Mohsen Assadi

Condition based maintenance of gas turbines using simulation data and artificial neural network: A demonstration of feasibility

Gas turbine maintenance is crucial due to high cost for the replacement of its components and associated loss of power during shutdown period. Conventional scheduled maintenance, based on equivalent operating hours, is not the best alternative as it can require unnecessary shut downs. Condition based maintenance is an attractive alternative as it decreases unnecessary shut downs and has other advantages for both the manufacturers and the plant owners. However, this has shown to be a complex/difficult task. A number of methods and approaches have been presented to develop condition monitoring tools during the past decade. Condition monitoring tools can e.g. be developed by means of training artificial neural networks (ANN) with historical operational data. Such tools can be used for online gas turbine performance prediction where input data from the plant is fed directly to the trained ANN models. The predicted outputs from the models are compared with corresponding measurements and possible deviations are evaluated. With this method both recoverable degradation, caused by fouling, and irrecoverable degradation, caused by wear, can be detected and hence both compressor wash and overhaul periods optimized. However, non-availability of operational data at the beginning of the gas turbine operation may cause problems for the development of ANN based condition monitoring tools. Simulation data, on the other hand, may be generated by using a manufacturer’s engine design program. This data can be used for training artificial neural networks to overcome the problem of non-availability of operational data. ANN models trained with simulation data could be used to monitor the engine from the very beginning of its operation. A demonstration case using a Siemens gas turbine has been shown for this proposed method by comparing two ANN models, one trained with operational data and the other with simulation data. For the comparison an arbitrary section of operational data was used to produce predictions from both models, whereupon these were plotted with corresponding measured data. The comparison shows that the trends are very similar but the parameter values for the measured and the simulated data are shifted by a constant. Using this knowledge, one can provide an ANN based engine monitoring tool that could be adjusted to a certain engine using engine performance test data. The study shows promising results and motivates further investigations in this field. (Less)

Operation and Performance Limitations for Solid Oxide Fuel Cells and Gas Turbines in a Hybrid System

Temperature limitations of Solid Oxide Fuel Cells (SOFC) in transient single operation and steady-state Hybrid System (HS) operation with Gas Turbines (GT) are presented. For transient SOFC simulations, an unsteady-state SOFC model was developed by upgrading a detailed validated steady-state model. As critical SOFC single operation modes, concerning the risk of material cracking due to exceeding SOFC transient temperature gradients, heat-up and cool-down are investigated. For minimization of transient SOFC temperature gradients at start-up and shut-down, a stepwise heat-up and cool-down procedure is proposed. Concerning HS off-design and part-load operation, the impact of SOFC temperature limitations on the operational window is investigated. Results show a reduced operational window due to exceeding local SOFC temperature gradients, which can be reduced by optimal adaptation of GT to SOFC size.

A Unique Correction Technique for Evaporative Gas Turbine (EVGT) Parameters

Modelling and data-normalization of a gas turbine process, called Evaporative Gas Turbine (EvGT) is studied here. The most important factor to achieve a high level of accuracy during the data normalization, is the consideration of changes in thermodynamic properties of the working medium at different environmental conditions. Performance of the EvGT, which is working with a mixture of air and steam, is strongly affected by the changes in the environmental conditions. When the properties of the working fluid such as the water content are continuously changing, the normalization process using conventional techniques becomes very difficult if not impossible. In this study, measured data from the worlds’ first Evaporative Gas Turbine at Lund University in Sweden have been used for generation of an empirical model by a single Artificial Neural Network system. Performance maps generated by ANN have been successfully used for data normalization and performance prediction of the Evaporative Gas Turbine. ANN predicted values are compared with experimental results, not used during the training, where very good correlation was observed.Copyright

Optimisation of HAT-Cycles – With and Without CO2 Capture

In a world where distributed power generation and deregulation of energy markets are on everyones agenda, the need for highly efficient power plants with short lead times is greater than ever. Although at present combined cycles provide a solution, development of ever more advanced machines to increase efficiency and lower the environmental impact has led to high maintenance costs and a decrease in availability. The Humid Air Turbine (HAT) represents a different approach, suitable for distributed power generation in the medium power range. The HAT cycle, and other wet gas turbine cycles, which have been extensively studied during the last ten years, show as high an efficiency as that of combined cycles, but at a lower specific cost and, with inherently low emissions of NOX. Despite all research done no full-scale plant has been built as yet. CO2 capture is another concept widely studied in recent years. In the present study three HAT cycle configurations are investigated, two of them connected to a post-combustion CO2-capture plant. Thermodynamic and thermoeconomic optimisation of the plants was performed using genetic algorithms (GA), a robust optimisation technique based on Darwinian evolution theories. The three configurations studied were 1) a standard inter-cooled HAT cycle, referred to as the reference cycle. 2) the reference cycle together with an integrated CO2-capture plant taking the energy needed for the CO2 separation from the exhaust heat of the turbine, and 3) the reference cycle together with a CO2 capture plant, in which the energy is supplied by a separate bio-fuelled boiler. This third configuration enables all fossil-based CO2 to be separated. All power cycles were modelled using IPSEpro, a heat-and mass-balance software, employing advanced component models developed by the authors. It has an interface for optimisation and the possibility of employing user-defined objective functions. The impact of CO2 taxation was studied to determine showing which configuration is the most economical at the current fuel-price and tax-level. (Less)

Sensitivity analysis of transient compressor operation behaviour in SOFC-GT hybrid systems

This paper presents a sensitivity analysis of unsteady-state SOFC-GT-HS operation based on two different characteristic maps of centrifugal compressor taken from open literature and scaled by the law of similitude to match the design point of the Hybrid System. The system layout under investigation is a pressurised type comprising a low and high temperature recuperator. Computations are based on a one-dimensional finite element model of planar high temperature SOFC, which is validated against open literature. The reduced Moore and Greitzer model is used for compressor modelling. Calculation results of the coupled SOFC-GT-Hybrid System show that unsteady-state part-load operation is sensitive to the characteristics of compressor speed-lines but also to the load change operation procedure. Copyright (Less)

A Novel Correction Technique for Simple Gas Turbine Parameters

Data normalization for gas turbines is necessary for comparison of test data collected at various environmental conditions. The normalization procedure is regulated by the ISO-standard. In this study, a single Artificial Neural Network is used to model the performance of a simple gas turbine (VT600) using measured data at various environmental and operational conditions. Consequently, engine performance maps covering a wide range of operational and environmental conditions have been generated. Comparison of the normalized/experimental data, results provided by thermodynamic models using heat and mass balance programs and results generated by the Artificial Neural Network (ANN) model shows a high level of consistency. The study presented here was performed as a pilot study, to investigate the applicability of an ANN model for data normalization applied to an Evaporative Gas Turbine (EvGT), since the ISO-standard normalization procedure is not applicable to the EvGT plant. Results of this work show that Artificial Neural Networks are powerful tools for performance prediction as well as generation of accurate power plant model of an specific simple gas turbine, and that data normalization can easily and accurately be carried out by using these performance maps.Copyright

Fault prediction/diagnosis and sensor validation technique for a steam power plant

An intelligent sensor validation and fault prediction/diagnosis technique for a typical steam power plant is proposed and studied. An auto-associative Artificial Neural Network (ANN) is trained to examine the consistency of the overall simulated data and allocate a confidence level to each signal.The same set is used to replace the missing or faulty data with a close approximation. For fault prediction and diagnostic system a feed-forward ANN with extra linear connections is trained to recognise faulty and healthy behaviour of the steam cycle for a wide range of operating conditions. Both ANNs are tested with unseen data sets, including combined scenarios of the partially failed system to assess fault prediction capability of the proposed ANN. It is concluded that a significantly more reliable sensor reading and a highly accurate fault prediction/diagnosis system is achieved.

Pre-Design Studies of Cycles With Energy, Exergy, Thermoeconomy and ELCA Analysis

High efficiency, environmental friendliness, low operation and maintenance (O&M) costs, and lowest possible impact on the surroundings are some requirements of sustainable energy production. In selection of new power generation systems, a number of steps have to be taken into account to meet these requirements. Here the first law analysis has been implemented and investigated followed by a combination of the first and second law analyses (exergy analysis), and thermoeconomics, and finally an Exergetic Life Cycle Assessment (ELCA) is carried out for two different power cycles. The two cycles, investigated here, are a two-pressure level combined cycle, hereafter called (CC), and a Humid Air Turbine or (HAT-cycle). The main goal of this study is to point out the advantages and the difficulties related to the usage of each and every method and their combinations, and to identify the target groups that can gain knowledge and information using these methods. Since the operators of power plants often do not have access to detailed information about component materials, characteristics, etc., of the power cycle, assumptions have to be made when comparing different cycle configuration with each other. This limited type of data and information has also been used here to create a plausible scenario of how different pre-design methods can differ from each other. One major conclusion that has been drawn is that the two cycles investigated here are favorable in different situations and that the results from application of the three methods mentioned above indicate differences in which cycle is the preferable one. However, using a combination of different analysis methods illuminates the plant strengths and limitations during pre-design studies, but conflicting results need to be resolved to obtain the most cost effective and environmentally-friendly power cycle.© 2002 ASME

Novel High-Perfoming Single-Pressure Combined Cycle With CO

The European electric power industry has undergone considerable changes over the past two decades as a result of more stringent laws concerning environmental protection along with the deregulation and liberalization of the electric power market. However, the pressure to deliver solutions in regard to the issue of climate change has increased dramatically in the last few years and given the rise to the possibility that future natural gas-fired combined cycle (NGCC) plants will also be subject to CO2 capture requirements. At the same time, the interest in combined cycles with their high efficiency, low capital costs and complexity has grown as a consequence of addressing new challenges posed by the need to operate according to market demand in order to be economically viable. Considering that these challenges will also be imposed on new natural gas-fired power plants in the foreseeable future, this study presents a new process concept for natural gas combined cycle power plants with CO2 capture. The simulation tool IPSEpro is used to model a 400 MW single-pressure NGCC with post-combustion CO2 capture, using an amine-based absorption process with Monoethanolamine. To improve the costs of capture the gas turbine, GE 109FB, is utilizing exhaust gas recirculation, thereby increasing the CO2 content in the gas turbine working fluid to almost double that of conventional operating gas turbines. In addition, the concept advantageously uses approximately 20% less steam for solvent regeneration by utilizing preheated water extracted from HRSG. The further recovery of heat from exhaust gases for water preheating by use of an increased economizer flow results in an outlet stack temperature comparable to those achieved in combined cycle plants with multiple pressure levels. As a result, overall power plant efficiency as high as that achieved for a triple-pressure reheated NGCC with corresponding CO2 removal facility is attained. The concept thus provides a more cost-efficient option to triple-pressure combined cycles since the number of heat exchangers, boilers, etc. is reduced considerably.Copyright

A Dynamic Model of an Atmospheric Solid Oxide Fuel Cell System for Stationary Power Generation

This paper focuses on the transient behavior of a solid oxide fuel cell system used for stationary power production. Dynamic modelling is applied to identify the characteristic time scales of the system components when introducing a disturbance in operational parameters of the system. The information on the response of the system may be used to specify the control loops needed to manage the changes with respect to safe component operation. The commercial process modelling tool gPROMS is used to perform the system simulations. The component library of the tool is completed with dynamic models of a fuel cell stack and a pre-reformer. The other components are modelled for steady state operation. For the fuel cell a detailed dynamic model is obtained by writing the constitutive laws for heat transfer in the solid part of the cell and conservation of heat and mass in the air and fuel channels. Comprehensive representation of resistive cell losses, reaction kinetics for the reforming and heat conduction through the solid part of the cell is also included in the model. The pre-reformer is described as a dynamic pseudo-homogeneous one-dimensional tubular reactor accounting for methane steam reforming and water-gas shift reaction. The differences in the transient behavior of the system components and their interaction are investigated under load changes and feed disturbances. Copyright (Less)