Segregation is due to a number of different reasons but the major contributor is the way the material enters and leaves the silo. In normal segregation the course particles end up on the sides and fine particles in the middle. In your case it is the opposite, this is due to inverse segregation which is related to pneumatic conveying. Inlet particle velocity normally gets the blame for it and it can be reduced by reducing the velocity of the particles entering the silo by using a stepped pipe. Cascade plates could be mounted in the silo but it will significantly increases particle breakage. ■

Be careful of the application of multiple tube blending silos for the use of different particle sizes. Typically, this type of silo is used for plastic pellets where particle size is uniform and the chemical composition (melt index) is different and a uniform mixture is required.

The problem with different particle sizes is that they often block the openings in the collecting tubes and render them useless. The structural loading inside a blending silo is often quite aggressive and with different particle sizes, the flow properties may be different enough to amplify that problem.

The use of a "target type" inlet box on the silo or, if there is adequate room on the silo roof, using multiple discharge points into the silo may be a better way to get more uniform distribution. ■

It may be initially observed that fines will be created during the lean phase conveying of granular sugar so, whatever the initial condition, there will be a greater percentage of fines in the silo than when the sugar started its journey. A second point is that the radial pattern of segregation described as ‘normal’, where the larger components congregate at the periphery of a pile, is usually associated with a single-point, gravity-fill stream, although the reverse effect can be caused by ‘impact penetration’ if there is a high proportion of fines in the material sufficient to form a dilute bed. A tangential input from a pneumatic conveyor can give this same pattern of coarse fractions at the centre of the bin.

It is vital to pay experienced attention to detail with any anti-segregation technique because the flow processes involved are phenomenological and sensitive to the physical properties of the bulk material and the geometry of the flow regime concerned. The consequences of segregation can be aggravated if the silo is re-filled before being completely emptied; and even more so if the silo is being filled and emptied at the same time. My book, ‘User Guide to Segregation’, published by The British Materials Handling Board, describes the various mechanisms and processes that take place and ways to deal with some of the problems encountered.

There are techniques for countering segregation during the filling of a container, but these are fraught with hazards and may even exacerbate the situation. Given an understanding of the general nature of dispersion pattern occurring during filling it is generally better to re-mix the contents during discharge. Mass Flow does this, to some extent, but if the hopper is not made to this design it is a specialised business to convert the flow pattern from Non-Mass Flow to one that does. It is rarely possible to make this change by fitting low friction liners or changing the surface contact conditions, although this can be done in very marginal cases. A partial solution of converting to an ‘Expanded Flow’ pattern may also be feasible in some limited instances. More professional conversions involve proprietary inserts systems, such as the ‘Cone-in-Cone’, ‘Binsert, ‘Bates’ type or ‘Lynflow’ inserts.

An alternative method is to employ ‘Tributary Collection’ inserts that, as implied by the name, create multiple flow channels within the bulk to draw product from various regions of the stored cross section. It should be emphasised that these combination inserts in the base region of the silo materially differ from variable length tubes that collect from different heights in the stored product. The tube method, apart from the drawbacks described by a previous contributor, does not address in depth the fundamental pattern of a radial segregated distribution in filling and is more appropriate for mixing a feed stream that varies in condition during different periods of filling the silo.

If the user would care to send particulars of the silo and operating details to me via lyn@ajax.co.uk, I would be pleased to review the options and discuss possible ways forward without obligation.

Lyn Bates ■

From your posting on Bulk Online I would like to get in touch with you to offer you our products. Please email me directly on geoty@bezeqint.net so that I can introduce you to our company and show you ways to obtain top quality products while saving you money.

Regards. ■