ABSTRACT
Cost, driven mostly by planned and unplanned maintenance, is the most significant barrier limiting widespread adoption of TECs. Accurate numerical models can be used to predict the structural loads on TECs and help improve their reliability. BEMT offers an efficient method for predicting the performance of TECs. A robust but unoptimized BEMT code, implemented in MATLAB, has been developed at Swansea University. Previous versions of the BEMT code only allowed for inputs of constant blade profile and Reynolds Number which contributed to the inaccuracies of the BEMT model in predicting experimental results. We present a faster and more accurate C++ implementation of the BEMT code. The error between the coefficient of power from the BEMT code and results of laboratory testing at optimum TSR was reduced by 13%. This should in turn improve the design of TEC, reducing their cost and increase their adoption as generators of low carbon energy.
