-
Study Group Problem 1: Diffuser optimum imbibition
Industry: Sugar Cane Processing
Industry Representative: Richard Loubser, Sugar Milling Research Institute NPC, c/o University of KwaZulu-Natal, Durban.
Moderator:
Problem Statement:
Typical sugar cane consists of 70% water, 15% insoluble material referred to as fibre and 15% dissolved solids. What is reported as fibre is mostly made up of fibrous material from the cane stalk and leaves, but also includes small amounts of soil and other insoluble components.
The dissolved solids are mostly made up of sucrose with smaller amounts of other components such as monosaccharides and inorganic salts.
The dissolved solids are extracted from the plant and concentrated by boiling off water to produce a syrup. Crystal sugar is produced during additional concentration steps in a process called pan boiling.
Before sugar-bearing juice can be extracted from the cane, cane stalks are prepared by shredding with a hammer-mill shredder. This reduces the cane to short fibres where the sap or juice within the cells is exposed. This juice contains the sucrose. The extraction process involves washing the sugar-containing juice out of the shredded cane. The most common type of extraction unit in South Africa is a diffuser, which is a staged counter-current leaching process. Additional water, called imbibition, is added to facilitate washing. The extraction process has two mechanisms: free juice is washed off the prepared cane in each stage of the diffuser by the liquid that passes over the fibres, and diffusion of sucrose through the walls of unbroken cells within the cane fibre. The free juice is much easier to separate from the fibre and dominates the extraction process. These processes are driven by the difference in concentration of juice attached to the fibre and the juice percolating through the bed as well as the contact between the flowing juice and the fibre.
The capacity of the extraction plant is considered to be limited by fibre throughput. Consequently, the amount of imbibition added is calculated and reported as a percentage of the fibre.
Any water added to the process needs to be evaporated before the sugar can be crystallised. This evaporation requires energy that comes from burning fuel in a boiler. Much of the energy is supplied by burning the spent fibre after sucrose has been extracted, known as bagasse. If there is insufficient bagasse, the fuel has to be supplemented with coal. Burning of coal adds to the production cost so the trend has been to reduce the amount of imbibition used so as to control energy costs.
Common wisdom suggests that reduction in the amount of added imbibition water results in reduced sucrose recovery efficiency, reported as extraction. Figure 1 shows how the reduction of imbibition has been associated with reduced extraction. The question is whether this is the consequence of a fundamental dependence of extraction on imbibition rate, or whether changes to diffuser operational parameters or configuration could allow high extraction efficiency at reduced imbibition rates.
Figure 1: Sucrose-based Extraction, [1]Corrected Reduced Extraction (CRE) and Imbibition % Fibre in South Africa for ten seasons.
The diffuser is divided into around 14 stages with prepared cane moving from stage 1 to stage 14. Imbibition is applied by pouring water over the cane at the start of the last stage. Counter current flow is achieved by collecting juice in a tray below the prepared cane bed in each stage and pumping it to a spray above the bed in advance of where it was collected. The spray is positioned so that the majority of the juice will percolate through the bed and fall into the previous stage. Some of the juice, however, returns to the same tray. This is known as recycle. The effect of the recycle is to increase the amount of water held up in the stage, i.e. the mass of juice resident in that stage at a particular point in time. Greater juice hold-up results in increased juice percolation rates (volume of juice per area of cane bed), a property that has been experimentally shown to influence extraction efficiency. The recycle of juice, however, may increase the amount of dissolved sugar in the juice within a stage, relative to a condition with less recycle. This may compromise the efficiency of sucrose extraction from the cane. Achieving optimum extraction therefore depends on a balance between factors such as quantity of juice hold-up and sucrose concentrations in that juice.
We are looking for an indication whether reducing imbibition will inevitably result in a penalty in extraction efficiency, or whether adjustments can be made to parameters, such as spray position (or the bed speed/height relationship), that will compensate for the reduction in imbibition.
[1] Extraction value adjusted to remove effects of cane quality
Presentation
Study Group Problem 1 Diffuser optimisation Presentation
Report-back Presentation
Study Group Problem 1 Report-back Presentation
-
Study Group Problem 3: Scaling up from experiment to industry
Industry: Carbon Capture
Industry Representatives:
Alba Cabrera: Laboratori d'Enginyeria Química i Ambiental, Girona, Spain
Tim Myers: Centre de Recerca Matemàtica, Barcelona, Spain
Moderator:
Problem Statement
When developing new processes which are to be employed on a large scale it is much more cost effective to test on small scale devices and subsequently scale. However, simply scaling dimensions seldom leads to the same behaviour predicted by the test device. Consequently the process of scale-up is non-trivial and often contentious.
For this investigation we will focus on a specific configuration, where a fluid is passed through a cylindrical column containing a porous adsorbent material. Column adsorption is employed in a variety of areas including carbon capture and the removal of many environmental contaminants.
Fig. 1 Industrial adsorption columns (taken from https://www.aiche.org/academy/webinars /pressure-swing-adsorption-ubiquitous-gas-separation-technology)
Fig. 2 Laboratory scale experiments.
Figure 1 shows a typical set of industrial adsorption columns, with dimensions of the order metres. Figure 2 shows a typical laboratory test, where the tube diameter is around 1.5cm.
Industrial columns typical contain pellets of adsorbent material, in Figure 3 we show pellets of activated carbon with a diameter of 4mm. For experiments these are ground into a powder with micron dimensions, as shown in Fig. 4.
Fig. 3 Pellets of activated carbon, diameter 4mm.
Fig. 4 Ground up pellets, with diameter on the micron scale.
The mathematical model for column adsorption is relatively straightforward, involving an advection-diffusion equation coupled to a linear mass sink which represents the adsorption. The question for the study group will be to define parameters, such as flow regime, particle size, column dimensions, and examine wall effects, which will allow results from small scale tests to guide and understand the large scale process. Alternatively, how can the tests be related to the industrial set-up?
Presentation
Study Group Problem 3 Presentation Industrial Maths
Report-back Presentation
Study Group Problem 3 Report-back Presentation
Supporting Material
