Uncertainty estimation of the efficiency of small-scale boilers

This article presents uncertainty estimation in measurements aimed at determining the thermal balance of a small-scale coal-fired boiler and its efficiency. In order to estimate uncertainty of boilers efficiency, the GUM uncertainty framework (GUF) analytical method and the numerical method (MCM) based on Monte Carlo simulation were used. This article shows that the MCM allows to estimate the uncertainty with accuracy similar to that obtained with the GUF method. It was proved that the quality of the fuel, i.e. its net calorific value and carbon content, were key factors in uncertainty estimation, 33% and 65% respectively. Taking into consideration the obtained results, a number of feasible simplifications regarding uncertainty evaluation were suggested. They included neglecting the effect of nonlinearity of the measurement model on efficiency uncertainty, neglecting correlation between parameters measured with a flue-gas analyser, and using the reference coverage factor equal to 2 to calculate the expanded uncertainty. These simplifications are not likely to diminish the accuracy of the results. The obtained results are within the measurements uncertainty range, so it could be concluded that the influence of fuel additives on boiler's efficiency is not obvious.