Analysis of Acoustic Problems with Different Boundary Conditions and Step Discontinuities
Underground coal gasification (UCG) is a promising clean coal technology to convert unmineable and deep coal reserves into syngas, which can be used in many industrial applications. The planning commission of Pakistan has initiated UCG Project Thar (UPT) in Block V of Thar coal fields located in Sindh. In a UCG field, real time monitoring of the hydrological and geological conditions like water influx rate, cavity growth and its interaction with overburden is a challenging task. Similarly, the development of a control system for a UCG field is also a formidable task due to numerous challenges, such as lack of instrumentation, installation of sensors at different locations, underground disturbances, process nonlinearities, and lack of direct control over the process parameters. This research work deals with the cavity prediction and the design of a multi-variable control system for the UCG field.
For this purpose, a 3D axisymmetric cavity simulation model (CAVSIM) is parameterized with operating conditions of UPT and properties of Lignite B coal of Thar coal fields. For model validation, a comparison has been made between simulated and the UPT field data for the composition and heating value of syngas. The results of CAVSIM are also compared with our previous ID packed bed model, which show the superiority of CAVSIM model. Moreover, a comprehensive simulation study has been carried out to predict the cavity growth and its interaction with overburden. The effect of operating parameters of UPT on volumetric cavity growth and heating value of syngas are also investigated.
The proposed research work highlights the significance of a model-based multivariable control system for the UCG field in general, and particularly for the UPT field. However, the CAVSIM can not be employed directly to design the modelbased control system due to its its complex and multidimensional dynamics. Thus, a simple multi-variable linear model is identified by employing the subspace-based system identification (N4SID) technique. The linear model is then employed to design the multi-variable linear and nonlinear robust control techniques. In the linear approach, an H∞ controller is designed using the S/KS method. The control >problem is formulated by using the standard approach. Moreover, the nonlinear robust control is designed by employing the sliding mode control (SMC) technique. The regular form of the linear model is formulated to design the conventional SMC and dynamic sliding mode control (DSMC). The stability of zero dynamics is shown on the approximate model of the CAVSIM. The designed controllers are implemented on the CAVSIM, and the simulation results of both the linear and nonlinear robust control techniques have been compared. It is observed that each controller has achieved the robust stability and performance in the presence of modeling inaccuracies and external disturbance. However, the performance of H∞ deteriorates when operated outside the operating range of the linear model. While the chattering is prominent for the SMC, whereas in case of DSMC the chattering is significantly reduced due to continuous control inputs. The DSMC has also consumed lesser control energy as compared to the SMC to achieve the desired objectives.