Process control and instrumentation methods for biomass fluidized bed gasifier operation

Abstract

A fluidized bed gasification (FBG) pilot plant was designed and constructed at the University of Saskatchewan Chemical Engineering Department Fluidization Laboratory. FBG is a thermo-chemical method for converting solid biomass to a gaseous fuel, termed syngas. Several instrumentation and control issues were particularly challenging with this pilot plant, including development of the fuel feeding system, pressure measurement of high temperature fluids, and metering of steam as a process reactant.

The fuel feeding system was tested using MBM (meat and bone meal) to determine the output rate stability, and predictability and measurability of the system as the components in the fuel feeding system were integrated. The fuel feeding system that was tested included a 150 mm primary metering screw conveyor, a 150 mm rotary airlock, and a 50 mm secondary injection screw conveyor. Each component of the system was fitted with a 3-phase electric motor and a variable speed drive to allow for a variable output rate. The weighing system that was integral to the metering conveyor was tested as well, but upon pressurizing the metering conveyor and hopper, the weighing system sustained an unreasonable amount of noise. Integrating a pneumatic injection nozzle with the injection conveyor was found to work effectively both under ambient temperatures and hot FBG conditions up to 725oC. Above 725oC, it was found that the test fuel would char and coat the nozzle, causing it to plug. Testing of the feeding system with the injection nozzle removed illustrated that the system could work well without it. It was determined that the injection conveyor speed to metering conveyor speed ratio that should be used for this system was 1:110 for absolute rotational speeds, or 1:1 of the full conveyor speeds. The complete system, including the injection nozzle, was analyzed and determined to produce a fuel output rate (FS) for % speeds from 5-25%, which roughly corresponded to the desired plant fuel feed rate of 1-5 g/s.

Techniques for remote pressure measurement of fluidized beds were examined as well, including the use of long tubes to cool hot gases and filters for blocking solid particles. The pressure measurement delay of these techniques was examined in comparison to a direct local measurement. This was conducted by comparing the pressure readings from two identical sensors; one mounted directly to a manifold, and the other mounted via a variable assembly (comprised of a variable length of 6.35 mm (1/4") PE tubing and a porous plate filter). Assemblies without a porous plate were found to have a minimal delay of up to 0.303 seconds for 30 m length of PE impulse tubing. More significant delays were found for systems using both a 10 media grade porous plate filter and impulse tubing; a 3 m tube length with filter has a delay of up to 0.221 s, and a 30 m impulse tube combined with the filter has a measurement delay of up to 1.915 s, a significant delay in cases where high-frequency analysis of pressure is used for bed agglomeration prediction, or systems where fast response is required to changing pressure conditions.

Additionally, a steam flow measurement system using an orifice plate and differential pressure sensor was installed and calibrated. By collecting time-based steam samples and process data, the physical system coefficients were determined for this system, allowing for steam flow measurement, accurate within 5% over a flow range of 0.5 to 2.0 g/s.