Energy Balances in the

Electrorefining and Electrowinning

of Cop per
T.B. Braun
Utah Refinery, Kennecott Minerals Company
Salt Lake City, Utah

SUMMARY boilers, and a package boiler located in the anode casting

Using basic heat transfer co"elations, the distribution of plant.
heat losses in a copper electrorefining tank house has been To facilitate the energy balance calculations, the tank
determined. The initial study led to programs to cover re- house was subdivided into ten subcircuits (Figures 1-10).
fining cells at both Kennecott refineries, with potential Subcircuits 1-4 constitute the commercial system, 5 and 6
energy savings of approximately 1 MM Btu/ ton cathode. the stripper system, 7 the slimes treatment system, and
Only partial implementation of cell covers has been pos- 8-10 the decopperization and nickel sulfate recovery sys-
sible to date at the Utah refinery due to water availability tems. The flows of solutionsand solids into and out of each
problems which developed. subcircuit are also indicated in the figures. Energy flow
The same heat balance approach can be used for copper and heat losses from cells and tanks have been calculated
electrowinning, although results differ significantly. In the according to the standard heat transfer equations listed in
design of a leach/ solvent extraction/ electrowinning facility, Table I. Flowrates, surface areas, weights, and temperatures
careful consideration must be given to the heat balance to were measured whenever possible; otherwise, the quantities
minimize cost while maximizing cathode quality. were estimated. All calculations were based on a 28-day
period, corresponding to the anode lifetime (i.e., two 14-day
INTRODUCTION cathode cycles).
With the rapidly rising costs of energy, increased emphasis Results
is being placed on minimizing energy requirements for the Overall flow and energy balances for each subcircuit are
electrorefining and electrowinning of copper. The distribu- indicated in Table 11. Balances could not be completed for
tion of heat losses at Kennecott's Baltimore refinery tank half of the subcircuits due to insufficient information on
house has been estimated and is discussed in terms of both
the general approach used and specific results obtained.
The finding that approximately 80% of all heat losses in
the tank house are due to evaporative, convective, and Table I: Heat Transfer I Heat Loss Equations l - 4 *
radiation losses from the tops of cells has led to a program
to cover cells at Kennecott's Baltimore and Utah refineries.
However, the covering of cells can lead to other operating Energy in Solution
problems such as a higher rate of loop corrosion and diffi- Streams
culties in maintaining the tank house water balance; these
Energy in Solid Streams q = M Cp(T-Tal
are also discussed.
While an electrorefining tank house requires the addition Electrical Energy Input q = 3.413 N I Vc
of energy beyond that supplied electrically to deposit copper,
an electrowinning tank house itself is energy self-sufficient Heat Loss from Cell
because of the much greater electrical power requirements. Side Walls
Nevertheless, proper design of the heating, cooling, or heat
exchange capabilities, in conjunction with solvent extraction Heat Loss from Vertical
and/or leaching operations, requires some knowledge of Tank Walls
temperatures, flow rates, required copper concentrations,
Heat Loss from Cell
and heat losses throughout the system. Factors influencing
Bottoms
the selection of an optimum design which maximizes cathode
quality and minimizes cost are illustrated with several
examples. Heat Loss from Gap
between Cells
ELECTROREFINING
Radiative Heat Loss q = euA(T:4-T~4l
General Approach
from Celll Tank Tops
The initial energy balance was conducted on the Kennecott
Refining Corporation tank house in Baltimore. The tank Convenctive Heat Loss
house is of conventional design, with cathode production from Celll Tank Tops
capacity of approximately 275,000 tons/ yr operating at a
current density of 20 A/ ft2. The plant includes electro- Evaporative Heat Loss
winning cells to control copper level in electrolyte, an evap- from Celll Tank Tops
orator for the production of nickel sulfate, and a circuit to
wash and dry anode slimes for subsequent processing at the Heat Loss from q = 0.0003 I N(Tc-Tal
Kennecott Utah Refinery. Steam for heating the tank house Oxygen Evolution
electrolyte is provided from waste he at boilers, auxiliary *See appendix for definition of terms.

JOURNAL OF METALS • February, 1981 59

 Table 11: Subclrcult Flow and Energy Balance Summary

Subcircuit Flow Balance (gal/ min) Energy Balance (MM Btu/ 28 Days)
Input* Output* Deviation, % Input* Output* Deviation, %
1 7700 7700 0 205,377 205,511 0.1
2 7737 7700 0.5 N/A N/A
3 107 107 0 2,198 1,954 11.1
4 N/A N/A N/A N/A
5 378 378 0 9,628 9,457 1.8
6 379 378 0.3 N/A N/A
7 15.9 15.3 3.8 3,239 3,232 0.2
8 70 70 0 2,264 1,999 11.7
9 70 70 0 N/A N/A
10 10.7 10.7 0 N/A N/A
*N / A: Insufficient information to complete overall balance.

steam and condensate flowrates. Since the main purpose of

the overall subcircuit balances was simply to aet as acheck .-_C~_C_EN_TM_T_ED_H_2_'0_,____ @
on the accuracy of the various heat loss estimates, the de- ~
@]
gree of closure of the balances was feIt to be sufficient. USED WATER FltOt!
tAlHODE "'ASN I MG
OECOPPERIZEO
ELECUOLYTE

Calculated heat loss rates in the four major tank house

systems are summarized in Table III. The evaporation losses
are, in actuality, the combined effects of evaporation, con- ".,
WASK WATER
".,
IUX""G

vection, and radiation. For particular tank house conditions WATER

and design, calculations show the following distribution of

"evaporation" losses: evaporation, 77%; convection, 12%; OILUTE ELEClRDL't'TE
radiation, 11%. Conduction losses might more properly be TO COHP1ERICAl STOAAGE

termed "conveetive" losses. Cell conduction losses combine o

three terms: vertical side walls, 55%; losses in the one-inch
gap between cells, 23%; losses from the cell bottoms, 22%. Figure 3. DIagram of subclrcult no. 3, wash water and mlxlng
Total heat losses for the four systems are also summarized tanks.
in Table III.

ELECIROl YTE TO COMEIICIAl

PD CEllS 0 ELECTROLYTE 0 OOf'IESTlC

e
I
H'
I
("THODES
ANO "00' SCM' G
FLASH STE,," STE""
COfrIDEMSER
@I

5TE"'" " CONOENSAT

FROM RECEIVERS HOT WATER
@]
".,
STORAGE
COMHERCIAl
tELL OVERflOW C[LLS CONOENSATE

o
~,6.T" Tn \lAST TANK HOT WATER TO
HE"1 BOI LEIIS CA THOOE WASH I NG
DTHER TANK HOUSE
USES

I SWtES TO (TI
UlKE UEATHENT PLANT
Figure 4. DIagram of subclrcult no. 4, condensate and hot
water storage tank.
Figure 1. DIagram of subclrcult no. 1, commerclal cells.

COHHERCIAl ELECTROLYTE
TO COHl4ERCIAl AHD
PD CE!lm 1
STIt,PPER AHODES ~
r-- I---

iI
CDM"'ERCIAL ANODE SCRAP ~ STlttPPER BLANKS ~

j
HEAD TANKS

ADDITION AGENTS "AR... ,"fE" j "Ee,"OL HE E!J

1
AHD HCl 0
COMERCIAl CELl 0 I1AKE uP
0
1
OVERflOW SOLUTION

STEAH [!J CEll OVERflOW STRIPPU

COMERCIAL COtI"'fRCIAl
~ CEllS

RECEIVER STORAGE
ONOENSATE [!] TANKS

)J-
T I BlEEO TO COIVtUC tAL AHO
SlIHES @]
Figure 2. DIagram of subclrcult no. 2, commercial head, re·
celver, and storage tanks. Figure 5. DIagram of subclrcult no. 5, stripper cells.

60 JOURNAL OF METALS • February, 1981

 ADDITION
-LAGENTS

ElECTROl'f'TE TO
STRIPPER CEllS
-
STRIPPER
I------
@J HEAD TANK
STRIPPER CEllS
DVR[FlOW' ~

WATER, H2S0,., HCl e!

_J
COOLING TOWER

§] STEMI TO llfE BLACK ACID TANK

STRIPPER
RECEIVER
"1-l STRIPPER
STORAG[
§] COND N AT TAA,
E\lAPORATER
DROP TANK

I -

Flgure 6. DIagram of subclrcult no. 6, stripper head, receiver, Figure 10. DIagram of subclrcult no. 10, NiSO, evaporator.
and storage tanks.

ANODE SliMES ANO ,CELL

Table 111: Calculated Heat Loss Rates
WASttIOLUTION

[TI AND[B MM % of % of
Btu/28 Total Total
Item Days System Plant
Commercial System (Subcircuits 1-4)
Cell Evaporation 36,196 89.0
SOLUTIONTO
SliMES
Cell Conduction 2,892 7.1
NATURAL GASIEJ
~LTERI
Receiver Tank Evaporation 927 2.3
DRY IliMEIT
Copper Products 346 0.8
uc. Receiver Tank Conduction 154 0.4
Flgure 7. DIagram of subclrcult no. 7, sllmes treatment. Hot Water Tank Conduction 70 0.2
Head Tank Conduction 38 0.1
Condensate Tank Conduction 23 0.1
Wash Water Tank Conduction 5 <0.1
Mixing Tank Conduction 2 <0.1
LEAD ANOOES
Subtotal 40,653 100.0 87.0
Stripper System (Subcircuits 5, 6)
Cell Evaporation 1,873 85.2
Cell Conduction 159 7.2
Receiver Tank Evaporation 72 3.3
Copper Products (and Blanks) 59 2.7
Receiver Tank Conduction 17 0.8
Head Tank Evaporation 10 0.4
Head Tank Conduction 8 0.4
Subtotal 2,198 100.0 4.7
ARSEN I tAL SLUOGE Slimes Treatment System (Subcircuit 7)
, CEll WASH ~
Heat Loss from Dryer 2,999 100
Subtotal 2,999 100 6.4
Flgure 8. DIagram of subclrcult no. 8, partial decopperlzation
cells. Decopperization and NiSO, System
Heat Losses (Subcircuits 8-10)
PD Cell Evaporation 332 37.9
Liberator Cell Evaporation 311 35.5
ADDITION AGENTS P 0 ELECTROL YTE [3
PD Tank Evaporation 62 7.1
BACK TO TANK HOUSE
Liberator Cell Conduction 49 5.6
TO TMk. HOUSE ~
PD Cell Conduction 36 4.1
LI8ERATOR Liberator Head Tank Evaporation 28 3.2
H[AO TANK
DlCO"U IlED CONOENS E CD Tank Conduction 16 1.8
ELECTROLYTE
STORAGE 8 Liberator Head Tank Conduction 12 1.4
PD Tank Conduction 11 1.3
C 0 ElECTItOL YTE ~ Intermediate Decopperization
TO TAHK HOUSE
Tank Evaporation 10 1.1
Intermediate Decopperization
C 0 EUtTROl YT
P 0 ElECTItOLYT
TAN'
Tank Conduction 4 0.5
STORAGE TANK
Copper products 2 0.2
PD Cell Oxygen Evolution 2 0.2
C 0 EUCTROL YTE
TO NI SO,. EVAPORATOR ~
Liberator Cell Oxygen Evolution 1 0.1
Subtotal 876 100.0 1.9
Flgure 9. DIagram of subclrcult no. 9, IIberator. TOTAL 46,726 100.0

JOURNAL OF METALS • February, 1981 61

 Examination of Table III indicates that heat losses from
the tops of cells and uncovered tanks ac count for approxi- Table IV: Heat Loss Rates Per Unit Area
mately 85% of the total tank house heat losses. In fact,
losses from the tops of commercial cells alone constitute
77% of the total, equivalent to a heat loss rate of 1.7 MM
Btu/ ton cathode produced. To determine where economic Itern $/ Yr/ft'"
energy conservation steps could be taken, the annual dollar Cell Walls 1.20
value of heat losses per unit area were calculated (Table IV). Cell Bottoms 0.30
This table indicates that while economic insulating measures Tank Walls 0.15-2.85""
could be implemented in most areas, both the relative and Cell and Tank Tops 20.75
absolute heat loss rates off the tops of commercial refining
cells dictate that they be given top priority. Also, since the
final results would not be significantly changed, it was not *Assume $3/MM Btu
""Varies depending on particular solution temperature
considered necessary to repeat the detailed study for the
Kennecott Utah Refinery, nor for any other refinery of con-
ventional design.
in fact, some operators claim that the infrared photographs
Cell Cover Implementation are easier to interpret when the covers are on.
The results of this study have led to programs to reduce A fourth consideration is the source of the steam used to
heat losses from the tops of cells at both Kennecott refin- heat the electrolyte. At Baltimore, steam is provided by a
eries. An early consideration was the choice of covering combination of two reverbatory-type anode furnace waste
material. The possibility of using plastic spheres like those heat boilers, two auxiliary boilers, and a small package boil-
used in electrowinning tank houses for acid mist suppression, er. Heat loss reductions through cell covers which reduce
was dismissed due to anticipated problems in handling the the demand on the auxiliary boilers or the package boiler
balls during removal ofanode slime from the cells. Both translate directly into energy savings whereas a reduction
polyethylene sheets and styrofoam boards were tested and of steam produced in the waste heat boilers would not. At
found to be effective for heat loss control, but polyethylene Utah, all steam is provided by an independent boiler plant;
has a tendency to melt on hot spots from short circuits, and thus, any steam demand reduction represents an energy
styrofoam is quite fragile. The material finally chosen was savings.
polyolefin, a. burlap-like material which is also used for cell A final consideration is the impact of reducing evaporation
covering at the Phelps Dodge EI Paso refinery. Polyolefin, rate on tank house water utilization and overall tank house
which costs less than 1cC/ft 2, was cut to length and attached water balance. At Baltimore, the water balances for heating
to a PVC pipe roller for ease of handling on each commercial and for the tank house are essentially separate systems
half-section (17 cells). . (Figure 11). Nearly all condensate produced in the tank
A second consideration in implementing a cell covering house heat exchangers is returned to the boiler plant as feed
program was the possible impact of covering on loop corro- to the boilers. The high quality of domestic water at Balti-
sion and subsequent failure of the loops connecting the more (Table V) allows for its direct use as tank house elec-
cathode starting sheet to the copper suspension bar. Exten- trolyte make-up. Water entering the tank house is distributed
sive testing showed that while the use of covers can aggravate into four major areas which enter the electrolyte system:
the problem of loop failures, these problems are not caused cathode washwater via the cathode washing machines; water
by the covers themselves. Rather, the temperature-depen- for anode scrap washing via the slimes treatment circuit;
dent corrosion processes will increase in rate if the electrolyte water for washing slimes out of the refining cells via the
temperature in the cells is allowed to increase. This will slimes treatment circuit; and other wash waters, including
occur if the cells are covered and the temperature of the direct water make-up, to maintain a constant overall elec-
electrolyte in the feed to the cells is maintained at the same trolyte volume. The three water loss routes are evaporation
level as when the cells are uncovered. Instead of adjusting losses, electrolyte bleed to nickel sulfate recovery, and other
the cell feed temperature up and down as the cells were un- miscellaneous losses of solution. As evaporation losses are
covered and covered, respectively, the approach taken at the reduced through cell covering, the water required for direct
two Kennecott refineries has been to ensure that the solution make-up is reduced accordingly. When direct make-up is
level is varied in a well-regulated pattern. At Baltimore this eliminated, further cell cover implementation can proceed
is accomplished with three riser levels (high, low, and out), only by improving wash water utilization. Other than by
while at Utah four riser levels (extra high, high, low, and directly reducing wash water usage, improved efficiencies
out) vary the solution level by about 1Vz inches overall. could be obtained by, for example, increased recycling of
Besides being designed to facilitate short clearing operations
early in the cathode cycle and to loosen deposits of gypsum
and copper sulfate late in the cycle, the riser schedules have
also been designed to maximize the difference between the
number of days a particular level has plated over and the
Table V: Domestle Water Analyses
number of days of corrosion at that level. The highest level,
of course, is never plated over consequently, the maximum
number of days at that level has been limited to three. Po-
tentialloop failure problems have been eliminated with these PPM
riser programs. Baltirnore Utah
A third consideration with a cell covering programis com- Refinery Refinery
patibility with tank house operating schedules and prac- Hardness 60 350
tices. Naturally, the covers must be removed when a section Bicarbonates 50 250
is being charged with anodes or when cathodes are being
Total Dissolved Solids 100 1500
pulled. In addition, the covers must be removed when short
clearing crews are working a section. The schedule adopted Calcium 20 70
at the Utah Refinery has been to cover each section two- Magnesium 5 40
thirds of each day with the remaining one-third uncovered Silica 5 50
and available for short clearing. Use of an infrared scanner Sodium 4 400
for short detection is not hampered by the polyolefin covers; Chloride 10 500

62 JOURNAL OF METALS • February, 1981

 Boilers
Blowdown Feed Water

Steam
Condensate Flgure 11. Water tlow diagram tor
Recycle
Kennecott Baltlmore retlnery.
T"House
I-
Hellt Elch_gers

Slime And Anode Sc rap Evaporation Lasses

Wash Via Filters
T_kHouse
T..k House
Wllter
Electrolyte
Oomestic Water Di stribution Cathode Washwater Nickel Sulfate
System System. Recovery

Oirect Make-up And Mise. Lasses

Other Washwater

F1
F3 Feed Water
Blowdown
Boilers

F2
Condensate

--
Steam F 4 f 5 Steam Recyc le
Flgure 12. Water tlow diagram tor
~ Kennecott Utah reflnery.
TnHouse
Heat Conden....
Exchanger.
:--- .---
Condensate F 6 F7 Condensate

Fg F 14
Slime Wash Via Silver
Fa Refinery Evaporation Losses

Domestic Water Tank House F10

Tank House
Anode Scrap Wash Via
Water Silver Refinery
F15
Electrolyte
Distribution Electrolyte Bleed
Fl1
Systems
F 13 System Cathode Washwater
Outside
C athode Wash F 12 F 16

Direct Make-up And Mise. Losses

Other Washwater

cathode wash water or using cathode wash water for washing
anode slimes.
At Utah, domestic water is drawn from deep wells located
a few miles from the Great Salt Lake, and water quality is
relatively poor (Table V). The water is unsuitable for direct
electrolyte make-up because of high calcium and chloride
levels. Consequently, essentially all condensate produced
from steam is used for electrolyte make-up, and domestic
water is used only when absolutely necessary. Since the tank
house depends upon condensate for all electrolyte make-up,
the water and heat balances at the Utah refinery are inti-
mately linked.
The water flows for the Utah refinery are shown in Figure
12, which illustrates the tie-in between boiler plant and tank
house. The overall tank house water balance is summarized
by the equation
F, + F 7 + F8 = F'3 + F 14 + F'5 + F'6 (1)
where F is the average flowrate in fta/hr. Note that flow
F7, condensate flow between boiler plant storage tanks and
tank house, can go in either direction: a positive flow is
towards the tank house while a negative flow is towards the
JOURNAL OF METALS • February, 1981
boiler plant. A simplified representation of the tank house
heat balance is shown in Figure 13. The overall tank house
heat balance is summarized by the equation
(2)
The first term is the steam energy input (Btu/hr), with
ßH, representing the he at of steam condensation in Btu/
ft 3 • The second term represents the electrical energy input
to the electrolyte, simplified as the constant C, (Btu/hr).
The third term represents the heat loss through evaporation
of water from electrolyte. The final term in Equation 2
combines all convective and radiative heat losses into a
single complex function simplistically expressed as the
product of a positive constant C2 (Btu/ hr / °F) and the dif-
ference between electrolyte and ambient temperatures.
Rearranging Equation 2,
(3)
63
 Steam Electrical
ßH 4 ) c,
Tank House
H'4) C2 (T c- Ta)
Ir
Evaporation T Related
Losses Heat Losses
Flgure 13. Heat flow diagram for Kennecott Utah reflnery tank
house.
Combining the water balance (Equation 1) and the heat
balance (Equation 3) results in
F,s + F'6 +[(1 -~~~~ F14]+
F 7 + F8 = F'3 +
[~~4] - [C2(~ILTa)]
(4)
with all terms having the units ft3/ hr. The water and heat-
ing systems will be in balance when Equation 4 is satisfied.
The left side of the equation has two terms: F 7, the flow of
condensate from the boiler condensers; and F 8 , domestic
water make-up. The boiler condensers are used to preheat
boiler feedwater and give flow F 7 a maximum value of 150
ft 3/ hr, corresponding to sending all condensate from this
source to the tank house. Domestic water make-up, F 8 , if
not zero, is maintained as low as possible because of its
high impurity levels.
The right side of Equation 4 contains six terms which are
discussed below:
• Outside Cathode Wash, F'3 Formerly, condensate was
utilized at the Utah refinery to give cathodes a final out-
doors wash with an averaged estimated consumption of
15 ft 3/hr. Recent installation of new three-stage cathode
wash machines in the tank house has eliminated this
practice.
• Electrolyte Bleed, FIs The completely de-copperized
electrolyte bleed amounts to 75 ft 3/hr and is used mainly
to controlarsenic, antimony, and bismuth levels in tank
house electrolyte. In contrast to practice at many other
refineries, it is not economically advantageous to recover
nickel at the Utah refinery because of its very low con-
tent in the anodes.
• Miscellaneous Losses, F I6 This term includes any water
64
not recovered for return to the electrolyte systems arising
from leaking condensate hoses, pump gland water, spills
on the basement floor, etc. Assuming a modest loss rate of
5 gaU min for the entire tank house would amount to 40
ft 3/ hr for this term.
• Evaporation Term This includes all evaporation losses
from electrolyte at 140°F, ~H14' is estimated at 63,300
Btu/ft3, while the heat gain from condensing steam at 30
psig down to electrolyte temperature, ~H4' is estimated
at 66,400 Btu/ft 3. Without cell covers, flowrate F14 has
a value of approximately 750 ft 3/hr, giving the entire
term a value of 35 ft3/hr.
• Electrical Term This ac counts for the electrical energy
that goes towards heating the electrolyte, and will be
relatively constant. For the Utah refinery, it has a magni-
tude of about 400 ft 3/ hr.
• ~T, Related Heat Loss Term This includes all heat losses
from the sides of cells and tanks as weH as convective and
radiative heat losses off the tops of cells. Due to the nega-
tive sign in front of the term, as its value increases more
condensate will be available. Without cell covers, the size
of the term is estimated to average 400 ft3/hr. Since the
he at losses are related to the difference between electrolyte
temperature and ambient temperature, the term will be
larger in winter than in summer. The range is estimated
at 300-500 ft 3/hr corresponding to summer and winter
respectively.
Combining the estimates for a11 terms indicates that during
the winter condensate is available for recycling to the boilers
while during the summer domestic water must be added to
the electrolyte. This corresponds to historical experience at
the Utah refinery.
Implementation of ce11 covers at the Utah refinery was
initiated during August, 1978, with coverage of one of the
36 commercial refining sections. Coverage was steadily in-
creased up to 28 sections in April, 1979. It was determined
that the covers reduced heat losses off the tops of ce11s by
roughly 75%; thus, with 16 hr/ day coverage, energy savings
approximated 3 MM Btu/ft2/year or roughly 1 MM Btu/
ton cathode. With the 28 sections covered, a condensate
shortage developed and so coverage was reduced. In addition
to reducing evaporation from the cells, the covers also re-
duceconvective and radiative losses off the cell tops. Re-
ferring to Equation 4, it is estimated that each section
covered reduced condensate availability by 4 ft3/ hr. Due
to the water availability problem, only partial implementa-
tion of ce11 covers, restricted to the cold weather months,
has been possible at the Utah refinery. To attain full imple-
mentation, installation of a 200 ft3/ hr reverse osmosis water
treatment system to supplement tank house water require-
ments is planned in 1981.
ELECTROWINNING
Heat balance for a copper electrowinning tank house dif-
fers significantly from that for an electrorefining tank house.
First, electrolyte temperatures for electrowinning are gen-
era11y lower by 20-50Fo, resulting in lower heat loss rates
per unit area. Second, the use of acid mist suppressants such
as Dowfax or polypropylene spheres essentially eliminates
evaporation off the top of cells. Finally, the electrical
energy requirements manifested as heating of the electrolyte
is roughly five times greater for electrowinning, amounting
to approximately 1600 Btu/lb of copper deposited.
Consider the case of an electrowinning tank house of con-
ventional design producing copper at the rate of 30 tons/
day. Using the heat balance approach discussed for electro-
refining, the rate of heat loss from the cells can be estimated
at approximately 0.6 MM Btu/hr. Since the rate of electri-
cal heat input to the electrolyte is 4.0 MM Btu/ hr, the net
rate of heat gain in the tank house will be 3.4 MM Btu/hr.
This results in an increase in electrolyte temperature as it
JOURNAL OF METALS • February, 1981
 ~- - TLO Loaded Organic
I f
.. .1 R_ _ _TsQ...
~ Stripped- ~ ---
El E2 S2
Organic
Tu:>
TPlS TR TLE
~
Pregnant Raffinate Lean
Leach Solution Electrolyte
passes through the tank house. The magnitude of the in-
crease is inversely proportional to the total· electrolyte
flow into or out of the tank house. For the 30 ton/ day exam-
pIe case, the temperature rise aeross the tank house would
be 6.7°F for a flowrate of 1000 gal! min, and 13.4 °F for 500
gal! min. The influence of flowrate on the drop in copper
concentration across the tank house similarly corresponds
to 5 g/l and 10 g/l at flowrates of 1000 gal! min and 500
gal! min respectively.
Despite the signifieant heat generation in an eleetrowin-
ning tank house, it is often necessary to supplement the
he at input to maintain desired electrolyte temperature.
The tank house represents the final unit operation for copper
recovery, usually from either direct eleetrowinning of eopper
from leaeh solutions or leach solutions purified and upgraded
by solvent extraetion. In general, the heat generated in the
tank house is lost upstream in the leaching process.
Consider now the heat balance on the solvent extraction
circuit shown in Figure 14, consisting of two extraction and
two stripping stages. For the example case, assurne that the
pregnant leach solution flows at 5000 gal! min at 75°F, and
that the organic/ aqueous flow ratio is 1:1 in the extraetion
stage and 5: 1 in the stripping stages. In other words, the
organic flowrate is also 5000 gal! min while the electrolyte
flowrate is 1000 gal! min. If the design and sizing of the
mixer / settlers is standard, the rate of heat loss from each
stage can be estimated at approximately 0.3 MM Btu/ hr,
using standard heat transfer equations. The heat capacities
of the various solutions are estimated as follows: leach solu-
tion, 7.4 Btu/ gal! °F; organic, 3.4 Btu/ gal! °F; elec-
trolyte, 8.4 Btu/ gal! °F. The overall solvent extraetion
circuit heat balance can be described by four equations,
each representing the heat balance for a particular stage.
Since the temperature of the aqueous and organic phases
leaving each stage will be identical, there are only five
temperatures to be determined (Figure 14). Combining the
four heat balance equations into a single equation yields
the relationship between the rich eleetrolyte temperature
Table VI: Solvent Extractlon-Electrowlnnlng Example ease Summary
Cu Cone. leaving Riehl Lean Heat
Case EW (g/l) Exehanger Area (ft')·
1 40 7,000
2 35 5,300
3 30 870
4 25 470
'Assurne U = 300 Btu/hr·ft2.oF
.. Assume $31 M Ib steam.
JOURNAL OF METALS • February, 1981
S1 Flgure 14. Solvent extractlon clrcult
flow diagram.
Tso
TRE
Rlch
Electrolyte
(T re ) and the lean electrolyte temperature (T le ). For the
example case this relationship is:
T re = 0.19 T le + 60.3 (5)
where temperatures are expressed in °F.
To finally determine the unknowns in Equation 5, the
solvent extraction circuit must be tied together with the
eleetrowinning tank house. In the design of a solvent extrac-
tion-electrowinning facility, this subjeet requires careful
consideration. Figures 15-18 show eight different methods
of tying together the example 30 ton/ day solvent extraetion-
electrowinning processes. In all eases it is assumed that the
rich electrolyte leaving solvent extraetion contains 45 g/l
Cu and that the lean electrolyte leaving electrowinning is
maintained at nO°F. Four of the methods rely on heat
transfer between rich and lean electrolyte by a he at ex-
changer (or exchangers), while the other four rely on steam
heat exchangers to supplement the rich electrolyte temper-
ature. In Figure 15 there is no intermingling of rich and
lean electrolyte; in Figure 16 rieh and lean electrolyte is
mixed on the eleetrowinning side of the heat exchanger; in
Figure 17 they are mixed on the solvent extraction side; and
in Figure 18, mixing occurs on both sides of the heat ex-
changer.
Heat exchanger requirements are given by the equation
q = VA .:lTlm (6)
where q is the overall rate of heat transfer (Btu/hr),
.:lTlm the log-mean temperature difference in °F, and the
heat exchanger size is represented by the product of the
overall heat transfer eoefficient V (Btu/ hr-ft2- °F) and
the heat exchange area A(ft2).
The solvent extraction-electrowinning example cases are
summarized in Table VI. The following are some of the
trade-offs which must be considered:
1. As a general rule, overall eathode quality will improve as
copper concentration inereases to a maximum of about
50 g/l. If one of the objectives of a solvent extraction-
electrowinning facility is to produce high-quality eathodes
suitable for direct sale, a design consideration should be
Steam Heating
Exehanger Area (ft')· Steam Cost (ei Ib Cu)"
250 1.3
190 1.0
50 0.3
20 0.1
65
 Rieh/Lean Exchange

76°F 103.3 °F
1000 GPM 45g11 ~
1000 GPM 45 gll ~

Rieh Eleel. HX
Figure 15. Solvent extractlon-electro- q" 13.8 MM BlU/hr EW
wlnnlng heat exchange requlrements, Lean EIeeI. UA=2.1 MM BTU/hr-oF
ease 1. 1000 GPM 40 gll 1000 GPM 40 gll
82.7 ° F 110°F

Sleam Heating
Condensale Sleam (10 Psig)

HX
81.2°F 103.3°F
q= 11.2 MM BlU/hr
1000 GPM 45 gll 1000 GPM 45911
UA 76 M BlU/hr ° F
Rieh Eleel.
EW
Lean Elecl.
1000 GPM 40 gll
11 0°

Ricbll Mn Exchange

76°F 9&.6 °F 103.3°F

-( )
1000 GPM 45 oll 1000 GPM 45 oll 1000 GPM 45 oll
Rich Eleel. HX
q = 10.4 MI! 8TU/hr EW
Figure 16. Solvent extractlon-electro- UA=1.6MM BTU/hr. °F
Le.. EIeeI.
wlnnlng heat exchange requlrements, 1000 GPM ° 40 gll 1000 GPM ° 40 g/ e 1000 GPM 35 gll
ease 2. 82.7 F 103.3 F 110 ° F

Steam Heating
Condensate SI.am (10psig)

HX
79.9°F q = 8.5 MM BlU-hr 96.6 °F 103.3°F
~1,-;0:::;00;,.G::;P:.,::M::....;:45"-,,,g1,-,I.....o-i UA = 56 MBlU/hr. °F 1000 GPM 45 oll 1000 GPM 40 gll
Rieh EIeeI.
EW
Lean Eleet.
1000GPM 40/ gl loo0GPM 35 oll
103.3°F 110°F

Rieh/tean Exchange

76°F 82.7 °F 96.6°F

-( )
1000 GPM 45.11 500 GPM 40 oll 500 GPM 40 oll
Rieh EIeeI. HK
q = 3.5 . . BlU/hr EW

Figure 17. Solvent extractlon-electro-

Lean EIBet.
1000 GPM 40 oll
82.7 ° F
- 500 GPM 30 oll
96.1 ° F
UA = 260 MBlU/hr.oF
500 GPM 30 oll
110°F

wlnnlng heat exchange requlrements,

ease 3.
Sleam H.ating Condensate SI.am (10 psig)

HK
77 °F SB °F q = 2.2 MM BTU/hr 9&.6 °F
1000 GPM 45 oll .........;5O=0..:G::.;PM=-.:;;4O:..g!!;/..:,I_+l UA =15 MBlU/hr. oF 500 GPM 40 gll
Rich EIeeI.

Lean Elecl. EW
1000 GPM 40 oll 500 GPM 30 gll
88°F 110°F

66 JOURNAL OF METALS • February, 1981

 RichlLe. EHtI!!!9!

':0 gli
103.3°F
.;;;1CIJI)~~~6~':S~~_"<-'=':=--=Jf.!...~r------, 500:;: 1000 GPM 30 gli
Ri eh EI ect. Hl
q = 1.8 II1II BlU/hr ew
UA =140 M BTU/hr.oF
500 GPM 30 gli
103.3 ' f
1000 am ,f' gli 1.-_ _----'

Flgure 18. Solvent extractlon-electro-

Condenute Ste..,.. (10 psig) wlnnlng heat exchange requlrements,
St..., He.ting
Ca se 4.
HK
q = 1.1" BlU/hr
76.5°F 89.9 °F
1000 GPM 45 gli UA '" 7 M BTU/hr. °F 500 GPM 40 g/I
Rich Eleet.
ew
Lein Eleet.
1000 GPM 40 gli 500 GPM 30 gli 1000 GPM 25 g/I
85.5°F 103,3°F 110' f 1.-_ _---'

maximizing the copper concentration in the cells. Table References

VI shows, however, that as the copper concentration is 1. R.B. Bird, W.E. Stewart, and E.N. Lightfoot, Transport Phenomena, John Wiley and
increased the heat exchanger area and/or steam require- Sons, Ine., New York, 1960.
2. J. Szekely and N.J. ThemeJis, Rate Phenomena in Process Metallurgy, John Wiley and
ments also increase. For example, increasing the copper Sons, Inc., New Vork, 1971.
concentration leaving the tank house from 35 to 40 g/ I 3. F. Kreith, Principles of Heat Transfer, International Text Book Co., Seranton, Penn,yl·
vania, 1967.
requires a 32% increase in heat exchanger area and, in 4. J.H. Perry, ed., Chemical Engineers' Handbook, MeGraw·HiII Book Co., Ine., New
steam heating, an additional $70,OOO/yr in energy costs. Vork, 1950.
2. Copper concentration in the cells could also be increased
by increasing cop per concentration in the rich electrolyte
leaving solvent extraction. Depending on the particular ABOUT THE AUTHOR
copper extractant used, this could require increased
reagent concentration, increased strip stage organic/ Tracy B. Braun received his BS in metallurgy
aqueous ratios, or increased staging. and materials science from the University of
3. Energy costs for steam heating can be substantial, ranging Pennsylvania (1970) and his ScD in metal-
to over let/lb copper in Case 1. This operating cost can lurgy from the Massachusetts Institute of
be eliminated by the rich/lean electrolyte heat exchange Technology (1974). He joined Kennecott in
1974 as associate scientist and later became
approach, but the required he at exchanger size is 20-30 a scientist in the Electrometallurgy Group of
times larger. the then Metal Mining Division Research Cen-
The most effective design must be determined case-by-case, ter in Salt Lake City, Utah. For the past three
depending on facility objectives, metallurgical considera- years he has been technical superintendent at the Kennecott
tions, and local factors such as fuel availability and cost. Minerals Company Utah Refinery in Salt Lake City.

Appendix
List ot Symbols tor Table I

A Area (fF) q Energy flow (Btu / hr)

b Width of gap between cells (ft) T Temperature (OF)
C" C2 Constants (Btul hr) To Ambient temperature (OF)
Cp Specific heat (Btu/lb-OF) Tc Inside cell temperature (OF)
e Emissivity (0.95 for H 2 0 at 145°F) Tw Outside cell wall temperature (OF)
F Flowrate (fpl hr) T' a Ambient temperature (OR)
~H Heat of evaporation I condensation (Btu l ft3) T' c Inside cell temperature (OR)
I Electric current (Amp) T,. Lean electrolyte temperature (OF)
Ka Thermal conductivity of air (0.016 Btu-ftl fF-hr- °F) T" Rich electrolyte temperature (OF)
Kw Thermal conductivity of concrete cell wall Tim Log-Mean temperature difference (OF)
(004 Btu-ftl ft 2-hr- °F) U Overall heat transfer coefficient (Btul hr-ft 2-°F)
L (Cell width X cell length) 'I' (ft) Vc Cell vol tage (V)
M Weight Ob) W Cell wall thickness (ft)
N Number of cells (I Stefan- Boltzmann constant
Pa Partial pressure of H 2 0 at ambient conditions (1.712 X 10- 9 Btu/hr-ft2-0R4)
(mm Hg) p Density Obi ft3)
Pu' Partial press ure of H2 0 at evaporating surface
(mm Hg)

JOURNAL OF METALS • February, 1981 67