Eventos Anais de eventos
ENCIT 2016
16th Brazilian Congress of Thermal Sciences and Engineering
THERMODYNAMIC DIAGNOSTIC METHODOLOGY USING THERMOECONOMIC AND THERMODYNAMIC INDICATORS IN CONJUNCTION WITH ARTIFICIAL NEURAL NETWORKS (ANN)
Submission Author:
Dimas Jose Rua Orozco , MG
Co-Authors:
Osvaldo Jose Venturini, José Carlos Escobar Palacio
Presenter: Dimas Jose Rua Orozco
doi://10.26678/ABCM.ENCIT2016.CIT2016-0483
Abstract
For decades several methodologies had been developed with the aim to solve the main problem of the thermodynamic diagnosis (TD), that is, identify which equipment has malfunctions and quantify in terms of additional fuel consumption such malfunctions. TD has two trends: one based on thermoeconomic indicators (exergetic cost) and another based on thermodynamic indicators (pressure, temperature, mass flow, etc.). However these methods often only achieve their objective partially, so sometimes it is necessary to use two or more of these methods together. In this paper a diagnostic methodology was developed for externally fired gas turbines (EFGT) using the thermoeconomic method in conjunction with artificial neural networks to identify not only components with malfunctions (intrinsic malfunctions) and their fuel impact, but also the fuel impact of the thermodynamic parameters (for example, fuel impact of temperature variation). An EFGT was simulated using the commercial software GateCycleTM using wood carbonization residual gas as fuel. An artificial neuronal network (ANN) was developed with the commercial software MATLAB®. To show the applicability of the methodology is considered 5% of degradation in the performance of the Compressor and the Burner. Under these conditions are obtained intrinsic malfunctions of 15,405kW and 5.11 kW in the burner and the compressor, respectively, where, the intrinsic malfunction of 5.11 kW in the compressor is represented by 8.6572 kW caused by a greater consumption of shaft work and -3.5474 kW caused by the compressor exit temperature. In the Burner the intrinsic malfunction of 15,405 kW is caused by the combustion temperature.
Keywords
artificial neuronal network, externally fired gas turbine, fuel impact, intrinsic malfunction, thermodynamic diagnosis., artificial neuronal network, externally fired gas turbine, fuel impact, intrinsic malfunction, thermodynamic diagnosis., artificial neuronal network, externally fired gas turbine, fuel impact, intrinsic malfunction, thermodynamic diagnosis., artificial neuronal network, externally fired gas turbine, fuel impact, intrinsic malfunction, thermodynamic diagnosis.
