FBA Issue 22: September / October 2008
Fundamentals of cassava pelleting, Part 3


by KOH Hui Ming



Having dealt with the unique issues created by cassava during the harvesting, drying, cutting and preparatory stages, we now go into detail with respect to its formation into pellets, focusing on how pre-conditioning, die configuration, peletting and cooling techniques affect pellet quality. Below we go into detail about how each of these factors can be used to tame this otherwise promising root's inherently large biochemical variability.


(1) Pre-conditioning

Preconditioning involves conditioning the cassava chips and mash into the most acceptable condition to form hard, smooth feed pellets. Traditionally this requires the addition of live steam to the feed stock to raise both its moisture and temperature to a point where optimal pellet quality can be attained.


When we raise the temperature and moisture by adding live steam the pellet hardness increases. In other words gelatinizations of the starch cells take place. Unit throughput also increases as the steam does some of the work done by the pellet mill under a no conditioning situation.


Moreover, raising the cassava mash's temperature in the presence of moisture explodes or breaks open a percentage of starch cells while at the same time decomposes some of the starch into simpler sugars. This set of reactions induces a state called gelatinization and as a result of it, the cassava paste when compressed, cooled and dried forms a good hard pellet.


Generally, as the temperature of the feed mash entering the die chamber increases from 45ËšC to 80ËšC, pellet hardness will increase from 13 kg to 18.5 kg if other conditions remain constant. However, there appears to be a limit to the amount that cassava can be conditioned in this manner. When pre-die temperatures reach 65ËšC, cassava pellets start  becoming too sticky to touch. This is due to cassava's high proportion of gelatinised starch cells and or their subsequent breakdown of starch to sugar. The temperature at which this will occur will vary depending on cassava chip quality.


The other limitation with cassava pre-conditioning is that when the temperature, moisture content, gelatinisation and simplification of starch to sugar increases beyond a certain point, cassava pellets acquire a rough texture this eventually produces a Pineapple type surface. This effect appears to be a side effect of cassava's tendency to become stickier as heat and moisture increase.


In such a circumstance, cassava mash passing through the die tends to adhere to the die's metal surface. In extreme circumstances, instead of being fully pelleted, cassava forms a layer within the die hole surfaces. The pelleting machine eventually jams up and the inferior feed pellets formed within the layered die walls lead to disastrous results.


Use of steam preconditioning also is subject to diminishing returns. According to standard estimates, total energy consumption is almost zero per ton from a situation of no steam to an addition of approximately 2 percent to raise the temperature of Chips into the die by approximately 28ËšC.


Once this figure is passed, there is a steady increase in the total energy required, that is steam and mechanical energy. Because of this, the most economical level of preconditioning must be calculated. Essentially, the operator faces a trade-off and must consider the cost of steam versus mechanical energy wear and tear on his die and pellet mill.


Yet, one of the biggest problems arising from traditional live steam preconditioning is that the critical moisture level of approximately 16 to 18 percent (depending on the chip quality) is rapidly reached - and thereafter, rough pellets are produced. Indeed, this preconditioning challenge brings us back to the issue of initial feedstock quality. Traditional, sun dried cassava is known for very large variations in moisture content, with moisture being the major cause of rough pellets and gummed up dies.


To remedy this, we suggest using a much larger conditioner that is steam jacketed as well as having live steam injection. Using this method, much better control of moisture and temperature can be obtained. This gives the mill operator a better chance of obtaining optimum levels of both moisture and temperature. In this way, pre-conditioning can be leveraged to optimize pellet quality.


Having addressed preconditioning issues, we now move on to the question of optimal cassava pellet die design.
(2) Die Configuration

The traditional pellet mill die shown on Table 2 can be varied in many different ways to effect pellet quality and be adapted to a great variety of different feed raw materials. At the same time, since dies are expensive and relatively hard to change, many compromises must be made when adapting a particular die for a given job.


Pellet hardness is directly related to the amount of energy that is put into the product in the form of steam heat. Frictional heat, compression pressure energy and length of time the product is subjected to heat can be modified. Indeed, almost everything except the preconditioning heat can be varied by changes in the die configuration. Below are the key factors which can be adapted to optimize the production of cassava pellets.


Inlet width (d), hole diameter (D) & friction length (l)

(Note: Please refer to table 2 when referring to various labeled die components)


We begin adapting a mill to the making of cassava pellets from within the die itself. First, we can alter the diameter of the inlet "D" in relation to the pellet size. Compression energy on the raw material is proportional to inlet width - the wider the inlet, the greater the compression energy.


Generally, a wide inlet is usually used when pelleting products that have a very low bulk density and open configuration such as straw.  It is not really applicable to cassava as the density of chips to pellets is only 1.29 to 1 as against 3.5 to 1 in the case of straw. A normal cassava die only has a short counter sink to ease the material into the hole. Hence, a mill should prepare for cassava pelleting by eliminating the inlet's additional width all together.


Second, we come to the issue of hole diameter itself (d). In general terms, it can be stated that the smaller the diameter of a pellet the better the quality. This is because the smaller the pellet the greater the surface area that is in contact with the die wall and therefore case hardened. This effect of surface area is ultimately related to the heat transfer from the die to the pellet. The traditional diameter for cassava is 8 mm. This is certainly a good compromise size and larger sized pellets are not recommended.


Third, effective friction length (l) is a pathway where newly formed pellets get frictional heat and case hardening. In the case of cassava, this is probably the most important section of the die. The longer it is, the harder the pellet. However, there are limitations, dependent on, in main, the quality and moisture content of the chips.


The fresher the chips, the longer the length should be and conversely the higher the number in visual classification Table One, the shorter the length. It is in this part of the die that the majority of the chemical changes take place. Surface temperature and pressures at this point of the die are very high and the conversion of starch to sugar can be rapid.


As sugars are very heat sensitive compounds, they react vigorously to heat and pressure. Over a short time, this may resulting in a dramatic increase in energy requirements to push the product through the die. That is basically why we must reduce compression ratios for older lower grade chips.


This is also the point in the die where pellets can and do turn brown, particularly on the surface. The brown colour is due to the caramelization of impure sugars. At this point, it would be appropriate to say that the black tarry substance found on the inside of pellet mills is in fact sucrose one of the disaccharides. This again is proof that this decomposition of starch into sugar does in fact take place under pelleting conditions.


On dies, without a compression inlet, as cassava dies are, the die compression ratio is calculated by dividing the effective length by the diameter: e.g. l /d. For example if die diameter (D) is 8mm and length (l) is 50mm, this would give a die compression ratio of 50mm/8mm, which equals 6.25 to 1.


The effective working length of the die holes can be varied by either (T), the total thickness of the die or by counterboring the die from the outside so as to reduce the pressure on the pellet in the counterbore.


Counterboring a die is mainly used when die thickness has to be reduced to such an extent that there is insufficient metal to withstand the pressures on the die due to pelleting. The disadvantage of the counterbore method is that shooters are produced - pellets break inside the die at the point where the counterbore starts and this results in inconsistent pellet length.


In most cases, counterboring is used to reduce compression ratios on cassava dies, as the initial thickness is at the point where die thickness cannot be reduced any further without it cracking or breaking.


(3) Die configuration issues, optimising quality


There are a number of other ways in which die configuration can affect both pellet harness and machine throughput. For instance:


(a)    The type of metal used and its surface finish will effect the frictional ratio between die and pellet, thereby determining pellet quality.
(b)   The type of steel hardening method used, case or through hardening and also the many variations in the degree of hardening.

(c)    Die hole pattern will affect throughput more than pellet quality. In addition, over time as a die wears out, the compression ratio falls to the point where either the die breaks or soft pellets are produced.


All this leads to the question of 'What is therefore the best die configuration for good quality cassava pellet production?'


First and foremost, the die must be made of good quality alloy steel that has been case hardened and this remains true regardless of chip quality. However, due to wide variations in raw cassava quality, it is impossible to produce a single, ideal die configuration that is suitable for all circumstances. 
The above are excerpts, full versions are only available in FEED Business Asia. For subscriptions enquiries, e-mail membership@efeedlink.com