SLUICES THEIR DESIGN, APPLICATION AND OPERATION
SAVANA MINING EQUIPMENT LLC
CALIFORNIA, USA
Placer System Design & Fabrication, Geological, Mining and Lands Consultants Phone: +353 83 452 452 5859 Europe +1 925 822 8852 USA www.savanamining.com Email:
[email protected]
PREPARED BY: Wyatt Yeager MSc – Savana Mining Equipment
OVERVIEW The dramatic increase in gold price during the years 2008 to 2010 created renewed interest in the development of alluvial gold mines and in particular has turned some previously uneconomic properties into viable mining ventures. The resulting interest has seen existing manufacturers of alluvial recovery equipment increase production and has introduced many new constructors into the market place. While conventional plant; screens, trommels, pumps, jigs and sluices remain popular there are many new innovations, while as yet untried in the long term, also available. The world economic downturn had resulted in equipment costs rising and many mine owners are thus forced, because of budgetary constraints, to seek less expensive recovery circuit equipment. The sluice remains remains one of the least expensive and reasonably efficient efficient “
” ”
means of recovering recovering heavy minerals. However, the capital savings savings do come with certain certain drawback’ drawback’s as sluices certainly do not offer the consistently high recoveries of machines such as jigs and centrifugal concentrators. This report is intended as a broad background to sluicing, a guide to the types of sluices available, their design, characteristics and and application. It is by no means intended to be a complete guide to sluicing and some comments and recommendations here may not be relevant to certain mining properties and their development. Sluices are one of the earliest means of recovering recovering minerals. Agricola in his De Re Metallica (1556) makes mention and describes in detail the use of strakes and sluices. The design of the sluice has altered little little from those times. Modern engineering practices have resulted resulted in more efficient riffle designs, the use of expanded mesh and 3M Nomad matting are two modern variations of the older wooden riffle and coconut fiber matting used up to as recently as the 1950’s. Sluices are at their most efficient when treating “Long Range” type feeds and do not respond well to “Short Range”, closely classified feeds. Having said that the combination of large boulders with fine sands is to be avoided. Similarly sluices are well suited to free running gravelly ground but not so well suited to fine sandy or clay rich feed.
While many operators claim high recoveries, +90%, when using sluices, this is not normally the case. Such high claims are often made as a result of the operator operator not actually knowing his average head grade or of his failure to account for losses of fine mineral across the sluice to tailings. Many operators never even know know they have fine mineral in in their feed stock. stock.
REPRODUCED FROM DE RE METALLICA, 1556 This report attempts to provide the reader with basic background information on the sluice as a gravity recovery mechanism. Much of the data contained in this report has been derived from old texts (See Bibliography), from the author’s author’s personal observations and from anecdotal information information provided by miners and prospectors. For those readers who require further information they are referred to the texts detailed later in this report and in particular to the very thorough study by Clarkson, 1990.
TABLE OF CONTENTS PAGE NO.
1.0
INTRODUCTION
1 – 2
2.0
COMPONENTS OF THE SLUICE
3 – 8
3.0
TYPES OF SLUICE
9 – 12
4.0
DESIGN CRITERIA
13 – 22
4.1
PRODUCTION RATE
14 - 15
4.2
SCREENING
4.3
FEED RATE / SLUICE CAPACITY AND DESIGN
4.4
WATER CONSUMPTION
5.0
GENERAL CONSIDERATIONS CONSIDERATIONS 5.1
SCREENING
5.2
MISCONCEPTIONS
15 15 – 15 – 22 22 22
23 – 26 23 24 – 24 – 26 26
6.0
SLUICE LOSSES
27 - 31
7.0
RECOMMENDATIONS
32 - 34
8.0
BIBLIOGRAPHY
35 - 36
LIST OF FIGURES PAGE NO.
FIGURE 1
COMPONENTS OF A HUNGARIAN RIFFLED SLUICE
3
FIGURE 2
EXPANDED MESH RIFFLES / MAT COMBINATION COMBINATION
8
FIGURE 3
LONG TOM AND RIFFLED SLUICE
10
FIGURE 4
KEENE TYPE UNDERCURRENT UNDERCURRENT SLUICE
10
FIGURE 5
OSCILLATING TYPE SLUICE
11
FIGURE 6
INFORMATION QUESTIONNAIRE QUESTIONNAIRE
13
FIGURE 7
GOLD SIZE ANALYSIS QUESTIONNAIRE QUESTIONNAIRE
14
FIGURE 8
CROSS SECTION OF RIFFLE RECOVERY MECHANISM AFTER CLARKSON, 1990
18
FIGURE 9
TYPES OF HUNGARIAN RIFFLES
18
FIGURE 10
BENT FLAT BAR RIFFLES
19
FIGURE 11
INCORRECT RIFFLE PLACEMENT (AFTER CLARKSON, 1990)
20
FIGURE 12
SELECTIVE DATA REPRODUCED FROM CLARKSON (1990)
26
LIST OF PHOTOS PAGE NO
PHOTO 1
PERUVIAN GROUND SLUICE, BOULDER RIFFLES
5
PHOTO 2
ARTISANAL SLUICE, MEXICO
5
PHOTO 3
ARTISANAL SLUICE, SULAWESI, INDONESIA
6
PHOTO 4
ARTISANAL SLUICE, CAMEROON WEST AFRICA
6
PHOTO 5
HUNGARIAN RIFFLED NUGGET TRAP SLUICE
7
PHOTO 6
EXPANDED MESH SLUICE RIFFLES
7
PHOTO 7
ARTISANAL SLUICE, SULAWESI, INDONESIA
27
PHOTO 8
INDONESIA, MINUTES AFTER PHOTO 7
28
PHOTO 9
CAMEROON, WEST AFRICA, NOTE EXTREME SLUICE ANGLE
29
PHOTO 10
CAMEROON, WEST AFRICA, NOTE LAMINAR FLOW ACROSS RIFFLES INDICATIVE OF POTENTIAL GOLD LOSSES
PHOTO 11
30
CAMEROON, REPLACEMENT SLUICE SYSTEM, EXCESSIVE WATER FLOW
30
1.0
INTRODUCTION:
Sluices are effectively a form of “stirred bed”. bed”. In other words the the feed, normally “Long Range”, is stirred or agitated at a rate insufficient to keep all the mineral grains and rock particles in suspension. This action, independent of the container in which the feed has been placed, results in reverse classification and gravitational stratification, the result being that the heaviest particles settle the most rapidly and form the base or basal layer in the container. In the case of a sluice it is the tumbling action imparted by the riffles coupled with the flow of the water across the riffles that effects separation and thus settling. Sluices have been used in various forms throughout the world for the recovery of gold, tin and tantalite and can be used for any mineral having a “Concentration Criteria” of 3.5 or greater. Other applications applications are washing pyrite pyrite or slate from from coal, lead shot shot recovery from rifle and gun ranges, etc. So what is a sluice? Basically it is an inclined trough or launder, usually on a slightly angled slope, into which the feed ore is placed and washed down the trough by a rapidly running stream of water. The sluice may have no riffles riffles (tin streaming box) or have riffles of which there is a wide variety that have been and are being used. Recoveries vary widely and depend to a large extent not only on the feed type, but the ability of the operator to monitor and adjust feed rate, water flow rate, sluice angle and sluice riffle load. No feed is consistently the the same, particularly alluvial or placer placer feeds, and sudden changes in feed size, feed rate, water flow and other such variables are the main cause of high losses over the sluice. Sluices were widely used in gold and tin mining by artisanal miners, in small scale mechanized operations and, up until the 1940’s, on large dredges. dredges. During that decade many of the larger dredge and even land based mining operations discarded the sluice in favor of modern pulsating jigs. jigs. The jig had much much to offer, a smaller smaller engineering footprint, footprint, more automated operation and much higher and more consistent recoveries.
Page 1
Notwithstanding these changes, the sluice is still popular and forms the basic tool of many of the world’s artisanal world’s artisanal miners. The Incas used natural rock sluices, see Photo 1, and wooden launders or sluices are seen seen in artisanal mining operations on every every continent. Photo 2. Recoveries are always problematic. Unless the operator has some some quantitative estimation of his head feed grade then it is impossible to determine sluice efficiency. Many claim recoveries of +90% and Clarkson, 1990, has documented very high recovery rates in studies he conducted in the Klondike and Yukon. Clarkson does, however, also report losses of between 0 and 71% in most operations, two operations he studied losing more gold than they recovered. Other mines with sophisticated screening systems prior to the sluice recorded recoveries of as high as 99%. Clarkson concluded that all the mines without screening equipment would pay back all their capital outlaid on screens in one season of operation. So do sluices have any application in modern mining? If operated carefully with adequate pre-screening and feed conditioning then they are capable of high recovery rates and the recovery to cost ratio in such cases is certainly better than other gravity machines. However sluices do not offer satisfactory recoveries of very fine gold particularly fine flattened particles. No writer appears comfortable in giving a lowest size limit for recoverability, some believe that sluices can can recover down to 200 mesh, this is unlikely! A conservative figure figure is more likely to be 50 mesh (300 (300 µm). Clarkson’s Clarkson’s data indicates that indicates that in the Yukon deposits there is little gold below 100 mesh and thus in that region the sluice is a popular recovery tool. The work by Clarkson, while restricted in its location is a valuable guide as to sluice efficiency and clearly demonstrates that for sluicing to be successful pre-sluice screening is required.
2.0
COMPONENTS OF THE SLUICE: Page 2
A sluice is effectively an open launder (or channel) into which are fitted riffles of one form or another. It can be a constructed box of metal, wood or plastic or a ditch excavated into bedrock (Inca Sluice). Sluice). Raw feed, usually screen underflow product, is introduced introduced into the head of the sluice and a flow of water applied so as to move the feed fe ed down the length of the launder across the riffles. riffles. The tumbling action imparted onto onto the feed by the water and the tumbling action of the coarse particles over the riffles results in separation of the mineral constituents in the feed, the heavy particles moving downward and becoming trapped on the floor of the sluice and behind beh ind the riffles. The various components of the sluice as set out in Figure 1 and described in more detail below.
FEED DISTRIBUTION BOX
PIVOT POINT NORMALLY A BOLT EITHER SIDE OF THE LAUNDER
HUNGARIAN RIFFLES CONSTRUCTED IN A STEEL FRAME
SLICK PLATE 600 MM TO 1,000 MM LONG
FASTENING CLIP EITHER SIDE OF LAUNDER
3M NOMAD MAT UNDER RIFFLED SECTION
FIGURE 1 – COMPONENTS OF A HUNGARIAN RIFFLED SLUICE
(I)
THE LAUNDER: Page 3
May be made from any material, normally it is made of steel or aluminum and apart from artisanal mining operations, rarely now of wood. The launder may be straight or tapered (Pinched Sluice). Old wooden sluices were tapered tapered so that each section of the sluice could be slotted together. Some modern sluices made of steel are also tapered. The “Long Tom” was tapered outwards in length; this effectively allowed the feed to spread out across the launder. This slowed down the flow rate, thus assisting the heavy minerals to settle more quickly.
(ii)
SLICK PLATE:
In most modern sluice boxes there is a plain flat section at the head or feed end of the sluice over which the feed passes prior prior to it reaching the the riffled section of the launder. launder. This blank section is normally 2 to 4 feet, 600 mm to 1.2 m in length, is designed so as to allow the heavy mineral in the feed stream to commence hindered settling prior to encountering the first riffle.
(iii)
THE RIFFLES:
There have been a wide wide variety of types of riffles riffles used in sluices. sluices. The Incas cut cut ground channels and used rock rock fill as the riffle medium. Photo 1 taken in Peru indicates indicates that this practice successfully continues even today. Other artisanal miners use more conventional wooden sluice arrangements, Photos 2, 3 and 4, taken in Mexico, Indonesia Indonesia and Cameroon, West Africa, respectively. Losses across these sluices are normally quite high. Riffles can thus be any material that imparts disturbance to the feed and promotes settling out of heavy minerals. Rocks, stones, and boulders, steel bars, wooden bars, wooden blocks, rails and poles (usually set longitudinally), angle iron, flat bar and expanded expanded mesh are and have been used. Modern sluices usually use a combination of “Hungarian Riffles” and “Expanded Mesh” riffles placed over synthetic matting such as 3M Nomad Nomad mat. See Figure 1.
Page 4
PHOTO 1 – PERUVIAN GROUND SLUICE, BOULDER RIFFLES
PHOTO 2 – ARTISANAL SLUICE, MEXICO
Page 5
PHOTO 3 – ARTISANAL SLUICE, SULAWESI, INDONESIA
PHOTO 4 – ARTISANAL SLUICE, CAMEROON WEST AFRICA
Page 6
PHOTO 5 – HUNGARIAN RIFFLED NUGGET TRAP SLUICE Hungarian Riffles are normally made of wood, steel or aluminum and are placed across the launder with an overhang on the downstream side. side. See Figure 1 and Photo 5. Alternately they can be made of angle iron, having the angle sloped slightly downstream. Expanded Mesh riffled sluices are conventionally used as scavenger sluices, they have a layer of expanded mesh mesh (walkway (walkway mesh) over Nomad mat. Figure 2 and and Photo 5. In this instance the slope of the mesh faces downstream.
PHOTO 6 – EXPANDED MESH SLUICE RIFFLES
Page 7
FIGURE 2 – EXPANDED MESH RIFFLES / MAT COMBINATION
Page 8
3.0
TYPES OF SLUICE:
There are many variations on the basic sluice, and one has only to access the internet to see the large number of groups now manufacturing commercial and hobby sluice boxes. Most modern sluices are constructed of steel or lightweight aluminum bodies and steel riffles. There are a number of manufacturers now constructing using plastics or Polyurethane with preformed or molded riffles. The list below is a fe w of the types of sluice in use, these being listed in no particular order of efficiency:
Streaming Box – effectively – effectively a non riffled sluice generally used for cleaning up tin or tantalum concentrates particularly when they contain accessory minerals such as zircon, rutile, ilmenite and monazite. The concentrate is placed at the head of the box into which is passed a regulated and evenly distributed flow of clean water. The operator stands stands in the box and using a flat mouthed shovel keeps turning the concentrate back on itself allowing the lighter impurities to move down the box to tails. While this does result in some some looses to tails along with the impurities the resultant product remaining at or near the head of the box can reach grades of over 74% 74 % Sn.
Long Tom – consists – consists of a tapered non riffled launder into which the feed and water are introduced. The box tapers outward at the discharge end and onto a punched plate screen. Large rocks are removed from this this section of launder by hand hand as is the screen oversize that accumulates. Fines are passed passed though to a wide riffled sluice. sluice. See Figure 3.
Undercurrent Sluice – Sluice – comes comes in many forms and is basically used to recover very fine gold that is conventionally lost to tailings. tailings. Figure 4 is of the Keene type Undercurrent where the -1/2” -1/2” material, the sands and fine gold are separated from the coarser material using a ½” punched plate screen.
Page 9
LONG TOM SCREEN
FIGURE 3 – LONG TOM AND RIFFLED SLUICE
Expanded Mesh Sluice – Sluice – is is depicted in Photo 6 and may be straight straight or tapered. tapered. They are normally a combination of expanded mesh over matting and used for recovery of fine gold lost across the primary sluice (Hungarian Riffled type).
NUGGET TRAP RIFFLES ½’ PUNCHED PLATE SCREEN
+½” SCREEN PRODUCT
KEENE TYPE 3-STAGE SLUICE BOX -½’ SCREEN UNDERFLOW PRODUCT
FIGURE 4 – KEENE TYPE UNDERCURRENT SLUICE
Triple Run Sluices - rely on a punched plate distribution system to separate the fine sands from the coarser coarser gravels much like like the simple Undercurrent Sluice. In this case the fines are distributed to two side sluices while the coarse material continues down the central run sluice. sluice. They are notoriously inefficient, inefficient, (Clarkson, 1990) with much much of the fine gold being trapped in the high and turbulent water volumes required to move the coarse material down the central sluice.
Page 10
Oscillating Sluice Boxes – – are being manufactured in a number of countries, some impart a side side to side motion, motion, others a circular motion motion similar to to panning.
The
efficiency of these units compared to simple straight run stationary units is not known. The diagram below illustrates the circular circular motion type sluice. sluice.
FIGURE 5 – OSCILLATING TYPE SLUICE
Hydraulic Riffled Sluices – – these have been tried at a number of locations with varying degrees of success. success. The normal installation sees every third riffle being replaced by a square tube in which are drilled small holes, water is introduced into the tube and thus thus out into the the area behind the riffle. This appears to rely on the settling velocity of the gold rather than creating disturbed vortices behind the riffles.
The riffles play an important role in sluicing, Peele notes three very important functions of riffles, specifically he states: “Riffles have three chief functions: (a) To retard material moving over them and give it a chance to settle; (b) To form pockets to retain gold which settles into them; (c) To form eddies which roughly classify the material in the riffle spaces.
Page 11
Their exact operation is not understood. Strength and shape of eddies (the “boil” of the riffle) is affected by shape and spacing of riffles, their position with respect to direction of flow, and the velocity of current. The boil must be strong enough to prevent the riffles from filling with heavy sand (packing), (packing), and not too strong strong to prevent lodgment of gold.”
Page 12
4.0
DESIGN CRITERIA:
Before considering the use of a sluice there are a number of parameters that must be taken into account. account . The first and and most important important is “Know the Ground”, specifically: What is the nature of the deposit, what are the average grade and feed size analysis and composition of the ground? The data set out in Figure Figure 6 is considered as minimum required required for adequate treatment plant design.
FIGURE 6 – INFORMATION QUESTIONNAIRE
Page 13
In addition it is also important to know the size analysis of the gold or mineral to be recovered. If for example the mineral is all in in the 100 to 200 mesh size range then selection of a system other than sluicing sluicing would be required. It is thus advisable to try try to determine grain size as follows:
FIGURE 7 – GOLD SIZE ANALYSIS QUESTIONNAIRE Size Analysis May Be Applied To Any Mineral Species
4.1
PRODUCTION RATE:
This is normally based on the economics of the venture, that is, how much feed is required to be processed to keep the operation profitable. profitable. Having settled on a figure then the design process can be commenced. commenced. Taken into consideration consideration must be:
Feed – Feed – what what is to be presented to the sluice, is it to be a “long range” feed, that is a feed with a wide size range or a “short range” feed, that is a more closely sized material;
Page 14
What is the nature of the feed material, is it sandy, cobble, clayey, etc.;
These factors determine the width and to some extent the length and the slope of the sluice.
4.2
SCREENING:
The screening of feed is of major importance. importance. Traditionally sluices were fed “long range” feeds and thus there was a requirement for large volumes of water to move the coarse gravels down and out of the sluice. This normally resulted in much, if not all, of the fine gold being carried away by the turbulent water flow and lost to tailings, in some cases as much gold was lost as retained. The use of modern earthmoving machinery and the availability of mechanical screens; trommels, vibrating screens, etc., meant that the sluice could be fed a ‘short range’ feed relatively easily. This resulted in improved improved recoveries and lowered water requirements. requirements. It should be noted however that an extremely “Short Range” feed is not desirable as this can lead to packing of the sluice by clays and fine sands, screening should be a happy compromise between a Long and Short range feed. Screening of feed is critical to good sluicing.
4.3
FEED RATE / SLUICE CAPACITY AND DESIGN:
Older texts relate feed rate to water carrying capacity, probably because most of the water was not pumped but flumed to the boxes. Modern pumps remove remove those concerns and getting water to the sluice is now considerably easier.
Calculation of Width: A simple simple rule of thumb used by some manufacturers is that you require 1” of sluice width per loose cubic yard of feed per hour.
Page 15
Note this is a loose cubic yard (lcy) (lcy) or material after excavation. excavation. In metric terms this equates to 0.30 lcm / hour per centimeter of width. Thus if you are intending to treat 10 lcy / hour across the sluice the width required would be 10”. This may well be an an oversimplification and and I have seen it misused on several occasions. The problem in applying a rather rigid rule of thumb is that the ground being treated is normally variable with size fractions changing both laterally and longitudinally within the deposit. Thus the feed rates to the sluice must change to reflect these variations. 3
It is often wise to add in a factor factor to the calculation, calculation, say 20%. Thus to achieve 10 y / hour would require a sluice width of 10” +2” = 12”. 3
Clarkson quotes feed rates of 0.65 to 1.30 y /hour per inch of width and makes a very valid point that Hungarian Riffled sluices would be the upper end and expanded mesh the lower end of those feed rates.
Calculation of Length: In most well operated sluice systems the bulk of the heavy mineral is retained in the upper 30% of the riffled area, however, this is not always the case and while shorter sluices are certainly less expensive to purchase or manufacture additional length is desirable. The author has observed losses of well rounded fines up to 1 oz nuggets across a 10 m sluice operating in the highlands of Papua New Guinea. “ Length Length does not affect capacity but does affect recovery rates” rates ” .
Many modern sluices are manufactured to material sheet sizes, thus 8’, 10’ and 12’ steel and aluminum sheet lengths usually control sluice length. On a small 30 lcy / hour plant where say 50% of the feed passes to the sluice after screening a length of 12’ would normally be re commended.
Page 16
It is a simple matter to add sluice modules to increase length however this should only be done where losses are high. These extra modules take take additional clean-out time and add greatly to plant operating workload and cost.
Riffle Action, Selection and Placement:
The Riffle Action or Mechanism:
It is important prior to selecting the riffle type to understand how the riffle works and why it works. Clarkson, 1990 summarized summarized very clearly the physical physical action within the the riffled bed and some of his comments are repeated here. Good riffle designs should ensure that there is a loose, active bed of sand in the interriffle spaces. This ensures that the the bed is maintained in in place while the only material being lost is from the upper layers in the sluice.
In this way heavy mineral is
continually being added to the inter-riffle active bed. Many early writers were of the opinion that it was the settling velocity of the gold that resulted in its capture by the riffles. Clarkson however likened likened the sluice to a centrifugal concentrator with the settling velocity only having the effect of allowing the heavy particle to to migrate to the the base of the slurry slurry column. At that point point the particles enter a turbulent vortex created by and in between the riffles. It is effectively this vortex, a low pressure system, which draws the slurry column down into the riffle riffle area. There, under ideal conditions, conditions, the slurry slurry is overturned and flows down the face of the forward riffle and in a circular motion to create and sustain the vortex. The gold and other heavies are are driven, by centrifugal force to the outside of the vortex and it is at the base of the vortex that centrifugal and gravitational forces forces drive the heavies out of the vortex and and into the matting. See Figure 8. To achieve this riffles should be of a height that does not interfere with the coarse upper layer to an extent that extreme turbulence results in the active inter-riffle bed being disturbed and lost to tailings.
Page 17
The most commonly used form of riffle is the so called called “Hungarian Riffle”. It comes in many forms but all have one common factor, a slight angling, against the flow in the case of angle iron and with with the flow in other cases. The riffles usually have an overhanging cap or bend. See Figure 9. STREAMLINES FREE WATER SURFACE
FINER / LIGHTER MATERIAL MATERIAL HEIGHT OF CUT TO VORTEX
LOW PRESSURE SEPARATION ZONE
COARSER / DENSER MATERIAL
EYE OF AREA OF PACKED SOLIDS
AREA OF
VORTEX
L I V E C R E S C E N T
PACKED SOLIDS
LIVE
MATTING
BASE OF SLUICE
BACK OF DEPOSITION
DOWNSTREAM
ZONE
RIFFLE SUPPORTING THE VORTEX
FIGURE 8 – CROSS SECTION OF RIFFLE RECOVERY MECHANISM AFTER CLARKSON, 1990 ANGLE IRON RIFFLES WITH MODIFIED AND SHORTENED CAP
BENT FLAT BAR RIFFLES
WOODEN BLOCK WITH STEEL CAP
FLOW
FLAT BAR
EXPANDED MESH RIFFLES
FIGURE 9 – TYPES OF HUNGARIAN RIFFLES Page 18
Riffle Selection:
Conventionally angle iron or angled Hungarian Riffles are used when the bulk of the heavy mineral is larger than than 1 mm (14 mesh). Where angle iron riffles are to be used they are normally made from 1” angle with the top leg modified or reduced in length to ½”. These types of riffle are are normally inclined upstream upstream at approximately approximately 15 degrees. Where bent flat bar is used riffles are normally inclined downstream, riffle heights O
should not exceed 1-1/2”. 1-1/2”. The riffles riffles are inclined forward at 15 and have a top O
section bent forward a further 35 . Figure 10.
35
O
" 0 5 . 0
O
105
" 5 7 . 0
3.0"
FIGURE 10 – BENT FLAT BAR RIFFLES
Straight flat bar riffles are normally only used for heavy minerals of greater than 2.5 mm (8 mesh) as they create greater turbulence and tend to dislodge upward all but the heaviest and coarsest coarsest material. They are ideally suited suited to the recovery of coarse coarse nugget gold, say +1/2”. Expanded mesh riffles are normally used where particle size is -1 mm (14 mesh). They tend because of their shape to have a small sorting area and are thus prone to variations in feed or water volume.
Page 19
Riffle Placement:
This is critically important as incorrect placement will result in the accumulation area either becoming packed with solids in the case of a narrow riffle spacing or either scoured or packed solid in the the case of extreme spacing. See Figure 11 (After Clarkson 1990).
1" ANGLE IRON MODIFIED TOP
LEG ½” VORTEX 15.0°
AREA OF PACKED SOLIDS AREA OF PACKED SOLIDS
1 I N NC H C T O OO N H O AR R RO W O W
3 I N NC H C E H S E S T O OO W O I D DE E
FIGURE 11 – INCORRECT RIFFLE PLACEMENT PLACEMENT (AFTER CLARKSON, 1990)
The ideal riffle placement where modified 1” angle iron is being used is 2” with a 15 O
degree angle upstream. Bent flat bar riffles riffles are angled downstream at 15 .
Sluice Operating Angle: There is widely divergent opinion on the ideal operating operating angle of sluice boxes. The operating angle and water volume will determine how the gravel is transported across the sluice, that is whether is slides, and rolls across the tops of the riffles in the case of the coarser particles or reacts in a turbulent or leaping motion (saltation) in the case of the finer material.
Page 20
As a general rule the steeper the operating angle the more violent the motion of the particles. A common rule of thumb used in old texts is a drop of 6” to 6.5” per 12’ of sluice length or an angle of 2.5
O
(Taggart).
Other texts such as Peele tabulate the
specifications of various mines; Peele notes angles varying from 2% to 15% of slope O
O
O
O
(1 to 8.5 ). In Guinea slopes were quoted quoted as 2.5 for loose sand, 5.75 for loose O
sand and gravel and up to 8.5 for coarse gravel and boulders. O
Clarkson (1990) quotes angles of 3” per foot (14 ) when using angle iron riffles while O
Savana prefer a 6 slope for the primary sluice and even ev en steeper for nugget trap riffle boxes. As the application of sluice boxes is so variable, each mine will have its inherent differences in alluvium type, size ranges, etc. The simplest arrangement is to have the sluice set up so that the angle can be easily altered, even while in operation. This arrangement would see variable slope adjustment as follows:
O
O
Nugget Trap Trap Sluice Sluice (0 to 15 );
Primary Sluice (5 to 12 ); and
Expanded mesh Sluice (2 to 10 )
O
O
O
O
The Matting: All modern sluices sluices use some form of matting matting under the the riffles. There are many many varieties available from simple coconut fiber, hessian sacking, astro turf, synthetic carpet and 3M Nomad mat. Each has its own characteristics, characteristics, specifically:
Coconut / Astro Turf – have – have very limited storage capacity and are extremely difficult to clean;
Page 21
Monsanto Matting - has a hard backing backing that makes collection collection of gold difficult and the long fiber strands project up into the vortex area and disrupt the action and collection ability in the vortex; and
3M Nomad Mat – Mat – is is the most widely widely used sluice matting. matting. It does not interfere interfere with the vortex, most of its volume is available for capture of mineral particles, and it has no stiff backing and is very easy to clean.
4.4
WATER CONSUMPTION:
In considering what water volume is required to operate the sluice successfully consideration must be given to what max. and min. water volumes are locally available and what is the quality of that water. Clearly the volume of water to be introduced into the sluice depends to a large extent on the type of ground being treated. Nominally Savana quote a rate of 10 gpm / cubic yard per hour however that rate may may be low for very rocky ground. Clarkson (1990) quotes volumes of 320 gpm per foot of width or 26.7 gpm / inch of width. Trying to cross reference these rather confusing figures indicates that in the Savana case the water requirement would only be 120 gpm for a 12” sluice and in the Clarkson case 320 gpm. Who is correct, probably neither neither as actual volume still remains remains dictated by the nature of the ground. In considering the volumes to be used it is important to determine the water source within the flowsheet. Where screening is being used the the screen underflow will contain contain all the wash water, for a 30 yph plant water volume is normally a maximum of 500 to 550 gpm. Clearly that would be far too great for a 12” sluice sluic e so there must be an effort to either dewater the feed or reduce the screen plant wash water. Every sluice should be equipped with its own wash water source, usually a 2” water line with valve for additional make-up water or for sluice clean-out.
Page 22
5.0
GENERAL CONSIDATIONS:
Given that there are some excellent publications and reports available about sluicing this section deals mainly with things the potential sluice operator should consider.
5.1
SCREENING:
It is quite clear from the material in the Clarkson (1990) report that pre-screening of the feed to a sluice results in far higher recovery rates. rates. The advantages of pre-screening pre-screening feed are:
Much less water is required;
Barren gravels are eliminated from the sluice box;
The removal of larger barren rock reduces the riffle and box wear and thus reduces maintenance and replacement costs;
The necessity for triple run and undercurrent sluices is removed and the difficulty encountered in splitting feeds in these types of sluices totally eliminated; and
Last but not least the feed enters the box in a well conditioned state, that it is prebroken and free running and does not rely on the action of the box to separate the mineral grains from the feed.
Numerous types of screen are available, trommels, vibrating screens, derockers, and hydraulic finger grizzlies. Each has its own application application to a particular operating scenario. The trommel or scrubber trommel is probably the most efficient unit for cemented or high clay gravels but it has high capital cost and usually requires the construction of high feed ramps. The introduction introduction of hopper – hopper – feeders feeders ahead of the trommel assist in reducing feed height. Multi deck vibrating screens are also an efficient screening mechanism; they require less feed height but are not suited to very clayey gravels or ground containing large boulders. The derocker is more a moving grizzly feeder that has limited throughput and cannot size feed to -2.5”. -2.5”. Screening is thus strongly recommended.
Page 23
5.2
MISCONCEPTIONS:
Testing of alluvial ground using sluice boxes is fraught with difficulty and many plant designs have failed because of errors errors in interpretation interpretation of test data. Specifically:
Fine Gold – – the presence or absence of fine gold in a sluice test must be treated carefully. Even the worst sluice sluice will recover fine gold and even the the best operator will lose fine gold;
Coarse Gold (Nuggets) - operators should be very aware of the “Nugget Effect”, one nugget per cubic yard does not necessarily mean that this result will be repeated for every yard treated. Expanded mesh test test sluices will normally discard most nugget material and retain the finer gold fractions while nugget trap sluices will efficiently retain nuggets at the expense of fine gold;
Concentration – Concentration – a a high concentration of gold in the first few riffles of a sluice is not an indicator of high efficiency, nuggets can be lost over even the most carefully operated sluice and for testing it is strongly recommended that the primary sluice be followed by a fines sluice and that tails be regularly sa mpled.;
Error – Error – efficiency efficiency of a sluice should not be based on the amount of gold recovered. Alluvial ground is highly variable both both in grade and in mineral grain grain size. Many samples and larger volume samples are usually required to satisfy sampling reliability;
The Gold Pan – is far from a quantitative testing mechanism, “Nugget Effect” and operator error make small volume dish samples at best qualitative;
The Bulk Test Plant – Plant – sample sample reliability can be calculated as a %, (See separate report) and it is a good rule to apply some statistical analysis before embarking on a lengthy and costly testing program. As a rule larger samples reduce sampling error particularly where there is a large large % of fine gold, or a very strong “Nugget Effect”. It would be nice to see a Normal Statistical Distribution Distribution of gold sizing but that is unusual to extremely rare.
Page 24
In operating situations the claim that the sluice is catching 100% of the gold is a major misconception. Many operators claim they have no fine gold because they just never see it, probably because their test work was was such that it was was never recovered. Clarkson (1990) has tabulated production and recovery rates from a series of mining operations, those data are somewhat of an eye opener. Losses are in some cases extreme, more gold being lost than recovered, losses of 0 to 71%, etc. The following Table, Figure 12, is taken from his work and is a snapshot snapshot of some of those data. It is clear from those data that prescreening of the feed prior to sluicing dramatically improved recovery rates; further the addition of screening devices to mines currently without that process improves recovery and reduces monetary monetary losses. Certainly you cannot capture all the gold or heavy mineral but the operator has choices which enable him to minimize losses and quite significantly improve his monetary returns. Sluices remain one of the least expensive gravity recovery processes available and while relatively easy and inexpensive to operate still remain one of the most misunderstood machines and processes available available to the mine operator. It should be remembered however that they are not suited to the recovery of extremely fine gold, that is, gold below 100 mesh Tyler. Many deposits that are now being developed that contain high percentages of very fine gold; the day of the sluice may well be numbered. The deposits with coarse nugget gold gold are rapidly disappearing.
Page 25
SIZE DISTRIBUTION AND RECOVERY SLUICE
S I NGLE RUN TYL ER ER
DI AM
MES H
MM
DISTRIBUTION %
LOCATION
HOMEMADE TRI PLE
RECOVERY %
S +4
DISTRIBUTION %
RECOVERY %
M
PEARS ON TRI PLE DISTRIBUTION %
RECOVERY
ROTATING ATING TROMM TROMMEL EL DISTRIBUTION
%
U
%
RECOVERY
VIBRATING ATING SCREEN SCREEN DISTRIBUTION
%
B
%
RECOVERY %
A
4.76
+8
2.38
+14
1. 1 .19
1
52
33
84
8
44
0
100
7
100
+28
0.59
13
68
37
88
28
68
35
100
39
100
+48
0.29
66
48
24
84
44
84
38
61
24
96
+100
0 .14
20
36
6
60
18
64
21
78
25
96
- 1 00
0 .9
0. 5
3
5
5
100.9
100.5
101
99
100
5
0.6
MONETARY MONETA RY VALUE OF GOLD LOSSES (GOLD CA $400.00 / OUNCE) 41
Raw Gold gm / hr $/Hour $ / 1,200 Hours Overall Recovery
14
37
$418.00
$145.00
$381.00
$47 .00
$6.0 0
$502,000.00
$174,000.00
$457,000.0 0
$57,00 0.00
$8,000.0 0
48%
84%
72%
79%
9 8%
4
0.3
RECOVERABLE GOLD LOSSES ( GOLD CA $400.00 / OUNCE) Raw Gold gm / hr $/Hour
38
11
36
$390.00
$113.00
$367.00
$44 .00
$3.0 0
$ / 1,200 Hours
$468,000.00
$135,000.00
$440,000.0 0
$53,00 0.00
$3,000.0 0
Capital Cost
$50,0 00.00
$100,000.00
$100,000.0 0
$1,000.00
$1,000.0 0
Operating Cost
$5,000.00
$10,00 0.00
$10,0 00.00
$O
$0
96
96
99
99
99
Overall Recovery
FIGURE 12 – SELECTIVE DATA REPRODUCED FROM CLARKSON (1990)
Page 26
6.0
SLUICE LOSSES:
Losses across sluice boxes vary widely but as a rule are reduced by pre-screening the feed material prior to sluicing. sluicing. Recorded losses (Clarkson, (Clarkson, 1990) varied from 0 to 71% with at least two operations losing more gold than they recovered. Unlike jigs, which in many respects are more forgiving in their application, sluices respond badly to variations in feed rate, water flow, slope and feed type. Water and feed flow are related. Periods of no, or low feed, usually see little change in the water flow, thus in this instance, the low feed causes the water to flush the sluice and some gold, particularly that material sitting exposed between the riffles, will be lost to tailings. Similarly surges in the supply of water will increase turbulence in the case of more water and cause losses of fines to tails or in the case of lower water flow cause general losses across the whole sluice. See Photos 7 and 8.
6.0
SLUICE LOSSES:
Losses across sluice boxes vary widely but as a rule are reduced by pre-screening the feed material prior to sluicing. sluicing. Recorded losses (Clarkson, (Clarkson, 1990) varied from 0 to 71% with at least two operations losing more gold than they recovered. Unlike jigs, which in many respects are more forgiving in their application, sluices respond badly to variations in feed rate, water flow, slope and feed type. Water and feed flow are related. Periods of no, or low feed, usually see little change in the water flow, thus in this instance, the low feed causes the water to flush the sluice and some gold, particularly that material sitting exposed between the riffles, will be lost to tailings. Similarly surges in the supply of water will increase turbulence in the case of more water and cause losses of fines to tails or in the case of lower water flow cause general losses across the whole sluice. See Photos 7 and 8.
PHOTO 7 – ARTISANAL SLUICE, SULAWESI, INDONESIA
Page 27
In Photo 7, a sluice fed by gravel pumping, several potential loss causing factors can be seen, specifically:
The sluice is not level;
The sluice has to few riffles, note the riffle f ree section in the centre of the sluice;
The sluice is being fed a “Long Range”, -4” -4” feed with no real screening and the operators are using highly turbid re-circulated water.
PHOTO 8 – INDONESIA, MINUTES AFTER PHOTO 7 Note in Photo 8, taken only minutes after Photo 7, which the flow rate across the sluice has dropped dramatically but that that a surge of water can be seen at the the head of the sluice. This surge will probably have resulted in gold be ing scoured from the riffles. In one instance the author has observed over-screening causing severe sluice losses, Photos 9, 10 and 11. In this instance in Cameroon, West Africa, the operators were trying to treat extremely clay rich ground and in the process screening to a -5 mm sluice feed.
Page 28
Apart from very poor trommel design that saw screens blind to about 20% open space availability, and fail to break up all the clay, the sluice was set at an extreme angle greater O
than 15 . This created the mini Niagara falls seen in Photo 9 and was made made worse by use of excessive water flow. The gold had little time to settle and no apparent vortex was able to form behind the riffles. The clays formed hard packed areas behind the riffles The solution was to install a dual sluice system of a Nugget Trap ahead of Expanded Mesh sluices. These units operated at flatter angles and resulted in better sluice performance although excessive water flow was a common problem.
PHOTO 9 – CAMEROON, WEST AFRICA, NOTE EXTREME SLUICE ANGLE
Page 29
PHOTO 10 – CAMEROON, WEST AFRICA, NOTE LAMINAR FLOW ACROSS RIFFLES INDICATIVE OF POTENTIAL GOLD LOSSES
PHOTO 11 – CAMEROON, REPLACEMENT SLUICE SYSTEM EXCESSIVE WATER FLOW
Page 30
It is also important to stress that sluices are not suited to recovery of very fine mineral product; sizes below 100 mesh are better recovered using other means. This fine size material is normally lost across the sluice in the turbulent flowing upper layers. Losses certainly can be minimized but not completely avoided.
Page 31
7.0
RECOMMENDATIONS:
The following are some recommendations for those considering using sluice boxes for the recovery of any heavy mineral, specifically: i.
“Know Your Ground”
The mine owner should make every effort to know the ground he intends to work. Simple procedures can be adopted such as:
Conduct feed size analyses at several locations throughout the area to be worked;
Determine the nature of the alluvium, whether it is sandy, clayey, rocky, etc.; and
Take several bulk samples and from the concentrate determine the size range of the mineral being sought.
Other information on items such as water supply, etc., is also valuable when considering pumping and pipeline requirements. ii.
Feeding and Screening Screening of the feed gravels is essential and where possible should be into three fractions; +1”, -1” to +1/2” and -1/2”. The 1/2”. The former is discarded to tailings while the other two are sluiced to recover the heavy minerals. Feed rates should be closely controlled either using mechanical feeders such as conveyors, apron feeders, vibrating grizzly feeders or even hand operated monitors. Sluices like constant feed without surging.
iii.
The Sluices Where a twin sluice system is to be used, that is where there are two size fractions the following should be adopted:
Page 32
The Coarse Fraction: -1” to +1/2” should be passed across a nugget trap sluice with a minimum O
length of 12’, 12’, set at a steep angle, at least 9 ; Ahead of the riffles the sluice should have a 3’ slick plate to assist in settling of gold or heavy mineral in the slurry column prior to hitting the first riffle; The sluice should be fitted with either angle iron or bent flat bar riffles as described in the text, riffle spacing of 2” for angle iron and 3” for flat bar; Water may have to be introduced and should not be less than 10 gpm / cubic yard of feed, but no less than 150 to 200 gpm per foot of sluice width. The operator may find higher water rates are required to move the gravel down the sluice and should avoid the phenomenon of “Rooster Tailing” by adjusting slope and water volume until a good turbulent flow is achieved; An over turbulent water motion (scoured bed) and / or a flat flowing water surface (indicating a packed sluice bed) will result in gold losses; and The riffles should be laid on Nomad Mat and clamped tightly down.
The -1/2” Fraction: Should be processed using two sluice runs in series, specifically:
Nugget Trap Section: A first nugget trap type sluice having a slick plate at its head and angle iron or bent bar riffles set tightly over Nomad Mat; M at; Riffle layout as described in the text, normally a 2” spacing or 3” in the case of flat bar riffles; The sluice run should be no less than 10’ and ideally 12’, width is determined by feed volume; O
O
Slope should be steep 5 to 9 depending on the wash type;
Page 33
The riffles clamped down on Nomad Mat; and Water volume determined by feed rate;
Expanded Mesh Sluice : A 12’ to 16’ section of sluice fitted with coarse expanded mesh tightly clamped on Nomad Mat; and The sluice should be wider than the Nugget Trap section, the difference should be around 20%, and the sluice set to operate at a flatter angle.
Ideally the sluices should be constructed so as to have adjustable operating angles and provision made to be able to introduce extra water at the head of the sluice. iv.
General: High sluice recoveries can only be achieved by good design and care in their operation. Sluices should not be allowed to pack with concentrate and should be cleaned out on a regular basis. Where nugget ground is encountered they can be fitted with locked mesh security screens.
Page 34
8.0
BIBLIOGRAPHY:
AGRICOLA, Georgius.
1556
De Re Metallica. Translation by Herbert Clark and Lou Henry Hoover, 1912 The Mining Magazine, London
BOWIE, Jr., Aug. J.
1895
A Practical Treatise On Hydraulic Mining In California, D. Van Nostrand Company, New York.
CLARKSON, Randy.
1990
Placer Gold recovery research, Final Summary. Klondike Placer Miners Association, New Era Engineering Corporation.
2010 The Use of Nuclear Tracers to Evaluate the Gold Recovery Efficiency of Sluice boxes
Gravity Gold, 2010, Proceedings AusIMM
GRIFFITH, S.V.
1938
Alluvial Prospecting and Mining. Mining Publications, Ltd, London
HARRISON, H.l.H.
1946
Examination, Boring and Valuation of Alluvial and Kindred Ore Deposits. Mining Publications Ltd., The Mining Magazine, London.
1962 Alluvial Mining for Tin and Gold. Mining Publications Ltd., The Mining Magazine, London.
LONGRIDGE, C.C.
1902
Hydraulic Mining, Part III The Mining Journal, London
Page 35
1906 Gold Dredging, Annual Supplement, 1906 The Mining Journal, London Gold and Tin Dredging and Mechanical Excavators The Mining Journal, London
MACDONALD, MACDONALD, Eion H.
1983
Alluvial Mining, The Geology, Technology and Economics of Placers. Chapman And Hall, London
PEELE, Robert.
1945
Mining Engineers Handbook, Third Edition. John Wiley & Sons, Inc. London
RICHARDS, Robert H.,
1903
Ore Dressing, Vol. I & II The Engineering & Mining Journal, London
TAGGART, Arthur F.
1945
Handbook of Mineral Dressing, Ores and Industrial Minerals John Wiley & Sons, Inc. London
WELLS, John H.
1969
Placer Examination, Principles and Practice, Bureau of Land Management, US Government Printing Office, Washington DC.
Page 36