Analysis of uncertainties in compact plate-fin recuperators for microturbines

The current study aims to perform a stochastic analysis on microturbine compact recuperators to evaluate the impact of uncertainties in design parameters on their cost and volume, by using two different probabilistic approaches: Monte Carlo (MC) and Response Sensitivity Analysis (RSA). These two methods have been developed in Matlab® and then coupled with CHEOPE (Compact Heat Exchanger Optimisation and Performance Evaluation) software, which allows to analyze two different types of recuperators, used in microturbine applications: the furnace-brazed plate-fin type and the welded primary surface type. This paper focuses on an analysis of plate-fin type recuperators, for which the cost function adopted was tuned and verified in a previous study. Three main parameters of the recuperator have been considered as uncertain: effectiveness, hot side and cold side pressure drops. The uncertainties associated with these three parameters are based on industrial knowledge. The aforementioned stochastic methods have been used to propagate such uncertainties on the relevant outputs, such as cost and volume, allowing us to evaluate the least expensive and the most compact recuperator among those analysed.