Logo image
Online estimation of crystal size distribution (CSD) within industrial gibbsite precipitation plants
Journal article   Peer reviewed

Online estimation of crystal size distribution (CSD) within industrial gibbsite precipitation plants

J.K. Hurst, P.A. Bahri and A. Nooraii
Computer Aided Chemical Engineering, Vol.29, pp.1638-1642
2011
url
Link to Published Version *Subscription may be requiredView

Abstract

Measurements of Crystal Size Distribution (CSD) within industrial gibbsite precipitation plants are difficult and generally not available online. Data from laboratory sampling provides infrequent measurements of CSD that is not frequent enough for online control purposes. More commonly available physical measurements include temperature and density of individual precipitation vessels, mass flow rates of seed crystals, and feed conditions such as alumina concentration, caustic concentration, and volumetric flow. A neural network based estimator is constructed from data generated from a first principles model of an industrial gibbsite precipitation circuit. The inputs to the estimator include readily available online measurements along with simulated laboratory sampled CSD measurements. The estimator predicts characteristic information about number, surface and mass distributions. The neural network estimator is employed at key locations in a typical industrial gibbsite precipitation circuit to provide estimation of fine seed and coarse seed CSD, final precipitator output CSD, and classified product CSD. These estimations can be readily used in a control scheme that aims to improve plant yield and product CSD.

Details

Metrics

40 Record Views
Logo image