EPA
TECHNOLOGY
TRANSFER
                         FOURTH PROGRESS REPORT:
FORCED-
OXIDATION
TEST RESULTS
ATTHE
EPA ALKALI
SCRUBBING
TEST FACILITY
 US EPA
 OFFICE OF
 RESEARCH AND
 DEVELOPMENT
 PROTOTYPE
 DEMONSTRATION
'FACILITY

-------

-------
        EPA
        TECHNOLOGY
        TRANSFER
                                   FOURTH PROGRESS REPORT:
EPA-625/2-78-018
FORCED-
OXIDATION
TEST RESULTS
AT THE
EPA ALKALI
SCRUBBING
TEST FACILITY
U.S. EPA
OFFICE OF
RESEARCH AND
DEVELOPMENT
PROTOTYPE
DEMONSTRATION
FACILITY

-------
Wet-Scrubbing Test Structure with Truck in Foreground for Waste Sludge Hauling

-------
INTRODUCTION

   This is the fourth of a series of capsule reports
describing the results of the Shawnee Lime and
Limestone Wet Scrubbing Test Program conducted
By EPA's Industrial Environmental Research Lab-
oratory, Research Triangle Park, North Carolina
(IERL-RTP). In this program, flue gas desulfuriza-
tion tests are being conducted at the Shawnee Lime
and Limestone Wet-Scrubbing Test Facility  located
at the Tennessee Valley Authority (TVA) coal-fired
Shawnee Power Station near Paducah, Kentucky.
Bechtel National, Inc. of San Francisco is the major
contractor and test director, and TVA is the con-
structor and facility operator.
   This report describes the results of forced-oxida-
tion testing at the Shawnee Test Facility from
January 1977 through June 1978 and summarizes
earlier forced-oxidation results, at the IERL-RTP
pilot plant, which led to the testing at Shawnee.
BACKGROUND

  Of the approximately 55,000 MW of coal-fired
electrical generating capacity in the United States
presently committed to the use of flue gas desul-
furization (FGD) systems, about 90 percent is com-
mitted to lime or limestone wet-scrubbing. A major
drawback of lime and limestone systems is the na-
ture of the waste material produced. In these sys-
tems, alkali (lime or limestone) reacts with the SO2
in the flue gas, producing a waste sludge contain-
ing calcium sulfite and calcium sulfate (gypsum)
plus collected fly ash. In most medium- to high-
sulfur applications, natural oxidation of sulfite to
sulfate in the scrubber system amounts to only 10
to 30 percent and calcium sulfite is the predomi-
nant material in the waste sludge.
  Calcium sulfite wastes present a serious disposal
problem because of the difficulty of dewatering.
The slurry can be dewatered only to about 50 to
60 percent solids, producing an unstable, thixo-
tropic material unsuitable for landfill. Where space
is available, ponding of the untreated sulfite sludge
has been practiced. But the pond area may be
impossible to reclaim, and in many locations suffi-
cient space is not available.
   Three procedures have been considered for con-
verting the sulfite wastes to material suitable for
landfill:
    Commercial fixation with additives

    Blending sulfite sludge with fly ash

    Forced oxidation of the calcium sulfite to a
     more tractable calcium sulfate (gypsum),
     which is easily dewatered to greater than
     80 percent solids

   In a recent TVA study of sludge disposal eco-
nomics, the incremental  increase in plant revenue
requirements with the above landfill disposal pro-
cedures over a base  case with pond disposal was
0.6 to 1.1 mill/kWh for commercial fixation, 0.7
mili/kWh for blending sludge with fly ash, and 0.2
mill/kWh for forced oxidation. Clearly, the eco-
nomics favor forced oxidation for upgrading pond
disposal to landfill.
   In Japan, where natural gypsum is not available,
forced oxidation in  scrubber systems has been
employed extensively to produce a high-quality
gypsum raw material for the cement and wallboard
industries. In the United States, scrubber gypsum
may be unable to compete extensively with the
widely available natural gypsum. Thus, the incen-
tive in the United States has been to develop sim-
plified forced-oxidation procedures, directed only
toward improving waste solids handling and dis-
posal properties. As a disposal material, the gypsum
sludge can have high fly ash content; moreover, the
oxidation reaction need be carried only to about
95 percent completion.

-------
  On the basis of the established need for better
waste material properties and reduced disposal costs
in lime and limestone wet-scrubbing, a program to
develop a simplified forced-oxidation technology
was initiated by the EPA in 1976.
  During 1976, initial studies conducted at the 0.1
MW IERL-RTP pilot plant showed that calcium sul-
fite can be readily oxidized to gypsum by simple
air/slurry contact in the hold tank of the scrubber
recirculation loop. Although the rate of oxidation
is maximum at a pH of 4.5, it was found that the
oxidation rate was adequate up to a hold tank pH
of about 6.0.
  Following tests at the IERL-RTP pilot plant, a
testing program was set upfortheShawneeTest Fa-
cility to demonstrate  promising scrubbing configu-
rations with forced oxidation and to generate de-
sign data. The Shawnee forced-oxidation test sched-
ule is shown in Figure 1. Tests were conducted over
a period of 18 months, from January 1977 through
June 1978. Additional testing to optimize these
processes is planned.

SHAWNEE TEST FACILITY

  At the Shawnee Test Facility, two parallel wet-
scrubber systems are in operation: a venturi/spray
tower system and a Turbulent Contact Absorber
(TCA) system. Each has its own slurry-handling fa-
cilities; each  is designed to remove both SO2 and
particulate from approximately 10 MW equivalent
of flue gas (up to 35,000 acfm at 300F)  contain-
ing 1,500 to  4,500 ppm SO2 and 3 to 6 grains/dry
scf of fly ash. The systems can also operate with
flue gas  having low fly ash loading by withdrawing
flue gas  downstream from the existing plant partic-
ulate removal equipment. Figure 2 shows  the test
facility control room.
FORCED-OXIDATION TEST BLOCK
WITH TWO SCRUBBER LOOPS
LIMESTONE
LIME
LIMESTONE/MAGNESIUM OXIDE
WITH ONE SCRUBBER LOOP
LIMESTONE (WITH EDUCTOR)
LIMESTONE (WITH SPARGER)
LIMESTONE/MAGNESIUM OXIDE
BLEED STREAM OXIDATION
LIMESTONE/MAGNESIUM OXIDE

J

hr-



F

- 

1977
M
A M
J
i
T

 -ma

,




SCK
MAIN


i
~~
T
*
ii
1EDULED
TENANC

_
BOI
EOU
J

Sgggg
KB

LER
TAG

A

.. gg
a-

E
s|o
|
j
ijgi 
 	 \+,
. .!_
!
f'''
H"*t
!
I
_4~
t
i
i
ij
N




D
la-

m


J
'-  
1978
F
M

A

! '

ESSSM,

i
I
1
1
- 4 	
i
i





M



^-j....

~~~~
'
m
J
ii
i
:
">
d
 	 7

 Figure 7.  Shawnee Forced-Oxidation Test Schedule

-------
TESTS WITH TWO SCRUBBER LOOPS

   The IERL-RTP pilot plant results showed that
forced oxidation in the hold tank of the first of
two scrubber loops in series was a promising con-
figuration for both lime and limestone scrubbing.
To reproduce this configuration at the Shawnee
Test Facility, the venturi/spray tower system was
repiped and a new  hold tank with an air sparger
was added to provide independent scrubber loops
for the venturi and the spray tower. The results of
forced oxidation testing with the two-loop venturi/
spray tower system are  discussed in Section 3.
TESTS WITH ONE SCRUBBER LOOP

  At the IERL-RTP pilot plant, successful opera-
tion of forced oxidation in the hold tank of a single
scrubber loop was restricted to limestone slurry.
Limestone systems depend on dissolving calcium
carbonate for reactive alkali. In lime systems, the
primary reacting component is calcium sulfite;
forced  oxidation eliminates most of this calcium
sulfite  from the slurry liquor and substantially re-
duces SO2 removal efficiency.
  Two methods were employed for modifying the
Shawnee TCA scrubber system for air/slurry contact
in the hold tank: one involved an eductor, the other
an air sparger. The results of forced-oxidation test-
ing with the one-loop TCA system are discussed in
Section 4.
BLEED STREAM OXIDATION TESTS

   Oxidation of the slurry bleed stream from a
scrubber system would have the least effect on the
scrubber operation, and facilities to implement such
oxidation would be the most practical to install
commercially. However, tests at the IERL-RTP pi-
lot plant indicated that such bleed stream oxida-
tion is made difficult by the dissolving residual al-
kali, which causes the pH to rise and slow down
the oxidation rate. Furthermore, bleed stream oxi-
dation did not produce as great an improvement in
solids dewatering properties as when the oxidation
was conducted within the scrubber loop.
  On the basis of promising laboratory results, lim-
ited bleed stream oxidation tests were conducted
at the Shawnee Test Facility during May and June
1978  in which limestone with added magnesium
oxide (MgO) was used as the alkali.  It was estab-
lished in these tests that magnesium ion increases
liquor sulfite concentration and buffers the pH, al-
lowing oxidation to take place. Waste solids pro-
duced in these tests had good dewatering proper-
ties. The results of these tests are described in
Section 5.
DEWATERING AND HANDLING PROPERTIES

   Regular dewatering and handling tests were con-
ducted on the waste sludge produced during the
forced-oxidation testing. In disposal tests, agypsum
sludge pond was created and a 10-foot gypsum pile
was made for the study of weathering properties.
The results of these tests and the morphology of
gypsum sludge are discussed in Section  6.
 Figure 2.  The Control Room

-------
                                           FLUE GAS OUT
                                                                  MAKEUP WATER
                                                                            CLARIFIED LIQUOR FROM
                                                                          SOLIDS DEW A TERING SYSTEM
                                                                        SYSTEM BLEED TO SOLIDS
                                                                         DEW A TERING SYSTEM
 Figure 3.   Typical Flow Diagram for Venturi'/Spray Tower System in Two-Scubber Loop
            Forced-Oxidation Service
   From early January 1977 through mid-May
1978, the venturi/spray tower system was oper-
ated in a two-scrubber-loop configuration with
forced oxidation of calcium sulfite to calcium
sulfate (gypsum) in the first scrubber slurry loop.
As the following data indicate, this operating mode
proved to be quite successful. Commercial applica-
tion of this type of operation should be feasible
since there are many FGD plants using two gas-
liquid contacting devices in series (e.g., venturi for
paniculate removal followed by an after scrubber
for SO2 removal), which, in some cases, can be
modified to a two-scrubber-loop operation.
   A typical  flow diagram is shown in Figure 3. The
spray tower slurry loop consists of the spray
tower; the effluent hold tank, to which the alkali
feed, additive (MgO), and makeup water (mist
eliminator wash water) are added; and the spray
tower feedlines. The venturi slurry loop comprises
the venturi scrubber, the oxidation tank, the de-
supersaturation tank, and the venturi feedline. The
low-pH spray tower effluent slurry is bled from the
downcomer to the oxidation tank. SO2 absorption
in the venturi scrubber acidifies this slurry, thereby
allowing the oxidation tank pH to be maintained
at a low value for oxidation by air sparging. The
desupersaturation tank provides extra time for the
growth of gypsum crystals. The system bleed
stream (oxidized slurry) is taken from the desuper-
saturation tank and sent to the solids dewatering
system, which consists of a clarifier and a filter in
series. Ail  liquor from  the solids dewatering system
is returned to the scrubbing system, resulting in a
very tight  water loop.
  The main advantage of this system, as far as
forced oxidation is concerned,  is that it permits

-------
the two scrubber loops to operate at different pH
values. Thus, the first loop pH can be controlled at
low values (4.5 optimum) for maximum oxidation
efficiency, while the second loop operates at the
higher pH needed for efficient SO2 absorption.
This configuration also maximizes the utilization
of limestone.
   A total of 29 limestone runs and 20 lime runs
were made. The average length  of test was about 6
days, except for one limestone  and  one lime long-
term (over 1 month) reliability demonstration run.
Tests were conducted with both high and low fly
ash loading in the flue gas. In addition, limestone
tests with high fly ash  loading were conducted
with MgO addition to enhance SO2 removal
efficiency.

LIMESTONE TESTS

   A typical  limestone run using flue gas with high
fly ash loading is presented in Table 1. In this run,
the venturi  inlet pH (oxidation tank pH) was con-
                                            trolled at 5.5 by adjusting the limestone slurry feed
                                            rate to the spray tower effluent hold tank. The
                                            slurry solids concentration in the venturi loop was
                                            controlled at 15 weight percent (including fly ash).
                                            The clarified liquor in excess of the mist eliminator
                                            bottom wash was returned to the oxidation tank.
                                            This resulted in a slurry solids concentration of 9.6
                                            weight percent (essentially no fly ash) in the spray
                                            tower loop.
                                              Near-complete sulfite oxidation  of 98 percent
                                            was achieved at an air stoichiometry of 1.7 atoms
                                            oxygen/mole SO2 absorbed. The filter cake con-
                                            tained 86 percent by weight of solids, and lime-
                                            stone utilization averaged 81  percent.
                                              During all the limestone and lime two-scrubber-
                                            loop forced-oxidation runs, it was observed that
                                            the venturi inlet slurry constantly exhibited a gyp-
                                            sum saturation of about 100  percent, which was
                                            below the incipient scaling level of 135 percent.
                                            This was undoubtedly caused by the abundance
                                            of gypsum crystal seeds produced by forced
                                            oxidation.
                                       Table!
       TYPICAL TWO-SCRUBBER-LOOP FORCED-OXIDATION LIMESTONE
                  USING FLUE GAS WITH HIGH FLY ASH LOADING
                                                                                  RUN
      It
     Lay Tower GquTcPto-GlasKatio. gal/Mcf
   OTWT 1 i- 6n Vs- iKSBt-^SuiaS;  la^^-ut^ta^ a^_mft3i^i^^
   /enturi Slurry Solids Concentration, wt %  (controlled)
55  -SK-i^** j-^ |Hmiiuj4L ~3>mSfiB-i*iUitr,tti
-------
   Tests were also conducted using flue gas with low
 fly ash loading. A comparison of the test results
 with high and low fly ash loading showed that the
 fly ash did not appear to affect the sulfite oxida-
 tion efficiency. Most fly ash is removed  by the ven-
 turi and ends up  in the first (oxidation)  slurry loop.
   The effects of pH and air stoichiometry on the
 oxidation efficiency were investigated. At an air
 stoichiometry of 1.7, the venturi inlet pH  (oxida-
 tion tank pH) was varied from 4.5 to 5.5 and the
 sulfite oxidation  remained greater than 96 percent.
 At a venturi  inlet pH of 4.5, 97 percent  sulfite oxi-
 dation was achieved at an air stoichiometry as low
 as 1.0. Further reduction of air stoichiometry to
 0.5 resulted in only 67 percent oxidation.
   A limestone run with low fly ash loading was
 made with thedesupersaturation tank excluded
 from the venturi  slurry loop. Figure 4 shows the
 oxidation tank and desupersaturation tank. No ad-
 verse effect was observed on either the filter cake
 solids content (86 weight percent) or venturi inlet
 liquor gypsum saturation (105 percent). Two runs
 were also made with the oxidation tank  level low-
 ered from 18 to 14 feet. Sulfite oxidation remained
 96 percent or higher at 5.5 venturi inlet  pH and 1.7
 air stoichiometry.
   Initially, an air sparger with 130 1 /8-inch holes
 was used in the oxidation tank. These small holes
 tended to plug easily because of slurry solids build-
 up on the wet/dry interface, even though the air
 had been prehumidified. When the sparger was later
 replaced by one with 40 1 /4-inch holes, the plug-
 ging tendency was reduced considerably. Finally, a
 simple 3-inch pipe was employed, which discharged
 air downward through an open elbow at the center
 of the oxidation tank 3 inches from the  bottom of
 the tank and  completely eliminated the plugging
 problem, even when the air was not humidified.
 Significantly, these changes did not affect the air/
 slurry contacting  efficiency.  It was theorized that
 the air/slurry contact was predominantly affected
 by the agitator in  the oxidation tank. The agitator
 used has two  turbines and operated at 56 rpm
 (fixed speed)  and  17 brake horsepower.
  A test was  conducted in an attempt to discover
the effect of agitator speed on oxidation efficiency
by turning off the agitator. Unfortunately, the test
failed because air  sparging alone could not prevent
the oxidized slurry from settling out in the oxida-
tion tank. Further testing will be conducted when
a variable-speed agitator is acquired for the oxida-
tion tank.
Figure 4.  Oxidation Tank and Desupersaturation
          Tank in the Two-Scrubber-Loop
          System
 LIME TESTS

   A typical lime run using flue gas with high fly
 ash loading is given in Table 2. The operating char-
 acteristics for a lime two-scrubber-loop oxidation
 run are quite similar to those of a limestone run.
 The major difference is that the lime system re-
 quires the control of both venturi inlet pH and
 spray tower inlet pH  by separate lime addition to
 avoid wide fluctuations in pH, whereas the lime-
 stone system requires the control of only the ven-
 turi inlet pH by limestone addition to the spray
 tower hold tank. This is because of the more buf-
 fered nature of the limestone slurry.
   For the run shown in Table 2, the venturi inlet
 pH and spray tower inlet  pH were separately con-
 trolled  at 5.5 and 7.8, respectively. The slurry
solids concentration in the venturi loop was con-
trolled  at 15 weight percent (including fly ash).
The slurry solids concentration in the spray tower
was maintained at approximately 10 weight per-
cent (no fly ash) by proper distribution of the
clarified liquor returned to the oxidation tank and

-------
spray tower hold tank. A high concentration of
solids in the spray tower makes it more difficult to
keep the mist eliminator clean, and a low solids
concentration  (below about 7 percent)  may cause
calcium sulfite and sulfate scaling.
  Sulfite oxidation was 98 percent at 1.7 air stoi-
chiometry, and the filter cake solids content was
85 percent. Near-complete lime utilization was
achieved.
  As in the limestone tests, no apparent effect on
oxidation efficiency between high and low fly ash
loading was detected.
  The effects  of pH and air stoichiometry on the
oxidation efficiency were  also studied.  At an air
stoichiometry  of about 1.7, sulfite oxidation re-
mained 97 percent or higher when the venturi inlet
pH  was varied  from 4.5 to 5.5. The oxidation effi-
ciency began to drop at this air stoichiometry
when the pH was increased above 6. At a venturi
inlet pH of 5.5, sulfite oxidation decreased from
99 percent to  76 percent when the air stoichio-
metry  was reduced from 1.5 to 1.2.
                                                 As in the case with the limestone tests, no dif-
                                              ference in the oxidation efficiency was observed
                                              for lime tests when the 130-hole (1/8-inch) air
                                              sparger, the 40-hole (1/4-inch) air sparger, or the
                                              3-inch air pipe was used in the oxidation tank.
                                                 Decreasing the oxidation tank level from 18 to
                                              14 feet did not change the degree of sulfite oxida-
                                              tion. Further decrease of the tank level to 10 feet
                                              resulted in about 85  percent oxidation. The oxida-
                                              tion at a 10-foot tank level, however, is not com-
                                              parable with those at 14 and 18 feet because the
                                              top turbine of the agitator was located above the
                                              slurry level (at 11 feet), thus giving a different
                                              agitation pattern.

                                              LIMESTONE TESTS WITH MgO ADDITION

                                                 Six two-scrubber-loop limestone runs were made
                                              in which MgO was added to the spray tower hold
                                              tank while oxidation was forced in the venturi slur-
                                              ry loop. The  primary purpose of the MgO addition
                                              was to enhance SO2 removal  efficiency in the spray
                                              Table 2
                TYPICAL TWO-SCRUBBER-LOOJP FORCED-OXIDATION LIME RUN
                         USING FLUE GAS WITH HIGH FLY ASH LOADING
       r~ *> ^+tr i73e-*T- - *T* /*~"
       pray Tower Gas velocity, ft/sec
       m >y"* r &"?%  j- f*f , % # * ** ^**  _sw *, .v:*^ *,.
                                                                                  (.>-* .
^sprayjlowe^uas^ve^ocay^j^/sec      __ _^_
IVentun Liquid-to-OTas SatioTgal/Mcf  "" "~
jji, ffiOT^K.^^'^t^^f'V^r^8d^sL*'t^?jin-i?" iw^iMSa*^sti(>*si
-------
tower by increasing the sulfite ion (SO3~) concen-
tration in the liquor for SO2 scrubbing. In the stan-
dard two-scrubber-loop configuration shown in Fig-
ure 3, the magnesium ion (Mg++) concentration in
the venturi loop is about twice that in the spray
tower loop because of the water loss to flue gas hu-
midification in the venturi loop. But because the
sulfite ion is converted into nonscrubbing sulfate
ion by forced oxidation, the higher Mg++ concen-
tration in the venturi loop does not improve the
SO2 removal in the venturi loop. Another purpose
of the MgO addition was to observe whether the
oxidation efficiency is adversely affected by the
MgO addition.
   A typical limestone run with MgO addition using
flue gas with high fly ash loading is shown in Table
3. The operating characteristics of this run were sim-
ilar to the one shown in Table 1, with the exception
of the effect of the 5,150 ppm effective Mg++ con-
                                                 centration in the spray tower. SO2 removal im-
                                                 proved from 86 percental 2,550 ppm average inlet
                                                 SO2 concentration to 96 percent at 2,250 ppm. Sul-
                                                 fite oxidation remained high at 98 percent with the
                                                 same 1.7 air stoichiometry. As expected, SO2 re-
                                                 moval by the venturi scrubber alone was 30 percent,
                                                 about the same as was the case without MgO addi-
                                                 tion. The filter cake solids content did not appear
                                                 to be affected by the MgO addition and  remained
                                                 high at 85 percent.
                                                   When the air stoichiometry  was reduced from 1.7
                                                 to 1.1, sulfite oxidation decreased from  98 to 92
                                                 percent and filter cake solids content dropped
                                                 slightly from 85 to 82 percent. When air to the oxi-
                                                 dizer was completely turned off, sulfite oxidation
                                                 dropped to 36  percent (natural system oxidation
                                                 level) and the filter cake solids content dropped to
                                                 only 63 percent.
                                              Table 3
             TYPICAL TWO-SCRUBBER-LOOP FORCED-OXIDATION LIMESTONE RUN
             WITH MgO ADDITION USING FLUE GAS WITH HIGH FLY ASH LOADING


,,   lilii'i" , ,  ,i',, ! 7  i,  " il siuf  I ,	   .  ,	  ,	 J[ ,,*J  ~ I ,
"""	:|'";~""::Spray  Tower Liquid-to-Gas Ratio, gal/Mcf
'	Ifc^ Venturi Slurry Solids Concentration, wt % (controlled)
	!!i Spray  Tower Slurry Solids Concentration, wt %
::::|f!:":;:yenturijiiplet pH (controlled)	
       ~oray  Tower Inlet pH
        '	>	i-s*wii	-:	IPT	J-.
   : Spray Tower Effective Mg++ Concentration, ppm*
   Ki~ "W^SS^SS,	P.ffiP).'1]- H2O _       ^" w _   _^
     xidation Tank Level, ft
     in nil iini)iiiiiii|iiiiiniiii  inn iiiiiij iiiiiinnw 111911141 briiiiiiinihii iniiiw INIJIIH iw t t  n  n liMHHti i ^   iet t/
     xidation Tank Residence Time, mm
                         j                   *    5  *"? i^   y fc>  SM1
     esupersaturation Tank Residence Time, min
     llllllV I 111 III *                                mi    i TI    "Mm   inww u "w*1-
 ;|;pray Tower Effluent Tank  Residence Time, min
   Percent S,P2 Removal
 ~; Average Inlet SO2	Co^e'iTtration,''p'jam
   Ptercent (Dxidatipn of Sulfite^to Sujfate ^         _^
   ^ir^Stoichiiometry,, atoms oxygen/mole SO2 absorbed
   Percent jLimestone" Uti'lfzation"" "      ""* ~~~_J  t "  ~^ ^  "_"",,
   Venturi l"nlet'Liquor Gypsum Saturation at SOC,"^"* ^  ***  **'
   Filter Cake Solids Content, wt %
      i  i  iiiii      iiini"! i	i" 'i" i         i  i-i       '  * * ^-t  rf
   _, ,'gctjye Mg++ concentration is defined as the total Mg++ minus
  [|n5f Mg^"*" concentration equivalent to total chlorides.
                                                                                   J3A
                                                                                    21
                                                                                    ^57
                                                                                  ,J5
                                                                                    6.0
                                                                                    5.5
                                                                                    6.0
                                                                                 8,410
                                                                                 5,150
                                                                                 , .. ..
                                                                                   11.3
                                                                                   4.7
                                                                                vi  tfrl -v f
                                                                                   14.7
                                                                                    96
                                                                                 2;250
                                                                                  .  98
                                                                                     -7
                                                                                   130
                                                                                   .85
                                                                                   I
                                                                                            *

                                                                                              1
                                                                                           fv^tsamem

                                                                                              t
                                                                                             i
                                                                                             i
                                                                                            
-------
                     Chevron Mist Eliminator'
                        Retaining Bar-Grids
                          FLUEGASIN
r\ /\ r\;
o cj
Q^Q-d
0o0i
sae&i
01
oo4
Q_Q. !
1
i
i
^^ 
1 [




^//WoW/e
Packing
Spheres















                                                                         CLARIFIED LIQUOR
                                  u
       COMPRESSED AIR
    HUMIDIFYING WA TER
                                                                           SYSTEM BLEED TO SOLIDS
                                                                           DEW A TERING SYSTEM
                            OX/D/l TION TANK    EFFLUENT HOLD TANK
 Figure 5.  Typical Diagram for TCA System in On'e-Scrubber-Loop Forced Oxidation Service
  The TCA system was operated in a one-scrubber-
loop forced-oxidation configuration from late June
through early October 1977 using an air eductor as
the oxidizer, and again from early December 1977
through late January 1978 usinga40-hole (1/4-inch)
air sparger. A typical flow diagram for this config-
uration is shown jn Figure 5. A total of 22 runs wer
made, including 15 runs with the air eductor and 7
runs with the air sparger. The average test length
was about 6 days. All runs were conducted with
limestone slurry with high fly ash loading.
   Both types of oxidation devices used at Shawnee
were set up with a minimum of system modifica-
tion. These devices should be readily applicable to
commercial-scale FGD plants, either for new or for
retrofit installation. Since an air eductor is more
prone to severe erosion in high-velocity slurry
service and consumes more energy than an air
sparger, it is less attractive as an oxidizer from  the
standpoint of both reliability and economics.

-------
OXIDATION WITH AN AIR SPARGER

   The flow diagram in Figure 5 shows the TCA
one-scrubber-loop forced-oxidation configuration
with two tanks in series and an air sparger. The
scrubber effluent slurry first enters the oxidation
tank into which air is introduced through a sparger
with 40 1/4-inch holes. Overflow from the oxida-
tion tank enters a regular effluent hold tank into
which the limestone slurry feed and clarified
liquor from the solids dewatering system are added.
The slurry  in the effluent hold  tank is then re-
cycled back to the scrubber. System bleed slurry is
withdrawn from the effluent hold tank and sent to
the solids dewatering system, which in these runs
usually consisted of a clarifier only. All liquor from
the dewatering system is returned to the scrubber.
   Operation with two tanks has the advantages of
providing oxidation of the scrubber effluent slurry
at low  pH and of increasing the utilization of
limestone. Typical run results using such a scheme
are shown in Table 4. In  this run, the  limestone
stoichiometry was controlled at 1.3, resulting in an
oxidation tank pH of 5.4; slurry solids concentra-
tion was controlled at 15 percent by weight (in-
cluding fly ash); sulfite oxidation was 94 percent
with 1.7 air stoichiometry; and actual limestone
utilization  was 81 percent.
                                                      A comparison of the results of this run and
                                                   those shown in Table 1  for a two-scrubber-loop
                                                   forced-oxidation limestone run shows that better
                                                   oxidation (98 percent) was achieved for the two-
                                                   loop run in a similar oxidation tank environment
                                                   (pH, air stoichiometry, tank level, percent slurry
                                                   solids, and limestone utilization). The major differ-
                                                   ence was the poorer agitation in the oxidation tank
                                                   for the run shown  in Table 4 (37 rpm, 3 hp agitator
                                                   motor), compared with the run described in Table
                                                   1 (56 rpm, 20 hp agitator motor). This may explain
                                                   the lower oxidation for the run described in Table
                                                   4. Figure 6 shows the air sparger and its position
                                                   relative to the agitator.
                                                      Since the existing Shawnee air compressor did
                                                   not have sufficient capacity to serve both the ven-
                                                   turi/spray tower and TCA simultaneously at full
                                                   gas load, no attempt was made to improve the de-
                                                   gree of sulfite oxidation in the TCA by  raising the
                                                   air stoichiometry. This limitation will be corrected
                                                   when a new air compressor is installed.
                                                      Tests were also conducted for the one-scrubber-
                                                   loop oxidation configuration using a single hold
                                                   tank. In this scheme, the effluent hold  tank in
                                                   FigureS was excluded from the scrubber slurry
                                                   loop. Thus, the system became a one-scrubber, one-
                                                   hold-tank configuration with air sparging in the
                                                   hold tank. The advantage of this system is, of
                                               Table 4
                      TYPICAL TCA FORCED-OXIDATION LIMESTONE RUN
                            USING TWO TANKS AND AN AIR SPARGER
                 .                  .       ..        .                                WS1**'1-^ ...... ;? ...... '.ITX
       rr .........  ........ ; ...... : ...... m-'"?,"^?? ......
                                                                                       "
                 -,
  : ..... ..... p-'-Scrubber Gas,, Velocity, ft/sec     ,    ^ ...... m ................ ^    ^   iu  ^   '    ;  ' / :^ ;";:_;" 8.4
  ......                ^                ................................. ' ..... ..... ................ '"" ......... '" """*'  "  " "''" ...... ""
      !	|ifl	"Ill	f	N|,j*	.liniiH	M,. if? ,	i,i.,,;;>n,,--^VrVT'h1^ii1iiiiiiiiiii	mm mm, mm 'i %-n9m^m^-^^^i^^>^K^r- a,w^ jafiimm m Ti^n^KmSmmK .-.^Sn^i.-	iilniii MnB..-.....--^.^^^ frm. ,j,,.m^w 	i
      I*Limestone Stoichiometrys moles Ca/mole SO2 absorbed (controlled)
      .:..	sIiJL'B'Uffiu  u :i_lll!>, U: oSi!IBisSS:m;*;tl^; J^BL^i^M^^^
     IL mesione iio cmomeiry, moies ^a/moie suy ao
     jIjJL'LMSU  u :i_lll!>, U: iiSii!BiIBaS:!SSJI!;*'it,%4s;UKj&ul>4i:>f>WHUiawa*iA.-4r.i.i
     IJotal	Static iBed	Fieight^ jncnes of 6.5 jm spheres

  |B^
   ''I?-	;Percerit SQ2 Removal
  h ilt;k-tl4y^!3sll!?!?!: Sp2'Concentration, ppm
  !;	f	TPercent"Oxidafion^of^Sulfite to^Sulfate^.^....^..^^'.., _rT..'. ,,,,.,,,.,. - .,,	,	r,,,Www?^-.. -Sa- ..KWKWH	tHM^uHBHHHHMH!	^^.^^a,,,,^^^^a.^^,
''IS'i'i^ScruSbTri'S                                           	.^...^.,,,,	...^...^	.,	"

^^r.:ziiLzi!^:^g,!1z T;::::g;a^::|||ga^g;ig5ig^
*-___:	.	,	,	,,.;.,_^_ ". 'mi ^ 1P~^""::"" "''"'.'". '-- 'Zw^r.j.-'-'i^v-, ?sUaU-J^2--^_^^^^jiaESajys^^

-------
course, its simplicity. But since the oxidation has
to take place at a higher pH in the hold tank (i.e.,
at scrubber inlet pH), a higher air stoichiometry is
required, as can be seen from  the following data.
With the same flue gas velocity of 8.4 ft/sec, a
liquid-to-gas ratio of 62 gal/Mcf, and  15 percent
slurry  solids  (see Table 4), one-tank  operation
yielded 94 percent oxidation at a somewhat higher
air stoichiometry of 1.9 and an oxidation pH
(scrubber inlet pH) of 5.5. Limestone utilization
dropped to 77 percent and SO2 removal decreased
to 75 percent at  2,800 ppm inlet SO2 concen-
tration. Attempts to increase the scrubber inlet
pH (to increase SQ2 removal), and thus oxidation
pH, yielded only 92 percent oxidation at 5.6 pH
with 2.0 air stoichiometry.
   A single test with one tank was made in June
1978 in which MgO was added to  maintain an
effective liquor Mg++ concentration of 5,000
ppm. Limestone slurry with high fly ash loading
was used in this test. Sulfite oxidation was 95
percent at 5.5 pH and 1.7 air stoichiometry. As
anticipated, these results were similar to those
obtained in tests with one-tank operation without
MgO addition. Because of the oxidation of the
sulfite ion to the sulfate ion, adding MgO  did
not increase the percent SO2 removal.
 Figure 6.  Air Sparger and Agitator on the TCA
           Oxidation Tank
OXIDATION WITH AN AIR EDUCTOR

  Oxidation tests were conducted using an air
eductor. The high velocity of the slurry pumped
through the eductor nozzle aspirates air into the
liquid steam. The high shear developed in the
throat of the eductor breaks air  into minute
bubbles which are ejected into the oxidation tank,
aerating the slurry in the tank. The main feature
of the eductor is its ability to create smaller air
bubbles than could be obtained with an air sparg-
er; the result is a higher mass transfer coefficient.
  The eductor employed was a Penberthy-
Houdaille Model ELL-10 Special, which normally
operated at 1,600 gpm at 40 psig head and  aspi-
rated 535 scfm of air from atmosphere when dis-
charged  against a 1 psig back pressure. This educ-
tor is shown in Figure 7.
  Both one-tank and two-tank schemes were used
in the tests. In the one-tank scheme, the eductor
took slurry feed  from the oxidation tank (effluent
hold tank) and discharged the air/slurry mixture
back to the same tank. In the two-tank scheme, the
eductor took slurry feed  from a separate low-pH
scrubber effluent hold tank and discharged the
air/slurry  mixture into a second tank. Limestone
feed was added to the second tank. The two-tank
configuration had the advantage of lower pH slurry
passing through the eductor and  better oxidation
efficiency.
 Figure 7.  Air Eductor

-------
   Initially, the eductor was mounted on top of the
oxidation tank in an off-center position and dis-
charged slurry vertically downward  2 feet below
the tank level. Tests were later made with the
eductor mounted outside the oxidation tank in a
horizontal, radial position, with the centerline of
the eductor 10 inches above the tank bottom. This
change was made in  an attempt to increase the
plume (air/slurry mixture discharged from the
eductor) residence time  in the oxidation tank,
thereby improving oxidation.
   Under typical operating conditions of 30,000
acfm flue gas flow rate (12.5 ft/sec scrubber gas
velocity), 1,200 gpm TCA slurry flow rate (50
gal/Mcf liquid-to-gas ratio), 1,600  gpm eductor
slurry flow rate, and  15 percent slurry solids con-
centration  (with fly ash), sulfite oxidation of
better than 91  percent was obtained at an eductor
inlet pH of 5 to 5.5, 8- to 12-foot oxidation
tank levels, and airstoichiometry of 2.0 or higher.
   With two-tank operation, a  higher pH to the
TCA was possible, and the result was better SO2
removal, normally over 80 percent. With one-tank
operation, low oxidation tank pH (scrubber inlet
pH) was necessary to achieve good oxidation, and
the result was poor SO2 removal, about 60 percent.
In the latter case, some SO2 removal efficiency was
recovered by adding limestone to the scrubber inlet
slurry stream instead of to the oxidation tank.
   No significant difference in  sulfite  oxidation
efficiency was observed between operations  with
the eductor in  the vertical position or in the hori-
zontal position. Air/slurry contact efficiency was
believed to be hampered  by the configuration of
the oxidation tank used, which was 20 feet in dia-
meter with only 8 to 12 feet of normal liquid
depth. A smaller diameter but deeper tank with
eductor discharge at the bottom would probably
improve the oxidation efficiency.
  Sulfite oxidation was found to be sensitive to
the slurry flow rate to the eductor.  In two runs in
which the eductor slurry flow rate was reduced
from 1,600 to 1,200 gpm, oxidation dropped to
about 60 percent. This was probably the combined
result of reduced air flow and decreased agitation
by eductor plume in the oxidation tank.
  Two runs were made with  the rotary-drum
vacuum filter in series with the clarifier for solids
dewatering. The filter cake solids content was 86
percent, which was typical for all oxidized slurry
at Shawnee.
  As far as the air flow rate was concerned, the
performance of the eductor was as predicted by
the manufacturer. Air flow rate was sensitive to
the depth of eductor submergence in the slurry,
ranging from 600 scfm at zero submergence to
200 scfm at 12 feet submergence, when operated
at 1,600 gpm slurry flow rate. An air stoichiometry
of 2.0, corresponding to about 200 scfm, was suf-
ficient to yield good oxidation at a 1,600 gpm
eductor flow rate.
  Erosion of the eductor in the limestone/fly
ash  slurry service was a major problem. The eductor
body was constructed of neoprene-lined carbon
steel. After 1,500 operating hours, the rubber near
the nozzle had eroded in a circular pattern; after
an additional 550 hours the carbon steel body had
eroded through. The Stellite nozzle  and the sta-
tionary whirler within the nozzle showed only
minor evidence of erosion.

-------
                                                                       CLARIFIED LIQUOR FROM
                                                                      SOLIDS DEWA TER1NG SYSTEM
                                                                    SYSTEM BLEED TO SOLIDS
                                                                      DEWA TERING SYSTEM
  Figure 8.   Typical Flow Diagram for Venturi/Spra_


  Starting in mid-May 1978, forced oxidation of
the scrubber bleed stream was conducted for about
1 month on  the venturi/spray tower system. This
operating mode was successful for limestone tests
with MgO addition. These tests yielded a filter cake
solids content of about 85 weight percent, approxi-
mately the same as the filter cake from two-scrubber-
loop operation, although the slurry solids settling
rate was not as good. A schematic flow diagram for
this mode of operation is presented in Figure 8.
  A single effluent hold tank was used for both the
venturi scrubber and spray tower recirculation slur-
ries. The limestone slurry feed, additive (MgO),
makeup water (mist eliminator wash), and  clarified
liquor were all added to the effluent hold tank. A
low-pH scrubber bleed slurry was withdrawn from
the spray tower downcomer and sent to  the oxida-
tion tank, into which air was introduced through a
3-inch pipe.  The final  system bleed stream  (oxidized
/ Tower System with Bleed Stream Oxidation


   slurry) was taken from the oxidation tank and sent
   to the solids dewatering system, which consisted of
   a clarifier and a filter in series.
      A major advantage of the bleed stream oxidation
   over the two-scrubber-loop oxidation is its simple
   flow configuration. In operation without forced oxi-
   dation, the scrubber bleed stream would be sent di-
   rectly to the solids dewatering system, bypassing
   the oxidation tank. Thus, the bleed stream oxida-
   tion scheme is particularly suited for retrofit forced-
   oxidation installation.
      Batch  oxidation tests at the Shawnee Laboratory
   indicated that near-complete sulfite oxidation could
   be achieved by simple air sparging of lime or lime-
   stone slurry when magnesium ion was present in
   concentrations of 1,000 ppm  or higher. Magnesium
   ion apparently has two effects: it tends to buffer the
   pH rise from dissolving residual alkali in the waste
   slurry solids; and it tends to promote dissolved

-------
sulfite availability, allowing oxidation to take place
at a higher pH. Tests without magnesium ion were
unsuccessful because the pH rise caused by the
residual alkali (above 7) made it difficult to redis-
solve the solid calcium sulfite. Consequently, bleed
stream oxidation is not expected to work without
MgO addition unless the bleed stream is acidified
by some means, such as by adding sulfuric acid.
  A total of four limestone test runs were made on
theventuri/spray tower system, averaging about 200
hours per test. All four runs were conducted with
MgO addition (5,000 ppm effective Mg++) using flue
gas with high fly ash loading. Table 5 lists the oper-
ating results of one such  run.
  From Figure 8, it is obvious that the operation
of the test run shown in Table 5 would be similar
in every respect, except for the oxidation tank, to
an ordinary venturi/spray tower run without forced
oxidation.
                                           Table 5 shows that sulfite oxidation was near
                                         complete at 97 percent, with an air stoichiometry
                                         of 1.6. The oxidation tank pH was 5.6, which was
                                         only 0.2 pH unit higher than the scrubber inlet pH,
                                         even though excessive limestone was present (low
                                         utilization in this run due to control  problems).
                                         Other test runs showed that, at the same air stoi-
                                         chiometry, near-complete oxidation could be
                                         achieved  at an oxidation tank pH as high as 6.7.
                                         Adequate oxidation at this high pH level is unlike-
                                         ly without MgO addition.
                                           The filter cake solids content was 85 percent, the
                                         same as the solids content in two-scrubber-loop oxi-
                                         dation. However, the slurry solids settling rate was ap-
                                         proximately 0.4to 0.5 cm/min, aboutone-half that
                                         of limestone/MgO slurry (with high fly ash loading)
                                         from two-scrubber-loop oxidation.
                                           These encouraging results warrant further testing
                                         of the bleed stream forced-oxidation concept.
                                             Table 5
                 RESULTS OF A BLEED STREAM OXIDATION LIMESTONE RUN
            WITH MgO ADDITION USING FLUE GAS WITH HIGH FLY ASH LOADING
"
        Spray Tower Gas Velocity, ft/sec
        Venfuri Liquid-to-Gas Ratio, gal/Mcf
  > ...... ==        .Tower Liquid-to-Gas Ratio, gal/Mcf
  :- ~~ r::i Scrubber Slurry Solids Concentration, wt % .'(controlled)
  , ...... "*"; ..... '"""Scrubber Inlet p'H"      "  '      ' '    '  " "     ' '     
                 T.a.nk pH              .......... ,
                              ^ moles Ca/mole SQ2 absorbed (controlled)
                ...... Jvjg++ Concentratipn, ppm*    ....... ' .....  _  _ ......
        Venturj ......... Pressure Drop, in. H2O
        Oxidation Tank Level, ft
        Efflqent,Hold Tank Residence Time, min
        Percent SO2 Removal
        Average Inlet SO2 Concentration, ppm
        Percent Oxidation of Sulfite to Sulfate
        Air Stoichiometry, atoms oxygen/mole SO2 absorbed
        Percent Limestone Utilization
                                                                           4.8
                                                                           42
                                                                          in
                                                                           15
                                                                          "'5.4'
                                                                           5.6
   Slit.''
        Oxidation Tank Liquor Gypsum Saturation at 50C, %
        filter Cake Solids Content, wt %
   :JH!',;* I  III III  II......     .....'
   J : .f Effective, Mg++ concentration is defined as the total Mg"*"1"
   If:!::::!1 that Mg++ concentration equivalent to total chlorides.

                                                                           1-4
                                                                        5,380
                                                                             9
                                                                           18
                                                                          11.2
                                                                           95
                                                                        2,500
                                                                           97
                                                                           1.6
                                                                           62
                                                                          115
                                                                           85*
                                                   mnus

-------
  At the Shawnee Test Facility, slurry solids de-
watering tests (cylinder settling and funnel filtra-
tion tests) are conducted routinely to evaluate the
settling and dewatering characteristics of the dis-
charge slurries. The general results of these tests an;
presented  in Table 6 and Figure 9. The initial set-
tling rate indicates the rate at which  solids fall
during the unhindered portion of settling. A know-
ledge of this rate is necessary for sizing the clarifi-
cation  section of a thickener. Ultimate settled
solids represents the highest achievable solids con-
centration in a thickener underflow. Funnel test
cake solids correlates with the solids concentration
achievable on a vacuum-filter-type dewatering
device.
  It is clear from Table 6 that the dewatering char
acteristics of oxidized sludge are markedly better
than those of unoxidized sludge. The initial  settling
rate is  higher by a factor of 4, and both the settled
and filtered solids concentrations are higher by a
factor  of 1.4.

INITIAL SETTLING RATE

  Without forced oxidation, the average initial
solids settling rate (for slurry containing 15 weight
percent solids) rarely exceeded about 0.2  cm/min.
With near-complete oxidation, the initial settling
rate varied witfvthe mode of oxidation and the
concentration of magnesium ion. However, in all
the oxidation cases, the average initial settling rate
ranged from 0.42 to 1.06 cm/min, as opposed to
0.05 to 0.23 cm/min for unoxidized slurry.
            20
                30   40   50   60

                 % SULFITE OXIDIZED
                                  70
                                      80
                                          90   100
 Figure 9.
Effect of Oxidation on Initial Solids
Settling Rate for Lime Slurry with
Low Fly Ash Loading
                                              Table 6
             DEWATERING CHARACTERISTICS OF SHAWNEE DISCHARGE SLURRY
X
ft-

fe
p-
if"
f-
] l~
I -

fe
f^~
 	 	 
Forced
Oxidation
Yes
Yes
Yes
Yes
No
No
No
* Effective

Oxidation
Mode
1-loop
2-loop
2-loop
bleed stream



Effective Mg++
Concentration*
ppm
^ ..;,. -
0
5,000
5,000
0
5,000
9,000
Mg++ concentration is defined
Average Initial
Settling Rate






'

cm/min
i';o~
-------
For further information:

  Detailed progress reports on the Shawnee Advanced Test Program (EPA-600/2-75-050, EPA-600/7-
76-008, and  EPA-600/7-77-105) and on the earlier test program (EPA-650/2-75-047) are available
from:
                             National Technical Information Service
                                  Springfield, Virginia 22151

  A further  detailed progress report and a final report on the advanced program will be prepared. If
you wish to  be notified when these reports become available, write:
                             Emissions/Effluent Technology Branch
                         Industrial Environmental Research Laboratory
                               Environmental Protection Agency
                         Research Triangle Park, North Carolina 27711

  Previous capsule reports summarizing the EPA test program at the Shawnee Test Facility are
available by  writing:
                          Environmental Research Information Center
                               Environmental Protection Agency
                                    Cincinnati, Ohio 45268

-------
                                        // ajnBij
              SP.'/S /fJJn/s
      jo ydDjBojDiw uojjod/3
                                        -Q /
                         SUOJOJUI 00 I OJQ3     .
   gzjs ui pgguEJ s|E}sAjo gjEj.|ns iumo|BD gqj fpgzjp
  -ixo jugojgd 56 gj9M spj|os gqj <[/L ajngj.) u| -jgj
-gUIEIp UJ SUOJOJUI M9J. B A|UO JO S9}}9SOJ pUB S9JB|d
   9jjj.|ns uimoiEO ||EUIS gqj gujgq juguoduioo UJEUI
gqj 'pgzipixo jugojgd OL 3J3M spjjos gqj 'QI 3Jni-j
   u| 'A|9Ajjo9ds9J 'Ajjnjs pgzjpjxo pun  pgzipjxoun
  J^ VA1 Aq apHOI SL|dEjgoJO!LU UOJJD9|9 SUJUUBOS
   9JB 1.1. pus OL sgjngj-j -SIBIJ9JEU4 jjguj jaqjo pun
-!P
  
 jo
         Lunp|Bo jonpojd uoijBpixo gqj '(
            HjpAqioi9q ajij.|ns ainp|BD pgzipixoun
             9JE 39UMEqs JE sP!|os jonpojd
                                                       SBM
                                                                                     |IOS JSIOLU p9|qU49S9J
                                                          spj|os p9Z|pjxo gqi -ju90J9d jqgi9M 58 punojB A]
                                                        -1U9JSJSUOO SHM UOJJHJ1U90UOO Spl|OS 9qj '/V\O|.)J9pUn
                                                                     LUOJJ 9pn|S p9ZlpjXO J9JEM9p OJ p9SP
                                                                        JO J9J|jj. H U9q/V\ 'pUBS^Dinb 9>)!| (0[
                                                                   gq oj pgpugj sp||os pgzipjxoun gq^ 'iu9D
                                                                     09 oj OS SBM uouBjjugouoD spi|os gqj
                                                                   J9IJ.UEP B LUOJJ. 9pn|s pazjpjxoun JSJBM
                                                         -gp oj pgsn SBM gSnjujugo B Jo J9j|ij. B ugqyy\ -
                                                        SujjEjgdo gqj qj|M jugjsjsuoo 9J9M sj|ns9j
                                                         -J9d
                                         'UOIJBpjXO
          . qjjM jugDjgd jqgpM gg 01 59 PUB'UOJJBP
  -ixo p9DJoj jnoqjjM spijos lugojgd jqgp/v\ Q9 oj 5^
   9J9M SJ|nS9J JS9J puunj 'Spl|OS }U9DJ9d jqg]9M 08
   oj 59 SBM ggiiEj gqj 'uojjBpjxo pgojoj. qjjM fgguBJ
     spj|os jugojgd jqgpM Q5 oj Q^ gqj ui A||Bjgu9g
   ajgM spjios pgjjjgs gjEaiij|n gqj 'uoijBpixo pgojoj
      jnoqjjyw -mnisaugEiu jo uoijBpixo /o gpoai SB
  qons SJOJOBJ, Aq p9j09j.j.E jou gj9M jn'q 'AJJPJS pgzjp
-ixoun pus pgzjpjxo ugg/wjgq sgougjgj,jjp juBDijiugjs
  pgMoqs OSJB sjsgj spijos g>)BD puunj puE sjsgjuoij
             spi|os                              '
                                                                         sanos
                                                                                                   QNV
                                                                                            3iviA(iin
                                                                                               UBSBq Sui
         86-0 j.o gjBj
  -pEO| 4SEA|j.qgj
      'UILU/UIO QI- i j.o gjBj gujijjgs |Biju  ggBjgAE UE
   SBq guipEO]  qsB A|j MO| qjjM AJJHJS giuj| pgzjpjxo
  'g|daiExg joj -Auqgjis gjEj Sujijjgs gqj g'sEgjogp oj
       Ajjnjs pgzipixo u| qsB AJJ. jo sougsgjd gqj_
                  spj|os pgAjossjp (Ejojjo junoaiE
           j gqj oj gnp Ajjsugp pus AJJSOOSJA jonbi|
  ui gsBgjouj UE j.o jjnsgj gqj AjqEqojd si jogjjg amis
  -gugBui gqi -gjBJ guj|j}gs |EJJJUJ gqj p'gonpgj 'jgAg
   -A\oq 'ujnisgugBoi jo gougsgjd gq^ -gdAj JIE>UE j.o
 ssgjpjBggj 'gjBj guj|}jgs [BJJJUJ gujEs gqj jnoqE gAEg
  UOJJEplXO pgOJOJ doO|-OA\J pUB -9UO '|BJ9U9g U|
                                          UILU/LUO
    5'3 SB qgjq SB SBM gjEj gui|jjgs gqj 'sjsgj g'ujujgs
jgpuj|Ao gqj jo gtuos u| -jugojgd Q6 spggoxg uoijBp
-ixo gqj SB A||EOjjEUJBjp sgsij gjEj uj|jjgs |Ejjjuj'gqj
 fjugpiAg AjjBgp si uoijBpjxo j.o iogjj.9 gqj^ :g"uipEO|
    qsE Ay. MO| qjjM uigjsAs gaiij B joj gjBj gui|jjgs
   snsjgA uojjEpjxo juaojgd jo jo|d B  sj 5 9-ingjj

-------
^-
V


	 ss
i-J
Lg
L
01
ez.
9L

SD11OC a'MEO
jsai puun j sSEjaAV a

snuiui ++
IV

Lti
IL
99
	 til
o/ JA\
SDIIOC P3UJ3Q
JElUlJin 3SEJ3AV
sspuojqo |EJO:
IAJ |EJOJ 3qj SE |
SO'O
OZ'O

SL'O
LO'L
90' L
UIU1/LUO
ajEy ui|jj3S
.Enm.a^AV
I 01 aU3|EAinb3
?'::;;;,',; !-iJEai 9JB 3pn|S pszipjxo j.o
                                                             UU3JEM9p 3qj JEqj 9 9|qEl UJOJ^ JE9|0 S! I]
                                                                                                  '90IA9P
                                                             SUU3JEM9P 9dAj-J311!J.-OinnOEA E UO 3)qEA9!qOE
                                                        UOIJEJJU90UOO SpnOS 3qj qjjM S3JE19JJOO SpljOS 9>]EO
                                                          JS9J |9UUnj 'M01.p9pun J3U3>)0!qj E Ul
                                                         -uoo spi|os 3iqEA9iqoE isgqSiq 9
                                                             pgiiiss gjEOiuin -J9U9>]0!qa E j.o UOIJ09S UOI;EO
                                                         -IJ.IJEP 9qi SUIZIS JOJ. AJESS903U SI 3JEJ  Sjqj J.O 3p3|
                                                        -M'OU>I V Suijiass j.o uoijjod psjspujqun sqjSuunp
                                                             ]|Ej. spijos qojqM IE  gjEj gqa ssjEOjpin SJEJ ui|a
                                                          -j3S |E!imi am "6 aJnSij  puE 9 ajq^i ui p9ju3S3jd
                                                        9JE sjsaj a'saqi J.o sjjnsgj  |EJ3U9 am -saujn|s 3JEqD
  9qj 91En|EA9 oa Apuunoj psjonpuoo SJE
   -EJH!! |3uunj. puE Suijuas japui|A3) sjsgj SuuajEM
   -ap spnos AJJHJS 'AiMpEj jssi 99UMEqs aqi W

-------
          gsn gAijonpojd oj gjjs |Esods|p gqj
   pus '|ios qjiM j| ss9jp-doj 'ggpnjs  pgjgjEMgp gqj
   joEdwoo oj gjqjssod 3q  p|noqs j| fpgAo|dujg |Eso]d
  -s|p jo poqjgui 341 jo ssgjpJEggy -AEM jEqj ji  dgg>|
     OJ pUE pgjgjEMgp Jj J3g OJ Sj H!JpUB| E SB ggpn
      pgzjpixo guisn ui juguiajmbgj A'jEiuud gqx
                                   jgqqnjos gqj oj
   >pEq BgjB puod guipunojjns gqj LUOJJ pgduind aq
             j9ij.jjH|o E LUOJJ AJJHJS L)JJM gpEoi 9||d
 '9|dlUEX9 JOJ '9gpn|S p9Zlp)XO 3LJJ JO
  gU!J9JEM9p p90UEL|U9 9L)J JO 9gEJUEApE
 UEO JELJJ S9jnp9oojd [Esodsip J9i|jo 9q AELU
     JE9A E JSOLU|E JOJ. WJOJ. OlSEq SJj  pgUjEJUJEUI
   9||d 9l]J 'SS9|9qj9UO|sj 'UOISOJ9 OJ JOU pUE 9gEddj
 IEUJ9JUJ OJ 9np 9q OJ JE9ddE 9||d 9l|J UJ SMOJJPJ 9LJ L
'P9SE3JOUJ qjgU9JJS guUE9q 9qj pUE J9AO p9JSHJO 9|jd
  9qj 'gUJ|JJ9S pUE UjEJ J9JJV 'qigU9JJS gU[JE9q JOOd
 PEL) pUE p9JEpj|OSUOOUn 9J9M Spl|OS 9S3qj 'A[|EJJJU|
                                                         9|jd siqj jo gjnjoid E s| n ajnSy -a|jd B ui p3j|sod
                                                           -9p SEM J9JJJJ OinnOBA E LUOJJ 9>)ED J9J||J SB Spj|OS
                                                          jugojad 58 JnoqB JE pgujBjqo gu|pEO| qsE Ajj MO|
                                                         qj|M ggpnjs pgzjpixo 'Apnjs (Bsodsip jgqjouE uj
                                                                                              pgjojjuooi
                                                             gU|9q A|JU9JjnO 9JE UIEJ JO SJ09JJ9 3qj fuOSE9S
                                                         AUJBJ jsjjj sj| qgnojqj gujog A|ju9S9jd si puod 9iy_
                                                         sABp J7 OJ  UiqjjM P9J9A009J SBM qjgU9JJS p9UjEJp
                                                         |EU|gjJO 9qi -JU90J9d Oe OJ ^l JO SSO| qjU9JJS gU!
                                                         -JB9q AjEJOdlU9J B p9JEOjpUJ 9>|BD p9U9JS(OOI9J pUB
                                                         p9U|BJp JO SJS9J AjOJEJOqE-] -J9J9LUOJJ9U9d p|9IJ B
                                                       Aq p9U|aiJ9jap SB 'isd Q0 JnoqE jo qjgu9Jjs guuB9q
                                                       E pEq ggpn|s 9qj 'gumiBjp J9jjy 'Asp E wqjiM spj|os
                                                        ju90J9d  98 oj 08 J spijos ju90J9d gg tuojj pgujEjp
                                                        ggpn|s 9L|j_ -ggpnjs jo J99j g puB 'PUBS jo sgqom 1 1
                                                        '|9AEjg E9d JO S9qOU| 9 A|9JEOIIXOjddB qjjM plB|J9AO
                                                           SEM gujdjd uiBJp 9qx 'urejpjgpun AjjAEjg E qjjM
                                                       P9ddmb9 puod E  ui pgoE|d SBM Moyjgpun JSIJUEP E
                                                      LUOJJ spj|os jugojgd jqg|9M gg jnoqB PUB gujpEoj qsE
                                                        A|j MO| qjjM 9gpn|s pgzipjxo 'Apnjs guipuod B u|
                                                                                                  'DldOJJ
                                                         -oxjqj si pUE spjjos jugojgd jqgi9M 99 oj Q? jnoqE
                                                         OJ A|UO SJ9J(!J |EU9JBLU pgZJpjXOUfl  '|!OS JSJOLU 9>JI|
                                                         S9|pUEq pUB 'SpjjOS JU90J9d 08 UEL)1 J3JJ9q OJ SJ9J
                                                         -|ij '|EU9jBUJ pgzjpjxoun UEqj jgjsEj gpnjmgEUJ-jo
                                                         -J9pjo UE SB qonuj SB Aq S9|jj9s ggpn|s
                                                              S3ianis nvsodsia aoanis anziaixo
     99UMEqs JE pgjo|dx9 Ajqgnojoqj 9q oj J9A SEq
J09JJ9 sjqi -|EjsAjo 9qj jggjEj gqj gujij gougp|S9J 9qj
 J9guo| 9qj : |EjsAjo ainsdAg gqj jo 9Zjs 9qj JD9JJB OSJE
   AELU >JUBJ uojjEpixo 9qj ui gujjj gougpjsgj gqx
                                   jgqgjq  U3A9 gq
  AELU 9jEJ guj|jj9s 9jEj]ns Lunp|ED gqj jo sjiwi] gqj
 'qsE A|j jugojgd jqgj9M  \_ oj dn UIEJUOO guipEoj qss
A|J M0| qjjM SEg 9n|J LUOJJ Sp||OS '93UMEq5 JE '90UJS
 UjUJ/UJO Y1 SBM JILUJI 9qj 'gUjpEOJ qSE A|J MO]
    Ajjn|s pgzjpjxo jo-j -UJLU/LUD g~i 9q oj
  guipEO|  qsB A|j qgjq  qjjM AJJHJS pgzjpjxo joj JJLUI|
 9qj 'Ajj|IDEJ JS9J_ 99UMEqS 9qj JV '3JHJ J9MOJS E JE
 S9|JJ9S 90U9q pUB S9|0!JJBd 9JBJ|nS LUnp|BO 9qj qjjM
             LigqM gujj Apujgjjxg si qss A|j -JOJOBJ
        ii  9qj gq ABLU qsE A|j gqj 'Ajjn|s  pgzipixo jo
     gqj ui jng -gjEJ guj[jj9s IEJJJUI 9qj ui JOJOEJ gui
 -JJLUJI gqj 9JE ssuij gjijjns LunpjEO  9qj 'Ajjnjs pgzjp
-|xoun jo gsBO gqj uj jBqj jqgnoqj  A|jugjjno sj j| '

-------
                                          !MO '!
                               AouaSy uojjoajojj |Ejuauiuo.nAug
                                 UOIJEUIJOJ.UI tpJEasa^j |Ejuaiuuo,iiAU3
                                                                           :UHIJM Aq
          A}||IOEJ JSSJL aauM^qs aqj JE uiEjSojd jsaj yjg aqj SujzuEiuuins sjjodaj 3|nscteD
                               AouaSy
                                             (EjuauiuojiAug |Eijjsnpu|
                                :ajUM 'a|qE|jEAE auiooaq sjjodaj asaqj uaqM pa^ijou aq oj qsjM noA
  ^1 -pgjEdajd sq ||IM LUEjSojd paouEAps aqj uo jjodaj JEUJ^ E pus jjodaj ssajSojd pa|iEjap jaqjjnj. y
                                                                                         tUIOJJ.
    3|qE|!EAE gas (yH/O'S L-Zl 'OS9-V ^d3) uiEjSojd jsaj aaijjEa aqj uo puE (sOL-Z.Z.-Z./009-VdH PUE '80Q-9Z.
-/./009-Vd3 'OSO-SI-Z:/009-Vd3) LUEJOJJ jsaj. paoueApy aauMEqs aqj uo sjjodaj ssajSojd panejaQ

                                                                        :UOIJEIUJO.|.UI jaqjjnj. uoj

-------