EPA-670/2-75-016
April 1975
Environmental Protection Technology Series
THE RECLAMATION OF
SULFURIC ACID FROM
WASTE STREAMS
National Environmental Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
EPA-670/2-75-016
April 1975
THE RECLAMATION OF SULFURIC ACID FROM WASTE STREAMS
By
Howard C. Peterson
Peter L. Kern
Research Department
The New Jersey Zinc Company
Palmerton, Pennsylvania 18071
Project No. S-801349
Program Element No. 1BB036
Project Officer
Herbert S. Skovronek
Industrial Waste Treatment Research Laboratory
Edison, New Jersey 08817
NATIONAL ENVIRONMENTAL RESEARCH CENTER
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
REVIEW NOTICE
The National Environmental Research Center — Cincinnati has re-
viewed this report and approved its publication. Approval does
not signify that the contents necessarily reflect the views and
policies of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute endorse-
ment or recommendation for use.
ii
-------
FOREWORD
Man and his environment must be protected from the adverse effects
of pesticides, radiation, noise and other forms of pollution, and
the unwise management of solid waste. Efforts to protect the en-
vironment require a focus that recognizes the interplay between the
components of our physical environment--air, water, and land. The
National Environmental Research Centers provide this multidisci-
plinary focus through programs engaged in
o studies on the effects of environmental
contaminants on man and the biosphere, and
o a search for ways to prevent contamination
and to recycle valuable resources.
Sulfuric acid, the largest volume chemical manufactured in the Na-
tion, is used in a broad range of industrial applications. Econom-
ical and environmentally sound disposal methods are still needed
for the residual, contaminated acids which must be replaced by
virgin acid. This project describes one approach developed within
a major acid-using industry, the titanium dioxide pigment industry,
to solve the disposal problem and simultaneously recover the acid
for reuse.
A. W. Breidenbach, Ph.D.
Director
National Environmental
Research Center, Cincinnati
iii
-------
ABSTRACT
This report presents the results of a pilot program to demonstrate
the technical feasibility of a novel approach to the recovery of
sulfuric acid from waste streams. The process involves the total
evaporation of the sulfuric acid and water in the waste stream to
effect the removal of dry, free-flowing solids. The resultant
acid-laden vapors are then partially condensed in multiple stages
to yield a highly concentrated acid product which is suitable for
reuse in the parent process.
The pilot plant was designed to process two tons per day of sul-
furic acid (100% basis) from the waste stream of a titanium dioxide
pigment plant. Spray evaporation effected the removal of 90% of
the dissolved solids and subsequent condensation of the acidic
vapors yielded a product acid having a concentration in excess of
85% H2S04. This acid was upgraded to 954-% 1^804 level through the
addition of oleum and was satisfactorily used to digest titanium
slag. Sulfuric acid mist and sulfur dioxide emissions were found
to be of reasonable magnitude and were responsive to efforts to
bring them into compliance with environmental standards.
Dry solids were generated at the rate of 0.66 kg/kg 100% I^SO^ pro-
cessed. The recommendations point up the need to develop a satis-
factory means of disposal of these highly soluble salts.
A commercial plant has been designed to process 345,000 metric tons
annually of 19.5% E^SO^ waste end liquor from a 38,100 metric ton
per year titanium dioxide plant. The estimated investment (as of
January 1, 1975) for such a plant is $7,800,000. The operating
costs would be approximately $77/metric ton (100% 112804 recovered)
of which fuel and electrical costs would account for 48%.
This report was submitted in fulfillment of Project Number S-801349
of the Environmental Protection Agency.
iv
-------
CONTENTS
Abstract
List of Figures
List of Tables
Sections
I. Conclusions and Recommendations ..................... ...... 1
II. Introduction and Process Background ........ . ........ ..... 2
III . Summary .................................................. 7
IV. Pilot Plant Design ....................................... 11
1. Spray Evaporator-Burner ............................. 11
2. Cyclone ......................................... :... 11
3. Partial Condensation Train .......................... 11
4. Mist Eliminators .................................... 21
5. S02 Scrubber ........................................ 21
V. Procedures .................................... ........... 23
1. Feed Preparation .................................... 23
2. Operating Procedures ................................ 23
3. Sampling of Stack Emissions ......................... 23
4. Special Analytical Techniques ....................... 24
VI. Discussion of Results .................... . ............... 25
1. Acid Feed to the Spray Evaporator ................... 25
2 . Nature of the Dried Solids .......................... 26
3. Quality of the Product Acid ......................... 30
4. Acid Mist Elimination ............................... 33
5, Scrubbing for Removal of Sulfur Dioxide ............. 34
6. Material Balance - Recovery and Loss of
Sulfuric Acid ..................................... 34
7 . Heat Balance ........................................ 35
8. Recovery of Sulfuric Acid from Pickle Liquor ........ 36
9. Preconcentration of Waste Acid ...................... 38
10. Materials of Construction ........................... 39
VII. Design of Commercial Plant ............................... 41
References
Acknowledgments
-------
FIGURES
No. Page
1 Schematic Diagram of the New Jersey Zinc Process 3
2 Pilot Plant Flowsheet - Evaporative Recovery of
Waste Sulfuric Acid 12
3 Spray Evaporator 13
4 Cyclone 14
5 Primary Condenser - Separator 15
6 Spray Tower 17
7 Venturi Scrubber 18
8 Two-Stage Condensation Train 19
9 Three-Stage Condensation Train 20
10 Two Plus-Stage Condensation Train 22
11 Product Acid Concentration as a Function of Conden-
sation Stages and Outlet Temperature of Last Stage 32
12 Spray Evaporator Heat Load Vs. Feed Acid Concentration... 37
vi
-------
TABLES
No._
1 Standardization of Oxalate Method for Determining
Free Sulfuric Acid 24
2 Typical Feed Acid Compositions 25
3 Bulk Density of Dried Solids 26
4 Free Acid Occluded in Solids 27
5 Oxidation of Ferrous Iron 28
<
6 Solids Collection as a Function of Particle Size 29
7 Average Size Distribution of Dried Solids 29
8 Typical Analysis of Product Acid 31
9 Acid Mist Loadings in Stack Gas 33
10 Average Sulfuric Acid Losses 35
11 Basis for Determining Operating Cost 41
12 Energy and Materials Impact of Commercial Plant 42
vii
-------
I. CONCLUSIONS AND RECOMMENDATIONS
The technical objectives of the pilot plant program were satisfied.
It was demonstrated that a solids-bearing sulfuric acid waste stream
could be totally evaporated to effect removal of 9070 of the solids
and that the sulfuric acid vapors could be condensed to yield a
highly concentrated product of strength and quality suitable for re-
cycle to the digestion stage of a titanium pigment plant. In addi-
tion, this was accomplished without an inordinate loss of acid due
to oxidation, decomposition, or simply as acid mist.
The results of the pilot program were sufficient to initiate the pre-
liminary design of a commercial plant which it is estimated would
cost $7,800,000. However, the viability of a recovery plant will be
dependent upon the ability to satisfactorily dispose of the dried
solids which are generated as a by-product. These materials are
acidic, are water soluble and are produced at the rate of 0.66 kg
for every kg of 100% t^SO^ which is fed to the evaporator. Although
the solids disposal problem does not affect the technical feasibil-
ity of the process, it must be resolved before the process can be
implemented commercially. Three possible approaches are: (1) par-
tial recovery of useful metallic values, such as vanadium; (2) use
as a flocculating agent in municipal sewerage treatment facilities;
and (3) neutralization and treatment to insolubilize the metallic
salts so that the resultant material can be used as landfill. For
purposes of estimating plant capital and operating costs, total
neutralization of the solids was assumed.
A second point which must be considered is the sensitivity of the
process to the cost of fuel. The total direct and indirect costs
associated with acid recovery amount to $77/MT of 100% l^SO^. At
$7.95/MM kcal ($2.00/MM Btu) energy accounts for 48% of the total
recovery cost. Since virtually no fuel is consumed in the produc-
tion of virgin sulfuric acid from sulfur and metal sulfides, the
evaporative recovery process is economically vulnerable to ever in-
creasing fuel costs. It is recommended that this drain on energy
resources be examined in detail as part of an overall study of the
environmental impact of the recovery process.
Although no insurmountable process difficulties were encountered in
the pilot program, scale-up to a commercial operation will require
careful attention. This is particularly true in the areas of solids
handling, overall acid-water balance, and emissions abatement. In
addition, improved cyclone design is desirable to reduce the carry-
over of solids to the product acid and thereby minimize the poten-
tially adverse effect of these solids on pigment quality.
-------
II. INTRODUCTION AND PROCESS BACKGROUND
The economical recovery of useful sulfuric acid from waste streams
has been of continuing concern to industry for many years, but it
has received greater emphasis recently due to the imposition of
Federal and State regulations on waste discharges. Domestic pro-
ducers of titanium dioxide by the sulfate process are particularly
aware of this problem since their process generates over 3-1/2 mil-
lion metric tons annually of wastes containing 18-20% free sulfuric
acid and 10-15% soluble sulfate salts. These salts, primarily of
iron (4.1% FeSO^), aluminum (2.9% A12(SOA)3) and magnesium (2.4%
MgSO^), contain minor quantities of vanadium, chromium and manganese
which, if recycled, would be highly detrimental to pigment quality.
Established techniques(1) for the recovery of acid, such as sub-
merged combustion and multiple effect evaporation, have certain dis-
advantages when used to recover sulfuric acid at the 80-90% concen-
tration level required for reuse in pigment plants:
a. The solubility of solid contaminants decreases with
increasing acid concentration. This results in
the precipitation of sulfates which must be removed
from the acid liquor by filtration or by centrifugal
separation.
b. The concentration of acid by liquid phase evaporation
presents serious materials of construction problems.
In multiple effect evaporation, for example, very
expensive vessels are required to resist the corro-
sion encountered at or near the boiling point of sul-
furic acid.
Because of these problems, processes to recover sulfuric acid from
waste streams by means of concentration and separation have been
limited in their application. Other commonly used techniques for
the control of acid pollution from this and other industries include
neutralization, ocean disposal and deep well disposal. However,
these methods do not present attractive long-term solutions to the
problem of acid disposal.
The New Jersey Zinc Company process(2) for evaporative recovery of
sulfuric acid from solids-bearing sulfuric acid wastes (cf. Figure 1)
is an attempt to circumvent the main problem areas cited above. In
the first place, solids removal is accomplished by the total evapora-
tion of finely atomized, waste acid feed to yield a dry, free-flowing
product which can be readily removed from the vapor phase. This
-------
COMBUSTION CASES
DRIED SOLDS
TO DIGESTION
FIGURE 1. SCHEMATIC FLOWSHEET - WASTE ACID RECOVERY PROCESS
-------
avoids filtration of precipitated solids and its attendant acid
losses. Secondly, highly concentrated sulfuric acid is condensed
from the vapor phase by direct cooling in refractory-lined vessels
thereby obviating the need for expensive, corrosion-resistant heat
transfer equipment.
The acid recovery process is best described by referring to the data
established for a commercial-scale operation designed to handle waste
acid from titanium slag digestion. For maximum fuel economy the
waste sulfuric acid stream is preconcentrated by means of submerged
combustion or venturi contactor(3) to achieve a target composition
of 40% l^SO^ and 23% solids. At this concentration and the operat-
ing temperature of the preconcentrator, the waste stream is still
single phase and no solids separation is encountered. This stream
is then evaporated in a spray evaporator to yield a vapor stream
theoretically containing 0.136 kg H2SO^/kg dry gas and 0.171 kg
H20/kg dry gas.
After removal of solids, this vapor stream, which has a sulfuric
acid dew point of approximately 215°C, enters a refractory-lined
condensation chamber at a temperature of about 315°C where it is
cooled by the direct introduction of a finely atomized dilute acid
stream recirculated from a secondary condenser. The rate at which
this recycled coolant is introduced depends on the approach to
thermal equilibrium in the condensers and the operating temperature
of the secondary condenser or spray tower. Under conditions of
ideal vapor-liquid equilibrium and an outlet temperature of 80°C,
the recycled coolant is returned at a theoretical rate of 0.024 kg
H2S04/kg dry gas and 0.103 kg E^O/kg dry gas. Thus, the overall
acid-water composition in the primary condenser is 0.160 kg I^SO^/kg
dry gas and 0.274 kg H^O/kg dry gas. At vapor-liquid equilibrium,
85% of the sulfuric acid will condense at an overall concentration
of 87% I^SO^. The exit gas, which is at its acid dew point, con-
tains 0.024 kg H2S04/kg dry gas and 0.254 kg H20/kg dry gas. This
exit gas then passes to the secondary condenser (and mist eliminator)
which serves to recover virtually all of the remaining acid by cool-
ing the gases to a stack temperature of 80°C.
In this approach to acid concentration, the partial condenser and
its vapor-liquid separator are the only vessels where hot, concen-
trated sulfuric acid is present. These vessels can be readily con-
structed of acid-resistant brick overlaying a lead membrane, thereby
substantially reducing the problems associated with acid corrosion.
The partial condensation technique employing two stages theoret-
ically permits recovery of acid having a concentration of 90% 112804.
-------
This can be done while maintaining the heat removal and water bal-
ance necessary to assure that there is essentially no acidic efflu-
ent other than the concentrated product.
Bench-scale work using solids-free sulfuric acid confirmed the fun-
damentals of this approach. However, several aspects remained to
be resolved by a pilot plant demonstration. These included the fol-
lowing:
a. The oxidation of ferrous iron to the ferric state.
This readily occurs in the presence of sulfuric
acid and oxygen as indicated by the following
chemical reactions:
(1) 2FeS04 + 2H2S04 •* Fe2(S04)3 + S02 + 2H20
(2) 2FeS04 + H2S04 + &>2 •* Fe2(S04)3 + H20
Since sulfur dioxide is a product of reaction (1),
iron oxidation is an important environmental con-
sideration.
b. The characteristics of the dried solids. The bulk
density, flowability, and the quantity of occluded
free acid had to be established.
c. The differences between waste acid from ilmenite
digestion and waste acid from titanium slag di-
gestion. The former has a higher ferrous iron
content (8.17. FeS04 as opposed to 4.1%) and this
might be significant in terms of sulfur dioxide
emissions resulting from oxidation.
d. The quality of the product acid. Both the solids
content and the H2S04 concentration of the product
acid are of major importance in determining the
suitability of the product for reuse in the pig-
ment process.
e. The acid mist emissions. It had been speculated
that large quantities of sulfuric acid would be
lost as acid mist.
f. The vapor-liquid equilibrium and water balance.
The condensation of high concentration sulfuric
acid must involve an efficient transfer of heat
-------
from the liquid to the vapor phase. If this is
not satisfactorily attained, water balance prob-
lems will prevent the desired acid concentration
from being reached.
These were the primary areas of interest in the pilot plant program,
and it was for the purpose of resolving these points that the demon-
stration unit was constructed and operated.
-------
III. SUMMARY
The piloting of The New Jersey Zinc Company process was done at
Palmerton, Pennsylvania, at a rate of two tons of 100% H^SOA per
day. The operations were carried out as part of a joint effort of
The New Jersey Zinc Company and Montedison S.p.A. of Italy, and
with the support of the United States Environmental Protection
Agency. The plant operated on waste sulfuric acid feed for a cumu-
lative total of 900 hours. The pertinent results from this period
of operation are summarized below.
1. ACID FEED AND PRECONCENTRATION
A typical sample of waste acid containing 18% H^SO^ and 11% dis-
solved solids was obtained from The New Jersey Zinc Company titanium
pigment plant at Gloucester City, New Jersey. This acid, which re-
sulted from titanium slag digestion, was processed in the as-received
condition and was also synthetically preconcentrated to 35% I^SO/ for
evaluation at this higher acid level. In addition, a waste acid
typical of that from ilmenite digestion was synthesized by adding
copperas (FeSO^'7H20), epsom salt (MgSO^-7^0), and virgin sulfuric
acid to Gloucester City end liquor. This was evaluated at a nominal
concentration of 30% ^SO^ and 17% dissolved solids.
Although preconcentration of the feed acid is necessary and desirable
from an economic standpoint, this aspect of the process was not inte-
grated into the Palmerton pilot plant since technology exists to ac-
complish this objective. However, tests were carried out at the
Selas Corporation pilot facilities in Dresher, Pennsylvania, to de-
termine stack emissions and iron oxidation, as well as to observe
the physical characteristics of the acid at various levels of pre-
concentration. No appreciable iron oxidation occurred and the acid
mist was amenable to control. However, some difficulty was en-
countered when solids periodically blocked part of the concentrator
throat. This resulted in a marked increase in pressure drop and an
attendant decrease in throughput. This problem will have to be re-
solved before this approach can be successfully incorporated into
the recovery process.
2. DRIED SOLIDS
Dry, free-flowing solids were produced in the spray evaporation arid
90% of these solids were removed from the sulfuric acid vapor stream.
The characteristics of the solids are summarized below:
-------
From Titanium Slag From Ilmenite
Apparent Bulk Density (g/cc) 0.41 0.45
Occluded H2S04 (% by Weight) 11.3 5.6
Iron Oxidation:
% of Ferrous to Ferric 66 52
7. Oxidized by Air 57 38
% Oxidized by H2S04 43 62
The presence of larger amounts of occluded sulfuric acid in the solids
from slag digestion is due to the presence of aluminum sulfate which
readily forms an acid salt. However, this higher free acid content
did not appreciably affect the flow of the solids under normal oper-
ating conditions.
3. PRODUCT ACID QUALITY
The presence of vanadium, chromium and manganese in the product acid
can be detrimental to pigment quality if the recovered acid is re-
cycled for digestion. The product acid contained only small quan-
tities of these contaminants as shown below:
Product Acid from Slag End Liquor
H2S04 80.8%
H20 13.2
FeS04 0.8
Fe2(S04)3 1.8
A12(S04)3 1.5
TiOS04 0.6
MgS04 1.2
VOS04 0.083
Cr2(S04)3 0.045
MnS04. 0.027
Laboratory digestions were made with product acid which was strength-
ened by the addition of 207. oleum. No adverse effects were observed
in digestion, filtration or in the recovery of titanium dioxide when
compared with the use of virgin acid. However, no pigment was ever
produced from the trial digestions and hence no conclusions can be
drawn as to pigment quality or the effect of continued recycle on
acid quality.
4. STACK EMISSIONS
The presence of sulfuric acid mist in the stack gas (after the mist
eliminator) varied inversely with stack temperature over the range
-------
73°C to 92°C. At the lower temperature, the mist averaged 645 mg/m3
(510 -*• 810 mg/m3), while at the higher temperature it was 250 mg/m3
(170 -*• 400 mg/m3). This difference can be attributed to the presence
of "water mist" at lower temperatures which effectively competes with
acid mist for condensation sites. The acid mist elimination effi-
ciency ranged from 80% at 73°C to 98% at 92°C. The mist loadings
were in excess of the established target of 80 mg/m3. A higher effi-
ciency mist eliminator would be required to reduce the loadings to
the target level.
Due to differences in iron content, sulfur dioxide emissions averaged
555 ppm for slag acid (149-932 ppm) and 865 ppm for ilmenite acid
(363-1674 ppm). The stack gases were scrubbed with a 5 wt. % slurry
of lime across a pressure drop of 125 mm 1^0 and the emissions were
reduced to 285 ppm and 290 ppm, respectively. The L/G ratio averaged
3.2 (m3 slurry/1000 actual m3 of gas).
5. ACID RECOVERY AND LOSSES
The overall recovery of sulfuric acid as product differed only
slightly for slag and ilmenite feeds. The losses are summarized
below:
Average Acid Losses and Recovery
7, HoSO/, As Using Slag Acid Using Ilmenite Acid
Free Acid in Solids 8.2 3.8
Iron Oxidation 4.6 5.8
Sulfur Dioxide 2.5 4.2
Acid Mist 0.5 0.6
Recovered as Product 84.2 85.6
6. HEAT BALANCE
The spray evaporation of waste acid which had been previously precon-
centrated to 35% I^SO^ required 2200 kcal/kg H2S04 in the feed. Pre-
concentration of the as-received 20% acid to this 35% level by sub-
merged combustion was calculated- to require an additional 2000 kcal/
kg H~SO,. On the basis of an average 85% recovery of I^SO^ as useful
product, the overall heat requirement is 5060 kcal/kg 100% I^SO^ re-
covered from the original waste stream. Since submerged combustion
is thermally more efficient than spray evaporation, the heat duty can
be reduced to 4560 kcal/kg if the feed acid can be preconcentrated to
40% H2S04 prior to spray evaporation.
-------
7. SPRAY EVAPORATION OF STEEL PICKLING LIQUOR
The spray evaporation of waste pickle liquor (2.1% H2SO^, 18.5%
FeSO^) resulted in poor acid recovery. The high iron content of the
feed resulted in a 46% loss of the sulfuric acid to ferric sulfate
and sulfur dioxide. Crystallization and separation of copperas
(FeS04«7H20) would be necessary prior to spray evaporation to mini-
mize oxidation to ferric iron and to ensure better acid recovery.
8. MATERIALS OF CONSTRUCTION
Evaluations were made of the numerous materials used in the construc-
tion and operation of the pilot plant. Plastics (including fiber re-
inforced materials), lead and acid-resistant alloys were suitable in
applications where acid concentration was less than 50% 1^804 and
temperatures were below 120°C. At higher acid concentrations and
temperatures, acid-resistant refractory, high silicon iron and Teflon
linings were satisfactorily employed.
9. COMMERCIAL PLANT DESIGN
A design was prepared for a commercially sized recovery plant to pro-
cess 345,000 metric tons annually of 19.5% H2SO^ waste liquor from a
38,100 metric ton per year titanium dioxide plant. The capital as of
January 1, 1975, was estimated at $7,800,000. The operating costs
were estimated to be $77/metric ton of recovered 100% sulfuric acid.
10
-------
IV. PILOT PLANT DESIGN
The pilot plant was designed, constructed and made ready for opera-
tion in the period February to June 1972. Several minor modifica-
tions were subsequently made to improve performance. Only the basic
components of the final design configuration are reported here (see
Figure 2).
1. SPRAY EVAPORATOR-BURNER
The refractory-lined spray evaporator (Figure 3) was 1.8 meters in
diameter and had a 6.1 meter high straight wall. The shell was con-
structed of Cor-Ten steel over which insulating refractory and a
monolithic, acid-resistant refractory lining (Sauereisen Cements
Co. #54) was gunned. The shell exterior was insulated to minimize
heat loss. Combustion gases were supplied to the evaporator from a
189,000 kcal/hour natural gas burner operating at a nominal tempera-
ture of 1000°C and at an inlet pressure of 760 mm t^O. The acid feed
to the evaporator was introduced through a Delavan two-fluid atomiz-
ing nozzle which was operated at a nominal air pressure of 3.5 kg/
cm2. Dried solids were collected batchwise at the bottom of the
evaporator in a heated steel drum. The exhaust gases passed through
an unlined Cor-Ten steel duct.
2. CYCLONE
The exhaust gases from the spray evaporator were sent to a cyclone
(Figure 4) for removal of entrained solids. The cyclone was con-
structed of Cor-Ten steel and was unlined. The exterior was insu-
lated to minimize heat loss and to maintain the gas temperature above
the acid dew point. In a manner similar to the evaporator, the dried
solids were collected at the bottom of the cyclone in a heated steel
drum.
3. PARTIAL CONDENSATION TRAIN
The condensation train consisted of three vessels - the primary con-
denser, the separator and the spray tower. These vessels are de-
scribed below.
A. Primary Condenser
The primary condenser (Figure 5), whose function it was to provide
residence for cooling and condensation of the acid-laden gases from
the evaporator, was fabricated of Cor-Ten steel. It was lined with
11
-------
BURNER
,SPRAY EVAPORATOR
CCLONE
\W\STE ACID
PRIMARY CONDENSER
OUTLET
FAN
MIST
ELIMINATOR
VENTURI
SCRUBBER
PREPARATION
TANK
PRODUCT
ACID
LIME SLURRY
TANK
FIGURE 2. PILOT PLANT FLOWSHEET - EVAPORATIVE RECOVERY OF WASTE SULFURIC ACID
-------
JL
o
1
( l j
•— - r
-ACC
DO<
14.3
200C
/
r
; l
:ESS
DR
).25 1.0. r
1
/
h .»,
\\
\\
\\
\\
V
jfek=
1
f
\
^
CO
•I ' k-152.4 !.D.
ALL DIMENSIONS IN MM.
FIGURE 3. SPRAY EVAPORATOR
13
-------
in
\f\
\ i—i
FIGURE 4. CYCLONE COLLECTOR
T
in
I.P.
klM,
14
-------
tO
3
v9
*
\
in
I. p.
cO
3
06
vO
•0
OJ
ALL
Ili MM.
FIGURE 5. PRIMARY CONDENSER - SEPARATOR
15
-------
acid-resistant brick (using a silicate mortar) over a lead membrane
4 mm thick. The dilute acid which was recycled from the spray tower
was introduced into the condenser by means of a two-fluid Delavan
atomizing nozzle. A venturi throat, which for pilot plant purposes
was constructed of lead-lined steel, was installed in the condenser
as a contactor to improve the approach to vapor-liquid equilibrium.
The pressure drop across the venturi was nominally 1QO mm of water.
The exterior of the condenser was insulated to minimize heat losses.
B. Separator
The separator (Figure 5) served to disentrain acid droplets from the
vapor phase and to improve the approach to vapor-liquid equilibrium.
The shell was of Cor-Ten steel over which was installed a lead mem-
brane and monolithically cast acid-resistant refractory. The vessel
was packed to a depth of one meter with 25 mm berl saddles. The ex-
terior of the separator was insulated to minimize heat losses.
C. Spray Tower
The spray tower (or secondary condenser) (Figure 6) was installed
after initial efforts using a venturi scrubber to cool the exhaust
gases from the separator proved unsatisfactory. The venturi scrub-
ber (Figure 7) was found to saturate the gas stream with water mist
which impaired the acid mist collection efficiency. The spray
tower, fabricated of lead-lined Cor-Ten steel and equipped with a
single Spraying Systems 1/4-JAG two-fluid nozzle, was installed to
permit operation at a temperature above the saturation point of the
exit gas. Furthermore, it provided operating latitude with regard
to the overall water balance on the process.
The vessels in the condensation train were grouped in three different
configurations to evaluate alternate condensation approaches. The
first of these is shown in Figure 8 and consisted of two stages - one
being the primary condenser and separator, and the second being the
spray tower. Dilute acid (approximately 20% 112804) was recycled from
the spray tower to the primary condenser to effect the cooling and
condensation of the acid vapors.
The second condensation configuration utilized three condensation
stages as shown in Figure 9. In this approach, the spray nozzle in
the primary condenser was replaced by a venturi throat and the con-
centrated product acid was recycled to this throat as a coolant at
the rate of 15-40 liters per minute. Although some heat was lost
from the pumps and piping, the system was basically adiabatic.
16
-------
ALL
III MM,
FIGURE 6. SPRAY TOWER
17
-------
oO
oO
oO
fO
1A
o
IM MM,
FIGURE 7. VENTURI SCRUBBER
18
-------
ACtc?
.1.
MIST
FIGURE 8. TWO-STAGE CONDENSATION TRAIN
-------
\
*
r-*>E
M
-^
s&\
+ ^
FIGURE 9. THREE-STAGE CONDENSATION TRAIN
-------
The third condensation configuration (Figure 10) was an attempt to
minimize the number of pumps used in the condensation train. This
approach, which was referred to as having two-plus condensation
stages, split the acid recycle between the venturi throat and the
tower packing. The total flow of recycled acid was approximately
40 liters per minute and was divided about equally between the two
vessels.
4. MIST ELIMINATORS
The initial design of the pilot plant relied upon Teflon pad mist
eliminators for acid mist control. These pads, which were 76 mm
thick, were evaluated at linear flow rates of 2.4 to 7.6 meters per
second and were found to be unable to reduce the acid mist loadings
to reasonable levels. A Brink Test Unit TU-102E was purchased from
the Enviro-Chem Systems division of Monsanto and was installed.
This unit was constructed of fiber reinforced polyester resin and
was designed for a mist elimination efficiency of 99%. This design
factor includes both acid mist and so-called water mist which is
present in gas streams close to the saturation temperature. Thus,
the sulfuric acid mist elimination will be less than the design
factor if the gas stream contains significant quantities of water
vapor.
5. SO2 SCRUBBER
The venturi scrubber (Figure 7) which was initially used to cool the
separator exit gases (and which was replaced for this purpose by the
spray tower) was evaluated as a scrubbing device to remove sulfur
dioxide from the exhaust gases. The scrubber body was fabricated of
fiber reinforced polyester and had an adjustable throat of Alloy 20
steel. The pressure drop across the venturi was capable of being ad-
justed between 100 and 400 mm of water, although limitations on fan
capacity restricted operation to a lower pressure drop of 100-125 mm
H2<>. The flow rate of liquid to the throat was 35-40 liters per
minute.
21
-------
to
V V
TAWl-<
MIST
ELIMINATOR
FIGURE 10. TWO-PLUS-STAGE CONDENSATION TRAIN
-------
V. PROCEDURES
1. FEED PREPARATION
Waste sulfuric acid from the Gloucester City, New Jersey plant of New
Jersey Zinc was shipped to Palmerton by tank trailer and stored on
site in a 6,000-gallon, rented stainless steel tank trailer. From
here it was transferred to a 2,000-gallon, lead-lined tank for use
in the pilot plant. It was in this tank that synthesized ilmenite
acid was also prepared. This was accomplished with the use of virgin
acid and dried metallic salts in addition to the waste liquor. The
tank was equipped with lead coils for heating or cooling the acid as
it was prepared.
2. OPERATING PROCEDURES
The pilot plant was prepared for operation by preheating the spray
evaporator to 850-900°C gas inlet temperature and 330-350°C outlet
temperature. The outlet temperature was maintained by spraying
water (instead of acid) into the evaporator. The acid surge tanks
and recycle tanks were filled with acid approximating the antici-
pated strength of the various outlet streams in order to minimize
the time necessary to come to steady state.
Upon reaching the desired operating temperatures, the water feed to
the evaporator was discontinued and acid feed was commenced. Levels
were measured in the various tanks and samples were taken for mate-
rial balance purposes. Operating data, such as pressures, tempera-
tures, tank levels and acid strengths were monitored and logged by
the operator once an hour. A continuous chart record of all operat-
ing temperatures was also maintained. Samples of product streams
and recycle streams were composited by shift. Solids flowed contin-
uously into 55-gallon drums mounted beneath the evaporator and cy-
clone. Drum warmers were used to maintain the drums at temperatures
in excess of 225°C in order to minimize acid condensation in the
drums. Upon termination of the run, samples and measurements were
taken for material balance purposes.
3. SAMPLING OF STACK EMISSIONS
Sulfur dioxide and sulfuric acid mist emissions were monitored with
stack sampling equipment. Method 8 of the EPA (Federal Register.
Vol. 36, No. 247, December 23, 1971) was used to determine sulfuric
acid mist and sulfur dioxide. However, since this technique re-
quired at least two hours to obtain a representative sample, the
more convenient Brink Model BMS-10 Sampling Kit by Monsanto was
used to obtain more frequent acid mist determinations.
23
-------
4. SPECIAL ANALYTICAL TECHNIQUES
The presence of salts in the feed and product acid (particularly iron
salts) interfered with the determination of the free sulfuric acid
level in samples. Free acidity was determined in the following man-
ner(4):
a. Two grams of sodium oxalate (Na2C204) was dissolved
in 100 ml of water and the pH adjusted to 7.0.
b. A 0.2-1.0 gram sample for analysis was dissolved in
the oxalate solution and was titrated potentiomet-
rically with 0.2 N NaOH to a pH of 7.0.
c. The end point of the free acid titration was taken
as the inflection point in the graph of pH versus
ml of NaOH solution.
The accuracy of this method for determining free acid is summarized
for various synthesized acids in Table 1.
Table 1. STANDARDIZATION OF OXALATE METHOD
FOR DETERMINING FREE SULFURIC ACID
% HgSO/,
Actual Oxalate Method
20% Acid with Solids "A" 19.98 19.54 -2.20
45% Acid with Solids "B" 44.96 44.74 -0.49
55% Acid with Solids "C" 54.94 55.44 +0.91
65% Acid with Solids "D" 64.90 65.11 +0.32
% MS04 in
Above Solutions
B
FeS04
Fe2(S04)3
A12(S04)3
MgS04
TiOS04
5.12
--
1.90
2.80
1.00
0.53
1.05
0.74
1.46
0.74
0.65
1.28
0.90
1.78
0.90
0.77
1.51
1.06
2.11
1.06
The compositions in Table 1 were chosen because they were typical of
acids which were encountered in the course of pilot plant operation.
No further significance should be attributed to them.
Samples of solids were also analyzed for free acid content using the
oxalate method. Standard wet chemical analytical methods were used
for iron, aluminum, magnesium and other components.
24
-------
VI. DISCUSSION OF RESULTS
For each run made during the pilot plant campaign, evaluations were
made of variations in acid feed composition, operating temperatures
and methods of acid condensation. The results of these evaluations
are discussed below.
1. ACID FEED TO THE SPRAY EVAPORATOR
The compositions of waste acid streams from sulfate Ti02 plants vary
because of the titanium raw materials used in the digestion. The two
primary sources of titanium, Sorel slag and ilmenite, have substan-
tial differences in the ratio of iron to titanium dioxide (approxi-
mately 0.14 and 0.5, respectively) thereby requiring individual eval-
uation of their corresponding end liquors. Slag-type end liquor was
made available from the Gloucester City operations of The New Jersey
Zinc Company and this was transported to the pilot plant site by tank
trailer. Ilmenite end liquors were synthetically prepared from the
slag end liquor, virgin acid and the appropriate metallic salts. The
compositions of typical acid feeds which were atomized in the spray
evaporator are listed in Table 2.
Table 2. TYPICAL FEED ACID COMPOSITIONS
Composition _ Slafc End Liquor _ Ilmenite End Liquor
_ Weight % _ As Received Preconcentrated Preconcentrated
Free H2S04 18.4 35.0 29.3
Total S04 26.5 50.4 40.6
Fe 1.59 3.02 3.90
Mg 0.67 1.27 0.79
Al 0.35 0.67 0.18
Ti 0.28 0.53 0.18
V 0.044 0.084 0.024
Cr 0.021 0.040 . 0.009
Mn 0.022 0.042 0.010
Water (Difference) 70.6 44.0 54.3
Preconcentrated acid feeds were prepared to nominal levels of 30%,
H2S04 for ilmenite liquor and 357., H2S04 for slag liquor. Care was
taken not to exceed these levels to minimize the precipitation of
salts, and the feed temperature was maintained at 70-80°C to assure
maximum solubility.
25
-------
2. NATURE OF THE DRIED SOLIDS
A. General
In most spray evaporation processes, the dried solids represent the
desired product. However, in the waste acid recovery process, these
solids are of secondary importance to the sulfuric acid. Nonethe-
less, since they do constitute a significant percentage of the total
weight of processed material, they must receive careful attention.
For example, for every 100 kilograms of 100% sulfuric acid present in
the feed liquor, there are almost 60 kilograms of metallic salts
present. Furthermore, after spray evaporation, some iron is oxidized
and some free acid remains occluded in the solids, thereby increasing
the total weight of solids to as much as 70 kilograms per 100 kg
H^SO/. Thus, these solids are a considerable by-product of sulfuric
acid recovery.
B. Bulk Density
The bulk density of the dried solids varied for the different acid
feeds, being somewhat higher for ilmenite solids as opposed to slag
solids. This can be attributed to the higher iron content of the
ilmenite solids. The bulk densities are summarized in Table 3.
Variations in bulk density of solids for a given feed are attrib-
utable to differences in size distribution and free acid content.
In general, solids with high free acid contents had higher bulk
densities than solids with low free acid contents.
Table 3. BULK DENSITY OF DRIED SOLIDS
(grams/100 cc)
From Slag End Liquor Range Average
Spray Evaporator Solids 28 -*• 54 43
Cyclone Solids 20 -»• 37 33
Composite Solids — 41
From Ilmenite End Liquor
Spray Evaporator 45 -»• 54 49
Cyclone 32 -»• 36 35
Composite — 45
26
-------
C. Free Acid in the Dried Salts
The extent to which free sulfuric acid is occluded in the dried
solids varied with acid feed and operating conditions, and the data
are summarized in Table 4. The higher free acid content of the slag
based solids is attributed to the greater quantity of aluminum sul-
fate present in the slag end liquor. Aluminum sulfate forms a rela-
tively stable double salt of A12(SO^)3 and H2SO^ in the presence of
sulfuric acid, while iron and magnesium do not.
Table 4. FREE ACID OCCLUDED IN SOLIDS
Weight % Free HoSO^
From Slag End Liquor Range Average
Spray Evaporator Solids 9.8 -> 16.6 14.1
Cyclone Solids 4.2+15.3 9.9
Composite Solids — 11.3
From Ilmenite End Liquor
/
Spray Evaporator Solids 1.4 -* 13.1 7.5
Cyclone Solids 1.2 ->• 8.9 4.3
Composite Solids -- 5.6
The "acid dryness" of the solids is also affected by operating con-
ditions. Higher atomizing pressures produced finer and drier par-
ticles than did low atomizing pressures. In one run, the solids
collected from the evaporator at 3.5 kg/cm2 (50 psi) atomizing pres-
sure had 6.8% free sulfuric acid while evaporator solids at 1.75 kg/
cm2 (25 psi) contained 15.7% acid. The corresponding cyclone-
collected solids contained 3.5% and 5.0% free acid, respectively.
(However, the degree of iron oxidation from ferrous to ferric in-
creased with increasing atomizing pressure - with 34% of all ferrous
iron being oxidized at 1.75 kg/cm2 and 40% being oxidized at 3.5 kg/
cm . Furthermore, high atomizing pressures generally resulted in
greater quantities of solids in the product acid.)
The outlet temperature of the spray evaporator was also a factor in
determining the amount of free acid in the solids. However, while
occluded acid losses were observed to decrease with increasing evap-
orator temperature, iron oxidation and sulfur dioxide formation were
observed to increase.
27
-------
D. Oxidation of Iron
The oxidation of ferrous iron in the spray evaporator can proceed in
two ways(4),(5):
(1) 2FeS04 + 2H2S04 + Fe2(S04>3 + S02 + 2H20
and
(2) 2FeS04 + H2S04 + %02 + Fe2(S04)3 + H20
The degree to which each of these reactions takes place can be estab-
lished by measuring the ferrous and ferric iron contents of the dried
solids and by observing the S02 in the stack gas. The oxidation of
the iron in the solids to the ferric state is summarized in Table 5.
This table shows that two-thirds of the iron present in slag based
liquor is oxidized while only slightly more than one-half of the iron
in ilmenite liquor is oxidized. However, since the iron content of
the ilmenite feed acid on the average was about one-third greater
than the iron content of the slag feed acid, the iron oxidation rates
were about equal.
Table 5. OXIDATION OF FERROUS IRON
% of Iron Oxidized
_ to Ferric State
Slag-Acid Ilmenite-Acid
Spray Evaporator Solids 67 58
Cyclone Solids 65 47
Composite Solids 66 52
Oxidized by H2S04* 43 62
Oxidized by Air 57 38
^Determined from amount of S02 present in the stack gas.
The mechanism of oxidation differed for the two acid feeds. Whereas
slag based acid appeared to favor oxidation by air (reaction 2), il-
menite based acid favored oxidation by sulfuric acid (reaction 1).
An explanation for this behavior was not apparent.
E. Particle Size Distribution
The particle size distribution of the dried solids was dependent upon
the atomizing nozzle, atomizing pressure, and feed rate (or air-to-
liquid ratio). Using the 1/4" atomizing nozzle, the ratio of cyclone
solids-to-evaporator solids averaged 1.2. With the 3/8" nozzle, the
28
-------
average ratio was 1.8. At an acid rate of 159 kg per hour, for ex-
ample, the air-to-liquid ratio was twice as great in the 3/8" nozzle
as in the 1/4" nozzle.
Finer particle size distribution at the atomizing nozzle is desirable
from the standpoint of evaporation rate. However, cyclone efficiency
and total solids removal from the acid suffers as particle distribu-
tion becomes finer. This is shown in Table 6.
Table 6. SOLIDS COLLECTION AS A FUNCTION
OF PARTICLE SIZE
Ratio of Cyclone Solids % Cyclone % of Solids Removed
to Evaporator Solids Efficiency From Acid*
1.20 89 94
1.75 81 87
*Evaporator solids and cyclone solids combined.
However, samples of solids obtained with the 1/4" nozzle were ex-
amined by optical microscopy and with an electron microscope and a
particle count was made on each sample. The average results are
summarized in Table 7.
Table 7. AVERAGE SIZE DISTRIBUTION OF DRIED SOLIDS
Diameter Particle Frequency (7») Weight (7.)
Evaporator Solids
<32 Microns 12 1
32- 96 Microns 50 13
96-160 Microns 29 44
160-320 Microns 8 30
>320 Microns 1 13
Cyclone Solids
<5 Microns 47 . 3
5-10 Microns 36 20
10-15 Microns 11 23
15-20 Microns 4 24
20-30 Microns 2 21
>30 Microns 1 9
29
-------
From these distributions, it is apparent that most particles greater
than about 35 microns were collected in the evaporator and particles
smaller than 35 microns passed to the cyclone.
3. QUALITY OF THE PRODUCT ACID
The quality of the product acid determines its suitability for re-
cycle to the digestion operation of the pigment process. The mea-
sures of product quality are (1) the cleanliness of the acid - that
is, the degree of contamination by solids which have escaped collec-
tion by the cyclone, and (2) its sulfuric acid concentration calcu-
lated on a solids-free basis. The former is important because of
the deleterious effects of small quantities of manganese, chromium
and vanadium on pigment color. The latter is important so as to
maintain a sulfuric acid balance in the pigment plant. Depending
on the source of titanium, acid for digestion is generally required
to contain 93-96% I^SO^ and is classified as Commercial Grade Acid.
In the case of acid recovery, this concentration requirement dan be
met by blending a lower grade product acid with oleum. For example,
in the case of The New Jersey Zinc Company plant at Gloucester City,
New Jersey, a sulfuric acid balance within the pigment plant can be
maintained if 86% l^SO^ acid (solids-free basis) is recovered and
blended with 20% oleum.
The pilot plant operations showed that three variables directly con-
trolled the final product acid strength: the sulfuric acid concentra-
tion of the feed to the evaporator, the number of condensation stages
and the exit gas temperature from the last condensation stage.
The acid concentration of the feed is subject to control in the pre-
concentration step of the recovery process. Maximum targets of 30%
I^SO^ and 35% I^SO^ were set for the ilmenite liquor and slag liquor,
respectively as described in Section VI-1. These values corresponded
to 35% H2SO/ and 447. H2SO^, respectively on a solids-free basis. The
number and description of the condensation stages are described in
Section IV-3.
The objective of the pilot plant operation was to achieve satisfac-
tory sulfuric acid concentration in the product while operating
within the constraint of a water balance. This was accomplished for
a given condensation configuration by controlling the outlet tempera-
ture from the spray tower. In the adiabatic cooling of the acid-
laden gases in the partial condenser, the ultimate source of coolant
was the water used to cool the vapor stream in the spray tower. The
dilute acid (10-25% t^SO^ depending on operating conditions) which
condensed in the spray tower was injected into the partial condenser
30
-------
which was the primary condensation stage. If the product acid con-
centration was too low, it indicated that too much coolant was being
used (that is, the spray tower outlet temperature was too low) and a
reduction in cooling water was in order.
The relationship between the three variables affecting product acid
concentration is shown in Figure 11. No distinction is made between
the spray evaporative cooling and the venturi contactor as a means
of condensing the acid vapors. There was no evidence to indicate
that for an equal number of stages one approach was better than the
other.
The typical solids removal which was observed in the pilot plant was
90%. The remaining 10% of the solids which were originally present
in the feed found their way to the product acid and gave it a milky-
white appearance. A typical analysis of this acid (obtained from
slag based end liquor) is given in Table 8.
Table 8. TYPICAL ANALYSIS OF PRODUCT ACID
(from slag based end liquor)
H2S04 (as is) .. 80.8%
H2S04 (solids free) 86.0
H20 (total) 13.2
FeS04 0.8
Fe2(S04)3 1.8
A12(S04)3 1.5
TiOS04 0.6
MgS04 1.2
VOS04 0.083
Cr2(S04) 3 0.045
MnS04 0.027
Experimental digestions of titanium slag were made with typical prod-
uct acid (79.31 H2S04, 7.4% solids) which was upgraded with 20% oleum
to 92.7% H2S04 (95.8% H2S04 on a solids-free basis). Attack acid of
this strength is required in the pigment plant to initiate the diges-
tion of the ore and to assure a high recovery of soluble titanium.
The upgraded acid was suitable for digesting the-titanium slag and
the resultant digestion liquor was comparable in filtration rate and
stability characteristics to liquor prepared with virgin acid. Ti-
tanium dioxide recovery levels were normal. Since it is difficult to
make meaningful pigment from a laboratory digestion, no pigment was
produced from the recycled acid.
The quantitative effect on pigment quality of recycling part of the
solids with the recovered acid cannot be predicted accurately. The
31
-------
T Tf MR *
FIGURE 11. PRODUCT ACID CONCENTRATION AS A FUNCTION OF CONDENSATION
STAGES AND OUTLET TEMPERATURE OF LAST STAGE
32
-------
recycled solids will increase the amounts of chromium, vanadium and
manganese in the hydrolysis liquor. Based on historical plant pig-
ment evaluations, the net effect of recycling 10% of the solids is
marginally tolerable. However, every effort should be made to re-
duce the solids loading in the acid vapor stream prior to the con-
densation stage.
4. ACID MIST ELIMINATION
In the initial stages of the pilot plant program, a Teflon mesh mist
pad was used to reduce acid mist emissions. This pad was situated
after the venturi scrubber and the gases which passed through the
pad were saturated at a temperature of approximately 73°C. The acid
mist loading in the effluent gases from the eliminator was 1660 mg/m3
(dry) which corresponded to an elimination efficiency of 50%. By re-
placing the Teflon pad with a Brink mist eliminator, while at the
same time maintaining the venturi scrubber, the mist loading was re-
duced to 645 mg/m3. This latter figure corresponded to an elimina-
tion efficiency of 80%, but was still below the Brink design value
of 99%. The discrepancy in the acid mist efficiency of the Brink
unit was attributed to the presence of water mist due to the satu-
rated gas stream entering the eliminator. The venturi was subse-
quently replaced by the spray tower to reduce water mist and thereby
increase the efficiency of acid mist elimination. This change per-
mitted independent control of the Brink inlet temperature and reduced
the water mist problem by keeping the gas stream above the saturation
temperature. The resultant mist loadings, which are summarized in
Table 9, confirmed the need for this additional degree of freedom.
Despite the fact that the mist loading to the eliminator was greater
at 92°C than at 75°C (15,900 mg/m3 versus 3,530 mg/m3), the elimina-
tion efficiency approached the 99% design level at the higher tem-
perature due to the virtual absence of water mist at that tempera-
ture.
Table 9. AVERAGE ACID MIST LOADINGS IN STACK GAS
Eliminator Type of Stack Emissions Elimination
Inlet Temp. °C Eliminator Acid Mist, mg/m3 EJEficiency^
73 Teflon Pad 1660 50
73 Brink 645 80
75 Brink 565 84
82 Brink 460 94
92 Brink 250 98
33
-------
5. SCRUBBING FOR REMOVAL OF SULFUR DIOXIDE
The sulfur dioxide emissions as measured by use of hydrogen peroxide
impingers averaged 555 ppm for preconcentrated slag acid and 865 ppm
for preconcentrated ilmenite acid. In an attempt to reduce these S02
levels, the stack gases (averaging 710 actual m-Vhr.) were contacted
with 2.27 m^/hr. of a 5 wt. % slurry of lime across a pressure drop
of 125 mm H20. The S02 levels were consequently reduced to 285 ppm
for slag acid and 290 ppm for ilmenite acid. Scrubbing efficiency
at higher pressure drops was not evaluated due to limitations on
blower capacity. No fouling of the venturi surfaces was observed,
but scrubbing tests were too short (generally about two hours) to
draw conclusions in this regard.
6. MATERIAL BALANCE - RECOVERY AND LOSS OF SULFURIC ACID
The losses of sulfuric acid associated with the acid recovery fall
into four categories:
Free acid occluded in the solids
Oxidation of ferrous iron
Formation of sulfur dioxide
Nonrecoverable acid mist
The magnitudes of these losses depend upon such operating variables
as temperatures, acid feed composition, feed rate and conditions of
feed atomization. In order to minimize the variations due to feed
composition, it was necessary to normalize the composition when cal-
culating losses. Slag based acid feed was normalized to a sulfuric
acid-to-iron ratio of 12 and ilmenite based acid was normaled to a
ratio of 7.5. Similarly, the ratio of total solids to iron was nor-
malized to values of 7.1 and 4.3 for slag acid and ilmenite acid,
respectively. With these adjustments, the average sulfuric acid
losses (as percentages of the acid fed to the evaporator) are re-
ported in Table 10.
34
-------
Table 10. AVERAGE SULFURIC ACID LOSSES
Slag Based
Liquor
70 of Acid Lost As
Free Acid in Solids
Ferric Sulfate
Sulfur Dioxide
Acid Mist
Total
% Recoverable As
Product Acid
7. HEAT BALANCE
Average
8.2
4.6
2.5
0.5
15.8
84.2
Range
5.5-10.2
4.0- 5.9
1.4- 4.2
0.1- 1.0
14.0-17.4
Ilmenite Based
Liquor
Average Range
14.4
85.6
0.9- 8.9
4.7- 6.9
1.2- 8.8
0.2- 0.8
9.1-24.7
A. Burner and Spray Evaporator
Heat balances were made on the system to determine the amount of heat
lost from the process equipment. These balances showed that 68,000
kcal/hr., or approximately one-third of the heat being supplied to
the system, was lost to the surroundings by the burner and the evap-
orator. The bulk of this loss (52,000 kcal/hr.) was lost through the
burner shell. In a commercial burner, where the combustion air is
preheated by the burner shell and where the relative surface area of
the shell is smaller, the burner loss would be a considerably smaller
percentage of the total heat input.
In order to establish the thermal requirements for evaporating acid
feed at various levels of concentration, the run data were correlated
in the following manner:
a. Since the burner operated at more or less constant
temperature throughout the campaign, a constant
heat loss of 68,000 kcal/hr. was subtracted from
the theoretical adiabatic heat input of the burner
(based on natural gas consumption). This yielded
the net heat available for evaporation of the acid
feed.
b. The acid feed was calculated on a solids-free basis
to factor out differences in the feed due to the
presence of greater or lesser amounts of solids.
35
-------
c. The net heat consumed per kilogram of 100% HoSO.
in the feed was plotted against acid feed con-
centration. This is shown in Figure 12 for an
average evaporator outlet temperature of 350°C.
The curve in Figure 12 points out the advantages (in heat duty and
evaporator size) in preconcentrating prior to spray evaporating.
Since the data used in generating this relationship had heat losses
factored out, a heat loss factor must be reintroduced for any com-
mercial scale-up. For example, for slag based acid which has been
preconcentrated to a level of 357. H2SO^, 21% solids, 44% water (44%
l^SO^ on a solids-free basis) the evaporator heat load is 2000 kcal/
kg H2SO^. If a 10% heat loss factor is used, the heat duty is in-
creased to 2200 kcal/kg HoSO^. (This heat duty is only for the evap-
orator. The heat required to preconcentrate the waste acid to 35%
H2SOA must be included to obtain the overall thermal requirements.
Basea on calculations, preconcentration to this level would require
an additional 2000 kcal/kg H2S04.)
B. Acid Condensation Train
In the pilot plant, the partial condenser and separator were insu-
lated to minimize heat loss. The purpose of this was to maintain as
adiabatic a system as possible in order to fully evaluate condensa-
tion of acid by spray cooling. Under typical conditions the enthalpy
entering the partial condenser was 265 kcal/kg of dry gas and the en-
thalpy of the gas stream entering the mist eliminator was 252 kcal/kg
dry gas for a heat loss of approximately 5% through the vessel walls.
8. RECOVERY OF SULFURIC ACID FROM PICKLE LIQUOR
The use of waste acid from steel pickling operations as feed to the
spray evaporator was evaluated apart from slag and ilmenite end li-
quors. A trailer load of typical waste acid was obtained from a
Pennsylvania steel mill. The recovery of useful acid from this mate-
rial was not encouraging due to the low sulfuric acid content and the
high quantities of ferrous iron (2.1% H2S04, 18.5% FeSO^). Only 54%
of the sulfuric acid was accounted for as low concentration product
acid (11% H2S04) and acid mist - almost all of it being as mist. The
balance of the sulfuric acid (46%) was lost as ferric sulfate and
sulfur dioxide. Significantly, no acid was lost as free acid in the
solids. The dried solids had the following composition:
36
-------
£000
Q
Lu
U)
IL
2
*
0
^4000
0
g
\D
!£
U
30OO
U
o
O
in
cQ
v
O
LU
2OOO
25> -3
-------
Iron As Evaporator Solids Cyclone Solids
FeS04 62% 72%
Fe2(S04)3 35 26
Fe203 3 2
Raw pickle liquor does not lend itself favorably to acid recovery by
spray evaporation. In addition to having a low ratio of acid to
solids, the crystals of ferrous sulfate had an eroding effect on the
nozzle. Blockages also occurred more frequently than with other
feeds. Consideration was given to a process whereby the pickle li-
quor could be concentrated to 20% H2S04 prior to evaporation. How-
ever, this approach would require cooling, crystallizing and concen-
trating the pickle liquor. Since the pilot plant was not equipped to
make this separation, no attempt was made to implement this modified
process.
9. PRECONCENTRATION OF WASTE ACID
From a standpoint of economics, spray evaporation of a waste acid
which has been previously concentrated from nominally 20% H2S04 to
30-40% H2S04 is desirable. At the 35% H2S04 level, for example, the
preconcentration step removes more than 50% of the water without in-
troducing any significant problems of iron oxidation (S02 formation)
or acid mist. This results in a reduction in size of the spray evap-
orator and other process vessels.
The preconcentration of waste acid was not integrated into the re-
covery pilot plant at Palmerton since it was felt that sufficient
technology existed elsewhere to accomplish this result(l). Among
suitable methods for preconcentrating acid feed are:
a. Vacuum evaporation of acid feed which has been pre-
viously cooled to crystallize and separate salts.
Montedison of Italy(6) currently use this technique
to obtain an acid which contains 45% H2S04 and 6%
solids.
b. Submerged combustion. Included are processes de-
veloped or proposed by Chemical Construction Cor-
poration(3), Selas Corporation(7) and Thermal Re-
search Corporation(S).
In order to evaluate some of the parameters associated with precon-
centration, a test program was run at the Selas Corporation pilot
facilities in Dresher, Pennsylvania. In these tests, waste acid was
38
-------
preconcentrated from 18% H2SC>4 and 10% solids to 40% H2S04 and 22%
solids. The acid was single phase at the normal operating tempera-
tures (110°C) and presented no difficulties in handling. The degree
of iron oxidation in these tests was nil. Acid mist in the stack ef-
fluent ranged from 1240 rag/standard m3 (dry) at 33% H2S04 to 4940
mg/standard m3 at 39% I^SO^ acid concentration. Tests indicated
that 75% of the acid mist was smaller than three microns in diameter.
It was concluded from the tests at Selas that no unusual problems
would be encountered in preconcentrating the waste acid stream prior
to introduction to the spray evaporator. However, some solids block-
age did occur periodically in the concentrator throat and this would
have to be resolved before commercial implementation.
10. MATERIALS OF CONSTRUCTION
The process to recover sulfuric acid by evaporation and subsequent
condensation encounters a wide range of acid strengths and tempera-
tures. In order to cope with the corrosion associated with these
conditions, a number of materials were evaluated and their suitabil-
ity is discussed below.
A. Cor-Ten Steel
The spray evaporator was constructed of Cor-Ten steel lined with acid-
resistant refractory, with the exception of the outlet gas duct which
was unprotected by refractory. The evaluation of this outlet duct at
the end of the 1972 campaign showed little evidence of corrosion de-
spite the fact that it was situated directly beneath the atomizing
nozzle and was subjected to a number of upsets. In order to confirm
this performance, four Cor-Ten test panels of approximately 0.5 m^
each were placed in the evaporator for the 1973 campaign. When these
panels were removed, they showed an average corrosion rate of only
0.38 millimeter per year. The conditions of operation in the pilot
plant were almost certainly more severe than would be encountered in
a commercial plant where start-up and shutdowns are not nearly so
frequent as in a pilot operation.
Cor-Ten steel was also used for the cyclone to collect solids from
the gas stream. No evidence of corrosion was apparent after either
of the two campaigns.
B. Lead
Lead was used as a lining in the spray tower and as a membrane be-
neath the acid-resistant refractory of the primary condenser and
39
-------
separator. With the exception of some minor acid leaks which devel-
oped in the duct work leading to spray tower (attributed to improper
installation) the lead functioned very well.
C. Acid-Resistant Refractory
Castable, acid-resistant refractory installed directly over Cor-Ten
steel was initially used in the primary condenser and separator.
Substantial downtime was experienced due to acid corrosion of the
shell in the areas where the refractory cracks developed. The de-
cision was made to replace the castable refractory in the primary
condenser with acid brick over a lead membrane, and no corrosion
problems were subsequently encountered. Similarly, the separator
was also provided with a lead membrane beneath its acid-resistant
refractory and no further corrosion was observed. An examination
of the acid-resistant refractories showed no apparent attack by the
85% H2S04 acid which was present at temperatures of up to 200°C.
The lead membrane is highly desirable insurance against improper
refractory installation and the vagaries associated with handling
sulfuric acid at elevated temperatures.
D. Plastics and Fiber-Reinforced Polyester
Kynar and FRP were used satisfactorily in applications up to 50%
I*2S04 an<* 100*C f°r lined pipes, pump housing, impellers, valves
and process vessels. Teflon-lined pipe was used satisfactorily
on all lines which carried concentrated acid (85% 112804) at ele-
vated temperatures (up to 200°C).
E. High Silicon Iron
A high silicon iron pump was used to recycle hot (up to 200°C), con-
centrated (85% 112804) acid to the primary condenser and separator.
This pump exhibited superior performance during the project, and an
examination of the housing and impeller at the end of the campaign
(400-500 operating hours) showed no apparent wear.
F. Special Alloys
Nickel and steel alloys were used in limited applications for atom-
izing nozzles and lances, gas and liquid distributors, thermowells,
and flowrators. These alloys, which included SS 316, Alloy 20,
Hastelloys B and C, and Inconel 625, performed predictably in the
several environments in which they were tested. In particular, Alloy
20 gave excellent performances in dilute acid applications (20% I^
at temperatures up to 80°C, but was found to be unsuitable at high
concentrations (80% 112804) and temperatures above 180°C.
40
-------
VII. DESIGN OF COMMERCIAL PLANT
As a result of the pilot plant performance, a preliminary engineering
design of a waste acid recovery plant using the evaporative technol-
ogy was made. The design specifications were for 15% excess capacity
beyond a base equivalent to 38,100 metric tons per year of titanium
dioxide pigment and 67,200 metric tons of waste sulfuric acid (100%
H2S04).
The capital cost of this recovery plant (as of January 1, 1975) is
estimated to be $7,800,000. This figure includes salts neutraliza-
tion and all normal site preparation, but does not include the cost
of land. Operating consumptions are summarized in Table 11.
Table 11. BASIS FOR DETERMINING OPERATING COST
Production
Titanium Dioxide
H2S04 (100%) as Waste Acid
H2SO^ (100%) Recovered Product
350 Operating Days Per Year
Utilities*
Fuel (No. 6, 2% Sulfur)
Steam
Process Water
Cooling Water
Electrical Power
Lime (as CaO)
Labor (Estimated)
Operating
Foreman
Analytical
Maintenance
Taxes, Insurance
*Based on preconcentration to 40%
oration.
109 Metric Tons/Day
192 Metric Tons/Day
165 Metric Tons/Day
3,090 Liters/Hour
757 kg/Hour
13,323 Liters/Hour
47,994 Liters/Hour
1,350 KVA
512 kg/Hour
7 Man-Days/Day
1.5 Man-Days/Day
0.5 Man-Days/Day
10% of Invested Capital
2% of Invested Capital
H2SOA prior to spray evap-
41
-------
At current values for fuel oil ($7.95/MM kcal or $2.00/MM Btu) and
assuming depreciation at 10% per year, the total operating cost of
such a plant is estimated at approximately $77 per metric ton of
100% I^SO^ recovered. This cost includes hauling and disposal of
the neutralized evaporator solids at an estimated cost of $3.00/MT
of solids which is equivalent to $4.80/MT of recovered acid.
Before such a commercial plant is constructed, an in-depth environ-
mental study of its potential effects should be made. A significant
amount of energy and material will be consumed in the recovery of
acid. This is summarized in Table 12.
Table 12. ENERGY AND MATERIALS IMPACT
OF COMMERCIAL PLANT
Per Metric Ton of
Annually Recovered H^SO,
^(^^^^t^™
Fuel Oil* 23,400 MT 0.40 MT
Electricity 11,350,000 Kwh 196 Kwh
Lime (as CaO) 4,300 MT 0.07 MT
Disposal of Neutralized, 92,400 MT 1.6 MT
Wet Solids Ǥ 50% H20)
*Does not include fuel for lime slaking or hauling of solids to
a disposal site.
The disposal of solids is of particular significance if the plant is
in an urban area. Approximately 30,000 m3 of wet sludge will have
to be disposed of annually, and this may require hauling over great
distances if landfill areas are not readily available.
42
-------
REFERENCES
1. Duecker, W. W., and J. R. West. The Manufacture of Sulfuric
Acid. American Chemical Society Monograph Series; New York,
Reinhold Publishing Corporation, 1959. pp. 297-345.
2. Umstead, C. H. Evaporative Sulfuric Acid Recovery from Sul-
furic Acids Containing Sulfates. U.S. Patent 3,713,786,
January 30, 1973.
3. Shah, I. S., and J. B. Rinckhoff. Method for Concentrating
Dilute Acidic Solutions. U.S. Patent 3,789,902. February 5,
1974.
4. Experimental work. Research Department, The New Jersey Zinc
Company.
5. Latimer, W. M. The Oxidation States of the Elements and Their
Potentials in Aqueous Solutions. Englewood Cliffs, N.J.,
Prentice-Hall, Inc., 1961. pp. 70-81.
6. Private communication from Montedison S.p.A. of Italy to The
New Jersey Zinc Company.
7. Selas Corporation of America, Dresher, Pa. 19025.
8. Thermal Research and Engineering Corporation, Conshohocken, Pa.
19428.
43
-------
ACKNOWLEDGMENTS
The authors wish to acknowledge with sincere thanks the support
granted this project by the U.S. Environmental Protection Agency
and the help provided by Dr. Herbert S. Skovronek, the USEPA
Project Officer.
44
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
REPORT NO.
EPA-670/2-75-016
2.
3. RECIPIENT'S ACCESSION-NO.
(.TITLE AND SUBTITLE
THE RECLAMATION OF SULFURIC ACID FROM WASTE STREAMS
5. REPORT DATE
April 1975; Issuing Date
6. PERFORMING ORGANIZATION CODE
. AUTHOR(S)
Howard C. Peterson and Peter L. Kern
8. PERFORMING ORGANIZATION REPORT NO.
3. PERFORMING ORGMMIZATION NAME AND ADDRESS
Research Department
The New Jersey Zinc Company
Palmerton, Pennsylvania 18071
10. PROGRAM ELEMENT NO.
1BB036; ROAP 21AZQ; Task 07/08
11. CONTRACT/GRANT NO.
S-801349
12. SPONSORING AGENCY NAME AND ADDRESS
National Environmental Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The New Jersey Zinc Company process for acid recovery employs spray evaporation to
separate sulfuric acid from metallic sulfates. The salts are removed as dry, free-
flowing solids and the acid-laden off-gas is directly cooled to partially condense
product acid having a concentration in excess of 85% ^SO^. The process was piloted
at Palmerton, Pennsylvania, at a rate of two tons per day of sulfuric aicd (100%
basis) using as feed the waste stream of a titanium dioxide pigment plant. On the
basis of the pilot work, a commercial plant was designed to process 345,000 metric
tons annually of 19.5% ^SO^ waste end liquor from a 38,100-metric-ton-per-year
pigment plant. The estimated investment (as of January 1, 1975) is $7,800,000.
Operating costs (including depreciation at 10%) would be approximately $77 per
metric ton of 100% ^SO^ recovered. This cost includes neutralization of the dried
solids and disposal in a landfill site.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
*Sulfuric acid
Reclamation
Waste treatment
Operating costs
*Evaporation
*Titanium dioxide
*Sulfuric acid reclamation
*Titanium dioxide manu-
facture
Titanium dioxide byprod-
ucts
13B
18. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (ThisReport)
UNCLASSIFIED
21. NO. OF PAGES
53
20. SECURITY CLASS (Thispage)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (9-73)
45
U. S. GOVERNMENT PRINTING OFFICE: 1975-657-592/5355 Region No. 5-11
------- |