- Email: [email protected]
Challenging the “Crago” double float process II. Amine-fatty acid flotation of siliceous phosphates
Pergamoln sow2-6875(97)ooo78-2
Minrrda thgineering, Vol. 10, No. 9, pp. 983-994, 199-l 0 1997 Elsevia Scitace Ltd Printed in Great Britain. All rights resavd 08924875/97 $17.00+0.00
CHALtLENGING m “CFtAGO” DOUBLE FL,OAT PROCESS II. AMINE-FATTY ACID FLOTATION OF SILICEOUS PHOSPHATES P. ZHANG, Y. YU and M. BOGAN Florida Institute of Phosphate Research, 1855 West Main Street, Battow, FL 33830, USA (Received 14 October 1996; accepted 19 May 1997)
ABSTRACT In the conventionalphosphate flotation (Crago) process, 3&40% by weightof the silica present in the feed are floated twice,first by fatty acid, and then by amine. The Crag0 process is therefore ineficient in terms of collectorefficiency.Also, the phosphatemining industry is faced with higher fatty acid prices, lower feed grade, and stricter environmentalregulations.For these reasons, the Florida Instituteof PhosphateResearch has developed an amine-fatty acid flotationflowsheet, the reverse “Crago”. In this process, fine sands are first floated witha minimumdosage of amine condensateadded stepwise. The concentratefrom prejoat is then floated with a blend of SurfactaruYfatty acioY’e1 oil. The technical and economicfeasibility of this process were evaluated on seven feeds of varying characteristics,producing concentratesanalyzing30-32% P205 and 4-10470Ins01withPZ05 recoveries of over 93%, at totalreagent costs below $2 per ton of concentrate.This novel process could simplifythe current processingflowsheet by eliminating the acid scrubbing circuit, reducing the sizing section, and reducing the number of conditioners. The process consumes about one third to half the amount of reagents required by the Crag0 process at signijcantly improved flotation recovery. The effect of slimes (in either the flotationfeed or water) and techniquesfor reducing this eflect were also investigated.0 1997 Elsevier Science Ltd Keywords Froth flotation; flocculation; flotation reagents; industriaI minerals
INTRODUCTION In 1994, the Floridia Institnte of Phosphate Research initiated an in-house research project “A Screening Study on Phosphate Depressants for Beneficiating Florida Phosphate Ores ( FIPR #94-02-101)“. The threeyear project was designed to screen the available phosphate depressants for the dual purposes of developing alternate phosphate beneficiation processes for the currently mined siliceous phosphate deposits as well as removing dolomite from the Florida phosphate reserves of the future. As a result, two alternate processes for Florida siliceous phosphates have been developed. In these processes, fine silica is first floated with an inexpensive amine, and the prefloat concentrate is further cleaned by either floating phosphate (the reverse Crag0 process) or floating silica (the all cationic process). This paper discusses the former. The all cationic process was published in the Processing Supplement to the September 1996 issue of Industrial Minerals [ 11. hesented at MinemLrEngineering‘96, Brisbane, Australia, August 2648.19%
983
984
P. zhang et al.
Analysis of the “Crago” Double Float Process
It may be of help to briefly describe how phosphate is processed in Florida After desliming, the phosphate ore is subjected to sizing, Figure 1 [2]. Typical sizing involves using a hydrosizer to size the d&imed feed into coarse (lOOOx42O micron) and fine (420x105micron) fractions. In some more sophisticated operations as shown in Figure 2, three fractions are produced (lOOOx707,707x420,and420x105 micron).
I Coarse Feed 1000x420 Micron (16x35 Mesh)
Fine Feed 420X105 Micron (35x150 Mesh)
Fig. 1 Sizing
Spiral Feed 1OOOx707Micron (16x24 Mesh)
Fig.2 Sizing II The sized feed is first subjected to rougher flotrqtion,Figure 3. In this prucess the sized feed is dewaBed and conditioned at about 70% or higher solids with fatty acid/fuel oil at pH about 9 for thw minutes. The phosphate is then floated. It must be emphasized that a significantakount (30-4096)of silica is also floated in this step.
Challenging the Crag0 double float process
985
Fig.3 Rougher flotation The mugher conentrate goes through a dewatering cyclone, an acid scrubber, and a wash box to remove the reagents from the phosphate surfaces, Figure 4. After rinsing, the feed is transported into flotation cells where amine (sometimes with diesel) is added. The silica is floated at neutral PH.
HzSO., -
I
I
Acid&xubbex
I
I
1
Amine Flotation
1 Amine+
1
Am&Flotation
]
I
1
Concentrate
Sand Tailings
Fig.4 Cleaner flotation In the conve&onal “Double Float” (Crago) process for phosph@e minerals, 3W% by weight of the sands present in the feed are floated twice, first by fatty acid and then by amine. The Crag0 ~KK‘RSS is, therefore, inefficient in terms of collector efficiency. Fatty acid dosage for floating “pure” phosphate was found to be about 0.18 kg per ton. The theoretical dosage for floating a feed of 6.86% P,O, (15% BPL) is only 0.027
986
P. zhllg
et al.
kg/ton of feed (TOF). Actual plant fatty acid consumption for such a feed is about 0.54 kg/TOF. Therefore, plant collector efficiency is merely 5% (0.027/0.54)! The rest of the reagents are wasted prhmuily because of silica. However, there were a number of reasons for the phosphate industry to endorse the process enthusiastically: 1) fatty acid was much cheaper than amine so that anionic flotation followed by amine flotation made more economic sense than otherwise, 2) desliming was not sophisticated, leaving significant amounts of clay in the flotation feed so that amine usage would have been prohibitive had silica been floated first, and 3) the ore was high in grade, so the adsorption of fatty acid on silica was tolerable compared with that on phosphate in the rougher flotation stage. The situation is quite different today: the amine price is approximately twice that of fatty acid compared to nearly 10 times in the 1950s; the desliming technology has been upgraded to reduce the fine slimes in the flotation feed; the phosphorus content in the currentlymined phosphate ores is about half that in the past. These trends do not favor the standard Crag0 process. Understanding Amine Flotation There are two primary reasons why the Crag0 process has never been seriously challenged. First, fatty acids were inexpensive in the past. The second reason is this conventional wisdom about amine flotation: Amine flotation requires clean feed and deep well water. Therefore, one would need more sophisticated desliming devices, and utilize more deep aquifer water in order to float silica first. To some extent, this conventional wisdom is true. But, we could use our knowledge about amine flotation to overcome this problem. For example, we know that amines are more selective than fatty acids, and that amine adsorbs instantaneously on sand. The fact that amine is more selective in floating silica than fatty acid in floating phosphate is demonstrated in the following Figures. Figure 5 [31 indicates that fatty acid adsorption on silica is significant.
0
Fig.5
10
20 30 czadbhglime,mln. .
40
60
60
Kinetics of oleic acid adsorption on flotation feed and concentrate at 0.27 kg/ton oleic acid and pH 9.5
Figure 6 [4] shows that amine can float more than 99% of silica from pH 3 to 12, while phosphate flotation by amine is minimal within this pH range. Figure 7 [4] shows that at near neutral pHs, there is a large difference in zeta potential between silica and phosphate. Therefore, it is ideal to se@rate silica f&m phosphate at neutral pHs. However, neutral pH is not an ideal condition for floating phosphate osing fatty acids: That leaves on@ one o#ion: floating silica tirst.
987
Challenging the ‘Cragodouble Boat process
Flotation Recovery 110, 100 s_ 90 P 80 g 70 880 b0 g 40 930 $20 10 0 0
2
4
pH
10
8
8
12
14
of Solution
Fig.6 Flotation recovery of apatite and silica with dodecylammonium chloride at different pH
PH
8
10
12
Fig.7 Effects of pH on zeta potentials for silica and apatite Since amine adsorbs on silica very rapidly, the effect of clay on amine consumption may be reduced by adding amine stagewise. Flotation is conducted in a series of banks of flotation cells. Each bank consists of four to six cells. In the conventional process, all the amine is added as one dose in the first flotation cell. If a small amount of amine is added in the first cell, this cell not only acts as a flotation machine, but also serves as a desliming device. Since no conditioning is required, the number of conditioners currently used for ilotation may be reduced by floating silica first. Because amine flotation is conducted at neutral pH, pH modiier consumption would be significantly reduced by floating silica first. Finally and perhaps more importantly, for amine is more selective than fatty acid, collector efficiency should be dramaticallyimproved by floating silica first.
988
P. Zhang et al.
Amine-Fatty Acid Flotation (Reverse Crago) Process
Based on the above understandingand in recognition of the changed conditions, FIPR has developed a much simplified flotation flowsheet designated as “reverse Crago”, or aminefatty acid flotation process, as is shown in Figure 8.
I UnsizedFeed
I
Dewatering
1
Fatty Acid RlmwlSoda Ash r
I
I
1
1
_ -._
I-
PhosphatePlotation
Underflow Sand Tailings
1
Concentrate Product
Fig.8 Basic amine-fatty acid flotation process In this process, fine sand is first floated with a minimal consumption of amine by adding amine stepwise, so that phosphate loss could be minimized without using a depressant The concentrate from prefloat is then floated with a blend of fatty acid/fuel oilkrfactants as a phosphate collector. This new process is unique in the following aspects: 1) stepwise addition of amine, 2) novel fatty acid flotationreagent scheme that improves recovery of coarse phosphate particles, 3) higher collector efficiency, 4) simplitied flowsheet, and 5) flotation of unsized feed without sa&lcing metallurgical recovery. For feeds containing more coarse particles, two additional steps may be added: sizing the prefloat tails at 1190micron (14 mesh) to get a mini pebble, and scavenging the coarse fraction, +420 micron (+35 mesh), of the fatty acid flotation tails. Figure 9 shows the flowsheet with scavenging. The benefits of the aminsfatty acid flotationprocess are evident 1) Reduced flotation reagent consumption because only a single mineral is floated in each step; 2) Maxim&d phosphate recovery and improved concentrate grade because phosphate is in a poor position to compete with fine sands for &sorption of amine, and coarse silica is not easy to float in the second stage; 3) Reduced energy cost due to the elimination of the acid scrubbing circuit, reduction of sizing equipment and number of condlticmers;4) Simplified flowsheet since it is a non-sizing and non-scrubbing process; and 5) Minimized inorganic chemical consumption, particularly sulfuric acid.
Challenging the
1
Crag0 double float process
989
Dewatering ]
1 Fatty Acid Blend Soda Ash Fuel Oil
Fatty Acid Blend SodacAsh Fuel Oil
Sand Tailings Fig.9 Reverse Crag0 flowsbeet with scavenging fatty acid tails
EXPERIMENTAL Materials Flotation feeds
This new flowsbeet was tested on numerous flotation feeds from FIorida. Tables 1 and 2 show the basic chemical and size analyses of the feeds. All the flotation fee& tested are unsized, and were collected f&m the secondary hydrocyclone underflows. Most of the flotation experiments wei-e conducted on the as-received feeds. In a few occasions, slight desliming (rinsing with tap water) was practiced for comparison.
P. Zhng et al.
990
TABLE 1 Basic chemical compositions of flotation feeds
II
Al
II
I
9.62
I
70.60
A2
9.44
70.45
B
5.43
83.49
Cl
8.60
74.86
c2
6.77
79.75
D
I
13.82
I
60.2
TABLE 2 Size distribution of flotation feeds
II
Sample ID
II
I
Al
I
+841 pm 595-841 pm 420-595 unr (+2OIvI) I (28~2OM) I (35-28M) 1.3
I
3.3
I
105-420 pm (150-35IvI) I
10.2
83.7
I
-105 luIl (- 15OM) 1.5
A2
4.7
6.1
15.6
73.1
0.5
B
3.0
3.1
7.8
82.6
3.5
Flotation Reagents
The quartz collector used is an amine condensate designated as Custamina738 and provided by Westvaco. A blend of fatty acid, surfactants and fuel oil was used as phosphate collector. These are all commercially available reagents. Soda ash and sulfuric acid were used as pH modifiers. Bartow tap water was utilized in the initial tests, and plant water was used for studying the slime effect. Flotation Procedure
All the flotation tests were conducted in a standard one-liter Denver cell with a charge of about 500 gram dry feed. Apart from the stepwise addition of amine, all the tests followed standard flotation procedures.
RESULTS AND DISCUSSION Initial Feasibility Evaluation
Table 3 summarizes the flotation test results. In every case, flotation recovery of over 90% was achieved at reagent costs of around $1.5 per ton of concentrate. In a typical indust&lopemtiotr, flotation recovery is about 80% at reagent costs of $2.5-3 per ton.
Challenging the Crago double float process
991
TABLE 3 Performance of the Reverse Crago Process
Feed1 Al A2 B Cl c2 D
Table 4 shows the phenomenal reduction in total chemical usage by adopting the reverse Crag0 process. The first row was obtained by averaging data from two plants during a two-week period. The numbers for the reverse Crag0 process were averaged on the six feeds evaluated.
TABLE 4
IReagent consumption (kg/h) lProceS
comparisoh between the “Crago” and Reverse Crag0
1 Amine 1 Soda ash 1 H,SO,
Process
Fatty acid + fuel oil
Crag0
7.35
0.54
2.49
2.90
IReverseCrag0
1.98
1.78
1.30
0
Dealing with the Clay (Slime) Problem All the above results were generated using plant feeds with tap water, There consumption would be economically prohibitive when slimy feed and plant Therefore, more &Nation tests were conducted for parallel comparison between reverse Crag0 process using plant feeds and plant waters. Table 5 shows similar on a low-grade (35% P,O,) fine feed. TABLE 5
Flotation hocesses
performance comparison between the “Crago” and Reverse using plant water on a plant feed
Item
30.04
30.32
Concenlrate % Insol.
10.15
9.08
P,O, Recovery
Crag0
1 Crag0 Process 1 Reverse Crag0
Concentrate %P,05
%
was a concern that amine recycled water are used the Crag0 process and the metallurgical performauce
I
84.94
I
84.49
However, the reagent consumption made a big difference in terms of both total kilograms and dollars, Table 6.
P. Zbang et al.
992
TABLE 6
Reagent consumption comparison between the Crag0 and Reverse Crago Processes using plant water on a plant fine feed
Reagent
Dosage, kg/ton concentrate Crag0 Process
Reverse Crag0
Fatty acid/fuel oil
11.20
1.89
Amine
0.33
2.18
-
Polymer
0.15
Sodium silicate
1.17
-
Starch
1.87
-
Sulfuric acid
2.79
-
Sodaash
2.79
0.71
TOTAL REAGENT USAGE
20.16
4.93
$3.95
$224
TOTALREAGENTCOST
It should be understood that amine flotation is sensitive to clay in the feed and suspended solids in the water. It must also be recognized that amine is still more expensive than fatty acids. Therefore, the success of the reverse Crag0 process depends solely on viable techniques for reducing the effect of slime on amine consumption. Three techniques have been developed for this purpose: adding polymers in the flotation water, adding amine stagewise, and adding amine continuously. Figure 10 shows the effect of stagewise addition of amine. It may be seen that 0.45 kilograms of amine added stagewise was equivalent to about 0.90 kilograms of amine added as one dose.
0.6
10.5 i 0.4 y
!I 0.3
i 0.2 0.1 0 0
10
20
30
40
50
60
70
w&%ofsmdPhbd
Fig.10 Comparison of one dose amine addition with stagewise addition
993
Challenging the Crago double float process
Another very effective way of reducing the slime effect on amine consumption is to add a small amount of polymer to the flotation feed. Figure 11 shows that the amine dosage could be reduced by as much as two thirds by adding the polymer for the same amount of silica floated. Here, 15 grams of polymer had a tradeoff of 900 grams of amine. This translates to a reagent saving of at least one dollar per ton of concentrate. 0.8,
0.7 -.
H P 0.6 -. Y _ f 0.5 --
5 0.4 -a 0.3 -.
0.2 s 30
50
40
50
70
80
90
Wt.%OfSWldFbltd
Fig.11 Effect of polymer (0.014 kg/ton) on amine consumption and wt.% of sand floated As is indicated in Figure 12, however, polymer could have a detrimental effect on amine flotation above a certain concentration. Overdose of polymer not only reduces phosphate recovery to an unacceptablelevel, but also reduces the total amount of sand floated. 1M
!i.
m
m
j_::
;
30 ..
20 + 0.006
0.01
0.015
0.02
0.025
polvmrrmmw
Fig.12 Effect of polymer dosage on wt.% of sand floated and P,oS recovery at 0.23 kg/ton of amine
994
P. zhaaget al.
CONCLUSIONS The reverse Crag0 process is much more efficient than “Crago”. The clay (slime) effect may be controlled using the following techniques: a) b) cl
Adding polymer to the flotation feed and/or water. Adding amine stagewise. Adding amine continuously.
The process would be more attractive in times of fatty acid shortage. Based on an annual production of 20 million tons of concentrate in Florida, the new process offers the following environmental benefits: Eliminating sulfuric acid usage by the amount of 60,000 tons per year. Reducing soda ash usage by 20,000 tons per year. Reducing organic reagents usage by up to 750,000 tons per year. The reverse Crag0 process also promises the following economic benefits: Reducing total reagent cost by a dollar per ton of product. Improving flotation recovery by 2-5%. Reducing number of conditioners by 50%. Eliminating acid scrubbing circuit. Reducing sizing equipment by Xl-100%.
ACKNOWLEDGMENTS The authors wish to acknowledge the Florida Institute of Phosphate Research (FIPR) for the financial support They would also like to thank members of the FIPR Beneficiation Technical Advisory Committee for their support and input. Great appreciation is due to Cargill Fertilizer, IMC-Agrico and PCS Phosphate for their assistance in sampling as well as evaluating the project.
REFERENCES 1.
2. 3. 4.
Snow, R., P. Zhang, and Bogan, M., Challenging the Crag0 Double Float Process, All-&ionic Flotation of Siliceous Phosphates, Processing Supplement, Industrial Minerals, 1996, September, 4w. Gieseke, E.W., Section editor, Florida Phosphate Rock, SME Mineral Processing Handbook, SMJZ Inc., Littleton, 1985, Chapter 21, pp. 2-5. Gruber, G., Somasundaran, P., Understanding the Basics of Anionic Conditioning in Phosphate Flotation, 1996, FIPR Publication No. 02-090-12I, P. 31. Tanaka, Y., et al., Reagents in Phosphate Flotation, Reagents in Mineral Technology, Swjhctant Science Series V. 27, ed. P. Somasundaran and B. Moudgil, Marcel Dekker, Inc., 1988, pp. 645662.
Minrrda thgineering, Vol. 10, No. 9, pp. 983-994, 199-l 0 1997 Elsevia Scitace Ltd Printed in Great Britain. All rights resavd 08924875/97 $17.00+0.00
CHALtLENGING m “CFtAGO” DOUBLE FL,OAT PROCESS II. AMINE-FATTY ACID FLOTATION OF SILICEOUS PHOSPHATES P. ZHANG, Y. YU and M. BOGAN Florida Institute of Phosphate Research, 1855 West Main Street, Battow, FL 33830, USA (Received 14 October 1996; accepted 19 May 1997)
ABSTRACT In the conventionalphosphate flotation (Crago) process, 3&40% by weightof the silica present in the feed are floated twice,first by fatty acid, and then by amine. The Crag0 process is therefore ineficient in terms of collectorefficiency.Also, the phosphatemining industry is faced with higher fatty acid prices, lower feed grade, and stricter environmentalregulations.For these reasons, the Florida Instituteof PhosphateResearch has developed an amine-fatty acid flotationflowsheet, the reverse “Crago”. In this process, fine sands are first floated witha minimumdosage of amine condensateadded stepwise. The concentratefrom prejoat is then floated with a blend of SurfactaruYfatty acioY’e1 oil. The technical and economicfeasibility of this process were evaluated on seven feeds of varying characteristics,producing concentratesanalyzing30-32% P205 and 4-10470Ins01withPZ05 recoveries of over 93%, at totalreagent costs below $2 per ton of concentrate.This novel process could simplifythe current processingflowsheet by eliminating the acid scrubbing circuit, reducing the sizing section, and reducing the number of conditioners. The process consumes about one third to half the amount of reagents required by the Crag0 process at signijcantly improved flotation recovery. The effect of slimes (in either the flotationfeed or water) and techniquesfor reducing this eflect were also investigated.0 1997 Elsevier Science Ltd Keywords Froth flotation; flocculation; flotation reagents; industriaI minerals
INTRODUCTION In 1994, the Floridia Institnte of Phosphate Research initiated an in-house research project “A Screening Study on Phosphate Depressants for Beneficiating Florida Phosphate Ores ( FIPR #94-02-101)“. The threeyear project was designed to screen the available phosphate depressants for the dual purposes of developing alternate phosphate beneficiation processes for the currently mined siliceous phosphate deposits as well as removing dolomite from the Florida phosphate reserves of the future. As a result, two alternate processes for Florida siliceous phosphates have been developed. In these processes, fine silica is first floated with an inexpensive amine, and the prefloat concentrate is further cleaned by either floating phosphate (the reverse Crag0 process) or floating silica (the all cationic process). This paper discusses the former. The all cationic process was published in the Processing Supplement to the September 1996 issue of Industrial Minerals [ 11. hesented at MinemLrEngineering‘96, Brisbane, Australia, August 2648.19%
983
984
P. zhang et al.
Analysis of the “Crago” Double Float Process
It may be of help to briefly describe how phosphate is processed in Florida After desliming, the phosphate ore is subjected to sizing, Figure 1 [2]. Typical sizing involves using a hydrosizer to size the d&imed feed into coarse (lOOOx42O micron) and fine (420x105micron) fractions. In some more sophisticated operations as shown in Figure 2, three fractions are produced (lOOOx707,707x420,and420x105 micron).
I Coarse Feed 1000x420 Micron (16x35 Mesh)
Fine Feed 420X105 Micron (35x150 Mesh)
Fig. 1 Sizing
Spiral Feed 1OOOx707Micron (16x24 Mesh)
Fig.2 Sizing II The sized feed is first subjected to rougher flotrqtion,Figure 3. In this prucess the sized feed is dewaBed and conditioned at about 70% or higher solids with fatty acid/fuel oil at pH about 9 for thw minutes. The phosphate is then floated. It must be emphasized that a significantakount (30-4096)of silica is also floated in this step.
Challenging the Crag0 double float process
985
Fig.3 Rougher flotation The mugher conentrate goes through a dewatering cyclone, an acid scrubber, and a wash box to remove the reagents from the phosphate surfaces, Figure 4. After rinsing, the feed is transported into flotation cells where amine (sometimes with diesel) is added. The silica is floated at neutral PH.
HzSO., -
I
I
Acid&xubbex
I
I
1
Amine Flotation
1 Amine+
1
Am&Flotation
]
I
1
Concentrate
Sand Tailings
Fig.4 Cleaner flotation In the conve&onal “Double Float” (Crago) process for phosph@e minerals, 3W% by weight of the sands present in the feed are floated twice, first by fatty acid and then by amine. The Crag0 ~KK‘RSS is, therefore, inefficient in terms of collector efficiency. Fatty acid dosage for floating “pure” phosphate was found to be about 0.18 kg per ton. The theoretical dosage for floating a feed of 6.86% P,O, (15% BPL) is only 0.027
986
P. zhllg
et al.
kg/ton of feed (TOF). Actual plant fatty acid consumption for such a feed is about 0.54 kg/TOF. Therefore, plant collector efficiency is merely 5% (0.027/0.54)! The rest of the reagents are wasted prhmuily because of silica. However, there were a number of reasons for the phosphate industry to endorse the process enthusiastically: 1) fatty acid was much cheaper than amine so that anionic flotation followed by amine flotation made more economic sense than otherwise, 2) desliming was not sophisticated, leaving significant amounts of clay in the flotation feed so that amine usage would have been prohibitive had silica been floated first, and 3) the ore was high in grade, so the adsorption of fatty acid on silica was tolerable compared with that on phosphate in the rougher flotation stage. The situation is quite different today: the amine price is approximately twice that of fatty acid compared to nearly 10 times in the 1950s; the desliming technology has been upgraded to reduce the fine slimes in the flotation feed; the phosphorus content in the currentlymined phosphate ores is about half that in the past. These trends do not favor the standard Crag0 process. Understanding Amine Flotation There are two primary reasons why the Crag0 process has never been seriously challenged. First, fatty acids were inexpensive in the past. The second reason is this conventional wisdom about amine flotation: Amine flotation requires clean feed and deep well water. Therefore, one would need more sophisticated desliming devices, and utilize more deep aquifer water in order to float silica first. To some extent, this conventional wisdom is true. But, we could use our knowledge about amine flotation to overcome this problem. For example, we know that amines are more selective than fatty acids, and that amine adsorbs instantaneously on sand. The fact that amine is more selective in floating silica than fatty acid in floating phosphate is demonstrated in the following Figures. Figure 5 [31 indicates that fatty acid adsorption on silica is significant.
0
Fig.5
10
20 30 czadbhglime,mln. .
40
60
60
Kinetics of oleic acid adsorption on flotation feed and concentrate at 0.27 kg/ton oleic acid and pH 9.5
Figure 6 [4] shows that amine can float more than 99% of silica from pH 3 to 12, while phosphate flotation by amine is minimal within this pH range. Figure 7 [4] shows that at near neutral pHs, there is a large difference in zeta potential between silica and phosphate. Therefore, it is ideal to se@rate silica f&m phosphate at neutral pHs. However, neutral pH is not an ideal condition for floating phosphate osing fatty acids: That leaves on@ one o#ion: floating silica tirst.
987
Challenging the ‘Cragodouble Boat process
Flotation Recovery 110, 100 s_ 90 P 80 g 70 880 b0 g 40 930 $20 10 0 0
2
4
pH
10
8
8
12
14
of Solution
Fig.6 Flotation recovery of apatite and silica with dodecylammonium chloride at different pH
PH
8
10
12
Fig.7 Effects of pH on zeta potentials for silica and apatite Since amine adsorbs on silica very rapidly, the effect of clay on amine consumption may be reduced by adding amine stagewise. Flotation is conducted in a series of banks of flotation cells. Each bank consists of four to six cells. In the conventional process, all the amine is added as one dose in the first flotation cell. If a small amount of amine is added in the first cell, this cell not only acts as a flotation machine, but also serves as a desliming device. Since no conditioning is required, the number of conditioners currently used for ilotation may be reduced by floating silica first. Because amine flotation is conducted at neutral pH, pH modiier consumption would be significantly reduced by floating silica first. Finally and perhaps more importantly, for amine is more selective than fatty acid, collector efficiency should be dramaticallyimproved by floating silica first.
988
P. Zhang et al.
Amine-Fatty Acid Flotation (Reverse Crago) Process
Based on the above understandingand in recognition of the changed conditions, FIPR has developed a much simplified flotation flowsheet designated as “reverse Crago”, or aminefatty acid flotation process, as is shown in Figure 8.
I UnsizedFeed
I
Dewatering
1
Fatty Acid RlmwlSoda Ash r
I
I
1
1
_ -._
I-
PhosphatePlotation
Underflow Sand Tailings
1
Concentrate Product
Fig.8 Basic amine-fatty acid flotation process In this process, fine sand is first floated with a minimal consumption of amine by adding amine stepwise, so that phosphate loss could be minimized without using a depressant The concentrate from prefloat is then floated with a blend of fatty acid/fuel oilkrfactants as a phosphate collector. This new process is unique in the following aspects: 1) stepwise addition of amine, 2) novel fatty acid flotationreagent scheme that improves recovery of coarse phosphate particles, 3) higher collector efficiency, 4) simplitied flowsheet, and 5) flotation of unsized feed without sa&lcing metallurgical recovery. For feeds containing more coarse particles, two additional steps may be added: sizing the prefloat tails at 1190micron (14 mesh) to get a mini pebble, and scavenging the coarse fraction, +420 micron (+35 mesh), of the fatty acid flotation tails. Figure 9 shows the flowsheet with scavenging. The benefits of the aminsfatty acid flotationprocess are evident 1) Reduced flotation reagent consumption because only a single mineral is floated in each step; 2) Maxim&d phosphate recovery and improved concentrate grade because phosphate is in a poor position to compete with fine sands for &sorption of amine, and coarse silica is not easy to float in the second stage; 3) Reduced energy cost due to the elimination of the acid scrubbing circuit, reduction of sizing equipment and number of condlticmers;4) Simplified flowsheet since it is a non-sizing and non-scrubbing process; and 5) Minimized inorganic chemical consumption, particularly sulfuric acid.
Challenging the
1
Crag0 double float process
989
Dewatering ]
1 Fatty Acid Blend Soda Ash Fuel Oil
Fatty Acid Blend SodacAsh Fuel Oil
Sand Tailings Fig.9 Reverse Crag0 flowsbeet with scavenging fatty acid tails
EXPERIMENTAL Materials Flotation feeds
This new flowsbeet was tested on numerous flotation feeds from FIorida. Tables 1 and 2 show the basic chemical and size analyses of the feeds. All the flotation fee& tested are unsized, and were collected f&m the secondary hydrocyclone underflows. Most of the flotation experiments wei-e conducted on the as-received feeds. In a few occasions, slight desliming (rinsing with tap water) was practiced for comparison.
P. Zhng et al.
990
TABLE 1 Basic chemical compositions of flotation feeds
II
Al
II
I
9.62
I
70.60
A2
9.44
70.45
B
5.43
83.49
Cl
8.60
74.86
c2
6.77
79.75
D
I
13.82
I
60.2
TABLE 2 Size distribution of flotation feeds
II
Sample ID
II
I
Al
I
+841 pm 595-841 pm 420-595 unr (+2OIvI) I (28~2OM) I (35-28M) 1.3
I
3.3
I
105-420 pm (150-35IvI) I
10.2
83.7
I
-105 luIl (- 15OM) 1.5
A2
4.7
6.1
15.6
73.1
0.5
B
3.0
3.1
7.8
82.6
3.5
Flotation Reagents
The quartz collector used is an amine condensate designated as Custamina738 and provided by Westvaco. A blend of fatty acid, surfactants and fuel oil was used as phosphate collector. These are all commercially available reagents. Soda ash and sulfuric acid were used as pH modifiers. Bartow tap water was utilized in the initial tests, and plant water was used for studying the slime effect. Flotation Procedure
All the flotation tests were conducted in a standard one-liter Denver cell with a charge of about 500 gram dry feed. Apart from the stepwise addition of amine, all the tests followed standard flotation procedures.
RESULTS AND DISCUSSION Initial Feasibility Evaluation
Table 3 summarizes the flotation test results. In every case, flotation recovery of over 90% was achieved at reagent costs of around $1.5 per ton of concentrate. In a typical indust&lopemtiotr, flotation recovery is about 80% at reagent costs of $2.5-3 per ton.
Challenging the Crago double float process
991
TABLE 3 Performance of the Reverse Crago Process
Feed1 Al A2 B Cl c2 D
Table 4 shows the phenomenal reduction in total chemical usage by adopting the reverse Crag0 process. The first row was obtained by averaging data from two plants during a two-week period. The numbers for the reverse Crag0 process were averaged on the six feeds evaluated.
TABLE 4
IReagent consumption (kg/h) lProceS
comparisoh between the “Crago” and Reverse Crag0
1 Amine 1 Soda ash 1 H,SO,
Process
Fatty acid + fuel oil
Crag0
7.35
0.54
2.49
2.90
IReverseCrag0
1.98
1.78
1.30
0
Dealing with the Clay (Slime) Problem All the above results were generated using plant feeds with tap water, There consumption would be economically prohibitive when slimy feed and plant Therefore, more &Nation tests were conducted for parallel comparison between reverse Crag0 process using plant feeds and plant waters. Table 5 shows similar on a low-grade (35% P,O,) fine feed. TABLE 5
Flotation hocesses
performance comparison between the “Crago” and Reverse using plant water on a plant feed
Item
30.04
30.32
Concenlrate % Insol.
10.15
9.08
P,O, Recovery
Crag0
1 Crag0 Process 1 Reverse Crag0
Concentrate %P,05
%
was a concern that amine recycled water are used the Crag0 process and the metallurgical performauce
I
84.94
I
84.49
However, the reagent consumption made a big difference in terms of both total kilograms and dollars, Table 6.
P. Zbang et al.
992
TABLE 6
Reagent consumption comparison between the Crag0 and Reverse Crago Processes using plant water on a plant fine feed
Reagent
Dosage, kg/ton concentrate Crag0 Process
Reverse Crag0
Fatty acid/fuel oil
11.20
1.89
Amine
0.33
2.18
-
Polymer
0.15
Sodium silicate
1.17
-
Starch
1.87
-
Sulfuric acid
2.79
-
Sodaash
2.79
0.71
TOTAL REAGENT USAGE
20.16
4.93
$3.95
$224
TOTALREAGENTCOST
It should be understood that amine flotation is sensitive to clay in the feed and suspended solids in the water. It must also be recognized that amine is still more expensive than fatty acids. Therefore, the success of the reverse Crag0 process depends solely on viable techniques for reducing the effect of slime on amine consumption. Three techniques have been developed for this purpose: adding polymers in the flotation water, adding amine stagewise, and adding amine continuously. Figure 10 shows the effect of stagewise addition of amine. It may be seen that 0.45 kilograms of amine added stagewise was equivalent to about 0.90 kilograms of amine added as one dose.
0.6
10.5 i 0.4 y
!I 0.3
i 0.2 0.1 0 0
10
20
30
40
50
60
70
w&%ofsmdPhbd
Fig.10 Comparison of one dose amine addition with stagewise addition
993
Challenging the Crago double float process
Another very effective way of reducing the slime effect on amine consumption is to add a small amount of polymer to the flotation feed. Figure 11 shows that the amine dosage could be reduced by as much as two thirds by adding the polymer for the same amount of silica floated. Here, 15 grams of polymer had a tradeoff of 900 grams of amine. This translates to a reagent saving of at least one dollar per ton of concentrate. 0.8,
0.7 -.
H P 0.6 -. Y _ f 0.5 --
5 0.4 -a 0.3 -.
0.2 s 30
50
40
50
70
80
90
Wt.%OfSWldFbltd
Fig.11 Effect of polymer (0.014 kg/ton) on amine consumption and wt.% of sand floated As is indicated in Figure 12, however, polymer could have a detrimental effect on amine flotation above a certain concentration. Overdose of polymer not only reduces phosphate recovery to an unacceptablelevel, but also reduces the total amount of sand floated. 1M
!i.
m
m
j_::
;
30 ..
20 + 0.006
0.01
0.015
0.02
0.025
polvmrrmmw
Fig.12 Effect of polymer dosage on wt.% of sand floated and P,oS recovery at 0.23 kg/ton of amine
994
P. zhaaget al.
CONCLUSIONS The reverse Crag0 process is much more efficient than “Crago”. The clay (slime) effect may be controlled using the following techniques: a) b) cl
Adding polymer to the flotation feed and/or water. Adding amine stagewise. Adding amine continuously.
The process would be more attractive in times of fatty acid shortage. Based on an annual production of 20 million tons of concentrate in Florida, the new process offers the following environmental benefits: Eliminating sulfuric acid usage by the amount of 60,000 tons per year. Reducing soda ash usage by 20,000 tons per year. Reducing organic reagents usage by up to 750,000 tons per year. The reverse Crag0 process also promises the following economic benefits: Reducing total reagent cost by a dollar per ton of product. Improving flotation recovery by 2-5%. Reducing number of conditioners by 50%. Eliminating acid scrubbing circuit. Reducing sizing equipment by Xl-100%.
ACKNOWLEDGMENTS The authors wish to acknowledge the Florida Institute of Phosphate Research (FIPR) for the financial support They would also like to thank members of the FIPR Beneficiation Technical Advisory Committee for their support and input. Great appreciation is due to Cargill Fertilizer, IMC-Agrico and PCS Phosphate for their assistance in sampling as well as evaluating the project.
REFERENCES 1.
2. 3. 4.
Snow, R., P. Zhang, and Bogan, M., Challenging the Crag0 Double Float Process, All-&ionic Flotation of Siliceous Phosphates, Processing Supplement, Industrial Minerals, 1996, September, 4w. Gieseke, E.W., Section editor, Florida Phosphate Rock, SME Mineral Processing Handbook, SMJZ Inc., Littleton, 1985, Chapter 21, pp. 2-5. Gruber, G., Somasundaran, P., Understanding the Basics of Anionic Conditioning in Phosphate Flotation, 1996, FIPR Publication No. 02-090-12I, P. 31. Tanaka, Y., et al., Reagents in Phosphate Flotation, Reagents in Mineral Technology, Swjhctant Science Series V. 27, ed. P. Somasundaran and B. Moudgil, Marcel Dekker, Inc., 1988, pp. 645662.