can be best characterised as action research (Park, 1999; Stringer, 1999). An initial phase of the research, comprising literature search, interviews, questionnaire surveys and process mapping exercises, acted to clarify research objectives and study scope. In the next step academics and practitioners conceptualised the problem area that then led on to the design of frameworks and tools to support decision-making within the new process and plant design activities. A limited number of major industry-based case studies were then undertaken sequentially to test and refine the conceptual approaches in an iterative manner.

4. Results and Discussion

The research is grounded in project management where new products and processes are designed, developed and introduced on a project basis. The developed framework comprises a set of tools to support individual project management but also has a set of tools to support the introduction of the framework to the company. The discussion here will focus on the former rather than the latter. Stage gate models (Stevens et al., 1999) are often deployed to provide a high level approach to managing projects and this was felt desirable in this case. A stage gate model was designed to meet the needs of the speciality chemical companies collaborating in the research – this was named the Business Gate Framework (BGF) (see Shaw et al., 2001).

Within this framework, the first step in dealing with the company’s potential projects is to evaluate the competing business opportunities and prioritise them such that attention is focused on the most beneficial areas. This is done through the Project Impact Priority Evaluation Tool. A project that becomes the centre of attention is specified in a brief, but essential, document (Project Aims and Description) that captures both the business and technical characteristics. An exercise is then conducted where business and technical personnel document the costs, benefits and risks related to the project in a Tool for Opportunity Focus, Improvement and Evaluation (TOFIE). This tool captures the key driving forces that then inform the succeeding design and evaluation activities.

The main technical activities of process and plant design are carried out in a high level process where alternative options, i.e. different outline designs, are generated. Different categories of project options can be identified in relation to the constraints set by existing plant, e.g. situations where minor modifications are projected for existing plant, or a new plant configuration is required but existing infrastructure is to be used, or both new plant and infrastructure are required. The design of plants for families of products or processes based on reconfigurable and interchangeable modules is one approach to achieving the agility required by the business. Plant reconfigurability can be at the level of modules or of equipment – in either case some standardisation of connections (type, sizes and positions in space) for process streams, utilities and control is required. The "plug and play" concept can then be adopted. One proposal is that of a spine containing all the utilities and modules that can be connected to that spine. A change in process can involve either a change of equipment within a module or the exchange of a complete module. Either way, standardisation is one of the requirements for speed of reconfiguration of the plant.

Generated technical options are then evaluated using a business-related framework based on a balanced scorecard approach, i.e. the evaluation philosophy is not dominated by purely financial considerations. Values are estimated and compared for key business performance metrics that operationalise multi-dimensional configurations of finance, speed, quality, flexibility and risk (see Figure 1) - a set of dimensions developed from typical ones encountered in the literature, e.g. Slack et al. (2001).

For the project option to be evaluated, metrics are first selected to represent each of the five dimensions. Values for each of the metrics are established from data and parameters manipulated within the analysis tools. The estimated values for the dimensions are then presented in an integrated manner using alternative schemes. As an illustration, a polar plot is shown in Figure 2.

A more involved analysis could utilise a multi-criteria decision approach where the sensitivity of the evaluation outcome to different weightings of variables can be explored using software (VISA, 2003).

The options, or alternatives to use the VISA terminology, to be evaluated are the different ways of producing the product under investigation. At the criteria level (i.e. the five dimensions mentioned earlier) each option (alternative) is assigned a score on that criterion. In this case the scoring scale adopted is from 0 to 5, where 1 represents the score for the existing option and 5 is the maximum that can be achieved[1]. Scores at the criteria level are weighted and aggregated up to the overall level to give a score for each option. For example, consider the pattern of scores in Table 1.

Table 1: Example of Scores Against a Project Option

Scores for the stated alternative on the stated criterion
Criteria / Financial / Speed / Quality / Flexibility / Risk
Weights / 0.25 / 0.125 / 0.2 / 0.125 / 0.3
Alternatives / Existing / 1 / 1 / 1 / 1 / 1
Improved / 1.2 / 1 / 1.2 / 0.7 / 1.8
Blue Sky / 1.5 / 3 / 5 / 1.3 / 0.7

If all criteria were equally weighed, as in the table above at 20% each, then the overall scores for the options would be as given in Table 2.

Table 2: Example of Comparing Scores For Competing Options

Alternatives / Overall score on the scale of 0 to 5 / Overall score with normalisation[2] to scale of 0 to 100
Existing / 1 / 20
Improved / 1.3 / 26
Blue Sky / 2.1 / 42

On the basis of the above analysis the Blue Sky option would be preferred. Of course the calculation above could just as easily have been carried out in Excel or any other spreadsheet[3]. The benefits of using VISA stem from aspects such as the ease of:

  • calculating and visually displaying the results
  • modifying the estimates used for both the scores and weights, and
  • examining the sensitivity of the decision to changes in the values for these estimates.


Figure 3 shows the screen shot for the above example and illustrates some of the basic features of the VISA software.

Figure 3 Screenshot of VISA

The top left of the screen displays the criteria hierarchy with the top level (overall) to the left, the five criteria to the right and the weights shown on the connecting link. The bottom part of the screen shows the alternatives window where the scores for each criterion for each alternative have been entered using a scale of 0 to 50. The window on the middle right of the screen contains the weights for each criterion (in this case expressed as a decimal fraction, e.g. 0.200 rather than 20%) while the top right window displays the overall scores for the three alternatives on a scale of 0 to 100.

The weights for a criterion can be altered by using the mouse cursor to grip the top of the on-screen bar representing the weight and moving the height of the bar up or down. The effect on the overall scores for the alternatives is shown immediately in the appropriate window. Thus the impact of the weights on the overall score can be seen interactively. For example, someone might argue that the score of 5.0 (or as expressed here, 50) against quality for the Blue Sky option is bound to ensure that this alternative scores best overall. However, it can be easily shown that removing the quality criteria from the analysis by altering the weight to zero does not alter the outcome where the Blue Sky alternative scores highest. This point can also be shown by producing the graph of overall score against the criterion weight, as in Figure 4.

Figure 4 shows the current value of the weight for quality as the chained, vertical line. The graph shows that irrespective of the value for the weight, the Blue Sky alternative will dominate the other alternatives.

This allows the introduction of judgemental inputs by members of the team and the sharing and refinement of that information. This then leads to a shared view – involving both business and technical staff - on the best option and ensures a consistent aim of the project team. Similarly, the analysis could be based on stochastic variables using, for example, the @RISK™ (@Risk, 2003) extension to the Excel™ (Excel, 2003) spreadsheet. The outcome of the evaluation is a selected option that can then be further specified in sufficient depth to go forward to the detailed processes of design, development and construction.

The tools described above were implemented in computer programs to assist rapid option evaluation, and applied successfully in three major industry case studies. The specific details of these studies must remain confidential to the companies and academic partners involved. In general terms, however, the innovative options identified using the framework and tools developed held out substantial benefits. In the first study, which focussed on the modernisation of processing methods that were felt to be dated, the potential for significant reductions in operating costs was identified, with over 8% reduction in materials plus labour costs if a new plant were to be constructed. Payback was also estimated to be within 12 months, with an increase of 30% in operating profits projected for the new plant. The second study, which addressed the production of a fine chemical intermediate from raw material, identified the potential for significant cost savings on the existing process with respect to waste treatment costs and improvements in yield. A new plant option was assessed to reduce operating costs by 64%, and to significantly improve both yield quality and processing speed. In the final study, which concerned the design of a new, purpose-built agile plant, a 40% increase in net income per year, with payback achieved in just over a year, was projected. These case studies demonstrated that the original objectives of the action research initiative were fully achieved, i.e. that the methods developed could halve the total project time from the start of process development to manufacture, and lead to significant reductions in the total manufacturing time and to 30 to 40% reductions in the capital cost of new investments. The final objective, to provide methods that lead to plant that is inherently more versatile, was also considered to have been achieved.

5. Conclusion

The novelty in this research stems from integrating business and technical aspects in a framework and tools that support the design of agile plant. The research outcomes arose from an action research collaboration of a novel nature in drawing on academics from business school and chemical engineering backgrounds. However, the collaboration of academics and practitioners also ensures as far as possible the utility of the outcome for the latter group. The ongoing nature of the research described here means that potential exists for additional work to further refine and develop the approach.

6. Acknowledgements

This work was supported, in part by a consortium of industrial partners, and in part by the EPSRC, Grant Reference GR/L65949/01.

7. References

Arora, A., Landau, R. and Rosenberg, N. (1998) ‘Introduction’, in Rosenberg, N. (ed.), Chemical and Long-term Growth: Insights form the Chemical Industry, John Wiley and Sons, London, pp: 3-23.

Barnett, B.D. and Clark, K.B. (1997) ‘Technological newness: an empirical study in the process industries’, Journal of Engineering and Technology Management, Vol. 13 pp: 263-282.

Burgess, T.F., Hwarng, H.B., Shaw, N.E. and de Mattos, C. (2002) ‘Enhancing value stream agility: the case of the UK speciality chemical industry’, European Management Journal, Vol. 20 pp: 197-210.

Chapman, K. and Edmond, H. (2000) ‘Mergers/ acquisitions and restructuring in the EU chemical industry: patterns and implications’, Regional Studies, Vol. 34(8) pp: 753-767.

Charpentier, J.-C. and Trambouze, P. (1998) ‘Process engineering and problems encountered by chemical and related industries in the near future. Revolution or continuity?’ Chemical Engineering and Processing, Vol. 37 pp: 559-565.