Replacement under ongoing technological progress

Abstract

This paper evaluates the cost savings achievable in the infinite horizon single-asset replacement problem by considering technological progress affecting assets available after multiple future replacements and allowing variable service lives. Using a simple geometric model to specify asset costs, constant service lives are shown to be suboptimal. It is shown by example that common replacement decision methods yield higher service lives for the first asset and substantially different discounted total costs for the series of assets. This effect is illustrated numerically using automobile cost data.