Quantifying Breakeven Price Distributions in Stochastic Techno-Economic Analysis — A Case of Cellulosic Biofuel Production from Fast Pyrolysis and Hydroprocessing Pathway

Techno-economic analysis (TEA) is a well-established modeling process in which benefit-cost analysis (BCA) is used to evaluate the economic feasibility of emerging technologies. Most previous TEA studies focused on creating reliable cost estimates but returned deterministic net present values (NPV) and deterministic breakeven prices. Nevertheless, the deterministic results cannot convey the considerable uncertainties embedded in techno-economic variables such as capital investment, conversion technology yield, and output prices. We obtain distributions of NPV, IRR, and breakeven price. The breakeven price is the most important indicator in TEA because it is independent of scale and communicates results effectively. The deterministic breakeven price is the price for which there is a 50 percent probability of earning more or less than the stipulated rate of return. For an investment under relatively high uncertainty, it is unlikely that investors would provide financing to a project with a 50 percent probability of loss. The point estimate breakeven price, therefore, does not represent the threshold under which investment would occur. In this study, we introduce the stochastic techno-economic analysis in which we incorporate Monte Carlo simulation into traditional TEA. A case of cellulosic biofuel production from fast pyrolysis and hydroprocessing pathway is used to illustrate the method of modeling stochastic TEA and quantifying the breakeven price distribution. The input uncertainties are translated to outputs so that the probability density distribution of both NPV and breakeven price are derived. Two methods, a mathematical method and a programming method, are developed to quantify breakeven price distribution in a way that can consider future price trend and uncertainty. We analyze two scenarios, one assuming constant real future output prices, and the other assuming that future prices follow an increasing trend with stochastic disturbances. We demonstrate that the breakeven price distributions derived using our methods are consistent with the corresponding NPV distributions regarding the percentile value and the probability of gain/loss.