Mathematical Modeling for Testing and Scale-up

Mathematical modeling of a fluidized multi-bed dryer is relatively easy.  A simple heat-and-material balance over each bed allows prediction of the bed-temperature.  The only assumption required is the humidity leaving the bed.  This can be estimated adequately from a little experience and feedback from a standard drying curve of moisture-content versus time.

The drying curve for wood chips can vary considerably with chip size and wood type, but the process to establish the curve is relatively easy.  Chips with measured moisture content are dried in a simple fluidized bed for a measured amount of time and then checked again for moisture content.  Several runs are necessary to get points at different times.  The air rate must be sufficient to remove the moisture as quickly as it is released from the chips. Equations to represent the drying curves are easily created using the Solver function in MS Excel so they can then be used in dryer model.

The upper limit of water removal with a given amount of dryer gas is to saturate it to 100% RH.  At the beginning of the dryer, liquid water is available on the chip surface, or readily available from the pores, and removal is fast, but the rate drops off considerably as the wood dries and diffusion of water vapor from within the chips becomes limiting. This pattern allows very hot dryer gas (up to 500F) to be used at the front end of the dryer without overheating the chips, but dictates much lower temperatures (200F) at the back end of the dryer to prevent “blue haze”.

The heat balance modeling is easily performed in MS Excel using the Goal-Seek feature to close the off-gas moisture assumption.  The calculation is performed sequentially from the wet end of the dryer, so the wood temperature to the next bed is known.  This ability to predict the performance of a unit with limited data is a major advantage of this technology.