Leaching (lixiviation) is commonly used for recovering a wide variety of metals from ores, including precious metals. A lixiviant dissolves the desired metals from the ore; cyanide leaching, for example, is widely used to extract gold.
Lixiviation is often followed by a Counter Current Decantation (CCD) circuit to recover the valuable mineral from the gangue. CCD consists of washing the feed slurry in a series of thickeners until the majority of the dissolved metals are removed.
The conventional choice for this process has been high-rate thickeners. However, using high-density or paste thickeners can be more cost efficient for many sites, as outlined here.
A review of CCD fundamentals is a good beginning. Emmett & Dahlstrom 1974 summarized it well by stating: “In its simplest form, washing in a thickener can consist of mixing solids and associated solution with water, settling the solids, decanting the clarified solution, and then repeating the process as required until the dissolved material is removed.
"The logical extension of this procedure is to use the more dilute, decanted solutions as wash liquid in earlier stages, a procedure which lends itself to continuous operation, and, hence, continuous countercurrent decantation.”
Figure 1 (above) provides a typical schematic for a CCD circuit. Feed slurry enters at Thickener #1 and wash water at Thickener #3. Flowing counter to each other, the overflows collect more of the valuable solution and the underflows more of the waste solids.
Emmett & Dahlstrom explained that the controlling factors of CCD are:
- Number of stages (number of thickeners)
- Liquid in overflow versus underflow
- Efficiency (thoroughness of mixing)
Therefore, the fundamental formulas for calculating recovery have nothing to do with the volume or mass being processed, but are dependent upon items 1 through 3 above.
Of course larger volumes of liquids and solids can increase the challenges of ensuring thorough mixing. So efficiency (item 3 above) is often improved or ensured by placing mix tanks between thickeners to blend wash water and slurry. For simplicity, our schematic in Figure 1 does not show the tanks, but the mixing stages are implied.
A properly designed feedwell also enhances mixing. It is an effective place to add flocculants (a key component in today’s thickening technology) and is important to good distribution in the thickener. Optimum distribution not only eliminates short-circuiting, but ensures use of the entire thickener volume (see Figure 2), which of course safeguards the performance of the CCD train (e.g. efficiency).
Figure 2, below, shows computational fluid dynamics comparisons of feedwells and their impact on thickener distribution.
To highlight the variables in designing a CCD circuit, we will consider different recovery estimates based on an ore leach followed by CCD. Table 1 outlines the base data for a three-stage CCD with high-rate thickeners.
In our example, a little over 86% of the dissolved values would be recovered. In most instances, that is unacceptably low and the system designer would look for ways to improve recovery, possibly by increasing wash or adding stages.
Increasing wash water to 3.5 tons of water per ton of solids, or a wash ratio of 1:1.1, raises recovery to about 95% (Table 2). Adding one stage (a four-thickener circuit) and keeping wash at 2.5 tons of water per ton of solids, provides a recovery of 96.2% (Table 3).
Using more water primarily impacts operating expenditures (OPEX), while adding thickeners also impacts capital expenditures (CAPEX); but is probably more significant for CAPEX. Of course there are other factors to consider, but weighing those details against the needs of the end user provides the necessary data to make sound design decisions.
The Premise for Using High-Density Thickeners
High-density and paste thickeners use deeper sidewalls and steeper cones to increase sludge bed depth and, thereby, increase solids concentration to produce a non-Newtonian slurry at the thickener underflow.
As a result, yield stress becomes a key design element for the thickener, pumps, etc. High-rate thickeners produce slurries that behave like Newtonian fluids and yield stress is not so critical.
Figure 3 provides a relative comparison of yield stress across thickener types. More applicable to our CCD discussion, the thickener profiles (while not to scale) also give a sense of the relative differences in depth versus diameter.
Increasing solids in thickener underflow means that more liquid reports to the overflow, enhancing efficiency of the CCD circuit. Improved efficiency can also reduce the first controlling factor, which is number of stages.
Using our example from Table 1, but substituting paste thickeners, we can change just the underflow assumption to 60% solids and dramatically impact recovery.
As can be seen in Table 4, a three-stage circuit with deep cone thickeners produces 97.6% recovery. In Table 5, a four-stage circuit yields 99.6%.
Under our assumptions, a three-stage CCD circuit using paste thickeners outperforms a four-stage CCD with high-rate thickeners (97.6% vs 96.2%). The process choice is clear, unless high volumes of wash water are justifiable; which of course depends greatly on the values contained in the pregnant liquor.
While process outcomes are key in making design decisions and often are the deciding factor, there is more to consider in a thorough evaluation.
The single largest operating cost for thickeners is use of polymers for flocculation. Polymer dosage is based on tons of solids in the feed slurry and, while non-Newtonian underflows may require slightly more polymer, the increase is not significant and is well within the margin of error one would expect from operating variances. For purposes of this discussion, we will assume that all thickeners use the same amount of flocculants for the same tonnage.
Of greater interest is power consumption. Perhaps it is obvious, but driving rakes through a denser sludge bed requires more power. However, the power differences are not as great as they might appear.
Let us assume a 35M high-rate thickener being driven by three 7.5 kW (10 HP) motors. Installed power then is 22.5 kW. We will further assume an alternate option of a paste thickener with twice the installed power, or 45 kW, which is a conservative estimate.
To ensure long life, gear drives are sized to handle larger loads than they are expected to see in normal operation. Thickeners can experience substantially high loading from upset conditions and a sound design means correspondingly larger drives to accommodate those upsets. While sizing practices vary a little between manufacturers, a 40% duty rating is a reasonable assumption.
Using mid-range power costs for the United States and assuming continuous operation for 365 days per year, we can see in Table 3 that each paste thickener would add a little over $6,000 per year to OPEX. An increase of 50%, versus the apparent 100% based on motor nameplate; and again, our 2X drive size for the paste thickener was conservatively high.
Power consumption in underflow pumps would likewise be impacted, but to a lesser degree. Mixers may also be a bit larger. Based on our example in Table 6, a reasonable assumption for total increased OPEX would be $10,000 per thickener per year. A four-stage circuit would then be $40,000 per year more in operating costs.
Put into perspective, we will say that our 35M tailings thickener processes 10,000 mT per day of solids and uses mid-range polymer dosage and costs. Table 7 then suggests that operating costs (not considering maintenance) would be in the range of $400,000 per year per thickener. So our increase in power consumption by selecting paste thickeners is about 10% of the OPEX.
Of course the estimates are general in nature and should not be considered a substitute for proper evaluation of a specific installation. Qualifiers that illustrate the point are:
- Paste or high-density will typically require one less thickener. Following our example, that would mean less than 8% OPEX increase.
- High-density thickeners require lower power than deep cone, so proper selection is key. Sedimentation tests are important.
- We did not address the option of increasing wash water, which could reduce the number of thickeners.
- Requirements for the final pump stage can vary a great deal, depending on the disposition of the underflow from the last thickener (#3 in Figure 1). We will discuss those challenges in more detail under the CAPEX assessment.
As mentioned, larger drives are required for thickeners capable of producing non-Newtonian underflow slurries. Hence sturdier rakes, shafts, and bridges are also required. A typical paste thickener can be about 70% more expensive than a high-rate. It is an approximate number, but sufficient to enable a comparison.
We can see then that elimination of one stage means CAPEX increases only 36%, even though the per-thickener increase is 70%. Still a substantial cost, but not as high as might be expected.
A less obvious benefit to choosing thickeners capable of producing non-Newtonian underflow is a substantial reduction in footprint.
In the graphic above, the red circles represent deep cone thickeners, green represents high-density, and blue is high-rate.
Footprint can become a significant CAPEX consideration when excavation costs are elevated. To illustrate, we will assume our high-rate thickeners in Table 8 cost $1 million each. Therefore, all thickeners for the high-rate option would be $5,000,000 and for the deep cone option $6,800,000; or $1,800,000 more.
If excavation costs range from $50/m3 to $200/m3 and 3m of material must be removed from under the high-rate thickeners, versus 8m of material from under the deep cone thickeners; we can use the previous total area calculations to produce a very rough comparison in Table 10.
Clearly the assessment is simplistic (perhaps absurdly so). But it illustrates the point: As the cost of groundwork increases, high-density and/or paste thickeners become more cost competitive. At the higher end of our example, the earthwork more than offsets the added cost of the thickener ($1,850,000 versus $1,800,000).
Footprint can additionally improve the CAPEX evaluation for deep cone and high-density thickeners when a CCD circuit is being installed in an existing mill, installed in a building, or when thickener roofs are required. Conversely, in active seismic zones, the taller tanks for paste and high-density thickeners may increase CAPEX, relative to high-rate thickeners.
A single high-density or paste unit in a CCD circuit may also have advantages. For example, it may be economically viable to use high-rate thickeners in all but the final stage. Such a configuration may be particularly attractive for existing installations where adding another stage with paste or high-density technology is justifiable, but replacing the entire circuit is not cost effective.
Considering Tailings Management
A hybrid configuration as outlined above can provide the lowest CAPEX, improve recovery, and lower costs for the tailings impoundment.
Tailings management can become a complex part of the CCD evaluation. Many studies have been done regarding tailings deposition (e.g. Fitton & Roshdieh 2013). It takes little reading to recognize that high level estimates, such as we have employed thus far in our discussion, are of little value in that arena. Site conditions and specific customer needs drive huge differences in outcomes.
Therefore, we will simply note that the tailings management study may outweigh all other CCD CAPEX considerations.
From a process perspective alone, high-density and deep cone thickeners have a clear advantage over high-rate thickeners. Because of the wide variance in metal values between processes (e.g. copper leach versus gold/silver), we did not attempt financial assessment of recoveries with different thickener options.
Depending on site details, high-density or paste may require slightly more or less OPEX. CAPEX can be greater as well, but there are a number of site-specific conditions that make the purchase of paste or high-density thickeners more attractive.
The decision-making process need not be a difficult one. Although some information must be provided by the process owners (primarily the value of improved recovery when comparing one option to the other), the majority of the data necessary for the economic assessment can be obtained from suppliers, an engineering company, and/or contractor.
So determining if high-rate, high-density, or paste is in your CCD future is a fairly straight-forward evaluation.
Schoenbrunn, Fred & Laros, Tim (2002). Design Features and Types of Sedimentation Equipment. In Mular, Andrew L., Halbe, Doug N. & Barratt, Derek J. Mineral Processing Plant Design, Practice, and Control Proceedings (p. 1339). Littleton, Colorado, USA: Society for Mining, Metallurgy, and Exploration, Inc.
Mulligan, M.C. & Bradford, L. (2009). Soluble metal recovery improvement using high density thickeners in a CCD circuit: Ruashi II case study. In (Volume 109) The Journal of The Southern African Institute of Mining and Metallurgy (pp. 665-669). Johannesburg, South Africa: The Southern African Institute of Mining and Metallurgy.
Klepper, R. & McCurdie, P. (2011). Leach Residue and Pregnant Liquor Separation – Process and Capital Comparison of Counter-Current Decantation and Counter-Current Washing with Vacuum Filtration. In MetPlant 2011 - Metallurgical Plant Design and Operating Strategies (pp. 488-505). Victoria, Australia: The Australasian Institute of Mining and Metallurgy.
McCaslin, M.L. & Johnson J. (2016). Liquid-Solid Separation in Gold Processing. In Adams, Mike D. Gold Ore Processing: Project Development and Operations (pp. 289-291). Amsterdam, Netherlands: Elsevier B.V.
Emmett, R.C. & Dahlstrom, D.A. (1974). Comparison of Countercurrent Decantation and Filtration in Solubles Recovery. (p. 1) Littleton, Colorado, USA: Society for Mining, Metallurgy, and Exploration, Inc.
Fitton, T.G. & Roshdieh, A. (2013). Filtered Tailings versus Thickened Slurry: Four Case Studies. In Jewell, R.J., Fourie, A.B., Caldwell J. & Pimenta, J. Paste 2013 (pp. 275-288). Perth, Australia: Australian Centre for Geomechanics.