|
|
Pilot Plants Look Different
In
order for pilot plants to work and extract metals, total residence times must be
about the same as full-scale plants. The
chemistry cannot be neglected. Typical
residence times are 2-3 minutes for the pumper and auxiliary mixing stages.
Equations 7 and 8 must also be adhered to. Example 6:
The pilot plant in Example 1, which was a true scale-down of the other
examples, only had a mean residence time of 9.7 seconds.
It is recommended to have a mean residence time of at least 30 seconds in
the pumper stage to achieve the desired mass transfer.
How? To do this, the flow rate must
be reduced. This can be done be
reducing the pumper speed or the pumper diameter.
Since the tank diameter is 0.305 m (12”) and Z/T=0.8 the volume is
V=22.7 liters. If the residence
time is to be 30 seconds the flow rate will be Q=0.757 L/s (12 GPM).
Based on Figure 6, the optimum efficiency (e=21.2%)
of a R300
with DO/D=0.33 is at Nq=0.065. With
D=0.152 m (6”) and Equation 1, N=197 RPM. From Figure 5, Nh(Nq=0.065)=0.705. Thus, using Equation 3, H=0.0892 m (3.51”) and using
Equation 2, P=3.13 Watts. The power
dissipation is only P/V=0.14 kW/m3 (0.69 Hp/1000 gallons).
The operating range according to Equations 7 and 8 is 0.52 kW/m3
< P/V < 0.83 kW/m3. Therefore,
there would be a lot of phase separation if the pilot plant ran at this
condition. This scenario is typical for pilot plants. In the past, the solution was to make the pumper as inefficient as possible. In other words, the pumper was designed to waste energy to achieve an operating condition, which could maintain a uniform dispersion with the right amount of residence time. Three methods were most common:
All
of these methods have an unfortunate drawback.
Because the power is increased using these methods, the hydraulic
efficiency is dramatically decreased and the concentration of fines increases,
as does the entrainment levels. To
fix this problem, optimum pilot plants have a minimum of two impellers in the
pump stage. The lower impeller is designed to do the pumping and the
upper impeller(s) are designed to do the mixing at the lowest possible power.
This method also works for full-scale plants with high Z/T as for example
at Sociedad Minera Cerro Verde in Peru [5]. The
upper impellers are the same type as those found in the auxiliary stages.
Over time, up-pumping A310s have proven to be ideal for this.
As a general rule of thumb, they require 1/10th of the power
of the pumpers to maintain a uniform dispersion.
They generate about 3-13 times the flow rate of the pumpers at very
little shear [3-6]. These need to
be designed in an optimum operating range, too.
The optimum P/V-range is not as easy to describe as the pumpers, since
now the distance to the surface, COV, is very important.
If two auxiliaries are used, the spacing, S, between the two impellers,
is used instead of COV for the lower one. The
following correlations can be used as guidelines.
These equations are valid for 0.18 < OB/D < 0.6, 0.5 < COV/D < 2.3, and 0.2 < COV/T < 0.9. Assuming that Z/T=0.5, D/T=0.375, COV/D=1.11, the operating range is 0.056 < P/V < 0.065 kW/m3. More power is required to keep a stable dispersion the taller the tank, the higher the coverage and the longer the residence time.
Designing the
Pumper Stage with Multiple Impellers
To overcome the problem when P/V is too low at optimum hydraulic efficiency, the pump stage is conceptually split into two or more zones. Figure 7 shows a schematic of this concept with 2 zones, each having one impeller. The pumper is designed to operate in the lower zone and the auxiliary in the upper zone. The boundary that splits the pumper stage in two zones is generally the location of the auxiliary impeller. The height of the lower zone is ZP and ZI is the height of the upper zone. In the case of a dual impeller pumper stage, COV = ZI. The
design that optimizes the hydraulic efficiency has already been done in Example
6. The problem was that P/V was too
low to maintain a stable dispersion. It
was only 0.14 kW/m3. The
optimum range was 0.65 kW/m3 < (P/V)Opt <
1.09 kW/m3. The
solution of Example 6 showed the power of the pumper to be PPumper=3.13
W. An iterative process begins to
find the optimum volume. The
procedure is to assume a ZP. Then
calculate ZP/T, which is the Z/T for the pumper zone.
Then calculate VP, which is the volume of the pumper zone.
For a square pump box, VP = T2*ZP.
The mean residence time, qRes,P,
of the pumper zone is now equal to the volume of the pumper zone divided by the
flow rate (0.757 L/s). The P/VP
of the pumper zone is then PPumper/VP.
Now, using Equations 7 and 8, determine the upper and lower range of P/V.
If PPumper/VP is within the range of operating P/V,
the assumption of ZP was okay. If
not, another ZP needs to be tested until PPumper/VP
is acceptable. A spreadsheet comes
in very handy now. For
this example, a possible solution is ZP = 0.113 m (4.4”). ZP/T=0.37.
VP=10.5 L. qRes=13.9
s. PPumper/VP=3.13/10.5
= 0.30 kW/m3. The
operating range is 0.29 kW/m3 < (P/V)Opt < 0.43 kW/m3. The
next step is to determine the auxiliary, upper zone of the pump box.
The A310 up-pumping auxiliary impeller is located at ZP. Similar
to the pumper zone, the first step is to determine ZI, ZI/T,
VI, qRes,I,
and PI/V of the auxiliary zone.
The flow and power numbers, 0.56 and 0.30, respectively, of the A310 are
standard and are not affected by the net flow through the pumper box [9].
Initially assume D/T=0.375 for the A310.
ZI=Z-ZP = 0.13 m (5.15”). ZI/T=0.43. COVI=0.13
m. COVI/D=1.15.
S/D= 0.99, VI=12.2 L. qRes=16.1
s. PI/VI=0.21/12.2
= 0.0172 kW/m3. Now
using Equations 9 and 10, the operating range is 0.075 kW/m3 < (P/V)Opt
< 0.093 kW/m3. Clearly,
PI/VI is much less than the lower operating range.
This would result in the disengagement of the two phases. Next
step is to either try a larger A310 impeller, or to see if a second auxiliary
impeller is needed. The A310s
cannot get too large. With a S/DI=0.99,
there is not enough room for another impeller.
If the level was higher, a second auxiliary should be examined. Increasing
the diameter of the A310 to 0.51T does work.
D=0.155 m (6.1”). COVI/D=0.84. COVI/T=0.86, S/D=0.73. PI/VI=0.97/12.2
= 0.080 kW/m3. Now
using Equations 9 and 10, the operating range is 0.075 kW/m3 < (P/V)Opt
< 0.093 kW/m3. The overall mean
residence time is still 30 seconds. The auxiliary impeller creates 6.5 times the
flow of the pumper resulting in a recirculation time of just 2.5 seconds, which
is ideal for high mass transfer with low shear. Thus,
this pilot plant will have one pumper and one auxiliary impeller in the pump
stage. When planning for a pilot
plant, variable speed drives and impellers that can be adjusted on the shaft are
very important, since equations 7-10 are approximate.
If air entrainment is noticed, the impeller speed can be reduced and/or
the upper impeller lowered. If
puddles of organic are noticeable at the surface, the impeller speed can be
increased, and/or the upper impeller put closer to the surface.
This
technique has also been successfully used on more than 8 full-scale SX plants in
order to maximize efficiency such as Sociedad Minera Cerro Verde, Mexicana del
Cobre, Soquimich, Quiborax, IMC, Phelps Dodge, CS Metals and Cal Energy [6].
It is conceivable to have just one mixing tank, with one pumper and several auxiliaries instead of having multiple tanks with one impeller in each. It is also conceivable to have a pumper box with the smallest possible volume (large pump) and do all the mixing in the auxiliaries. Both methods would increase hydraulic efficiency by reducing the overall power consumption. Scale-up of the Auxiliary TanksClick this title to continue with this article. Click here to go to the first page of this article. Click here to go to the previous page of this article. |
|
Send mail to webmaster@postmixing.com with questions or comments about this web site. |