##
466913 Global Sensitivity Analysis of Economic Assessment of Early Stage Process Design: The Case of the Glycerol Biorefinery

The risk assessment tool described above provide valuable information on how uncertainties (technical versus external such as feed/product prices) impact the decision of optimal process concept and how to better mitigate these impacts by improving technology development. In this contribution, we take one step further and ask the question, a given a certain economic risk associated with a conceptual process alternative, what is the contribution of input uncertainties((e.g. feed composition, yield, feedstock prices, product prices, etc)) to the risk? If risk is defined as variance (deviation from an expected value), what is the individual variance attributed to each input uncertainty? And how we can use this information to further minimize economic risk and improve technology advancement? To answer these questions, we use sensitivity analysis methods including local methods such as one-factor-at-a-time (OAT) as well as global methods such as variance decomposition to compute first order as well as total sensitivity indices. We implemented algorithms in Matlab (R2015a) based on Monte Carlo sampling using Latin Hypercube Sampling with Iman Conover correlation control (Sin et al 2009) and sampling based on conditional probability of dependent variables proposed by Kucherenko et al 2012 to take into account the correlation structure of input uncertainties (especially price correlation matrix). The economic model used in this work is the discounted cash-flow rate of return (DCFR). The OAT sensitivity analysis has shown that deviations in the product’s and feedstock prices, total production cost, fixed capital investment as well as discount rate, among others, have a high impact on the project’s profitability (quantified with NPV and MSPA global sensitivity analysis of NPV indicated two major factors contributing to its variance (risk): uncertainty in fixed capital cost estimation (82%), and, uncertainty in product prices (14%). The NPV meta-model showed that improving the accuracy of fixed capital cost estimation (5% *std*) leads to a 95% reduction in the NPV’s uncertainty. This global sensitivity analysis confirms the process engineering expectation that more detailed capital cost estimation supported through pilot scale studies or more rigorous model analysis is expected to decrease the economic risk and better judge the transferability of technology to the market. Global sensitivity analysis is an important and complementary tool to study and decompose impact of uncertainties to economic assessment of conceptual process design studies.

Keywords: techno-economic assessment, uncertainty analysis, glycerol-based biorefinery concepts, risk-based decision making, global sensitivity analysis.

**References:**

Kucherenko, S., Tarantola, S., & Annoni, P. (2012). Estimation of global sensitivity indices for models with dependent variables. *Computer Physics Communications*, *183*(4), 937-946.

Sin, G., Gernaey, K. V., & Lantz, A. E. (2009). Good modeling practice for PAT applications: Propagation of input uncertainty and sensitivity analysis. *Biotechnology progress*, *25*(4), 1043-1053.

Cheali, P., Gargalo, C., Gernaey, K. V., & Sin, G. (2015). A Framework for Sustainable Design of Algal Biorefineries: Economic Aspects and Life Cycle Analysis. In *Algal Biorefineries* (pp. 511-535). Springer International Publishing.

**Extended Abstract:**File Not Uploaded

See more of this Group/Topical: Computing and Systems Technology Division