1	The Science of Separation The Elimination of Extraneous Material From Sugarcane Billets
2	Billet Harvester Since 1995, there has been a major shift in the Louisiana sugarcane industry away from whole- stalk harvesting. Today more than 90% of the cane in Louisiana is harvested as billets. A billet harvester cuts the cane into pieces ranging from 6 to 14 inches in length and discharges it into a trailer moving alongside.
3	Cameco
4	Inorganic Debris The billet harvester knocks the cane down and bites into the surface of the row, taking along with the cane an abundance of trash such as tramp iron, sand, clay, clay balls, stones, bricks, leaves and tops. In the best case, a ton of cane contains 13% trash (8% inorganic and 5 % organic). Under wet conditions, a ton of cane may contain up to 40% trash (30% inorganic and 10% organic).
5	Loss of Recoverable Sucrose Not only does this trash not contain sugar, but as it leaves the mill in the form of filter cake or bagasse, it carries away sugar at a rate of roughly 0.75 lbs per percentage point of trash. Each percentage point of trash represents a loss in production of about 3 lbs of sugar.
6	The High Cost of Trash In the case of 13% trash, production losses exceed 16%, and in the case of 40% trash, production losses exceed 50%. More trash means more maintenance, more flocculent, more lime, more natural gas, more unburned bagasse, more filter cake, more front-end loaders, more water treatment, more settling basins, more draglines to clean out settling basins, more haulage, more inversion,
7	The High Cost of Trash more molasses, less sugar and more cost. Surely there has to be a way to maximize the recovery of sugar and minimize the cost of producing it.
8	Intrusion of Foreign Matter All mechanical harvesters whether of potatoes, beets, carrots, salsifies and peas, face the same problem of the intrusion of foreign matter that varies widely in terms of grain size and density. For over 40 years, the vegetable industry in Europe battled this problem by means of vibratory screens, rotary trommels, optical sorters, air separators and a variety of dynamic effect separators employing water.
9	Intrusion of Foreign Matter None of these technologies have worked. Likewise, we may try to screen, rinse, wash or scrub the billet, but all of these steps fall far short of producing an acceptable product. The average washing efficiency for a single deck washing table is about 53%, for a dual deck washing table, it is about 59%, and for a cane washing drum, it is about 44%. In a wet season as we have had this
10	Intrusion of Foreign Matter past year, these efficiencies would decline considerably. If a sugar mill wants to make money, produce a consistent product and eliminate the randomness associated with bad weather, there is only one process that will do the job: dense medium separation.
11	Bi-Directional DMS The bi-directional dense medium separator patented by ESR is the most accurate separator ever used in the pre-processing of root vegetables. The smallest dense medium separator handles 100 TPH of vegetables with an accuracy of separation unparalleled in the history of vegetable separation. Not only does it separate out the trash, but it can also separate a good potato from a bad potato,
12	Dense Medium Separator a good carrot from a bad carrot, a fresh carrot from a partially dehydrated carrot. It separates accurately, for example, a carrot of a 1.030 density from a carrot of a 1.033 density. Some ESR separators have been in operation for over 15 years. Most have been in operation for over 10 years. During this entire time, no one has ever established a single separation error in the processed vegetables.
13	Without Error Most ESR separators are in continuous operation, night and day, 8 months per year. Nestle operates two potato separators at their Sitpa site in Rosiere, France, every day of the year, except on holidays and weekends.
14	Potato Separator
15	15 Units Sold As many as 15 vegetable separators were sold to the largest canneries, farming co-operatives and vegetable wholesalers in Western Europe. Bonduelle, the largest vegetable processor in Western Europe, bought four vegetable separators, and they are now placing an order for a fifth. But someone might argue that even though this technology is successful in Europe, it is totally unproven with respect to sugarcane billets.
16	The Size and Shape of the Billet It terms of size and shape, the billet is no different from a large carrot or a salsify. The salsify is a long root vegetable grown in a soil that is predominately clay, and since it is harvested in the middle of winter when it rains continuously, it generally comes from the field with over 60% clay. Nothing of this clay, not a single ounce, can ever be found in the salsifies exiting an ESR separator.
17	The Density of the Billet The density of a sugarcane billet is no different from that of a sugar beet or potato. Variety 384 accounts for 85% of what is currently planted today in Louisiana. The 321, 845 and the 555 make up the rest. The 540 will be released this year. Let us look at their densities:
18	The Density of the Billet
19	Laws of Physics Since the size, shape and density of the billet appear quite normal; since metal, stone and clay all have densities twice that of the billet; since Newton’s laws continue to operate at sugar mills, nothing is unproven about ESR technology. This technology has been around for over 18 years, and anyone trained in the art of dense medium separation would find it
20	Laws of Physics impossible to explain how it would not work in sugarcane. What is dense medium separation?
21	What is Dense Medium Separation ? Here we speak of the dynamic of a quiescent bath, where floats float and sinks sink. The specific gravity of water is changed by means of fine particles in suspension. In the case of vegetable, the suspension fines, for the most part, are derived from the fine sands and clays naturally found in the harvested vegetables.
22	Bi-Directional Dense Medium Most horizontal dense medium separators on the market today are mono-directional: floats and sinks move in the same direction. The ESR dense medium separator by contrast is bi-directional: floats move in one direction, while sinks move in the opposite direction. Bi- directionality brings, as we shall see, a large number of significant advantages.
23	Cross-Sectional View
24	Four Critical Steps There are four critical steps in making a good dense medium separation:
25	1) Correct Injection In the ESR process, the material drops from a conveyor belt into an injector filled with a fast- moving stream of medium. The medium plus solids enter the separation zone, where they are well distributed over a broad 3-dimensional plane. This prevents floats, at the critical moment of entry into the separator, from being buried or entrapped with sinks.
26	2) Good Medium Next: we need a stable yet non-viscous medium throughout bath. If the fines drop out of suspension, this leaves water at the top of the bath and a high density sludge at the bottom of the bath. In this situation, obviously no separation takes place. In billet separation, we do not have to worry about stability, since clay constitutes a large component of the inorganic debris.
27	2) A Good Medium But since there is so much clay accompanying the billets, we should be concerned about viscosity. If the medium becomes too viscous, then it is impossible for particles to float or sink as rapidly as they should, and, once again, a bad separation takes place. That is why we foresee a series of classifying cyclones that continually reject clay but retain fine sand between 15 and 100 microns.
28	Classifying Cyclones
29	Fine Silt/Sand ESR has mastered the art of working with fines naturally found in the material to be separated. In vegetable separation, salt in solution is commonly used, and in automobile shredder residue, a very expensive magnetite powder is commonly used. ESR pioneered the use of fine sand or silt to reach separating densities as high as 1.6 RD.
30	Organic Contaminates The knives of the billet harvester generate large quantities of organic fines. These fines can easily compromise the fluidity of the separating medium and must be removed from the circuit as quickly as they enter. At each pass through the separating drum, it is absolutely essential that these organic fines be filtered out of the medium. A large rotary trommel screen, fitted with stainless steel wedge
31	An Ideal Medium wire panels, is ideal in this regard. Once the billets are injected into a drum filled with a stable, non-viscous medium free of organic contaminates, the drum must offer, under quiescent and tranquil conditions, sufficient space and time for a separation to take place.
32	3) Floats Dynamic A particle must have adequate time to float or sink. The length of the separation zone, therefore, is critical in assuring adequate residence time. The rise rate or sink rate of a near-gravity particle can be very low, and if it exits the separation zone before it has had a chance to float or sink, then it easily reports in the wrong direction. In contrast to many dense medium manufacturers,
33	3) Floats Dynamic ESR put forward the principle or rule that the length of the separation zone must be twice that of the barrel diameter. Most dense medium drum manufacturers design drums of a large diameter but short length. ESR reversed this trend, offering a much smaller diameter relative to barrel length. Only in this way do we have a near-perfect floats dynamic.
34	3) Floats Dynamic In order to minimize turbulence in the separation zone, there should be no paddles moving floats along, as we see in the classical Drewboy separator. There should be no lifting of sinks within the separation zone, as we see in the conventional Wemco separator. There should be no curtains within the separation zone, as in the Wemco separator. The Wemco curtains, in
35	3) Floats Dynamic conjunction with the lifters passing underneath, create a great deal of turbulence, that, all too often, completely destroys the accuracy of separation. Since Wemco lifts and evacuates sinks while they are in the separation zone, Wemco is obliged to put a sinks evacuation chute within the separation zone. Since Wemco lifts and evacuates sinks while they are in the separation zone,
36	3) Floats Dynamic Wemco cannot rotate the separation drum more than 3 or 4 RPM. If the Wemco drum were to be rotated any faster, the turbulence created between the lifters and curtains would totally destroy the accuracy of separation. Since the inside of the Wemco separator is completely filled with lifters, curtains and chutes, the operator can see nothing of what is actually taking place within the
37	3) Floats Dynamic separator. For example, if a curtain should break or tear, the operator sees nothing. But even more problematic is the fact the large width of Wemco separator is not actually used, since all separation is confined to the narrow zone between the two curtains running the full length of the separation zone.
38	3) Floats Dynamic In contrast, the ESR separator has no lifters, curtains or chutes within the separation zone. All is clearly visible and transparent to the operator. The full surface of the bath is available for separation. An 8-foot diameter ESR drum can handle up to 200 TPH of floats, while an 11-foot drum can handle up to 400 TPH of floats.
39	3) Floats Dynamic Since, in the ESR drum, sinks are given adequate space and time to sink, there is very little danger of finding sinks in floats. But how about floats in sinks? In the ESR drum, the curtain to prevent floats from reporting with sinks is situated completely outside the separation zone.
40	4) Sinks Dynamic Sinks are lifted up by means of a scrolled cone, but only when completely outside of the separation zone. This assures an enormous sinks evacuation capacity. The ESR drum can be rotated at 20 RPM if necessary, with no danger of floats reporting with sinks. An 8ft barrel can evacuate more than 50 TPH of sinks, while an 11ft barrel can evacuate more than 150 TPH of sinks.
41	US Patent 5,373,946 It is this idea of placing the curtain completely outside the separation zone that forms the heart and core of the ESR patent. To accomplish this, it was necessary to increase the diameter of the sinks evacuation cone relative to the diameter of the mid-section of the drum.
42	Top View
43	Side View
44	High Capacity Since the ESR drum is bi-directional, dewatering and rinsing devices can be placed on both sides, easily handling a large volume of either floats or sinks. ESR no longer uses vibratory dewatering screens. Large rotary screens fitted with stainless steel wedge wire panels are ideal, since they are robust and self-cleaning. Let us review once again the dynamics of the ESR drum.
45	Side View
46	Chaparral Steel 10 ft Diameter Drum
47	100 TPH Billet Separator
48	200 TPH Billet Separator
49	400 TPH Billet Separator
50	Largest and Smallest
51	The Billet in Water Perhaps the biggest objection to the dense medium separation of sugarcane billets concerns the loss of sugar associated with putting them in water. The washing of whole stalk cane had been in practice in Louisiana for about 50 years. But since the harvesting of a whole stalk involves but one or two cuts, the loss of sugar here does not exceed about one pound of sugar/TC.
52	Sugar Losses There are four factors that determine the loss of sugar with respect to billeted cane: the number of harvester cuts (billet length), the precision of the harvester cut (clean or frayed), the forces applied to the billet (agitation/tumbling), the amount of time the billet is subjected to these forces.
53	Short - Long Billets In general, it is fairly safe to say that if all other factors are the same, twice the number of harvester cuts will yield twice the loss of sugar. In the study by Birkett and Stein, we see that short billets of an average length of 5.25 inches lose twice as much sugar as long billets of an average length of 9.55 inches (see Cane Washing Losses, pp. 3 & 4).
54	The Precision of the Cut If the billet harvester cutting blades are not sharp, they will damage and fray the ends of the billet. The same study by Birkett and Stein shows that sugar losses increase almost threefold if the billet is damaged (see p. 3). Therefore it is very important to change “the cutter blades on the combine harvester at the frequency recommended by the harvester manufacturer” (p.6).
55	The Forces Applied to the Billet In vegetable separation, the handling of the carrot or salsify within the dense medium separation process is a very important issue. If, for example, a carrot or salsify is broken, it has little or no value. Several studies conducted in Europe have shown that the carrot remains undamaged throughout the entire DMS process. Likewise, if a billet is agitated or tumbled, either before or during the cleaning
56	The Handling of the Billet process, it will lose sugar and drop in value. That is why the scrubbing of the billet, as in a typical wash drum, is far from ideal. It would take at least 5 minutes of violent scrubbing to break down the large clay balls that often accompany the billets. But a retention time of five minutes would require a scrub drum of an enormous length, and during this time, far too much sugar would be lost.
57	The Importance of Time The study by Birkett and Stein shows that if billets are placed in a cane preparation index apparatus and rotated at 19 RPM, a lot of sugar is released by this tumbling action. The longer the billet is tumbled, the greater the loss of sugar. In the first 30 seconds, we see a loss of about one pound of sugar/TC, while after 15 minutes, this loss increases to almost 8 lbs/TC (see ibid., Figure 15).
58	The Handling of the Billet It would appear that the simple act of putting the billet in water releases very little sugar from within the billet. To determine the density of the billet, I once spent an entire day at the USDA lab in Houma weighing billets underwater. The billets were left in a tranquil bath of water for several hours at a time, and throughout the entire day, the water used in this test was never changed.
59	The Handling of the Billet At the end of the day, the amount of sugar in this water was indistinguishable from the amount of sugar present in the original tap water used for this density test. What can we learn from all of this? Water does not diffuse sugar out of a billet. Since the cell membrane of cane is resistant to osmotic effects, the cell wall must be ruptured and force must be applied to the wall of the billet.
60	What Can We Learn? Once the billet is cut, every effort should be made to minimize all aspects of its handling. From field to factory, it should be loaded and unloaded but one time. It should never be dumped on a concrete slab or handled by a wheel loader. In the cleaning process, it should not be scrubbed, agitated or tumbled in any way, and it should not remain in a separator for more than about 10 seconds.
61	Loss of Sugar If we do all of the above, then we have every reason to expect a sugar loss of less than 3 lbs/TC. This is roughly equivalent to the loss of production associated with only one per cent of trash/TC. The sugarcane industry in Louisiana has a lot to learn from the vegetable industry in Europe. If it handles billets with the same respect that the vegetable industry in Europe handles carrots, and if it
62	European Model separates billets with the same technology that Europe employs to separate carrots, then most of the problems created by the introduction of billet harvester in Louisiana are easily solved.
63	The Billet in Water Putting a billet in water will not cause sugar to come out of a billet. Putting a billet in water simply shows us how much we have mishandled it prior to putting it in water. The extremely gentle action of a dense medium separator can never be a major factor in sugar loss. Perhaps we need to adjust the length of the billet, change cutter blades more often, reduce the number of times we load and unload
64	Takes Time the billet, eliminate the use of wheel loaders, reduce the distance we allow a billet to fall, and so forth. It has taken 8 years for the billet harvester to dominate the sugarcane industry in Louisiana, and no doubt, it will take the same amount of time for this industry to correct the problems that the billet harvester has created.
65	Adhering Water When billets are put in water, a considerable amount of water adheres to the cane, and this water increases the load on the evaporators. In the Birkett & Stein study, we see that single deck tables average 4.7% adhering water by weight, dual deck tables average 11.3%, and a washing drum averages 16.3%. In this dense medium process, the outgoing cane is rinsed twice: once with
66	Adhering Water re-circulating rinse water, and then with a final clean water. It would be easy to equip the exit side of the floats trommel with lifters so as to tumble and dry the billets, but this might release a lot of sugar. We propose instead that the exit side of the trommel be fitted with a series of air knives that would remove adhering water without impacting the billet in any way.
67	Air Separation The leaves and tops of the sugarcane plant contain no sugar, and those that accompany the billet into the mill can be easily removed by means of an air separator. Since it is very important in air separation to minimize particle interference, the billets should be spread out over a broad two- dimensional plane, and the blower should be situated in-line with the flow of billets.
68	Air Separation This air separator can be situated right above the injector of the dense medium separator. After air separation, the billets should slide directly into the mouth of the injector. This would reduce the impact on the billet and minimize the loss of sugar.
69	Dual Strategy Hopefully it is clear that this dual strategy of air separation and dense medium separation can eliminate virtually all organic and inorganic trash. What mill in Louisiana has ever considered the possibility of processing billets without extraneous material? It is hard to imagine what this should mean. For the first time ever, a mill will be able to operate under constraints that are truly internal.
70	Elimination of Externalities A mill would possess a level of control in daily operations that it has never before experienced. More sugar would be recovered. Production would stabilize at a very high rate. Operating and maintenance cost would decrease considerably. Rain and even hurricanes would have minimal impact on profitability. And if we turn once again to Europe,
71	The European Model we find to our amazement, that some vegetable brokers there have even figured out how to avoid trucking extraneous material to a factory and hauling it away. Taking full advantage of this dense medium separation technology, they set up de- centralized pre-processing centers in proximity to clusters of growers, and only clean vegetables are transported to the cannery.
72	Leaving Trash Where It Belongs What would such a distribution model mean to the Louisiana sugar industry? What if relatively small dense medium separators were strategically situated so that only clean billets would be transported to the mill? Leaving trash where it belongs cannot be such a bad idea, and surely the grower is better equipped than anyone else to deal with it. Each ton that the grower delivers to the mill would be a ton of billets
73	Harvester Capacity and nothing else. A modern billet harvester in full operation has a capacity of 200 TPH, but when we look at the total number of harvesters and the total tonnage of cane harvested in a given season, we see that a harvester averages less than 20 TPH. The Louisiana cane industry has invested over $80 million dollars in harvester technology, more than twice what is demanded to meet its harvesting
74	Investment in Separators needs. At the same time, the cane industry would have to invest roughly $30 million dollars in dense medium separation technology. A 50% reduction in the number of harvesters equates to a savings of over $40 million dollars and easily covers the cost of this dense medium separation technology. The savings in the transport of trash to and from the mill would be well over $10 M dollars per season.
75	Farmers Working Together Imagine a group of cane farmers working together: with one farmer specializing in harvesting, a second in dense medium separation, and a third in transport. If the primary objective is to make money, then a collaborative effort among farmers in the harvesting, cleaning and transport of the billet would be the most logical way to proceed.
76	Conclusion Such an approach would simplify every aspect of this complex sugar-making enterprise. It would offer growers and factories a myriad of new options and new economic opportunities, and at the same time, it would secure, safeguard and stabilize one of our state’s oldest and most important industries.