C
0;^
|°-
»•
O C\
•D ft O
4> 00 (-0
l/> 1
3 •-
tf) r— C
•r- O I
ECU.
UJ *- 1
0
o •*•
u. E <*•
0) u.
4J
E 1/1
O X a-
I- l/) C
Li_ O
•o
«t c >
E.= i
o a. —
LL.
0)
1
OJ
>
«c
II
X C
£§
a;
>
«t
x 5
£ J
Average
i I
E E
x c
g £
•M
3
O
C
C«J
• o
e (/i
•i
'Q_
t—
CM
O
l/l
1
0
no.
h-
CM
J?
O
Q.
3 1—
.?
« LO
!i
- cx
1—
r
>5
Z
U-
O
0)
CM
O
*s
=:
1/1
0)
C"
z
c
z_
z
0
l/>
E
o
l/l
Blending Conditions
VO CSI i— ••-
^- m m ir>
ro m «j- LO
10 o ^ ^
*s- * •» «*
o o o o
CM CM OJ CM
ii — r^ r-~
if) U3
m oc ^o r^.
ID CM f» CO
CM
t in ui ID
f CM CM CM
s s s
I C^ O O
i tr> ir) ID
ui m ir>
If) If) If) If) If) Uf)
I 00 ,— 00 ^ CC ^
1 CM CM CM CM CM CM
8O O O O O
O O 0 O O
I O O O O O O
i in ifi if) in if) if)
LD LT) If) LT) If) Lf)
CM CM
r*> LD CM ey>
f*^ *f in m
*— -— co'S' co oc"
\oo^ ^-o OOCM *r^
^•CM l/l^- Ifllf) VOVO
s s § s
O O f*l r*l
r-^ CM ^- m
CO CO C^ 00
o o o o o o o
O O O •— ^- CJi OJ
I*-. CM O VO O ro f—
r»- csj ^ fi LO CM ^~
b
A. RF + Conv SC/SA Acid Plant
B. [RF (Lime/Limestone)]:
Conv
C. [RF MgO] + Conv
D. (RF Citrate) + Conv
o
to
E
O
o
o
-o
to
E
•o
CU
to
(SI
o
o
Q.
CU
-o
O)
o
T3
E
CU
to
cu
JC
o
to
CU
to
CU
ts>
M- Q.
T3 (J
O) 10
T3
E U
CU •!-
S 3
E
ID
Q.
T3
U
10
u
S-
3
U
fO
o
o
a;
en
E
to
"O
QJ
to
-------
6.1.4 OTHER BLENDING CONSIDERATIONS
This section will examine blending scenarios which will enrich
the S02 concentrations of the gases sent to the acid plant as well as
decrease the volumes to be handled by such systems. The strategy
used will entail processing different combinations of smelting equipment
offgases, other than only the RF, in FGD systems prior to blending
in hopes of finding an economic optimization. The FGD systems
applicable to such an approach include only the MgO and citrate
since these systems not only reduce the offgases but also concen-
trate the resulting offgas SOp. Figures 6-10 through 6-19 show
these various systems devised to handle offgases from the copper
smelter assumed for this study.
The lime or limestone FGD systems are excluded since their
application for processing stronger SOp concentrations may be
limited both economically and physically (Section 5.2). The scenarios,
as discussed herein, typically will process offgases from MHR, FBR,
or converter gases along with or without those of the RF. Thus,
these processed gases will have higher SOp concentrations due to
the assumed S02 concentrations from these process equipments.
Table 6-6 is similar to Table 6-4 and considers the charac-
teristics of the blended systems as indicated. All assumptions
considered for Table 6-4 are applicable here (see Appendix Y).
Table 6-6 indicates that the design capacity for the acid
plants (based on maximum volume flow) will change considerably
from the base cases, depending on the scenario. Reductions
in volume achieved are as much as 79 percent (i.e., case No. 50);
however, the Op/SOp ratios of the resulting blended gases may not
be sufficient for conversion of SOp to S03 (refer to Section 4).
Table 6-7 lists the 02/S02 ratios for those scenarios presented
in Table 6-6. As shown, many of these ratios fall below the minimal
1.2 and, therefore, would require additional oxygen for proper
conversion. Case Nos. 19, 23, 27, 30, 31, 32, 34, 35, 36, 38, 39,
40, 42, 43, 44, 46, 47, 48, 50, 51, 53, 55, 56, 57, and 58 are all
250
-------
CASE 19 \ 27 - MHR
CASE 35 I 43 - FBR
Figure 6-10. Flowsheet for Blending Schemes Using
Regenerative FGD Systems
CASE 23 i 31 - MHR
CASE 39 I 47 - FBR
Figure 6-11. Flowsheet for Blending Copper Smelter
Offgases Using Oxygen Enrichment and
Regenerative FGD Systems
251
-------
CASE 20 I 30 - MHR
CASE 36 > 46 - FBR
Figure 6-12.
Flowsheet for Blending Schemes Using
Regenerative FGD Systems
CASE 24 t 34 - HHR
CASE «0 i 50 - FBR
Figure 6-13.
Flowsheet for Blending Copper Smelter
Offgases Using Oxygen Enrichment and
Regenerative FGD Systems
252
-------
Figure 6-14.
Flowsheet for Blending Schemes Using
Regenerative FGD Systems
CASE 25 I 33 - MHR
CASE 41 S 49 - FBR
Figure 6-15.
Flowsheet for Blending Copper Smelter
Offgases Using Oxygen Enrichment and
Regenerative FGD Systems
253
-------
CASE 22 t 28 - MHR
CASE 38 1 44 - FBR
Figure 6-16. Flowsheet for Blending Schemes Using
Regenerative FGD Systems
\F.M.
/""I'M W* — ***** S. ^ i f|
COWL. — *• ^u^u^ ~* M/
Yfil
E5P
SE 26 > 32 - MHP
SE 42 t 48 - F3R p>
, '
i
UVHCMMOn
w
am
•
WWfl
,
WOT
ESP
1 — 3. — 1 * A050U
C/7WWJ *' 1 1
»
r
.— »
.37WH
COvV
i
C.A
wor
ESP
8.5.
• AT
A. C.
, ACID
fLAHT
(STACK)
Figure 6-17.
Flowsheet for Blending Copper Smelter
Offgases Using Oxygen Enrichment and
Regenerative FGD Systems
254
-------
!£&*;
COWT _»"*•***• _.. iy/
WJIVA,. * iv«ive "'
i
W.
E5P
CASE 51 i 52 - HHR
CASE 55 S 56 - FBR
j— i
— «
/^v£ea
1
WHB.
|
-J«T-
£5P
Eo=jSS
jc^euz *> — j — 1
— »
— »
COW?
1
CD.
\
HOT
ESP
lyJWHt
1
-•
8.5.
^ A f
^ A. C.
, ACIO
PLAHT
Figure 6-18. Flowsheet for Blending Schemes Using
Regenerative FGD Systems
CASE S3 S 54 - MHR
CASE 57 t 58 - FBR
Figure 6-19.
Flowsheet for Blending Copper Smelter
Offgases Using Oxygen Enrichment and
Regenerative FGD Systems
255
-------
co
3 Z
Z t-H
I—
-«
o: a. LU
0 I—
U_ CD
O Z
co o i— i
O «* 1—
1 — 1 * 1
ry i — LU
O
,- oS
3 |
•o
OJ
£o~-2
O -r-
Emissions
From FGD
so2
Emissions
i
•4-*
l/l
O
c:
o
+j
u
4->
t.
a>
U
X
•M$ C
C id
I- 4->
£
•o
U
«r
>o
!c
"3
^
3 CM
ss
=s .
o t-
£
a. x <
(/> «
0.0,.
(J 3
it
•M LU
V}
>>
O
ON
•4-J
C X
a. -*^-
£
i~
QJ
>
<:
ET
X C
?f
OP
>
«t
Maximum
(Minimum)
0
c
•o
c
QJ
oc
J
3
a
a.
a
i.
CM
3
i/)
Q
Q.
-
o4
to
i.
h-
CM
S
X
U
(/)
CM
O
LO
*<
JE;
u
at
o>
2
_^^
E
C_J
L/)
Z
li_
CJ
l/l
o
+J
c
o
— »«o »— o m r^ (*•> •—
r^CD CMOr^ O CM O eg •— CM O co ^~
& a* encnao o^ cr> ovoocneyiCTNco cr»
Sr^ »jo o ^- m vo e7*oui^-coco ^i
oo Scoo oo ^o r-.Of-mr^°N r-
So; oio;^ <^ w o^wo^o^cj;^ ^
,_. ^. m «tf- ^ r^ rx ^ ^ r^-Lnr^^-^^1 f^
r>-in ro«*cc to m «*r-*^-o->uir^ •*
,— O mOesi O> rn cscMvoroeMCM ID
^- *t \oirir-»i — vo ur>
§o o o *o o *o jooooo^c o
o oow o 01 oooooo^ o
o ooo oo oo oooooo o
lt> \O I^.CM^- CO CO CMinCMt^iD^f CM
CMCO OAOOCM CNJ CO COCMOOCTirnvM CO
r-^ 0*0^ ino Or- ror-. »— i— CNJ inoo^oo-—
£ OCM oort ^-O mm «tf-in co^- ^-o coin «-o oor-~ OCM coin - in ^^SS^^S 2
^cco r«-^or- ro r-. r— r*. cs% CM vo «3- co
IA r-t in ^ »XS r«- m moiunr-^CMfM vo
o^co t— •— OOCM vo«r r^.o <-in OOCM COCM CMP- CMCM Or— COCM r^-r^. vo*±
i^i 9 oovo vo^o inco f-.m co vo vo vo vo^ mv.o co^~ P^.CT» r^.co vooo crt^r
gg 22S C 2 SSS5BS £
ino CMCMCT> in CM f— to *± m o CM m
* » •,«•> " » *•.«•••«• «
ve.— vooro O »O r-^OiO^-r-^r-* £.
oco omcr» co o »— moco^rf co
CMO CMO OO CMO mm r-O CMOO COvo com i— vo vo^- ^-vo uoo I^CM
Sm mco onro ONO co«a- Or- votn cocn r^. un ^-r— covo CM«a- ro«— <*•>•—
r». COLO com CM«* o«— «J-o cft^- r-CM CMUT* ts*in» l/1^, ^^ »• "~1 .
cMr-T fTt-^" vom vo»— om r-»r- voco coco cMr^ coco CSJCM CTI^T ouo °l£?
coco ovo ^-vo in*— cMr-. ovo ^-vo ^-ON voi/i OCTI CM^- ^-«a- in^r crioo
xfe +m |> x x+ |, x_ +u_ + a _ fe si-
*CTC/1 (/? t/l'trt 'a) ^i00-" •" 5 X -"2 O ^O^ -^ £ "- % CD «- »Z ZOL. C — "Z *-— O SE •— ' 0V to i-*J 0-M t_)CJ>E cjuj
A^1 -&1 'i.10 AK 4^£ £A+ £A+ 4^ *G fu. ^S *£ f*u f .v,
'? f" /k° Ti 'A ~ ^1 -J^i I" '* 1= t- I- l!'i l£
>" t>I TA >+ >'S SfS Si-S i* >' »+ ti ii >|- >-
ct^ 'cZ 4 = = STj? SI Si SI Si Si S,7o~ S +
(jio r-CMPO •* Lf) VOh^COCT»O<— CM
,— CM CMCMfM CM CM CMCMCMCsjmro fO
256
-------
OO
3 Z
LJ O
O CQ
Ss:
u. o
oo
o ^
I—I I—I "—.
OO ryf Q
OO UJ UJ
I—l Q ^
2:1-1 -z.
UJ OO i—<
2: h-
CMO Z
OOO
OO O
UJ
CD I—
_j I— oo
_i ^ <
O UJ CJ3
o: o u.
i— z u_
z o o
o o
O I I—
Q -Z.
cc a. uj
o i— s
U_ D-
O 1-1
oo o ^
o «* o-
i— I » UJ
ea: "~" C3
Z CD "Z.
C/0 OO _ I
00 UJ
CD uj s;
z: o oo
1-1 O
Q Qi _J
Z Q -- 1
CQ UJ U.
I— O
o oo
Q Q.
Q Q-
< O
CJ
UD
(1)
S
(O
4-> •— t
s- c T:
C '1-
O (.
O f
to E
i E
UJ U
*
O
CD
S
U
irt
U
U
|Q
JE
O
CM
O
u o CL.
•o
u
C
8 =
3 CM
*• o
0 i/l
O
c a.
OP-
e
-C
(/I
o :
&£
>>
CO S4
i;
4->
C >
Ji
i.
a*
<£
j~.
E E
E E
X C
""
1
X
<
E
X
-
4->
C
CM
CO
-0
0.
p-
••— *
CM
0
a.
^_^
rT
O
O.
CM
C O
LO
a.
H-
CM
O
CO
x;
u.
to
sf
^
u.
CJ
J CJ)
5 c
-£
u
7>^
> CO
z
5 U.
I <_>
— CO
o> c
C O
Q 4->
v -o
— c
U 0
0
m 10 ^-tn o vovo CM o ps.^-^moo
a, » eo* a^ 0,00 <*
r^ r-. CMOO oo com ro oo coro^-e^^
S S Ooo S f**O CO ^» r^Or^^Of**-
ps. P^ 00 0 00 0 0 00000
a^ o^ oo o oo o o ooopo
ID 01 r-^- if) ^s-in e^ in *r •— «y in «*
m S rCS S £S S n ^r^^rmin
*t T— i— *t n *fO *r ro ^-CMO^-m
O ^ i— f-^ csj t— o rv CM omr^-ocsi
s: 5 s^ » mS 5 - ^s^i?^
i I § § 1 i § I i § § 1 § 1
**> m in to r-T CMUI en •* S^S&^iS
CO CM OfO r>. 000 £J «3 COOOOr^M
r— " mr*. o d (?i O ^DP-» a»p-» POP*. r-» i— —, °* ^>^S^?lS
rn^ Sin oooOco inm ^a-oey*co *3-m CMO ^ooco^Oi^rooeM
mco ^-CM \Oin ^-CM rof> m^- min -SI-CM *^- m^- >oin LTIO- rom ^-CM
— ,_, — — ^_, — *_. ^—
si if 1§ if §1 §1 If 11 ii §§ §§ §§ i| II
m PO ror-^ inrovO"* r^P*- CMOtfifo ro^- ^T'J ^SiS^^StiE^fSSi
coco 1000 «-\o rs.cn I-.P-. CM^- ^ \o ^oco tnio cs*^- ^-^D CM<- r^r* r^trt
*-* *-" ^-^— ^-- «— ^- -— *—
PO O VP-. in CSIP-S «> m ^IT^jrilri
(M CT> v\O i*- *j-f— >0 r*. l~^^'~;^.
SS SS SR SS SS S? £2 SE 2S 5° E£ S? SS S£
^£ tS "S "2. "S ^^ "" *£ "iS. ^S "~ °S ^S SS
S 1 S 5 ? 5 § 5 2 1 § § S S
SVD iO»— ID OCO *— "^ ^^dlS^J^
5f 0*0 0 ^cr. «5 0 cnincOQOCM
1— 1— i— •—
coST ^^ OGO'CSJO' O^OT" OToLr-aT ^^" COM' CMO vo o j- « 5£[£^ ^^
§S S§ SK 5? SS S§ 5^ 55 S3 2S RS S= ES SS
a"s" s's" S'B" ss" 52" ss §s §ss ss ss ss ss ss ss
— — - — — — — • — — ~_^- — —
^ If f: if i~ ! f-i- ^4 ^| ^ 1 i 1 1
£^> (O*^ tOWl 4JU1 4->(/l CO I^W£ PJ11- pX +§ ?4. ? ^?c £-h
U<— ' 0 4-> ft I 0 00 >,0 >>tO 0 0 CO U U 0 + U CD 00 4J 4J 4J 0 *-
•F-ft)O'^*EZ O* I/I O* tO CT> CTl -^ •*" tJ •"- ^ li- "' ^ OjJ O i- C_34->
ti5'S t I'a I* ?T f f > T£ T?^~ "tss, "^s> T ~ T-- T" i" T-,
itl p il lis 11? j; lit m if? k j! i« I! il
o-) m mro ro mt'o «r T •O-^-ST ^
257
-------
oo
3 Z
LiJ O
O CQ
LL. O
CO
O Z.
I—I I—< -—•
co a: o
co LU LU
LU co i—<
•z. i—
CNJO 'Z.
O C_> O
CO O
LU-—
03 h-
"Z.
(_) I I—
a -s^.
rv* Q_ 111
o i— s:
LU a.
O >-"
CO O ZD
O ** cr
i—I •> LU
a: ^
•a; o
LU "Z. i—i
o I-H i—
co co —i
co LU
z o co
M O
ca Q; i
•Z. Q 1
LU o:
_i a;
co LU LL.
I— O
<; LU
z s:
o co
Q Q.
a D.
<: o
o
10
i
10
(U
c Q -o «
OJ t- 4) C
(J 4-» trt
i- C IO
(U O CO
0_<_J
8 §
.;: c
i— VI 4-
5 £
o E
h- LU
K 5
UJ
(/I "C
O L
CM ••- <
irt 1/1 E
tu u
1
§
s
I-
1
0™
00 .^
*£ **"*
OJ
0 •*•>
tj
Is
o
3 CM
3 (/>
1
O
§0.
1-
tt*
ex >
2 *;
«
CT* QJ
C >
•5 *-
3 •«-
7,2
C U.
™
£«*•
0) V
+J U.
X
c
cr
c >
en
o>
E^
K C
<
<
5
1
z*;
a
|
o
to
o
Q-
t-
CM
O
t/1
ro
o_
t—
CM
O
to
o
t-
0^
*?
a.
i—
CM
O
to
**
z
u.
VI
CM
O
Z
u_
o
t/1
„&
C i-
IT
=s
5 tj
f
z
iu.
-"io
to
Cf> C
c o
O *->
a; -o
— c
D O
(_>
SOr-cMo'cOOOO '"cOO'iS
i^ m a* « N « •- lommvo
crr-.icr^or-crt r^cfvr^cn
O000^r~h- r-000
CM (VJ C«J CSJ t— *— >— t— CXJCSJCSJ
in «• in ot vo « w mvoo>o
SSSSi-^SS SSSSS
cvju,«-ino>mc, « « n 0
v in 0 N oo « in oIoda^CT,
^ ^' in "O *o m in ^~
tn m i— «*• m in co co O o m
ui«s-*rmrorom co^^r^s-
oooo^rotc untnco^o
,— ocnco^co^ C^CM^O
OCO^OCMLOO^- CftPOCOCM
a»r^" "~*r^^' r>rT rs.co' rv.co ui in coco coco oo
mm m^ ^2: ^^i ^^2. *C. ^C *^2. *2- *S S
oo oo oo oo y> ud incr> mo IDCM ^vo o in CMO
U-»PO CMOV^- f>i — i— ro CMO VOCJ* COiO O**CM CMO roco
— — ^— '^*— ^- *-* ^ ^ *-*
*OCOCOf— C3OOO
CMCOr^»-COCOCO COCOCOCO
*-*. f-^ r— , f— . *— » *— H ^— * •"-«• .^^ *— * ^~"
cril? CMO CMCM coai oo oo oo oo oo oo oo
loao CMior^CT COLO co oo coco ao co coco coco coco coco
sls2s§i R§SS
r^. oo *Y r^ CM ^ (si inro m PO
r^o Sr^cM^o f^-cn oo roco ^-r— LO co ind CMin mcM
oco i — ^""^i"* CT»O roo CTitsj vco iin f> «*"O ro^ o^—
*•! .AC* ••» - » •• »» •» •>* *B **
S-o- i-v«3 CM** CMCSJ o^- — *r o<* ^"* o^- — o- o^-
^^, ^^, +^, ^_^ ^^, ^^- ^_*. -^-- *_•• *_^ ^-«
L_ A -f -0 C C
^-^ ^7 ^ DC 'f*-*^ Q£. OJCM-O O>OJ tJ +J1O
O>4-> • "— 4J 08 E -CO)
*j ^^, — . *Jc d)Li_. *ja^Li_ «e oe-ona • — • c >cow B "*S c E -^
.,- t. E •*- o •*- v *-) -f- •«-rQ? 01 z co (_> -D s_ - — -o a> z to ••- ~o
i'.'^'o <»'-' c«+ ^AC^O oTSo c/-oo cc^o^ oTco ocmo QC-D-M cec.0)
TO-r- JU- llll-,— * »UjE Z *J IC-U X. JUJ4J CO *J 000)10 COUJf-
1 4 c K o^ 5 I > i-- T i- > - — - T c > a) c>ai c-i-r- C"— 'ii c>at cr u c— Z
coccMcc£ T-*-* ^ f a> c i- DCS- a* i_ co o ai i- a* c i- a> TJ *> *o
OCOO OCQ U-I— OlJ_] »— O*J r-O-w •— C O> i— U. +J ^-O+J • — CO i-U.CC
r^ CO CT> O •— CM CO ^j-unwDf^
^j- ^ ^ in m in un in in ur> in
258
-------
oo
3 -Zi
LJ O
O CQ
CC-SL
LuO
O
o
(/)UJ LU
i-HQ =3
•SL^-Z.
LU{/) i— I
z: h-
(MO ^
o o o
oo o
1 1 \ ^ __ '
C5 1—
z <: oo
t-i C£ LU
_J I— 00
_J Z -, O
o o: _j
Z Q -- 1
LU <£.
_j o;
CQ LU LL.
I— O
_1 _1
e£ LU
z s:
O oo
i— I LU
a a.
a a.
ID
I
0)
(O
4J i— O CM
C O ~O ^O
U t. Ol COCO
? C* * o
u o eo co.
b. O Of-
Ol
CM I/I L.
O C O»
OO O -C '
•,- O O.X
° LU «:
2
i— C C >
O CM O T- •«-
L. 0 - T) 4-)
-> OO W) 3-*-
5 ••- *o :
u E cu-
C UJ i-t
cS ««-
O U- E •«-
CM -f- OJ LU
4-1
3
c
—
CM
o
1
a.
t—
CM
o
CO
D tfl +^ <
co wi E v> O
•t- o x»* a.
wut^cn
c
O U •*-> • CM
CM ••- «S C X O
3 V» <0 fC
I
W)
CO
o
o
LL.
s
u
VI
O)
ID
X 3 C
I-.
O) O
> CO
I'
X
£
z:
I-
E"
— to
tA
C O
f -t-
C i-
a> x>
-— c
CO O
« CO en
i-^ r-^ r-
en co co
en en «
S | 5
O •— «•*
M CO OO
1 1
5
^r co CM
en co CM
i i
o
*"2. i •
S
f_ "^
• • i t
in ^>
$ro
CO f**
o vo in
CO ^ in
O O co in O CO
CO CO "^ *, *°i2.
CO 00
S 0 ^O
in in in
I-* CM O
f^.^ O O^ O O
*o »— o o o o
r— f- o o LD to
c\j en in ro in co
— • — • >—
£ ai t- J^
> -i- -O * "O U- "O
C i. OJ S- LL. C C£ C
O C QC IT) *O
ecu •- aj or) >cy(/i
• CNJ > i— > *—
oiO'OO c f— cri C^CT>
ca— c 4J ooc ooc
c *^ n o ~*~ o •*—
aj -o c u ceux ce u x
i— C O) 4J XCLU CQ C UJ
co «j E uo !:=>— U-ZD —
Sen o
in vo
c
o
u
o
T3
IO
C
•o
in
01
o
S
a.
e
m
TJ
at
IO
u
a.
•o
o
in
ai
"c
o
10
o
L
o
(O
c
o
o
in i—
>*- a.
O T3
T3 U
ai 10
•a
c u
01 -i-
IJD 3
Ol
u
£
a.
01
tn
10
en
T3
Ol
•o
i 03
JD
259
-------
Table 6-7. OXYGEN TO SULFUR DIOXIDE (02/S02) RATIO
FOR ADDITIONAL SCENARIOS AS PRESENTED IN TABLE 6.5
Blending Scenario
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
02/S02 Ratio3
2.74
2.63
2.71
2.57
2.50
2.71
2.54
2.49
2.19
1.92
1.92
1.83
1.77
1.93
1.83
1.77
2.99
2.19
1.00
1.45
2.00
1.64
0.91
1.44
2.00
1.53
0.28
1.20
1.68
>1.20b
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
<1.20C
X
X
X
Calculated based on average gas characteristics to acid plant
Desired minimum ratio for processing in a double contact sulfuric
acid plant
Considered too low for processing in a double contact sulfuric
acid plant
260
-------
Table 6-7. OXYGEN TO SULFUR DIOXIDE (02/S02) RATIO
FOR ADDITIONAL SCENARIOS AS PRESENTED IN TABLE 6-5 (CONCLUDED)
Blending Scenario
Number
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
02/S02 Ratio3
0.94
0.18
1.09
1.68
0.92
1.00
0.67
2.00
0.87
0.91
0.67
2.00
0.78
0.280
0.44
1.68
0.16
0.19
0.33
1.68
0.16
1.07
1.16
1.09
1.21
1.00
0.96
1.02
1.00
>1.20b
X
X
X
X
X
X
<1.20C
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Calculated based on average gas characteristics to acid plant
Desired minimum ratio for processing in a double contact sulfuric
acid plant
Considered too low for processing in a double contact sulfuric
acid plant
261
-------
considered to have ratios too low for proper conversion without
dilution air or oxygen. Consequently, they would not follow the
strategy of this blenging section to reduce the volumes of offgases
to the acid plants.
The remaining scenarios tend to reduce offgas volumes to the
acid plant while increasing SOo concentrations and still maintaining
adequate Oo/SCU ratios. The actual amounts depend upon the scenario
in question, however, those that reduce the volumes to about 110,000
or less may be attractive from an economic viewpoint. This occurs
because these volumes can be processed in one acid plant while
higher volumes usually require two acid plants (assumed as current
standard practice in the industry).
The gas volume flow rates and S02 concentrations to be pro-
cessed in FGD systems are all higher than those included in Table
6-4. The exact values depend upon the scenario in question but
generally are about triple those in Table 6-4.
Total S02 control efficiencies for these scenarios are somewhat
lower than those presented in Table 6-4. The major difference in
control efficiencies for scenarios in Table 6-4 and Table 6-6 is
that acid plant control is over 99 percent, while FGD systems are
assumed to control only about 90 percent. Thus, increasing volumes
to the FGD systems while decreasing the acid plant handling capacity
will markedly affect the total S02 control efficiency based on assumed
control efficiencies.
A comprehensive cost evaluation is warranted to fully assess the
economic viability of the above scenarios and those presented in
Section 6.1.2. Although it is expected that some of the scenarios
may be more cost-effective than the control approaches presently used,
additional cost data is necessary.
The final control system choice, whether it be blending, tradi-
tional acid plant control, or new smelting technology, will be based
on both economic and control capability considerations.
262
-------
SECTION 7
APPLICATION TO EXISTING SMELTERS
7.1 GENERAL
The approaches used to control weak S02 offgas streams from
new copper smelters can, in general, be applied to existing smelters.
However, retrofitting of any control approach will generally in-
troduce untque problems for each specific smelter.
As discussed in previous sections, the accepted control approach
in the copper smelter industry is to send S02 laden offgases to a sul-
furic acid plant. All smelters in the United States using reverbera-
tory furnaces produce an offgas stream with an SOp concentration too
low to serve as feed gas to a sulfuric acid plant. Thus, the major
type of smelting equipment that must be controlled is the reverbera-
tory furnace.
With the older smelters using multihearth roasters, an additional
low SOp concentration offgas stream is produced. The major reason for
this is that multihearth roasters and connecting duct work, currently
in operation, are quite old and have many leakage points. In some
cases, the duct work being used may be constructed of brick and has
been in place for over 50 years. With all these leakage points and
with the systems generally operating at negative pressure, the induced
or dilution air that enters the system reduces the S02 concentration
to well below the required acid plant input value. Furthermore, the
increase in gas volume as a result of the infiltrating air makes control
of the final offgas considerably more costly by increasing the size
of any control system.
Thus, in the case of existing smelters, thare are two major types
of equipment that must be controlled, the multihearth roasters and
263
-------
the reverberatory furnace. These are continuously operating devices
and can be adapted to add-on controls, to produce low volume high
SCL concentration streams. These streams can then be blended with
converter gases. Close attention to operating of the converters
including programming, offgas pickup hood design, and proper exhaust
system design and operating is currently required to maintain steady
feed conditions to the acid plant.
Another major problem that is encountered in some smelters when
retrofit control systems are considered is the available space to
place control equipment. Many smelters have equipment that is very
closely spaced and in order to find room for control equipment, such
as precipitators or baghouses, it Is necessary to place these units
some distance from the equipment to be controlled.
7.2 PROCESSING TECHNIQUE MODIFICATIONS
The major first step that must be taken by an existing smelter
to approach weak S02 stream control from either the reverberatory
furnace or the multihearth roasters is to seal up the system. This
includes not only sealing leaks in the reverberatory furnace and the
multihearth roasters but, even more important, leaks in the connect-
ing duct work that transfer the gases to the acid plant. It has
ro
been estimated that downstream leakage from reverberatory furnaces
may easily be as much as 100 percent or more. Thus, where current
typical reverberatory furnaces with leakage generate over 100,000
scfm, the reverberatory furnace system in Japan generates only 55,000
scfm after extensive sealing efforts.
The second most applicable technique, one that has been demon-
strated to produce a reverberatory furnace offgas of sufficient
strength for direct processing in a sulfuric acid plant, is oxygen
enrichment (Section 3). The development work at the Caletones
smelter in Chile (over 4 years) converted a typical green charge
reverberatory furnace operation to total pure oxygen/fuel burning.
This resulted in SO- offgas concentrations of well over 5 percent
indicating that reverberatory furnace S02 control with a sulfuric
264
-------
acid plant is feasible. The increase in production, at least for
that smelter, resulted in an economic advantage over conventional
operation. The copper smelting industry is familiar with the handling
and use of pure oxygen since this is an accepted technique for addition
to converter air. Oxygen/fuel burners are available and have actually
been used on an experimental basis in some reverberatory furnaces in
the United States (Section 3). Since the Caletones experience
was with a green charge furnace, some question remains as to the
potential for control of calcine charge furnaces. Some work is
currently being performed in Canada using oxygen/fuel burners with
calcine charge furnaces.
Relatively smaller yet significant gains can, in some cases,
be accomplished by other processing technique modifications such as
elimination of converter slag return, operation at lower air-to-fuel
ratio, instrumental control particularly pressure control, and con-
tinuous furnace charging. Blending of reduced volume flows that
result from system sealing also may be a viable technique in particu-
lar situations. Converter scheduling and improvement of converter
offgas collection systems, specifically including installation of
shutoff and modulating dampers, can increase SCL concentration from
offgases in those smelters that do not have a modern system design
allowing greater blending flexibility.
7.3 FLUE GAS DESULFURIZATION SYSTEMS
The use of the lime/limestone gypsum neutralization and the
magnesium oxide concentration system have been demonstrated in
Japan (Section 4) for control of reverberatory furnace offgas. The
gypsum producing system with lime or limestone has the advantage of
removing the SCL in solid form that can be "thrown away" without
introducing any additional pollution problems. This system may be
the most logical for existing smelters to control calcine charge
reverberatory furnaces because of the smaller amount of S(L generated
compared to green charge furnaces.
265
-------
The magnesium oxide concentration system generates an S(L stream
in the 10 to 13 percent concentration range. While this obviously
will allow direct processing in a sulfuric acid plant, it is still
relatively low. The citrate concentration system, on the other hand,
generates an SCL offgas stream over 90 percent concentration which
considerably reduces the size of the entire subsequent gas handling
system. The magnesium oxide system has had several years experience
in Japan on a full-scale basis operating with smelter reverberatory
furnace offgases. Conversely, the citrate system has only had pilot
plant experience with metallurgical gases at smelters in the United
States and in Sweden. If the choice were to be made between these
two systems, the citrate system would have the obvious advantage
except for the fact that actual full-scale experience is not yet
available. In addition, the use of this concentrated S02 stream
can readily allow production of liquid SO,, or perhaps even sulfur
with subsequent processes. The liquid S02 production would be
quite simplified in using this concentrated stream.
The use of systems to produce sulfur introduces the problem of
providing a reducing agent which is in all cases a fuel. The
experience with coal as a reducing agent (which is the most logical
in view of the current energy problem) has been demonstrated but
has not been shown to be a highly efficient process.
The Cominco ammonia scrubbing system has had extensive experience
under full-scale smelting conditions with low SO,, concentration. A
major problem of plume opacity is apparently one that appears to be
solvable. Cominco is currently constructing equipment to solve the
problem as a result of successful pilot plant testing.
Thus, several FGD systems are available with either actual full-
scale operating experience at smelters or with sufficiently promising
pilot testing to indicate a potential for application to specific
smelters. The local conditions at each smelter will, of course, con-
siderably influence the selection of any given system.
266
-------
7.4 BLENDING
Blending of gases to generate a rich enough stream to be pro-
cessed directly in a sulfuric acid plant is more difficult with
existing smelters. This is primarily because of the additional
weak streams particularly from the multihearth roasters. The local
conditions at the smelter and the potential for minimizing in-leakage
of air to the system will have a major influence on the capability
for blending of gas streams. It should be noted, however, that
there is a considerable range of options that must be considered
to determine the optimum blending design at any specific smelter,
as evidenced by the scenarios shown in Section 6.
7.5 ECONOMICS
Average values for the costs of controlling of weak SOp streams
from existing smelters are not sufficient to determine acceptability
of any given approach. It is necessary to review each individual
smelter on a case-by-case basis because of all of the unique prob-
lems that occur with each smelter. For example, the installation of
specific duct work, sealing problems, metallurgical conditions, and
the general feed and production situation of any particular smelter
will strongly influence the selection of a minimum cost approach.
However, it should be emphasized that to obtain a minimum cost approach,
all factors must be considered in relation to each other and to the
specific local conditions. The entire smelting process, from feed
constituents, either present or possible future, and their metallurgical
requirements through to the potential marketing or disposal of pollu-
tion control products, must be evaluated concurrently to determine
the optimum economic situation. The pollution control considerations
are so extensive and influence the operation of the smelter to such
a great extent, that they must be included as part of the overall
copper production system to obtain or define the most economical path
to follow for any specific smelter.
267
-------
SECTION 8
CONCLUSIONS
The major weak S02 gas stream from copper smelters is the offgas
generated by the reverberatory furnace. This gas concentration is
too low to process directly in a contact sulfuric acid plant, the
accepted sulfur S02 control system. Two major basic approaches
have been considered and are indicated for controlling the reverberatory
furnace. The first is to modify or improve furnace operating techni-
ques to provide more favorable offgas characteristics facilitating
control. The second is to introduce either neutralization or concen-
tration systems for the weak S02 gas stream.
Combinations of the various control approaches discussed in this
study can result in minimizing overall costs and increasing effective-
ness. The major cost-effective approach appears to be sealing of all
systems either with new or existing smelters to minimize offgas volumes
that must be processed.
Using oxygen enrichment for reverberatory furnaces with a green
charge appears to be a demonstrated technique for increasing the
weak S02 stream to a high enough value to process directly in a sulfuric
acid plant. Neutralizing the weaker S02 offgases with a lime or
limestone system for calcine charge reverberatory furnaces also appears
to be an economic approach, although only full scale green charge
systems have been completely demonstrated. The magnesium oxide S02
concentration system has been demonstrated on a full-scale smelter
and can produce gases with an S02 concentration of 10 percent allowing
direct processing in a sulfuric acid plant. The citrate concentration
system has been demonstrated on a pilot-scale with metallurgical gases
indicating it can concentrate weak S02 streams up to the 90 percent
level which provides greater economic advantages. The ammonia
268
-------
scrubbing system has been used with weak S02 streams from a full-scale
smelter. The plume opacity problem encountered when using this system
has been eliminated on a pilot scale and is currently being installed
at full scale.
The combination of a FGD or concentration system with a sulfuric
acid plant may introduce economies as far as capital costs are concerned
Retrofitting any of the approaches for existing smelters must be
determined on an individual local condition basis.
In general, from a technical standpoint, it can be stated that
with all of the considered approaches discussed in this study that
weak S02 offgas streams can be controlled from copper smelters.
The cost and system to retrofit these approaches must be determined
on a site specific basis. The cost to apply to new smelters may be
of the same magnitude as the increase in cost for changing to flash
smelting or some other new smelting process rather than using the
conventional roaster, reverberatory furnace, and converter technique.
269
-------
REFERENCES
1. "Background Information for New Source Performance Standards:
Primary Copper, Zinc, and Lead Smelters Proposed Standards,"
Vol. 1, EPA Publication No. 450/2-74-002a, October 1974
2. Reference 1
3. Hayward, C.R., "An Outline of Metallurgical Practice," D.
Van Nostrand Company, Inc., 1952
4. Reference 3
5. Ellis, O.W., "Copper and Copper Alloys," American Society for
Metals, 1947
6. Weisenberg, I.J., and Umlauf, G.E., "Evaluation of the Controlla-
bility of.S02 Emissions from Copper Smelters in the State of
Arizona," Pacific Environmental Services, EPA, Contract
68-02-1354, Task 8, 1975
7. ASARCO, Inc.
8. Visit to Onahama Copper Smelter, January 25, 1977
9. Anderson, R.J., "Operations at Utah Copper Division Smelter,"
in R.P. Ehrlich (Editor) Copper Metallurgy, Extractive
Metallurgy Division of the Metallurgical Society, Denver,
Colorado, February 15-19, 1970
10. Rinckhoff, J.B., "Sulfuric Acid Plants for Copper Converter
Gas," ACS Paper, April 4, 1974
11. Reference 1
12. Personal Communication with Tim J. Browder, Metallurgical
Acid Plant Consultant
13. Itakura, K., Nagano, T., and Sasakura, J., Converter Slag
Flotation—Its Effects on Copper Reverberatory Smelting
Process. J. Metals, v. 21, July 1969, 5 pp.
14. Jackman, R.B., and Hayward, C.R., "Forms of Copper Found in
Reverberatory Slags." Trans. AIME, v. 106, 1936, 11 pp.
270
-------
REFERENCES (Continued)
15. Itakura, K., Ikuda, H., Goto, M., "Double Expansion of Onahama
Smelter and Refinery," Paper No. A 74-11, The Metallurgical
Society of AIME, 1974
16. Niimura, M., Konada, T., Kojima, R., "Control of Emissions at
Onahama Copper Smelter," Joint meeting MMIJ - AIME 1972,
Tokyo, Japan
17. Weisenberg, I.J., Umlauf, 6.E., "Evaluation of the Controlla-
bility of S02 Emissions From Copper Smelters in the State of
Arizona." Pacific Environmental Services, Inc., EPA Contract
No. 68-02-1354, Task 8, May 1975
18. Pawson, H.E., "Giants Milling Operation," Paper presented at
Canadian Mineral Processors Meeting, January 23-26, 1973,
Ottawa, Ontario, Canada
19. Review comments by J. Henderson, ASARCO
20. Anderson, R.J., "Operations at Kennecott's Utah Copper Division
Smelter Reverberatory Department," Copper Metallurgy,
Extractive Metallurgy Division of the Metallurgical Society,
Denver, Colorado, February 15-19, 1970, pp. 146-147
21. Saddington, R., Curlook, W., and Queneau, P., "Tonnage Oxygen
for Nickel and Copper Smelting at Copper Cliff," Journal of
Metals 18 (4), pp. 440-452
22. Reference 15
23. Goto, M., "Green-Charge Reverberatory Furnace Practice at
Onahama Smelter," Extractive Metallurgy, International
Symposium on Copper Extraction and Refining, Las Vegas,
Nevada, February 22-26, 1976
24. Eastwood, W.B., Thixton, J.S. and Young, T.M., "Recent Develop-
ments in the Smelting Practice of Nchanga Consolidated Copper
Mines' Rokana Smelter," TMS Paper No. A71-75, Metallurgical
Society of AIME, 32 p, Pamphlet (1971)
25. Pluzhnikov, A.I., et al., "The Possibility of Using Roof Firing
of Reverberatory Furnaces," Tsvetnye Metally, Vol. 12, No. 10,
October, 1971, pp. 7-11
26. Kupryakov, Y.P., et al., "Operations of Reverberatory Furnaces
on Air-Oyxgen Blasts," Tsvetnye Metally, July 1972, pp. 13-16
27. Reference 26
271
-------
REFERENCES (Continued)
28. Wrampe, P., and Nollmann, E.G., "Oxygen Utilization in the
Copper Reverberatory Furnace: Theory and Practice," IMS
Paper No. A74-25, Metallurgical Society of AIME, 18 p,
pamphlet (1974)
29. Beals, G.C., Kocherhans, J., and Ogilvie, K.M., "Reverberatory
Matte - Smelting Process," United States Patent Office
3.222,162 Patented December 7, 1965
30. Zhuravlev, Y.A., et al., "Selection of an Efficient Thermal
Load in a Reverberatory Copper Smelting Furnace in the Case
of Oxygen Enrichment," Tsvetnye Metallurgiya, No. 6, 1975,
pp. 120-125
31. Reference 30
32. Reference 19
33. Chizhikov, D.D., "Present State of the Problem of the Use of
Oxygen-Enriched Air in Non-Ferrous Metallurgy," A.A. Baikov
Metallurgical Institute, Academy of Sciences of the U.S.S.R.
34. Reference 28
35. Reference 29
36. Goto, M., "Green-Charge Reverberatory Furnace Practice at
Onahama Smelter," Extractive Metallurgy, International
Symposium on Copper Extraction and Refining, Las Vegas,
Nevada, February 22-26, 1976
37. "Use of New Technologies at Caletones Smelter," H. Schwarze
D. G. Vera B., F. Pino 0., TMS Paper Selection No. A 77-90
Metallurgical Society of AIME, New York, New York, 1977
38. Reference 21
39. Reference 26
40. Reference 26
41. Reference 30
42. Reference 28
43. Itakura, L., Ikeda, H., and Goto, M., "Double Expansion of
Onahama Smelter and Refinery," TMS Paper No. A74-11,
Metallurgical Society of AIME, 1974, p. 29
272
-------
REFERENCES (Continued)
44. Reference 23
45. Reference 24
46. Reference 37
47. "Improvements in Full Use of Oxygen in Reverb Furnaces at
Caletones Smelter," J. Achurra H. R. Expinosa E., L. Torres J.,
IMS Paper Selection No. A77-91, Metallurgical Society of AIME,
New York, New York, 1977
48. Smirnov, V.I., "Possibilities of Technical Progress in
Reverberatory Smelting of Copper Concentrates," Tsvetnye
Metal1y. pp. 5-7
49. Reference 14
50. Otvagina, M.I., et al., "Sulfuric Acid Production from Rever-
beratory Furnace Gases," Tsvetnye Metal1y. Vol. 12, No. 7,
July 1971, pp. 5-7
51. Reference 28
52. Reference 30
53. Reference 26
54. Reference 50
55. Reference 50
56. Reference 29
57. Reference 47
58. Reference 24
59. Reference 21
60. Reference 26
61. Reference 21
62. Reference 28
63. Reference 29
273
-------
REFERENCES (Continued)
64. Reference 19
65. Reference 26
66. Reference 21
67. Reference 26
68. Reference 26
69. Reference 15
70. Reference 28
71. Reference 37
72. Reference 47
73. Reference 28
74. Reference 26
75. Reference 15
76. Reference 3
77. Reference 28
78. Reference 28
79. Reference 19
80. Reference 16
81. Reference 16
82. Nissen, W.I., et al., "Citrate Process for Flue Gas Desulfuri-
zation, a Status Report," Paper presented at the 6th Symposium
of Flue Gas Desulfurization in New Orleans, La., March 8-11,
1976
83. McKinney, W.A., et al., "Design and Testing of a Pilot Plant
for S02 Removal from Smelter Gas," Paper presented at the
Annual AIME Meeting, Dallas, Texas, February 23-28, 1974
84. Reference 83
274
-------
REFERENCE (Continued)
85. Zhuravlev, Y.A., "Zonal Three-Dimensional Model and Calculation
of Heat Exchange in a Reverberatory Copper Smelting Furnace,"
Tsvetnye Metal1y, No. 2, 1975, pp. 91-96
86. Reference 43
87. Reference 48
88. Personal Communication with Flakt
89. Reference 88
90. Reference 88
91. Reference 88
92. Reference 88
93. Personal Communication with Mr. Gunnnar Wai in of Flakt, Sweden
94. Reference 88
95. Reference 88
96. Reference 88
97. Reference 93
98. Reference 93
99. Ramsey, "Use of NH3 SOo-HoO System as a Cyclic Recovery Method,"
British Patent 1,427
>uo-ri9U
(T883)
100. King, R.A., "Economic Utilization of Sulfur Dioxide from
Metallurgical Gases," Industrial and Engineering Chemistry,
Vol. 42, No. 11, November 1950, pp. 2241-2248
101. Reference 83
102. Burgess, W.D., "S02 Recovery Process as Applied to Acid Plant
Tail Gas," Chemistry in Canada. June 1956, pp. 116-119
103. Report to the U.S. Bureau of Mines by the Smelter Control
Research Association, "Engineering Evaluation of Soluble
Scrubbing Systems," March 1974
275
-------
REFERENCE (Continued)
104. La Mantia, C.R., Lunt, R.R. and Shah, I.S., "Dual Alkali
Process for S02 Control," Paper presented at the 66th Annual
Meeting of the American Institute of Chemical Engineers at
Philadelphia, Pennsylvania, November 1973
105. Pedroso, R.I., "An Update of the Wellman-Lord Flue Gas
Desulfurization Process," Paper presented at Symposium of
Flue Gas Desulfurization, New Orleans, March 1976
106. Reference 100
107. Infra-red Remote Sensing and Determination of Pollutants in
Gas Plumes, H.W. Preagle, et al.,Environmental Science and
Technology, May 1973, p. 417
108. MHI Flue Gas Desulfurization Systems Applied to Several
Emission Sources, N. Hirai et al.
109. Personal Communication, Mr. Marvin Smith, Gypsum Association,
Los Angeles, California
110. Reference 5
111. Sulfur in 1975, Mineral Industry Surveys, U.S. Department of
the Interior, Bureau of Mines
112. Mineral Facts and Problems, 1975 Edition, U.S. Department of
the Interior, Bureau of Mines
113. Reference 112
114. Arthur D. Little, Inc, "Evaluation of S0?/as Controls
Economic Impact on ASARCO Smelting and Refining at Tacoma,"
Report to EPA, 1976
115. Reference 21
116. Reference 28
117. Personal Communication with Mr. A.J. Kroha, of ASARCo
118. Waitzman, D.A., et al., "Marketing H2S04 from S02 Abatement
Sources — The TVA Hypothesis," EPA Publication 650/2-73-051
119. Commodity Year Book, 1976
276
-------
REFERENCES (Continued)
120. McGlamery, G.G., et al., Detailed Cost Estimates for Advanced
Effluent Desulfurization Processes, EPA Publication 650/2-75-
006 (TVA Bulletin Y-90), January 1975
121. Reference 118
122. Chemical Marketing Reporter, October 25, 1976
123. Reference 114
124. Reference 114
125. Reference 114
126. Sulfur in 1975, Mineral Industry Surveys. U.S. Department of
the Interior, Bureau of Mines
127. Reference 112
128. Reference 122
129. "Standards of Performance for New Stationary Sources: Primary
Copper, Zinc, and Lead Smelters," Federal Register, Vol. 41
No. 10, January 15, 1976
130. Reference 1
131. Reference 6
132. "A Report on Removal of S02 From Copper Reverberatory Furnace
Gas with Ammonia Double-Alkali Process," SCRA, Inc.,
December 1977
133. Correspondence with Jan H. Reimers and Associates, Mettal-
lurgical Consulting Engineers
134. Reference 1
135. Reference 1
136. Reference 133
137. Reference 1
138. Carpenter, B.H., "Nonferrous Smelter Studies: Investigation
of the Role of Multihearth Roaster Operations in Copper
Smelter Gas Blending Schemes for Control of SOo; Part 1,"
RTI, EPA Contract No. 68-02-1325, Task 35, May 1976
277
-------
REFERENCES (Concluded)
139. Merkle and Associates, Inc. Engineers, Suspended Refractory
Designs
140. Reference 133
141. Personal Communication with Tim J. Browder, Acid Plant
Metallurgical Consultant
142. Reference 133
143. Reference 1
144. Weisenberg, I.J., et al., "Appendices: S02 Control for the
Primary Copper Smelter Reverberatory Furnace," Pacific Environ-
mental Services, EPA Contract No. 68-03-2398, August 1977
145. Weisenberg, I.J., et al., "S02 Control for the Primary Copper
Smelter Reverberatory Furnace," Pacific Environmental Services,
EPA Contract No. 68-03-2398, August 1977
146. Weisenberg, I.J., and P.S. Bakshi, "Process Parameters for
Primary Copper Smelters and Their Effects on Arsenic Emissions,"
EPA Contract No. 68-02-2606, PES Project No. 266, July 1978
278
-------
APPENDIX A
MULTIHEARTH AND FLUID-BED ROASTERS
A.I MULTIHEARTH ROASTER DESIGN AND OPERATING CHARACTERISTICS
A.1.1 INTRODUCTION
The blending of lean SOo gas streams from reverberatory fur-
naces with stronger gases from other smelting equipment to obtain an
SOp concentration capable of direct processing in a sulfuric acid
plant is one technique for increasing sulfur capture in the primary
copper smelting industry. Roasters, as indicated in Section 2
facilitate the smelting process by introducing the flexibility of
adjusting sulfur and impurity elimination. While fluid-bed roasters
are known to emit high SO- concentrations (8 to 10 percent), a
considerably lower value is emitted from some multihearth roasters
in current United States production because of the excessive air
dilution resulting* from the poor condition of many units. The pur-
pose of this section is to define realistic multihearth units.
Analysis will include theoretical considerations as well as review
of actual S02 concentrations from a number of multiherath roasters.
A.1.2 BACKGROUND INFORMATION
The conventional copper smelting process involves three indi-
vidual process steps -- roasting, smelting, and converting. Roast-
ing drives off a portion of sulfur from the charge producing cal-
cines that yield a desired copper matte (30- to 45-percent copper)
during subsequent smelting operations using the reverberatory
furnaces. The converting process upgrades the matte to blister
copper (98 percent).
279
-------
Typically copper ores are low grade and are not economically
suitable for direct smelting. Copper ore-bodies in the United
States have a low overall copper content (1 percent or less).
Additionally, most of the copper ores and concentrates are rela-
tively high in sulfur content and if fused directly to matte, will
form a low-grade product high in impurities. Thus, it is a common
practice to concentrate these low-grade ores to produce a higher
copper content (15- to 35-percent copper) material prior to
roasting.l
A copper content of 30- to 45-percent in the matte is commonly
used. A higher percent tends to result in a higher percent of
copper retained in the furnace slag while a lower percent requires
more oxidation of iron and sulfur in the matte during the subsequent
and more expensive converting stage.2 Thus, ore concentration and
sulfur removal in the roasting operation must be controlled to put
the reverberatory and converting operations in the best economical
balance.
In addition to the aforementioned considerations, multihearth
roasters have been indicated by some sources in the industry to be
effective in removing arsenic and other impurities; hence, it is
preferred for roasting concentrates that contain high concentrations
of such impurities.3,33 Thus, multihearth roasters can, under
certain conditions, be the more desirable approach for treating
copper ores in the United States.
A.1.3 MULTIHEARTH ROASTERS
Figure A-l4 shows a typical multihearth roaster which is
essentially a cylindrical, bricklined vessel divided from top to
bottom by horizontal brick hearth. Each hearth has one or several
drop holes in the brickwork, leading to the hearth below. The drop
holes are located alternately on the inner and outer peripheries of
successive hearths. A central, rotating brick surfaced steel column
extends vertically through the center of the roaster. On each
280
-------
r"A
BasfGartrmthCfiangr \ •
»_..&/!-.
-••:.-l-i~>-kfJK
jMerfiamt
FttdPbk
Figure A-l. Typical Herreshoff Multihearth Furnace
281
-------
hearth, arms equipped with rabbles are fixed to the rotating central
shaft. The charge is supplied to the top hearth near the central
shaft and rabbled to the periphery where it falls through the drop
holes to the hearth below. The rabbles on this hearth push the feed
toward the central shaft where it drops to the next hearth. This
zigzag pattern continues until calcine exits from the bottom hearth.
There are several makes of multihearth roasters which differ
slightly in design but all are based on the general principles of
the McDougal furnace. The main points of difference in
construction of these multihearth roasters are the construction of
central shaft, method of attaching rabble arms to the shaft, method
of attaching t'eeth to the rabble arms, method of cooling shafts and
rabbles, method of introducing air to the hearths , and number of
hearths. Table A-l lists some of the design parameters of the
common roasters used in the copper smelting industry. At present
practically all roasters have at least six or seven hearths with a
tendency toward increasing the number to eight, nine or more.
Diameters vary from 19 to 25 feet and the heights vary from 18 to
over 47 feet. The central shaft not only supports the rabble arms,
but contains pipes which conduct the cooling air or water to the
rabble arms. Central shafts range in diameters up to 4 feet which
enables a man to enter it for repairs. The speed of this shaft,
which is driven by a motor and train of gears beneath the furnace,
is on the order of one to two RPM. The methods of attaching rabble
arms to the shaft and teeth to the rabble arms vary, but they all
have the objective of easy replacement with minimum loss of time.
The cooling of the shafts and rabbles are done either by water or
air, with air cooling being the most common practice. The methods
of air introduction vary but the main objective is to promote accur-
ate temperature and atmosphere control to ensure proper roasting.
During roasting of copper ores it is not desirable to
completely remove the sulfur in the feed because control of the
matte is primarily accomplished by providing specific copper to
282
-------
Table A-l. SUMMARY OF S00 EMISSIONS FROM MULTIHEARTH ROASTERS
Ref-
er-
ence
1
2
3
4
5
i
7
Coa»any
Phelps
Dodge.
Douglas.
Arizona
ASARCO,
Tacona,
Hash-
Ington
ASARCO,
Hayden,
Arizona
ASARCO,
El Paso,
Texas
Horanda,
Duebec,
Canada
Kashoe.
Anaconda.
Montana
801-5.
Yugo-
slavia
Type of
Roaster
(Hwfcer)
Herreshoff
(24)
Herreshoff
(10)
HcDougal
(5)
HcDougal
( 3)
Herreshoff
( 3)
Bartlett-
SMM
Pacific
t 11
I >)
Herreshoff
I 4)
Hedge (8)
HcDougal
(56)
K.«,ln
(5)
Mutter
of
Hearths
7
H/A'
e
7
7
H/A
7
7"
6
11
Dimensions
of
Roasters
Helgnt
(ft)
18.50*
25
23.75
22.75
8.50
34.75
H/A
H/A
H/A1
47.6
Dim-
eter
(ft)
21.58
19
24
19
21.58
25
22.5
25
H/A*
21.3
HP
H/A
H/A
IS
10
10
40
H/A
15
H/A
«/A
RPM
H/A
H/A
2
2
2
2
H/A
1.09
H/A
1 5*
Feed Rate
Tons/
Day
175"
250b
225'
250'
292b
40"
zoo"
Sulfur
Percent
29 "
29. £
28 "
26 '
24.3b
34 »
40 c
Exhaust Gas
SCFH
,6.11!'
20.000C
I5.000C
15.000°
15.000C
10, mf
!0,014*
«>,420C
6,667C
7,560*'
Percent
S02
(by
Volume)
1.7"
1.5"
1.8 "
.79
i
2.25
4«d
Hornal
Roasters
In
Service
18
5
•'
H/A
H/A
H/A
H/A
Ton/
SCFH
(Sulfur
In
feed)
0.003
.004
.004
.003
.003
.002-
.011
Footnotes
* Value taken for standard
Herreshoff roaster
Average per roaster
c Average per roaster based on
18 roasters In service
d Low - probably due to consid-
erable air In leakage, (toast-
ers arranged In double row.
4 Typically 7 hearths arranged
In double row
Average capacity per roaster
c Average per roaster for range
of 30.000 - 70.000 SCFH total
d Average. Subject to air
In leakage.
Average per roaster
Average per roaster based on
28 percent sulfur
c Average per roaster
d Average based on 1.1 - 2.5
percent sulfur
* Average based on 6 - 10. All
but one roaster subject to air
leakage. Arranged In double
row.
Average per roaster based on
maximum roast feed
Arranged in double rows.
* 7 internal and 1 external
drying
Average per roastar from Janu-
ary 1 - June 30. 1933
c Value taken fro* approximate
roaster treater handling
450.000 ft'/min, therefore
average per roaster based on
SCFH at 1 .000DF
* Typically 24' diameter x
23.75'
Average per roaster
c Estimation (ft3/m1n) per
roaster at 48
d Based on (a)
14 rows. e«ch with 4 furoices
* Average (1 - 2 RPM)
b Design for each roaster
c Average from test 3.5. 6 per
roaster
d Average from test 2.3.4,5,
per roaster
Tests 1 and 6 were ruled out
due to fluctuations in roaster
performance and feed variances.
Note: These data cover nearly an 80 year time period and should
be considered as indicative of the technology and not
necessarily current operations.
283
-------
sulfur ratios in the calcine feed to the reverberatory furnaces.
Partial roasting is characterized by oxidation of a portion of
the sulfur content of the feed and a conversion of part of the iron
sulfides to iron and sulfur oxides. Air is usually introduced at
the bottom of the roaster and passes up through heated chambers
where oxygen in the airstream reacts with iron and sulfur in the
feed to liberate heat which sustains the roasting hearth tempera-
tures. Under such conditions, the tendency is for iron sulfide
to be oxidized in preference to copper sulfides and since only
a partial roast is obtained, it follows that most of the copper
will remain as copper sulfides or copper sulfates.
There is a limit to the amount of sulfide exposed per unit
time to the action of oxygen and, therefore, to the amount of
heat liberated per unit of time. If a large excess of air is
used, it may absorb so much heat that there is insufficient heat
available to keep the roasting ore above its ignition point
(about 300°C). Thus, the oxidation of sulfide minerals will
cease as soon as this temperature is no longer maintained.
Similarly, if there is too good a contact with oxygen, there is
the possibility of the temperature rising to the melting point
of the ore which excessively reduces the sulfur in the calcine
product.
Additionally, temperatures affect the various chemical reac-
tions occurring in the roaster and ultimately the SOo concentra-
tion in the exhaust gases. The heat of formation of sulfates is
higher than that of corresponding oxides so if temperatures are
low enough, sulfates will be formed with an obvious decrease in
the elimination of sulfur. The equation CuO + SOQ—^CuSCL is
7
driven to the left at higher temperatures. Thus, toward the
end of the roasting step, it is necessary to maintain higher
temperatures than at the beginning in order to decompose the sul-
fates which have a tendency to form. Figure A-2 shows the pro-
gressive removal of sulfur and the flame temperature profile for
a McDougal roaster with six hearths.
284
-------
0% 20
1st Hearth
2nd Hearth
3rd Hearth
4th Hearth
5th Hearth
1
6th Hearth
40 60
80 100%
'
-------
it is brighter near the outer periphery. Here the maximum
temperature of 960°C is attained. On the sixth, the final, hearth
heat has become uniform but lower (860°C). As the ore leaves the
hearth it seems brighter, but speedily cools off to 660°C as it
falls, smoking freely, into the hopper."
Current operating temperatures tend to be lower than discussed
above, generally in the 550 to 700°C range during the first stages,
increasing to a maximum of 800 to 850°C1Q depending upon the low
melting point constituents of the charge. The strategy is to keep
the charge below the sintering temperature on the upper hearths and
maintain as high a heat as possible on the lower hearths. Tempera-
tures are influenced and controlled by furnace design, air regulation,
rate of feed and amount of sulfur in the ore.
Particle size and percent sulfur in the feed are known to
influence roasting. Large particles roast slowly and sometimes
incompletely due to limited surface area and diffusional limita-
tions. On the other hand, fines can cause dust entrainment and
roast so fast that sintering may occur. When the initial sulfur
content is below 24 percent, it is necessary to finish roasting
Q
with extraneous heat. The resulting exhaust gases, diluted
with products of combustion, are leaner in S02 concentrations.
The depth of roaster bed, number and design of drop holes and
number of hearths have a definite effect on the roasting process.
It is assumed that about 60 percent of the sulfur removal takes
place on the hearth bed while about 40 percent takes place as the
ore drops through drop holes. Thus, if the roast bed is too
thick there will be diffusional limitations with a decrease in
roasting resulting. Townsend et al. investigated the roasting
of lumpy chalcopyrite in an air stream at temperatures between 550°C
and 750°C and followed the transformations with a microscope and
microprobe. The roasted particles had an extremely porous zone of
hematite covering a compact layer of magnetite. The thickness of
the layers was independent of the roasting temperature. Traces of
286
-------
copper ferrite (CuFe204) were found in the hematite and were
presumed to have been formed as result of establishment of a local
equilibrium. Details of the transport of oxygen and sulfur through
the impervious magnetite layer could not be explained. Winterharger
et alJ^ found that during the roasting of copper containing
pyrrhotite, various copper-iron sulfides occurred as sequential
phases in the receding sulfide kernal, and copper sulfides were con-
verted to cuprite (Cu20). As roasting continued (in air or in
S02-rich gas), there was a zonal progression of these sequential
phases, as would occur if the reaction rate-determining step were
the gas-phase diffusion processes occurring in the pores of the
roasting product. This rate-determining step has also been shown by
other investigators to approximate closely the burning of iron sul-
13
fides in air. The rate of oxidation appears to be determined
by the rate of diffusion of gaseous reactants through the solid
crust which covers the core of the sulfide particles. Chemical re-
action occurs at the boundary of solid metal sulfides and metal ox-
ides. Oxygen diffuses in; S02 diffuses out. Unconsumed oxygen in
the gases provides the driving force. As oxidation proceeds, dif-
fusion becomes more difficult and the reactions slow down. Addi-
tionally, at oxygen concentrations lower than about 13 percent in
the gases next to the solid particles, the fire goes out. Data in-
dicates that active oxidation of the charge requires about 13 per-
14
cent oxygen in the gases and temperatures of 640° to 760°C.
Drop hole area will affect the velocity of the gases in contact
with falling ores. If the holes have a large area then the velocity
of the gas in contact with the ore will diminish—facilitating a
better roast. Since some roasting takes place during the drop from
hearth to hearth it is obvious that a larger number of hearths will
facilitate a more complete roast.
Some furnaces are designed with special air ports on each hearth
while all furnaces have doors on each hearth which can be used in
air regulation. In those furnaces with air-cooled rabble arms, air
can be admitted to the hearths through holes in the arms. Another
287
-------
method commonly used with air-cooled rabble arms, is to use this air
as preheated air for the roasting process. For these roasters the
the central shaft is constructed in sections consisting of an inner,
cylindrical part (cold air tube) and an outer annular part (hot air
compartment). Cold air is forced in through the cold air tube and
passes from here into hollow rabble arms, thus serving to keep
them cool. The heated air coming from the rabble arms enters the
hot air compartment, and from here it may be discharged to waste
at either the top or bottom of the shaft, or admitted to hearths
as preheated combustion air. It is the usual practice to admit
most of the air on the bottom hearth which brings an excess of
oxygen upon the ore as it is ready to leave the furnace. By the
time the air reaches the upper hearth much of the oxygen has been
replaced by S02 causing less rapid roasting of the fresh ore and
consequently minimizing the danger of sintering. The hearth doors
may be partly opened to supply additional oxygen to prevent sul-
fate formation or opened wide to allow a large excess of air to
rush in and cool an overheated charge.
Temperatures are ideally set from 600° to 700°C during the
first stages of roasting and increase to 800° to 850°C by the final
stages. The strategy is to keep the charge below the sinter-
ing temperature on the upper hearths and maintain as high a heat
as possible on the lower hearths. The need to remove arsenic
from some concentrates calls for consideration of the temperature
to which the solids should be heated during the roast. Among the
arsenic sulfides that may be present in copper ores are FeSAs2$3,
ASpSo, and As^S,-. The As^S., boils at 707°C, while ASpSc sublimes
at 500°C with decomposition. Arsenic trioxide melts at 310°C and
boils at 475°C. Thus, a portion of the arsenic will be vola-
tilized at operational temperatures within the roaster. Also,
the amount of oxygen present affects arsenic removal. As20~
is preferred since it is quite volatile and will pass off with
the roaster gases; however, in an oxidizing atmosphere, much of
the arsenic will oxidize to As205 which is less volatile and
288
-------
form stable, nonvolatile arsenates with other metallic oxides.
Usually, it is necessary to alternate oxidation and reduction
several times to completely remove arsenic. A similar phenomenon
applies to antimony which goes from SbpO, to SbpOj- in an oxidizing
atmosphere. Thus, roasting can be ideally controlled for some
impurity removal through proper temperature and air regulation.
Finally, typical roasters handle from 125 to 150 ton/day using
high sulfur charge. If copper is high in the charge the capacity
may be increased since less roasting is required (i.e., there is
more sulfur left in the product). Roasters can handle up to
350 ton/day.17'18
A.1.4 S02 EMISSIONS FROM MULTIHEARTH ROASTERS
Table A-l lists data obtained from several multihearth roasters
IQ ?n ?i ?? ?? ?A ?£>
in the United States and one from Yugoslavia. '*"''"'"'*' °
These data indicate that the SOp concentrations (by volume) vary
from less than 1 to 5 percent. It should be noted that roasters
1 to 4 in Table A-l were reported to have considerable air
leakages which would result in low S0« offgas concentrations.
Additionally, roasting eliminates a portion of sulfur in the
feed which is based on the required Cu/S ratio for the subsequent
smelting operations. Thus, if a feed contains a small percent cop-
per and another feed contains a larger percent copper, then in order
to obtain the same Cu/S ratio for both feeds, a larger percent of
sulfur must be eliminated in the former. Roasters 2 and 5 in
Table A-l had a feed copper content of 22.8 and 5 percent, respec-
tively. Thus, to obtain similar Cu/S ratios, it is apparent that
roaster 5 had to eliminate more sulfur indicating a more complete
roast and a subsequent higher S02 concentration. Nevertheless,
average SO? concentrations of 4 percent by volume have been reported
7ft ?7
for multihearth roasters in the United States.
An accurate calculation of the oxygen consumed for any specific
roasting case requires knowledge of the mineralogical character
289
-------
and degree of transformation of the particular charge. However, a
general formula for estimation of oxygen consumption using the
predominant chemical, metallurgical, and thermodynamic aspects of
28
the process has been proposed.
Table A-2 lists the mineral conposition of the roaster charge
for the Bor and Majdanpek ore concentrates and the quantities of
the minerals in the calcine produce therefrom. The aforementioned
reference (28) calculated heat balances based on reactions
considered to be most significant for the Bor roaster. The accuracy
of the calculated results depends upon accuracy of the required
input data; the weight fraction of components in the calcine; the
fraction of dissociable sulfur and arsenic (or other minor elements)
actually removed; and the fraction of copper oxidized. The results
are shown in Tables A-3 and A-4. These calculations were based
upon data for the Bor, Yugoslavia roasters. It was also assumed
that the ratio:
FeS2 to
to
is constant over a broad range of sulfur removed in the offgas,
that the fraction of copper oxidized was about 0.8 the ratio of
sulfur removed by dissociation/dissociable sulfur in charge,
and the fuel oil was assumed to contain 1 percent sulfur and
provide about 150,000 Btu/gallon when burned with 15 percent
excess air. Additionally, these values for S02 concentrations may
be somewhat lower for charges containing significant quantities
of Sb, Pb, Bi, and Zn since, theoretically, these compounds can
affect the heat balance and subsequent gas S02 concentrations
primarily because their elimination requires oxygen. The conclu-
sion reached in the study was that the S02 concentration in exit
gases from multihearth roasters can be held above 5 percent pro-
vided at least 6 percent of the sulfur is removed from the charge.
Also maintaining this concentration level requires monitoring the
290
-------
Table A-2. MINERAL COMPOSITION OF ROASTER
CHARGE AND CALCINE, CLEAN CONCENTRATE28
Mineral Components
Clean Charge
FeS2
CuFeS2
ZnS
Inerts
FeS
FeO
Fe3°4
CuFeS2
CuOFe203
ZnO
Total
Kkg per 100 kkg of Charge
In Charge
25.18
49.01
1.52
24.29
100
In Calcine
1.05
24.29
6.78
0.21
2.31
32.28
21.83
1.27
90.02
291
-------
>- Z
C£ LLJ I—
O I— Z
Z UJ
C3 O O
^ o a;
at
3
« 3 «
I
10 »•
01 0>
SS
UJ
oo »
p— i oo
00 . UJ
-<
O I-H _l
o o
Z _J O
« o
UJ O
_
"X Q-
00 UJ
•z. z
o •• o
i— i (_j i— <
I— o I—
< o <:
_l O i— i
ro od a
0 Al <
_i a;
<; oo
_i z
<: I— L±J
s: 1-1 o
a: x a:
<_) I—
o 3
• LT> CT>
CO PvJ
ca: ., ..
UJ UJ
«
« •« *
115 3
-
f 1
292
-------
C£ OO
LU LU
• •i—oo
oo <: oo
i—i 3 O
00 _l
<
CD . CJ3
O Z
LU O i—I
CJ O —I
cz «a- o
— < A I L"-1 tr>
(— O r-H
< oo on
—I LU
•— t—
J3 >- Z
03 CC O
h- Q O
i!ii!iIMtli!§u
!i*^??S5*IIS«*ll.
293
-------
exit gas 0? concentration and controlling it at 12 percent along
with controlling heat losses to a maximum of 15 percent and calcine
exit temperatures to a maximum of 1,364°F.
A.1.5 CONCLUSIONS
Based on actual SOp offgas emissions (see Table A-l), multi-
hearth roasters are capable of releasing S02 concentrations up to
4 percent volume. Likewise, the above study indicated that at
least 5 percent S02 (up to 7 percent) offgas concentration can be
maintained provided:
1. At least 60 kg of sulfur is removed per kkg charge
2. Exit gas Q£ concentration does not exceed 12 percent
(assuming oxygen is supplied by air addition with
consequent increase in the accompanying nitrogen dilution
effect)
3. Controlling heat losses to a maximum of 15 percent
4. Controlling the calcine exit temperature to a maximum of
740°C
The S0£ concentration will vary depending on the amount of
sulfur contained in the ore and the actual operating parameters.
Nevertheless, a 5 percent SO? offgas concentration seems feasible
in discussing the capabilities of the multihearth roaster.
Additional means may be employed to upgrade the S0£ concentra-
tion from multihearth roasters. These points may be summarized as
follows:2^
1. Operate with the lowest oxygen concentration consistent
with smelter calcine requirements
2. Drying in a separate operation
3. Better insulation and recovery of heat from rabble-arm
cooling air
4. Preheat incoming air
5. Keep exit calcine temperatures lower
6. Modification: reduce air infiltration to a controllable
level; effect control through the use of oxygen monitors in
the exit gas duct.
294
-------
A.2 FLUID-BED ROASTER DESIGN AND OPERATING CHARACTERISTICS
Fluid-bed roasting involves the autogenous oxidizing of sulfide
particles while they are suspended in an evenly distributed stream
of air. It is based upon the principle that the blowing of air
through a bed of fine solids tends to support the particles at mod-
erate velocities. These particles may be permanently suspended in
an expanded or fluidized bed. The particles are essentially
surrounded by air so that rates of gas/solid roasting reactions
are high. The reactions occurring are similar to those occurring
in multihearth roasters (refer to Appendix A.I).
Figure A-3 shows a cutaway of a typical fluid-bed roaster.
Air is blown into the roaster by means of a tuyere plate at the
bottom, and concentrates are added in particulate or slurry form
near the top of the roaster. The roasting operation is begun by
heating the roaster (usually containing an inert bed of sand or
calcines) to the temperature at which the concentrate will ignite
by air. The temperatures are maintained between 930° and 1,300°F.
The concentrates are then added, slowly at first to begin the
roasting and to make the operation autogenous.
Reaction rates within the roaster are rapid and an important
consequence is the high efficiency of oxygen utilization by the
roasting reactions. This leads to air requirements only slightly
in excess of the stoichiometric amount. SCL concentrations in the
effluent roaster gases are considerably higher than those of the
multihearth roaster, it is 8 percent compared to 4 percent (new units
Figure A-3 shows a cutaway of a typical fluid-bed roaster. The
fluidized state is accomplished by considerable agitation of the
particles in the bed which results in efficient heat transfer and a
uniform temperature across the roaster. This, in turn, permits
accurate control of the roasting temperature.
One problem, however, caused by the high chemical efficiency
of the fluid-bed is that the roaster tends to overheat due to the
295
-------
OFF-OS
SLURRY
FEED
TUYERE
HEADS
PRODUCT
Figure A-3. Typical Fluid-Bed Roasted (Text)
exothermic oxidation reactions. This could result in sintering,
overoxidation, or agglomeration which could collapse the bed. This
problem is generally rectified through the addition of water or
inert fluxes (for use in subsequent smelting) with the concentrates.
During copper roasting, a major portion of the fluidized solids
are carried out through the top of the roaster. These are collected
in cyclones above the roaster while the remainder of the solids over-
flow the fluid-bed portion of the roaster. Thus, the bottom portion
of the roaster contains a stable fluid-bed in which the larger par-
296
-------
tides are oxidized. In design, the larger particles which require
lengthy oxidation times have a long residence time in the stable bed
portion, while smaller particles are blown out before they have time
to overoxidize. Thus, the fluid-bed roaster offers a means of pro-
ducing an even roast.
A critical factor in fluid-bed roaster design is that the lar-
gest particles in the concentrate must become.fluidized, thus pre-
venting tuyere clogging and eventual collapsing of the bed. Hence,
the velocity of the gas must be considered of prime importance in
actual operational practices.
The residence times of particles control, in part, the extent
of the oxidation reactions taking place. The residence times are
controlled by varying the depth of the stable bed, the rate of con-
centrate feed, and air flow rate.
Advantages of fluidized roaster over conventional multihearth
roasters include:
1. Roaster offgases are ideal for making sulfuric acid
since the roaster operates under continuous steady
state conditions with a relatively high S02 concen-
tration in the offgas.
2. Offgas volumes are reduced significantly due to reduced
air requirements.
3. Reduced residence time and consequently fewer pieces
of equipment needed for the same production.
While it appears that the current philosophy in the copper in-
dustry is to use fluid-bed roasters, some problems may exist that
would inhibit their use. Fine grind is needed and with underground
ores it might be necessary to reduce the size so much that the grind-
ing costs could be excessive. Also, there may be an increase in
magnetite formation depending upon the composition of the charge.
Finally, there is some concern over problems with impurity removal
due to the decreased residence time within the roaster.
Fluid-bed roasters are reported to be undesirable for process-
ing concentrate ores containing high impurity levels. These impur-
297
-------
ities are relatively difficult to separate by volatilization, and a
fluid-bed roaster may not provide sufficient residence time for com-
plete separation. However, this is entirely dependent upon the
impurities and their concentrations in the feed. Multihearth roasters,
on the other hand, provide a much longer residence time as the ore
travels through each hearth.
In a recent study, it was found that the amount of arsenic
volatilized in the roaster and furnace combined appear to be the
same for both the fluid-bed roaster-reverberatory furnace combina-
tion and the multihearth roaster-reverberatory furnace combination,
although the fluid-bed roaster volatilizes less arsenic than a multi-
hearth roaster for a given set of conditions.
Although the multihearth roaster is capable of removing more
arsenic than a fluid-bed roaster, a fluid-bed roaster is being used
at the Anaconda smelter where the feed input is high in arsenic con-
tent. This is contrary to the belief that fluid-bed roasters should
not be used when the arsenic content in smelter feed is high. How-
ever, the Anaconda fluid-bed roaster, in conjunction with an electric
furnace, produces a matte grade above 50 percent. The choice of
the type of roaster used with a "high" arsenic content feed does
not seem to be governed by the high amount of arsenic in the feed,
but by the quality of matte desired from the smelting furnace.
Since multihearth roasters seem to yield calcine which produces a
lower grade matte than fluid-bed roasters for any given amount of
arsenic elimination, they tend to be preferred whenever the above
two qualities are desired simultaneously. This is particularly
true when lead and antimony are present. Complete proof of the
above observation has not been established.
Nevertheless, fluid-bed roasters warrant consideration for
weak stream S02 control since the blending of fluid-bed roaster,
reverberatory furnace, and converter offgas has been reported
feasible and produces a considerably higher combined S0? concen-
32
tration.
298
-------
APPENDIX B
REVERBERATORY FURNACE HISTORY, DESIGN AND OPERATION
B.I FURNACE FUNCTION
The primary function of the reverberatory furnace is to
economically smelt the required copper-bearing charge into a
molten mass. The objective of the furnace operation is to pro-
duce separate fluid layers of lower specific gravity discardable
slag (iron-gangue comprising the top of the bath) and a higher
specific gravity matte (consisting theoretically of a mixture of
Cu2S'FeS comprising the lower fluid layer). The matte can be
separated by selecting tapping levels. It is then further treated
in other furnaces for separation of Cu from the Fe and S. The slag
is tapped from the furnace at a higher level and usually discarded.
B.2 BRIEF HISTORY
Previous to the development of the first reverberatory in
America at Colorado Smelting and Mining Company in Butte, Montana
33
in the year 1879, most smelting was done in blast furnaces.
The ore content of oxides and pyrites, available at that time,
were more readily processed in blast furnaces. With sulfide ores
becoming more prevalent, the reverberatory furnace became the
more efficient processing device. The first reverberatory fur-
naces were of the batch type and the charge was smelted by grate-
burning of wood or charcoal. From the first small 10 ton per day
batch furnace, size gradually increased up to lengths of 50
meters (160 feet) and widths of 11 meters (36 feet). As heat
loss increased with the area of refractory arch and side-wall
surfaces, the largest furnaces proved less economical, and
present-day dimensions are slightly over 30 meters (100 feet) in
299
-------
length and 8.5 to 10.4 meters (28 to 34 feet) in width in most
instances.
B.3 MODERN CONSTRUCTION
Reverberator!es are large horizontal chambers constructed of
refractory material throughout, capable of being fired through
one end and containing an outlet flue at the opposite end to
allow exit gaseous products of smelting and combustion.
There are two types of roofs or arches in current use as
illustrated in Figure B-l. The original arch-type is constructed
of silica brick and is sprung between skews and buckstays at the
sides so that it is suspended in its entirety, independent of
the walls or ends. The original arch thickness is 0.51 meters
(20 inches) and is maintained by slurring on the underside
through a nozzle and a pipe connected to a Quigley-type gun.
Ribs above the original arch are sometimes installed so that an
upper or relieving arch may be installed above the original, when
it becomes thin. The maximum allowable width of a sprung arch
is about 8.5 meters (28 feet) due to inherent brick strengths.
The second and newer type of arch is made of brick indivi-
dually suspended by means of hangers attached to the brick and to
the superstructure. Refractories may be of either silica or basic
35
brick. This type of arch is generally panelized into relatively
small removable sections (typically 3 by 10 feet) bound together
with new panels which can be done in as short a time as a half
hour. Individual brick or pairs may also be replaced without
being part of the panel. The principle advantage of suspended
arches is that any width of arch may be used and cross-sections
can be installed at different elevations to allow any desirable
internal contouring. There generally is a much greater heat loss
through an arc constructed of basic material, but bricks wear
longer under furnace conditions. The modern trend is towards the
use of basic suspended arches.
300
-------
OJ
CO
I
S-
Ol
D.
Q.
O
O
ai
s_ jc
O O
4- i-
•i- 0)
(/> T3
O) C
O (LI
Q.
O) CO
O 3
fO OO
S- 08
3
Li_ O)
O Q.
•M 00
1C
S- cn
CU £
O
0) oo
CD
i— C
O •!->
.
>> E
I— oo
CQ
-------
The sprung silica arch would remain on a furnace practically
forever, with the excellent continuous hot strength of silica,
if it were not affected by splashed matte and slag and the
corrosive fumes of the furnace atmosphere. Unfortunately, the
chemically acid silica brick can be quite readily corroded by
the primary iron-silicate and iron-lime-silicate constituents
of the furnace fumes and slags. The continual use of silica
spraying furnaces has prolonged their life to as much as 12 years.
The magnesite-chrome system of refractories (basic brick)
offers significant improvements in roof life with the typical
iron and iron silicate slags of the reverberatory furnace but is
more expensive. However, the hot strength of basic brick is
much less than silica. This coupled with the permanent expansion
characteristic of basic brick, makes it unsuitable for tradition-
al sprung arch construction; therefore, a suspended system must be
used on most reverbs.
The grade of basic brick most commonly used is the chemically
bonded type with an MgO content of 40 to 60 percent. The shapes
are normally steel encased to reduce spalling. The steel is also
thought to form complex compounds with refractory-iron ingredients
that are beneficial to brick life.
Direct bonded and rebonded fused grain basic refractories
are premium type products which can give increased performance
in the high wear area of the reverb roof and have come into general
use in certain areas -- usually in the smelting zone from 10 to
40 feet beyond the burners. However, their increased cost should
be economically justified by careful analysis of their performance.
Their are various types of brick suspension as shown in
Figures B-2a through B-2d. The first and most common involves
suspension from rods hooked into tops of bricks, usually using
one rod for two brick. A more recent innovation involves
suspension of brick by interlocking with a refractory to which
the suspension rod is attached.
302
-------
Figure B-2a
Typical Tile Support System
Figure B-2b
Panel Construction
Figure B-2c
Typical Roof
Cross
Figure B-2d
Sections
Figure B-2. Suspended Arch Brick Assembly and Mounting Design
303
-------
Sectionalizing of brickwork into panels is now commonly
practiced in the smelting zone. It is claimed that an average
time required to replace a 3 by 10 foot panel is about a half hour
as compared to several hours to replace a similar area with
individual or paired brick. Replacement panels are prefabricated
and bound beforehand and effect a saving in hangers required.
Furnace sidewalls are constructed of refractory material
in brick form that may be of silica, clay or basic material
although the basic type is generally used on the inner face of
the furnace and below the bathline. Walls may be corbelled
(bricks set so as to form short supporting brackets) because
of the temperatures and liquid attack -beneath the bath line
and under the burners. Hater or air cooled jackets of copper,
steel, or cast iroja are generally used around tap holes, skim
bays, bridge walls and, in the deep bath furnace, generally
all along the entire bath line.
External insulation of the brickwork has been tried but is
not extensively used as fusing and melting of the internal face
of the brickwork is greatly increased with its use.
Most furnace bottoms rest upon a tamped sand, clay, or
concrete base of several feet thickness. On this base will be
layers of magnetite-slag, chrome ore, quartz, converter slag,
used in combination or singly, and all fused into place. Even-
tually bottom thicknesses are several feet deep, the final thick-
ness being determined by type of furnace, desired bath depth and
temperature and peculiarities of the charge to be smelted. Bot-
tom cooling may be accomplished by means of air piped under or
through the bottoms or just above the foundations. Typical fur-
nace across sections are shown in Figure B-3.
Most of the actual smelting is done in the first 70% of the
furnace length, allowing almost complete separation of the matte
and slag layers in the balance of the furnace length. The most
intense smelting zone is usually about 15 to 20 feet from the
qc
burners and temperatures will range from 1500°C (2800°F)
304
-------
Holf Section~A-A
Half Section "B-B"
Figure B-3. Transverse Half Sections of Furnace9
305
-------
to 1700°C (3100°F) in this area. Bath temperatures in the furnace
will range from 1300°C (2300°F) down to 1100°C (2100°F), usually
near the skim bay. As the slag forms in the top layer of the
bath, it receives more heat reflected from the flame and the
arch and is the hottest portion of the bath. It also serves,
undesirably, as a protective barrier for heat transfer to the matte
layer beneath. The height of the furnace arch above the bath-
line is determined by the heat and combustion requirements of the
smelting, plus heat losses through the refractories and in slag
and matte removal, so as to allow approximately 1,000 Btu per
square foot of cross-sectional area.
t
Either sprung or suspended arches may be used in furnaces
of either side-charged or bath smelting types. Installation plus
maintenance costs are comparable for either sprung or suspended
arch types and, today installation would be between $4 and 5
million for a furnace complete with auxiliaries. The
advantages of the reverberatory furnace lie in its flexibility
and in its ability to handle up to 1600 tons of hot charge or
1200 tons of green charge per day. Use of oxygen or preheated
combustion air would raise these capacities considerably.
B.4 FUELS
Originally, wood or charcoal was the only fuel used, but
in the early 1900's coal came into prominent use, being first
burned on grates and later pulverized and blown into the fur-
nace with pressurized air. This was gradually replaced by oil
fuels in the 1920's,and in the 1930's the use of natural gas
firing became most common. With the shortages of natural gas
in the early 1970's, all plants are reverting back to some oil
firing and many are planning on eventually returning to the use
of pulverized coal. Gaseous fuels give a clear nearly invisible
flame, oil fuels produce a semi-luminous flame and pulverized
coals produce a luminous flame.
306
-------
With the rapidly developing energy shortage, the only large
reserve of fossil-fuel energy in the United States is coal, so
it is inevitable that coal will be the furnace fuel of the
future. Its use gives a much more luminous flame than does oil
or gas and the heat transfer from flame to charge is thought to
be more efficient. The two main disadvantages of its use is
that it must be finely pulverized just before use, and the pre-
sence of fly ash after its combustion may either form a thin
insulating blanket on the furnace bath and/or contaminate the
gases with additional particulate matter that is hard to remove.
B.5 FURNACE BURNERS
A vast conglomeration of burner types and positions have
been used in various reverberatories in past installations.
Most of the "odd-ball" burners have fallen by the wayside, and
the use of standardized main burners through the bridgewall
(using only pressurized primary and secondary air for idealized
combustion} is now becoming almost universal.
Burners must be large enough and numerous enough to supply the
heat necessary to smelt the tonnage of charge desired-insofar
as cross-sectional area is available for proper combustion. The
trajectory of the flame from each burner must be such that the
maximum heat is transferred directly to the charge and a minimum
to the furnace arch. The use of a combustion chamber upstream
of the furnace has generally been abandoned as uneconomical.
The judicious use of oxy-fuel burners through the arch is becom-
ing more prominent and this is desirable to "level-out" the
temperatures in the charge smelting zone.
In general, furnaces are equipped with five to eight main
burners through the bridge-wall and the ports are usually
jacketed to protect both burners and wall. Each burner is
capable of slight adjustment (movement) to direct its flame in
the desired trajectory. Proper mixture of fuel and air is
essential at all levels of burner capacity.
307
-------
B.6 FURNACE CHARGING CONFIGURATION
There are two types of reverberator!es in current use and
they differ principally in the amount of bath and the method
charging.
The first type in use, and still extensively used, is the
deep-bath furnace, wherein a molten bath of several feet depth
is maintained throughout the furnace, covering the entire
bottom. The charge may be admitted to the furnace by dropping
through the arch, through retractable or fixed charge guns,
discharging through ports in the sidewalls which extend some
distance towards the middle of the furnace, or in one instance
through green charge slingers that cover the entire bath in
the smelting zone. A calcined charge must be used through
charge guns or through arch drop holes in the center or along
the furnace sidewalls. Green charge may also be dropped
through the arch or introduced through belt slingers. In all
cases, every effort is made to distribute the charge on top
of the deep molten bath in the smelting area, to expose the
largest possible area to the heat from the burners and the arch.
Water or aircooled jackets in the sidewalls are used because the
molten bath is in direct contact with the sidewalls.
The second type is the side-charged furnace, (see FigureB-4)
which is more predominately in use today and usually has a bath
depth of three feet or less. Basically, the green or calcined charge
is dropped through the arch along the walls to form banks of material
that slowly melts, giving sufficient protection to the sidewalls to
generally eliminate special cooling requirements. The smelted mater-
ial forms a liquid bath throughout the center length of the
furnace. It is adaptable for either calcine or green feed
charging or a combination of both. Typically, feed is gravity-
charged into hoppers staggered along each side of the furnace and
located above drop holes through the arch, a few inches in from
the sidewalls. The hoppers may be fed by calcine car discharge,
drag chain distribution of calcine or green feed, or by conveyor
308
-------
x
^= U-
o
I C
OJ O)
-O (U
I
QQ
(U
309
-------
belt and tripper car for green feed. As the angle of repose of
the charge bank is lesser with calcine charging, the height of
the charge pile must be lower or the width of the furnace must
be greater than for the bath smelting type. A particular hazard
in side charging is the possibility of a portion of the charge
bank flowing, caving in, or sloughing into the molten bath and
creating a rapid or even explosive reaction between charge and bath
with an accompanying boiling and rapid gas evolution, particularly
with regards to green or wet feed charging. When using the latter
there have been rare instances where the generated pressure was
strong enough to damage the furnace arch and even blow off portions
or sections.
Either spring or suspended arches may be used in furnaces of
either type. If width of arch is greater than twenty-eight feet,
a suspended type arch must be used as hot brick strengths limit
the width of sprung arches.
B.7 SMELTING CONSIDERATIONS
In past years smelter feed consisted primarily of direct smelt-
ing ores high in gangue material, generally low in copper (3 to 10%}
and comparatively high in iron and sulfur. Roasting was thus desir-
able to help concentrate the copper by removing some of the sulfur
and oxidizing some of the iron for subsequent slag formation by com-
bination with the silicious gangue material. In more recent years
flotation methods have been developed that now eliminate most of
the worthless gangue materials as well as some of the excess iron
and sulfur that is not combined with the copper thus furnishing a
concentrated smelter feed that can directly produce a matte of de-
sirable grade without requiring pre-roasting. Of course, if higher
grade mattes are desired, this concentrate can be roasted to remove
additional sulfur and produce matte grades as high as 65% copper
and high grade S02 gas that can be utilized for production of
sulfuric acid, elemental sulfur, etc.
310
-------
B.7.1 CALCINE SMELTING
The grade of matte desired and the sulfur content of the smelter
feed determines the need for pre-roasting, although desired production
of sulfuric acid, or need for increased furnace capacity are other
considerations.
The smelter feed may be calcined in flash, multi-hearth, or
fluo-solid roasters. The primary objective is the elimination of
the first atom of sulfur combined with the iron. Over-roasting pro-
duces sulfates and magnetite and results in a "dead" or unreactive
furnace charge. Inevitably some metal sulfides are oxidized, prin-
cipally the iron and part of qny zinc or lead present. In ores
containing arsenic and antimony, a percent of these elements are
generally fumed off. Controls are more positive in the fluo-solids
roaster, which accounts for its present popularity in new installa-
tions. Elements volatilized or fumed off in the roasters are often
recoverable from the particulate matter collected in the flues by
electrostatic precip^tators or baghouses, and may thus be reclaim-
able. Most roaster gases contain strengths of S02 high enough for
acid production and may run as high as 15 to 29 percent SO,, in the
fluo-solids type of furnace.
Anywhere from 11 to 16 percent of the sulfur content in calcine
from the roasters is released in the reverberatory furnace in the
ratio of about 97 percent in the exit gases and the balance in the
furnace slag. Normal calcine-fed reverberatory exit gas contains
about 0.4 to 1.0 percent SO^ under ordinary operating conditions.
Calcine-fed furnaces generally require between 2.5 to 4.0 million
BTU's per ton of charge to smelt, the variance depending upon the
refractoriness of the charge.
B.7.2 GREEN FEED (WET) CHARGE SMELTING
With advances made in hydrometallurgical and concentrating
processes in recent years, a smelter feed with sufficiently high
copper content usually can be obtained for direct smelting and thus
311
-------
eliminate the added expense of pre-roasting. As long as the charge
does not contain an excessive amount of sulfur (preventing proper
matte grade formation), no roasting is required.
B.7.3 PRODUCTS
B.7.3.1 Matte
Matte, which contains the metallic sulfides, is obtained al-
most simultaneously with the smelting of the charge and, containing
the heavier base metals (Fe and Cu), settles in the lowest layer
of the molten bath. At times, when more than 10 percent of the
furnace charge consists of copper precipitates, a layer of molten
copper has been known to lie beneath the matte. A matte grade of
40 to 45 percent Cu is most economical for subsequent converter
treatment but older smelters have been operated with matte grades
as low as 16 percent Cu. Lower matte grades are desirable when
there is much secondary copper-bearing material such as reverts and
scrap brass to be smelted in the converters and there is sufficient
converter capacity available. Specialized smelting may produce a
matte grade as high as 75 to 85 percent, which is difficult to treat
in the converters.
The matte should be tapped from near the bottom of the molten
bath, preferably through the sidewalls, in the area comprising the
front 30 percent of the furnace length. Tapping techniques vary
in different plants and have been covered by many articles written
in the past. A matte temperature of around 1,000°C (1,800°F) is
desirable for further treatment processes.
B.7.3.2 Slag
Slag, which contains the iron, most of the other base metals
present and the gangue or worthless portion of the feed, forms when
iron in the oxide form contacts silica in the near molten state.
Traditional slag consists of a sesquisilicate or a fayalite-type
slag in the SiO^-FeD-Fe^O, system. Copper is seldom oxidized at
312
-------
this stage, but if copper oxide is present, it follows the iron and
the other base metals. Limestone is usually added to the charge to
lower the melting point, increase the fluidity of the slag and
react with alumina present to form a discardable slag wherein matte-
copper entrainment is minimized. A common factor used in charge
computation is the sidewall silicate degree. This is determined by
the ratio of oxygen in siliceous (acid) materials in the charge to
the oxygen in bases (such as FeO, CaO, Cu^O, PbO, MnO, ZnO, MgO,
BaO, K203, Na20, and usually AlpCL). Slags of these types have
proven to have a relatively low melting point, are low in specific
gravity, and separate easily from matte components.
Normal slags should not contain more than 0.4 percent copper
but laxity of controls and improper operating techniques such as
slag skim level too low, charge rate too high, poor metallurgical
composition, or charge drop in location not giving uniform coverage,
often results in over 1.0 percent copper content in some slags.
In one large western copper plant, some previously discarded slag
containing over 0.7 percent copper is being treated by flotation
for further copper removal.
Generally, the higher "the grade of matte produced, the higher
the natural copper slag loss, within reasonable limits.
In some locations the slag is marketed to cement plants, as
ballast for highways and railroads, or even as abrasive material.
Slag itself does not deteriorate to any marked extent upon exposure
to the atmosphere as illustrated by several dumps still in exist-
ence from abandoned smelters operated as far back as the early 1900's,
B.7.3.3 Furnace Gases
Furnace gases are not treated generally except for heat and
particulate removal. In the interest of economics, the gases are
invariably passed through waste heat boilers which recover in the
range of 25 to 40 percent of the heat value in the fuel, as super-
heated steam. This steam is then used for generation of electricity
313
-------
and compressed air, etc. At one time in an Arizona location, nearly
enough electricity was generated from waste heat steam to operate
the smelter and most of the time the concentrator, mine, and a large
amount of auxiliary equipment such as lighting, etc. This is not
generally the case today, refer to Section B.10.
Often upward of 30,000 cubic meters (100 million cubic feet)
of gas per furnace per day at temperatures in excess of 200°C (400°F)
is involved. Gases are principally introduced into the atmosphere
through stacks upward of 150 meters (500 feet) in height and the
trend is now toward much higher stacks such as the $13 million
380-meter (1,250-foot) high stack at International Nickel in Sudbury,
Ontario, and somewhat similar stacks at ASARCO's Hayden and Tacoma
Smelters.
B.7.3.4 Dusts
In modern smelters, it is almost universal practice to remove
up to 99.8 percent of the contained particulate matter in furnace
gases by means of electrostatic precipitators. The solids thus
separated from the gases are generally then added to the reverbs
or converters as they contain enough copper to make this practice
economical. Modern Cottrells are of the plate, rod, or wire types
and are sectionalized so that one section can be isolated by means
of dampers for "rapping" (dust removal from electrodes), without
interfering with furnace draft requirements or allowing reintrain-
ment of dust.
B.8 CONTROL INSTRUMENTATION
Within a relatively few years vast advances have been made
in the instrumentation available to smelter operators. The follow-
ing instrumentation type and controls are necessary to maximize
reverberatory furnace operation from a centrally-located control
room :
314
-------
Burner Control - Proper combustion should be maintained
on each burner at all times. In order to achieve this,
air and fuel feed rates should be automatically con-
trolled by use of continuous recorded temperature and
gas composition sampling at a few selected points in
the furnace and flues. Manual overrides must be
available to the operator in order to adjust for
abnormal conditions.
Draft Control - Two sets of automated dampers installed
in the flues are desirable. The first would probably
be a butterfly type automatically controlled for a
slightly negative pressure. It would be followed
by a vane-type, also automatically controlled to allow
for wide variances in furnace conditions and draft
that cannot be handled solely by the preceding damper.
Each boiler must be equipped with a water-cooled inlet
drop type flue damper to allow for complete isolation
of the boiler for repairs, cleaning, etc.
Automatic Charging Control - A variable-rate sealed
charging apparatus should be installed at each charg-
ing port and the control should have variable rate-
settings for each unit. In this way the feed-rate
at each charge port can be set so as to be automatic-
ally controlled in accordance with the smelting rate
in that location. Manual overrides must be available
to the operator. Continuous charging is desirable
insofar as it can be accomplished with proper maint-
enance of the charge banks or, in the case of deep
bath smelting, an even spreading over the bath surface
in the smelting zone.
• Observation - Constant observation of furnace
interior is necessary for the operator to
advantageously use the controls and overrides
available to him. This is modernly accomplish-
ed by the use of closed-circuit television
screens in front of the control operator in-
stead of the old method of opening various
observation ports in the furnace walls. Tele-
vision viewing when coal is fired may be dif-
ficult because of greater flame luminosity.
315
-------
Furnace Atmosphere - Recorded gas analysis
observations should be obtained at three or
more locations within the furnace and outlet
flues. If subsequent treatment of gases is
to be practiced, a reading of the gas analysis
entering the treatment unit should be chart
recorded. Oxygen, carbon monoxide and sulfur
dioxide are the primary constituents needed
for control purposes.
Temperatures - Temperature recording is needed
for flame, arch, bath, matte, slag, boiler
inlet and boiler outlet points. It is also
desirable to know Cottrell inlet and outlet
temperatures and occasional bottom temperatures.
Additionally, temperatures of brickwork in speci-
fic areas should be available when desired
to determine refractory thicknesses, develop-
ing hot-spots, etc.
Manual Controls - Unless complete furnace
operations are computer-controlled, it is
sufficient to provide manual controls for
placing of needed fettling material, fluxing
variations or addition of other desired materials
as dictated by observance of furnace interior
and recorded data. Custom smelters will have
greater difficulty in establishing computer control
because of feed type variances.
Bath Measurement - These measurements are needed at
periodic intervals to determine proportion of matte
to slag and to determine bottom elevations and bath
depths. So far this is probably not automated any-
where but can periodically be done manually by use
of sounding bars. The data should be recorded on the
furnace log at desired intervals, usually no less than
twice each shift. A computer printout is used at the
Onahama Smelter in Japan.
Many Other Controls - Many other controls are avail-
able, both automatic and manual, in addition to the
principal ones listed. One recently available tool
is a quick-analyzing procedural method for matte and
slag wherein component analyses are available within
an hour after taking the samples.
316
-------
B.9 ENERGY CONSUMPTION FOR COPPER PRODUCTION
,38
The energy consumption in domestic copper production was
175 trillion Btu in 1973. Energy required to produce one pound of
copper was 50,000 Btu in 1973; this included mining, beneficiating,
smelting, and refining stages. The copper production could be di-
vided into two main stages: Mining-Beneficiation, and Smelting-
Refining. The data for 1973 for these two main stages is given
below in Table B-l.
Table B-l. ENERGY CONSUMPTION FOR COPPER PRODUCTION
Stage
Mining-Beneficiation
Smel ting-Refining
Btu
Consumed
(billions)
87,603
87,773
Btu per
Pound of
Copper
25,497
23,935
Percent
of
Total
49.95
50.05
These two main stages can be further divided into four stages:
Mining, Beneficiation, Smelting, and Refining. The energy consump-
tion for these four stages is given in Table B-2.
Table B-2. ENERGY CONSUMPTION BY PROCESS OPERATION
Operati on
Energy
Consumed
Btu/lb of
Copper
Percent
of
Total
Mining (average of underground and
open pit mines)
Beneficiation (average of flotation,
leaching, and precipitation)
Smelting
Refining
Grand Total
7,560
17,937
17,923
6,012
49,432
15.29
36.29
36.26
12.16
100.00
317
-------
The smelting operation can be further divided into roasting,
shops, and miscellaneous; reverberatory furnace; acid plant; con-
verter; and anode casting. The energy consumption for all these
operations for the year 1973 is given in Table B-3.
Table B-3. ENERGY CONSUMPTION BY EQUIPMENT USED
Equipment Used
Roasting, shops, and miscellaneous
Reverberatory furnace
Acid plant
Converter
Anode casting
Grand Total
Energy
Consumed
Btu/lb of
Copper
2,737
11,932
1,084
1,204
966
17,923
Percent
of
Total
15.27
66.57
6.04
6.72
5.4
100.00
318
-------
APPENDIX C
CONTINUOUS SIDE WALL CHARGING
Jan H. Reimers and Associates Limited
GENERAL
Continuous side wall charging of either green or calcine charge
is one of the methods used for reverberatory furnaces. A bank of
charge is built up along each wall of the furnace and the feed rate
at each point along the wall adjusted to maintain a uniform bank.
This has the advantage of protecting the side walls and exposing a
large surface area to the flame. Smelting then is carried out at
a uniform rate and the variations in the SC^ contained in the off
gas are minimized as compared to the gases from furnaces fed inter-
mittently by such devices as Wagstaff guns or slingers.
Two examples of continuous side wall charging are at Inco's Copper
Cliff smelter Figures C-l and C-2 and Noranda's Gaspe Smelter Figures
C-3 and C-4 copper smelter.
CONTINUOUS GREEN CHARGE
The Gaspe copper smelter went into production in 1955 and is
39
described in a paper presented to the AIME in 1957. This smelter
was designed to treat 450 STPD of concentrate and employed green
side wall charging up until 1973 when a fluid bed roaster was
installed.
40
The initial installation is described in the 1957 paper and
shown in the above mentioned figures. The green charge is fed to
94.5 foot long drag conveyors located on each side of the furnace,
which feed vertical charge pipes spaced at 3 ft. intervals along
the furnace. Concentrates are charged for a distance of 70 feet
319
-------
Q_
Q.
O
CJ
00
-------
o
Q.
Q.
O
O
oo
LLJ
5:
o
O)
a
(T3
c
3
O
4-J
(O
a>
-Q
s_
O)
>
O)
c:
o
o
OJ
oo
to
10
o
S-
CM
I
OJ
CD
321
-------
GO
O
CO
GO
CO
l/J
OJ
O)
D-
Q.
O
o
O)
o.
C3
CO
O
-------
I-']
U.
s-
(U
(1)
co
(U
S-
0)
Q.
Q.
O
O
O)
Q.
V)
(O
C3
I
O
d)
s-
3
CD
323
-------
from the firing wall, whereas the remaining 25 feet are fettled
with siliceous flux.
Shortly after start-up the drag conveyors were replaced with a
belt conveyor system to reduce maintenance. With this arrangement
a reversible belt was used to supply the charge pipes. The concen-
trate filter cake, averaging 9 to 10% moisture, was fed to the
furnace at rates from 500 up to 700 STPD.
Although some operators, using green charge reverbs, have
in the past experienced severe sluffing of the charge bank and
puffing of SO^, this was not the case at Gaspe. In fact, some
sluffing was preferred to prevent a bench building up on the bank.
The system worked well up to 1973 when a fluid bed roaster and
calcine charge system was installed as part of a program to
increase throughput. A similar green charge system, however,
is still in use on the two reverberatory furnaces in operation
at the Noranda smelter in Noranda, Quebec.
CONTINUOUS CALCINE CHARGE
The reverberatory furnaces at the Copper Cliff smelter of Inco
Metals Ltd. are side wall charged with calcine on a continuous
basis. This smelter has been in operation for a number of years
and is described in two special issues of the Canadian Mining
Journal.41'42
Copper-nickel concentrate and silica sand flux are partially
roasted in Herreschoff roasters, located above the furnaces, to
produce a calcine which is fed directly to surge bins. These
feed the drag conveyors on either side of the furnace. The overall
cross-section of the reverberatory building and a furnace are shown
in the attached figures. Cottrell flue dust is added to the drag
conveyor and the charge distributed t& the drop pipes. With dry
charge it is possible to use a slide gate in each pipe to control
the charge distribution and maintain a uniform bank along each
side of the furnace. The banks cover approximately 70 feet of the
324
-------
side wall, measured from the burner end.
Approximately 1500 STPD of dry solid charge is smelted in each
furnace at Copper Cliff. By locating the roasters directly over
the furnaces the calcine enters the furnace at approximately 1000°
F and this provides some saving in furnace fuel.
Other than maintenance on the drag conveyors, no serious
problems have been experienced over the years with continuous
calcine charging at Copper Cliff. On the contrary, it has the
advantage of enabling the operator to adjust feed distribution along
the furnace and maintaining a uniform bank, through adjustments
to the fettling dampers. Inspection of banks indicate no excessive
puffing of S02 or sluffing of the charge.
SUMMARY
Continuous side wall charging is one of the practical and
proven methods of charging a reverberatory furnace. It has the
advantage of protecting the side walls, an important consideration
if oxygen enrichment is used, and evens out variations in the SO^
contained in the off gas. Whether green or calcine charged, there
will be some inleakage of false air due to the number of feed
pipe openings along the furnace. This inleakage, however, can
be minimized by keeping inspection ports closed, when not in
use, and by careful control of furnace draft.
325
-------
APPENDIX D
RECOVERY OF COPPER CONVERTER SLAGS BY FLOTATION
A report by the U.S. Department of the Interior Bureau of Mines
entitled "Recovery of Copper from Converter Slags by Flotation" by
V.E. Edlund and S.J. Hussey of Salt Lake City Metallurgical Research
Center in Salt Lake City, Utah discusses laboratory batch flotation
tests, grindability studies, and cost studies for slag processing.
This is Report of Investigations 7562 (Revised) 1972.
Laboratory batch flotation tests were conducted on copper con-
verter slags to evaluate the relative merits of recovering copper
from slow-cooled versus water-quenched slags. Three slags containing
1.6, 5.0, and 61.6 percent copper were used. More than 90 percent
of the copper was recovered in a rougher concentrate leaving a 0.2
to 0.3 percent copper tailings when treating slow-cooled slag.
Lower recovery and higher copper tailings ranging from 0.5 to 0.6
percent were obtained from quenched slag.
Grindability studies were made on the respective heat-treated
slags. Quenched slags proved more difficult to grind than slow-
cooled slags.
Cost studies showed that quenched slags can be treated at
slightly lower costs than slow-cooled slags. However, the cost
advantage of processing quenched slags is more than offset by
the higher copper recovery obtained from slow-cooled slags.
The relative merits of treating converter slag by water-quenching
versus slow-cooling indicated the following conclusions by the authors
1. Flotation of slow-cooled slag yields a higher copper
recovery and a lower copper tailing.
326
-------
2. Total investment and processing costs would be slightly
higher for treating slow-cooled slag. These higher
costs, however, would be offset by higher copper recov-
eries and slow-cooling would be more economical because
of the additional copper recovered.
3. Quenched slags are more difficult to grind than slow-
cooled slag.
4. The true economics of a method for re-treating con-
verter slag separately would require consideration of
benefits due to increased reverberatory furnace capa-
cities, simplified furnace operations, and possible
lower copper content in the reverberatory slags.
5. The amount of copper in the slag has little influence
on the residual copper content of the flotation tailing.
327
-------
APPENDIX E
ONAHAMA SMELTER REVERBERATORY
FURNACE FLUE GAS CALCULATIONS
General Conditions''^0
Fuel Consumption = 6Kl/Hr = 6,000 1/Hr = 100 1/min
Sp. Gravity of Fuel Oil = 0.90 (@ 100°C)
Oil Analysis: S - 2.5%, C - 87.25%, H - 10.25%
Quantity of Sulfur Oxidized in Reverb. Bath: 3.12 T/Hr
Quantity of Combustion Components
Sulfur in Oil: 6,000 x 0.9 x 0.025 = 135 kg/hr =
2.25 kg/min = 4.96 Ib/min
Carbon in Oil: 6,000 x 0.9 x 0.8725 = 4711.5 kg/hr =
78.53 kg/min = 173.13 Ib/min
Hydrogen in Oil: 6,000 x 0.9 x 0.1025 = 553.5 kg/hr =
9.23 kg/min = 20.35 Ib/min
Sulfur in Bath: 3'1| gg = 317° k9/nr =
52.8 kg/min = 116.40 Ib/min
Theoretical Air Required for Oxidation of 1 Ib:
Sulfur = 4.29 Ibs (1.946 kg)
Carbon = 11.53 Ibs (5.23 kg)
Hydrogen = 34.34 Ibs (15.58 kg)
Amount of Air Required for Oxidation
Sulfur: (116.40 + 4.96) 4.29 = 520.63 Ib/min = 236.16 kg/min
= 6467.5 ft3/min = 183.14 Nm3/min
Carbon: 173.13 x 11.53 = 1996.2 Ibs/min = 905.5 kg/min
= 24797.5 ft3/min = 702.2 Nm3/min
328
-------
Hydrogen: 20.35 x 34.34 = 698.82 Ibs/min = 316.98 kg/rnin
= 8681.0 ft /min = 245.82 Nm /min
Total Quantity of Air Required for Combustion and Sulfur Oxidation
" 183.14 = 702.2 = 245.82 = 1131.2 Nm3/min
Quantity of Flue Gas = Air Required for Oxidation = Combustion
Components
Volume of Combustion Components @ 0°C STP
Sulfur =
' = 1363.60 SCFM = 38.61 Nm/min
Carbon = 173.13
Hydrogen
. 0056
= 3633.93 SCFM = 102.90 Nmmin
ino Q 3
Wet Total = 1131.2 + ~~ = 1182. 7 Nm /min
Volume of .Water = 102.9 NM /min
Dry Total = 1132.7 - 102.9 ~ 1079.8
n
2 Concentration
SO <;as
2
. 40 + 4. 96)
. 1784
32
38 5
SO Concentration =
= 3.62% Wet
53
'
SO Concentration =
3.57% Dry
329
-------
APPENDIX F
JAPANESE REVERBERATORY FURNACE PRACTICE AFFECTING S02 EMISSIONS
F.I ONAHAMA COPPER SMELTER
F.1.1 REVERBERATORY FURNACE DESCRIPTION
There are two reverberatory furnaces currently operating at
the Onahama smelter in Japan. With the exception of the one
reverberatory furnace at the Naoshima smelter these are the only
furnaces of this type in operation in Japan. Table F-l summarizes
the furnace design configuration.
Typical Onahama smelter sulfur balance is shown below in
Table F-2 indicating 99.7 percent sulfur-eliminated from the
stack gas.
F.I.2 METALLURGICAL PROCESSING TECHNIQUES INFLUENCING SO-
EMISSIONS FROM THE ONAHAMA FURNACE
The green charge furnaces are used at Onahama as a result
of a study that was made comparing this approach with a calcine
charge reverberatory furnace, flash smelting furnaces, and the
Mitsubishi continuous smelting process. Economy, actual results,
risk factor, and technical level of the company were all considered
at the time the smelter was redesigned for increased capacity.
Expansion was also limited by the smeTter layout and available
space. Since any new smelting system would necessitate additional
new preparatory systems., ft was determined that the conventional
green charge reverberatory furnace would be used. A bedding
system is used to control charge composition. A total of 55,000
tons of concentrate are handled per month plus scrap and blister
of an additional 4,000 tons per month.
330
-------
Table F-l. ONAHAMA SMELTER REVERBERATORY FURNACES
Fuel
Length3
Width9
Height9 (burner side)
(boiler side)
Bath depth
Section area of flue
Number of burners
Waste heat boilers
Number
Capacity
Steam generation (maximum)
Age
Number 1
Bunker C oil
34 m
10 m
3.6 m
3.4 m
0.8 m
10 m2
8
2
70% (x2)
32T/H (x2)
12 years
Number 2
Bunker C oil
34 m
11 m
4 m
3.4 m
1.1 m
12.5 m2
8
2
100% (x2)
47T/H (x2)
3 years
Inside brickwork
Table F-2. SULFUR BALANCE FOR THE ONAHAMA SMELTER
INPUT
Concentrate
and ore
Fuel oil
Reverts
(recir-
culating)
Tons
Per
Month
15,554
243
793
16,590
Percent
93.9
1.5
4.6
100.0
OUTPUT'
Acid
Gypsum
Slag
Convert Dust
Reverts
(recircu-
lating)
Granulating &
wash water
To atmosphere
Tons
Per
Month
12,562
2,721
308
51
793
108
47
16,590
Percent
75.5
16.4
1.9
0.3
4.7
0.7
0.3
100.0
331
-------
It is necessary to repair and replace portions of the furnace
roof every six months to minimize leakage. This is particularly true
within an area of approximately 10 meters of length from the burner
end.
There have been some problems with a small area in the arch at
the point where the oxygen-fuel burners pass through. This resulted
from the fact that this point was cooled by the burner feed line to a
point where it was 150°C below the dew point resulting in corrosion.
No cooling jackets are used for the ports.
3
The amount of 140 Nm of oxygen per ton of charge was the maxi-
mum used in a test furnace. Oxygen is only used in the main furnace
when an increase in capacity is required.
The following describes the sealing and oxygen enrichment tech-
niques for increasing S02 concentration in the reverberatory furnace
offgas at Onahama, and is taken directly from Reference 47.
"Efforts to enrich S02 concentration of the reverb offgas were
made in two ways; one to minimize air infiltration, the other to apply
oxygen to the reverberatory furnace. Air infiltration into the offgas,
which had amounted to approximately 50 percent of the furnace exit gas
at the uptake, was reduced to less than 15 percent by eliminating air
leaks through crevices, clearances and openings of the furnace roof,
side walls, fettling chutes, damper slots, expansion joints, peep holes,
cleaning doors and especially dust discharging hoppers of the boilers
and the Cottrell treaters. Careful and accurate draft control of the
furnace also served for this purpose to a large extent.
The tedinitpre -ef -oxygen application is different from the con-
ventional way of using oxygen enriched air for combustion of the main
burners. This technique is in some ways similar to the oxygen-fuel
roof burner applied to the open hearth of steel making works. Two
oxy-oil burners are installed vertically penetrating the roof of the
reverb and the intense heat of the flame is directly transferred to
the exposed slopes of the banked charge around the midway of the
reverb, where the melting rate is low in customary operations.
332
-------
Converter slag is returned to the reverberatory furnace. Ap-
proximately 49 ladles per day are returned. Each ladle contains
nine cubic meters. When converter slag is returned to the reverb-
eratory furnace the slag return door is opened, the molten slag
poured in and the door closed.
There is approximately 15 to 20 percent magnetite in the con-
verter slag. Nearly half of this is reduced in the reverberatory
furnace. Therefore the converter slag does contribute to the S02
emissions. Some of this increase will be balanced out by dilution
when air enters the slag return.
SCL fluctuations appear to be primarily influenced by the
charging rate and frequency of charging. The chart in Figure F-l
shows output SCL versus time. The range generally is from 2.0 to
3.2 percent.
The use of preheated air will increase the S02 concentration
by a small amount. Approximately 1 percent of the total sulfur
input comes from the furnace fuel oil.
Pressure control is set by sensor just before the uptake which
is connected to a damper downstream of the boiler. This control
pressure is reset every week or so by observation of leakage out
the roof. Figure F-2 shows the variation of pressure versus the
length in the furnace.
Temperature is controlled and measured by optical pyrometers
or Temp Stiks. The pyrometer is used at the matte tap hole and
the uptake. Furnace temperature is controlled by changing the fuel
rate.
Vertical stratification of S02 has been detected in the furnace.
The maximum amount of S02 appears approximately one meter above
the bath and decreases from this to zero at the roof.
The major leakage points in the reverberatory furnaces are at
the two slag return doors, the charge ports, the burner entry (eight
burners), cracks, the oxy-fuel burner ports and the bath measuring ports.
333
-------
Lt-
334
-------
(U
Q£
OJ
^:
•M
0) "oJ
O)
U
re
I/)
4->
C
•^»
<2
(O
0)
01
cc
0)
O ro
O) (O
c .c:
O O)
-------
Firing rate of individual burners, oxygen-oil ratio, shape and speed
of flames, angle and position of burners and the method of furnace
feed have proven to be of vital importance through three years' opera-
tion in a pilot furnace. Inappropriateness of these conditions causes
increase slag loss, accretion of furnace hearth, brick damage to walls
and roofs and other unusual and hazardous problems. Application of
oxy-oil burners to the reverb furnace has increased by about 15 percent
both smelting rate and S02 concentration of offgas."
"The fuel of the burners is Bunker C oil. High pressure pure
oxygen atomizes and burns the fuel in the furnace upon the fettling
slope. The very short high temperature flame directly melts down the
charged concentrate at an extremely rapid rate. Table F-3 shows the
specifications on these oxygen oil burners. Through extensive test
operations over several years, it has been found that by using oxy-
gen-oil burners, 3,000 tons of concentrate can be smelted additionally
per month and the S02 strength in the exhaust gas can be increased by
0.3 percent per one burner. These auxiliary burners are suitable for
use with green charged reverberatory furnaces to increase capacity
without increasing exhaust gas volume. Therefore, theoretically, it is
possible to increase the smelting capacity merely by increasing the
number of burners, and finally the exhaust gas from the reverberatory
furnace can be handled by contact type acid plant without any heat
supply from outside. From the viewpoint of the damage to the furn-
ace roof brick, this system is undoubtedly superior to the oxygen
enrichment system for the main burners. However, still many prob-
lems remain. For example, there must be a heat balance between auxil-
iary burners to ensure smooth operation."
Table F-3. OXYGEN-FUEL BURNER
Fuel
Fuel consumption (max.)
Oxygen consumption (max.)
Oxygen pressure
Length
Diameter
Cooling system
336
Bunker C oil
0.4 m3/Hr
1,200 m3/Hr
5.0 kg/cm 2
1,930 mm
150 mm
Water Jacket
-------
F.2 NAQSHIMA COPPER SMELTER
F.2.1 DESCRIPTION
The Naoshima copper smelter is located on Naoshima Island
which lies in the Seto inland sea of Japan about 300 kilometers
west of Osaka.48 In addition to the conventional calcine charge
(from fluid-bed roasters) reverberatory furnace, an unconventional
series furnace that combines three metallurgical stages (which are
normally carried out in separate furnaces) as one continuous line
has been placed in operation. A detailed description of this
smelter is included in Appendix G.
The furnace charge prepared in a bedding yard, using computer
analysis, is conveyed to a fluid-bed roaster to be partially roasted.
At least 10 different concentrates are processed at the present
time with no noticeable variations in S0? emissions due to this
factor. Flux for the reverberatory furnace is also added to improve
fluid-bed operation and to heat up the material prior to furnace
feed. About 40 to 45 percent of the sulfur in the charge is elim-
inated in the roaster. Gases from the roaster are sent to the acid
plant for direct processing for S02 control; the principal reason
why this roaster is used.
F.2.2 FACTORS AFFECTING S02 EMISSIONS AT THE NAOSHIMA SMELTER
The average sulfur dioxide volume percent on a dry basis from
the reverberatory furnace is 1.5. Maximum S02 occurs approximately
one minute after dropping a charge through one of the six Wagstaff
guns at approximately 2.7 percent S02. Minimum concentration is
approximately 1.0 percent. No oxygen enrichment is used at the
present time. A charge is dropped at approximately every five min-
utes from one of the guns. The sulfur dioxide sensor is located
downstream of the reverberatory furnace precipitators.
Matte averages 42 to 43 percent copper, 25 percent sulfur,
and 29 percent iron. Sulfur elimination occurs at 40 percent in
337
-------
the roaster, 5 to 10 percent in the reverberatory furnace, and 50
to 55 percent in the converters.
Reverberatory furnace offgas composition includes 3 percent
oxygen at the uptake and this is increased to approximately 5 per-
cent oxygen at the acid plant inlet due to leakage.
Reduction of calcine magnetite by oxidizing of concentrate
does tend to produce a small additional amount of sulfur dioxide
in the furnace. The exact amount has not been determined.
338
-------
APPENDIX G
NAOSHIMA SMELTER
Reverberatory furnace smelting of calcine from hearth roasting
49 50
of copper concentrate started at Naoshima in 1918. ' Both domestic
and foreign concentrates are processed. A smelter addition of
slightly greater capacity than the original smelter and using a
fluidized bed roaster, Figure G-l, was constructed in 1969. This
addition effectively added an entirely new conventional smelter in-
cluding a second reverberatory furnace. Except for minor differ-
ces, the second smelter was essentially a duplicate of the old
smelter with roasters, reverberatory furnace, and converters.
Development work on the new continuous Mitsubishi process
(Section 8.6) began in 1961 and a prototype pilot plant was constructed
and operated at the Onahama smelter. When the pilot plant operation
proved the technical feasibility of the process, Mitsubishi built
a semi-commercial plant at Onahama which was started up in Novem-
ber, 1971. The old original smelter section at Naoshima was closed
in 1973 and a prototype continuous smelter of 50,000 TPY capacity
was started up in 1974 in its place.
The new continuous smelting process off-gases are now combined
with the presently operating (No. 2) reverberatory furnace offgases.
These gases may be directed to one of three sulfuric acid plants.
Roaster and converter gases are blended and passed to a double
contact acid plant.
The roaster gases are passed to a cyclone dust collection system
Figure G-2.
Operating data for the dust collection system is:
339
-------
Figure G-1. Lateral View of Fluid Bed Roaster
1 -- GAS CARRIER OUTLET; 2 -- FEED PORT; 3 -- AIR FOR DISPERSION;
4 — AIR; 5 — MATERIAL UNDER FLOW; 6 — WATER SPRAY
Figure G-2. Fluid Bed Roaster Gas Treatment System
1 — FLUID FURNACE; 2, 3, and 4 -- PRIMARY, SECONDARY, AND TERTI-
ARY CYCLONE; 5 — PREHEAT BOILER; 6 — COTTRELL; 7 — TO
SULFURIC ACID TREATMENT
340
-------
• Primary cyclone efficiency - 90 percent
• Secondary cyclone efficiency - 70 percent
• Steam generated by roaster waste heat
boilers - 7.5 tons per hour
• Electrostatic precipitator efficiency -
99.6 percent.
The following shows typical operating data for the fluosolids
roaster:
• Roaster Charge - 53 wet tons per hour
• Blown in air - 300 Nm3/min (10,593 SCFM)
• Bed space velocity - 80 cm per second
• Tuyere pressure drop - 780 mm water
• Bed pressure drop - 1,500 mm water
• Bed temperature - 630°C
• Freeboard temperature - 560°C
• Freeboard draft - +_ 0 mm water
• Exit gas to acid plant
Volume 530 Nm3 (18,715 SCFM)
S02 concentration - 11 percent
Temperature - 300°C
Dust content - .0047 gr/scf
• Thermal Output
Gas heat content - 57 percent
Calcine heat content - 38 percent
Other losses - 5 percent
G.I REVERBERATORY FURNACE
The partially roasted calcine is conveyed by chain conveyor
from the roaster to the calcine hoppers above the reverberatory
furnace, Figure G-3. The calcine is then charged into the furnace
through the side wall by Wagstaff guns with direct television view-
ing of the bath by the operator. These Wagstaff guns are arranged
in such a way as to allow an even distribution of the calcine over
the bath in the furnace. The furnace is fired with Bunker C oil
through six low pressure burners.
341
-------
Figure G-3a. Vertical Cross Section
of Reverberatory Furnace:
1) ORE CHARGE; 2) WATER JACKET;
3) CLAY LAYER; 4) OPERATIONAL
PLATFORM; 5) BURNER; 6) SLAG LAYER;
7) MATTE LAYER; 8) CHROME ORE
LAYER; 9) CAST SLAG; 10) SAND
LAYER; 11) CONCRETE;
12) CONCRETE MAT
Figure D-3b. Lateral View
of Reverberatory Furnace
1) CONCRETE;
2) CONCRETE MAT;
3) SAND LAYER;
4) REMAINDER ILLEGIBLE
Figure G-3c. Reverberatory Furnace Overhead View
1) OFFGAS DUCT TO BOILER-NEGATIVE PRESSURE; 2) JACKET; 3) SLAG
OPENING; 4) MATTE OPENING; 5) BURNER; 6) RETURN SLAG OPENING;
7) WAGSTAFF GUN FOR CHARGING FURNACE
Figure G-3. Structural Views of Naoshima Reverberatory Furnace
342
-------
Matte averaging 42 to 43 percent copper, 25 percent sulfur,
and 29 percent iron, is tapped from any of four tap holes, located
two on each side, and transported to the converter aisle by elec-
trically operated cars. Slag is skimmed from the skimming hole
near the uptake end of the furnace, is granulated by seawater, and
conveyed to the slag stockyard. The gas from the furnace is cooled
in waste heat boilers and treated by the electrostatic precipitator
which reduces the dust content to less than 0.003 gr/scf.
Specifications for the reverberatory furnace are as follows:
• Length - 33 m
• Width -9m
• Height - 3.76 m
• Smelting capacity - 8,000 tons/month
• Bunker C fuel consumption - 100 liters per ton solid charge
• Steam generated in waste heat boiler - 33 tons per hour
• Precipitator collection efficiency - 98 percent
The structural characteristics of this bath-type furnace are
described as follows:
1. The floor of the reverberatory furnace is composed
of solute slag 2.3 meters thick above a 2-meter
thick layer of sand set on the inner surface of
concrete constructed above rock.
2. To prevent seepage of matte downwards, chrome ore
and magnetite layers are above this. Thus, an
imprevious furnace floor is produced.
3. The shape of the bath portion is that of a crucible.
To prevent erosion of the furnace wall near the
slag line, a water-cooled jacket of cast copper
section 510 mm high is extended along the entire slag
1 i ne.
4. The crucible wall is made of clay. The settling zone
wall, the burner wall and the vertical wall are lined
on the inside with chrome-magnesite brick (magnesia-
bricks near the matte and slag apertures), while the
outside is chrome-magnesite brick. The ceiling is a
suspended structure of chrome-magnesite brick. A por-
tion of the ceiling panel can be adjusted to control
its temperature.
343
-------
5. The path for the gases to the boiler is inclined at
an angle of 26°, and is made as broad and as short
as possible to prevent blockage from accidentally
occuring.
6. In order to improve the thermal efficiency and the
draft, the ceiling at the burner end is raised.
7. There is a slag return opening in the roof above the
burner. This opening is normally closed and not used.
The walls and roof of the furnace appeared to be very tightly
sealed. The roof had gaps between the bricks at the outer sur-
face but were sealed at the inner hot-side. The outer gaps are
presented to allow expansion and contraction of the bricks within
the reverb arch. On walking around the roof there were a few areas
where gas could be detected but, in general, most of the area was
free of noticable odors.
The point at which the burners enter the burner wall is tightly
sealed and there are not other openings. There are some points where
a charge could be dropped through the roof, however, these are closed
by dampers.
Oxygen enrichment has not been used in this furnace. They do
have an oxygen plant at the smelter and do have the capability of
using it if desired. They also have the capability of using pre-
heated air. What is used depends upon specific smelting problems
they have, primarily related to the production rate.
G.2 REVERBERATORY FURNACE WASTE HEAT BOILER
Because the exhaust gas of the reverberatory furnace is at the
high temperature of 1,250°C, thermal recovery is carried out in a
waste heat boiler, and cooling to 350°C results. Two boilers in
parallel operation are usually provided, and they are designed so
that one can maintain the average operational output of the reverb-
eratory furnace when one is shut down for inspection at regular
intervals. Table G-l illustrates the characteristics of the waste
heat boiler.
344
-------
Table 6-1. REVERBERATORY FURNACE WASTE HEAT BOILER CHARACTERISTICS
Item
Inlet gas amount
Inlet gas temperature
Outlet gas temperature
Supplied water temperature
Amount of vapor generated
Vapor pressure
Vapor temperature
Steam uses
Units
3
Nm /mi n
°C
°C
°C
t/h
kg/cm G
°C
—
Design Figures
1,030
1,250
350 + 30
120
33
40
420
Electricity generation
The gas is cooled in the boiler and undergoes 90.6 percent
dust removal through two sets of four horizontal Cottrells.
6.2.1 REVERBERATORY FURNACE
The objectives of the reverberatury furnace feeding system include
preventing as much as possible the temperature decline of the calcine
emitted from the fliro-solids roaster, the maintenance of low fuel con-
sumption of the reverberatory furnace and the preservation of dis-
persibility and fluidity of calcine. To minimize calcine heat loss
and facilitate handling the roaster is placed as close as possible
to the reverberatory furnace.
On each side of the reverberatory furnace, there are three calcine
weighing hoppers and the Wagstaff gun is located below. Upon calcine
charging, the charging window of the reverberatory furnace is auto-
matically opened, the gun is moved pneumatically into the furnace and
the calcine hopper is opened. High temperature calcine is instantly
distributed in a wide area, and it is rapidly smelted. High efficiency
decomposition is carried out with little dust dispersion.
A television camera is used to observe the molten bath within the
reverberatory furnace. When is appears that the last charge material
has been melted in a given area, a new charge is introduced through
the nearest Wagstaff gun. The gun position is selected by the operator.
The frequency of charges is approximately once every five minutes from
345
-------
one gun. Each charge through the gun at 5 to 6 tons is weighed and
recorded. The black and white television picture is very clear in
determining when the last charge had been melted by observation of
dark points and bubbles. The next charge is introduced by simply
pressing a button, with everything automatic. It is a very smooth
operation.
The oil buners are fitted with six low pressure jets. Further,
there are two silicated ore gun feeders above the burner walls, allowing
the slag composition to be controlled depending on furnace conditions.
Generally, a minus 1 millimeter of water pressure in the uptake
of the reverberatory furnace is maintained. Observation of the
chart during operation indicated a variation of approximately +0.2
millimeters of water. The damper pressure control for the furnace
is located at the inlet of the Cottrell. There are also manually
water-cooled drop dampers at the inlet to the waste heat boilers to
allow isolation for maintenance work.
There are two matte extraction outlets in both walls 24 m and
27 m away from the burner walls, the individual slag extraction out-
lets are similarly placed in positions 31 meters away. In usual
operations, the depths of the matte and slag layers are held at
660 mm and 400 mm, respectively.
The slag is granulated by pouring into a large stream of water
directly at the launder exit.
Once or twice every month a maintenance operation is conducted
to replace damaged brick work. The bricks in the roof are hung by
pairs with two bricks per hanger. Some arch sections are panelized.
Maintenance is probably considerably more extensive than in the U.S.
G.3 GAS EMISSIONS PROCESSING
The average sulfur dioxide volume percent on a dry basis from
the reverberatory furnace is 1.5. Maximum S0? occurs approximately
1 minute after dropping a charge through one of the six Wagstaff
guns at approximately 2.7 percent SO,,. Minimum concentration is
346
-------
approximately 1.0 percent. No oxygen enrichment is used at the
present time. A charge is dropped at approximately every 5 minutes
from one of the guns. The sulfur dioxide sensor is located down-
stream of the reverberatory furnace precipitators.
Matte averages 42-43 percent copper, 25 percent sulfur, and 29
percent iron. Sulfur elimination occurs at 40 percent in the roaster,
5-10 percent in the reverberatory furnace and 50-55 percent in the
converters.
Reverberatory furnace offgas composition includes three percent
oxygen at the uptake and this is increased to approximately 5 percent
oxygen at the acid plant inlet due to leakage.
Reduction of magnetite by oxidizing concentrate does tend to
produce an additional amount of sulfur dioxide in the furnace. The
exact amount has not been determined.
Cleaned and blended gases from the continuous and reverberatory
furnaces are treated in one of three sulfuric acid plants as determined
by the operator. There are two single contact and one double contact
acid plants producing 40,000 tons/month of acid which is shipped by
boat to as far away as Australia.
A part of the manufactured acid (5,000 tons/month) at about
50 percent strength is bled from the drying tower and reacted with
the limestone milk in reactors to make wall board grade gypsum.
Three milimeter maximum size limestone is milled in a ballmill and
minus 200 mesh limestone milk is prepared. The gypsum formed is
separated by centrifuge.
The lime scrubber could be described as a type of Venturi scrubber,
however, the water is injected at the top along the centerline of the
Venturi and the gas is injected at right angles to the centerline.
Sea water is now used but this will be changed in the future. No
sludge occurs within this system because the liquid is maintained
at below saturation (calcium sulfate) condition.
347
-------
Total gypsum manufacture is 10,000 tons per month. About
2,000 tons per month of fuming grade acid (oleum) is also produced.
About 98 percent of the smelter feed sulfur is fixed in slag or
captured in gypsum and acid.
The Lurgi double contact acid plant produces 98-99 percent
sulfuric acid. All of the exit gases from the converters and fluo-
solids roaster are treated in this plant. The plant consists of
three semi-venturi scrubbing towers, eight electrostatic mist
precipitators, ten gas coolers, a drying tower, a converter with
heat exchangers and two absorbing towers.
To meet fluctuations of gas volume and S02 concentration, the
acid plants are equipped with closed circuit TV sets, telephone
and light signals so that the operator can always keep track of
smelter operations.
348
-------
APPENDIX H
CONVERTER PROGRAMMING
The converter programming used for this study was based on a
typical converting process as utilized at the Ilo Smelter in Peru
and discussed by Reference 52. This smelter uses three Peirce-Smith
converters plus maintains one on reserve.
The typical converter charge consists of sixteen ladles of matte
added in a 5-2-2-2-2-2-1 sequence; a ladle contains approximately
16.6 tons of matte (32- to 33-percent Cu at Ilo). The first 82.5
tons of matte (5 ladles) are added to a converter containing 5.9
tons of flux remaining from the previous charge. This flux had been
used to cool the bath surface and to get a clean blister copper pour.
Blowing is commenced (416.5 SCFM average). The bath temperature,
initially at 2,012°F, rises until it reaches 2,147°F, at this point
more flux is added, aiming at a slag assay of 24 percent silica
(Si02).
The blowing continues until the temperature reaches 2,300° to
2,372°F. The first slag skim is made. At this point, there is still
some iron sulfide remaining in the matte to impede excessive magnet-
ite formation.
The second slag blow begins with the addition of 33 tons of
matte and more flux (at temperature of 2147°F). Again some of the
iron sulfide is retained in the matte. The slag skim is made. The
converter is skimmed 6 to 7 times and only in the last slag blow is
all the iron sulfide removed.
After the slag blows are completed the copper blow begins.
Scrap copper and reject blister bars are added for temperature con-
trol. When this blow is completed, a boat of flux (5.4 tons) is
added and the copper is poured. The air flow is about 450.1 SCFM
349
-------
during the copper blow. The product is approximately 84.7 tons of
copper.
The time required for this process is 11 hours and 10 minutes
per charge, including 6 hours and 10 minutes for slag blowing, one
hour and 50 minutes for copper blow, and 3 hours and 10 minutes of
down time. This means blow time is 72 percent of the time required
per charging cycle.
This is considered to represent a typical converter operation
as used in the copper industry. Based on this, a basic model for
estimating the S02 concentration and offgas volumes for converter
programming was proposed. The following is a direct quote describing
the model used for programming:
"The basic model for estimation of the SOp concentration of the
converter exit gas has been derived by considering the applicable
thermodynamics, as given primarily in references [53-58],
The nine chemical reactions shown in the Introduction form the
proposed model . Based on the data presented by Schuhmann [53] and
Ruddle [54], the partial pressure of oxygen does not appear ample to
form appreciable amounts of Cu20 during the slag blows; therefore,
equation (6) shall be neglected. This was justified further by Ref-
erence [55]. The slag blow is then described by the following
equations:
2FeS + 302 - ^2FeO + 2S02 + 223,880 Cal
6FeO + 0 - ^2Fe° + 151,800 Cal
2FeO + Si02 - ^(2FeO) ' Si02 + 22,200 Cal
FeS + 3Fe304 + 5Si02 - ^5[(FeO)2 • Si02] + S02 + 4,760 Cal
By the same token, if caution is exercised not to overblow the white
metal-to-blister copper transition, relatively small amounts of Cu20
will form in the blister copper, and the copper blow is described by:
350
-------
Cu2S + 02 *-2Cu + S02 + 183,600 Cal
The following assumptions are made:
1. 02 efficiency will be taken as 75 percent, defined as
kg 09 consumed
C. •
kg 02 blown
2. Charging sequence used at Ilo will be followed. Ilo
typically begins with six tons of flux used in the pre-
vious charge for cooling the bath surface and getting a
clean copper pour.
3. The iron (Fe) in the matte shall be assumed to go to
80 percent FeO and 20 percent Fe304 in the slag.
4. Copper slag losses typically run from 3-5 percent, but
these shall be ignored because of insufficient data on
loss mechanisms.
5. A matte composition (weight, %) is assumed to be 35 per-
cent Cu, 27 percent S, 32 percent Fe, 5 percent Fe~04,
and 1 percent impurities (As, Bi, Pb, Sb, Se, Te).
6. The final Fe30, level in the slag approaches 25 percent
(including the 5 percent from the matte). This tacitly
assumes that the reduction of Fe304 by FeS is equal to
its production by oxidation of FeO.
7. The silica content of the slag is assumed to be 24-27
percent.
8. Following the work of Korakas[593, the (wt % Fe)/(wt % Cu)
in the matte will be 0.025 when all the slag blows are
stopped (except final blow which removes all iron from
matte). This means that approximately 1 percent of Fe will
remain until this final slag blow.
9. Impurities will be ignored quantitatively.
Using these assumptions and keeping the 5-2-2-2-2-2-1 sequence
presented above in mind, calculations can be made using the
model.
Materials/charge:
Total charge = 16 ladles x 15.091
351
-------
= 241,456 kg/charge
Total Cu/charge = (241,456) x (0.35) * 84,509 kg/charge
Total Fe/charge = (241,456) x (0.32) = 77,265 kg/charge
Total S/charge = (241,456) x (0.27) = 65,193 kg/charge
Total Impurity/charge = 2,415 kg/charge
Consider now the first slag blow. This blow processes 5/16 of
the charge or
5/16 x 241,456 kg - 75,455 kg
Cu = 26,409 kg
Fe = 241,456 kg
S = 20,373 kg
Fe304 = 3,773 kg
Impurities = 775 kg
Remembering that the blow will be stopped when wt % Fe (matte)
wt % Cu
» 0.025, 23,485.2 kg of Fe will be converted to FeO and Fe304 in
the ratio of 80 percent to 20 percent respectively, and 13,483.8
kg of S will be converted to S02. This leaves 660.2 kg Fe and
6,988.8 kg S in the matte. Therefore, 18,788 kg of Fe are used
to produce FeO, and 4,597 kg are used to produce Fe304. From
this it can be seen that the slag contains 24,172 kg of F«0 and
6,491 kg of Fe^O. plus the 3,773 kg which was in the matte origin-
ally, for a total of 10,264 kg of Fe-jO^. The flux remaining in
the converter from the previous copper pour is 5,455 kg (66 per-
cent Si02), yielding 3,600 kg of Si02. In order to produce a slag
of 25 percent Si02 additional flux must be added.
.66 X _ n nr
(24,171.7 + 10,264) + x ~ u<"
x = 20,997 kg flux
20,997 - 3,600 = 17,397 kg flux
352
-------
The first slag skim will remove approximately the following amount
of material:
34,436 (FeO + Fe304)
20,997 (flux)
55,433 kg (material removed)
Total oxygen requirement will equal the oxygen which combines
with Fe to form FeO and Fe304 and S to form S02 divided by the
efficiency, as follows:
5,391 kg (from FeO formation)
1,794 kg (from Fe304 formation)
13,484 kg (from S02 formation)
20,669 kg
Total 02 required = -Jfyrp - 27,559
Air is blown during the slag blows at 707.9 SCMM (25,000 SCF) or
148.7 SCMM (5,250 SCFM) of oxygen. This is equal to 212.3 kg/min.
Therefore, it is seen that using air 27,559 kg/212.3 kg/min =
130 minutes are required for the first slag blow. The S02 concen-
tration of offgases can be determined in the following manner.
The average S consumption during the 130 minute blow period is:
13,484 kg/130 min = 103.7 kg/min
This S produces about 207.2 kg/min (72.5 SCMM) of S02 on the aver-
age. The exit stream is, then:
559.3 SCMM NZ
37.3 SCMM 0
2
72.5 SCMM S0
2
669.1 SCMM Total
This produces average S02 concentrations on the order of 10.8
percent, or, with 100 percent dilution, 5.4 percent.
353
-------
Consider now the copper blow. Neglecting the addition of
concentrates and scrap copper used for temperature control , the
converter contains only the white metal (all iron removed on last
slag blow). The slag has been skimmed.
Material in Converter:
84,509 kg CU
21,322 kg S
Material Converted:
21 ,322 kg S
Oxygen Required:
(21,322 x 32)7(32.064 x 0.75) = 28,372 kg of 02
Blow Time Using 40% 02:
28,372/437 = 65 min
Rate S Consumption:
40% 02: =328.1 kg/rain
S02 Production:
40% 02: 655.6 kg/mi n = 229.4 SCMM
S02 Concentrations (No Infiltration):
40% 02: 28.1%
S02 Concentration (with 100% Infiltration):
40% 02: 14.4% "
Scheduling of multiple converters has been examined. The
S02 vs time for a single converter is shown in Figure H-l using
the assumptions of the simple model presented. Based on the
354
-------
cu
(O
o
CXJ U-
o
GO
O O
r- CO
S- C7)
•i—
X.
WW3S '31V« SV9 11X3
355
-------
calculations, five converters would be necessary to have one
converter in the copper blow at all times.
Figure H-2 shows the scheduling for three converters
and Figure H-l is an indication of the percent SC^ at the outlet
by joining the offgases of the three converters. Figure H-l
also shows the flow rate fluctuations during the operations.
These figures reflect conditions after 100 percent infiltration.
Based on the figure, exit gas flow rate varies from 1,416 to
4,361 SCMM; S02 concentration (vol percent dry) varies from
5.4 to 7.8."
Using the above model as a guide, a converter program was
carried out based on the blowing times as given above. The con-
verter program thus used in this study is presented in Figure H-2.
Initially S02 concentrations of 7.8 and 5.4 percent were used for
the copper and slag blows, respectively, as used in the basic
model. Similarly volumes of 52,000 and 50,000 SCFM were used for
the copper and slag blows, respectively. The resulting gas char-
acteristics versus time were similar to those obtained by the model
The values originally obtained for the base smelter system,
Appendix Y-l, proposed converter offgas averaging 5.7 percent S02
at 64,500 SCFM for the smelter system under consideration. This
implied each converter produced an offgas of approximately 30,000
SCFM. Also, the average gas characteristics are maintained from a
sulfur balance based on 1,400 TPD concentrate.
For this study, gas characteristics between the two values
are used. Thus 40,000 SCFM for the slag blow and 42,000 SCFM for
the copper blow are considered representative for converter off-
gases. Since there are three converters, the average number
blowing at any one time is two. Therefore, 82,000 SCFM is used as
the average converter offgas volume. A sulfur balance indicates
that 4.5 percent S02 at 82,000 SCFM corresponds well with the 5.7
percent S02 at 64,500 SCFM. Finally, the S02 concentrations were
adjusted to give the average S0« concentration of 4.5 percent.
356
-------
* ,2
O CO
CO S_
OJ
en Q-
(C O.
i— O
co o
o
o
VO
o
s
o
CM
O
VO
CO
O
O
CO
O
«*
CM
CO
8
s-
o
o
CO
o
o
CO
o
o
CO
o
o
o
CM
en
O (O
CO i—
•— CO
o
CO
-------
The actual values used in this study are shown in Figure H-3 and
H-4 while numerical values are presented in Table H-l. As can be seen,
the SCL concentration varies from 4.0 to 6.6 percent while the flow
rates vary from 40,000 to 122,000 SCFM.
These values appear to be more representative of current operating
practice corresponding to data reported from actual copper converters
using efficient hooding and offgas systems.
This adjusted model seems feasible even for processed ores con-
taining high impurity levels of As, Sb, Pb, and Zn. Assuming that
there are still traces of these impurities in the matte, the longer
slag blowing times are around 1,200°C will insure removal prior to
the blister forming blow. Thus, oxygen enrichment for the copper blow
should not affect the final copper quality since these impurities
will be removed during the slag blows.
358
-------
SCFM
xlOOO
LI
30 60 90 120 150 180 210 240 270 300 330 360 390 420 450 480 510 540 570 600 630 660
TIME (MIN)
Figure H-3. Converter Offgas Volume vs Time
Characteristics for Programming
359
-------
6.6
' i
5.7
5.4
II
4.0
0 30 60 90 120 150 180 210 240 270 300 330 360 390 420 450 480 510 540 570 600 630 660
TIME (MIN)
Figure H-4. Converter %S02 Offgas
Characteristics for Programming
360
-------
Table H-l. CONVERTER OFFGAS VOLUME AND S02 PROFILES
02 and no 02 Enrichment
Time
(M1n)
5
15
11
14
40
10
15
9
10
20
5
7
10
5
10
20
29
5
5
4
24
15
25
10
Percent
(so2)
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
6.6
5.7
5.4
5.7
5.4
4.0
4.0
4.0
4.0
4.0
4.0
Volume
(SCFM)
120,000
80,000
40,000
80,000
120,000
80,000
40,000
80,000
120,000
80,000
40,000
80,000
40,000
42,000
82,000
122,000
82,000
122,000
120,000
80,000
40,000
120,000
80,000
120,000
Time
(Min)
25
15
10
11
14
5
20
10
25
5
16
14
15
9
26
19
5
19
5
5
29
20
10
5
Percent
(so2)
4.0
4.0
4.0
4.0
4.0
5.4
5.7
5.4
5.7
5.4
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
4.0
5.4
5.7
6.6
5.7
Volume
(SCFM)
80,000
120,000
80,000
40,000
80,000
122,000
82,000
122,000
82,000
122,000
80,000
40,000
80,000
120,000
80,000
120,000
80,000
40,000
80,000
120,000
122,000
82,000
42,000
82,000
361
-------
APPENDIX I
CONVERTER SLAG RETURN: ADVANTAGES AND DISADVANTAGES
More and more smelters are weighing the economic advantages and
disadvantages of returning converter slag to the reverberatories. For
those smelters which have been plagued by magnetite and its build-up
in furnace bottoms or hearths and copper entrainment in furnace slags,
the eliminatfon of the magnetite source in the converter slag has re-
sulted in much improved furnace conditions. Some smelters, generally
using green feed, have been able to practice magnetite control in the
furnaces and still prefer the return of hot converter slag to the fur-
naces.
Among the advantages of converter slag return to reverberatories
are (13):
a. Retention of latent heat in converter slag for addition to
furnace heat reserve.
b. Rapid chemo-thermal reaction of the slag constituents with
the molten bath, resulting in an increased smelting rate.
c. Economic furnace separation of copper and precious metals
in the converter slag while molten and their return to the
matte in the furnace.
d. General economy of converter slag treatment and handling
while still in the molten state.
e. Minimization of equipment involved in returning slag to the
furnace as compared to the equipment required for the flo-
tation process. For example, the power requirement for
crushing and grinding is approximately 40 KWH per ton.
f. Cheaper disposal of undesirable elements by their elimination
in the reverberatory slag.
g. Assistance in maintaining an open flow channel in the furnace
bath from the bridgewall to the skimming end, washing away
"floaters" and charge bridges. This is principally due to a
slight rise in the bath elevation in the burner end when slag
362
-------
is received and which is eventually transmitted down through
the length of the furnace to the skimming bay.
Among the advantages of not returning converter slag to the re-
verberator ies are:
a. Elimination of most build-up of magnetite in the reverb-con-
verter system and its many detrimental effects on hearth and
bottoms, higher slag losses, additional fluxing required,
higher slag temperatures required for separation, etc.
Note: Part of the problem concerning magnetite is that
there has been much difficulty encountered concerning rou-
tine percent determination by analytical methods. With
the advent of modern laboratory equipment, this has become
more rational and it is found converter slags normally con-
tain 40 to 50 percent Fe (an easy laboratory determination)
and that a larger percentage of the iron may be in the mag-
netite form than formerly thought. This can result in an
analysis as high as 53 percent Fe30. in some cases.
b. Prevention of surges in gas flow at the furnace outlet,
usually accompanied by increased S02 concentrations, caused
by the rapid chemo-thermal reaction between the slag and
the furnace bath.
c. The reduction of air infiltration from slag receiving-launder
openings which causes undesirable cyclic presence of oxygen
and reduction of S02 concentration in the furnace atmosphere.
The volume of the furnace exit gases is also increased by
this excess infiltrated air, resulting in greater heat loss
from the furnace. Those problems are thus eliminated.
d. Elimination of the oftentimes excessive heat generation in
the immediate contact area of converter slag and furnace
bath, with accompanying splashing of the bath on bridge and
sidewalls, causing damage to the brickwork.
e. Elimination of the higher labor and maintenance cost involved
in the upkeep of slag return launders. Usually one operative
furnace crewmember is required solely for launder cleaning on
each furnace shift as well as some mechanical maintenance re-
quirements.
f. Elimination of the high capital, upkeep and replacement cost
of equipment involved and even more so if retractable laun-
ders are used.
g. Reduction of spillage of converter slag. Spillage requires
that more reverts be smelted and that labor be involved in
cleanup in the converter aisle.
h. Without slag return, fluxing requirements to treat the mag-
netite are lower resulting in less material passing through
363
-------
the furnace. Also, less reyerberatory slag is produced (only
half) and cost of disposal is proportionately lower. This
results in lower copper-pound loss due to the decreased vol-
ume. In some locations the percentage of copper loss in the
furnace slag is higher than normal, mainly due to the presence
of excess magnetite. Copper content of furnace slags without
converter slag returns have decreased as much as 0.2 percent.
364
-------
APPENDIX J
OXYGEN ENRICHMENT EXPERIENCE AT THE CALETONES SMELTER
Experimental work on the use of oxygen in green feed reverbera-
tory furnaces was started at the Caletones smelter in Chile in 1971.
This development work was continued through to 1976. During this
time, a gradual transition was made from supplying fuel convention-
ally through air-oil burners at the end of the furnace to totally
through oxygen-fuel burners positioned over the furnace area and sus-
pended from the roof.
Eventually one furnace was converted to 12 oxy-fuel burners
installed through the roof of the furnace so that the flame impacted
on the charge banks. A total of 380 long tons per day of oxygen
allowed a smelting rate of 1,520 dry tons per day with decreased fuel
consumption in the range of 0.81 x 10 Kcal.
Matte grade tended to increase from 38 percent copper before
major use of oxygen enrichment to 49 percent copper with full oxygen
usage. Copper in the slag reduced from 1 percent without oxygen
enrichment to 0.7 percent with full oxygen. Another additional
effect with the slag was noted that because of its higher temperature,
it was possible to eliminate calcium carbonate used as flux (approxi-
mately 4 percent of the charge). The increase in matte temperature
also minimized the furnace bottom buildup with magnetite. The
smelting level and production increased significantly to values over
100 percent of those with no oxygen enrichment.
The most significant result from a pollution control standpoint
was that the S0? concentration range in the offgases increased to
365
-------
5.8 - 7.3 percent. This is sufficiently high, of course, to allow
direct processing of the gases in a sulfuric acid plant.
It was necessary to use a bottom ventilation system when full
oxygen was used for this reverberatory furnace. However, furnace
wear on a per-ton of copper basis was either less or the same as
previously encountered.
366
-------
APPENDIX K-l
PROCESS DESIGN FOR DIRECT PROCESSING OF REVERBERATORY FURNACE GASES
by Tim J. Browder
Tim J. Browder Company
K.I GENERAL
The Browder reverberatory furnace gas sulfuric acid plant is
designed in this study to produce 192.6 short tons per day of sul-
furic acid (based on 100% H2S04) for each 1.0% of S02 in the feed
stream entering the S02 converter when operated at a maximum flow rate
of 170,000 Nm3/hr (100,000 SCFM) of dry gas to the converter. Assumed
characteristics of the reverberatory furnace effluent are:
Volume Flow Rat
0.5% to 1.5% SO
Volume Flow Rate - 170,000 Nm3/hr (100,000 SCFM)
J2
2.0 - 9.0% C02
0.03% SO,
'3
).
Balance N
4.0 - 20.0% H20
2
Acid plant emission SOp = 450 ppm (maximum)
Assume acid plant adjacent to reverberatory furnace
Altitude 1525 meters (5,000 ft.)
Gas temperature out of reverberatory furnace 1200°C (2200°F)
Gas temperature out of W.H.B. 400°C (750°F)
Dust Load 50 grain/SCF
Figure K-l presents the flow system. Metallurgical feed gas
containing a weak stream of S02, 02> COp and N2 and some acid mist
in the presence of water vapor is collected from the exit side of a
reverberatory furnace gas hot Cottrell precipitator and is conveyed
in a gas flue to a humidifying tower. The humidifying tower humidi-
fies and cools the gas and removes some of the particulate matter.
The gas next passes to a washing tower in which further cooling of the
367
-------
->
u
I
0 o
0 U
i "
s *» =».
it*
8 J i
K sS
W " -5
8.So'
O •; E
O e
X o. o
— «n
cn
< M
o «
>• Ul
(E (/)
o:
ui
m
cc
ui
ui
u.
o
ui !Z
x<
W -I
^ o-
So
o
Uj g
u. cr
(O
I
363
-------
gas is accomplished by recycled cool liquor contacting the counter-
current flow of gas leaving the humidifying tower. After being washed
to remove a large portion of the dust and some sulfuric acid mist, the
gas passes to the wet ESP (electrostatic mist precipitators).
The gas passes from the mist precipitators to a series of
condensers. In the first series of condensers, the gas is indirectly
cooled in the condenser tubes using cooling water. The second
series of condensers has a closed circuit refrigeration unit whereby
chilled water is used to further condense the water vapor from the
gas in order to achieve water balance before passing to the acid
plant's drying tower.
The clean refrigerated gas, with the proper moisture content
(water balance), then flows through the drying tower which removes
the remaining moisture from the gas. The drying tower gas is then
passed through a two-stage demister on the top of the drying tower
and then flows to one operating (of two installed) main gas blower.
The gas blower propels the gas through the plant to finally accom-
plish the manufacture of sulfuric acid.
The gas is next heated in order to be able to pass to the
converter. The preheating of this gas is accomplished in two
stages. The blower discharged gas is passed first into the top
vestibule of a cold gas heat exchanger flowing down through the tubes.
This gas is indirectly heated by a counter-current stream of hot gas
leaving the converter. The gas which has then been preheated in
the cold heat exchanger next flows to a hot heat exchanger before
passing to the converter. The gas is thus indirectly heated to the
converter catalyst ignition temperature in the hot heat exchanger
with a stream of hot gas. The temperature of this gas is controlled
by mixing reverberatory furnace offgas from upstream and downstream
of the waste heat boiler. Mixing is accomplished in the jug damper
(Figure K-2). The hot preheated gas at approximately 438°C (820°F)
next passes into a multi-stage converter. The converter processes
the S02 to SOg. The gas leaving the converter stages is cooled as
required before passing back into additional catalyst stages.
369
-------
y*5 \ ' ££
-------
From the final catalyst bed, the gas is then passed to the shell
side of the cold heat exchanger where the gas is cooled to approxi-
mately 440°F before entering the Freon superheater and then flows
to the absorption tower. The absorption tower absorbs the SO.,
from the gas and the remaining gas is then passed through a high
efficiency demister (mist eliminator) and then vented to the
atmosphere through an exhaust stack.
K.I.I GAS CLEANING
K.I. 1.1 General
Reverberatory gas which has been passed through the rever-
beratory furnace waste heat boilers and through a hot Cottrell to
remove some of the particulate matter is diverted by a new damper
installed in the main line to the humidifying tower. The gas
is conducted through a carbon steel duct, which is externally
insulated with rockwool blanket insulation, to the humidifying
tower. The humidifying tower is a carbon steel vertical mounted
vessel with conical bottom containing teflon and polypropylene
linings. Inside of these linings are placed acid-proof brick lin-
ing unless the gas contains fluorine, then carbon brick will be
substituted. The inlet nozzle of the tower which is close to the
tower base and above the conical bottom is water jacketed and cooled
in order to prevent stress problems from occuring in the steel shell.
The hot gas enters this humidifying tower at approximately 370-400°C
(700-750°F) through the gas nozzle, and turns and flows up through
the empty tower. The tower shell contains a series of spray nozzles
through which weak acidic recycled liquor is sprayed into the tower
counter-current to the gas flow stream. This spraying of the
liquor into the tower evaporates some of the water into the gas
stream thereby humidifying the gas and reducing the temperature
to approximately 66-68°C (150-155°F). The top dome of this tower
also contains spray nozzles and emergency spray nozzles which are
used in case of power or pump failure. These emergency spray
nozzles allow the towers to have water sprayed continuously to
371
-------
prevent damage to the tower in case there is a power failure. The
gas leaving the top of the tower is then conveyed down through
fiberglass reinforced plastic (FRP) duct in where there is a swirler
knockout to reduce the mechanical water droplet carryover from the
humidifying tower. The gas then enters a washing tower for further
cooling and further removal of dust and acid mist. The gas which
has been cooled to approximately 66-68°C (150-155°F), then flows
up through the washing tower which contains 3-1/2 x 3 1/2 inch
plastic (PPE) Pall Rings. Over this tower is circulated weak
recycled cooled liquor which accomplishes further scrubbing (wash-
ing) of the gas to remove particulate matter, and acid mist which
is present in the gas, before passing to the wet electrostatic
acid mist.
The down flowing acidic liquor leaving the washing tower is
pumped through a series of Carbate coolers to remove the heat that
has been recovered in the washing tower. This heat removal occurs
as a result of water vapor being condensed from the gas. This water
vapor has previously been added to the gas as a result of humidifi-
cation in the previous humidifying tower,
K.I.1.2 Acid Mist Precipitators
The scrubbed, humidified and washed process gas then passes
through the acid mist precipitator system. The electrostatic type
acid precipitators are comprised of lead tubes with a lead covered
electrical charging wire located in the center of each tube. The
processed gas passes up through the annular space between the cen-
ter wire and the tube. Liquid acid droplets and particles
are electrostatically charged and are collected on the inner
surface of the tube and drained to the bottom of the precipitators.
The acid mist precipitators are equipped with water sprays which
may be used to clean the precipitator tubes while the plant is in
operation. Effluent is then directed to an acidic disposal system
outside of the battery limits of the plant.
372
-------
K.I.1.3 Gas Dehumidification by Cooling
The clean process gas then flows to two sets of condensers.
The gas leaving the acid mist precipitators is conveyed in an FRP
duct into the top section of the condensers. The condensers are
vertical shell and tube Carbate units in which the saturated gas
enters the top FRP vestibule flowing down through the Carbate
tubes. Cooling is accomplished in the first set of condensers
using cooling water recycled from a cooling tower outside of the
plant's battery limits. The gas is then collected in the bottom
vestibule of the first set of condensers and condensate is separated
from the gas. The gas is then conveyed through an FRP duct and to
the second bank of condensers. The second bank of condensers
contains a closed circuit refrigeration unit in which chilled
water is used to further reduce the gas temperature in the
condensers to approximately 70C (45°F) before passing to the
drying tower in the sulfuric acid plant.
K.I.2 REVERBERATORY FURNACE SULFURIC ACID PLANT
K.I.2.1 Drying Tower
The clean wet feed gas from the gas cleaning plant is now
drawn to a packed drying tower over which 94% sulfuric acid is
circulated in order to completely dry the gas. The dry SOp gas
leaves the drying tower through a two stage demister (mist elimi-
nator) located in the top of the drying tower. The 94% sulfuric
acid for the drying tower is supplied from the drying tower sump
and flows by gravity to the drying tower pump tank from which it
is recirculated by an acid pump to the top of the drying tower
weirs from the drying tower acid coolers. Since the acid in the
drying tower is continuously being diluted, the strength of this
acid is maintained by cross-transfering the required amount of
strong 98.5% acid from the absorption tower system to the drying
tower pump tank. The change of level that results from the cross-
transfer is controlled automatically by level control instrumentation
and valving.
373
-------
K.I.2.2 Drying Tower Demister
The drying tower demister consists of two horizontal pads of
teflon media supported by CA-20 grids. These horizontal pads are
located in a chamber mounted directly on top of the drying tower.
The purpose of these demisters is to eliminate any acid mist carry-
over from the drying tower and to protect the blower and the remain-
ing down stream equipment in the acid plant from corrosion.
K.I.2.3 Main Gas Blower
The clean-dry process gas which has been pulled through the
system and through the drying tower demisters now enters one of two
main gas blowers (one blower being a spare). These blowers supply
the necessary pressure for the gas to flow through the remaining
equipment in the acid plant. The suction and discharge lines of the
blowers are carbon steel ducts which are normally only painted
externally and not insulated. The gas then flows from the blower
into the heat exchangers, to be preheated to allow the gas to obtain
the proper catalyst ignition temperature, before being passed to
the converter system.
K.I.2.4 Cold Heat Exchanger
The cold heat exchanger is a vertical carbon steel vessel
mounted on steel grillage. The blower discharge gas enters the
top vestibule of this cold heat exchanger flowing down through the
tubes thereby being preheated in a counter-current flow with hot
gas exiting the last stage of the converter which is passed on
the shell side across the tubes of the heat exchanger. The gas
leaves this heat exchanger and then flows to a shell side of the
hot heat exchanger. This entire cold heat exchanger is constructed
of carbon steel plate with internal aluminized parts and contains
a top and bottom tube sheet with the necessary carbon steel tubes.
Additional service life can be obtained from this heat exchanger
by using Alonized tubes. This heat exchanger is insulated with
two inches of rockwool insulation on the outside, and in addition
374
-------
contains weather seal protection. The blower discharge duct lead-
ing into the top vestibule of this heat exchanger is carbon steel.
The hot gas leaving the lower vestibule of this cold heat exchanger
passes into a duct which is aluminized, and the gas is conveyed
through this duct over to the hot heat exchanger.
K.I.2.5 Hot Heat Exchanger
The hot heat exchanger is a specially designed vertical unit
which allows the gas which has been previously heated in the cold
heat exchanger to flow on the shell side of the hot heat exchanger
up across a series of tubes, leaving the top of the shell side of
the heat exchanger and then flowing to the converter. This gas is
heated counter-currently with mixed hot gas exiting the reverberatory
furnace and waste heat boilers. The top vestibule of this heat
exchanger contains refractory to protect the heat exchanger from
extremely high temperature which can be encountered on startup
since this unit is also used as a preheater during startup
operations. This heat exchanger has a special bottom vestibule
with a conical bottom and contains the tubes extended out approxi-
mately six feet beyond the lower tube sheet to act as knockouts
for any course dust which is conveyed down through the pipes. The
conical bottom of the heat exchanger acts as a collection device and
contains two double weighted valves to remove any dust which settles
out in the bottom of this conical hopper. The heated gas is then
conveyed to the converter. This is accomplished in a carbon steel
duct externally insulated and internally aluminized.
K.I.2.6 Converter
The previously preheated clean dry process gas enters the
top bed of the converter and passes down through a series of
vanadium pentoxide catalyst beds. The process gas, after being
heated as a result of the exothermic reaction that takes place in
the catalyst bed, leaves the converter chamber and passes externally
to the horizontally-mounted gas cooler on the side of the converter.
375
-------
The gas, after passing across the tubes of this gas cooler, returns
to an additional catalyst bed and then flows through this bed where
further conversion takes place with some slight temperature rise.
The gas from other successive beds may also be heated as a result of
this exothermic reaction and cooling with gas coolers as required.
Gas finally leaves the bottom of the converter and flows to the lower
shell side of the previously mentioned cold heat exchanger.
The catalyst bed design considerations must include the con-
ditions generated by the plant duty cycle. Turn down ratio of
170,000 Nm /hr (100,000 SCFM) system design herein may be as low
as 40% over a range of SOp concentration from 0% to 2.5%.
Reverberatory furnace S02 emission concentration will vary in rela-
tion to charge frequency with maximum S02 occurring at the time the
charge is dropped. Some furnaces have been known to generate peak
S02 concentrations considerably above 2.5% but usually for only
periods of 2 to 5 minutes.
The large duct volumes downstream of the furnace will tend
to mix the high concentration to more nearly the average values
of 0.5 to>1.5%. The catalyst tends to adsorb S02 and will serve
as a "sink" under short term peak conditions to minimize acid
plant peak emissions. The catalyst and steel in the converter
and associated ductwork will serve as a thermal "sink" to mini-
mize temperature peaks so that the heat exchanger system opera-
tion will not be effected to any great extent. If, despite the
above, S02 surges do effect the plant operation, additional
catalyst can be added (it should be noted that a large extra
quantity has been included in the present design to provide an
operating margin).
The converter is a vertical carbon steel vessel mounted on
steel grillage beams and contains a dish top head. The internal
beds of the converter are supported by a vertical cast iron post.
Each catalyst bed consists of catalyst and quartz, and is supported
on triangular cast iron grid sections which in turn are supported
by the cast iron post. The internal surface of the converter is
376
-------
completely aluminized to protect the converter shell from scaling.
The external surface of the converter contains four inches of rock-
wool blanket insulation and is protected from the weather by weather
seal.
K.I.2.7 Absorption Tower
The SO, gas, after leaving the final stage of the converter,
passes to the shell side of the cold heat exchanger. The con-
verter exit gas is cooled by the blower discharge gas inside the
tubes of the cold heat exchanger. The gas at approximately
221-227°C (430-440°F) is then passed to the Freon superheater and
then to the absorption tower. The SCL is removed in the absorption
tower by 98.5% sulfuric acid continuously recycled over this tower
and down through the packing. The acid leaves the bottom of the
absorption tower and flows to an acid pump tank. From the pump
tank, the acid is conveyed by one of two installed pumps (one
being a spare), and conveyed to the acid coolers to remove the
heat generated in the tower before the acid flows back to the top
of the absorption tower. The absorption tower is similar to the
drying tower and contains cast iron acid distributors.
K.I.2.8 Absorption Tower Demister
The top of the absorption tower contains a high efficiency
demister, usually a York type SA. This unit will remove any acid
mist which has been formed previously in the system and return it
as droplets to the absorption tower to be recovered.
K.I.2.9 Stack
The residual gas leaving the absorption tower is conveyed in
a carbon steel duct leading to a vent or exhaust stack. The gas
is then discharged to the atmosphere. The S02 concentration of
this gas will be less than 300 parts per million at all times when
initially starting with 1% SO,, gas. The gas, however, may be ducted
back to the original flue and put into the main exhaust stack for
377
-------
the smelter if required.
K.I.2.10 Acid Pump Tanks and Pumps
In order to maintain sufficient acid quantity and to main-
tain the required strength in both the drying and absorption tower
system, pump tanks are included. The pump tanks serve as a surge
volume and also contain the acid which will be recirculated over
the towers. The drying tower pump tank contains two acid pumps.
One pump is a spare. The acid flows from the towers into the pump
tanks. The acid is then pumped through a cast iron pipe to the
acid coolers and from the acid coolers back to the tops of the
towers. The absorption tower pump tank is similar in design to
the drying tower pump tank.
K.I.2.11 Acid Cooler System
Each acid system contains a separate pump tank and recircu-
lating pumps as well as separate acid coolers. The acid cooler
system consists of a horizontal stainless steel shell and tube
unit. The acid flows through the tube side of the acid cooler
and Freon is on the shell side.
K.I.2.12 Acid Product Transfer and Drain
Product acid is pumped to OBL (outside battery limits) and
into an acid storage area. The product pump also serves as a spare
for the drain pump which may be used to drain and pump acid from the
system for servicing or for maintenance.
K.2 OUTSIDE BATTERY LIMITS (OBL) PLANT FACILITIES
K.2.1 COOLING TOWER
A three cell double cross-flow cooling tower is designed to
accommodate the cooling loads for the weak acid scrubber system,
the coolers for the drying and absorption tower systems, the con-
densers and refrigeration unit for the plant. The cooling tower is
378
-------
normally not considered a portion of the process part of the acid
plant and is usually therefore outside battery limits (OBL) of the
plant. The cooling tower is usually constructed of wood and mounted
on a concrete water sump base. The cooling water after accomplishing
its function of cooling various components in the plant is pumped
while still under pressure to the top of the cooling tower and flows
down a series of grids to the bottom of the tower. Air is conveyed
through the bottom of the tower and flows crosswise to a center
phleum section and vented out through propellers mounted on the top
of the cooling tower.
K.2.2 COOLING WATER PUMPS
The cooling water pumps are mounted in a sump adjacent to
the cooling tower. There are three cooling water pumps, normally
two in operation, with one spare. The two pumps circulate the
cooling water from the base of the cooling tower through the respec-
tive acid coolers, refrigeration unit, condenser, and weak acid
coolers. The water is then returned to the top of the cooling tower.
K.2.3 PRODUCT STORAGE TANKS, LOADING PUMPS AND ASSOCIATED EQUIPMENT
The processed loading system tanks and other units are normally
outside of the battery limits of a sulfuric acid facility and many
times are not included in the process design since many companies
already have acid storage systems.
K.3 PROCESS GUARANTEES
Sulfuric acid plants are usually guaranteed based on the
production of sulfuric acid for a fixed gas flow and for a fixed
percent gas strength. For varying gas flows and varying gas
strengths, the usual design then encompasses the maximum gas flow
and the maximum gas strength in order to establish the required
quantity of catalyst to be charged to the converter. The rever-
beratory furnace gas sulfuric acid plant which, for this analysis,
contains a maximum of 1.5% SOo is based on a design total gas flow
379
-------
rate of 170,000 m3/hr (100,000 CFM) which requires about 165,000
liters of catalyst. A catalyst manufacturer would supply the cata-
lyst guaranteed based on the gas flows, pressures, altitude and
operating conditions both in flow and SOp concentration of the
gases. Appendix K-3 includes a typical computer run showing these
calculations, the stack gas analysis, and the required catalyst
loading and operating temperatures for such an installation.
Under the conditions of the design of reverberatory furnace
gas plants, it is very easy to obtain stack gas effluent of less
than 450 parts per million in most cases and with low gas strengths
the stack gas could be as low as 50 parts per million with a single
contact system.
K.4 MECHANICAL DESIGN
K.4.1 GENERAL
This section presents a brief description of mechanical de-
sign, major equipment and supplementary facilities proposed for
the reverberatory gas sulfuric acid plant. The design data repre-
sented herein and included in Appendix K-2 are based on the plant
operated at an average flow of 170,000 Nm3/hr (100,000 SCFM) of
clean dry gas entering the sulfuric acid plant converter and when
supplied with all utilities and materials required at the use con-
dition with the proper proportion of sulfur dioxide, oxygen, and
water. Design, fabrication and installation of the equipment and
facilities is in accordance with the standard practices of the sul-
furic acid industry and OSHA standards. A material list is in-
cluded in Table K-l and utility requirements are summarized in
Table K-2.
380
-------
TABLE K-l
REVERBERATORY FURNACE GAS
SULFURIC ACID PLANT
(100,000 SCFM)
EQUIPMENT LIST
ITEM NO.
Towers
T-1101
T-1102
T-1103
T-1104
T-1105
SERVICE
Humidifying Tower
Washing Tower
Drying Tower
Absorption Tower
Lime Silo
DESCRIPTION
18 - 0" dia. x 40'0"
with Cone Bottom,
Carbon Steel, Teflon,
Ppe and Carbon Brick-
lining.
22' - 0" x 36' - 0"
High, FRP
20 - 0" dia. x 25' -
0" High, C.S., Teflon,
and Acid Proof Brick-
lined. Double Gas In-
let Nozzles
20 - 0" dia. x 25' -
0" High,C.S., Teflon,
and Acid Proof Brick-
lined. Double Gas In-
let Nozzles
61 - 0" dia. x 30' -
0" High, C.S. with 5
HP feed screw
Heat Exchangers
& Coolers
HE-1301
Cold H.E.
4,000 1 1/2" dia.
tubes, 10 BWG,
381
-------
ITEM NO.
Table K-l (continued)
SERVICE
He-1302
HE-1303
HE-1304
HE-1305
HE-1306
HE-1307
HE-1308
HE-1309
HE-1310
Hot H.E.
Weak Acid Coolers
Condensers
(Water Cooled)
Condensers
(Refrigeration)
Drying Tower
Cooler
Absorber Cooler
Product Acid Cooler
Superheater
Converter Gas Cooler
DESCRIPTION
C.S. 30' - 0" long.
Shell 13' - 6"
1,000 - 3" dia. tubes,
14 BW6 C.S., 30' - 0"
Long, Shell, 12-0"
dia. (Stainless Steel
Tubes)
Impervious Graphite
Tubes 7/8" I.D., Steel
Shells, 12,000 Sq. Ft.
Impervious Graphite
Tubes 7/8" I.D., Steel
Shells - 24,400 Sq. Ft.
S.S. Tubes - Steel
Shell - 16,700 Sq. Ft.
Chemetics - 500 Sq.
Ft. S.S., Anodically
Protected
Chemetics - 2,100 Sq.
Ft. S.S., Anodically
Protected
Chemetics - 200 Sq.
Ft. S.S., Anodically
Protected
3000 Sq. Ft.
2750 Sq. Ft.
Pump & Drives
P-1501A&B
P-1502A&B
P-1503A&B
Humidifying Tower
(2)
Washing Tower Pumps
(2)
Drying Tower Pump
3,600 GPM - 100' TDH
200 HP
5,500 GPM - 150' TDH
350 HP
Lewis Size 7, -
2,000 GPM
100 HP
382
-------
ITEM NO.
P-1504A&B
P-1505A&B
P-1506A&B
P-1507 A,B,
& C
Table K-l (continued)
SERVICE DESCRIPTION
Absorber Pump
Product & Drain
Neutralization Tank (2)
Cooling Water Pumps
Lewis Size 7, 2,000
GPM
100 HP
200 GPM, 120' TDH,
15 HP
200 GPM, 100 TDH,
10 HP
7,000 GPM, 130' TDH,
250 HP
Cooling Towers
CT-1701 Cooling Tower
13,000 GPM, 190 MM
BTU/Hr. (106.2 - 77°F.)
150 HP
Blowers
B-1801 A,B,
Main Gas Blowers
100,000 SCFM, 7.03
psi delta P, 4,500
HP Each
Tanks
TK-l901
TK-1902
TK-1903
TK-1904
TK-1905
Humidifying Tower
Pump Tank
Seal Pot
Drying Tower Pump Tank
Absorber Pump Tank
Neutralization Pit (Tank)
20' - 0" and 14' -
0" x 10 Deep
6' - 0" dia. x 5' •
0" Deep
22' - 0" dia. x 8'
deep Teflon & A.P.
bricklined
22' - 0" dia. x 8'
deep Teflon & A.P.
brick-lined
16' - 0" x 16' -
0" x 10' deep
383
-------
Table K-l (continued)
ITEM NO.
Reactors
R-2501
SERVICE
Converter
DESCRIPTION
40' - 0" dia. x 48'
0" High, 4 Catalyst
Beds
Separation Equipment
S-2801
S-2802
A, B, C, D,
E & F
S-2803
S-2804
S-2805
Gas Liquid Separator FRP - Integral in Duct
Electrostatic P
(Each 2 Series)
Drying Tower Demister
Electrostatic Precipitators 158 tubes x 17' -
0" dia. 3 Parallel =
6 Total
18' - 0" dia. - 2
Stage 1 - 18' dia. pad
& 1-16' dia. pad
Absorber Demister
Neutralization Tank Mixer 3 HP
York Type S,17' -
0" dia.
Stack
ST-2901 Stack
Refrigeration
Cooler
RC-3101, A,B.C. Water Chillers
RC - 3101, A,B, Water Chillers
& C
F-3101
Preheater Furnace
12' - 0" dia. x 149' -
0" High Self Standing-
Spoilers at Top 20'
York, Open
Turbopak - 3
Units (1500 Tons
Total)
York, Open
Turbopak - 3
Units (1,500 Tons
Total)
6' x 20'
Bricklined
384
-------
TABLE K-2
UTILITIES
REVERBERATORY FURNACE GAS
SULFURIC ACID PLANT
(100,000 SCFM GAS FLOW)
Sulfuric Acid Plant
REV = 0
9-30-76
Case I
Name
Lime Silo Feed Screw
Humidifying Tower Circ. Pump
(Spare)
Washing Tower Circulating Pump
(Spare)
Neutralization Tk. Pumps (Sump)
(Spare)
ESP - Seals Fans
Drying Tower Pump & Spare
Absorption Pump & Spare
Main Gas Blower
(Spare)
^t" A y*"f~ — ti n 1 iihi* Piimnc iMj^in RTnwpv*)
(Spare)
Drain (Product) Pump
(Spare)
Blower Room Exhauster
Utility Air Compressor
6 Mist Precipitators
Acid Plant (No Control Room,
No Lights, No Cooling Tower)
(A)
Normal Average
(Acid Plant)
Motor List
Connected
5
200
200
350
350
10
10
(3) 3/Each
100 (2)
100 (2)
4,500
4,500
in
1 U
10
15
15
2
20
10,606 HP
(3) 50 KVA/
Each
Connected
10,606 HP
(150 KVA)
Average
Operating
3
180
300
6
6 Total
70
75
3,700
12
1 1/2
5
4358.5 HP
90 KVA Total
Connected
4358.5 HP
(90 KVA)
385
-------
Table K-2 (continued)
Name
Water Cooling Tower
C.T. Well Pump (7,000 GPM)
C.T. Well Pump (7,000 GPM)
Spare
Cooling Tower Fan
Cooling Tower Fan
(B)
Connected
250
250
250
150
150
1,050 HP
Average
Operating
200
200
100
100
600 HP
Control Room '(Motors)
Instruments 16 x 1/8 = 2 11/2
Air Conditioning (Heating) 20 10
"7C5
22 HP 11.5 HP
Total Plant Lighting
All Lights 100 KW 60 KW
386
-------
K.5 GAS CLEANING AND CONDITIONING PLANT
• Process Gas Inlet Flue Connection
Gas after passing through a hot Cottrell precipitator, is
diverted by a new damper through a carbon steel externally insulated
duct into the humidifying tower, (see Appendix K-l, page K-39).
• Humidifying Tower, T-1101
The purpose of the humidifying tower is to saturate and cool
the dirty reverberatory gas before this gas enters a packed washing
tower. Gas flows into the bottom of the empty humidifying tower and
upwards against a series of spray nozzles circulating liquid counter-
current to the up flowing gas stream. The bottom of the tower has a
conical bottom which collects the recycled liquid which feeds an FRP
pump tank which in turn feeds the rubber lined recirculating pumps.
The liquor is then recycled back to the spray nozzles in the walls
and top dome of the humidifying tower. This tower is a vertical
carbon steel vessel, teflon line, polypropylene (PPE) lined and acid
proof brick lined. If the gas contains fluorine, the brick lining
will be carbon or graphite brick, otherwise if weak acid is used,
standard acid proof brick is the construction material of this tower.
The hot gas inlet nozzle of the tower will be externally jacketed
and cooled with cooling water.
• Washing Tower, T-1102
Construction of this tower usually is of fiber glass reinforced
plastic (FRP). The gas inlet nozzle is in the lower section of this
vertical tower and the tower contains approximately 10 feet of packing
consisting of 3-1/2 x 3-1/2 inch PPE Pall Rings. The top of the tower
has a series of spray nozzles which allow weak liquor or water to be
sprayed into the top of the tower and distributed uniformly over the
packing. The washing of the gas occurs as the gas flows up through
the packing and contracts the recycled cooled liquor flowing down over
the packing. The liquor is collected in the bottom of the tower and
is pumped by external rubber lined pumps through a series of carbate
coolers and recycled back to the top of the tower.
387
-------
• Acid Mist Precipitator (Electrostatic Precipitators)
(S-2802, A, B, C, D, E, and F)
The electrostatic acid mist precipitators will be the lead
tubuler type, with structural steel supports, and frames.
The top high tension insulator compartments will be provided
with air sweep seals. This equipment includes a total of three
complete high voltage electrical sets, necessary transformers,
rectifiers and automatic controls. A water spray flushing nozzle
system is included to permit the unit to be flushed out (washed)
and cleaned while in operation. The precipitators are arranged
three in parallel flow, followed in series by three more in
parallel flow, a total of six units. The overall operation and
clearance efficiency for acid mist and dust carried into the unit
is estimated in excess of 98% of the rated flow capacity of the
unit.
• Condensers (HE 1304, HE 1305)
Two series of sets of parallel condensers are installed to
lower the temperature of the gas to the required conditions in
order to achieve the correct ratio of water to S0? (water
balance). Design requirements have been calculated in Appendix
K-l, page K-69 and condenser design on page K-74. The first set
of parallel flow condensers are cooled with water recycled
continuously from the cooling tower system. Cooling water require-
ment calculation has been included in Appendix K-l, page K-82.
The second set of condensers are cooled with chilled refrigerated
water in closed circuit with the refrigeration unit
The condensers are constructed of carbon steel shells with
impervious graphite tube sheets and tubes. The top and bottom heads
or vestibule of these vertically mounted units are FRP (fiber-
glass reinforced plastic) and the lower section of the vestibule
contain a separating section to remove condensate leaving the
388
-------
system. The flow of the gas is down through the tubes. After the
first bank of parallel condensers the gas is collected and passed
by an FRP duct up to the top vestibule for down flow through the ":
second bank of parallel condensers which are water chilled using '*
refrigerated water. The design of the second bank of condensers
is similar to that of the bank used in using cooling water as the
cooling media.
The Allied-IHI Freon driven turbine system used to drive the
refrigeration system in this analysis has had considerable plant
experience by Ishikawajima-Harima Heavy Industries in Japan and
the Allied Chemical Corporation.
K.6 SULFURIC ACID PLANT EQUIPMENT
• Drying Tower, T-1303
The drying tower is fabricated of carbon steel plate with
minimal wall thicknesses of carbon steel as follows:
LOCATION CARBON STEEL THICKNESS. INCHES
Shell 3/8
Bottom 5/8
Top 3/8
The bottoms and sides of the tower are lined with teflon
sheet covering with a minimum of one layer acid proof brick over
the entire tower; however, the bottom section up to just above
the gas inlet nozzles will contain one additional layer of acid
proof brick. The packed section is approximately six feet deep
and consists of acid proof grid tile and special Intalock saddles
or Cascade Minirings. The packed section is supported on acid
proof brick arches. Acid distribution over the tower is accomplished
389
-------
using troughs and spouts fabricated of cast iron. One or more
spouts are provided for each square foot of cross sectional area
for the packed section. The drying tower is provided with a
required number of nozzles and manholes, and mounted on the top of
the tower is a two stage acid mist eliminator demister (S-2803).
The flat bottom of the vertical tower is supported on steel high
beam grillage which carries the load of the packed tower. The
grillage allows the dissipation of heat from the bottom of the
tower.
• Cold Heat Exchanger, HE-1301
The cold heat exchanger is a carbon steel vertical shell
mounted on grillage and the unit contains 4,000-1-1/2 inch tubes,
10 BWG - 30 feet long. The entire shell is 1/2 inch thick carbon
steel grillage and 1 inch thick tube sheets. The gas enters just
above the top tube sheet in the top vestibule and flows down through
the tubes. The cold gas is heated in the tube by the shell side
upflow hot converter exit gas. The entire inner surface of the heat
exchanger is aluminized. The external surface of the heat exchanger
has 2 inches of rockwool blanket insulation and a weather seal to
protect it from the weather.
• Hot Heat Exchangers, HE-1302
The gas which has discharged from the blower and passed
through the cold heat exchanger for initial preheating, is passed
out of the bottom vestibule of the cold heat exchanger into the
lower shell side of the hot heat exchanger. The gas flows up
around the tubes of the hot heat exchanger discharging at the
top of the vertical heat exchanger just under the tube sheet. The
exiting temperature of the gas leaving the hot heat exchanger is
sufficient to reach the catalytic conversion temperature required
in the converter. The hot heat exchanger is a vertical carbon
steel vessel containing stainless steel tubes with the top and
bottom tube sheets stainless steel. This unit contains 1,000,
3 inch diameter tubes, 14 BWG-24 feet long between the tube sheets,
390
-------
with 6 feet of extension into the lower vestibule to act as dust
knockouts. The bottom vestibule of this heat exchanger is a
conical bottom with a center collection nozzle and double dump
valves to allow any trapped dust to be removed from the system.
The entire inner surface of this heat exchanger is aluminized.
The top vestibule is lined with 2-1/4 or 2-1/2 inches of insulated
refractory to protect the vessel while it is used as a preheater
(during startup). The entire outer surface of this heat exchanger
is covered with 2-1/2 to 3 inches to rockwool blanket insulation
and is normally installed inside of the building; however, if it
is installed outside, it will also contain external weather seal.
• Converter R-2501
The converter is a vertical carbon steel vessel with a dish top
head and flat bottom. The converter is provided with cast iron
grillage grates to support the catalyst. The catalyst grates are
supported by cast iron pipe columns to support a triangle grid
grate section for heat catalyst bed. Each catalyst bed contains
two inches of quartz above and below each catalyst bed to maintain
and stabilize the catalyst.
Gas, leaving the hot heat exchanger HE-1302, enters the top
dish head of the converter and flows down through the first catalyst
bed passing to the outside, to an external cooler, and then return-
ing to the second catalyst bed. There is a minimum of three
catalyst beds in this plant. The internal surfaces of this converter
vessel are aluminized and covered externally with rockwool blanket
insulation and weather seal. Each catalyst bed contains sufficient
catalyst to convert the gas to the required sulfur trioxide content.
Each catalyst bed has two thermocouples to indicate and record bed
temperatures and also contains pressure connections for taking the
pressure drop across each catalyst bed.
• Catalyst
The catalyst is a vanadium pentoxide (V^O,-) hard non-fusing
low ignition type catalyst and is installed in the converter. This
391
-------
catalyst meets the conversion conditions required at the flow and
the gas strength conditions.
Twelve cases of catalyst information were computed by Catalyst
and Chemicals, Inc. These various catalyst cases were computer run
showing various gas strengths, oxygen contents and quench between
catalyst beds. The catalyst data was selected on the basis of
0.5, 1.0, and 1.5% SOo gas with varying oxygen compositions of
11% and 14%. Also separate cases were used for by-passing cold
gas around the first catalyst bed and using it for quench cooling.
In each case it can be seen that the catalyst loading does not
substantially affect the capital cost of the plant. The curve in
Figure K-3 shows that the differential price of investment from
one case to the next is only approximately a maximum of 25,000
dollars. This is based on using 16,000 differential liters of
catalyst from one case to the next. Therefore, it can be seen
that the effects of using quench gas in by-pass, or using oxygen
for dilution, or for quenching has substantially very little effect
on the capital cost of the plant. The design is made for a total
of three catalyst beds to handle large fluctuations of gas.
• Absorption Tower, T-1104
The design of the absorption tower is similar to that used
in the design of the drying tower. The difference will be the
demister located in the top of the tower which is a high efficiency
absorption tower demister (S-2804). This demister is a York type S
unit.
• Pump Tanks, TK-1904
The pump tanks are constructed of 3/8 inch carbon steel plate.
The tops and bottoms are flat. The bottoms and sides of the tank
are lined with three or four mil thickness of teflon sheeting and
a minimum of 3-3/4 inch course of acid proof brick. Each tank is
supported vertically on steel grillage beams. The top of each tank
contains two rectangular flanged nozzles to hold acid circul-
ating pumps. One pump circulates acid continuously from the pump
392
-------
10
«3
CD
O
c
CD
3
cr
o:
z
it
O O
c c
O) OJ
3 3
o- o-
CM
O
CM +
O
-C
+ o
-c CD
° £
ii
n
ii
CM CM
o
o
O CM
U," °
£ "g
•to- o-
< O
CM Z
O
CD
^ S.
•— O
O O
r— LO
O
ur>
o
cr
/
^
m
\
\
^
0
0
o
««
iO
V
\
v \
^N.^ CM CM
T«>eyz:
o
o
CM
\
^
5.
o.
in
CD
T3
10
O
4J
to
. o
1 1
S
1 1
E i
o
LO
to
s-
o
CD
S-
CD
1
tn " o
. . (/) o
•— CM S- O
O O) •>
CO -P O
o
o
o
t/1
Q)
o
o
o
tO II
O
X CO
to to
to s-
o o
•M
X
CD
c
0) CD
CD
CD
I
to
•o
CD
JO
fO
o
oo
to
CD
en
1C
4->
fB
0
o
(C
Q
CO
CD
S-
1SA1V1V3 1V101
393
-------
tank through the coolers and over the tower and one pump is instal-
led as a spare. The pump tanks each has manholes in the top and
drain nozzles as required for the necessary piping, and instru-
mentation connections.
• Drying Tower Cooler, HE-1306
These coolers are stainless steel horizontally mounted units.
The acid flows through the shell side diverted by baffles across
the tubes. The cooling flows through the tubes. Each unit is in-
dividually anodically protected. The purpose of the acid coolers
is to cool the acid which has increased in temperature as a result
of flowing through the tower, and mixing with cross-transfer acid.
The acid is pumped continuously from the pump tank through the
cooler and over the drying tower.
• Absorption Cooler, HE-1307
The design of the absorption tower cooler is similar to that
of the drying tower, however cooling is accomplished with Freon.
• Product Acid Cooler, HE-1308
This unit is considerably smaller but is similar in design to
the above coolers.
• Converter Gas Cooler
The converter gas cooler is provided to compensate for possible
maximum ranges of SCL entering the acid plant system from the rever-
beratory furnace. Normal operation in the 0.5 to 1.5% S02 concen-
tration will not require this unit.
Design calculations are included in Appendix K-l, page K-84.
• Pumps and Drives
The pumps are listed on the equipment list, Table K-l. The
pumps in the gas cleaning portion of the plant (front end) handling
weak acidic liquor and solids are horizontal rubber lined slurry
type pumps. The strong acid pumps used in the system are Lewis
vertically submerged pumps that hang down into the pump tank.
394
-------
Cooling water pumps are vertically hung pumps mounted in the cold-
well. All the horizontal pumps have horizontal motors and vertical
pumps have vertical motors.
K.7 EFFECT OF PLANT LAYOUT AND EQUIPMENT POSITION
In comparing Case I (Onahama) and Case II (Browder) it can
be seen that one of the major differences results because in Case
II a mixture of reverberatory boiler exit and reverberatory gases
are mixed in the jog damper to give the required temperature enter-
ing the top vestibule of the hot heat exchanger. This scheme per-
mits the use of the hot gas from the reverberatory to replace a
fired furnace which has been used at Onahama (Case I), where weak
gases have been encountered and autothermal conditions have not
been achieved in sulfuric acid plants. The distance that the sul-
furic acid plant can be removed from the reverberatory furnace is
critical.
By reviewing known smelters, it can be seen that usually the
reverberatory furnace building contains boilers and in these
buildings there is usually sufficient head space adjacent to or
above the reverberatory boilers to install gas-to-gas exchangers.
The gas-to-gas exchangers normally would not occupy a circular
space of more than 6 - 7-1/2 meters (20-24 feet) in diameter even
for larger plants. These heat exchangers can be installed in a
vertical position any place in a building adjacent to the reverbera-
tory furnace or the reverberatory boiler.
In the past, sulfuric acid plants had been built essentially
at one level, in which all of the equipment is installed at the
ground level location. This is not necessary. Sulfuric acid
plants could be built on more than one level, having the converter,
cold heat exchanger and hot heat exchangers mounted adjacent to a
reverberatory building or within the building itself and above the
ground level. Location of the rest of the sulfuric acid plant is
not critical and may be remotely placed. It is essential, however,
395
-------
that the acid plant converter hot heat exchanger and cold heat
exchanger be adjacent to the reverberatory furnace when energy from
reverberatory off gas is used. A typical case might be taking
reverberatory furnace off gas from the exit of the Cottrell precipi-
tator, further cleaning and humidifying it by passing through a wash
tower, demisters, and condensers, and then ducting it to a drying
tower adjacent to the absorption tower. The drying tower, demisters
and blowers could be somewhat removed from the reverberatory furnace.
The gas could then be passed through a duct to a cold heat exchanger,
hot heat exchanger and the converter in an area in the reverberatory
building. The gas leaving the cold ;heat exchanger could be re-piped
some distance back into the absorption tower and then vented to the
atmosphere through a stack.
If this scheme, using the converter hot and cold heat exchanger
adjacent to the reverberatory furnace, is not used and these three
components are isolated from the reverberatory building, then the
system used at Onahama would apply, i.e., the hot heat exchanger
would have to be fired with an auxilliary combustion furnace
supplying the heat for the final pre-heating of the gas going to
the converter. In each and every plant, it is not known if the
flow scheme in Case II could definitely be used. Each would have
to be reviewed on a case by case basis to determine if the converter,
hot, and cold heat exchangers could be located adjacent to or in-
side of the reverberatory building. There is even a possibility
where low head room buildings exist that extensions of the column
line could be made and the converter and heat exchangers located
on top or above the building.
Plan views of Case I (Onahama) and Case II (Browder), to
scale are shown in Figures K-3a and 3b.
K.8 COST
While it is not the object of this project to include detailed
cost estimates, a brief comment on installation schedule and system
396
-------
«• =
Si?
iiiiitiihffl!
co
o>
Z3
Dl
397
-------
co
J_
3
cn
•i—
u_
398
-------
costs is included here to provide a comparative review pointing out
design/cost relationship.
Project construction schedule of 28 months is shown in Figure
K-4. Capital costs vs. gas flow for various SCL concentrations is
3
shown in Figure K-5. It can be expected that this 170,000 Mm /hr
(100,000 SCFM) plant will cost approximately $22,000,000 and pro-
duce slightly less than 200 tons/day sulfuric acid.
Figure K-5 also shows comparable costs of conventional auto-
thermal metallurgical single contact acid plants operating at 4%
and 9% S02 gas. The trend in all cases is for costs to decrease as
S02 concentration increases. Also the Browder system costs are
considerably less than conventional single contact systems. The
major reason for these cost variations is primarily because heat
exchanger requirements are affected by the amount of energy available
for heating the gas to the ignition temperature of 440°C (830°F).
The Browder process provides a large LMTD thereby minimizing heat
exchange surface area which can cost $15.00 per square foot.
K.9 ENERGY COMPARISON
The following results indicate the additional energy in terms
of motor horse power and fuel oil required for Case I when compared
to Case II. Additional details are given in Appendix K-l, page K-86.
Connected
Average Operating
Electrically Driven Units
Chilled Water Refrigera-
tion Units
Preheater Blower
(Add to Case I)
Oil-Fired Preheater
Fuel Oil
1,750 HP
250 HP
2,000 HP
410 gal/hr
1,625 HP
235 HP
1,860 HP
399
-------
0
c"
f-
c*
3
it*
0*
•^t
c<
A
c1
C<
C*
?
c
c-
O;
O
|V-
«
O
^
0
N
a
o
u22 *J a
IA*U i»n M r*
8* % 3 1
§1 §a| i §
aifi OK>£! a o
•a
OJ
o
oo
o
0)
•r-J
O
a_
0)
s-
400
-------
PLANT COST
FURNACE: 5&S SULPUFUC ACID PLANTS
O.S"- ?,S% 502. - ,5/NSlE CONTACT
£55 5 "at? 7r~_V3_
H..- .T|~-J .-.-_.-.- -:. -^-^.-..--^^--i-.|_,.H-
--M- . 'rH '
zr^-.j-i.:!- -^r^'pr^ptivv::-:^::
"l ' "' ~""~ h-r-T-r T:-M
:t:|—- t--^-! J-
i I — -i —\^ :-
TIM I. BROWDER COMPANY
Post Office Box 8473
San Manno, CA 91103, U.SA
Telephone: (213) 287-7709
_Rt=VER.BERATDRY FLOW
- CLEANED GAS ( DR.T.)
Figure K-5
401
-------
APPENDIX L
LIME/LIMESTONE GYPSUM SO? CONTROL SYSTEM FOR REVERBERATORY
FURNACE OFFGASES AT THE ONAHAMA COPPER SMELTER
The development of the lime/limestone gypsum S02 control
system has been carried out over a period of approximately 20 years
at the Hiroshima Technical Institute which is one of three R & D
facilities operated by Mitsubishi in Japan. ' Their work
was initiated by reviewing scrubbing for all pollutants including
particulates and S02«
As the main interest turned to S02 removal, attention was cen-
tered on the control of a sintering plant in the year 1957. Pilot
plant tests at approximatley 3,000 Nnr/hr were conducted for a
period of three months. Pyrite sinter was used with ammonia as the
feed materials to obtain ammonium sulfate as a product. The main
problem at that time was converting ammonium sulfite to ammonium
sulfate because of the demand for ammonium sulfate fertilizer. The
process was intended to be installed at an industrial complex where
there was an iron and steel plant within 6 kilometers of a chemical
plant which made acid and fertilizer. A pipe line was planned be-
tween the two plants. This development work occurred during the
period from 1957 to 1960. However, when the market dropped for
fertilizer, the project was dropped after it had been carried
through the design stage.
In 1962 pilot tests were conducted at the Kansai Electric Power
Station at Amagasaki City near Osaka to select the absorbent and
type of scrubber. The tests were conducted at the No. 3 power sta-
tion which had an oil fired boiler with a gas flow of 5,400 Nm^/hr
(3,350 scfm). During this test program, ammonia, MgO, lime, lime-
stone and red mud waste from an aluminum plant were considered as
402
-------
possible absorbents. Bayer process aluminum plants waste consisted
of approximately 20 percent sodium along with aluminum, iron, titanium,
silicon, and sulfur. Sodium aluminum silicate was obtained as a com-
plex salt from the red mud by first separating out the iron oxide pro-
ducing a white product. Using MgO to get sulfuric acid directly was
not too successful because the best they could do at that time was
to obtain 50 percent concentration.
It was determined, after the 1962 test program, that the lime/
limestone system was the best because it was simple, the lime was
readily available in Japan and the demand for gypsum was increasing.
The requirement for gypsum in Japan is approximately one-third that
of the U.S.
During the period between 1968 and 1972, the Hiroshima Techni-
cal Institute carried out a $5 million program to investigate the
scaling problem which was occurring in the scrubber and the mist
eliminator of the pilot plant lime/limestone system. The pilot
tests carried out in 1962 had not uncovered this problem because of
their relatively short duration, minimizing scale buildup. Funda-
mental research at the bench scale, as well as pilot scale of 2,000
Nrrr/hr (1,250 scfm), was included. Figure L-l is a photograph of
the pilot plant at the Hiroshima Technical Institute.
The sulfite scale was easy to remove but the sulfate was very
hard. It was found that preparation and operating conditions of
the absorbent were critical and required the proper range of pH,
temperature, concentration, construction material of the scrubber,
L/G, uniformity of the stream, and prevention of carry-over of mist
to the eliminator. Their scaling work was reported at the EPA New
Orleans meeting in 1971 including the development of the seed
crystal technique.
Research and development work on a dry activated manganese
oxide system during the period between 1964 and 1970 was also
conducted. Pilot test work at the 1 MW, 55 MW and 110 MW levels
was conducted. The 55 megawatt system was sponsored by MITI.
403
-------
c
ro
D_
O
D-
in
a.
O)
C
o
4J
CO
0)
0)
01
404
-------
This work was then followed by a 220 megawatt installation at
the Yokaichi Power Station on an oil fired system. The development
was completed satisfactorily from a technical standpoint but: did
not go commercial because it was not economical.
Solid absorbents were also investigated. Handling the solid
material absorbent was more difficult than the liquid, and it was
also discovered that the wet system would take a higher S02 con-
centration. Furthermore, the byproduct was ammonium sulfate which
is currently not being used very extensively in Japan as a
fertilizer, having been replaced by urea. The solid or dry systems
were reported at the 7th World Congress in Mexico City and also at
the Los Angeles Chemical Engineering Society, at the present time
Mitsubishi does not recommend dry systems particularly for rever-
beratory furnaces because the wet system can better handle the
higher S02 concentrations even though the dry system does not re-
quire water or stack reheat. However, with the reverberatory fur-
nace gases, it is not expected that large quantities of water will
be required.
There was some bench scale testing at S02 concentrations of
20,000 ppm before the system for the Onahama smelter reverberatory
furnace was constructed. However, no pilot plant work was done
using the higher concentration compared to utility concentration.
405
-------
APPENDIX M
COMPONENTS INCLUDED IN THE ONAHAMA
LIME/LIMESTONE GYPSUM CONTROL SYSTEM
Washing Tower
The initial washing tower includes inlet ducting with internal
sprays leading to a cylindrical vessel with a horizontal axis.
Sprays are located along the top side. This unit is constructed of
lead lined brick to provide resistance to corrosion from the low
pH (1-2) water resulting from absorption of a slight amount of S02
and SO,. Internal baffles are placed vertically to the axis at
several intervals to cause mixing of the gas and spray by disturbing
the flow pattern of the gas as it passes through the chamber. The
gas leaving the washing tower is near saturation temperature. The
pressure drop of this unit is approximately 70 mm FLO.
Gas'Coolers
The five gas coolers are sea water indirectly cooled heat
exchangers.* These coolers drop the gypsum system inlet gas temp-
erature in preparation from entry into the absorption towers.
3
Cooling water required is 930 m /hour.
Absorption Towers
Two plastic grid packed absorbers in series are used. Their
external shell is made of steel plates lined with synthetic rubber.
The grid packing has large openings to minimize plugging and pres-
sure drop.
The grid packed tower was adopted for the absorption units
*Ten coolers are commonly used when both the gypsum plant and the
MgO plant are in operation.
406
-------
because its mass transfer coefficient for SCL absorption is high,
its pressure loss low, the structure simple, its performance
stability against load variation is large and in addition it can
be scaled up or down easily.
The lower the pH of the absorbing solution the higher the
Ca utilization. Conversely the lower the pH of the absorbing solu-
tion the higher will be the S02 concentration in the offgas. To
lower S02 concentration at the outlet it is necessary to increase
pH considerably. Therefore the rate of SCL absorption and Ca
utilization can be satisfied only when a multistage absorber is
used. The absorbers are designed so that the pH of the circulating
solution in the #2 absorber is kept at 7, that of the absorbing
solution to be taken out of #1 is decreased to 4 or less, which
is also desirable for operation in the oxidizing section.
It is important to obtain an internal configuration producing
an operating condition that prevents the slurry from stagnating and
ensuring that all the inner surfaces of the absorber are sufficient-
ly sprayed or covered with absorbing solution to avoid any pH
variation which could occur locally which may provide a condition
for scale generation. The addition of gypsum seed contributed to
lowering of super saturated concentration of gypsum and makes a
sacrificial surface to crystal precipitation. The piping systems
are designed such that a suitable flow rate is maintained to prevent
sedimentation of particles in the slurry as well as preventing
clogging by installing strainers at suitable places.
Mist Eliminator
The mist eliminator located in the gas circuit downstream
of the absorbers is of the Chevron type. It is internally coated
with synthetic resin and made of steel.
Oxidizing Section
Three oxidizing towers are also internally coated with
synthetic resin. Figure M-l shows a schematic diagram of the rotary
atomizers used to promote oxidation inside these towers. Absorbing
407
-------
solution taken out of the absorbing section is fed to the oxidizing
tower where it is oxidized by air at pH 3.0 to 4.5 according to the
follow equation:
CaS03.l/2 H20 + 1/2 ^ solution" CaS04.2H20
It is believed that the reaction of calcium sulfite advances
through the medium of bisulfite ion. This reaction mechanism has
also been proven by a correlation of the mass transfer coefficient
of the oxidizing tower, pH value and bisulfite ion concentration.
The oxidation reaction depends on dissolution and diffusion
of oxygen into the slurry. Decrease in air bubble diameter and
increase in the amount of gas in the solution were the objectives
of the rotary atomizer developed. Figure M-l also shows the tower used
for atomizer development. This atomizer tears with a shearing force
caused by rotation between the surrounding solution and air layer.
The air layer formed on the external surface of the cylinder which
rotates at about 500 to 1000 rpm generates fine air bubbles of
0.1 to 1.0 mm.
The oxidizing towers are indirectly cooled with the cooling
water going to a cooling tower for reuse in this circuit. The
speed of oxidation is controlled by the pressure of the oxidizer.
The oxidizer tower was designed for 70 psi and the actual nominal
2
operating pressure of the air is between 40 and 45 psi (3 kg/cm ).
Slurry Section
Slurry flowing out of the oxidizing tower is concentrated in
a thickener. The underflow slurry from the thickener is treated by
a basket type centrifugal separator to separate the dihydrate crystal
of gypsum. The percentage of water content in the centrifuge
material is 5-10 percent. Overflow liquid from the thickener
and filtrate from the centrifuge is used for preparation of lime
slurry and for other purposes.
Centrifuges
A vertical basket type centrifuge design is used. There are
408
-------
fin* •» kvbMt
Figure M-l. Rotary Atomizer Schematic, Test Tower and Spray Pattern
409
-------
14 centrifuges all operating continuously with no spares.
The smaller the crystal size the more difficult it is to
separate in the centrifuge. Also small size tends to produce vibra-
tion.
Excessive amounts of silica in the gypsum tend to plug the
centrifuge screens.
Slaking Section
Quick lime is fed to the slaker from a hopper and is mixed with
liquid from the thickener overflow. From the slaker, the material
goes to a ball mill where it is ground and passed through a liquid
cyclone. Underflow from the liquid cyclone is recycled to the ball
mill and overflow goes to the milk holder for use in the lime feed
system to the absorber.
410
-------
APPENDIX N
WATER BALANCE GYPSUM SYSTEM
This balance is for the lime-limestone-gypsum system being
used at the Onahama copper smelter in Japan. Information received
from the Onahama smelter indicated the water flowrates at different
process locations. This information has been utilized to obtain
an approximate water balance shown in Figure M-l.
Washing Tower and Gas Cooling Section
a) Quantity of water in flue gas
Flue gas rate at Onahama - 2,000 Nm3/min = 70,600 SCFM
Amount of water in the flue gas = 0.116 Ibs/lb of dry gas
Assuming the density of the flue gas
= 0.0808 lbs/ft3 (Density of air)
/.Mass rate of flue gas = 0.0808 x 70,600
= 5,704 Ibs/min
Quantity of water in the flue gas = 5,704 Ibs/min
y 0.116 Ibs H20
* 1.116 Ibs gas
= 593 Ibs/min
.'.Quantity of dry gas = 5,704 = 593 = 5,111 Ibs/min
b) Quantity of water in gas at outlet of Washing Tower
Assuming the gas is saturated with water vapor at the outlet
of the washing tower and is at 60°C:
Humidity ratio = 0 623 x Partial Pressure of Water Vapor (Pv)
Humiany ratio u.b^j x Absolute Pressure of Vapor (Pa)
Pv at 60°C from steam tables = 2.890 Ibs/sq in
Pa = 14.696 - 2.890 = 11.806 Ibs/sq in
411
-------
Humidity ratio = 0.623 x - 0.1525 jjg »?° gas
.'.Water in gas at the outlet ,. ,,, Ibs dry gas Y n -,„,- Ibs H?0
of the washing tower °'"' min OD lbs dry gas
.'.Quantity of water added to
flue gas in the washing = 785-593
tower
= 192 lbs.
c) Quantity of water in gas at outlet of Gas Coolers
Quantity of dry gas = 5,111 lbs dry gas/min
Assuming the gas to be saturated at the outlet of the gas
coolers and at 45°C:
Humidity ratio = 0.623 p^-
-L39I= 0.0651 lbs H?°
.'.Quantity of water in gas at the outlet of the gas coolers
= 5,111 x 0.0651
= 333 Ibs of H20/min
d) Water balance
Therefore, in the Washing Tower and Gas Cooling System
785-332 = 453 lbs of H^O/min is removed in the gas cooling system,
whereas 192 lbs of HLO/min is added in the washing tower.
Hence, in the Washing Tower and Gas Cooling Section, 260 Ibs
of HgO/min excess water is removed from the flue gas and 324 lbs
of H0/min remains in the gas to the Absorbers.
The Absorber Slurry System
Water Outlet Sources
a) Water going out to the ocean from the neutralization tank
= 6,636 Ibs/min (4,336 M3/day)
b) Some water is removed from the system by the mass of gypsum
being produced by two processes; i) surface moisture, and, ii) water
412
-------
of crystallization.
Assuming the gypsum contains 10 percent surface moisture:
.'.Water going out of the system as:
(i) surface moisture with gypsum = 450 x 0.1 = 45 TPD
= 62.5 Ibs of H20/min
(ii) Mass of water of cyrstallization with gypsum can be
obtained from the chemical formula of gypsum
Molecular weight of CaSo4 .2H20 = 172.17
where molecular weight of 2H20 = 36
oe
.'.Water of crystallization in gypsum = ,72 17 x (450 ~45) TPD
= 84.7 TPD
= 118 Ibs/min
.'.Total mass of water in gypsum = 118 +62
= 180 Ibs/min
c) The moisture in the exhaust gases through the main stack is
another source of moisture loss.
Assuming the gas is saturated with water vapor at the outlet
of the main stack and is at 45°C
Pv
.'.The Humidity ratio = 0.623 ^
- n fi?? 1.391 = 0.0651 Ibs H20
u 13.305 Ibs dry gas
.'.Quantity of water in gas at the outlet of the main stack
= 5,111 x 0.0651
= 333 Ibs of H20/min
d) Total water outlet from system = 6,636 + 180 + 333
= 7,149 Ib/min
Water Inlet Sources
a) Water in flue gas at inlet to Absorbers = 333 Ibs/min
413
-------
b) Water inlet due to pump seal leakage = 3,673 Ibs/min
(2,400 M3/day)
o
c) Water from atomizer leaks = 459 Ibs/min (300 M /day)
d) Water from TCA spill gas system = 1,613 Ibs/min (1,054 M /day)
e) Water from Miscellaneous Sources being input to the neutral-
ization tank:
= 1,071 Ibs/min (700 m3/day)
f) Total water inlet to system 7,149 Ibs/min
/.Water inlet (7,149 Ibs/min) = Water outlet (7,149 Ibs/min)
414
-------
APPENDIX 0
MAGNESIUM OXIDE SYSTEM COMPONENT AND COST CONSIDERATIONS
The TCA absorber is 26.5 meters high and 4 meters in diameter.
The rotary calciner is 52 meters long and 3.4 meters in diameter.
Each of the two dryers used contains 900 sq. meters of steam drying
surface area.
3
The blower is designed for 2200 m /minute at 900 KW. One
spare blower is also available. The slaker is 35 m and the ball
mill is 2 meters by 4 meters.
The cost of operation of the MgO system is higher than the
gypsum system but is still in the range of four to five cents per
pound of copper, assuming no credit for product sales. The MgO
system would have an advantage over the gypsum system in terms of
the operating cost providing that there is an existing acid plant
to treat the S02 gas from the MgO plant and the acid credit is
reasonable. Capital costs of both systems are the same and present-
ly considered to be $7,500,000 for the 1500 Nm /minute size system
at Onahama.
415
-------
APPENDIX P
CHEMISTRY INVOLVED IN MAGNESIUM OXIDE SYSTEM
The simplified chemical reactions taking place in the process
may be written as follows:
Absorber:
MgO + S02 + 3H20 - ^ MgS03 • 3H20 (very little formed)
MgO + S02 + 6H20 - +• MgS03 • 6H20
MgS03 + %02 + 7H2 0 - +> MgS04 • 7H20
Dryer:
MgS03 • 3H20 heat » MgS03 + 3H20
MgS03 • 6H20 heat >
MgS04 • 7H20 - - ^ MgS0
Calciner:
MgS03 heat» MgO + S02
MgS04 + %C heat •> MgO + S02
Input Material and Preparation
The magnesium hydroxide is purchased as a liquid containing
416
-------
20-30 percent solids of Mg(OH)2. Analysis of this material is:
• CaO 0.97%
• S04 0.30%
• MgO 17.68%
• H?0 remainder
417
-------
APPENDIX Q
WATER BALANCE IN THE MgO SYSTEM
Washing Tower and Gas Cooling Section
a) Flue gas
Flue gas flow rate = 1,500 Nm3/min = 52,970 SCFM
Humidity ratio of flue gas = 0.116 Ibs of H20/lb of dry gas
Density of flue gas = 0.0808 lbs/ft3 (Density of air)
/.Mass rate of flue gas = 52,970 SCFM x 0.0808
= 4,280 Ibs/min
.'.Quantity of water in the flue gas
. 4.2SO ,bs/»1n x 0.1J6 jbs H,0
= 445.87 Ibs/min
.'.Quantity of dry gas = 4,280 -445
= 3,835 Ibs/min
b) Washing Tower
Assuming that the gas is saturated with water vapor at the
outlet of the washing tower and is at 60°C
Humidity ratio of gas at outlet
Pv (Partial pressure of vapor)
Pa (Absolute partial pressure)
= 0.623 x ^"gog = 0.1534 = .1525
.'.Water in gas at the outlet of the washing tower
= 3,835 x 0.1525 = 585 Ibs H20/min
418
-------
.'.Quantity of water added to the flue gas in the washing tower
= 585 -445 = 140 Ibs/min
c) Gas coolers
Assuming that the gas is saturated with water vapor at the
outlet of the gas coolers and is at 45°C:
Humidity ratio of gas at
outlet of the gas coolers = 0.623 x 1.391/13.305
= o 0651 1bs H2°
u'ubbl IDS dry gas
.'.The total quantity of water at exit of the gas coolers
= °'0651 bs d gas x 3'835 1bs/min
= 250 Ibs H20/min
d) Water balance
Water inlet to washing tower in flue gas = 445 Ibs/min
Water outlet at gas coolers exit = 250 Ibs/min
.'.Total water lost by flue gas in the washing tower and gas
cooling section is
= 445 -250 = 195 Ibs/min
The Absorber Slurry System
a) Water in Make up Slurry
The composition of the make up slurry is given as follows:
CaO = 0.97% MgO = 17.68%
S04 = 0.3% H20 = 81.05%
Also, quantity of Mg (OH)2 being input is given: 1,128 Ibs/hr
= 18.8 Ibs/min
40
.'.MgO being input = 18.8 x Jg- = 12.97 Ibs/min
.'.Quantity of H20 in slurry = 12.97 x ^=|
= 59.46 Ibs/min
419
-------
b) Water loss in dryer
To determine the quantity of water evaporated in the dryer,
the following assumptions were made:
1. The cake going to the dryer is assumed to be MgSO.,. GhLO
since the MgO and MgSCL. 7HLO in the cake are negligible
compared to MgSO,. 6hLO.
2. The quantity of MgS(L. 6HLO is calculated from the reaction
between MgO and SCL.
3. It is assumed that the cake contains 10% surface moisture.
MgS03 is formed in the absorber by the following reaction.
MgO + S02 - *> MgS03
40 + 64 - *> 104
It is also known that:
3
Volume flow of gas = 1,500 Nm /min
Volume flow of S02 = 1,500 x 0.026 = 39 Nm3/min = 1,377 SCFM
.'.Mass flow of S02 = 1,377 scfm x 0.16867 lbs/ft3 = 232.3 Ibs/min
Ratio of Mg(PH)2 makeup to S02 absorbed = 0.08 kg Mg(OH)2/kg
of S02
We know the quantity of SOp in the flue gas. Assuming 100%
S02 absorption, quantity of MgO used is equal to:
232.3
64
x 40 = 145.2 Ibs/min
?
.'.Quantity of MgS03 produced ^Q x 104
= 377.5 Ibs/min
Amount of water of crystallization in MgSCL. 6HLO
377.5
104
x 108 = 392.0 Ibs/min
420
-------
.'.Total amount of MgSO,. 60 produced = sum of MgSO., produced
plus water of crystallization
= 377.5 +392.0 = 769.5 Ibs/min
.'.The total quantity of the cake going through the dryer
= 769.5/0.90 = 855.0 Ibs/min
All the surface moisture and water of crystallization is driven
off in the dryer
.'.The quantity of water lost from the dryer
= 855.0 =377.5 = 477.5 Ibs/min
c) Water Balance
The quantity of water coming into the system from the pump
seals is given as 11 T/hr = 411 Ibs/min.
The quantity of slurry purged = 10 T/day 15.6 Ibs/min. Assum-
ing 90 percent of the slurry purged is water, the quantity of water
purged = 15.6 x 0.9 = 14.04 Ibs/min.
.'.Water coming into the system = water through the pump seals
+ water in makeup slurry =411 + 59 = 470 Ibs/min
Water coming out of the system = water lost from the dryer +
water lost in the purged slurry = 477.5 + 14.04 = 492 Ibs/min.
Note: the overall water balance indicates more water leaving the
system than being charged to the system. The difference can be
attributed to combustion products, miscellaneous sources of water
of which data were not available and measurement errors.
421
-------
APPENDIX R
PROCESS CHEMISTRY FOR THE CITRATE PROCESS62
Absorption of S02 in aqueous solution is pH dependent,
increasing at higher pH. Dissolution of S02 forms bisulfite ion
with resultant decrease in pH by the following reaction:
However, by incorporating a buffering agent in the solution
to inhibit pH decrease, high S02 loading and substantially
complete S02 removal from the waste gases can be attained. This
citrate ion performs this buffering action by the following
reaction:
The chemistry for the production of sulfur and regeneration
of absorbent by reacting H/>S with the S02 in the aqueous solution
is complex, but the overall reaction is as follows:
Hydrogen sulfide for regenerating the absorbent and pre-
cipitating elemental sulfur can be produced by reacting sulfur
with methane and steam by the following reaction:
422
-------
APPENDIX S
PROCESS CHEMISTRY FOR FLAKT CITRATE PROCESS63
The chemical reactions in the absorption stripping operation
follows the simplified reaction scheme given below.
S02 (g)^Z±S02 (aq) (1)
(2)
(2-n)
(3)
Ci denotes the citrate ion, with n = 0, 1 or 2. The forward
reactions take place during absorption, and the reverse reactions
during stripping.
Absorption of S02 in aqueous solution is pH dependent,
increasing at higher pH. Because dissolution of S02 forms
bisulfite (HS03~ ) ion with the resultant decrease in pH by the
reaction 2, the absorption of S02 in aqueous solution is self-
limiting. However, by incorporating a buffering agent in the
solution to inhibit pH decreasd (remove the hydrogen ions
formed in reaction 2), high S02 loadings and substantially complete
S02 removal can be attained. In the citrate process this is
accomplished by the buffering action of the various citrate species
by the reaction 3.
The buffering capacity is naturally dependent on the concen-
tration of citric acid and sodium hydroxide and the relation
between them. In the Flakt-Boliden citrate process, the concen-
tration of sodium hydroxide is between once and twice that of
citric acid. (The relationship chosen is dependent upon the raw
gas composition). In most cases this results in the absorbent pH
of 4 to 5. 54
423
-------
APPENDIX T
PROCESS CHEMISTRY FOR THE COMMINCO AMMONIA S02 CONTROL SYSTEM
The process consists of absorbing the S02 from the flue gas
in aqua ammonia, forming a solution that is essentially ammonium
bisulfite according to the following relations:
2NH4OH + S02 - ^(NH4)2S03 + H20
(NH4)2 S03 + S02 + H20 - ^2NH4HS03
In addition to ammonium bisulfite, some ammonium sulfate
is also formed primarily by the reaction:
(NH4)2S03 + 1/2
Since the aim of the L.C. process is to make ammonium
sulfate to be used as a fertilizer, the ammonium bisulfate com
bines with sulfuric acid to produce ammonium sulfate by the
reaction:
2NH4HS03
(NH4)2S03 + H2S04 - MNH4)2S04
424
-------
APPENDIX U
PROCESS CHEMISTRY FOR WELLMAN-LORD SYSTEM64
The process is based on the chemistry of the sodium sulfite/
bisulfite system. Flue gas containing SOp is scrubbed with a
solution consisting of soluble Na2S03, NaHS03> and Na,,S04. The
S02 reacts with sodium sulfite to form sodium bisulfite according
to the following reaction.
S02 + Na2S03 + H20 *>2 NaHS03
In the regeneration cycle, the above reaction is reversed
by the application of heat releasing sulfur dioxide and regenerat-
ing the sodium sulfite.
2 NaHS0 ^NaS0 + S0 + H0
425
-------
APPENDIX V
RECOVERY OF SULFUR FROM SMELTER GASES BY THE ORKLA PROCESS AT RIO TINTO
Submitted for discussion, 21«t April, 1949.
fRecovery of Sulphur from Smelter Gases by the
Orkla Process at Rio Tinto*
By H. R. PoTTaf, Member, and E. G. LAwronoJ, A.K.S.M., Member
INTRODUCTION
NUMEROUS patents have beon granted for processes designed to
recover sulphur from industrial gases, but so far as the authors arc
aware only two or three of these processes have been successfully
worked on a large scale ; the object of this paper is to describe one
that has given very successful results in at least three different*
countries—namely, the process which was worked out in Norway
by Mr. N. E. Lonandur of the Orkla Grube Aktiobolag and which
has tho distinguishing feature of recovering sulphur as a by-product
of copper smelting.
Sulphur was first successfully recovered as a by-product from tho
blast-furnace smelting of pyritic copper ore at a small plant at
Lokken, Norway, about tho year 1928. Tho success of this pilot
plant led to the construction of a large modern smelter with four
blast-furnaces at Thamshaven, near Lokken, which was completed
about 1932.
The first unit (a single blast-furnace) of tho Rio Tinto plant
wont into production in August, 1980, and tho plant has since been
gradually expanded by the modification of two more furnaces of
the original smelter, so that there are now, in all, three furnaces
specially equipped for tho recovery of sulphur.
Reference will, from time to time, bo made to the plants of both
the Orkla Company, Norway, and of Mina de S. Domingos, Portugal,
operated by Mason &, Barry, Ltd., but the main purpose of these-
notes is to describe the work at Rio Tinto, as tho conditions arc, in
certain respects, markedly different from those prevailing at Orkla.
For convenience of presentation tho paper is divided into throe
Hoctions—nanwly, Section 1—Principles of the process ; Section 2—
The plant; Section 8—Practice of tho process.
•Paper received on 12th January, 1049.
tSinelter superintendent, Rio Tinto Co., Ltd., 1034-1948.
JTnuhnical staff, Rio Tinto Co., Ltd.
426
-------
428 IT. n. POTTS AND E. 0. tAWFORD : RECOVERY OF BUM'llt'tt
SECTION 1— PRINCIPLES OP THE PROCESS
It will be remembered that in tho smelting of pyritio copper
ores the loose ntom sulphur is distilled off at tho top of tho chargn
column and burns in atmospheric air. The iron monosulpliido
descends and at or near tho focus burns in the oxygen of tho Must,
forming iron oxido (which is at once slagged) and sulphur dioxide
gas ; unoxidizod monosulphidc, together with tho mtlphidu of
copper, forms tho matte, which nlso contains Bomo magnetite nnd
the sulphides of load, zinc, etc. Tho authors have referred l»
tho compound remaining after tho distillation of tho loosu ntom
Biilphur as monosulpliido of iron, but actually this compound is
generally considered to bo FonSn + i, in which n is gonornlly tul««n
to bo about 7. The reactions (lien are :
(1)
(2}
It has long been realized that under certain conditions it in
possible effectively to reduce SO, by solid carbon according to the
equation —
SOt-fC=CO,+ iS, [[[ (3)
The inventor of tho process therefore conceived the idea that if
an ordinary blast-furnace, primarily designed for the smelting of
sulphide ores to matte, could bo fitted with a closed top very
similar to that used on a conventional iron blasl-furnnco nnd that
if, furthermore, a substantial excess of coke wero added to tho
charge, two results would follow : First, tho looso atom sulphur
would not burn, as thoro would bo no oxygen present and, secondly,
the S0t produced at the focus would bo reduced in the course of its
ascent through tbo charge column, by the colte present. In this
fashion a g.is would bo produced containing only nitrogen, carbon
dioxide, and sulphur vapour, tho last-named derived partly from
tiio loose atom sulphur distilled from tho pyrites while descending
towards tho focus, and partly from the reduction of tho SOj
formed near the focus. On cooling tho gas tho sulphur vapour
would condense and >l>o sulphur could thus bo rocovorod.
Tho Norwegian cxpertmonlers found that it was quite a simple
matter to close in tho top of tbo furn.ico and to use for charging a
line of bells similar to those used on an iron blast-furnaco. They
also found that when using 100 Itg. colco per ton* of pyrites in tho
charge the effect was ns expected arid a substantial recovery of
sulphur resulted.
Ideally, then, tho aim of tho process is to produce as much
sulphur dioxide as possible at tho focus and to havo this travel up
the column in such a way as to effect complete reduction by solid
carbon before tho gases loavo tho furnace. It is usually assumed
that this reaction takes place for the most part in throo stages.
-------
FROM flMEWBR OASBB BY TltH ORTCLA, PROCESS AT WO T1STO 429
The first in represented by equation (3). In the second stage the
C0t formed is reduced by coke a little higher up the furnaco :
CO,+C-2CO (4)
The CO at once reacts with SO, thus :
80,4-200.= §S.+2CO, (5)
This last reaction goes towards completion at temperature* below
COO°C. It must bo carefully noted that if equations (4) and (5) bo
added together and like molecules subtracted from each side of the
equation tho result is equation (8). Therefore, from the point of
view of coke consumption per ton of SO. reduced, it matters not
at all whether reduction proceeds in one step or in three, provided
that all the CO involved is derived from COZ produced by equation
(8) and not from other sources which will now bo discussed.
Unfortunately tho reduction of SO, by solid carbon—equation
(8)—will only take place with reasonable rapidity at a tompnraturo
of about 1,200°C. and in tho copper blast-furnace the high
temperature zone does not extend for a sufficient distance above
the focus to give efficient reduction. In other words, there is
insufficient contact time at the temperatures prevailing for tho
coke to burn in S0» and much of it passes unburnt down to tho
focus, whoro it burns in oxygon to CO,. Tho greater part of (his
00t is reduced to CO just above tho focus and then reduces SO.
in accordance with equation (5). This burning of coke in air nt
tho focus has two very ill effects : First, it moans that reduction of
SO. is dono by CO, using 750 kg. of carbon per ton of sulphur
instead of only 875 kg., tho quantity required for reduction by
solid carbon whether directly by equation (3) or in stages by
equations (8), (4) and (5) ; secondly, the carbon consumes oxygen,
which would otherwise bo available for oxidizing FeS and producing
SO,. Tho practical manifestation of this effect is a largo matte
' fall' of poor grade.
Tho fact has to bo faced that, in attempting to carry out (he
reduction of nulphur dioxide by coke in n wator-jackotod furnace
within a few feet of tho focus, (he aim is to superimpose a reduction
process on one which is essentially intensely oxidizing. Tho result
is necessarily a compromise, in which tho pyritic smelting is not
very good and the reduction of sulphur dioxide somewhat inefficient.
A cotnploto carbon and sulphur balance is sot out in what follows
and a quantitative analysis of the theory developed in an endeavour
to assess l,ho amount of SOZ reduced by C and by CO respectively.
Tho authors must, however, state at tho outset that in developing
a quantitative analysis of tho thoory, (hey do so with great reserve,
because of (ho enormous difficulties involved in obtaining accurate
samples and analyses of tho furnace gases; this subject deserves
a papor to itself. Suffice it to say that oven to-day it is not
absolutory certain that tho sampling and nnolysis at Rio Tin to gives
the correct distribution of sulphur in tho guses as between thr
various compounds SO., COS, CS2, etc.
I thnli 'i sp'1 ' thi- —orta:"*" and '" ^ito of the
428
-------
430 E. R. POTTS AND E. O. LAWFORD : RECOVERY OF SULPHUR
TABLE I
TYHCAI. FURNACE FEED AND PRODUCTS
(in metric tons por furnace day)
Tout
Cu
per cent
S
per cent
Ft
per cent
SiOt
percent
At
percent
Afh
per cent
ftSKO
Pyrites
Quartz
Converter slag
Limestone ...
Sulphur
Coko
Total
PRODUCTS
Matte
Slag
Dust
Crude sulphur
188-6
51-8
23-0
16-6
5-0
20-0
305-0
40-0
101-0
1-5
68-5
1-70
2-75
48-12
1-60
00-01
41-25
6-33
55-53
2-60
88-30
18-00
0-71
2-72
0-23
0-30
25-32
2-54
06-91
61-45
40-78
33-30
2-12
0-20
Per cent
by Vol.
CASES
CO,
CO
SO, ,
H,S ,
cs,
COS
3 ...
13-8
0-9
0-5
3-0
0-41
Total
-------
PROM SMELTER OASES BY THE ORKLA PROCESS AT RIO TINTO 481
fact that they make certain broad assumptions the validity of
which may possibly be questioned, the authors think it is worth
•while to sot out a quantitative analysis of the reduction theory.
The data required are given in Table I.
The first step is the determination of the volume of gas per ton
of pyrites smelted by means of the carbon balance.
leg. of Carbon per
1,000 kg. pyrites
Charged to the furnace as coke 07-01
Carbon charged to furnace in limestone 0-32
Totnl carbon charged to furnace 100-93
Carbon in exit gases :
g. Carbon
per cv. m.
CO, 13-8 por cent 73-98
CO 0-0 4-82
COS 2-UO
CS 3-84
Total 85-54
Therefore the volume of gas por 1,000 kg. pyrites is—
108-03 x 1.000 , „„
— 1.200 cu.ni.
The total sulphur in tho exit gases per 1,000 kg. pyrites will
therefore be:
In SO8 ... 42-7 g. S por cu. m. X 1,250 03-3 kg.
.. H2S ... 0-1 x 7-7 „
.. CS 20-5 X , 20-0 „
.. COS ... 7-7 x 9-6 „
.. S 3-0 X , 3-8 „
Total... 80-0 .. „ „ „ X , 100-0 „
Tho sulphur balance can now bo constructed as in .Tnblo II on
page 483.
The first step in the quantitative analysis of the theory is to
determine tho quantities of sulphur produced respectively by the
reduction of SOZ and by the volatilization of volatile sulphur. Tho
calculation follows:
kg. S per
1,000 kg.
pyrites
1. Calculation of volatile tulphur
Total sulphur in pyrites 482-2
Leas Sulphur combined with Cu, Pb, Zn, etc 19-0
Sulphur in FoS, 403-2
Therefore, volatile sulphur (42 per cent of sulphur in
FoSj) 194-5
Add Sulphur in residues 25-4
Total volatile sulphur 219-0
430
-------
432 ir. n. POTTS AXD E. o. LAWPORTI : nEcovEnr OF sotnrun
leg. Sjtrr
1.000 l-g.
j>yr,te»
Z. Calculation of fulptiur recovered by reduction ofSOt
Total lulphur cho.rpcod to fwrrmco 500-0
Dtflnet volatile su'.phui-, volnl ilizc-d 21'J-O
Fixed milpliur (tutoring mm
-------
FROM SMBVrnn OASBS m TUB ORKLA PUOCBSS AT RIO TINTO 433
TABM II
CifAnoieo TO PuiWAOie
Pyrito*
TOT AL . . . • . .
CoUo
PRODUCED
Mntto
Stac
Dust
Oanes
TOTAL
Tout
188-0
n-o
23-0
51-8
10-0
'85-0
20-9
C8-5
49-0
101-0
1-5
Per cent
,n/p~ ''*"' ralf-
432
-------
484 . H. H. POTTS AND B. O. LAWlfORD ! KECOVERY OP
i.e., adding more coke to tho charge — would effect a greater dogrou
of reduction and thus increixso the recovery of sulphur. In practice,
however, although it ia possible that tho amount of reduction in
increased, there is no improvement in recovery for any increase in
carbon beyond about 80 kg. pov ton. Tho total sulphur in tho
j;ascs remains tho same, but with a different distribution between
tho various compounds ; S02 in diminished, but CS, and COS are
increased.
It would seem in fact that the- CS, and COS in the exit gases
from tho furnaco are dependent on the amount of coko present, tho
concentration of tho sulphur vapour and the quantity of CO in tho
#ises. For example, the COS is formed by the action of CO on
sulphur vapour at temperatures below 800°C., whereas CSj is
formed nearer tho focus by reaction between sulphur and hot
carbon. The hydrogen sulphido in tho gases depends very much
upon the amount of sulphur dioxido in tho exit gases and tho con-
centration of water vapour present, n small reduction in tho amount
of sulphur dioxido present increases the H4S, tho increase being
roughly in the proportion of
of SO, for any
given concentration of water vapour.
Tho effects of varying the carbon/pyrites ratio (i.e. tho percentage
of coke on tho burden) ion charged par ton pyrite, kg
„ CS,
. COS
.. ..H.S
Total milphur
COt Vol. per cent
CO
1
104-0
39-2
18-0
7-0
0-9
71-7
14-3
0-8
0-5
2
77-3
51-6
5-0
0-2
7-3
71-0
12-8
0-5
1-0
3
61-0
70-8
5-8
4-8
4-8
80-0
Jl-0
0-0
0-8
The conclusion is, therefore, that, whcro conditions make
impossible tho further treatment of the exit gases, tho most
433
-------
VBOX BMBtTBn OA8K9 BT THE ORKtA PIIOCE8S AT RIO TISTO 435
economical percentage of coke on pyrite will correspond to about
80-90 Jig. carbon per ton of ore. Whore, howpvor, gases are free
of arsenic nnd can be subjected to catalysis after thoy leave the
furnace, the coke can bo profitably increased because tho catalysing
of Gas 1 (Table III) will yield fl total of 84 g. sulphur por cu. m.,
•whereas tho catalysis of Gas 2 would only yield '21 g. por cu. m.
•In othor words, with 1,250 cu. m. gas por ton pyrites, the extra
27 kg. of carbon would produce an additional 1C kg. of sulphur.
Tho reactions upon which catalysis of tho exit gases depends are :
SO.-fCS, =COZ -f-'S, (6)
SOl-f 2COS = L>eo, + JS, (7)
SOJ4-2H,S = 2HsO + iS, (8)
At Orkla catalysis of tho exit gases has always boon most success-
fully practised and has resulted in a much higher overall recovery
of sulphur than has so far been possiblo when treating tho arsenical
Iberian ores. Tho precise offoct of arsenic is discussed later.
Tho type of smelting dono in tho Orkla process is pyritic, as it H
the oxidation of a sulphide oro, and nearly 70 por cent of tho heat
generated is derived from the oxidation of iron sulphide. For the
process to work well and smoothly it is essential that a high rate of
oxidation should bo at all times maintained and to ensuro this a
llux containing a high percentage of froo silica is essential.
From what has been written it can bo inferred that tho con-
ventional typo of wntor-jackotod copper blast-furnace may not
b» tho most suitable apparatus in which to carry out this process,
in which a reaction xono 1ms to bo maintained at over 1,200°C. if
tho sulphur dioxide is to bo reduced in tho contact time available.
This unsuitability has boon recognized for a long time, but so far
the problem of reconciling copper smelting with high sulphur
recovery in ono apparatus has not, in tho opinion of tbe authors,
boon fully solved.
Notwithstanding all its admitted shortcomings, tho Orkla
process usos much loss coke per ton of sulphur produced than any
of the other reduction processes. The reason, of course, is that
nbout 05 por cont of the sulphur recovered is tho loose atom sulphur
nnd only 85 por cent is derived from tho reduction of SO., whereas
Jill othor processes first burn tho raw material to SO... Thus at
Hio Tinto slightly over 8 kg. crude sulphur per kg. of carbon is
obtained, whereas tho theoretical equivalent when reducing S02 by
€ is 2-Gfi; in practice the figure is almost certainly less than 2-00 kg.
crudo sulphur per kg. of carbon.
The comparative inefficiency of S0t reduction in the Orkla
furnaco is therefore much loss important from tho economic stand-
point than would bo the case if all tho sulphides in the food to the
furnaco had first to bo oxidized to S02. This point has an important
hearing on tho treatment of sulphides othor than FoSz.
SuLPiiun PURIFICATION
Tho crudo sulphur coining from tho condensers and mist Cottrells
contains 1-5-2 pnr cont of arsenic and 0-2-0-3 per cont of nsb.
434
-------
436 H. B. POTT9 AND B. 0. tAWronD : RECOVKUY OF SUtPHUR
Arsenic is removed by circulating milk of Hmo through tho molf.tm
'ulpb'jr in autoclaves. When llio original experiments wore carried
out to determine tho most efficient method ol refining, various soda
compounds worn tried, with tho idea of forming sodium thio-
ixrfioniU.o. It •wn<» found, that when caustic soda was used mixad
with lime i\ much quicker removal of arsenic resulted, but NaOH
tjlwuys gav
arsenic. It will bo seen from Table IV that the actual consumption
is between 1-8 and 2-2 tons free CaO per ton arsenic. Thero is
always a portion of tho CaO which fails to react because it is inort
through being overburnt, also calcium polysulphidcs are formed.
The treated sulphur contains some 8 to 10 parts per million of
ursenic nnd is pnssed through ordinary steam-heated filter presses,
these filtors removing the ash so that tho fina. product is sulphur
of high purity.
SECTION 2—THE PLANT
Tho process flow-sheet is comparatively simple (Fig. 109), and
various sections of the plant will bo described in some detail boforo
proceeding to (bo discussion of metallurgical results.
Before describing th<» individual units of tho plant, a very brief
summary of the sequwu of operations will bo given.
Sulphur-bearing gases from blast-furnaces nro cleaned by passing
through Cottrell electrostatic precipitator units ; they then pass
first through condonsors<, where tho larger part of the sulphur is
recovered, and then through a second set of Cottrolls, where a
further recovery of sulphur ii m.a'lo. Tho rriido sulphur contains
arsenic, which is removed by a washing process employing quirk
lime.
It should bo clearly understood that tho llio Tinto plant is very
much an improvization and that it certainly cnnnot bo described as
an ideal Orkla process layout. In this respect tho Norwegian
plants and that of Minn de S. Domi'igos, 1'orlugal, aro both much
435
-------
rnoM SMBLTBH OASES BY TITE OHKI.A TOOCESM AT mo TINTO 487
superior ixnd it is hopod that n fnll description trill be given during
the discussion of this paper. Tho operators referred to had the
advantage of boinp unencumbered with nn existing copper smoltor
JIIK! RO wore able to bnild from tho fonndntions with tbo Orkln
smelting system in view. At Jlio Tinto, on tho other hand, there
w:w already n Pinoldir ninny years old nnd in many r«'spt'cl.s already
Miiliqiiiilod, bul. from which tho How of copper had l.o be iniiinliiinrd.
Jt wns, thoroforo, noressury l.o crcaLu an Urkla Hinvlting plant by a
f
-------
433 H. It. POTTB AND B. O. I.AWPOM) : HECOVRIIY OF SULPHUR
a i
1*
enw
CO fi
e> 2
jj.
?
.8
||
5 8
S 2
*T **
^ **•*
o ^
i"
i?
0
k
a
1
V
0
6
2
*.
•s
to
.J
V.
O
H-4
S
W
S
s
!a
^
•e
e
^
1"
u
j S
1
3
s
!^
J
o
,5
Sj.
ftl
!
*-
^
a> '
c» i O
o 1 in
CO OS
to
IO
10
«•'. 1 1
« I 1
Cl
? 1 1
o 1 1
CO *
10"
CO
10
O CO
IO IO
O
oo
00 O |
tbio '
•* n
«0 IO
'n* CO .
coci 1
M* IO
CO
to
m
V
KARGE
Pyrites
Concentrat
Precipitate
0
t-
JO
OS
M4
1
m
cl
IO
1
O
m
S
CO
0
2
1
00
en
o>
Cl
1
e
CO
0
O
CO
o
lr c.'
O l>-
O 00
ss
-8
»•*?>
- to e-<
10
t- •*
•f 1-
ooo
CO
c> o
10 O»
00
oo ci
Cl
i US to
9 1-4 eo
5 Cl
>co t»
> « C-l
i •* to
1 00 C)
bo
a
'Ji
' °«
•*
«.
•H<
S
"•
?,
00
,_
Cl
CO
to
o
O)
n
CO
00
n
j§
w
H
1 1
1 1
1 1
1 1
g.0*
-
IO Cl
r-"ob"
"
SI
_
CO
CO
Cl
1
e>
'8
to
Cl
1
s
2
•«
o
to
f*
1
1-
o
00
CO
1
eo
o
i-
us
o
1
! I
1 1
1 1 1
1 1
"jO-s,
So g«
>o
ta
1 I
1 1
1 1 I
I i
1 1
1 i 1
xP"v
(A *»*ifl
1 Is
:i
° Is'c
« " § 5
•*• fe 8 S
6^1 £
437
-------
PROM SULPHUn OASES BY THK OHKLA PROCESS AT RIO TINTO 489
00 CO-" Ok
f gp ' "*
, OOO-J« I I
1 OO CO 00 I |
o cr> co 10 o IN
•* •* Cl iO if O
t-« i-l OO CO C4 P-
cb co" 10
coo
«?-«?•? 1 1
r- 10 c< o> ' '
O>CN — 1
Ok tf CO CO
*-» OS CO I— I |
coco_io_co_ 1 |
!*• IO ^t* CO
o ioco co
IO IO O> Cl
1 '".'"I 1 *
rH
r-»
cfco A
00 CO i-l O O Cl
ci —i co !*< co o
00 — 1 CO — 'O 00
CO Cl IO i— 1 Cl 00
GO CO -H CO
CJ t^ CXN 1 I
i^ co c) -H 1 1
o>e» ci
co oo o O)
CO M* C-l 00 , i
^."."l-i i
00 if3 iO IO
OO CD CO
ft »— < IQ
F4
00
°i
1
0>
o
IO
CO
o i< ci co
O> O 00 (N CO
CI I CO i-l CI IO
*N 1 CO O -* 00
CP CO i-l
"1 a c»o oo
00 ^ CO IO OO .
CO CO CO C) 00 [
co'co co" O*ef
CO CO OJ —
CO
* "* °* 1
0
*» *
H-Jj ? 1 ^c5 "^'? 1 1
£ M ,? * 5
CO t- C- CO CO
•*r oo -i 10 oo
CO i O_ lO^COCO^
00 O M
CI | <3> ~o2 I
i- 1 c» ~Tci 10 '
a> <3> <>. i-» oo
10 1- oo 10 o
'Ji CM CO CD to i
*"1""1 *. e'l."H 1
OO 00 CO O O)
CO «0 00 X,
p- 1 •— «
2
-It, E
03 c
FA o q
^ 1* H &.I
tfpj« o^ "-a
* S| § a 1- «| k
•JjE-S o3 jS*!^
ijjOh-l p;W >(\A<
phur entering tho furnaces in tho two periods, 15,927 and 11,117 tons of sulphur arc from
when refining sulphur which do not represent fresh sulphur entering tho process.
n tho pyrites and concentrates and perhaps from the converter slag and matte.
om the above-mentioned four materials only, the sulphur recovery ia ;
1934-1939 133.360 _ „.,„ „_
5"3 5 c e
°!£ f J
Il| I
o £ 5 .a
O S O j_
£ a .a 1 S
i^j S5"
-^t:
» i I I S
« = & 8 2
°^ - s A
j: « ~ o< o
-^^2
0 Hfc
s i
i, * S,
5 I? S
: I 3
S «
1= S 7
= •§. 3
7 "317
^Isi
S | " 3
2. -5 ««.
n>7 C)
12
•« i
c J-
§ «
o e>
o
c>
•O
a
S
1
o
f.
I
V
'J
o
o
_>»
»• ° .
3 -B *?
s s s
i £
5 | e.
! i S
i t «
P « 3"
O«; r-
01 ,2 O
0 ** 5 eT
7 1§-^
S SSg
O^ w> p-, C^
00^ *• 0^
^-i (D »-*
2. b ;2.
fc °
C T*
27
25
"~ 0
t_- —
l-«
438
-------
440 n. n. POTTS AWD 8. o. &AWFOITD : nECOVunv OF mn.rmm
idenl sites and properly connected to the furnaces but on situs
dictated by the availability of space in the existing plant.
There is no doubt that the ideal layout consists of smelting
furnaces each coupled to its own dust procipitator and sulphur
condenser, all precipitators and condensers being placed as close
ns possible to the furnaces. It will bo seen from the description
which follows how very far it 1m been necessary in tho present,
cn=e to depart from thin ideal.
Bl.AST-FunXACES
Three out of the original six open-lopped furnaces have been
modified to nllow sulphur recovery ; tho first being No. 6 fnrnaco,
the most easterly one and, therefore, nearest to the hot Cotlrolls.
Tho conversion of Nos. 5 and 4 furnaces followed later. Tho dates
upon which these three furnaces were first blown in after being
modified were : No. C, I Aug. 1MO : No. 5, 5 Sept. 1932 ; No. 4,
10 Feb. 1942.
Tho normal copper-smelting furnace is preserved in its usual
form up to tho level of tho charging floor, but instead of tho top
being left open and charging effected through side doors on the
level of the charging doors, the (op wns heightened, completely
closed in, and equipped with four sets of doublo-vixlved charging
bells, very similar to the gas-tight, bulls used on tho normal iron
blast-furnaco; by this means charge can bo introduced into the
furnace without allowing tiny access of atmospheric air. l''ig. 110
(Plate XXI) shows the const motion.
It will bo seen from Tnblo V t.lmt tho dimensions of tho three
furnaces are not identical. All of them have tho same height.'and
use standard 12-ft. side jackets, but the rake of the jackets is
different, being much greater in No. C ; this largo flare was origin-
ally given with the idea of reducing tho gas speeds in the furnace
and thereby effecting a better reduction of SO^; the authors
think that in many ways No. G has been consistently the most
satisfactory furnace, perhaps because, having tho shortest flue
connecting it with tho hot Cottrolls, less back pressure is developed
and a faster smelling rato results. However, the wide flaro has
the. disadvantage of needing specially largo, non-standard end and
breast jackets. Tho jackets :•'•(> all made entirely by electric
welding, tho fire sheets " ->ing of a Jj-in. mild steM, and tho back
xheots of -ft-in. Tho 12-ft. jackets nro satisfactory on tho open-
topped furnaces, but they are not sufficiently stiff to resist the
heavy thrust imposed by an Ift-ft. charge column and they tend
to buckle, even though they are roinhict'd by tlirec heavy buck
stays bolted across f-iirh side of the furnac<'. Leading particulars
aro given in Tablo V.
Tito upper part of nil tho furnaces, abovo tho jackets, is supported
on a framework <>f box girders and consists of a steol box welded at
tho joints. This in solidly lined with ISin. ot good-quality Jiro-
brick ; the roof of tho furnace between the charging bell* is formed
by an arch made of carefully cut- and lilted lire-brickH. Tho fttrnnco
439
-------
TROM 8M1!T,Tl?n OA8EB BY TltB ORKIA TBOCFSS AT RIO TINTO 441
TABLE V
Ko. 6 Furnace Ifo». 4 and 5 Fitrnaix*
•Length 18ft. 18ft.
Number of charging bolls ... 4 4
Height from closed bolls to
floor 21 ft. 7 in. 22ft.
Height from top of jnokots
1o roof 10 ft. 2 In. Oft. 11 In.
Width nt top of jackets ... 0 ft. 4 in. fi ft. 15 in.
11 t> toyorca 4 ft, B in. 3 ft. 5 in.
Hearth nrcn 82-0 sq. ft. 01 nq. ft.
N'imbor iif tiiy«ro» 30 30
Dinmolor of tuyeres 4 in. 4 in.
Ttiyftro nrot» 3-l4«q. ft. :M4sq. ft,
Jtivtio—Hearth to tuyoro
urea 20 to 1 19 to 1
Hoii>ht o,/l of tnyta* to floor 2 fl." 0 In. 2 ft. 0 in.
Number of gos ports ft 18
Height of pns portd 1 ft. 0 in. I ft.. 4 in.
Width of gns ports 3ft. 7} in.
Total nroa of EOS port* 22-5 sq. ft. 14-4 sq. ft.
Number of Hiflo ji\rltots 12 12
Diinrntinn* of side jacket*... 12 ft.. 3} in. by 3 ft. 12 ft. 3} in. by 3 ft.
J'lnro of jnnkotH from rontro
linn of liiyorra 1 ft. in 2 ft. 1 ft. in 4 ft.'0 in.
bottom in put. in with ono lowor row of firebrick nnd an upper ono
•of magnosito brick ; tho use of firebrick alono for tlio furnncr
bottoms has boon found unsatisfactory, becamo of tho scouring
vlTccl of U\o Ring.
Tho praclico of lining tho inner side of tho water jackets com-
pletely with good-grade firebricks for (ho full height of tho jacket,
baa been made standard in the Norwegian plants, upon tho
supposition that this will maintain a higher temperature next to
tho jacket and therefore load to better reduction of S02, and also
that tho hot surface of tho brickwork may have a catalysing effect.
This lining has been tried onco or twice ivt Rio Tinto, but its uso
has been discontinued for many years past, as it was found that the
lining is soon fluxed away, leaving the jackets bare ngnin.
Tho gases leave tho top of tho furnaces through ports and arc
plant, but because of the exisling layout of tlui
original furnace plant (hey nro unavoidable. They form a great
iilmtuclo on (bo fiirnaco feed Moor, hindering ready access to the
furnaces ; (hoy also slowl) fill up with a layered deposit of dust
END OF COPY
440
-------
APPENDIX W
FW-BF DRY ADSORPTION SYSTEM66'67
W.I HISTORY
A 20 MW demonstration FW-BF dry adsorption system was put into
operations at the Scholz Steam Plant of Gulf Power Company near
Sneads, Florida by Foster Wheeler Energy Corporation in 1975. The
technical basis for this system is derived from a license with
Bergbau-Forschung GmbH, the research grouo for the German Bituminous
coal industry. The basis of design for both the Gulf Power Demon-
stration Unit and the STEAG Demonstration Unit at Lunen, West
Germany, is the very extensive work done by Bergbau-Forschung over
a two year period at their pilot unit at Welheim, Germany. This
plant ran for one continuous period of 6,000 hours, and the results
of the two year testing period were published in 1970 at the Second
International Clean Air Congress in Washington, D.C.
Bergbau-Forschung is presently ooerating a demonstration plant
for the desulfurization of flue gases. The plant accepts 88,275 cfm
of flue gas in the form of a slipstream from a 350 MN coal fired
boiler owned and operated by STEAG at Kellermann Power Station in
Lunen, Federal Republic of Germany. This prototype demonstration
unit consists of an adsorption section, a regeneration section and
a modified claus unit for reducing S02 to elemental sulfur. The
demonstration unit at Sneads, Florida consists of an adsorption
section, a regeneration section and a RESOX* unit for reducing S02
to elemental sulfur.
*Trademark of Foster Wheeler Energy Corporation,
441
-------
W.2 DEVELOPMENT OF THE FW-BF PROCESS
The S02 removal system developed in the laboratories of
Bergbau Forschung in Essen, West Germany is based on and designed
CO
for a special activated coke adsorbent. The activated coke, the
most critical ingredient in the system, is characterized by
excellent SCL adsorption, high ignition temperature, and good
physical strength and is the result of a research and develop-
ment program initiated in the late 1950's.
The basic system consists of a gas solid contacting device,
the adsorber, and a regenerator which is the desorber. Figure W-l
shows the schematic of the adsorption section. Within the adsorber
r
the activated coke moves downward in the plug flow contained by
permanently fixed steel louvers on the gas entrance and exit sides
of the unit. The polluted gas stream is passed in through the
louvers, through the adsorbent and out of the adsorbent through
louvers on the opposite side of the adsorber. The SO^ contained
in the gas stream is initially adsorbed on the inner surface of
the activated coke and undergoes subsequent oxidation to sul-
furic acid in the presence of the oxygen and water vapor also
existent in the polluted gas. Coincidentally, the adsorber also
adsorbs NO and functions as a panel bed filter for the removal
A
of particulates entrained in the gas stream. The sulfuric acid
content of the activated coke increases as a function of coke
dwell time in the adsorber; therefore, the coke discharged at
the bottom of the adsorber contains the highest possible amount
of sulfuric acid for the given conditions and adsorber geometry.
After the adsorbent discharged from the adsorber is
separated from particulates by a vibrating sieve, it is regener-
ated.
The regeneration is effected thermally by heating the sul-
furic acid loaded adsorbent in an inert atmosphere. The condi-
tions of regeneration are selected to be sufficient to cause a
directional change in the driving forces governing the reactions
in this system, whereby the participants undergo a modified
442
-------
activated coke
from regeneration
from the power plant
to the stack
blower
to regeneration
Figure W-l. Adsorption Section - Lunen
443
-------
reversal of the adsorption reaction, resulting in the reduction
of sulfuric acid by the fixed carbon of the adsorbent to yield
sulfur dioxide.
The regeneration is carried out in a moving bed reactor
utilizing sand as a direct heat carrier to heat the adsorbent
to around 650°C (1200°F). The effluent gas of the regeneration
contains 20 to 30 percent sulfur dioxide by volume and also HLO
and C02. It can then be fed directly to Foster Wheeler's RESOX
process for its sulfur dioxide content to be converted to sulfur.
The RESOX process uses coal as a reducing agent to pro-
duce elemental-sulfur. It was developed in Foster Wheeler
Corporation's John Blizard Research Center, and is the result
of a research program initiated at the end of the 1960's.
The process is designed to reduce sulfur dioxide contained
in an offgas stream to sulfur, and to condense the so produced
sulfur product from the gas stream. The RESOX process is
capable of handling a wide range of inlet gas compositions,
and does not require gas cleaning, drying, or dust removal systems.
Crushed coal is the only material and the only catalyst consumed
in the process. The process itself represents a new way to
achieve the desired degree of reaction between sulfur dioxide
and crushed coal at temperatures as low as 650°C (1200°F).
The major process equipment consists of a reactor vessel
and a sulfur condenser. In the reactor vessel sulfur dioxide-
rich gases are reacted with crushed coal to yield gaseous elemental
sulfur. This sulfur is condensed from the gas stream in the sulfur
condenser. The high purity liquid sulfur effluent of the process
represents a non-polluting by-product.
W.3 PROCESS CHEMISTRY
The process chemistry of the FW-BF Dry Adsorption System
for the Adsorption, Regeneration and RESOX sections are described
below.
444
-------
The sulfur is produced in the form of a gas which is sub-
sequently condensed. The nitrogen and carbon dioxide constituents
of the Regenerator offgas pass through the RESOX reactor without
taking part in the reactions.
The process configuration for the FW-BF Dry Adsorption
System at the Scholz Steam Plant is shown schematically in Figure
W-2. The system is installed on Boiler No. 2, a nominal 40MW
pulverized coal-fired boiler, which has been retrofitted with a
sectionalized, high efficiency electrostatic precipitator capable
of 99.7% particulate removal.70
The system is sized to handle one-half of the total flue
gas from this boiler. The 20 MW equivalent in flue gas entering
the system is 85,600 ACFM. The flue gas leaving the boiler air
pre-heater is at a maximum of 350°F. The system is nominally
designed to meet Florida SOp reduction codes of 1.2 Ib. SO^/MMBTU
heat input which equates to a 74.5% removal efficiency require-
ment for 3% sulfur, 12,200 BTU/lb fuel. The actual performance
of the unit during initial operation far exceeded this requirement.
The equipment utilization and operation of adsorption,
Regeneration and RESOX sections is discussed below.
Adsorption is accomplished by passing the flue gas horizon-
tally through vertical columns of activated char in the adsorber.
Scholz Steam Plant Demonstration Unit adsorber consists of two
vertical stages of char beds designed in a modular fashion. There
are eight 6' x 6' beds in the first stage and four 4' x 4' beds
in the second stage. All beds are approximately 40 feet high.
The char in the beds is continuously recycled. A conveyor
at the top of the adsorber feeds regenerated char into a holding
tank which has discharge tubes that gravity feed the regenerated
char into the individual char beds. The char moves downward in
mass flow adsorbing S09 and NO as it travels. The char flow
c. A
445
-------
CD
C
o
CL
S-
o
to
CO
I
CM
I
o>
S-
446
-------
rate is controlled by a vibromagnetic feeder at the hopper out-
let below each char bed. The saturated char is collected at
the discharge of these feeders and sent by a combination of
natural frequency conveyors and bucket elevators to the regen-
eration section of the system.
The flue gas entering the adsorber is tempered, if neces-
sary, by the use of a dilution air fan (vane axial type) which
maintains an inlet flue gas temperature of approximately 140°C
(284°F). The adsorber discharge fans, one per stage, restore the
pressure drop suffered by the flue gas during its passage through
the adsorber and associated ductwork.
The regeneration section provides for the continuous on-
site regeneration of char which has been loaded with SCL in the
form of H?S(L. Regeneration is achieved by contacting the load-
ed char with hot sand. Sand is utilized as an inert heat transfer
media and as such does not take part in the reactions occurring
within the regenerator. Its sole function is to supply heat so
that the reactions may take place. The mixture of hot sand and
char at 650°C (1200°F) flow slowly downward through the regenerator.
Their flow is controlled at the discharge of the regenerator by
a char-sand separator-feeder. The char and sand are physically
separated by means of a vibrating screen deck. The char is spray
cooled to 220°F and returned to the adsorber. The sand is conveyed
to a fluidized bed sand heater where heat is added by direct com-
bustion of No. 2 fuel oil. Both the char and sand streams are
closed loop operations.
The flue gas produced by the fluidized bed sand heater is
used to preheat the incoming combustion and fluidizing air to
this heater. After preheating the incoming combustion air, the
flue gas goes to the boiler air preheater flue gas inlet for
additional heat recovery. It is then injected into the main flue
gas stream entering the adsorber, thereby assuring closed loop
operation of this gas stream.
447
-------
The RESOX section provides for the continuous on-site
reduction of sulfur dioxide to elemental sulfur. The low
volume S02 rich offgas stream is directed from the regenerator
to the RESOX reactor which is filled with crushed coal. The SCL
stream is reduced to gaseous elemental sulfur and the liberated
oxygen combines with a portion of the coal carbon to form carbon
dioxide. The gases leaving the reactor enter a sulfur condenser
where the sulfur is condensed to molten elemental sulfur. The
sulfur is collected and stored in an insulated tank (which is
equipped with steam heating coils to make up for heat losses
through the insulation system) to maintain the sulfur in a molten
form pending shipment via tank truck. The mass-flow coal move-
ment through the reactor is controlled by a discharge feeder.
The combination of non-consumed coal and ash is fed into a
receiver vessel for ultimate disposal after cooling. The tail
gases leaving the sulfur condenser consist of CCL, FLO, N,, and
those remaining "S" values not converted to elemental sulfur.
The gases are recycled to the boiler via a centrifugal blower
where the sulfur values are oxidized to S02 and then re-enter
the adsorber allowing complete closed loop operation of the unit.
W.4 PW-BF SCHOLZ STATION DRY ADSORPTION SYSTEM PERFORMANCE
During the start up of the Lunen demonstration unit, some
mechanical problems were encountered in the adsorption and the
regeneration sections. But according to Mr. W.F. Bischoff of
Foster Wheeler Energy Corporation, the adsorption and regen-
eration sections are reasonably well optimized now. He adds
that the only major problem existing at Lunen is the proper
functioning of the modified Claus unit. Foster Wheeler is cur-
rently looking for funding to install a RESOX unit to replace
the Claus unit at Lunen and then test the integrated system.
72
Foster Wheeler reports the power consumption for the
whole system to be between 0.6 and 1.5 percent of steam generator
name plate rating depending on the system design and the mode of
448
-------
operation. Another point of interest is the S0? removal efficiency.
Removal efficiency is a function of several parameters including
SCL concentration at inlet, inlet flue gas temperature, char dwell
time and gas residence time. In the brief period of operation
with diverse operating conditions, S02 removal efficiencies were
between 96 and 100 percent consistently. Another point noted
during the operation was that the exit gases from the adsorber were
0 to 30°F higher than inlet to the adsorber, thereby increasing
the buoyancy of the stack gases.
Based on present information the FW-BF Dry Adsorption
System appears technically feasible but requires full-scale
operating time and data to prove itself.
449
-------
APPENDIX X
SULFUR/SULFURIC ACID PRESENT AND FUTURE USES
AND MARKETS
X.I SOURCES
Over the past centuries, there have been many fundamental
changes in sulfur/sulfuric acid supply sources. Early civiliza-
tions obtairfed their meager requirements from native sulfur
deposits in or near volcanoes. The increase in demands in the
late 1700's and early 1800's was largely satisfied by the devel-
opment of native sulfur deposits in Sicily. Monopolistic prac-
tices by the Sicilians, which resulted in exorbitant price levels,
caused consumers to shift to pyrites as their major source of
supply during the second half of the 1800's.
For many years pyrite, iron disulfide, was the main sulfur-
containing material in the manufacture of sulfuric acid. In
1895, Herman Frasch developed his process for extracting sulfur
from underground deposits by injecting hot water into the deposit,
melting the sulfur, and recovering it in liquid form. The
exceptional purity and quality of sulfur appealed to the chemical
industry, particularly to the manufacturers of sulfuric acid.
As a consequency, Frasch's sulfur supplanted practically all other
raw materials formerly used in the manufacture of sulfuric acid.
The development of the Frasch process for mining the large native
sulfur deposits associated with the salt domes in Texas and
Louisiana created a new and important source of high purity
elemental sulfur for domestic and world markets. As a result, the
United States became the world leader in sulfur production in
1913, and it has never relinquished this lead.
450
-------
Frasch sulfur and pyrites have continued to maintain their joint
predominant positions as world sources of sulfur to the present time,
but they are being rapidly challenged by new sources of sulfur pro-
duced as co-products or byproducts. The trend started about 1950
with rapid increases in the production of elemental sulfur at
refineries and natural gas treatment plants and of byproduct sulfuric
acid at nonferrous smelters. The pyrites sector of the sulfur
industry is being seriously affected by the latter developments.
X.2 PRESENT USES
Sulfur is unusual as compared with most major mineral com-
modities in that by far the largest portion of it is used as an
intermediate in the manufacture of other chemical products.
Sulfuric acid is the most important of these intermediate
products. Ninety precent of the sulfur consumed in the
United States in 1975 was either converted to sulfuric acid
or produced directly in this form. Other intermediate products
were carbon disulfide and sulfur dioxide, each of which accounted
for 3% of the total sulfur consumption. Only 4% of the total con-
sumption was used directly in the elemental form. The use of
sulfur in U.S. by end-use categories is as follows:
• Agriculture (Fertilizers) - This category is by far the
most important, accounting for 55% of domestic sulfur demand. The
principal individual requirement is for the manufacture of phos-
phatic fertilizers, with sulfuric acid being the essential inter-
mediate sulfur product being used. Another individual, but
relatively small, end-use within this category is the production
of ammonium sulfate by the reaction of sulfuric acid with ammonia.
Additionally, a small amount of sulfur in elemental form or in the
form of gypsum is used as a soil conditioner and plant nutrient.
This latter use of sulfur is growing in importance because the
higher grade phosphatic fertilizers now being produced by the wet
phosphoric acid process do not have the needed sulfur content that
was a component of the older, lower grade fertilizer products.
451
-------
• Plastic and Synthetic Products - This category covers a wide
range of synthetics including acetate, cellophane, rayon, and various
products, fibers and testiles. The sulfur intermediates involved in
their manufacture were equally divided between sulfuric acid and
carbon disulfide.
• Paper Products - In this category, the largest single segment
of demand is in the manufacture of wood pulp by the sulfate process.
In this process, the major sulfur intermediate is sulfur dioxide,
generally produced at the plant site by burning elemental sulfur,
although some sulfur dioxide is produced as a byproduct at smelter
operations, purified and liquified, and shipped to the pulp mills.
• Paints - The major sulfur use in this category is for the
production of titanium dioxide pigment by the sulfate process.
Difficulties in the disposal of ferrous sulfate waste product has
led to the development of the hydrochloric acid process.
• Nonferrous Metal Production - This category covers the leach-
ing of copper and uranium ores with sulfuric acid. In the case of
copper, it is used for the extraction of metal occurring in deposits,
mine dumps, and wastes whose copper contents are too low to justify
concentrations by conventional floatation techniques, or for recovery
of copper from ores containing copper carbonate and silicate minerals,
which generally cannot readily be treated by floatation processes.
The sulfuric acid required for copper leaching is invariably the
byproduct sulfuric acid produced by the copper smelters in the area.
Sulfuric acid is the most commonly used reagent for the
recovery of uranium from ores. The sulfuric acid used is either the
byproduct sulfuric acid produced at smelters or sulfuric acid pro-
duced from elemental sulfur.
• Explosives - In this category sulfur is used entirely in the
form of sulfuric acid. Sulfuric acid is not only used in direct
manufacture of explosives, but numerous related nitration processes.
452
-------
• Petroleum Refining - This category includes not only
petroleum refining as such, but associated chemical processes
where process streams may serve both the refinery and the chemical
complex. The major idetnified end use for sulfuric acid is as a
catalyst for alkylation, a process by which liquid high-octane
gasoline components with very good stability may be produced by
a combination of gaseous streams. Sulfuric acid and hydrofluoric
acid are competing catalysts in this process. Sulfuric acid for
refinery processes is manufactured from recovered sulfur produced
at the refinery and from contaminated acid (acid sludge) returned
to the acid plants for reconstitution.
• Iron and Steel Production - Sulfuric acid is used as a
pickling agent to remove mill scale, rust, dirt, and grease from
the surface of steel products prior to further processing. The
sulfuric acid pickling process faces increasing competition from
hydrochloric acid, largely because of the problem of disposing
of the ferrous sulfate waste product. Although there are both
advantages and disadvantages in the use of hydrochloric acid,
it appears that it will largely replace sulfuric acid over the
long range. There is no well-defined source of sulfuric acid for
steel pickling since it is generally obtained from merchant
sulfuric acid plants that use the cheapest form of sulfur avail-
able in the area in which it is produced.
• Other Uses - This general category covers a wide variety
of end uses, including intermediate chemical products. These
miscellaneous uses, especially those involving sulfuric acid, are
intimately associated with practically all elements of the nation's
industrial and chemical complexes.
X.3 NEW USES FOR SULFUR73'74
For some time now, organizations in the sulfur industry have
been involved in the development of new uses for sulfur. The re-
search is directed towards the developments of new uses that would
453
-------
(a) consume large volumes of sulfur, (b) are economically favorable
to attract industry interest and capital, (3) are non-polluting,
and (d) are not costly to produce.
One of the major developments for new uses in recent years
has been that of a sulur-asphalt paving material called "Thermo-
pave" by Shell Canada.76'77 The product contains 13% sulfur,
6% asphalt and 81% sand, and possess considerable commercial
acceptance. It is economically attractive, has superior properties
when compared to the conventional asphalt-aggregate paving materials
and would use large amounts of sulfur.
A foam which is lighter than water with compressive strength
comparatively higher than that of typical organic polymers has
78
also been developed. Its potential uses are, for instance,
thermal insulation and general building insulation.
The manufacture of sulfur concrete involves the mixing of
molten sulfur (350°F) with sand or aggregate, in the ratio 30:70,
and then letting the mixture cool and harden, either to form
bricks, blocks, tiles or other structural materials. Other
potential uses of this product are concrete pipes, slip-forming
79
of street curbs, traffic barriers, etc. Indeed there is a huge
potential for exploiting and establishing markets for such a
product, but as yet it is still in the early stages of development.
Research in various other uses of sulfur has been carried
80
out. Included in these are: (a) the use of sulfur as a coating
material providing resistance against corrosion and errosion,
and (b) as a surface bonding material, providing a hard and
impervious surface on a wall that would not require mortar or
joining material between blocks.
However, while some of these products have already been
developed and tested successfully, it is highly unlikely that
any of them will be commercially produced on a significant scale
prior to the end of this decade. It should be noted that the
454
-------
research in the development of new products such as discussed
above was stimulated by the prospects of a sulfur surplus.
The use of sulfuric acid injected directly into the soils
81
in Arizona looks very promising. This procedure tends to break
up the adobe soil as well as decrease the pH to a more favorable
range for agriculture. Yields have definitely increased and it
is considered one of the promising future potential markets.
Approximately 4,000-5,000 tons per month of acid are presently
being used for this purpose.
X.4 PRESENT PRODUCTION AND CONSUMPTION
Sulfur in its different forms is produced worldwide with no
one country being a producer or supplier to world markets. In
1974, world production of sulfur in terms of sulfur content of the
82
product produced amount to 50.9 million tons. The United States
was the leading producer, accounting for 22% of the output. A
tablulation of world sulfur production in 1973 and capacities
for 1973, 1974 and 1980 is shown in Table X-l. The total U.S.
production of sulfur in 1975 was 11.26 million tons. The salient
sulfur statistics for the years 1971 to 1975 is given in Table X-2.
Sulfur for domestic consumption was obtained mainly from
domestic sources: Frasch 45%, recovered elemental 27%, and sulfur
in other forms 10%. The remaining 18% of the sulfur was obtained
by imports of Frasch and recovered elemental sulfur. In 1975,
go
Frasch sulfur accounted for 64% of the domestic production.
Frasch sulfur was produced at 13 mines in Texas and Louisiana.
Duval Corporation, with one mine in Texas, Freeport Minerals
Company, with four mines in Louisiana, and Texas Gulf, Inc., with
five mines in Texas and one mine in Louisiana, accounted for most
of the Frasch production. The five largest mines, with a production
rate in excess of one-half million tons per year each, accounted for
84% of the total Frasch sulfur output. They also accounted for 54% of
the total production of sulfur in all forms during 1975.
455
-------
Recovered elemental sulfur accounted for 26% of the total
84
domestic production of sulfur in all forms . It was produced
by fifty-six companies at 140 plants in 28 states, one plant in
the Virgin Islands, and one in Puerto Rico. Most of the plants
were relatively small size, with only five of them reporting an
annual production exceeding 100,000 tons. By source, 55% was
produced by 38 companies at 79 refineries or satellite plants
treating refinery gases, and two coking operations, and 45% was
produced by 29 companies at 59 natural gas treatment plants.
Table X-l. WORLD SULFUR PRODUCTION 1973 AND CAPACITY 1973,
1974, AND 1980
(Thousand long tons)
Produc-
tion Capacity
1973 1973 1974 1980
North America:
United States 10,921 12,000 13.000 15,000
Canada 7,779 8,000 8,000 7,000
Mexico 1,650 1,900 2.500 3,000
Other 100 200 200 200
Total 20,450 22,100 23.700 25,200
South America — 300 400 600 1,100
Europe:
U.S.S.R 7,500 8,000 8,500 12,000
Poland 3,600 4,000 4,000 4,000
France 2,000 2,500 2,500 2,500
West Germany 1,050 1,200 1,400 1,600
Spain 1,000 1,200 1,400 1,400
Italy 800 900 900 900
Other 2,700 3,000 3,300 3,800
Total 18,650 20,800 22,000 26,200
Africa 600 700 700 800
Asia:
Japan 3,000 3,500 4,000 5,000
Near East 1,250 1,500 2,000 5,000
China -- 1,200 1,300 1,300 1,800
Other 250 300 300 400
Total - 5,700 6,600 7,600 12,200
Oceania 300 400 400 500
Uorld Total 46,000 51,000 55,000 66,000
Source: Ref. 85
456
-------
Table X-2. SALIENT SULFUR STATISTICS
(Thousand long tons, sulfur content)
Production:
Shipments (Sold or used):
Byproduct sulfuric acid
Imports:
Pyrites (Canada)
Total
Exports :
Crude, recovered elemental, from the
Virgin Islands
Apparent Consumption 4,':
Recovered elemental
Other forms I/
Yearend Producers1 Stocks 5/:
1971
9,580
518
9,280
130
1,429
1,536
1,582
126
„„ • A1-S
97
-— L i ?n
1972
7,290
1,950
546
283
149
10,218
7,613
1,927
546
283
149
10,518
269
868
1
50
1,188
1,847
5
1,852
5,761
269
1,927
869
546
283
50
149
9,854
3,665
131
3,796
1973
7,605
2,416
600
212
88
10,921
7,438
2,451
600
212
88
10,789
302
905
15
1,222
1,771
5
1,776
5,662
302
2,451
920
600
212
88
10,235
3,816
111
3,927
1974
7,901
2,632
654
162
70
11,419
7,898
2,547
654
162
70
11,331
954
1,194
2
2,150
2,580
21
62
r/2,663
5,297
954
r/2,485
1,196
654
162
70
1/10,818
3,744
213
3,957
1975
7,211
2,969
767
237
75
11,259
6,077
2,902
767
237
75
10,058
967
930
y
1,897
1,288
7
57
1,352
4,782
967
2,845
930
767
237
75
10,603
4,857
269
5,126
r/ Revised.
I/ Hydrogen sulfide and liquid sulfur dioxide.
27 Less than 1/2 unit.
3_/ Accounted for as Frasch sulfur.
ftj Measured as shipments, plus imports, minus exports.
_5/ Reported producers' stocks after invertory adjustments.
Source: Reference 86
457
-------
Sulfur contained in by product sulfuric acid produced at
copper, lead and zinc roasters and smelters during 1975 amounted
87
to 7% of the total domestic production of sulfur in all forms.
It was produced by 12 companies at 22 plants in 13 states. Twelve
acid plants operated in conjunction with copper smelters, and 10
plants operated as accessories to lead and zone roasting and
smelting operations. The five largest producers of byproduct
sulfuric acid were American Smelting and Refining Company, Magma
Copper Company, Kennecott Copper Corporation, Phelps Dodge Corp-
oration, and St. Joe Minerals Corporation. Together, their 13
plants produced 68% of the output during 1975.
Contained sulfur in pyrites, hydrogen sulfide, and sulfur
dioxide amounted to 3% of the total production in all forms during
QQ
1975. Pyrites were produced by three companies at three mines
in three states; hydrogen sulfide by four companies of five
plants in four states; and sulfur dioxide by two companies at two
plants in two states. The three largest producers of these
products were Cities Service Company (pyrites, hydrogen sulfide
and sulfur dioxide), American Smelting and Refining Company
(sulfur dioxide), and Shell Oil Company (hydrogen sulfide).
Together, their one mine and five plants accounted for 93% of the
contained sulfur produced in the form of these products.
The trends in the sulfur industry in the United States for
the past twenty-five years is given in Figures X-l, X-2, and X-3.
The sulfur produced and shipped from Frasch mines in the United
States, by-product sulfuric acid produced in the United States,
and pyrites, hydrogen sulfide and sulfur dioxide sold or used in
United States is given in Tables X-3, X-4, X-5, and X-6 respectively.
The trends in the sulfur production and consumption in the United
States for the past twenty-five years is given in Figures X-2 and
X-3 respectively. The sulfur supply-demand relationships and
end-use for the years 1964 and 1974 is given in Table X-7.
92
Arthur D. Little gives the U.S. sulfuric acid capacity
for 1975 in Table X-8. As indicated in Table X-8, U.S. sulfuric
458
-------
c
o
a
c
o
u.
3
— < — \ — \~"\ i I I I I i I I i I I I I
I I i
19SO
1955
1960
1965
1970
1975
198C
c
"5 o
11
" 1
S g
O. *o
•u
50
40
30
20
10
0
9
F.O.B. mi"e or plont .*^t A
- „ x* \ /
.^•n*««»w ^ ^ *
,«>•» "••«*,•** *% ^
-
(ill
1 1 1 ' 1 ! 1 1 1 1 1 1 I J 1 1 ! 1 1 1 1 1 1 1
1 I 1 1
1950
1955
1960
1965
1970
19X5
1981
Source: Reference 39
Figure X-l. TRENDS IN THE SULFUR INDUSTRY IN THE UNITED STATES
459
-------
o
o
c
JO
o
t/l
Domestic fraseh
i i (
1950
1955
1960
1965
1970
1975
1980
o
|
a
u.
_l
100
90
80
70
60
50
40
30
20
10
Domestic frasch
Domestic other forms
UtlllllMintlMIIMItllllllll Illllllt
r^-T--r-T""r i i ]
Imports
Domestic recovered
1950 1955 1960
Source: Reference 90
1965
1970
1975
1980
Figure X-2. TRENDS IN THE CONSUMPTION OF SULFUR IN THE UNITED STATES
460
-------
.0
1
=>
Recovered
Other Forms
_J I I I I ! I I
1950
1955
1960
1965
1970
1975
1980
100
90
BO
70
60
50
40
30
20
10
Frotch
Other Forms
Recovered
1 J II I I I I
j I i i i i
j i 1
1950
1955
1960
1965
1970
1975
1980
Source: Reference 91
Figure X-3. TRENDS IN THE PRODUCTION OF SULFUR IN THE UNITED STATES
461
-------
UJ
£
^^
CO
a
uf
i— i
z
=>
UJ
mr"
£
^"•^
Z
I-H S-
ro
OOr-
Ljj r-
Z O
1-1 -a
•o
31 C
o ro
oo
5 ~*
U- -C
•M
ll
1 1 f0
O CO
UJ C
0- O
a. 4->
oo c
o
O i—
O 3
=3 0
a j=
§t
o.
a;
ii
_i
Z5
OO
ro
i
X
at
_a
n
c
a
e
•H
CO
iction
>-f
•n
o
M
*
"1
a)
,3
(4
>
^
4J
•H
4J
S
3
^*
^•^
•-•1
r-l
cd
O
H
1
3
O
01
s
01
H
at
4)
SH
sr m
O1 CO
00 CO
f*. CM
iH i-l
CO CO
r*^ ^o
VA f^
*w r^
in o
CM &
0 CM
r> r—
\O -4-
i-l co
vO i/l
CO CO
00 >O
O in
m, «
CO CO
1 1
1 1
! i
1 1
j j
j I
j j
1
iH CM
r^ r»
H H
CO O CO
m o oo
•M <->.•&
CO *f O
^ CM CO
oo oo r*
-* 00 O
rx r*. \O
m rH iH
0 O r-l
SO O>. CM
r<- r^ r»
rH 00 O
1-1 ° C
ro ro ^y
f^ CO fO
ST CO iH
C\ ON -4-
» M 0
<»• *T -*
1
j
! i
i i
i i
1
j
j
i
j
j
co -^ m
l-H r-< H
•
2
£
•o
g
0
^J
B
s
T3
C
0)
a,
4)
•o
^H
O)
u
1
q-
(U
a:
* •
a>
u
i-
3
O
«/)
462
-------
CO
UJ
r™
^t
CO
0
UJ
1—
IJ
iff
^3
UJ t/>
re j-
,—
^P» _„
i-i 0
•o
o
UJ -O
a. c
o. to
•— i
CO O
._ "^
E~ T*
^z
^C "O
O (O
UJ
O (/)
o o
O +J
a:
a. a>
c
ID f~-
1 1
CO
[^
H
(0
~J
id 00
-» O 00 ui CO
r^ O r» O
»-l f^ rH CM
ON t-l CO -» VO
*. * * A
iH tH F"! t—l
•»^
rO|
CO O\ vO tf\ O4
en H •* r-i -s-
NO 00 O CM Cl
f~4 tH rH
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
1 1 1 1 1
: : : i
! ! i i
I I I !
i
i i i i i
i i i i i
tH CM en •» IA
C< ON S\ ON ON
r-l t-l t-l rH iH
,
00
•H
•O
C
3
O
4J
41
•O
4>
S*
1
•rl
O
0)
3
•O
|
01
01
,_!
o
u
o
4J
1
JJ
g
(d
4J
«t
Q
.
m
§
***!
-------
co TO
U- UJ CO TJ M
i tij X; c *J
=> < o * C
co j— .c _ J5
" * *" "*"* W CX
~~ )^ (/)
—1C cu u
=> Z O Q. C
co i— 1 1— o. nj
O iH
h- T3 U O.
O C
=> (O
O CO
O 3
f\^ O
a. ^=
i i—
OQ
LO
1
X
» in o \D
C7\ C^ P^ f*^ l^
CN OO H ro CTi
* " " " *
«N T CD l/^ vD
m 10 vo vo r-»
-* H 0-4 rH vO
oo i/> oo oo rH
CO OS rH ^» CsJ
«M CM in
1
i— I CN ro
rH t-l r-| rH ^H
CO
cu
4J
CO
4J
C .
• CU M
CO O 3
C9 O rH
fH 0 3
M CO
CU CO
w cu en
B -H -H cr>
J^ 4J
fi >> n) nj
60 a d o
-------
a>
3
O r» co 01 t^
co e-4 oo m cr>
m «M ^H o O
•> •» •* *» *
o> r*. vo t^
t-j
vr
- co a c
a: rs M- o
i— -M
«o; 3
co o co en
^—• (^ ^J
Qi O '"'
>- CO -O
Q_ C
LU ra
• O
cu
o
o
CO
ON
r^. r-
ON ON
465
-------
T-
•^
UD
2
A
CO
0-
1— 1
3:
CO
1— «/>
«C c
_J O
LU 4-5
O>
O E
2 O
LU ~o
0 C
i re
Q. O
Q. .C
CO
ID
ii
_J
CO
*
1
X
Ol
XI
£
-
0
»
R
*•
£
O)
1
-
s
*~
i
s
i
I
1
O)
£
-
O
o
00
**
o
c
o>
CO
o
ff
0
U)
CM
*
ci
o
CM
CO
*-"
CO
sS
01 O>
CM
m
§
CO
§
a
8
«'
o
rt
§
S S *
*• *- o*
^ N ™
»~ CM
-- III ftl
o •-' V
§ss
Ol r-* CO
K r^ co
m 10 co
in cq rt
01 - CO
in in in
XOt if)
fw UJ
in « •{
^-" co"
co" »-" V
88.8
K »• •»
K K
5
h«" CO" CM
S 8
0. 01
tft n
«) K
tf> CO
fc
K
*•"
2
"~
I
0
s
CM
0
J
CO.
O*
+
|
S
§ ISS?
«r ^
""
» «
t CO
r.
3
*"
oi
a
oi
P
s 85*?
o n
9 18
„ *•
CM
¥ ft
*H **
"~
sig
O*l»»" +
St 8|
_r »- . o
Wo 9 "i
S 3 | 3 °> -j
s -. e s ;3 <
J O y M *. M
(£ Soil |i
f S
U O
1
*
UJ
m ^r
y> xJ
rJ 1
j
£
§289885^8
Oi
* CM
S888SSSSR!
^« mcO'»C'JCM«-'-CN(
V CM
in Q !'^r<>
v, *n^incocM»-CM —
^ CM
u>w>* S CM R * CM o»
o »•"
eo" ^
^
s
•5
T
^ I
~ S o
H
o
X
o
CD
O>
O
O)
S
Ol
o>
&
o>
a
Ol
?
Ok
i
in
r*.
5
a
c
^ 1 ?'
1 Is >.
1 -«i ^
s •* ; s 1
I IfliflpSfe I
i V^iL5l?ZuS^5O
w
*
2
w
i
i
X
1
•s
§ ,
i O
« o;
(0 O
=> C
S S_
i ^
2 Ol
i Qi
1 «
•6 O
^ 3
O
3 CO
466
-------
Table X-8. U.S0 SULFURIC ACID CAPACITY - 1975
Producer Annual Capacity
(Thousand tons)
A. Northwest 2,158
B. California 1,702
C. Southwest 3,787
(Western U.S.Subtotal 7,647)
D. Other U.S. 38,748
46,395
Source: Reference 102
467
-------
acid capacity in 1975 was approximately 46 million tons. The
vast majority of this capacity (84%) is in the Gulf Coast and
eastern U.S. The Northwest, the Pacific Coast, and the Southwest,
together accounted for only 16 percent of the U.S. sulfuric acid
capacity.
United States Department of Commerce in their Current Industrial
93
Reports report sulfuric acid production in U.S. by area for the years
1973 and 1974. This production distribution is given in Table X-9.
There are a wide range of government programs that are designed not
only to increase the production of sulfur but to protect the environ-
ment from the effects of sulfur dioxide emissions and to develop new
94
uses for sulfur. The Bureau of Land management has a leasing
program for native sulfur on public lands in Louisiana, New Mexico,
and the outer continental shelf. The Office of Minerals Exploration
lends up to 50% of approval costs for sulfur exploration. The Bureau
95
of Mines is conducting research on the recovery of sulfur from
smelter gases and industrial stack gases. The Department of Energy
is doing extensive research and development on gasification and
liquefaction of coal. The sulfur recovery potential from these
processes would be very great if commercialized. The Environmental
Protection Agency has a wide range of research programs primarily
aimed at reducing sulfur dioxide emissions. The Bureau of Mines
has a research program aimed at developing new uses for sulfur of a
magnitude that would alleviate a potential over supply problem.
X.5 FUTURE MARKET PROJECTIONS
World sulfur or "brimstone" trading is expected to maintain a
steady growth between 1977-80 to reach 21,2 million tons. The higher
rate of growth up to 1980 is largely attributable to the anticipated
recurrence in the strength of demand from the fertilizer industry.
Almost 57% of the projected 12.7 million tons increase in brimstone
output between 1976 and 1985 is forecast to be in the form of recovered
(not mined) sulfur. In fact, for the first time recovered sulfur
output will exceed that of Frasch and native refined sulfur.
468
-------
Z3
(O
p
a
•
a
i.
•r
g
«
«
a
IN
r
•
t
7
•4
t
j
S
PI
S
•rt
i
:
I
•
a*
|
.
«
f
O
*ff
!S
r
IN
•
p
J
O) V Ol
1* aff ,TI
!r* t* 01
48 rt V
t« tO *
?
1
1 " t
^1 ."
e *
C f 1
if ;
•i
o in r- o
(0 ?i If N * N Is1
ip H m H N f-i n
«oaDaeti»orsAO)«r» o
HsSslpsss i
A
MMiiiJi
::::»£:,•
JilBis^ai"1-
i 5S=*Jis*3< ]
ssSS
w o * n
§;«g
....
• S u.!
• 2
SS
•* *» a
! |5
i "i
I i1.
ause of independent rounding J
•rated plants-
chusetts, and Rhode Island.
inneeota, Iowa, and Kanaas.
a, Kentucky, Tennessee, Missll
nt«na, WyosUnc, New Mexico, rt
iirurvs Bay not add to totala bee
r Government-owned, privately op
ir plants located in Maine, Ha*aa
>r plants located In fti«oonsin, H
ir plants located In West Vlrglnl
r plant i located in Colorado, Mo
,4 «*«**•*•«»
* slllll
CO
o
-------
Canadian and French recovered sulfur output is projected
to decline by 2.2 million tons but this fall will be more than
compensated by an increase of some 3.9 million tons in the Arabian
105
Gulf recovered sulfur output. Frasch and/or native sulfur mining
in Mexico, Poland, Iraq and the USSR will maintain a steady growth.
The brimstone supply/demand situation in the 1980's may give rise
to large scale exploitation of sulfur deposits in other areas, such
as Bolivia.
In the September 27, 1976 issue of Chemical Engineering, the
conflicting views of experts on the sulfur supply and demand
situation in the coming years is presented. Marked change in the
balance of supply of brimstone from mining, hydrocarbon sweetening
and pollution control, as well as restructuring of the market for
the material, complicates sulfur's outlook over the next decade.
Controversy over sulfur is not a new phenomenon. While the
material finds customers in a broad range of sectors within the
chemical process industries and elsewhere, its demand has remained
cyclical, tied to the fortunes of the key-market-phosphatic
fertilizers. Now, the picture is clouded by a number of new
factors.
On the surplus side: existing world stockpiles of a record
26 million tons of surplus sulfur; the likelihood of large ton-
nages of recovered sulfur being produced in the Middle East as
a result of sweetening operations at natural-gas-exploitation
project underway there; the probability of more sulfur being
recovered as a result of U.S. and Western European environ-
mental legislation targeted at power-plant stack-gas emissions;
the virtual certainty that more high-sulfur crude will be proces-
sed in the U.S.; and the expected growth of recovered sulfur
supplies following commercialization of projects for synthetic
natural gas, coal liquefaction, shale oil and other alternative
107
energy sources.
470
-------
On the shortage side: predicted declines in sulfur associated
with dwindling gas-fields in currently strong producing areas such
as France and Canada; doubts about the rate of growth of recovered
sulfur in the next decade; lack of strong economic incentives for
continued growth of Frasch (mined) sulfur production in U.S.;
reservations about the marketability of small tonnages of recovered
sulfur from scattered individual sources; and rapidly inflating
capital costs for Frasch facilities, recovery units and shipping
I QQ
and transportation.
On the consumption side, some of the sectors in industry will
cut their sulfuric acid requirements and consequent demand for
suflur. In the wood pulp industry, there is a swing to making less
use of the sulfate process. In the paint industry, there is a swing
to making titanium dioxide pigment by the chloride route, phasing out
the sulfate route (with its byproduct problems). And in steel pickling.
hydrochloric acid is rapidly replacing sulfuric acid.
One expanding industry that will call for more sulfuric acid
is non-ferrous-metal production (especially copper and uranium),
where annual growth of sulfur demand will probably hold at 4.2%.109
Most of the sulfuric acid in the copper industry will come from
recovery plants at the smelter.
Recovered-sulfur output figures are a strong factor support-
ing theorists who predict a surplus. In the U.S., according to
one industry source, elemental sulfur from petroleum and natural
gas operations may mount to 4.6 million tons annually by the end
of the decade and continue to grow in subsequent years, perhaps
reaching 4.9 million tons/hr by 1985, in another, more conservative
estimate.110
But at last October's London meeting of the European Chemical
Market Research Association, Martin Horseman, a director of the
British Sulfur Corporation (London) reported that the world growth
in recovered sulfur would be far less spectacular during the next
ten years than it was during the 1965-74 period. He added that
471
-------
between 1975-85, sulfur production from hydrocarbon process is
anticipated to increase at a rate of some 3.6% per annum, compared
with a rate of 11.7% per annum between 1965-74.
112
Texas Gulf's Rittenhouse, speaking at a May meeting of the
U.S.'s Chemical Market Research Association (CMRA) predicted that
byproduct sulfur production in U.S. may total only 8 million tons/
yr as compared to earlier preductions of 20 million tons.
113
U.S. Bureau of Mines on the other hand, is still confident
that recovered sulfur will continue to carve itself a large share
of the worldwide market. Subject to suitable technology emerging,
and continued environmental constrains, the Bureau predicts that
co-product sulfur output in the U.S. will rise from 31% of total
production to 83% by the year 2000, accounting for 87% of U.S.
supplies.
The big question mark hanging over recovered-sulfur output
is that of environmental legislation. The basic imponderables
here are: When will coal gasification start to make inroads as
an energy source, and how much sulfur will be recovered in a
usable form from stack gases, rather than lost because of the use
of cleanup technology that discards sulfur values?
114
Exxon estimates that, by 1985, coal gasification will
contribute a maximum 200,000 tons/yr of sulfur, and shale oil will
yield only 100,000 tons/yr because sulfur in shale oil is mainly
retained in the shale as sulfate. Texas Gulf and the Sulfur
Institute are equally pessimistic, seeing these sources developing
slowly, too. Texas Gulf's Rittenhouse wonders whether environmental
legislation will have much effect on sulfur recovery at power plants,
since at the present there are hardly any commercial power plants
that are recovering sulfur.
Figures from the National Coal Association (Washington, D.C.)
based on Project Independence projections, show that 1985 total
coal consumption will reach 1,140 millon tons against 1975's
620.2 million tons. Electric utilities will be burning at least
472
-------
715 million tons/yr of coal in 1985, and more probably as much
as 826 million tons, says NCA. But how much sulfur will be con-
tained in the coal, or how much of this material will actually
be recovered, no one is inclined to speculate about.
Canadian sulfur production, from Western Canadian sour-gas
plants, will decline as a major source, all authorities agree.
Output of Canadian sulfur peaked in 1973-74 at just over 7 million
tons/yr, and fell in 1975 to 6.5 million tons. Production will
drop to 5.7 million tons/yr in 1980 and 3.7 million tons/yr in 1985,
Texas Gulf estimates. The chances of new discoveries halting this
slide are slim too.
The impact of increased Middle East recovered-sulfur output
is difficult to estimate. Projects to harness previously-flared
natural gas will start up in the 1980's, turning out large tonnages
of sulfur too. In Saudi Arabia, for example, Aramco plans sulfur-
118
recovery facilities turning out 7,000 tons/d. Rittenhouse pegs
current production in the area at 1.4 million tons/hr, mostly in
Iraq, which has a domestic Frasch industry, and Iran. He estimates
that production should double by 1980 and that by 1985, output
could be as much as 4.0 million, but will certainly reach 3.0
million tons.
The U.S. Frasch industry has viewed the growth of recovered
sulfur with concern intensified by soaring Frasch process energy
costs, and increasing capital costs for new mining and trans-
portation facilities. Increasing costs or exhaustion of reserves
have closed ten mines in the 1970's, including five within the
last year.119
Predictions for future Frasch output vary, but none are
120
optimistic. According to Exxon, U.S. Frasch production may
fall to 6.6 million tons/yr or less by 1985. Another source
predicts an even sharper falloff: output at 7.4 million tons/
yr in 1980, dropping to a tentative 5.7 million tons by 1985.
121
The views of Westinform Service Ltd. on sulfur also tend to
support reduction or minimum expansion of Frasch.
473
-------
Forecasts of U.S. and rest-of-world sulfur demand for 1985
and 2000 is given in Table X-10. The U.S. sulfur demand in 2000
122
is forecast to range between 18 and 26 million tons. The
probable range within this range is set at 14.5 million tons in
1985 and 23 million tons in 2000. This represents a probable
average annual growth rate of 3%. The growth rate of the probable
demand in the rest of the world is expected to be somewhat greater,
about 3.7%. This faster rate is due to relatively more rapid
industrial expansion in the developing countries of the world.
In summation, the most probably world demand in 2000 is estimated
to be 110 million tons at a growth rate of 3.5%.
f
The demand forecasts for domestic end uses are shown in
124
Table X-ll. They were developed by U.S. Department of Interior
by applying the growth rate of selected economic indicators to the
patterns of end uses and projecting to 2000 to establish a fore-
cast base for each use. The economic indicators used for estab-
lishing the forecast bases were as follows: phosphate rock con-
sumption in fertilizers as an agricultural indicator; population
as a paper products indicator; and the Federal Reserve Board
Index of Industrial Production as an indicator for all other end
uses. In the case of paints, and iron and steel production,
neither a forecast base or forecast range was established because
it is anticipated that these uses will be phased out by 2000.
The sulfur supply pattern is expected to be drastically
restructured by 2000. Production from primary (Frasch) sources
will be gradually phased out and replaced by production from
125
environmental-related co-product sources. Table X-12 shows
an assessment of domestic co-product sulfur production and
potential for 1974 and 2000 by seven types of sulfur sources. For
1974, it includes the actual production from these sources and the
estimated potential capability for production. For 2000, it
includes an estimate of the probable forecast production from an
estimated potential capability for production.
474
-------
o
o
o
cvj
CO
U.
00
a
I
u_
o
is:
*«!
• o ?
oo $-£.
• 3
Z3 O
OO
00
oo
5
UJ
a:
o
O
>-
oo
x
a>
§.§ §| |§
w ?5 co o *- rt
?8" S'8 SSJ
»- H to
88 88 83
^ ^ h. V- r~
*
I
o
o
o
PJ
TJ
(O
CO
r*.
en
LU
00
oo
oo
00
o
LU
•-3
§
a.
X
-------
O (O f-
O ••- 1—
O 4-> -i-
CM c .a
Q.
z o (d
> o
o^ **"**
i— 0
CVI
o c
U_ +J O
_i IS -Jj
4 * * • % »^_
p • ^"* T^ ^^^
I— c ^a
o en 0 1— 3
v_^ T3
r- 0
0 S-
=3 Q-
O
Q£
a.
1
O
O
CM
'T
X
(U
JD
0 O 0 0 0 0 O
O O CM CO LO <3- VO
vo r^ ^^ o^ LO n^
»l •» »t *t
O CO i— CO
CO
o o o o to LO LO
O O O LO O CM CVJ
CM LO r— LO LO r-
* A *\ *
CTl LO r- CO
0 0 0 0 O 0 0
O CD 00 r**> O CD VJD
CT^ r*x CM CO CO
•t *t A fv
O LO r— r—
CM CO CM CM CT« LO
1 VO i — LO M- CO
<• CM LO CM
rt n
^~ ^^
o
o
f—
*l
LO
^~
o
o
o
#*
o
CVJ
o
o
CM
9*
o
CM
CO
r-"~
IT)
n
en
i/)
$_
3
M-
lo
W) i — I —
fO fO '^
E to +J +*
3 0)0
tt) r— E 1—
•— (0 S-
2S-
O. C «O -C
O O> «O O -i- (!) 4->
u a. z o M _i o
CT^
»
M-
Q^
a:
O)
o
S-
3
o
00
476
-------
The forecast production for 2000 is based upon the ongoing
development of the required technology for the production of
marketable sulfur or sulfuric acid from the various sources, and
environmental restraints that will enforce the production of
these sulfur products regardless of normal economic considerations.
During 1974, the co-product sulfur production was only 17% of the
potential capability and ranged from nothing (in case of coal) to
I p£
95% (in the case of nature gas) By 2000, it is forecast that
this type of sulfur recovery will increase to 44% of the potential
capability. It will range from 30% of the estimated potential
capability for coal to 98% for natural gas. Coal is clearly
indicated as being the principal source of co-product sulfur in
2000, accounting for 46% of the estimated forecast production,
followed by petroleum with 27.5%.127
In 1974, coproduct sulfur production accounted for 31% of
the production of sulfur in all forms and met 32% of U.S.
128
demand. By 2000, it is projected that co-product sulfur pro-
duction will account for 83% of U.S. sulfur production in all
forms and supply 87% of U.S. demand. With forecast domestic co-
product sulfur production amount to 20 million tons per year by
2000, the remaining 3 million tons per year required to cover
domestic requirements can be obtained as domestic native sulfur
by the Frasch process, by imports of recovered elemental sulfur
from Canada, and by importing Frasch sulfur from Mexico.
The foreseeable changes in the sources of future sulfur
production have important implications that concern both producers
and consumers. It will certainly change the existing marketing
patterns and price structure to a point where they will be entirely
different from what they are now. It is predicted that the net
effect of these changes will be to break down the sulfur-producing
and sulfur consuming industries into regional segments, each with
its own supply-demand relationships that will be largely independ-
ent of other regions.
477
-------
Westinform Service Ltd. (London) in a report entitled
129
"Sulfur '85" predicts an acute brimstone shortage of 8.8
million tons, almost 15% of the potential demand, by 1985 due to the
inability of producers to match a rapid growth in world demand.
They add that constraints on Canadian exports may cause this deficit
to occur even earlier perhaps by the late 1970's or early 1980's.
The world brimstone shortage by the regions for 1985 is given in
Table X-13.
The tight brimstone supply conditions can be expected to give
rise to a rapid increase in demand for imported sulfuric acid of
smelter or pyritic origin, and/or the resurgence in the use of
pyrites and hence trading in pyrites. It is predicted that sul-
furic acid trading could increase from 1.5 million tons in 1976
to 2.1 million tons in 1980 and 6.3 million tons by 1985.
Table X-13. WORLD BRIMSTONE SHORTAGE BY REGIONS: 1985
West Europe
East Europe
Africa
Asia/Oceanic
South America
North America
Middle East
Total
Volume (1000 tonnes)
2608
259
1140
1564
939
1984
325
8819
Source: Ref. 129
Cost and Pricing
Operating factors in the sulfur industry are diverse be-
cause of the widely differing manners in which marketable pro-
ducts are produced and sold. Basically, the industry is divided
478
-------
into two sectors. In one, the production of sulfur is the sole
objective, as typified by the Frasch process; in the other, sulfur
or sulfuric acid is recovered as a byproduct or coproduct, as in
the case of the natural gas, petroleum, and nonferrous smelter
industries.
The Frasch sector of the sulfur industry is based on the
orderly exploitation of discreet deposits of native sulfur, with
the objective of obtaining as nearly a complete recovery of the
resource as economic conditions will permit. This requires an
assured market and a stable, viable price structure that will be
attractive to both producers and consumers. Under normal conditions,
the assured market is based on the relatively close proximity of
these deposits to the Nation's fertilizer production centers, and
the availability of cheap water transportation to these points.
The major items of operating expense in Frasch production
are the fuel for heating the process water, chemicals for treat-
ing (softening) the water, and the cost of drilling wells for
extracting the sulfur. The major possible constraints to a
sustained orderly development of the Frasch sector of industry
would be a long period of depressed prices and a continued in-
crease in fuel prices. This would restrict production to the
higher grade sections of operating properties, and the lower
grade ores probably would be lost.
The operating factors in the co-product sector of the
industry are much more complex. This is because the sulfur
revenues represent only a small portion of the revenues from the
primary mineral production. In fact, the sulfur production may
be enforced by the need to remove sulfur from the primary pro-
duct so as to be able to market it or by environmental restraints
on the release of sulfur compounds. Additionally, there may be
no ready market for the sulfur or sulfuric acid that may be
produced. Under these conditions, the economic desirability of
producing sulfur may be subordinate to the necessity of producing it.
479
-------
The major expense in the production of co-product sulfur is
the amortization of the large capital expenditures required.
Operating costs will range from low to medium, depending upon the
type of feed to the recovery plants.
The U.S. Bureau of Mines in their Mineral Industry Survey
gives the average net shipment value f.o.b. mine/plant for Frasch
and recovered elemental sulfur of $45.63 per ton in 1975. The
year-end price for Frasch sulfur was $65 per ton. There were
corresponding increases in both export and import prices. The
average sales value of shipment of Frasch sulfur f.l.b. mine for
both domestic consumption and exports during 1975 rose to $50.16
per ton. The average price of recovered elemental sulfur was
lower than Frasch sulfur. As a nondiscretionary byproduct, there
was a general tendency for the industry to sell in local markets
at prices that were competitively lower than sulfur from other
sources. Prices varied widely in different regions of the nation
as a resulf of these competitive factors. Reported unit shipment
values of recovered sulfur f.o.b. plant in 1975 were $36.14 per
ton. The time-price relationship for sulfur from 1954 to present
is given in Table X-14.
The prices of sulfur and sulfuric acid as reported by the
130
Chemical Marketing Reporter dated October 25, 1976, are as
follows:
• Sulfur crude, molten, f.o.b. vessels, Gulf port - $60-61/ton
• Canadian f.o.b., Alberta for U.S. delivery - $25/ton
• Dark sulfur, Ex-Tampa, Florida - $65-66/ton
• Sulfuric acid (100%)
East Coast - $47.70-53.25/ton
Gulf Coast - $44.95-50.25/ton
West Coast - $49.75-55.307ton
Other areas - $32.60-53.40/ton
These prices are not firm indicators of actual prices paid,
however, since discounts, variability in credit terms to buyers,
480
-------
Table X-14. TIME-PRICE RELATIONSHIP FOR SULFUR
Average annual pnce. dolltrt per long ton
Ye«r
Actual prices Constant 1973 dolars
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975'
1976 • (first quarter)
26.65
27.94
26.49
24.41
23.82
23.46
23.13
23.12
21.75
19.99
20.19
22.47
25.77
32.64
40.12
27.05
23.14
17.47
17.03
17.84
28.88
46.50
55.00
45.67
4744
43.50
38.62
36.76
35.60
34.57
34.10
31.70
2876
28.64
31.30
34.92
42.83
50.59
32.55
26.42
1907
17.96
17.84
2618
3872
4355
• Estimate • Prelminary.
' Frasch and recovered sulfur.
Source: Ref.131
481
-------
and service fees combine to determine the realized price available
to the producer. Local situations can result in lower prices.
Sulfuric acid production has primarily been sited adjacent to
the consuming industries it serves. Because of the high shipping
costs relative to the price of sulfuric acid, most sulfuric acid
is consumed within several hundred miles of the producing point.
Only with the recent advent of regional inbalance in sulfuric
acid supply and demand created due to the stringent air pollution
regulations and high Frasch sulfur price, has more distant rail
shipment and to some extent, ocean transport, became more important.
The selling price for acid can vary considerably depending
upon the seller's situation and also the buyer's. In the case
of the seller, if acid is being produced as a resulf of pollution
control techniques then the quantity available for sale is dependent
upon production of other products rather than the acid market. Thus
if the seller can arrange for a long term (and usually low selling
price) contract, he would prefer this rather than trying to change
his customers depending upon the acid market price at any current
time. As mentioned, the cost of shipping sulfuric acid is very
high in proportion to the selling price and therefore, its shipping
range is limited. Currently 500 to 1,000 miles is the maximum
shipping range.
The smelters in the southwest of the United States are sell-
ing acid in the $8-15 per ton range. Long term contracts and/or
increased acid availability could reduce this price. Smelter
utility and refinery acid projected increase in supply seems to
indicate reduced prices unless proposed market demand and/or
Frasch sulfur prices change considerably.
482
-------
APPENDIX Y
BACKGROUND MATERIAL, ASSUMPTIONS, AND CALCULATIONS
FOR BLENDING SCENARIOS
Y-l REVERBERATORY FURNACE S02 OFFGAS CHARACTERISTICS
Reverberatory SOp offgas concentrations fluctuate due to inter-
mittent calcine charginq. The amount of fluctuation varies from
smelter to smelter depending on both age of smelter and local smelt-
ing practice. Since this report deals with a new smelting unit, it
follows that many of the problems with older smelters will be elimi-
nated, i.e., excess air dilution, large fluctuations, leaks, etc. As
a result, the fluctuations will be minimized while S02 offgas strengths
will be slightly increased.
Hith the above in mind, it was decided to choose an example
smelter that has minimal fluctuations and higher S02 strengths to
serve as a model for the RF offgases used in this study. The Onahama
in Japan is reoresentative since it was constructed relatively re-
cently. Although they process green charge, the RF offgases were
used to serve as a basis for the new smelter considered in this study.
By using a green charge smelter, larger S02 fluctuations might be
experienced than would be the case for a calcine charge RF. However,
this tends to add conservatism for the S02 profile chosen for this
study.
Figure Y-l shows the actual S02 emissions from the Onahama
smelter versus time. As can be seen, these fluctuations are not as
great as some (see text, Figure 2-4).
Figure Y-2 shows a rough aonroximation of such emissions that
were made so calculations could be carried out more efficiently.
483
-------
The above figure (Y-2) was used to calculate emissions from the
new smelter under consideration for this study. Since the adjusted
Onahama's average S02 concentration is 2.5 percent, the average con-
centration chosen for this study (1 percent) was scaled down by a
factor of 2.5 percent to be more representative of U.S. practice.
The results are summarized in Table Y-l.
484
-------
(O
>
sr rd
o s-
•r- 0)
co -Q
CO 5-
•r- <1J
E >
uj oj
CM
O
OO
a;
o
0)
o.
cu
485
-------
t-
€
S
vo
o
m
U3
"X~
v^
o
s
s
o
r^
n
o
o
n
o
r-
CN
o
T
CM
U-
01
4->
3
C
o
tO
X
o
S-
Q.
Q.
O)
CO
s_
0)
(O
O)
o
(S)
(O
CD
oo
-------
O
o:
o.
CXI
o
OO
CO
<
CJ3
O
>-
o
UJ
CQ
OL
I
>-
JQ
03
a
CO
a:
LU
LU
z:
CO
a
o
z:
z.
o
c
CU CM
PS
CU
trt
e 3
r- 1-
c
CU CM
0 O
S- CO
cu
cu
•r- C
"5
•u
c
CU CM
u o
cu
Q.
wl
cu
g£
i- C
s:
+J
c
01 CM
cu
CU 4->
I- E
^~ *r-
O
,—
o
3
o
LO
I-.
o
LO
CM
O
CO
CM
LO
1*^
O
,_
o
CM
CO
o
r-
O
r—
co
CM
O
CM
OO
LO
CM
O
r~ •
S;
o
o
S
o
o
CM
CM
•d-
CM
0
CO
CM
CM
O
CM
cn
o
,_
LO
CM
CO
O
r_
CO
CO
CM
LO
CM
00
LO
CM
LO
t-~
CO
oo
o
o
CM
CM
cn
o
LO
CM
LO
CM
0
CM
CO
CM
LO
CM
r~~
00
cn
o
LO
CM
g
o
o
$
CM
LO
CM
cn
CM
o
^—
cn
oo
0
LO
CM
o
,_
o
CM
CO
CM
CM
LO
CM
LO
CM
O
CM
O
,
0
"3-
cn
o
LO
i-^
0
cn
CM
o
CO
CM
LO
PN.
r—
co LO cn LO
o cn oo o
i— O O •—
O LO O CO
CO i— i— «J-
OOi— LOOCOOOCO^J-r- COr-CMCO
ocnooocoor-ocoocno
r— Or— r- r— Or— r— r—Or— Or—
LO
LoooOLOOLOoooor^LO
CMCMCMCMr— CMi— CMCOCMi— r—
LO 00 CO CM
CO CO CM CO
CM CM CM CM
O LO O CO
CO r— r— ^~
en r^ CM cnoiLO^trOLoroi— LO
COCMCOLOCOi — COOOLOl — COCOVO
CMCMCMCMCMCMCMCMCSJCVICMCMCM
CO
co o o o LO o co o o o o r** LO
CMCMCMCMr— CMr-CMCOCMi— r—
^.
^
LO
LO
00
CM
LO
487
-------
APPENDIX Y-2
CALCULATED GAS CHARACTERISTICS FOR ONE CONVERTER CYCLE (11 HOURS)
FOR SCENARIOS HANDLING RF OFFGASES IN A FGD SYSTEM
Table Y-2. BLENDED: MHR + RF (02 ENRICH) + CONVERTERS
Time
(minutes) Vol - S02
Time
(minutes) Vol - S02 S 02
488
-------
Table Y-2 . BLENDED: MHR + RF (02 ENRICH) + CONVERTERS
Time
(minutes)
Time
(minutes)
489
-------
Table Y-3. BLENDED: MHR + CONVERTERS
Time
(minutes) Vol - S02
Time
(minutes) Vol - S02 % 02
490
-------
Table Y-4. BLENDED: (MHR + CONVERTERS) + (RF < . • MgO SYSTEM)
Time
(minutes) Vol - S02 % 02
Time
(minutes)
Vol - S02 % 02
i
2
3
4
5
6
7
6
9
10
11
12
13
14
15
16
II
19
K>
21
22
23
24
25
26
27
26
29
30
31
32
33
34
35
36
37
36
39
40
f *
I'I
J
WSkc- t^l,
&
to.-it
^L
Ml?
I'fL
Vit.
:"?
<
-------
Table Y-5. BLENDED: (MHR + CONVERTER)+(RF
CITRATE SYSTEM)
Time
(minutes)
Vol - S02
Time
(minutes) Vol - S02 % 02
/MS ID -
-5
l»rjp-¥Ta
•Got
*
J-SSV
>V
'r
•i
>-Jtol
«»
/fi
"t
7*
h'rhJ
?-g?7
tft
/>L»
Tr'rD'
/K
492
-------
Table Y-6. BLENDED: (MHR + CONVERTER) WHEN RF HAS 02 ENRICHMENT
Time
(minutes)
Vol - SO,
Time
(minutes) Vol - S02 * S02
f?
$L*L
lol
&
ife
r^
J/,03
/£
Wefc-k
-590
LL
Li-
7*7
/b.TJf
n-
I*
1^
1±
tl>
_Ll±±
m
iiA
n
493
-------
Table Y-7. BLENDED: (MHR + CONVERTER)+[02 ENRICHMENT) ^MgO SYSTEM]
Time
(minutes)
Vol - SO,
(minutes) Vol - S02 X 02
494
-------
TABLE Y-8-. BLENDED: (MHR + CONVERTER)+[0? ENRICHMENT)-
SYSTEM] *•
CITRATE
Time
(minutes)
Vol - SO,
Time
(minutes) Vol - S02 % 02
i
2
3
4
5
6
7
8
9
10
11
12
13
U
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
3'
35
36
37
38
39
40
.V
ifU^-S
/Z*ft-tf
JW_
cJEl
^&p
S'n
75
-v
ite
-fZQ
'*'5"
-Ste
1-&M
LL
V ^ifo
if Mi
^
I I
j^TT
/v,/*/
±5.k4
dSfi
z£.
4.-X
/a:
495
-------
Table Y-9 . BLENDED: FBR + RF + CONVERTERS
Time
(minutes)
Vol - SO,
* 0,
Time
(minutes)
Vol - S02 X 02
1
2
3
4
5
6
7
e
9
10
11
12
13
14
li
16
17
18
19
20
21
22
23
24
2!
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
rf'r-
H£
Lief
iis
bw
TT
77
K*A£W
J8,
/n*te-l
7*
,1fr
4^
mux!-ft\tf
14
w
57
DO
irbUDLJ
t>t
M
»>-»/> ft*
i(it& PT
4.
I*.
•<•[•-
£g
71
i7
tel
4
3I7/
-i"*
(,*•
5?
-
496
-------
Table Y-10. BLENDED: FBR + (02 ENRICHMENT) + CONVERTER
Time
(minutes)
Vol - SO,
XO,
Time
(minutes)
Vol - SO,
1
2
3
4
5
6
7
e
9
10
11
12
13
14
15
16
17
IB
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
36
40
I
fjf -
'ILl
,5?
(&&
E
£Z
I-/37
.•rfl03
Ui
ii
I^K>."fl_Dk
.$7*
..,;
Ob
/£
-K/C5
IB
ft,'
4>£H
*-3Ll
fA
iyrt
W'>7
r7
il£
39
-WTtf
M*ft-^
17
^S
'0
:?
M
f
-CKLii
. n-t:
fH
•ft
S
¥6
-#9/
^
iHȣj
f
i?-
r-so?
7-
497
-------
Table Y-ll, BLENDED: FBR + CONVERTER
Time
(minutes)
Vol - SO,
Time
(mfnutes) Vol - S02 % 02
f.'LStk
m
H
i$Wr
SB
Z'±.l!L
Xi
li
ML
r2^
£*'+£
i
$k
ifauk J- ^
ijj
b'-s
£
itt>L'>
ti
tfta -1 *L
\ IJ.K
/J^'
/^j^T^J j
/^alfl?!-
£2
35
s/t
36
37
36
1
if
\'-i
39
40
H*#-\
498
-------
Table Y-12- BLENDED: (FBR + CONVERTER)+(RF—«-MgO SYSTEM)
Time
(minutes) Vol - S02
Time
(minutes) Vol - S02 % 02
i
2
3
4
5
6
7
e
9
10
11
12
13
14
li
16
17
18
19
20
21
22
23
24
25
26
27
26
29
30
31
32
33
34
35
36
37
36
39
40
^L-
i$4
0&
I I
1&.L
m
Wi
IkWL
•
L1A°---.Z
n*&i
'tt
-5
3S
:m.
:3
/a
c-
/£
£L^
/i
lT-l>
''i
iXHi
#a
li-
-,\t
'Alt.
;EU4
/^3
IT
ill
_ /^i^^
cl
499
-------
Table Y-13. BLENDED: (FBR + CONVERTER) + (RF——CITRATE SYSTEM)
Time
(minutes)
Vol - SO,
X 0,
Time
(minutes) Vol - S02 % 02
1- 5 /t
?
i
500
-------
Table Y-14. BLENDED: (FBR + CONVERTER - ADJUSTED FOR RF 02
ENRICHMENT)
Time
(minutes)
Vol - SO,
Time
(minutes) Vol - SO,
i
2
3
4
5
6
7
e
9
10
II
12
13
14
15
16
17
IB
19
30
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
36
39
40
1.10
/f
l-^ici-r
/v'
/'t
&
'•L&
«/ W
1 n./
fs
_&i
w
'^
La?
-J5
r?
1 7*P
"bffi
1Q\
djti
tk'/ff(
W-r
fi
'j&fa
i-
Mr
^L
^fda
-•5pJ
$7
/Y
jftti-i
/**•/
±»
a.
3.
S
'sw
JL
^
501
-------
Table Y-15. BLENDED: (FBR + CONVERTER)+(FR(0~ ENRICHMENT)——MgO
SYSTEM) L
Time
(minutes)
Vol - S02 * 02
Time
(minutes) Vol - S02 3! 02
:&SS.
-------
Table Y-16. BLENDED: (FBR + CONVERTER)+[CL ENRICHMENT ^CITRATE
SYSTEM]
Time
(minutes) Vol - S02
Time
(minutes) Vol - S02 X 02
i
2
3
4
i
6
7
8
9
10
11
12
13
14
li
16
17
IB
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
'Zli
mo-,5
li.
j'/
r.
.at>
\i-ki
I
fr.frg
*}
t&
M:
l-Kt'-A-
\i*>
£,:
|B/7
Eiti
1C
o
Itsise
fML
*.<
'?/-
/
503
-------
APPENDIX Y-3
ASSUMPTIONS FOR BLENDING SCENARIOS
(TABLE 6-4 IN TEXT)
SUMMARY OF ASSUMPTIONS FOR TABLE 6-4 IN TEXT
A. BACKGROUND
1. SO- offgas concentrations are constant for multihearth
roasters (MHR) and fluid-bed roasters (FBR)
2. Offgas volumes are constant for reverberatory furnace
(RF), MHR, and FBR
3. Offgas volumes and S02 concentrations fluctuate for
converters. Also S02 concentrations fluctuate for RF.
4. FGD concentration systems are assumed to be able to
process fluctuating offgas volumes and S02 concentration
and to produce an enriched offgas of constant volume and
S02 concentration (FGD system design based on maximum
volume input). Thus, the offgas volume and S02 concen-
tration from the FGD system will be based on the average
offgas volume and S02 concentration processed in such a
system. Such offgas characteristics will be based on an
assumed 95 percent S02 and 10 percent S02 for the citrate
and magnesium oxide (MgO) system, respectively. The
corresponding FGD concentrated $62 product volume will
be calculated by:
A (B) ^ = Volume
Where
A = Average volume of gas processed
B = Average percent SO,, in A
C = FGD S02 offgas percent, and 0.9 = 90 percent efficient
504
-------
B. VALUES USED FOR THE VARIOUS BLENDING STREAMS BASED ON BACKGROUND
INFORMATION
Equipment
1 . No FGD Control :
Reverb Furnace
MHR
Converter
Fluid Bed Roaster
2. No FGD Control:
Reverb Furnace
MHR
Converter
Fluid Bed Roaster
3. MgO System: No
Reverb Furnace
MHR
Converter
Fluid Bed Roaster
4. MgO System: 02
Reverb Furnace
MHR
Converter
Fluid Bed Roaster
5. Citrate System:
Reverb Furnace
MHR
Converter
Fluid Bed Roaster
6. Citrate System:
Reverb Furnace
MHR
Converter
Fluid Bed Roaster
MAXIMUM
% S02 @ SCFM
M I N
% S02
I M U M
@ SCFM
A
V E
so2
R A
G E
SCFM
No Reverb 02 enrichment
1.14 54,000
5.0 43,000
5.4 122,000
9.1 23,600
0.
5.
4.
9.
86
0
0
1
54
43
40
23
,000
,000
,000
,600
1
5
4
9
.0
.0
.5
.1
54
43
82
23
,000
,000
,000
,600
Reverb 02 Enrichment
1.6 41,800
5.0 42,096
5.4 122,000
9.1 23,100
02 Enrichment
10.0 5,540.4
10.0 19,350
10.0 59,292
10.0 19,328.4
Enrichment
10.0 6,019.2
10.0 18,943.2
10.0 59,292
10.0 18,918.9
No Oo Enrichment
95.0 583.2
95.0 2,036.84
95.0 6,241.26
95.0 2,034.57
02 Enrichment
95.0 633.6
95.0 1,994.02
95.0 6,241.26
95.0 1,991.46
505
1.
5.
4.
9.
10.
10.
10.
10.
10.
10.
10.
10.
95.
95.
95.
95.
95.
95.
95.
95.
2
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
41
42
40
23
4
19
14
19
4
18
14
18
2
1
2
1
1
1
,800
,096
,000
,100
,179
,350
,400
,328
,514
,943
,400
,918
430
,036
,515
,034
475
,994
,515
,991
.6
.4
.4
.2
.9
.48
.84
.79
.57
.2
.02
.79
.46
1
5
4
9
10
10
10
10
10
10
10
10
95
95
95
95
95
95
95
95
.4
.0
.5
.1
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
41
42
82
23
4
19
33
19
5
18
33
18
2
3
2
1
3
1
,800
,096
,000
,100
,860
,350
,210
,328
,266
,943
,210
,918
511
,036
,495
,034
554
,994
,495
,991
.4
.8
.2
.9
.58
.84
.79
.57
.4
.02
.79
.46
-------
C. Acid plant designed for maximum flow to acid plant — rounded
to nearest 1,000 SCFM.
D. Maximum S02 emissions from acid plant based on 2,600 ppm (0.26
percent) for single contact acid plant and 650 ppm (0.065 per-
cent for double contact acid plant. Thus, maximum SOo emissions
will be based on maximum volume to acid olant.
E. SOo emissions from the FGD systems are based on a 90 percent
efficiency for the system; thus, 10 percent of the sulfur pro-
cessed in such systems will be released to the atmosphere.
F. The uncontrolled S02 emissions represent fuqitive emissions
based on, a sulfur balance for 1,400 TPD concentrate containing
30 percent sulfur:
2 Ib SO?
1,400 TPD x 0.3 S x = 840 TPD - SOo.
1 Ib S e'
Thus, for the two cases:
Case 1:
MHR: 43,000 SCFM at 5.0% S02 = 276.43 TPD - S02 = 32.91%
RF: 54,000 SCFM at 1.0% S02 = 69.43 TPD - S02 = 8.26%
Conv.: 82,000 SCFM at 4.5% S02 = 474.43 TPD - S02 = 56.48%
97.65%
This implies 2.35% fugitive emissions = 19.7 TPD - S02-
Case 2:
FBR: 23,600 SCFM at 9.1% SOo = 276.12 TPD - SOo = 32.88%
RF: 54,000 SCFM at 1.0% S00 = 69.43 TPD - S02 = 8.26%
Conv.: 82,000 SCFM at 4.5% SOo = 474.43 TPD - SOo = 56.48%
97.62%
This implies 2.38% fuqitive emissions = 20.0 TPD - S02.
Additionally, for Case Nos. 17 and 18 the fugitive emissions
are added to the uncontrolled reverberatory S02 emissions.
G. The total S02 in TPD charged to the system is 840 TPD - S02.
506
-------
APPENDIX Y-4
SPECIFIC ASSUMPTIONS FOR ADDITIONAL BLENDING SCENARIOS
(TABLE 6-6 IN TEXT)
All assumptions considered for Table 5-4 are applicable here
with the additional assumption that:
For Numbers 51-58, the strategy was to blend all offgas streams
and bleed a portion off to send to a FGD system. The volume of the
bleed-off stream was determined by calculating the amount required
to achieve an 8.0 oercent SOo stream which could be sent to the acid
plant. The 8.0 percent S02 is usually the maximum that can be pro-
cessed in an acid plant without adding dilution air; this strategy
results in a constant S02 concentrated stream of reduced volume.
The following equations were used:
For MgO = 10% S02 at X = 0.1X where X = concentrated S02 volume
of FGD system
Y = volume of bleed
Y (7\ j_9_ _ v stream to FGD
• I
F = volume of initial
blended stream
n ns - (F - Y) Z + 0.1 X
u-uo ~ (F - Y) + X Z = percent S02 of initial
blended stream
For Citrate: 95% S02 at X = 0.95 X where X, Y, F, and Z are
the same as above.
Y (Z) = X
n nfi - (F - Y) (Z) + .95 X
u.uo - (F - Y) + X
These equations assume that the volumes off the FGD systems are
507
-------
constant (95 percent S02 for citrate and 10 percent 3^2 for MgO) and
that a 90 percent efficiency is realized.
508
-------
SULFURIC ACID PLANT COST ANALYSIS
Capital Cost Analysis
Detailed cost analyses were prepared for two different acid plant
gas handling capacities. The analyses were made for double contact/
double absorption (DC/DA) acid plants handling 85,000 and 50,000 scfm
of gas at 4.25 percent S0?. From these costs, Figure 1 was generated
for DC/DA acid plants.
Annual Operating Costs
Figure 2 indicates the annual operating costs for DC/DA acid
plants. When calculating operating costs for two acid plants, i.e.,
when the feed volume is greater than about 110,000 scfm, annual
operating costs will not be double the value obtained from the figure;
rather, operating labor costs for the second plant must be subtracted
since it is considered that the same crew can operate two side-by-side
plants as one.
509
-------
10C
I/I
o
•o
en
i
•o
iS)
o
LU
O.
-------
10C
0\
r^
cr>
i
TJ
oo
O
O
0.
O
10'
9
(20)
I
14
(30)
I
19 24
(40) (50)
4.25% S0
I
S0
S0
S0
I I
I .
47
(100)
71
(150)
SMELTER GAS FOLW RATE, Nm 3/s (103 scfm)
Figure 2. Annual Operating Costs for Sulfuric Acid Plants —
Double Contact Acid Plants (DC/DA)
94
(200)
511
-------
REFERENCES TO APPENDICES
1. Treilhard, D.6., "Copper Smelting Today: The State of the
Art," E/MJ, April 16, 1973, Paqe P
2. Reference 1, Page Q
3. EPA Public Comment Summary, Primary Copper Zinc and Lead
Smelters, Tab B
3A. Weisenbera, I.J. and P.S. Bakshi, "Process Parameters for
Primary Copper Smelters and Their Effects on Arsenic Emis-
sions," PES, 1978, EPA Contract Mo. 68-02-2606, Task 8
4. Hayward, C.R., An Outline of Metallurgical Practice, Second
Edition, D. Van Nostrand Comoany, Incorporated, New York,
1940, Page 40
5. Reference 4, Page 35
6. Austin, L.S., "The Washoe Plant of the Anaconda Copper-Mining
Company in 1905," Trans. AIME, Volume 37, 1942, Page 66
7. Newton, J. and Wilson, C.L., "Metallurgy of Copper," John
Wiley and Sons, Incorporated, New York, 1942, Page 66
8. Reference 6, Paqe 466
9. Bray, J.L., "Nonferrous Production Metallurgy," Second Edition,
John Wiley and Sons, Incoroorated, New York, 1947, Page 139
10. Carpenter, B.H., "Nonferrous Smelter Studies: Theoretical
Investigation of Role of Multihearth Roaster Operations in
Copper Smelting Gas Blending Schemes for Control of S02,"
Environmental Science and Technology, Volume 12, Number 1,
January, 1978, Paqe 58
11. Townend, R., et al., Amdel Bull. 2, Austral. Min. Dev. Lab.,
1966
12. Vlingterharger, H., et al., J. Ore Min. Nonferrous Metal 1.,
27(5), 225-32, 1974
13. "Chemical Engineering Handbook," 5th ed., pp. 4-9, McGraw-Hill,
New York, New York
512
-------
REFERENCES (Continued)
14. Reference 11, Page 60
15. Reference 4, Page 46
16. Reference 11, Page 60
17. Reference 10, Page 143
18. Reference 4, Page 47
19. Weisenberg, I.'J'., and Seme, J.C., "Design and Operating
Parameters for Emission Control Studies: Phelps Dodge,
Douglas, Copper Smelter," February, 1976, EPA-600/2-76-036h
20. Weisenberg, I.J., and Serne, J.C., "Design and Operating
Parameters for Emission Control Studies: ASARCO, Tacoma,
Copper Smelter," February, 1976, EPA-600/2-76-036k
21. Weisenberg, I.J., and Serne, J.C., "Compilation and Analysis
of Design and Operating Parameters of the ASARCO, Incorporated,
Hayden Plant, Hayden, Arizona, for Emission Control Studies,"
November, 1975, EPA Contract No. 68-02-1405, Task Order No. 5
22. Matthews, J.C., et al., "SO? Control Processes for Nonferrous
Smelter," January, 1976, EPA-600/2-76-008
23. Boggs, M.B., and Anderson, J.N., "The Noranda Smelter,"
Trans. AIME, Volume 106, Page 189
24. Reference 6
25. Stankovic, D., "Air Pollution Caused by Copper Metallurgy
Assemblies in Bor: Volume 11 - Multilevel Roasting: Furnaces,"
Project Study Rep. PL-480, Number 2-513-1, Bor, Yugoslavia,
August, 1975
26. Oldright, et al., "Production of Ferric Sulfate and H2S04
from Roaster Gas," Trans. AIME, Volume 73, Page 87
27. Reference 1, Page S
28. Reference 11, Pages 57-62
29. Carpenter, B.H., et al., "Copper Smelter Emissions Control
Study," EPA Contract No. 68-02-1325, Task Nos. 43 and 60
30. Reference 6, Page 466
31. Reference 6, Page 466
513
-------
REFERENCES (Continued)
32. Reference 9
33. Laist, F., AIME Transactions, Vol. 106, History of Reverberatory
Smelting in Montana
34. Winkler, Mooney, Kuzell and Mounts, -Patent #2, 124, 865
July, 1938
35. Merkle & Associates, Inc. Engineers, Suspended Refractory Designs
36. Modern Refractory Practice, Harbison-Walker Refractories, 1961
37. Jan Reimers and Associates
38. Roserykranz, R.D., "Energy Consumption in Domestic Primary Cooper
Production," Information Circular 8698, U.S. Department of the
Interior, 1976
39. McKerrow, G.C., Gaspe Copper Mines Smelter, AIME Annual Meeting,
New Orleans, February 25-28, 1957
40. Reference 39
41. Staff, Canadian Mining Journal, International Nickel Co. Issue,
Volume 6, No. 6, pp. 431-435
42. Staff, Canadian Mining Journal, Sudbury Operations of Inco.
May 1977, pp. 63-65
43. Edlund, V.E., and S.J. Hussey, "Recovery of Copper from Converter
Slags by Flotation," Report of Investigations 7562 (Revised);
U.S. Department of the Interior
44. Visit to Mitsubishi Metals Co., January 17 and 28, 1977
45. Visit to Mitsubishi Metals Co., January 17 and 285 1977
46. Itakura, K., Ikuda, H., Goto, M., "Double Exoansion of Onahama
Smelter and Refinery," Paper Mo. A74-11, The Metallurgical
Society of AIME, 1974
47. Niimura, M., Konada, T., Kojima, R., "Control of Emissions at
Onahama Cooper Smelter," Joint meeting MMIJ - AIME 1972, Tokyo,
Japan
48. Visit to Naoshima Copper Smelter, January 25, 1977
49. "Naoshima Copper Smelter and Refinery Complex," Guidebook, 1972
514
-------
REFERENCES (Continued)
50. Ichida, Norimitsu, "Design and Construction of New Naoshima
Refinery Using High Efficiency Reverberatory Furnace," Journal
of the Mining and Metallurgical Institute of Japan, Vol. 87,
No. 1001, 1971, pp. 509-14
51. Visit to Naoshima Copper Smelter, January 25, 1977
52. Carpenter, B.H., "Nonferrous Smelter Studies: Investigation of
the Role of Multihearth Roaster Operations in Coooer Smelter Gas
Blending Schemes for Control of SO?: Part III," RTI, EPA
Contract No. 68-02-1325
53. Schuhmann, R., "A Survey of the Thermodynamics of Copper
Smelting," Journal of Metals. 188; pp. 873-884, 1950
54. Ruddle, R.W., The Physical Chemistry of Copper Smelting,
Institute of Mining Metallurgy, London (1953)
55. Mynnykyj, J.R., "Thermodynamic Constraints on the Cerbothermic
and Matte Smelting Process," Canadian Mining Metallurgy Bulletin
56. Johansen, E.B., et al, "On the Thermodynamics of Continuous
Copper Smelting," Journal of Metals, pp. 39-47, September 1970
57. Toguri, J.M., et al., "A Review of Recent Studies on Copper
Smelting," Canadian Metallurgy Ouarterly, 3 (3); pp. 197-221,
July-September, 1964
58. Jeffes, J.H.E. and C. Diaz, "Physical Chemistry of One-Step
Copper Production From a Chalcopyrite Concentrate," Trans.
Inst. Win. Met., pp. C1-C6, 1971
59. Karakas, N., "Magnetite Formation During Copper Matte
Converting," Trans. Inst. Min. and Met., 1972, pp. 35-53, 1962
60. Visit to Mitsubishi Heavy Industries, Ltd., January 19, 1977
61. Visit to Hiroshima Technical Institute, January 24, 1977
62. Korosy, L., et al., "Chemistry of S02 Absorption and Conversion
to Sulfur by the Citrate Process," Paper presented at the
Symposium on Sulfur Removal and Recovery From Industrial Sources,
16th American Chemical Society National Meeting, Los Angeles,
California, April 5, 1974
63. "S02 Removal by a Sodium Citrate Solution Scrubbing," Sweden
Pacer prepared by Olar Erqa, Associated Professor in Chemical
Engineering Norwegian Institute of Technology, Norway
515
-------
REFERENCES (Continued)
64. Pedroso, R.I., "An Update of the Wellman-Lord Flue Gas Desulfuri-
zation Process," Pacer oresented at Symposium on Flue Gas Desul-
furization, New Orleans, March 1976
65. Potts, H.R., E.G. Lawford, "Recovery of Sulfur Gases by the Orkla
Process at Rio Tinto."
66. Bischoff, M.F., Foster Mheeler Eneray Corporation, Livingston,
New Jersey, Private Communications.
67. Bischoff, W.F., et al., "BF Dry Adsorotion System," Paper
presented at the Symposium on Flue Gas Desulfurization, New
Orleans, March 1976
68. Steiner, P., and Juntgen, H., "Prpcess for Removal and Production
of Sulfur Dioxides From Polluted Gas Streams," American Chemical
Society
69. Reference 67
70. Reference 68
71. Reference 66
72. Reference 67
73. Reference 33
74. Rosenkranz, R.D., "Energy Consumption in Domestic Primary Copper
Production," Information Circular 8698, U.S. Department of the
Interior, 1976
75. Westinform Shipping Report No. 310, Sulfur '85 May 1976,
Published by the Mestinform Service, London
76. Reference 75
77. Sulfur in 1975, Mineral Industry Surveys, U.S. Department of the
Interior, Burea of Mines
78. Reference 75
79. Reference 75
80. Reference 75
81. Personal communication with Mr. A. J. Kroha, of ASARCO
516
-------
REFERENCES (Continued)
82. Mineral Facts and Problems, 1975 Edition, U.S. Department of the
Interior, Bureau of Mines
83. Reference 77
84. Reference 77
85. Reference 34
86. Reference 33
87. Reference 77
88. Reference 77
89. Reference 77
90. Merkle & Associates, Inc. Engineers, Suspended Refractory Designs
91. Reference 90
92. McGlamery, G.6., et al., Detailed Cost Estimates for Advanced
Effluent Desulfurization Processes, EPA publication 600/2-75-006
(TVA Bulletin Y-90), January 1975
93. Bucy, J.I., et al., "Potential Utilization of Controlled SOX
Emissions From Power Plants in Eastern United States," Proceed-
ings: Symposium on Flue Gas Desulfurization, New Orleans, March
1976, Vol. II.
94. Shreve, R.N., Chemical Process Industries, McGraw Hill, 1967
95. Reference 77
96. Reference 77
97. Reference 90
98. Reference 90
99. Reference 90
100. Reference 90
101. Reference 36
102. Reference 64
103. Reference 66
517
-------
REFERENCES (Continued)
104. Reference 75
105. Reference 75
106. Savage, P.R., Sulfur: 1980s Shortage or Glut, Chemical Engineer-
ing. September 27, 1976
107. Riegel, Emil Raymond, Riegel's Handbook of Industrial Chemistry
(Seventh edition) (ed. James A. Kent). Van Nostrand Reinhold Co.,
New York, New York 1974
108. Reference 107
109. Reference 106
110. Reference 106
111. Reference 106
112. Reference 106
113. Reference 106
114. Reference 106
115. Reference 106
116. Reference 106
117. Reference 106
118. Reference 106
119. Reference 106
120. Reference 106
121. Reference 75
122. Reference 82
123. Reference 82
124. Reference 82
125. Reference 36
126. Reference 82
518
-------
REFERENCES (Continued)
127. Reference 82
128. Reference 82
129. Reference 75
130. Reference 35
131. Reference 77
132. Tim Browder
519
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-80-152
2.
4. TITLE AND SUBTITLE
Feasibility of Primary Copper Smelter Weak
Dioxide Stream Control
7. AUTHOR(S)
I.J. Weisenberg, T. Archer
A. Prem
Sulfur
, P.M. Winkler, T.J. Browder
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Pacific Environmental Services, Inc.
1930 14th St.
Santa Monica, California 90404
12. SPONSORING AGENCY NAME AND ADDRESS
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
15. SUPPLEMENTARY NOTES
Project Officer: John 0.
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
July 1980 Issuing Date
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
IAB604
11. CONTRACT/GRANT NO.
EPA Contract No. 68-03-2398
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA/ 600/12
Burckle, Nonferrous Metals ft Minerals Branch
16. ABSTRACT
The major source of uncontrolled emissions of S02 from primary copper smelters in
the U.S. is the reverberatory furnace because gas strength is too low for direct pro-
cessing in a sulfuric acid plant, the accepted control approach in this industry. Sys-
tems and techniques that experience indicates, either singly or in combination, can be
used to control weak S02 emissions from copper smelters are identified, analyzed and
discussed.
Two overall approaches to weak S02 stream control are (1) increasing the concen-
tration of S02 to a range where it is feasible to produce sulfuric acid or other use-
able byproducts or (2) neutralizing the effluent as a waste product. Process modifi-
cations to minimize the use of air such as in-leakage control and oxygen enrichment can
significantly increase S02 concentration. The use of add on systems to concentrate the
weak S02 such as the magnesium oxide, ammonia and citrate systems have demonstrated
applicability. The lime or limestone neutralization process where forced oxidation
is used to produce gypsum has been demonstrated as an approach to producing a "throw
away" product. Coal reduction to sulfur also shows sufficient promise for serious
consideration. Product markets are discussed.
17.
KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Exhaust Emissions
Smelting
Pollution
18. DISTRIBUTION STATEMENT
Release to Public
b. IDENTIFIERS/OPEN ENDED TERMS C. COSATI I leld/Group
Sulfur Dioxide Control 13B
Primary Copper Smelting
19. SECURITY CLASS (This Report) 21. NO. OF PAGES
Unclassified 540
20. SECURITY CLASS (This page) 22. PRICE
Unclassified
EPA Form 2220-1 (Rev. 4-77) PREVIOUS EDITION is OBSOLETE
U.3. GOVERliMKNT PRINTING OFFICE: 1980—6^7-1 ft/0090
520
-------