MANAGING ADVANCED TECHNOLOGY SYSTEM DEPLOYMENT: AN OPTIMAL ALLOCATION BETWEEN R&D AND PROTOTYPE FUNDING

Abstract

The research and development manager allocates R&D funds to maximize the expected discounted net value of the R&D programme. Because public R&D managers do not have the same market discipline (or rewards) as private R&D managers, public R&D managers require a methodology for maximizing the expected net benefits. The US National Research Council of the National Academies in Prospective Evaluation of Applied Energy Research and Development at DOE (Phase One): A First Look Forward (2005) proposed a cost-benefit methodology to evaluate US Department of Energy's Research, Development, and Demonstration (RD&D) programmes. This paper specifies and extends this methodology, e.g., by adding cost targets into each stage of the RD&D process. Expected benefits are modelled as a function of funding levels, stage durations, stage transition probabilities, and target costs. With this method, the paper determines an optimal allocation of pre-prototype R&D funding, given a total funding constraint for an advanced energy system.

Keywords:

• Decision analysis
• Research and development
• R&D management
• RD&D
• RDD&D
• Technology development
• Innovation
• Energy policy
• Advanced energy systems

Acknowledgements

This paper was written under contract with the U.S. DOE-NE (DE-FC07-03ID14448). I thank D. Foray, O. Granstrand, R. Graber, D. Korn, J. Stamos, R. Versluis, participants at “Developing a Framework for Assessing Spillovers among Multiple Nuclear Energy System R&D Programs” at Ecole Polytechnique Fédérale de Lausanne, and members of the Economic Modeling Working Group (EMWG) of the Generation IV International Forum for their encouragement or comments. This paper reflects the views and conclusions of the author and not those of the employers, sponsors, or publishers. An earlier version of this paper is available as a CEMI (Chair in Economics and Management of Innovation) working paper with the College du Management de la Technologie of EPFL: http://cdm.epfl.ch/wpe.php.

Notes

¹Following ‘Methodology for Prospective Evaluation of DOE Programs’ NRC-NA (2005), the expected benefits should include economic, environmental, and security benefits. This paper focuses on economic benefits. However, the methodology could be extended to environmental benefits, for example, by assuming that environmental benefits are highly correlated with lower carbon content, or security benefits, for example, by assuming that security benefits are highly correlated with lower oil consumption. These benefits could then be compared to the economic benefits, and tradeoffs among these benefits could be calculated.

²This is a narrow definition of success. At the R&D stage, the most important result is a reliable information about the technology that can be used to decide whether to build a prototype. This definition of success will be explored in the portfolio version of the model.

³On target costing, see references in Everaert and Bruggmn (2002, p. 1339): ‘Current case study researchers have found that assigning a cost target (or target cost) to design engineers during [new product development] leads to future products with lower production costs than when design engineers have no specific cost target.’

⁴A spreadsheet allowing different funding levels and parameter values is available from the author at [email protected].

⁵‘Evaluating the probabilities associated with the different branches on the tree —the likelihood of technical success of the program and the likelihood it will achieve different levels of market penetration —is a critical part of the work of expert panels… To guide the expert panels, the committee developed the guidelines in the following subsections for assessing probabilities for each of the nodes in the decision tree.’ (p. 48) These guidelines would be useful for anyone assessing RD&D probabilities for any advanced technology system. Here, the technical success of the programme is P ₁ and the likelihood of market penetration is P ₂.