Minerals Engineering, Vol. 13, No. 14-15, pp.

1533-1542, 2000
Pergamon © 2000 Elsevier Science Ltd
All rights reserved
0892--6875(00)00136-9 0892-6875/00/$- see front matter

CALCULATION OF LOCKED CYCLE FLOTATION TEST RESULTS

G.E. AGAR

University of Toronto, Lassonde Program in Mineral Engineering, Rm 119,

170 College St., Toronto M5S 3E3, Canada. E-mail [email protected]
(Received 23 April 2000; accepted 21 September 2000)

ABSTRACT

Locked cycle test are frequently used in mineral processing laboratories during the flowsheet
development exercise in order to confirm batch results and to get an approximation of a continuous
circuit material balance. A review of the literature indicates that there are at least four different
methods for calculating the results of a locked cycle test. Only two of these methods offer the
opportunity to examine whether or not the test reached steady state and only the cycle-by-cycle
method intrinsically calculates the circulating streams. Because of the difficulty and expense of
executing locked cycle tests and the frequent failure of the tests to reach steady state it should be of
paramount importance to determine whether or not the test reached steady state; thus, the preferred
method of calculation is the cycle-by-cycle technique. © 2000 Elsevier Science Ltd. All rights
reserved.

Keywords
Mineral processing; froth flotation; mass balancing

INTRODUCTION

An experimental simulation of a continuous circuit that utilizes repetitive batch tests in a cyclic manner is
universally referred to in the mineral industry as a locked cycle test (LCT). These tests are begun by doing
a complete batch test in the first cycle then a second cycle is done by adding the material from the first
cycle to the appropriate location in the second cycle. Batch tests are continued in this iterative manner for
an arbitrary number of cycles then the terminal products are dried, weighed and subjected to chemical
analysis. Usually only one concentrate and one tails are collected from each cycle and because the
intermediate tailings are recycled only after the final cycle are the recycling intermediate products available
for analysis. This type of test is commonly used in flotation investigation but is also used in grindability
determinations (Michaelson 1945, Yancey 1933), bailing (Sastry and Fuerstenau 1975) and in leaching
tests.

Surprisingly there are at least four methods for the representation of the results from flotation locked cycle
tests, obviously not all of equal technical merit. This presentation examines these methods and does a
critique of each. The grindability and bailing test procedures are not executed in the same manner as the

1533
1534 G.E. Agar

flotation tests. This is due to the simple fact that the product from each cycle of the grindability or balling
test is evaluated immediately and the results are used to determine the conditions of the next cycle. By
contrast, the flotation locked cycle test can only be fully evaluated after the entire test has been completed.
Only the flotation test calculations are discussed in this text.

BACKGROUND

One of the first recorded reports of a locked cycle test in flotation investigations is that of Dietrich et al.
(1934). Taggart (1945) referred to this report in the Handbook of Mineral Dressing in a rather derogatory
manner. "It is questionable whether in any case it approximates mill results any more closely than the
standard batch test". However, Taggart offered no alternative.

Balance based on the two-product formula

Macdonald and Brison (1962) also give an account of locked cycle tests where they show their calculations
for such a test. Their account indicates some of the uncertainty regarding the calculation of the results. The
fact is that their calculations are erroneous. They calculate the recovery in each of the cycles from what is
known as the two-product formula. The two-product formula is a simple material balance that applies to a
single batch test or a steady state continuous circuit that produces only two products, a concentrate and a
tails. Clearly, the first cycle of a locked cycle test meets neither of the criteria for the application of the
two-product formula. Since they had the weight and the analysis of the feed, the concentrate and the tails
from each cycle the calculation of the recovery should have been straightforward. The discrepancy between
their recovery calculations and the correct calculation is shown in Table 1.

TABLE 1 Recovery per cycle calculated by Macdonald and Brison

Cycle 1 2 3 4 5 6 7
M &B 94.8 92.9 91.1 89.9 91.2 91.2 91.2
Correct 67.0 87.5 90.1 87.9 86.4 87.4 89.8

One aspect that Macdonald and Brison did correctly was to calculate the magnitude of the circulating load.
As will be shown in this case and others, the magnitude of the circulating load is a good indicator of the
stability of a locked cycle test. The danger is that with the experimental conditions chosen that the test will
not reach a steady state, a circumstance that usually indicates that the circulating load is increasing with
each cycle or vice versa. The data from Macdonald and Brison is shown in Figure 1.

140

.E

120

n,,

100
1 2 3 4 5 6 7

Fig. 1 Recycle weight for test of Macdonald and Brison.

This figure shows that the recycle weight may have stabilized at the sixth cycle and therefore this test may
have produced valid results; however, this simple examination was not included in the original text.
Clearly, the method used by Macdonald and BrisOn for calculating the locked cycle test results is
inadequate if not actually misleading.
 Calculation of lockedcycleflotationtest results 1535

Every cycle yields 100%

A second method of calculation assumes that the output of each cycle without regard to the recycle is
100%, This method cannot possibly be correct because of the lack of accountability of the recycle that is
beginning to accumulate with the first cycle. An interesting variant of this method is used by De Cuyper
(1987). Although De Cupyer attempts a cycle-by-cycle balance he ignores the recycle material until the
very end. He calculates what he calls a "fictive feed grade" by adding up the products from each cycle
without the recycle. The recovery in each cycle is based on this fictive feed grade and the weights and
analyses of the products actually removed from the cycle. He then plots this fictive feed grade against this
fictive recovery. A plot that he produces is shown in Figure 2.

e-.
--,, 2.9
6O

"-" 2.4
-g
r3 1.9

~ 1.4

Lr_. 0.9
70.0 80.0 90.0 100.0

Fig.2 De Cuyper plot of locked cycle test results.

In this case the feed grade was 1.93% and De Cuyper claims that when the "fictive feed grade" equals the
actual feed grade the test has reached steady state. Although this representation appears to be related to
Brownian motion maybe in cycles 7, 8 and 9 the test did approach his criterion for steady state. As often
happens cycle 10 seems to have gone off track. In this method as in other variants of the calculation of the
material balance the sum of the distribution of all elements in the products discharged from each cycle is
attributed 100%. Since recycle is generated in the first cycle it is totally impossible for the distribution in
the concentrate, and tails from this cycle to be 100% of what was in the feed. In the second and all
succeeding cycles, recycle from the previous cycle is added along with a new batch of feed. It is
conceivable, although highly unlikely, that the test could achieve a steady state in the second cycle and the
distribution of material in the products may equal 100% of that material in the new feed batch. Depending
on the complexity of the circuit, the robustness of the circuit design and the experimental skill exercised a
locked cycle test will usually require at least four cycles to reach steady state and may require as many as
twelve cycles. An estimate of the number of cycles required to reach steady state can be obtained from a
numerical simulation of the material balance because the numerical simulation emulates the locked cycle
test but does not have experimental error (Agar and Kipkie 1978)

Finally, there is always the possibility that the test will never reach steady state.

Combined products technique

Lakefield Research and others (Coleman 1978, MacDonald et al. 1985, Chong et al. 1999, Tan and Wright
2000) use this method of calculating the material balance for a locked cycle test but its origin is unknown.
In this technique all the products are weighed and analyzed, the metal units are calculated for each
individual product then the distribution of the elements into the respective products is accumulated. Thus,
in Coleman's example, the lead concentrates for all six cycles executed are summed together as are the six
zinc concentrates and then the six zinc rougher tails (the final tails for the circuit). Although the approach to
steady state can be monitored with this calculation routine neither Coleman nor MacDonald et al. (1985)
made the examination. To determine whether or not the test approached steady state the sum of each of the
components out of each cycle must be calculated without regard to the recycle streams. For example, in the
Coleman test the distributions of the weight, the silver, the lead, and the zinc in the products from each
cycle is calculated based on the average feed per cycle. It is important that the feed to each cycle is the
same; i.e., the feed to each cycle must be identical in weight and mineral composition, a normal
requirement of sampling. In the first cycle the material that went into the recycling streams is unaccounted
1536 G.E. Agar

therefore the total cannot equal one hundred per cent; however, as the test progresses the accountability will
increase until hopefully all of the feed material is accounted for in the final products. Such an accounting is
shown in Table 2.

Clearly, this test did not reach steady state during the six cycles performed and projections based on this
incomplete test would be of dubious value. This type of examination reveals that the test did not reach
steady state but it does not show clearly the accumulation of material in the circulating load. It is possible
to follow this accumulation by summing the unaccounted constituents as the test progressed. For instance
for the data in Table 2 the unaccounted zinc in cycle A is (1- 0.7284) 0.2716. In cycle B a further 0.1514
(1-0.8486) of the zinc is added to the recycle streams and so on. Thus, the accumulation of each of the
components could be followed but no example of this calculation was found in the literature.

TABLE 2 Accountability in Coleman's test

Accountability
Cycle Weight A~ Pb Zn
A 0.738 0.598 0.597 0.728
B 0.915 0.936 0.937 0.849
C 0.916 0.936 0.980 0.850
D 0.947 0.885 0.906 0.935
E 0.979 0.907 0.865 0.982
F 0.974 0.956 0.947 0.972

While a poorly designed circuit may come to steady state eventually, if it equilibrates with a large
circulating load it would be difficult or impossible to operate in a real concentrator. Too many of us are
familiar with the cyclic filling and overflowing of sumps in concentrators, a problem that is contributed to
by inadequate circuit design. It is a fascinating observation that the result that Coleman used are reproduced
by MacDonald et al. (1985), perhaps an indication of the paucity of acceptable data available from locked
cycle tests.

Cycle-by-cycle balance

The calculation of the balance on a cycle-by-cycle is relatively straightforward. The major stipulation is
that, like the combined-products method of calculation, the feed to each cycle must be the same weight and
must be true samples, i.e. they must be identical in mineral composition. Under ordinary laboratory
conditions this is not a severe restriction; indeed, it is a standard procedure in any well run laboratory. The
only limitation may be if the samples are taken from an operating circuit in which case some special care
must be taken to ensure that all the wet samples contain the same amount of dry solids.

The procedure begins with a calculation similar to the combined-products technique in which the analyses
of all products out of the test are combined to reconstruct the total feed to the locked cycle test. From this
sum the average feed per cycle is determined and this average feed is the basis for the subsequent cycle-by-
cycle balance; thus, the critical importance that all feed samples are alike. Once the average feed per cycle
has been calculated the distribution for the first cycle is readily determined from the sum of the products
and the remainder is attributed to the recycle. In all subsequent cycles the calculation is the same: the feed
to the cycle is the sum of the new feed and the recycle from the previous cycle. As in the first cycle the sum
of the products out of the cycle provides the basis for calculating the distribution and as in the first cycle
the remainder is assigned to the recycle. This iterative calculation is continued throughout the test until the
final cycle is reached. Since the circulating steams are analyzed from the final cycle there exists the
possibility of checking the recycle. If no arithmetic errors have been made then the calculated circulating
load from the final cycle will equal the actual circulating load.

This calculation of the material balance in this cycle-by-cycle mode provides the opportunity to examine the
approach to steady state by plotting the material accounted for from each cycle. Obviously, in the first cycle
some of the material has to go into the circulating streams so the accountability will be less than one
hundred per cent. As the test progresses the circulating steams should reach a steady value and the sum of
the products should reach one hundred per cent. A plot of the distribution of the sum of the elements
 Calculation of locked cycle flotation test results 1537

accounted for out of each cycle is one of the means by which the approach to steady state can be examined.
An even more sensitive measure of the approach to steady state is provided by the circulating streams. A
simple plot of the weight of the circulating streams will determine whether or not the test came to steady
state.

Coleman provides a complete material balance for a relatively complex circuit and he made the following
statement..."If equilibrium has not been reached, it cannot be assumed that the intermediate products will
end up in the desirable concentrates in the ratio indicated by the overall test because some of these
products may not be upgradeable". Despite this reasonable rhetoric Coleman made no attempt to determine
if equilibrium or steady state had been reached. The results of Coleman's test have been calculated using
the cycle-by-cycle technique (see Appendix 1) and the appropriate plots are shown in Figures 3 and 4.
Figure 3 shows the sum of the elements accounted in each cycle. Figure 4 shows the corresponding values
for the circulating streams.

1.00

0.90
t-

.o
~Wt
0.80
---II- Ag
£3
--A.- Pb
0.70
X Zn
Cy~;~mber
0.60 ~"
1 2 3 4 5 6

Fig.3 Approach to steady state for Coleman's test results.

Clearly none of the elements nor the total weight in the products equaled the feed in any of the cycles and
there is only a hint that with continued cycles that the accountability would have improved. This suggests
that the circulating load was accumulating during the six cycles of the test. The recycle data has been
plotted in Figure 4 to examine this aspect in greater detail.

1 ~

1 2 3 4 5 6

Fig.4 Recycle in Coleman's test.

The plots in figure 4 confirm the suspicion that the circulating load was increasing as the test progressed. In
fact by the sixth cycle there was between 50 and 80% of the material in the circulating load. Clearly this
circuit was poorly designed and would be disastrous in a continuous circuit. The most significant
1538 G.E. Agar

observation is that none of this information was revealed by the combined-products method of calculation
as used by either Coleman or MacDonald et al.

In the report of Chong et al. there are two locked cycle tests for which results are given. The first of these
(Table 6 in their paper) had a different weight to the first cycle than to other cycles. Perhaps the feed weight
was not controlled to any of the cycles, which makes the interpretation of the results very difficult and
perhaps impossible. The second test result given in Table 7 of their report apparently had a controlled
amount of feed to each cycle. The circuit for this test appears to be quite simple: a gravity separation that
yielded what they identified as a pan concentrate, then a rougher-scavenger flotation on the remainder. The
scavenger concentrate was sent to a cleaner with the cleaner concentrate joining the rougher concentrate to
form the final concentrate. The scavenger-cleaner tails returned to the scavenger feed. The data plotted in
Figure 5 indicates that this circuit stabilized very quickly and that the accountability of gold was close to
one hundred per cent even from the beginning of the test. The only unusual item in this test was the
negative distribution of gold in the recycle. Probably that was due to the difficulty of collecting and
analyzing the free gold in the pan concentrate. It is significant that this examination of the approach to
steady state was not included in the paper by Chong et al.

E
0

"E:

Cycle Number

Fig.5 Approach to steady state for the test of Chong et al.

A more recent example of a test that was evaluated only by the combined-products technique is that of Tan
and Wright. Their data has been recalculated using the cycle-by-cycle technique and the approach to steady
state has been examined by plotting the distribution of elements and weight as a function of cycle number
in Figure 6.
1-'-

f,

[]

Fig.6 Approach to Steady State for the Results of Tan and Wright.
 Calculation of locked cycle flotation test results 1539

These data show that this test produced reasonable results except for the regular increase of the circulating
load as the test progressed. The interesting negative distribution of gold in the recycle in the first two cycles
is probably due to the difficulty of assaying the high grade gravity concentrate. As with Chong's data this
analysis of the approach to steady state was not presented by Tan and Wright.

The results of a locked cycle test on bulk matte from the INCO smelter that used a four stage counter-
current flowsheet was published in 1996. (Agar et al). The complete material balance is shown in Appendix
2 and the approach to steady state is shown in Figures 7a and 7b.

14t't /'l

1
v
e-

.(2_
"5
"t-

C3

1 3 5 7 9
Cycle Number

800
700
600
I

v 500
t-
400
300
200
100
0
1 3 5 7 9
Cycle Number

Fig.7a (top) and b (bottom) Approach to steady state for matte flotation (Agar et al. 1996).

These results demonstrate that this test readily reached steady state. The calculation of the results on a
cycle-by-cycle basis permits this examination of the approach to steady state that other methods of
calculation do not facilitate.

DISCUSSION

Locked cycle tests are used in the mineral industry to generate an experimental simulation of a continuous
circuit. These tests have been used for many years and are still commonly employed to authenticate the
choice of a circuit. The tests are costly because they require considerable time to execute and there is a
substantial analytical component required. There are methods that can be used to enhance the likelihood
that the test will reach steady state but they are not popular despite the potential improvement in the actual
locked cycle test and economy of effort that they offer.

The attainment of steady state in a locked cycle test (experimental simulation of a continuous circuit) is
critically important when evaluating the utility of the results. Randomly designed circuits are not
1540 G.E. Agar

guaranteed to reach steady state or they may reach an unsatisfactory steady state, in either case the results
are of little use. Thus, the examination of the approach of the test to steady state should be the first step in
the evaluation. Such an examination is facilitated by what is designated here as the cycle-by-cycle
calculation technique. In fact, of the four calculation schemes that have been used for locked cycle tests
only the cycle-by-cycle method includes the calculation of the recycle streams as part of the calculation. A
review of the few published results reveals that some of those did not reach steady state and this was
unknown to the people who published them. Utilization of the cycle-by-cycle calculation procedure would
have avoided this uncomfortable situation.

CONCLUSIONS

Four methods for calculating the results of a locked cycle flotation test have been examined and it has been
shown that only one of these, the cycle-by-cycle technique, intrinsically includes the calculation of the
circulating streams which facilitates the examination of the approach to steady state. Since the viability of
the locked cycle test results is so critically dependent upon having reached steady state, the cycle-by-cycle
technique is regarded as the most appropriate method for calculating the balance.

ACKNOWLEDGEMENTS

Mike Ounpuu read a draft of this paper and made many valuable suggestions that improved the text greatly.
Dr. G. Tan graciously provided data on the recycle streams that were not included in his original paper.
This help is greatly appreciated.

REFERENCES

Agar, G. E. and Kipkie, W.B., Predicting Locked Cycle Flotation Test Results from Batch Data CIM
Bulletin 1978, Vol 71, No. 824, pp 140-147
Agar, G. E., Khan, F., Markovich, B., Mukherjee, A., Shea, B. and Kelly, C. Laboratory Flotation
Separation of INCO Bulk Matte Minerals Engineering, 1996 Vol. 9, No. 12 pp 1215-1226
Coleman~ Richard L., Metallurgical Testing Procedures Chapter 9 in Mineral Processing Plant Design
Society of Mining Engineers N.Y. 1978 ed. A. L. Mular and R. Bhappu
Chong, T.Y, Folinsbee, J.A., Robitaille, A., Odo, T. and Wickens, G., Flotation Improvements at Placer
Dome's Campbell Mine, CIM Bulletin 1999, Vol. 92, No 1032, pp 60-70
Dietrich, W.F., Engel, A.L., and Guggenhiem, M., Ore Dressing Tests and Their Significance, USBM RI
3328, 1934
De Cuyper, J.A., (1987) Mineral Processing Flowsheet Development, in Mineral Processing Design, Ed.
Yarar and Dogan Martinus Nijhoff Pub.
MacDonald, R.D and Brison, R.J., Applied Research in Flotation, Chapter 12 in Froth Flotation AIME,
N.Y. 1962 ed. D.W. Fuerstenau
MacDonald, R.D., Hellyer, W.C., and Harper, R.W., Process development testing in SME Mineral
Processing Handbook, 1985, pp.30-117, ed. N. L. Weiss
Michaleson, S.D., Determination of Ball-mill Size from Grindability Data, AIME Tech. Pub. No. 1844,
Class B, Milling and Concentration, 1945
Sastry, K.V.S. and Fuerstenau, D.W., Laboratory simulation of closed-circuit bailing drum operation by
locked-cycle experiments, Trans. SME 1975, vol. 258, pp 335-340
Taggart, A.F., Handbook of Mineral Dressing 1945, pp. 19-181, John Wiley and Sons, New York
Tan, K. Ghie and Wright, F. (2000) Metallurgical Testwork on Harmony Project Ore Proc. CMP 2000, pp.
363-374
Yancey, H.F., Furse, O.L. and Blackburn, R.A., (1933) Trans. A1ME, Vol. 108,267

Correspondence on papers published in Minerals Engineering is invited by e-mail to

bwills @min-eng.com
 Calculation of locked cycle flotation test results 1541

APPENDIX 1

Assays (g/t or %) Diskibution

cycle # Prod. Wt (g) Ag Pb Zn Wt Ag Pb Zn
1 Pb Conc 27.8 31.04 72.06 5.46 0.028 0.446 0.508 0.018
Zn Conc 104.6 1.62 0.99 54.18 0.107 0.088 0.026 0.679
Zn Ro TI 590.0 0.21 0.42 0.45 0.603 0.064 0.063 0.032
Total 722.4 1.60 3.26 8.42 0.738 0.598 0.597 0,729
Recycle 256.6 3.03 6.20 8.84 0.262 0.402 0.403 0.271

2 Pb Conc 56.2 24.89 58.33 8.21 0.057 0.724 0.831 0.055

Zn Conc 113.8 1.57 0,92 55.37 0.116 0.092 0.027 0.754
Zn Ro TI 726.0 0.32 0.43 0.45 0.742 0.120 0.079 0.039
Total 896.0 2.02 4.12 7.91 0.915 0.936 0.937 0.849
Recycle 339.5 2.65 5,42 10,40 0.347 0.465 0.467 0.423

3 Pb Conc 59.8 24.79 56.63 8.95 0.061 0.767 0.858 0.064

Zn Conc 116.3 1.51 0.91 53.47 0.t19 0.091 0.027 0.745
Zn Ro TI 721,0 0.21 0.52 0,48 0.737 0.078 0.095 0.041
Total 897.1 2.02 4.31 7.91 0.916 0.936 0,980 0.850
Recycle 421.4 2.43 4.56 11.35 0.430 0,529 0.486 0.573

4 Pb Conc 54.6 25.11 55,90 8.25 0.056 0.709 0.774 0.054

Zn Conc 131.8 1.40 1.03 52.56 0.135 0.095 0.034 0.829
Zn Ro TI 741.0 0.21 0.52 0.56 0.757 0.081 0.098 0.050
Total 927.4 1.85 3.85 8.40 0.947 0.885 0.906 0.933
Recycle 472.9 2.63 4.85 11.30 0.483 0.644 0.581 0.640

5 Pb Conc 54.9 23,06 51.90 8.74 0.056 0.655 0.722 0.057

Zn Conc 140,9 1.73 1.18 51.46 0.144 0.126 0.042 0.868
Zn Ro TI 763.0 0.32 0.52 0.62 0.779 0.126 0.101 0.057
Total 958.8 1.83 3.56 8.56 0.979 0.907 0.865 0.982
Recycle 493.1 2.89 5.73 11.14 0.504 0.736 0.716 0.657

6 Pb Conc 57.8 23.82 54.30 8.14 0.059 0.712 0.795 0.056

Zn Conc 130.7 1.73 1.24 55.04 0.134 0.117 0.041 0.861
Zn Ro TI 765.0 0.32 0.57 0.60 0.781 0.127 0.111 0.055
Total 953.5 1.94 3.92 8.52 0.974 0.956 0.947 0.973
Recycle 518.5 2.91 5.85 11.03 0.530 0.781 0.769 0.685

Pb Scav Conc 73.6 2.48 3.60 12.02 0.075 0.094 0.067 0.106
Pb 1st CI TI 108.1 3.66 7.18 13.18 0.110 0.205 0.197 0.171
Pb 2nd CI TI 28.3 5.92 19.10 13.04 0.029 0.087 0.137 0.044
Pb 3rd CI TI 28.0 7.98 23.04 13.11 0.029 0.116 0.164 0.044
Pb 4th CI TI 16.1 9.59 21,93 12.83 0.016 0.080 0.089 0.025
Zn Scav Conc 53.4 1.08 1.36 4.02 0.055 0.030 0.018 0.026
Zn 1st CI TI 168.5 1.29 1.51 8.49 0.172 0.112 0,064 0.171
Zn 2nd CI TI 21.5 2.38 3.05 14.10 0.022 0,026 0.017 0.036
Zn 3rd CI TI 21.0 2.80 2.90 24.71 0.021 0.030 0.015 0,062
1542 G. E. Agar

APPENDIX 2

BULK MA'I-rE LCT 13348

Assays (%) Distribution (%)
CYCLE Wt-(g) Cu Ni Fe S Wt Cu Ni Fe S
1 Cu Conc 454.2 74.9 2.2 0.1 20,1 35.2 62.2 2.3 15.1 31.7
Ni Conc 533.5 4.7 70.3 0.4 25.0 41.3 4.6 85.8 61.2 46.2
TOTAL 987.7 37.0 39.0 0.3 22.8 76.5 66.8 86.0 76.3 77.9
*Cir Ld 303.4 59.8 17.3 0.3 21.0

2 Cu Conc 672.5 75.7 3.2 0.1 20.3 52.1 93.1 4.9 28.4 47.3
Ni Conc 551.7 5.4 67.7 0.4 24.8 42.7 5.5 85.4 64.9 47.4
TOTAL 1224.2 44.0 32.3 0.3 22.3 94.8 98.5 90.3 93.4 94.8
*Cir Ld 370.4 51.2 25.7 0.3 21.3

3 Cu Conc 675.4 73,9 3.6 0.2 20.1 52.3 91.2 5.5 32.6 47.1
Ni Conc 617.0 8.3 65.7 0.4 24.7 47.8 9.3 92.7 70.8 52.8
TOTAL 1292.4 42.6 33.2 0.3 22.3 100.1 100.6 98.2 103.4 99.9
*Cir Ld 369.1 50.5 27.9 0.2 21.4

4 Cu Conc 654.9 74.3 3.4 0.1 20.0 50.7 89.0 5.1 25.7 45.4
Ni Conc 643.4 9.3 65.0 0.4 24.8 49.8 11.0 95.6 73.8 55.3
TOTAL 1298.3 42.1 33,9 0.3 22.4 100.6 99.9 100.7 99.5 100.7
*Cir Ld 362.0 51.7 27.6 0.3 21.3

5 Cu Conc 631.0 75.6 3.0 0.1 20.1 48.9 87.2 4,4 19.0 44
Ni Conc 688.4 13.1 61.8 0.4 24.6 53.3 16.5 97.3 72.7 58.7
TOTAL 1319.4 43.0 33.7 0.2 22.5 102.2 103.7 101.6 91.8 102.7
*Cir Ld 333,7 50.0 27,8 0.4 20.7

6 Cu Conc 646.2 75.3 3.3 0.1 20.3 50.0 89.0 4.9 27.3 45.5
Ni Conc 645.1 9.7 64.6 0.4 24.7 50.0 11.4 95.3 72.0 55.2
TOTAL 1291.3 42.5 33.9 0.3 22.5 100.0 100.3 100.2 99.4 100.7
*Cir Ld 333.6 49.4 27.5 0.4 20.1

7 Cu Conc 654.2 74.1 3.4 0.2 20.1 50.7 88.6 5.1 29.6 45.6
Ni Conc 638.1 9.3 64.6 0.4 24.5 49.4 10.9 94.2 77.0 54.2
TOTAL 1292.3 42.1 33.6 0.3 22.3 100.1 99,5 99.3 106.7 99.8
*Cir Ld 332.4 50.4 28.6 0.3 20.4

8 Cu Conc 648.6 75.6 3.5 0.1 20.1 50.2 89.6 5.2 27.4 45.2
Ni Conc 663.1 10.1 64.2 0.4 24.6 51.4 12.2 97.3 74.1 56.6
TOTAL 1311.7 42.5 34.2 0,3 22.4 101.6 101.9 102.5 101.5 101.8
*Cir Ld 311.9 50.5 26.9 0.3 20.1

9 Cu Conc 651.3 75.3 3.5 0.1 19.8 50.4 89.7 5.3 23.6 44.7
Ni Conc 598.6 9.3 64.8 0.4 24.4 46.4 10.1 88.7 68.7 50.6
TOTAL 1249.9 43.7 32.9 0.2 22.0 96.6 99.8 93.9 92.3 95.4
3rd Clnr T1 88.9 63.2 14.3 0.2 20.4 6.9 10.3 2.9 6.4 6.3
2nd Clnr TI 104.7 52.1 24.4 0.3 20.9 8.1 10.0 5.8 10.7 7.6
1st Clnr TI 159.5 29.9 45.2 0.4 22.6 12.4 8.7 16.5 16.8 12.5
Cir Ld 353.1 44.9 31.3 0.3 21.5

Av Fd I Cycle = 1291.1 42.4 33.9 0.3 22.3 * = Calculated values