Model-based online estimation of the product gas water content and the product gas mass flow in a dual fluidized bed steam gasifier using the coarse product gas cleaning section

Abstract

Replacement of fossil fuels and base chemicals by sustainable alternatives is one of today’s greatest challenges for economy and society. Dual fluidized bed steam gasification with downstream synthesis processes is a promising technology to generate versatile energy carriers and platform chemicals from a broad variety of feedstock. However, an economically viable operation of such process chains in an industrial scale poses a challenge. A key limitation is the lack of reliable, and real-time available information on process and product quantities, which are required for an efficient plant operation. In particular, the product gas water content and mass flow are essential, but difficult to determine reliably in real-time in an industrial setting. As a solution, typically available measurements, first principle models of the coarse product gas cleaning section, and an unscented Kalman filter as estimation algorithm are combined in two soft sensors to estimate the product gas water content and mass flow in real-time in this work. Experimental data obtained from a 1MW demonstration-scale dual fluidized bed steam gasification plant showed that the proposed method can estimate the product gas water content online with a mean estimation error of −1.82 vol.%, qualifying the soft sensor for monitoring and control applications in industrial plants. The soft sensor yielding the product gas mass flow could provide a more reliable estimate than the typically installed orifice plate by adapting the orifice plate’s parameter online. Furthermore, pollution of the orifice plate could be detected by the soft sensor, qualifying it for predictive maintenance.