ABSTRACT Operational KPIs play an extremely important role in the process industry, aiding in decision making. However, they need to be reliably calculated to be representative. The present work presents a schematic methodology for the calculation of these KPIs, including techniques of steady-state detection, denoising, error propagation and sensitivity analysis, presented, as far as it is known, in the form of a new graphical tool proposed by the authors named StatSSCandlePlot. The methodology was applied in a real case study of a gas fired boiler in which the indicator studied was its efficiency evaluated by the Stack Loss Method. From the StatSSCandlePlot it was possible to identify the trends of the indicator, the portion of each window in the steady-state, the values to be considered from the indicator and, in a complementary way, to identify the variable that most influences the variation of the indicator, through the sensitivity analysis.

ABSTRACT Sweet potato is an attractive feedstock for ethanol production due to its high starch content and favorable agronomic characteristics. This paper proposes a simple equation to estimate the total reducing sugars (including glucose from starch) in sweet potatoes based on their moisture content (low cost and simple measurement). It allows the calculation of the ethanol production potential of a given sweet potato mash. According to the equation, the ethanol potential increases non-linearly with increasing concentrations of sweet potato mash in the fermenting medium (w/v), reaching a constant value for high concentrations (22 % of ethanol to 10 kg: L of a sweet potato with a moisture content of 66 %). Additionally, the ethanol yield potential is very sensitive to the sweet potato moisture, increasing linearly when the moisture decreases. We emphasize that the relations proposed in this paper can be used by other researchers, who can apply them to their specific cases.

ABSTRACT Y-rank can present faults when dealing with non-linear problems. A methodology is proposed to improve the selection of data in situations where y-rank is fragile. The proposed alternative, called k-rank, consists of splitting the data set into clusters using the k-means algorithm, and then apply y-rank to the generated clusters. Models were calibrated and tested with subsets split by y-rank and k-rank. For the Heating Tank case study, in 59% of the simulations, models calibrated with k-rank subsets achieved better results. For the Propylene / Propane Separation Unit case, when dealing with a small number of sample points, the y-rank models had errors almost three times higher than the k-rank models for the test subset, meaning that the fitted model could not deal properly with new unseen data. The proposed methodology was successful in splitting the data, especially in cases with a limited amount of samples.