(247i) Estimation of Gasifier Wall Profile Using Measurements from a Wireless Sensor Network | AIChE

(247i) Estimation of Gasifier Wall Profile Using Measurements from a Wireless Sensor Network

Authors 

Huang, Q. - Presenter, West Virginia University
Bhattacharyya, D., West Virginia University
Sabolsky, E., West Virginia University
Temperature is the most important variable in efficient operation of the entrained-flow gasifiers. In addition, slag penetration into the refractory wall of these gasifiers leads to stress buildup and spalling of the refractory bricks reducing the plant availability. An estimation of the slag penetration profile and stress profile can be very useful for monitoring, maintenance planning, and determining operating conditions for enhancing the refractory life. However, the state-of-the-art measurement technologies are inadequate for measuring any variable inside the gasifier due to the high-temperature, high-pressure gasifier environment and corrosive properties of the slag. In addition, the existing technologies need open access port in refractory lining, which provides a penetration path for slag and reduces lifetime of the refractory further. With this motivation, novel smart refractory bricks with embedded sensors are being developed at West Virginia University. However, as the slag penetrates into these smart bricks, their electrical and thermal properties change. In addition, a steep temperature gradient exists along the length of these sensors. Therefore, the existing approach where the sensor output is directly correlated to the measurement variable is inadequate for this system. In addition, slag penetration depth cannot be directly correlated to the sensor output due to complex effect that the slag penetration has on the temperature profile as well as the refractory and sensor properties. To address these issues, a model-based estimation approach has been developed. The measurements from the embedded sensors are noisy. Due to temporal change in the material properties, it is likely that the discrepancy of the process and measurement models would also evolve with time. In addition, the measurement data are available from a wireless sensor network that suffers from communication constraints, packet dropouts, and synchronization errors. Due to the limitations in the communication bandwidth, only limited packets can be transmitted to the estimation algorithm at any transmission instant. In addition, random packet dropouts can occur in wireless network caused by the unavoidable transmission error. Furthermore, the estimation algorithm may receive out-of-sequence measurements. Therefore, a modified nonlinear unscented Kalman filter (UKF) algorithm for this differential-algebraic equation (DAE) system has been developed for estimating the temperature profile, stress profile, and slag penetration profile with due consideration of the issues with the wireless sensor network. Our study shows that the existing approach can increase the refractory life considerably.