United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
EPA-600/7-80-030
February 1980
Survey of Dry S(>2
Control Systems
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, u-.e-
Protection Agency, have been grouped into nine series. These nine oroa
gories were established to facilitate further development and appucauon
vironmental technology. Elimination of traditional grouping was (J°"*r " ,ds'
planned to foster technology transfer and a maximum interface in relate
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from i
effort funded under the 17-agency Federal Energy/Environment Researcn
Development Program. These studies relate to EPA's mission to protect the Pu°
health and welfare from adverse effects of pollutants associated with energy sy -
terns. The goal of the Program is to assure the rapid development of d°mes
energy supplies in an environmentally-compatible manner by providing the
essary environmental data and control technology. Investigations include ana y-
ses of the transport of energy-related pollutants and their health and ecologic
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield. Virginia 22161.
-------
EPA-600/7-80-030
February 1980
Survey of Dry SO2
Control Systems
by
G.M. Blythe, J.C. Dickerman,
and M.E. Kelly
Radian Corporation
P.O. Box 8837
Durham, North Carolina 27707
Contract No. 68-02-2608
Task No. 71
Program Element No. INE827
EPA Project Officer: Theodore G. Brna
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ABSTRACT
The status of dry flue gas desulfurizatlon (FGD) processes In the
United States for both industrial and utility applications is assessed. The
assessment is based on reviews of past and current research, development, and
commercial activities. Systems covered include: (1) spray dryers with
either baghouse or electrostatic particulate (ESP) collectors, (2) dry
injection of alkaline material followed by baghouse or ESP collection of
wastes, and (3) various other systems, such as coal-alkaline material feeds
to a combustor and passage of flue gas through a fixed bed of alkaline
material.
A summary of dry FGD processes, including key features of three types
of dry systems and commercial systems, is provided. Limited economic data
are also presented. Conclusions and recommendations are given on the
potential role EPA can take to advance the overall environmental acceptability
of dry FGD systems as viable SC>2 control alternatives.
ii
-------
CONTENTS
Abstract ii
Tables vi
Figures v
Acknowledgement vii
1.0 Introduction 1
1.1 Report Organization 1
1.2 Technical Glossary ..... 2
2.0 Summary 4
2.1 Process Assessment 9
2.1.1 Spray Dryer Based Systems 9
2.1.2 Dry Injection Process 12
2.1.3 Combustion of a Coal/Limestone Fuel Mixture 14
2.2 Comparison of Dry and Wet Scrubbing for S02 Removal 15
3.0 Conclusions and Recommendations 17
3.1 Spray Drying/Particulate Collection 17
3.2 Dry Injection/Particulate Collection 18
3.3 Combustion of Limestone/Coal Fuel Mixtures . 19
4.0 Dry FGD Research Review 20
4.1 Dry Injection/Particulate Collection 23
4.1.1 Owens-Corning Fiberglass, Laboratory Tests,
August 1970 23
4.1.2 Air Preheater at Mercer Station, March 1971 27
4.1.3 Wheelabrator-Frye at Nucla Station, July 1974 29
4.1.4 American Air Filter, Laboratory Tests, 1976 32
4.1.5 Wheelabrator-Frye at Leland Olds, March 1977 34
4.1.6 Grand Forks Energy Development Center, DOE: Bench-
scale Dry Injection/ESP or Baghouse Collection,
1975 to Present 36
4.1.7 KVB, Incorporated, Bench-scale Tests, Late 1977 to
Present 37
4.1.8 Carborundum, Dry Injection/Baghouse Collection
Pilot on Stoker-Fired Boiler. 1976 - Present .... 38
4.2 Spray Drying/Particulate Collection 39
4.2.1 Atomics International (Rockwell) at Mohave
Station, 1972 39
4.2.2 Koyo Iron Works, Pilot-Unit, 1973 41
4.2.3 Rockwell/Wheelabrator-Frye at Leland Olds
Station, 1977-78 43
4.2.4 Joy/Niro at Hoot Lake Station, 1977-1978 46
iii
-------
CONTENTS (Cont.)
4.2.5 Babcock and Wilcox, Pilot Scale Spray Dryer/ESP
at Velva, 1978 52
4.2.6 Carborundum/DeLaval Spray Dryer Pilot Plant at
Leland Olds, 1978 54
4.2.7 Bechtel Power Company, Conceptual Spray Dryer/
Baghouse & Dry Injection/Baghouse Study, August 1978
to Present 55
4.3 Other Research 55
4.3.1 IMC Corporation, Bench-scale Fixed Bed,
June 1970 55
4.3.2 Nagoya Institute of Technology, Multi-Stage
Bed, September 1973 '. 56
4.3.3 Stearns-Roger/Superior Oil, Fixed-Bed &
Countercurrent Reactors, 1974 60
4.3.4 R.W.E. Tests in Germany 63
5.0 Current & On-going Activities 65
5.1 Babcock & Wilcox 65
5.2 Buell/Anhydro & EPA/Buell 72
5.3 Combustion Engineering 74
5.4 DDE/Grand Forks Energy Technology Center 75
5.5 DOE/Morgantown Energy Technology Center 77
5.6 DDE/Pittsburgh Energy Technology Center 80
5.7 Ecolaire 81
5.8 Energy & Pollution Controls, Inc 83
5.9 EPA/Battelle-Columbus Labs 86
5.10 EPA/Energy & Environmental Research Corp. (EERC) 87
5.11 EPA/Kerr Industries 90
5.12 Joy/Niro Joint Venture 91
5.13 Kennecott Development Company (Environmental Products
Division) 95
5.14 Koch Engineering 97
5.15 Mikropul 97
5.16 Research-Cottrell 102
5.17 Rockwell International/Wheelabrator-Frye Joint Venture . . . 103
i
References 109
Appendix Ill
iv
-------
FIGURES
Number Page
2-1 Typical spray dryer/particulate collection flow diagram 11
2-2 Nahcolite dry injection flow diagram 13
4-1 Process schematic - gas and nahcolite into baghouse.
Wheelabrator-Frye at Nucla Station 30
4-2 Hoot Lake pilot plant flow sheet 47
4-3 Flowsheet of pilot plant operation with partial solids
recycle 49
4-4 Y-jet slurry atomizer 53
5-1 Laramie River Station flow diagram .' 70
5-2 Flowsheet for bulk evaluation studies of modified dry
limestone process 78
5-3 Effect of moisture on S02 removal in a fixed limestone bed ... 79
5-4 Ecolaire's mobile demonstration dry FGD unit 82
5-5 Air pollution control (SO^) reactor for dry reagent ....... 84
5-6 Antelope Valley station gas cleaning system 93
5-7 Mikropul spray dryer/baghouse dry FGD system 101
-------
TABLES
Number Page
2-1 Summary of Key Features of Dry FGD Systems 5
2-2 Summary of Key Features of Commercial Spray Drying Systems .... 7
4-1 Listing of Research Activities Conducted Prior to April, 1979 . . 21
4-2 Variable Test Conditions - OCF Laboratory Testing 24
4-3 Percentage SC>2 Removal with Continuous Sorbent Injection -
OCF Laboratory Tests 26
4-4 Summary of Average SC^ Removal and Sorbent Utilization
Air Preheater Corp. at Mercer Station . 28
4-5 Summary of Results of Nahcolite and Trona Tests -
American Air Filter 33
4-6 Temperature Effects on SC^ Removal - Koyo Spray Dryer 42
4-7 Open Loop ACP Pilot-Plant Test Conditions -
Rockwell at Leland Olds 44
4-8 The Relationship between Spray Dryer Temperature Drop and S02
Removal at a Constant Stoichiometric Ratio of 2.5
Joy/Niro at Hoot Lake 50
4-9 Results of Screening Tests on Sodium Carbonate and Sodium
Sesquicarbonate - FMC Corporation 57
4-10 Experimental Schedule - Nagoya Multi-Stage Bed 59
4-11 Experimental Results - Nagoya Multi-Stage Bed 59
4-12 Test Results With Pilot-Scale Countercurrent Reactor 62
5-1 Summary of Current R & D and Commercial Activities 66
5-2 Dry Reactor Chronological Test Summary, August 1978-
October 1978 85
5-3 S02 Removal in the EERC Low-N0x Coal Burner 89
5-4 Spray Dryer Test Variables - Strathmore Paper Company 100
vi
-------
ACKNOWLEDGEMENTS
The authors of this report appreciate the cooperation of several
process vendors and research organizations involved with dry SO- control
technology development. The information they provided has been instrumental
in preparing a complete and accurate assessment of the status of dry FGD
technology. The contributors are too numerous to list here, but have bet-n
cited in the appropriate sections of the report. The partial support of
this survey by the Department of Energy via pass-through funds to the
United States Environmental Protection Agency in fiscal year 1979 Is
gratefully acknowledged.
vii
-------
SECTION 1
INTRODUCTION
The purpose of this report Is to summarize the status of dry flue gas
desulfurization (FGD) processes in the United States, for both utility and
industrial application. Throughout this report, dry FGD will be defined
as any process which involves contacting a sulfur-containing flue gas with
an alkaline material and which results in a dry waste product for disposal.
This includes (1) systems which use spray dryers for a contactor, with
subsequent baghouse or electrostatic precipitator (ESP) collection of
waste products; (2) systems which involve dry injection of alkaline material
into contact with flue gas, and subsequent baghouse or ESP collection; and
(3) other varied dry systems which include concepts such as addition of
alkaline material to a fuel prior to combustion or contacting flue gas
with a fixed bed of alkaline material.
This definition of dry systems excludes several dry adsorption or
"acceptance" processes, such as the Shell/UOP copper oxide process, or
the Bergbau-Forschung adsorptive char process. It was felt that the status
of these processes has been documented in other EPA reports and further
documentation would not be necessary here.
Also excluded was the regenerable Rockwell Aqueous Carbonate Process
(ACP) which, although it does use a spray dryer for a flue gas contactor,
does not fit the limitation of this study as being a "throwaway" system.
However, the open loop, spray dryer contactor portion of the Rockwell
process has been adapted for a "throwaway" system, and as such as been
included here.
1.1 REPORT ORGANIZATION
Section 2 summarizes the key features of the three major types of dry
FGD processes considered in this survey and includes:
A discussion of the state-of-the-art for each type of process
(spray dying, dry injection, and combustion of a coal/limestone
fuel mixture).
A process assessment for each type of dry FGD control technology
that includes a general process description, key design parameters,
and special considerations for the application of the technology on
a commercial scale.
A comparison with conventional wet lime/limestone scrubbing
technology.
-------
Section 3 presents conclusions and recommendations resulting from this
study. Section 4 reviews research efforts in dry FGD prior to April of
1979 in considerable detail, and Section 5 presents a summary of current
and on-going dry FGD activities for both research and development and
commercial projects. In Section 5 each company involved with dry FGD systems
is discussed with respect to the type(s) of systems being developed and
marketed; past, current, and future research and development programs; and
commercial sales summaries. This report also contains a reference section
and an appendix listing metric conversions since virtually all work has been
reported in engineering units.*
1.2 TECHNICAL GLOSSARY
Definitions for several terms that are used frequently throughout this
report to describe the operation dry FGD systems are defined as follows:
Stoichiometry for dry scrubbing is defined as the moles of fresh
sorbent introduced to the system divided by the moles theoretically
required for complete reaction with all of the SO. entering the system
whether or not it is all removed. This is opposed to wet scrubbing
where Stoichiometry is generally based on moles of SO. removed by
the system.
Sorbent utilization is defined as the percent SO- removal by the
system divided by the Stoichiometry:
moles SO2 removed
moles S02 entering system X
[moj-e^ sorbent entering system ~|
moles sorbent required to react with SO. entering
= (percent) utilization
If one defines the sorbent for a calcium-based system as CaO and
the sorbent for a sodium-based system as Na.O, one mole of sorbent
reacts with one mole of SO . Consequently, the above expression
reduces to:
(percent) utilization - moles S02 removed
moles sorbent entering system
It is EPA policy to report measurements in the International System (SI)
of units. For clarity of presentation, units used in this report are
those commonly used in engineering calculations in the U.S. Conversion
factors are presented in the Appendix.
-------
Since the moles of sorbent do not include alkalinity from other
sources such as recycled fly ash, it is possible to see apparent
utilizations of greater than 100 percent. That is, the alka-
linity in recycled fly ash can react to remove SO,,, so that
there are more moles of S0» removed than moles of fresh sorbent
feed.
A spray dryer is defined as any apparatus in which flue gas
is contacted with a slurry or solution such that the flue gas is
adiabatically humidified and the slurry or solution is
evaporated to apparent dryness. For FGD applications the
material dried is often a calcium-based slurry or a sodium
solution which reacts with flue gas sulfur during and following
the drying process. The spray dryer can use rotary, two-fluid or
nozzle atomization, and the vessel can be anything from the
back-mix reactor typically used in spray dryer technology to
a large horizontal duct.
Dry injection is defined as the process of introducing a dry
sorbent into a flue gas stream. This can take the form of
pneumatically injecting sorbent into a flue gas duct, pre-
coating or continuously feeding sorbent onto a fabric filter surface,
or any similar form of mechanically introducing a dry alkaline
sorbent to a flue gas stream.
Coal/limestone combustion is defined as the process of burning
a. mixture of coal and limestone whereby the S02 released
from the coal reacts with the limestone to form solid calcium
salts that are collected with the ash. Two specific combustion
processes are discussed: one involves burning a coal/limestone
pellet in a stoker fired boiler, and the other involves burning
a pulverized coal/limestone mixture in a low NO burner.
X
-------
SECTION 2
SUMMARY
There are currently three major types of dry FGD systems being
developed today: spray drying, dry injection, and combustion of fuel/
limestone mixtures. Of these systems, spray drying is currently the
only one being developed on a commercial scale. Table 2-1 summarizes
the key features of these three main types of dry FGD systems.
Four companies have sold commercial spray dryer-based systems. The
Rockwell International/Wheelabrator-Frye joint venture sold the first
utility system for the 410 MW Coyote Station to Otter Tail Power Co.
Rockwell has also sold a 65,000 acfm industrial boiler system to Celanese.
The Joy/Niro joint venture sold a 440 MW system for Basin Electric Co.'s
Antelope Valley Station} B&W sold a 500 MW system to Basin Electric
for their Laramie River Station. Mikropul Corporation has also sold a
dry FGD system, a 40,000 acfm unit, for Strathmore Paper Company's power boiler.
The utility systems are expected to start-up in 1981 and 1982, while the
systems on the industrial boilers are expected to be operational in 1980.
Table 2-2 summarizes the important features of the five current commercial
spray drying systems.
Research on dry Injection of alkali powders into the flue gas stream
showed this technique to be a viable S02 control method that required little
additional process equipment. However, development and marketing of
commercial dry injection systems has not been forthcoming due primarily
to sorbent (nahcolite) availability problems.
Preliminary studies have been conducted to determine the potential
for reducing SO^ emissions by firing a coal/limestone pellet in a
spreader-stoker boiler. Initial results are promising (75 to 80 percent
retention of the available fuel sulfur). EPA funded tests are continuing on
industrial size boilers. Another recent study has shown in preliminary
results that pulverizing limestone and coal together before combustion
in a low NO burner can be effective in controlling S02 emissions (up
to 88 percent S02 removal).
Experimental studies have also been carried out to Investigate
removal of SO- by passing the flue gas through a fixed or fluidized bed
of alkali sorbent. The relatively large pressure drops encountered in
these types of systems make them undesirable for commercial-scale
applications. Most fixed bed studies have been aimed at investigating
the sorption reaction kinetics.
-------
TABLE 2-1. SUMMARY OF KEY FEATURES OF DRY FGD SYSTEMS
PROCESS TYPE
IPIUY DIVING/
PAITICUUTE COIUCTKM
OIT
PAKTICUUn COU.ICTIOH
COMUITION OF COAL/
L1MUTONI rUII. MIX
UNIQUE DESIGN rEATUUS
bploy* • epray dryer equipped with
eto*iior(e) to epray eorbent eolutlon
or slurry Into Incoming SO, laden flu*
tea. Tha •pray dryer .• coufled with •
baihouae (or poaflbly ESP) to provide
collection of fly ..eh >nd .mtralTi..d
product .lolidB.
Pn.M.m..tle lojtctlon of dry 43!kail
•orbant into • flu* §•• itrw.. with
•ubi«qu«nt p4irticul.it* collection. In-
jection point variM fro* iMMdiattly
•fttr the boilgr to juit upetr«-w of
tlM COllmeCtlan drntVlf* (mtaghOUmttt OT
ISP). A b*|houM !• uMally tvploytd
•• contld«rabl«90, rnovt.1 occur*
•croM th* filter cak* collected in
th* bag aurfac*.
Th* ...oat pro*ii*ln| t»chnuli>«l«i in
thli *r*a appear tu be 1) coabutu.ttn
of a coal/ll»e*tone p*ll*t in • epr»ed-
er »tok*r boil*r. and 2) co*>bu«ti«n of
a coal/llaeeton* fuel ftixtu.-* in • low-
MO burner. Th*. lan.tr adlabatlr fl-*r
tadperatur* reaultlni fron th* twt>-
etasa coaibuetion ech..** aaiployvd In
both technoloflea appear* to increae*
the available It-tee ton* reecilvlty.
REAGENT(a) USED
Nn.S (IK 8EACENT
UTILIZATION
RAMC.ES OF SO, HI310VAI.
PARTICULATE REMOVAL
SPECIAL PROBLEMS OR
ADVAMA'.tS
IIEVKUIPMENI STATUS
REPORTED CAPITAL
COST ESTIMATES
Sodiun carbonate, lime, trona (NaCO.
NaHCO, 2H,0) and llMlton* hav* all
b»n teetod. Planned commercial
•yetema will u** aodlum carbonate or
Ha*.
Sodiuai-ba»d aik.ll.: aodlu* carbo-
nate, .odium blcarbonat*. trona and
nahcollt* (60-701 N.HCO,). Lima «nd
limeiton* have bean lnv*.ti|atcd but
both require tMH'P flue fa. for
elinlflc.nl SO reeov.l.
Llmeetone (Pellet *l»o requlrp.
•om« type of binder).
SO to IOCS far eadlum-b**ed .1U11.. (>e|hnu.e *y.t*m>) 40 to 60J for Ce i S r.tl». of 'tl h*» be>n »»d In
30 to )OX for Urn* on * "once-through" nehrollte, eodlum btcerhonet* el htah- voel/1 Imeittme pellet, while « III
ba*ia. 80 to 901 for lime with parti.I eet SO. removal condition., 301 01 rellu wee need fur the low-M^ Ihirnet
recycle of product aolida. 201 or leaa lea. lir llm.eton. even it hl.h flue leel..
a atroni function of tha outlet temper- at higher aaa temperaturea and I* •
etur* of the gee; utlllaetlon Incree.e. function of eorbent feedlni eelluxl.
e. the dryer outlet temper.tur. of the
g.e approachee It* adlabatlc eeturatltm
temperature.
(For Inlet SO,-1000 to 2000 pom) 10 60 to «01 for aodlum-baaed ellull Cuel/l leeelone p.I I.I. repulr.il )'•
to 905 for eodlum-baaad alk*lli. *^ ayaleme d.pendlm on atochlometrIt to mix uf the ivillehlv .iillui In
to 60! for lime on e "onc.-lhrou»h" ratio, flue |e. tmp.r.tur. «nd n-.tluiJ Ihr fuel. Pr.1 Imln.rv reeull. In !••!«
bull. 80 ti> 8it for lima with p.rll.l of feeding. »OJ rmmuvala hev. been wllli lou W< burner* lndlr.tr thui Hut
recycle of product *olld*. L**e than achieved with nahcollte at ?90*r' r.t.ntlon |S erhlrv.hlr.
30!; for limeaton*.1 tmmperaCuree. 20 to 30X for llmeelone
at hl|h temper.tur.. (600*'M.
Both baghoueea and ESP'a have conala- Dry injection ey.tem. with baghou.ee Completion i'l thr co.l/llmr>tonr
tently achieved »9+I remov.l of en- removt 99+1 of entrained product fuel ml»ture will reeull lmire>e>l
trained product aolida and fly aih. eolld. and fly aah. ESP', demonetr.te parttcul.te loedlng.
Seghouee. have the advantage of provl- 9942 remov.l aleo, but SO removel le
ding for additional SO, removal .croa* much lower thai In beghoulrn, Aleo thr
the filter c*k. that cflll*cta on th* increeeed inlet-gr.ln loading will
fabric aurface. However aoie reporti affect ESP ailing.
claim it 1. poaalble to more clo.ely
epproach the ediebatlc eaturatlon temp-
erature of the g» with an ESP down-
•treem of the .pr.y dryer.
Spray drying re.ulta in a dry »a.y The drv product r.eultlni I. qultr The .ddltlon.1 rn.te of pr*i>eilnp
to handle wa.te product. However, when weter aoluble and teachability anil thr rnal/lliiealnnr lurl and nwinii
.odium alkali, ar* u.ed th. product. etablHtv problem, .re likely to nrrur urr.t.t .eunnti> ot ..h .rr .inni! <
are quite water aoluble, creating dl«- In dl.poaing the wait* eolld>. The u*e r.nilv I.IP, th.n , unvenllonal v>t
poaal problem.. Water and energy re- ol relatively Innp.n.lvr re.u'nt errul.blng .y.lm m.t.. 11,-wry.t
qulrvient. ar. 1... then for conv.n- (nehcollt.) end mlnlnel .qulp**nl r.- th.nr t.chnolool.. nev. only l.pr,,
tlonal "wet" llme/llmeatone eyitem.. requirement, make dry Inlecllon eronn- applied on a em.ll ereli Iralu.11 ul
High .ultur roel application, may hr mlrally .ttrarllve. The two me lor tyre l>i>llrr evulemi,.
limited but ere being inve.tfg.ttd drewberk. ere the .v.llehlllty ol
further. nehcollte Inenount. required for com-
mercial eppllcetlon. etwt th. we.t»
.llepaMl problem.
Spray drying 1. currently the only Alllnnigh dry Injerl Ion ha> heen i:,m.ld»r«|.lr woi k tmelne u> ,lr\.lti|.
commerrlally applied dry FGD technology «hown to be technlc.lly fea.lMx Mir lerhnoloilli'i li" rnmmen I.I M -il,
with 1 utility (400-100 HW e.ch) iomm.rrl.1 ippllcatlon I. at e «|.|,|h .1 u.n.. .Itlxuith Imlu.irl.l . ,«.-
ayalaie being conatructed (elartup In atandatlll due to unrertelnlle. In merrul .ppllr.tlon. loitk i>nNni.lii|t.
19«1; 12 and (3) and two Induatrlal aorbent availability. F.PA I. currently lundlnu contlmml
•yatam* to atartup in late 1979. pilot plant te.tlnt on lndu.lt|M|
Several other companlea ar* conduct Inn holler, and more complete t..t wink m,
exlenilve »»P. program, toward . commer- low-Nil burner, lie. been nropo.,,1
clal ayatam. lnd ^ under review hv lltf KI'A.
MO to 1120/KV for utility ayatmma 12) u 130 million (SJO/KV lor M10 Sn/ton to uroducr I Imeelimr/i ".I
(•00-tOO MV).° HW) for It S coal with a heeling pellet.'
Sl.O to M.J million for indu.trl.l v"" of I0-500 >w"-'
.y.temi (10-25 MW).«
Baaed on pilot plant acale-up
deelgne.
-------
TABLE 2-1 (Cont.)
«a
Utilization and removal figures are quite site specific and the given
values should only be taken as a general indication of system performance.
Removal and reagent utilizations may be lower or higher depending on
system design, SC>2 inlet concentration, stoichiometric ratio, flue gas
inlet temperature, temperature drop over the spray dryer and fly ash
alkalinity.
Source: Janssen and Eriksen, "Basin Electrics Involvement with Dry FGD"
presented at EPA Symposium on FGD, Las Vegas, Nevada. March, 1979.
Source: PedCo Inc., "Survey 01
1979". EPA No. 600/7-79-067B.
Source: Lutz, S.J., et al., "]
Filtration for FGD". EPA No. 600/7-79-005. January, 1979.
° Source: PedCo Inc., "Survey of Industrial Boiler Dry FGD: First Quarter,
Source: Lutz, S.J., et al., "Evaluation of Dry Sorbents and Fabric
Source: Dickerman, J.C., personal communication with Jack Wasser, EPA.
-------
TABLE 2-2. SUMMARY OF KEY FEATURES OF COMMERCIAL SPRAY DRYING SYSTEMS
VTSTBI
iDcbell Afceelabr. tor-frye
fotter Tall Pu»€J Co. '» Coyote
Statloa. Dolt 1. Bold, m
Start «p dice: Jwae. 1H1
Jay/BUro flaela Electric'*
Antelope Valle* Static*.
Halt 1. Bealah, »
Surt ep Hate: April H62.
Babcoek end trllcoa flaau
Electric 'a Laraml« liver
Statloa. Halt 1, wheat land.
«yo»ln».
Hlkropul estratbaore
Paper Co.. Uoronoco. tfaaa.
Start up date: Late 1«'»
Hockwell/Uhtelibrater-rrye
•Celaaete Corporation
GEra»TDE
CAJACITT sTSTm DEsanrnoR
410 NU Four Boven 46 ft. alaOMter apray
(l.«»0.00e acfai) tovera la parallel, each equipped
with three ee»tritaf/al atoaiiien.
coolaed ITC/lb-lowT
'-, AJb
Llae liorth Dakota
llcnlte. Average
S-^}.78t Maxlaua
5-1. I2t
Ll«e Hroalaf ana-
*itu«laana
Averate 5-0.5*1
n»wai S-C.Mt
«!iO HT/lb-ltW
8- Art
lla>e : t? : 1/21 S
llee 1 to 71 S
erww SOBBFJTf
rfAtairrrF CTn.i2*nc* EFTIMTTT. rsTElAro
u.«aa»ltl. OCA*A!.TEE OUT7AL COTT OPEMTtllC COSTS
7« for all !»•"!« (.r-; SJ2.OOO.MO S4.SaO.OOO/-n S2.5 .11. /
<5?«/tV) kwhr) Doea act Include
waate dlspoeal coat.
*2I for avg. S coal No, ;«porte.' S-».«*5,10P S2.270.834/yr (O.S mil./
781 for _u. S coal (S113/IV)* kwtir)" aorbem coat
(liae) - SJ, 102. 500
(SMVton baala) Doea not
include va*te diapoaal cost
• 51 for avj. S coal Kaiiaua atochloejetrtc $4*. 807. 000 S2.571.000/yr (0.7 all./
•01 for au. S coal r.tlc o, ah-ut 1.12 (585 fix) kwhr)k aorbeac coat
(llae) ' S1.3W.570
(S6O/ton baala) Doea not
nc u e wa> e apoaa
75Z on 35 S coal tfcnlnua atochloaetrlc S1.40O.OOOC S162,OOO/yr (2.3 alia/
ratio of 2.7S kwhr)0
85, Hot Reported Sl.250.ODO NOT AVAILABLE
d, MnryUtatl
St«rtlii« B*te: L*t* 1979
Dry M«t« product vltl
be MlMed with llaTM and trucktH to
area.
-------
TABLE 2-2 (Cont.)
a
Capital cost for complete turnkey installation from air preheater outlet
to stack connection, excluding I.D. fans (1977$). Source: Johnson, O.B.,
et al., "Coyote Station, First Commercial Dry FGD System", Presented at
41st Annual American Power Conference Meeting. Chicago, IL April 23-25,
1979.
b Evaluation based on 35-year life, annual plant factor of 75% (1981$).
Source: Janssen and Eriksen.
Total installed cost of entire FGD System (1979$). Source: PedCo Inc.,
"EPA Industrial Boiler FGD Survey: First Quarter, 1979". EPA No. 600/
7/79/067B.
Purchased cost of equipment, silos, internal ductwork, slaker, pumps
(excludes I.D. fan, installation, ash handling system, and electrical).
(1979$).
-------
2.1 PROCESS ASSESSMENT
2.1.1 Spray Dryer-Based Systems
In these systems, flue gas at air preheater outlet temperatures
(generally 250 to 400°F) is contacted with a solution or slurry of
alkaline material in a vessel of relatively long resident time (5 to 10
seconds). The flue gas is adiabatically humidified to within 50°F of
its saturation temperature by the water evaporated from the solution or
slurry. As the slurry or solution is evaporated, liquid phase salts
are precipitated and remaining solids are dried to generally less than
one percent free moisture. These solids, along with fly ash, are
entrained in the flue gas and carried out of the dryer to a particulate
collection device. Reaction between the alkaline material and flue gas
S02 proceeds both during and following the drying process. The mechanisms
of the S02 removal reactions are not well understood, so it has not been
determined whether S0« removal occurs predominantly in the liquid phase,
by adsorption into the finely atomized droplets being dried, or by
reaction between gas phase SO- and the slightly moist spray dried solids.
Sodium carbonate solutions and lime slurries are common sorbents.
A sodium carbonate solution will generally achieve a higher level of
SO- removal than a lime slurry at similar conditions of inlet and outlet
flue gas temperatures, SO. level, sorbent stoichiometry, etc. Lime,
however, has become the sorbent of choice in many circumstances because
of the cost advantage it enjoys over sodium carbonate and because the
reaction products are not as water soluble. Through the use of
performance enhancing process modifications, such as sorbent recycle
and hot or warm gas bypass, lime sorbent has been demonstrated at the
pilot scale to achieve high levels of removal (85 percent and greater) at
sorbent utilization near 100 percent.
Using a spray dryer for a flue gas contactor involves adiabatically
humidifying the flue gas to within some approach to saturation. With set
conditions for inlet flue gas temperature and humidity and for a specified
approach to saturation temperature, the amount of water which can be
evaporated into this flue gas is set by heat balance considerations.
Liquid to gas ratios are generally in the range of 0.2 to 0.3 gal/MCF.
The sorbent stoichiometry is varied by raising or lowering the concentration
of a solution or weight percent solids of a slurry containing this set
amount of water. While holding other parameters such as temperature
constant, the obvious way to increase SO- removal is to increase sorbent
stoichiometry. However, as sorbent stoichiometry is Increased to raise
the level of S02 removal, two limiting factors are approached:
1) Sorbent utilization decreases, raising sorbent and disposal
costs on the basis of SO. removed.
2) An upper limit is reached on the solubility of the sorbent
in the solution, or on the weight percent of sorbent solids
in a slurry.
-------
There are at least two methods of circumventing these limitations.
One method is to initiate sorbent recycle, either from solids dropped out
in the spray dryer or from the particulate collection device catch. This
has the advantage of increasing the sorbent utilization, plus it can
increase the opportunity for utilization of any alkalinity in the fly ash.
The second method of avoiding the above limitations on SO. removal
is to operate the spray dryer at a lower outlet temperature; tnat is, a
closer approach to saturation. Operating the spray dryer outlet at a
closer approach to saturation has the effect of both increasing the
residence time of the liquid droplets and Increasing the residual moisture
level in the dried solids. As the approach to saturation is narrowed,
S0_ removal rates and sorbent utilization generally Increase dramatically.
Since the mechanisms for SCL removal do not appear to be well understood,
it is not obvious whether it is the increase in liquid phase (droplet)
residence time, or the increase in residual moisture in the solids, or
both which accounts for the increased removal.
Unfortunately, the approach to saturation at the spray dryer outlet
is set by either the requirement for a margin of safety to avoid condensa-
tion in downstream equipment or restrictions on stack temperatures. The
spray dryer outlet can be operated at temperatures lower than these
restrictions would otherwise allow if some warm or hot gas Is bypassed
around the spray dryer and used to reheat the dryer outlet. Warm gas
(downstream of the boiler air heater) can be used at no energy penalty,
but the amount of untreated gas involved in reheating begins to limit
overall S02 removal efficiencies. Significantly less hot gas (upstream
of the air heater) is required to heat, but an energy penalty associated
with the decrease in heat load to the air heater comes with bypassing
the plant air heater. Figure 2-1, a general flow diagram of a spray
dryer based system, illustrates these two "reheat" options.
The spray dryer design can be affected by the choice of particulate
collection device. Bag collectors have an inherent advantage in that
unreacted alkalinity in the collected waste on the bag surface can react
with remaining S02 in the flue gas. Some process developers have
reported SO, removal on bag surfaces on the order of 10 percent (Ref. 13).
A disadvantage of using a bag collector is that since the fabric is
somewhat sensitive to wetting, a margin above saturation temperature
(on the order of 25 to 35°F) must be maintained for bag protection. ESP
collectors have not been demonstrated to achieve significant SO- removal.
However, some vendors claim that the ESP is less sensitive to condensation
and hence can be operated closer to saturation (less than a 25°F approach)
with the associated increase in spray dryer performance.
The choice between sorbent types, use of recycle, use of warm or hot
gas bypass, and types of particulate collection device tends to be rather
site specific. Vendor and customer preferences, system performance require-
ments, and site-specific economic factors tend to dictate the system design
for each individual application.
10
-------
CLEAN GAS TO
ATMOSPHERE
HOT OR WARM GAS BYPASS ^
4
\
1 CLEAN GAS / \
... ». ._. . i ... . /**-^\ ""/ \
I JFLUEGAS / \
1 nnn rn ............
|— •"! PREHEATER
/U
AIR
i
COMBUSTION AIR
f 1 U STACK
fc. 1
* /K
SPRAY
DRYER r-
l J i »
Ysr —-
SPENT 1
SOLIDS
PARTIAL RECYCLE OF SOLIDS '
(LIME REAGENT) 1
)W
SORBENT PRODUCT SOLIDS *
SLURRY FLY ASH DISPOSAL
TANK
X
SORBENT STORAGE
70
Figure 2-1. Typical spray dryer/particulate collection flow diagram.
-------
2.1.2 Dry Injection Process
Dry injection schemes generally Involve pneumatically introducing a
dry, powdery alkaline material into a flue gas stream with subsequent
particulate collection. A generalized flow diagram of this process is
shown in Figure 2-2. The injection point has been varied from the
boiler furnace area all the way to the flue gas entrance to an ESP or
bag collector. Most dry injection schemes use a sodium-based sorbent.
Lime has been tested but has not been demonstrated with much success.
Many dry injection programs have used nahcolite as a sorbent. Nahcolite
is a naturally occurring mineral, associated with western oil shale
reserves, and is about 80 percent sodium bicarbonate. Sodium bicarbonate
appears to be more reactive than sodium carbonate, because it loses both
two moles of CO- and one of water in reaction, while sodium carbonate
loses only one mole of CO- in reaction with SO-. The following overall
reactions illustrate this point:
2NaHC03 + S02 Na2S03 + H20 (1-1)
Na2C03 + S02 Na2S03 + C02 (1-2)
Since bicarbonate loses three moles for every mole of SCL removal, bicar-
bonate particles tend to have larger pore volumes and are apparently less
susceptible to blinding on reaction than are sodium carbonate particles.
Unfortunately, the availability of raw nahcolite in commercial quantities in
the near future is questionable due to the substantial investment necessary
before commercial scale mining can begin. Since the potentially favorable
economics of dry injection are based to some extent on the use of inexpen-
sive sorbents, the use of commercially refined sodium bicarbonate is
prohibitively expensive. Recent research has been aimed at studying the
use of raw trona ore, which is currently mined in large quantities both in
the Green River, Wyoming area and the Owens Lake, California area. The
mineral trona contains one mole of sodium carbonate, one mole of sodium
bicarbonate and two waters of hydration (Na2CO,»NaHCO~* 2H20). Trona has the
potential for providing a good compromise between reactivity, cost, and
availability for use in dry injection schemes.
An unresolved problem with this technology Is disposal of the sodium-
based waste materials in an environmentally acceptable manner. As mentioned
previously, sodium waste materials are highly soluble and can result in
contamination of aqueous streams. Disposal of sodium compounds is an
area requiring further investigation.
Both baghouse and ESP collection devices have been tested with dry
injection processes. However, the effect of the reaction between unspent
sorbent on collecting bag surfaces and SO, remaining in the flue gas seems
to overwhelmingly favor the bag collector (Ref. 15). Since a major portion
of the SO. removal reaction appears to take place on the bag surface,
various methods of feeding have been tested:
12
-------
BAGHOUSE COMPARTMENT*
AIR PREHEATER
CONVEYOR
NAHCOLITE
STORAGE
NAHCOUTE
HOLDING
BIN
INJECTION
FAN
A.
A
BAGHOUSE COMPARTMENTS
70 ISM 1
Figure 2-2. Nahcolite dry injection flow diagram.
-------
1) Continuous. After the bag is cleaned, sorbent entrained with
the flue gas is added to the bag surface continuously from
injection points located upstream of the baghouse.
2) Batch. After the bag is cleaned, all sorbent is added to the
bag as a precoat before flue gas flow is resumed.
3) Semi-batch. This feeding method is a compromise between types
1 and 2. After bag cleaning some sorbent is added initially as
a precoat and the remainder is added continuously throughout
the bag cycle from an upstream injection point.
Sorbent stoichiotnetry, sorbent particle size, point and temperature
of injection, baghouse air-to-cloth ratio, and bag cleaning frequency, are
also varied in dry injection programs.
2.1.3 Combustion of Coal/Limestone Fuel Mixture
The current research on combustion of a coal/limestone fuel mixture
has taken two forms:
1) Combustion of a coal/limestone pellet in an industrial spreader-
stoker boiler.
2) Combustion of a pulverized coal/limestone mixture in a low-NO
burner system. x
Preliminary results of bench-scale test work on both processes have
indicated that up to 80% of the available sulfur in the fuel can be retained
by the limestone. The ratio of calcium to sulfur in the coal/limestone fuel
mixture is important in determining how much sulfur is retained.
A spreader-stoker boiler (20bhp) has been used in testing the
combustion and sulfur retention characteristics of the coal/limestone
pellet. A Ca:S mole ration of 7:1 has been used so far, but further work
with a 3:1 Ca:S pellet is planned. The emissions generated are dependent
upon burner design, coal properties and combustion operating parameters.
The inherent staged combustion of the stoker-fired boiler (accomplished by
supplying the total combustion air as primary air through the grates and
secondary air through over-fire jets above the bed) results in lower NO
emissions relative to conventional pulverized coal-fired boilers. x
The two-staged combustion concept was employed by Babcock & Wilcox
(B&W) to design an advanced low-NO burner system. EPA has funded test work
to develop a concept of firing a coal/limestone fuel mixture in B&W low-
NO burners to reduce SO- emissions. Tests conducted on a 12 x 10 BTU/hr
un^t by the Energy and Environmental Research Corporatio (EERC), with a
Utah low sulfur coal, have demonstrated 88 percent SO- removal with a 3:1
Ca:S mole ratio. This high S02 removal has been attributed to the lower flame
temperature found in the low-Nox burner which may help maintain limestone
reactivity. EERC has reported that S02 removal increased substantially when
the reagent was passed through the pulverizer with the coal.
14
-------
Further research on a larger scale for both systems is needed to
determine the effects of combustion of a coal/limestone fuel mixture on
boiler operation and maintenance. Collection of the increased ash loading
and investigation of the properties and disposal of the waste products
must also be studied.
2.2 COMPARISON OF DRY AND WET SCRUBBING FOR S02 REMOVAL
Comparisons between dry and wet scrubbing systems can be drawn in five
major areas: waste disposal, reagent requirements, operation and
maintenance,energy requirements, and economics. This comparison will
focus on general aspects of dry FGD systems as compared to conventional
lime/limestone wet scrubbing systems.
With regard to waste disposal, dry FGD systems have an inherent
advantage over wet lime/limestone systems in that they produce a dry,
solid waste product that can be handled by conventional fly ash handling
systems, eliminating requirements for a sludge handling system. However,
the waste solids from sodium-based dry FGD systems are quite water soluble
and can lead to leachability and waste stability problems. Waste solids
from lime spray drying systems and coal/limestone fuel systems should have
similar environmental impacts as waste from lime/limestone wet systems,
for which waste disposal technology is better defined.
In general, dry FGD systems require a higher stoichiometric ratio
of sorbent to entering S02 to achieve the desired removal efficiency than
do conventional limestone wet scrubbing systems. In addition, the reagents
employed in spray drying and dry injection systems (soda ash, lime,
commercial and naturally occurring sodium carbonates and bicarbonates, such
as nahcolite and trona) are significantly more expensive than limestone.
Consequently, limestone wet scrubbing systems will have an advantage with
regard to reagent utilization and sorbent-related operating costs.
It has been claimed that dry system will have lower maintenance
requirements than comparable wet systems. Dry systems require less
equipment than wet systems as the thickeners, centrifuges, vacuum filters
and mixers required to handle the wet sludge waste product from wet systems
are eliminated. In addition, slurry pumping requirements are much lower
for spray drying and are eliminated in dry injection and combustion of
coal/limestone fuel systems. This is important because wet systems have
reported high maintenance requirements associated with large slurry
circulation equipment. Finally, the scaling potential in limestone wet
systems requires extra effort to maintain proper scrubber operation and
possibly makes dry systems somewhat more flexible as far as their ability
to adjust process operations to respond to variations in inlet SOj concen-
trations and flue gas flow rates.
-------
With regard to energy requirements, dry FGD systems appear to have a
significant advantage over wet systems due to savings in reheat and pumping
requirements. Many wet FGD systems reheat the flue gas before it enters
any downstream equipment to prevent corrosion. This reheat requirement is
eliminated in most dry system configurations and results in considerable
energy savings. Spray dryer systems are usually designed with a 30° to 50°F
approach to the adiabatic saturation temperature of the flue gas at the
outlet of the spray dryer. Energy savings from reduced pumping requirements
result from the fact that wet scrubbers may require liquid to gas (L/C)
pumping rates of up to 100 gallons per 1000 acfm of gas whereas spray dryers
only require an L/G rate of 0.2 to 0.3.
One of the major driving forces for development of dry SO^ removal
systems is the opportunity for reduction in both capital and operating
costs. Although costs are quite site specific, the three types of dry FGD
technologies considered here offer several potential possibilities for
cost savings. This is due to the reduction in equipment and operation and
maintenance requirements relative to conventional wet lime/limestone systems,
especially in utility applications. Basin Electric evaluated the costs of
the two spray drying systems they have purchased (Antelope Valley and Laramie
River Stations) to be about 20 to 30% less costly over the 35-year life
of the plant than comparable wet systems (See Section 5.12). However, it
should also be noted that these economics are based on pilot scale data and
should be better determined after the operation of commercial systems has
begun. Reagent costs for sodium-based spray dryer systems will be consider-
ably higher than for lime-based systems although vendors claim the capital
costs, excluding waste disposal, will be lower for sodium-based systems.
The minimal equipment and operating requirements for dry injection systems
make the process economically attractive as far as capital costs are
concerned, but high sorbent requirements and uncertainties in sorbent
availability and cost are slowing further development of the technology on
a commercial scale. Capital costs for both the pellet and low-NO burner
coal/limestone fuel mixture systems should also be low since they will con-
sist mainly of the equipment needed to produce the mixtures. However,
since these systems have the potential for impacting the design and/or
operation of the boiler, more information on the overall operability of
these systems is needed before overall operating costs can be estimated.
In summary, dry systems do offer potential advantages over wet systems,
especially in the areas of energy savings and costs. However, crucial
issues such as waste disposal and demonstration of commercial-scale systems,
which may continue to limit the overall acceptability of this technology,
remain to be answered.
16
-------
SECTION 3
CONCLUSIONS AND RECOMMENDATIONS
The primary intent of this report is to describe the status of the
development and commercial application of dry FGD in the United'States,
in both utility and industrial applications. However, after reviewing
the current status of dry FGD, the following conclusions and recommendations
have been made concerning the role EPA could assume to advance the acceptance
and application of dry FGD technologies as a viable alternative for controll-
ing S0» emissions. Because the three types of dry FGD technologies
considered here are in quite different stages of development, the conclusions
and recommendations are necessarily technology specific. Section 3.1
dicusses spray drying/particulate collection, Section 3.2 discusses dry
injection/particulate collection, and Section 3.3 discusses combustion of
pulverized coal/limestone fuel mixtures in low-NO staged combustion systems
and pelletized coal/limestone mixtures in stoker-rired boilers-
3.1 SPRAY DRYING/PARTICULATE COLLECTION
With regard to process development, there does not appear to be a need
for EPA to fund programs aimed at the development of spray-dryer based
technology since a significant amount of commercial interest currently
exists in these systems. Three Western utilities and two industrial FGD
spray dryer systems are under construction and reportedly as many as six more
utility applications will consider dry systems by the end of 1979. Conse-
quently, numerous vendors of dry processes are devoting large research
budgets to the development of spray dryer-based FGD systems, and it appears
that this technology will be developed regardless of EPA involvement.
However, the potential does exist for EPA to assist in helping prospective
FGD users to better evaluate this technology by funding development programs
to evaluate dry FGD overall environmental acceptability and to answer several
unresolved technical questions. Recommendations on the direction of these
development programs in four specific areas are discussed below-
1) Answer Unresolved Technical Questions
Waste Disposal— One major unresolved technical issue is that of dry
waste disposal, especially for sodium-based systems. The production of
a dry waste rather than a sludge for disposal is seen as a major advan-
tage of the dry systems over wet systems. However, no major development
program has been aimed at optimizing disposal techniques for either
sodium-based or calcium-based scrubber waste/fly ash mixtures. Although
waste disposal is not seen as a technology limiting area, many unknowns
such as water consumption rates, waste stability, load bearing qualities,
and leachability have not been quantified.
17
-------
Reaction Mechanism—Another key technical issue is that of defining
the reaction mechanisms for SC^ removal in the spray dryer. Until the
reaction mechanisms are well understood, methods of improving spray
dryer performance and expanding applicability can only be determined
by an empirical approach. If the reaction mechanisms are well
understood, the methods for improving spray dryer performance and
the limits of spray dryer applicability will be better defined.
High Sulfur Coal Applications—As discussed previously in this
section, most current spray dryer development work applies to low
sulfur coals. An EPA-funded program to test this technology on
higher sulfur coals could promote application of spray dryer technology
for Eastern installations. Of course, such a program would be supported
by a better understanding of the mechanisms of the SCL removal reactions
in the spray dryer, as discussed above.
Limestone reagent investigations—The application of spray dryer
technology to higher sulfur coals may be subject more to economic
rather than technological limitations. Wet FGD systems can use
limestone as a reagent, while spray dryer based systems are
currently limited to the use of a more expensive lime reagent.
For high sulfur applications, where the reagent quantities involved
tend to be large, the reagent cost differential between lime and
limestone alone may make a spray dryer-based system uneconomical.
Emphasis should be placed on the development of an effective spray
dryer process using limestone as a reagent which would greatly improve
these economics.
2) Demonstration of Technology
Of the three utility spray dryer-based FGD systems sold
to date, the earliest commercial operation is slated for start up
in mid-1981. The first industrial spray dryer based FGD system is
scheduled for startup in late 1979. While these industrial applications
should begin to demonstrate the operability and reliability of spray
dryer-based systems almost immediately, these results may only be avail-
able to the process vendors and industrial clients involved. EPA fund-
ing of a commercial industrial or demo-scale utility spray dryer-based
dry FGD system prior to the 1981 startup of the first utility system
could provide data for evaluation of this concept by the utility
industry as a whole.
3.2 DRY IMJECTION/PARTICULATE COLLECTION
There is a significant data base for dry injection technology that
indicates its technical feasibility as an S02 control alternative. Major
restraints to development of commercial systems have been uncertainty in
sorbent (nahcolite) availability, and uncertainty regarding acceptable
disposal practices for the sodium-based waste material. EPA could help to
18
-------
these questions by funding programs to quantitatively answer both of these
questions.
3.3 COMBUSTION OF LIMESTONE/COAL FUEL MIXTURES
To date only preliminary data exist for the S02 control effectiveness
and operation of boilers firing either coal/limestone pellets or a pulverized
coal/limestone fuel mixture. It is recommended that EPA continue to fund
development programs in both of these areas with emphasis on validating the
technology, determining the effects on boiler operation and maintenance, and
ascertaining costs as a function of S02 removal.
19
-------
SECTION 4
DRY FGD RESEARCH REVIEW
This section provides a summary of past research conducted on dry FGD
methods. The research activities have been divided into three areas:
1) dry injection/particulate collection systems (4.1), 2) spray drying/
particulate collection systems (4.2),and 3) other research including
combustion of a coal/limestone fuel mixture and fixed and fluidized bed
reactors (4.3). Table 4-1 presents a summary and brief description of
the research activities reviewed in this section.
The research activities in each area are listed in chronological order
as reported. The reviews cover the following areas:
• scope of the project,
• process description,
•parameters investigated,
•test conditions,
•conclusions and comments, and
• references.
For those activities where only limited information is available, the review
covers only those entries in the above list which are applicable.
The results of previous test work have generally been published and
are available in great detail. Results from current research activities
are often not readily available because test work may not be completed,
conclusions may not have been finalized, etc. Since it was the intent of
this document to report as much information regarding dry FGD research as
was available, the results of older, completed studies presented here are
generally more complete and detailed than those of current or recent studies.
It was not intended, however, to bias the content of the report toward
earlier studies. Unfortunately, although the more recent studies are
discussed in less detail, these are probably most indicative of the state
of the art of dry FGD. Hopefully in future quarterly updates, more de-
tails of the current and recent research efforts will be available for
presentation.
20
-------
TABLE 4-1. LISTING OF RESEARCH ACTIVITIES CONDUCTED PRIOR TO APRIL 1979
rowutou
LoctTin/nu
wunn TOTD
Cftl TD1TKU11U
nairiourr umn
•n l«J«cttoWC«»rlc
(llKwt k^c%-««.l. lab
•Iliw: »U»c-«e«l«
(lit—. »!>«-«•
III*
.11*. U.tr.1 mtmlml-
•tr.tlM IMM ml
Iteicn
IMM
»4 C
If II
IC.IOT.M niiitr
UMtrlc «.OTC.)I
J»lf lt»«
AAT UMl H'«
JUI.«M KAlli
••rvlc*
••ht.lltt (kl|h tMr^r.ttf.
l.«tt^ ter 1IM 994 m*i»mtt*
b» Hc.rt»««l«. M
i.t.l Ittkullu
«.».» IWmllK «M !>•«•
9W u let* r
Ml
f?«*r
M< «M*r
Il.tX ll«. .M «.lMltr
M. «l«f wkc.Itt. M. .IM, Ml
• lit >1 1. »1 W4tll>. .IIIII.IIM
«t joo-r.
Ml
Ml «0, r*M«l vlik Ma , lit
xiti«»f«> t.i i>vr>. '
»l U >n~.) .Ill, MM. 111.;
UI >tll!uil«i I.I IWf).
IM M.
>OI «•,
•lllllvttM.
(O.tl > c«d)
S«M ItM. With
tn f*> n.ll l>
»i*r
K«* » IM * r .MklanrK 'ill* III) .< i.) •.!..
M.CO^/^I. I fn .1 1VOT. 0.1.i.r.
CMIttl .r .1 I M.I 11 rrwl»4 IM
.ffmm M, tra*p.l. H t. tn i.-
—..I ..IM InM. (II > l.<) .
»f t.< (Ml
^ I. DM rr»)
ni i. Mrr
CMOTflM
*rr l»J««tiWt
— ' ••» •»!•«•—'
r..»rtk IM1.
(Will
>~.«f. — «.Hl«. If—..
•Ilk i
!• fO. n..i.l
•tr u cl.tk rail.* •! J.
IM M >M mil k. 4IM.
rr*/M> i* Mr i. i.i. I»M.
)ie*r
..
>"•«"'•'
flll*T .• Blrtcr-llr*
Wlln; »ll«-«.l.
t.l.l t^la kl«rk«.l..
«*ltt«.
(contlnuecf)
-------
TABLE 4-1. (continued)
mwlcm
MTMIMf*"
«•»!«• IMicmtMl
(•UI4M* *f btkv.ll, IK.
ten, UlMt«MMl*kf •<•*.•»•
H.tto.
baU.
Itll
c..): itti-tm
4.1.1
«• '*" »• •— '
t. 1il,l,- M
J4f»
!>•• *• «io-r"
lMI. M
•JI «l l«let SO. M. I
4TF«r olch .• *MltUMl !• 1. 101 MT.
Ck. k»«l»_».i
•Itk U» U4ltl« .tlllMIIM, iltk Ito.
; »llM-MU.
N>
N>
M«« L»fc« bat
I«I«T Tall t~«
e>.|. im-itn
tlmru): Itlt
4.I.S
»**r
4kMt •» fftx rwv.l .cliln^ wlik M.V
IkMC.tlc.l lllllH •( ll.I v*«« 14CIU1
r«y«l. «f Mlt4» M»l*fM. *0t f^«4l
•Ckt^^ vllk M«C tkMt.tlC.l Ml lit. •(
IIM VICkMl f««rCl*.
IIM. ..... u tUItM Mtk lb»
rt.>i..Mj/»rt»
; 1*M
nc latos 1*M-
»r>r« uk»i mi
4.1.*
4.J.I
4.1.1
>«• •• »**• «••
CU .U)/MO. M(
Uxtlci;
••V.I.
»t~ •.l. vtlk f»it
u *l 1 Xll u.1 .t 4.4 w *.n 1.
-------
4.1 DRY INJECTION/PARTICULATE COLLECTION
4.1.1 Owens-Corning Fiberglass, Laboratory Tests, August 1970
Scope
In an effort to develop fabric filters that would withstand high
temperatures, Owens-Corning Fiberglass (OCF) developed an S-glass cloth
treated with a proprietary inorganic finish. Under contract with the
National Air Pollution Control Administration, now EPA, a stainless steel
test facility to assess the filtration and contacting characteristics of the
new fabric was constructed and operated during 1970. The emphasis was on
investigating various sorbents for SOo removal from flue gas at high
temperatures.
Process Description
The test facility was a single compartment stainless steel baghouse
with 6 bags. Two methods of sorbent feeding were tested: 1) continuous
injection of the dry sorbent into the simulated flue gas, upstream of the
baghouse and 2) pre-coating of the bag by injecting the sorbent into S0?-
free gas to build up a dust cake on the bag before the regular gas stream
(containing SO,,) was started.
The synthetic flue gas was supplied by a natural gas boiler-heat
exchanger and spiked with SO-. Hot and warm flue gas streams were
controlled to give a gas temperature between 250 and 1000°F at varying
flow rates and constant composition.
Bag cleaning was accomplished by conventional reverse-air shaking
methods. Inlet and outlet S(>2 concentrations, pressure drop across tin-
bags, and baghouse temperatures were continuously monitored.
Parameters Investigated
The tests were divided into four series: an additive study, a flue gas
flow rate study, a precoated bag study, and a fly ash study. Parameters
investigated within these series included:
• type of sorbent,
• gas stream temperature,
* flue gas flow rate,
• stoichiometric ratio,and
• additive feeding method.
Test Conditions
Inlet S02 concentration was maintained at 2800 ppm. The test cyle was
usually 60 minutes. The average flue gas composition was 3.5 percent H20,
81.0 percent NZ, 5.6 percent 0,, 1.0 percent Ar, and 9.0 percent C0_. See
Table 4-2 for a summary of the variable test conditions.
23
-------
TABLE 4-2 VARIABLE TEST CONDITIONS - OCF LABORATORY TESTING21
Series
Additive
Study
Flue Gas
Flow Hate
Study
Fly Ash
Study
Feed Method Sorbent
1 In alaked llaie
all caaea
promoted anr as above
all cases
1 in slaked lime
all cases
promoted slakrd 1 inn-
Klakrd dolnmllP
promoted slaVi'd dolnraltr
""'"
2 In iilaktxl li».-
all c««t«
promoted slnkrd 1 liw<4
slaked dolumtn*
pr»mot ed Nlaki'd dtilontllr
MiiO,
StoichloMtrir
Ratio
1.0-3.0
1.0-1.0
1.0-1.0
1.0-1.0
1.0-3.0
i.n
1.0
Kame /i.s
NK
NK
NK
NH
1.0 1.0
lor j>) 1
I'llhrli
Ta»p. Kaiige flur Caa riox
7OO-900'F 190 <(• In all
tuaea
700-900-F
• Ir-ololh rn in -
<> Oil/Bin
700-900' F
>OO-VOO'F
SOO-700'K
300- 500' F
300- 500' F
««•«• *n W> cf« (A/C * 40 fi/aHn)
«hnve (i>mrpt iinlicol 1 tc vlilrli
«^» 68-. r(ni)
BOO'F JH1. ifn |,,r nl 1 I«M.J.
BOH" K
BIM,"F
•u.r»
mxi"F
/Od-Wd'F Mil unit IHS ,l» lot
fill It 1 IIMC
/(Id "Mid" F
/(III 'M!U"K
100-VHI'F
M - continuous (wd to flue gas flow; 2 • prr-coat n(
NK * not reported.
-------
Results
The percentage S0~ removal was calculated using conventional break-
through curves. Some of the results for runs with continuous injection of
sorbent are summarized in Table 4-3.
Slaked lime and dolomite achieved maximum SO removals at 800°F.
However, additive utilization was low, with a value of only 40 percent at
81 percent S(>2 removal.
Promoted slaked limes did not perform significantly better than slaked
limes. Both MnO and alkalized alumina gave good SO, removals (70 percent for
MnO~ and 70 to 90 percent for alkalized alumina depending on test conditions).
Again, additive utilization was low and a regenerative process would
definitely be required for these expensive sorbents.
Increasing turbulence by increasing the flue gas flow rate improved
SO removal and additive utilization when lime sorbents were used. This
indicates that lime sorption of SO- may be gas-phase mass transfer limited
at the high temperatures used in these tests.
In the tests using nahcolite sorbent, an increase in flow rate did not
enhance S02 removal, but it did improve additive utilization. Overall,
nahcolite utilization was greater than lime utilization 75 percent versus
40 percent at 290 cfm.
Precoating of the bags was not an effective method of removing SO. when
lime sorbents were used. This was attributed to the absorption of the (^
present in the SO -free flue gas used in coating the bags, thus reducing the
amount of lime available to react with the SO,.
Fly ash addition did not affect SO. removals.
Conclusions and Comments
Slaked lime and dolomite achieved satisfactory SO removals at 800°F
but with low reagent utilization. These high temperatures are not likely
to be encountered in commercial dry injection FGD systems, as the hot flue
gas passes through an air preheater upstream of the injection point and/or
fabric filter.
Removal of SOj using nahcolite was about 60 percent with 75 to 85 percent
additive utilization at 300°F (the inlet SO, concentration was 2800 ppm). Also,
the dry product formed is quite water soluble and waste disposal problems may
occur.
Information Sources
Veazie, F. M., and W. H. Kielmeyer, "Feasibility of Fabric Filter as Gas-
Solid Contactor to Control Gaseous Pollutants". Report No. APTD-0595
for the National Air Pollution Control Administration, HEW Contract No.
Ph-22-68-64, Owens-Corning Fiberglass Corporation, Granville, OH.
August 1970.
25
-------
TABLE 4-3. PERCENTAGE S02 REMOVAL WITH CONTINUOUS SORBENT
INJECTION - OCF LABORATORY TESTS21
Reactant: Slaked Lime
Flow Rate:
Stoichiometric Ratio:
Bag Temperature
700°F
800° F
900°F
290 cfro
385 cfm
1.0 2.0 3.0 2.0 3.0
46% 64% 57% 74%
40% 65% 78% 81% 91%
52% 70% 65% 81%
Sorbent
Utilization* :
maximum 40%
minimum 26%
Reactant: Slaked Dolomite
Flow Rate:
Stoichiometric Ratio: 1.0
Bag Temperature
700°F
800°F 30%
900°F
Reactant: Nahcolite
Flow Rate:
Stoichiometric Ratio:
(Temp - 300°F)
290 cfm
2.0 3.0
40% 54%
52% 64%
34% 49%
290 cfm
0.8
59%
385 cfm
2.0 3.0
Sorbent
40% 52% Utilization8: 25%
55% 70%
44% 60%
385 cfm
0-7 0 u
Sorbent
60% Utilization8 :
75% @ 290 cfm
80% @ 685 cfm
Sorbent Utilization - percent S02 removal/Stoichiometric Ratio
26
-------
4.1.2 Air Preheater at Mercer Station, Match 1971
The Air Preheater Corporation was contracted by the National Air Pollu-
tion Control Association (now EPA) to conduct tests to investigate tilt-
potential of the top inlet fabric filterhouse as a chemical contactcir for
SOj removal from flue gas. The study was carried out at a pilot test
facility at the Mercer Station of Public Service Electric and Gas Company
in Trenton, New Jersey during 19/1.
Process Description
The filterhouse (baghouse) test facility of top entry design had four
compartments, each containing nine fiberglass filter bags. The total filter
area was 4300 ft2. The bags were cleaned by the reverse-air deflation method.
In addition, two different types of baghouse operation, cyclic and parallel,
and both continuous and batch sorbent feeding methods were tested. Sorbent
could be added upstream of the baghouse, just before the baghouse or precoated
onto the fiter bags.
Flue gas was available at temperatures from 270°F (leaving the air
preheater) to 680°F (just ahead of the air preheater) and was mixed to give
a variety of operating temperatures. Inlet and outlet S02 concentrations,
as well as temperatures and pressure drops, were continuously monitored.
Parameters Investi gated
The study was performed In two phases. The first phase evaluated tin-
feasibility of the process usinn sodium bicarbonate, NaHCO-j. The second
phase investigated the process variables at higher temperatures using dif-
ferent sorbents (nahcolite, hydrated dolomite, promoted hydrated dolomite,
and hydrated lime). The process variables investigated were:
• stoichiometric ratio ,
• operating temperaturej
•filter air-cloth ratio (flue gas flow),
•method of sorbent feeding, and
•mode of baghouse operation.
Test Conditions
Flue-gas flow rates of 7,500, 10,000 or 15,000 acfm were used to give
corresponding air-to-cloth filter ratios of 1.75, 2.3, or 3.5, respectively.
The flue gas temperatures ranged from 270 to 650°F. No S02 inlet concent ra-
tions were reported. The moisture level was given as 5 percent. The tests
were usually conducted in 40-minute cycles.
Results
About half of the total 62 tests were run with sodium bicarbonate at
various operating conditions. Increasing the stoichiometric ratio increased
27
-------
the S0? removal with accompanying decreased sorbent utilization. An increase
in operating temperature also increased S02 removal. Table 4-4 provides
a summary of the tests results with NaHCO^, nahcolite, and hydrated dolomite,
although the results are of limited value without inlet S02 concentrations.
TABLE 4-4. SUMMARY OF AVERAGE 50^ REMOVAL AND SORBENT UTILIZATION
AIR PREHEATER CORP. AT MERCER STATION 1A
Stoichlo»«'trK Kntlo • 1 Stolchloml rlc Mtlii • 1
Additive
Sodium Blcirboiwte
Sodlun Bicarbonate
Sodlua Bicarbonjilr
Nihcolit*
Nalirolttc
Dolonlte
Do 1 omit*
Trap. (*F>
270
3»0
600
350
600
350
600
Coflv.
Efficiency
3291
««X
90S
6>X
94 X
,
201
Additive
Utllliaclen
32X
*BX
»OX
SIX
9«X
_
20X
Conv.
Efficiency
*8X
761
-
sn
-
20X
38X
Addlllve
UlllUetlon
I6X
m
-
tot
-
71
111
Nahcolite tests results showed trends similar to those in sodium bi-
carbonate tests. Nahcolite gave a somewhat higher S02 removal than sodium
bicarbonate at the same temperature and stoichiotne-ric ratio. The investi-
gators claim this may have been due to the smaller particle size of the nah-
colite. Sulfur dioxide removal efficiencies were low in tests with lime at
350°F.
As to the variation with mode of baghouse operation, parallel operation
resulted in higher removals for continuous additive feeding. This seems to
be a result of the lower air-to-cloth ratio of parallel operation, where all
the compartments are on-stream during the cycle, as compared to the higher
ratio resulting when one of the compartments is always off-stream for cleaning
during cyclic operation. This is in agreement with results of tests with
increased air-to-cloth ratios that show a decrease in S02 removal and additive
utilization for both NaHC03 and nahcolite at temperatures from 280° to 350°F.
When lime was used, the removal efficiency and additive utilization increased
with higher flue gas flow rates.
Data from the chemical analyses of fallout and shakedown samples were
used to approximate the conversion distribution in the filterhouse.
Although only approximations could be made, it appeared that most of the
conversion, up to 80%, occurred in the filter cake for tests with sodium
bicarbonate and nahcolite at 350°F. However, for lime the bulk of the
conversion was in the gas suspension. These calculations wcrr based on the
a«sumpHon that no chemical reaction took place in the fallout collected from
the hopper.
28
-------
Conclusions and Comments
The use of a fabric filter as a chemical contactor to remove SO through
reaction with a dry sorbent was proven feasible. The additive utilization
was fairly low at flue gas temperatures below 350°F. Higher temperatures
(650°F) are required for lime to be an effective sorbent.
Stoichiometric ratio and operating temperature appear to have the
strongest effect on SO removal. The inlet SO concentration was not
specified for any of tne tests. The moisture Content (5 percent) of the flue
gas was somewhat lower than most of the other similar studies on the dry in-
jection/baghouse process.
In tests with nahcolite and sodium bicarbonate at flue R;IS temperatures
of about 300°F, the effect of increasing flue gas flow was small. This
suggests that the reaction is not gas-phase mass transfer limited at these
temperatures. However, in tests with lime at 650°F, the increase Jn flue
gas flow rate substantially increased the S02 removal efficiency, Indicating
that this reaction may be gas-phase mass transfer limited at higher tempera-
tures.
Information Sources
Liu, Han and R. Chafee, "Evaluation of Fabric Filter as a Chemical
Contactor for Control of S02 in Flue Gas," presented at the Air
Pollution Control Office Fabric Filter Symposium, Charleston, S.C.
March 1971.
4.1.3 Wheelabrator-Frye at Nucla Station, July 1974
Scope
In July 1974 Wheelabrator-Frye conducted tests at the Nucla
Station of the Colorado Ute Electric Association. The tests were performed
to demonstrate the applicability of dry nahcolite injection into a
commercially available baghouse to remove SO? from the flue gas.
Process Description
The tests were conducted on a unit comprised of an 11 MW spreader-
stoker-fired boiler burning low sulfur coal and a Wheelabrator-Frye
continuous automatic fabric filter collector. Figure 4-1 shows the
schematic process flow diagram. Fifteen of sixteen tests were conducted
in a batch feeding mode. The most frequently employed procedure was to
inject all the nahcolite at the beginning of the test cycle by blowing
it into the bags through a common inlet manifold using the injection fan.
In some runs the nahcolite was injected into a single compartment, in
which case the injection fan was used to overcome the added pressure drop
that resulted across that compartment.
29
-------
BAGHOUSE COMPARTMENTS
AIRPREHEATER
CONVEYOR
NACHOLITE
STORAGE
IN DRUMS
U)
o
2 ,
NAHCOLITE
HOLDING
BINS
INJECTION FAN
u
u
BAGHOUSE COMPARTMENTS
FUNCTION
VALVE NO.
VALVE NO.
TYPE
• NO NACHOLITE INJECTION
• NACHOLITE FEED TO COMMON INLET
• NACHOLITE FEED TO INDIVIDUAL
COMPARTMENTS
• PURGE SYSTEM OF NACHOLITE
•VALVE OPERATION: O = OPEN
X- CLOSE
1 2 3 4 5* 6
X X X X X X
O O X O X X
o o o
O X O
O X
X O
BUTTERFLY MANUAL OPERATED
ROTARY-ELECTRIC MOTOR OPERATED
BUTTERFLY-MANUAL
BUTTERFLY MANUAL
BUTTERFLY-SOLENOID
BUTTERFLY-SOLENOID
•ONLY ONE NO.5 VALVE CAN OPEN AT A TIME.
Figure 4-1. Process Schematic - gas and nahcolite into baghouse.
Wheelabrator-Frye at Nucla Station.16
roi SMI
-------
Information Sources
Bechtel Corp., Evaluation of Dry Alkalis for Removing SO9 from Boiler
Flue Gases. EPRI Final Report FP-207, October 1976.
4.1.4 American Air Filter, Laboratory Tests, 1976
Scope
In 1972 American Air Filter, with substantial funding from Arizona
Public Service, conducted bench-scale studies to investigate the use of
fabric filters as chemical contactors for S02 removal from flue gas.
Process Description
Reactant and fly ash were introduced into synthetic flue gas flowing
in a vertical tube 8.3" in diameter and 29.5" long. A filter made of
commercial fiberglass bag fabric was clamped and sealed across the bottom
of the tube. After investigating both batch and continuous modes of
reactant/fly ash feeding and observing no difference in performance, most
of the remaining tests were conducted using the easier-to-control
continuous mode.
Parameters Investigated
Tests were conducted to evaluate the following parameters over the
given ranges:
•sorbent type,
•flue gas veloticy (2.0 to 4.4 ft/min),
•sorbent stoichiometric ratio (0.75 to 4.5),
•temperature (200 to 250°F),
•moisture content of flue gas (1.1 to 7.7%), and
•fly ash concentration (0.18 to 15.6 gr/ft3).
The stoichiometric ratio was defined as moles of Na^COo feed per mole of S
in the inlet flue gas stream.
Test Conditions
The primary sorbents investigated were nahcolite and trona (Na2C03«
NaHC03•2H20). The S(>2 concentration ranged from 450 to 600 ppm. The flue
gas averaged 1.5 percent 0 , 14 to 15 percent C02, and 76 to 82 percent N-,
depending on the moisture content.
Results
Removal efficiencies of up to 90 percent were achieved when the moisture
content was greater than 5 percent. Removal decreased significantly below
this moisture level, but little effect was observed upon Increasing the
32
-------
moisture content above 5 percent. Increasing the stoichiometric ratio
increased SO- removal. Higher operating temperatures resulted in better
additive utilization as well as greater SO- removal. Fly ash content had
no effect on performance. Table 4-5 provides a summary of results
illustrating these trends. It is not clear on what basi.s additive
utilizations were calculated.
TABLE 4-5. SUMMARY OF RESULTS OF NAHCOLITE AND TRONA TESTS -
AMERICAN AIR FILTER
16
Inlet SO: Cone, Temperature Stoichiometric Water Vapor Air to Cloth SO, Mem0
250
et at
2.3 5.5
1.15 7.4
2.2 2.2
7.? 6.3
2.1 6.6
1.9 5./
l.V o.fc
1.2 7.8
have been recalculated from a wclg
3.7
4.0
4.15
3.1'.
1.H
3.9
4.1
4.1
ht baala to a
92
65
62
K;
70
H7
71
51
volume basli*.
40
56
2H
4(1
14
4b
)R
42
*Stolchlomitrlc ratio li defined ai mole* N«;COi fed pir mole S fed.
After 60 minutes of operation.
In a test conducted without a filter essentially no S02 was removed.
This indicated that reaction occurred primarily on the filter cake and not in
the gas-solid suspension.
Conclusions and Comments
Both nahcolite and trona were shown to be effective SO removal agents.
Trona was found to be less reactive than nahcolite, but theZauthora nuggesU'd
that reactivity could be improved by such means as partit-lu size optimization.
A moisture content of at least 5 percent was required for effective SO
removal. Increases in moisture content above 5 percent did not significantly
increase removal.
The observation that no reaction occurred in the gas-solid suspension may
be explained in part by the laminar nature of flow in the test apparatus.
Because the relative gas/particle velocity is small, a stagnant film of gas
33
-------
surrounds the particle and mass transfer can only take place by molecular
diffusion (no convective diffusion by turbulence). This requires high bulk
SO- concentrations to overcome film resistance.
Information Sources
Rivers, R.D., et al, "The Role of Fabric Collectors in Removing SO "
presented at the 1st National Fabric Alternatives Forum, Denver, CO,
July 1976.
A.1.5 Wheelabrator-Frye at Leland Olds, March 1977
Scope
Wheelabrator-Frye conducted pilot testing of a dry injection process
using nahcolite for SO. removal with a baghouse collection device. The
tests were carried out at Basin Electric*s Leland Olds Station in Stanton,
ND during the period of January through March 1977. Bechtel Power Corporation
acting for the Otter Tail Power Company, coordinated the overall effort.
The tests were performed to demonstrate the viability of the dry injection
process for use at the full-scale Coyote Station being planned by Otter Tail
Power Company. Process conditions at Leland Olds were similar to those
expected at the new facility.
Process Description
Leland Olds Unit 2 is a 440 MW facility with a North Dakota lignite-
fired boiler. The Wheelabrator-Frye process, based on test work at the
Nucla Station in 1974, involved injecting dry nahcolite and collecting the
solids with a baghouse filter. The S02 is removed through chemical
reaction with the nahcolite. Several possible feeding procedures were
considered: 1) feeding all of the nahcolite at the start of the cyple in the
batch mode, 2) feeding nahcolite continuously throughout the cycle, and
3) combinations and variations of the two methods. Provisions were made
for SO spiking of the flue gas and for high temperature testing by injecting
the nancolite into hot flue gas upstream of the air preheater. The use of
a booster fan to blow the gas that, depending on the feeding method, may or
may not have contained SO-* helped to disperse the nahcolite if it was present.
The Wheelabrator-Frye twelve-compartment baghouse contained fiberglass
bags with a silicone-graphite finish that comprised a total filter area of
1080 ft . The cleaning of the filters was accomplished with conventional
deflation-shaker methods.
Temperatures and S02 inlet and outlet concentrations were continuously
monitored.
-------
Parameters Investigated
The primary parameters investigated were nahcolite feeding method,
stoichiometric ratio, and S0_ inlet concentration. The effect of pre-
treating the nahcolite by heating to decompose NaHC03 to more porous Na^CO.,
was also examined.
Test Conditions
The air-to-cloth ratio in the baghouse was 3:1 with the normal flue gas
flow of 3100 acfm. The flue gas composition ranges were given as 10.5 to
15 percent CO , 77 to 83 percent N , and 5.5 to 8 percent 0 on a dry basis
with a moisture content of 13 to 16 percent. The stoichiometric ratio varied
between 0.8 and 1.7. The SO inlet concentration was usually between 850
and 1000 ppm with extremes of 830 and 2700 ppm SO . Henhouse temperatures
ranged from 286 to 301°F.
Results
It was found that 90 percent S(>2 removal could be achieved at a stoichio-
metric ratio of 1.6 under optimum conditions. Variations in S02 inlet concen-
tration appeared to have only a small effect on nahcolite performance with
removal at higher concentrations being slightly better. Collection of filter-
able particulates was 99+ percent in all cases, and baghouse performance
remained satisfactory even when large quantities of nahcolite were injected.
It was also reported that feeding procedures had a significant effect on
additive utilization. However, these effects were not detailed. No results
of high temperature testing were given.
Conclusions and Comments
The pilot plant test work at Leland Olds by Wheelabrator-Frye appeared
to show that substantial S02 removal by nahcolite injection is possible
under optimum conditions. These "optimum" conditions* were not spec i I 'led in
the available literature. However, additive utilization at 90 percent So.,
removal is only about 56 percent, even under optimum conditions. The resulting
large nahcolite requirements pose a serious problem due to great uncertainties
in nahcolite availability (see Bechtel report listed below).
Although it was noted that nahcolite feeding procedures had a signifi-
cant effect on additive utilization, these effects were not discussed nor
was an optimum method of feeding given.
Information Sources
Bechtel Corp., Evaluation of Dry Alkalis for Removing SO^ from Boiler
Flue Gases. EPRI Final Report FP-207, October 1976.
Wheelabrator-Frye, Nahcolite Pilot Baghouse Study - Leland Olds
Station, Non-Confidential Test Data, unpublished, March 1977.
35
-------
4.1.6 Grand Forks Energy Development Center, DOE; Bench-scale Dry
Injection/ESP or Baghouse Collection, 1975 to Present
DDE-Grand Forks dry FGD work began in 1975 during bench scale combustor/
ESP testing of North Dakota lignite when sulfur retention was observed in
fly ash alkalinity. The current dry PGD program was begun around the first
of 1978, with the testing of nahcolite injection and ESP collection. The
scope of the project was broadened to include Green River area trona as a
sorbent. Later, when it was apparent that high S0? removal efficiencies
were not demonstrable with this configuration, the ESP collector was
replaced by a bag collector and testing was resumed.
Process Description
The Grand Forks bench scale coal combustor produces a nominal 200
scfm flue gas flow rate. Ductwork downstream of the combustor is
equipped with water cooling so that the time/temperature profile of
the flue gas can be varied. Injection points are also variable.
Parameters Investigated
The test program with a bag collector has so far included over 30 tests,
where injection and bag temperatures,sequencing of sorbent addition and bag
cleaning, bag materials, sorbent (nahcolite and trona), and bag air-to-cloth
ratios were varied.
Re_sul_ts
The dry injection/ESP system was not effective for SC^ removal. The dry
injection/bag collector was much more suitable. S02 removal efficiencies on
the order of 90 percent at 50 to 60 percent utilization were demonstrated
at an air-to-cloth ratio of 3 ft/min. At higher air-to-cloth ratios (up to
6 ft/min), lower sorbent utilization resulted, as low as 10 to 30 percent.
Current plans are to operate the dry Injection/baghouse collection system
another 6 months at various selected "optimum" conditions. Detailed results
will be available when the final report for the study is published after this
additional test period. Plans call for a final report around the end of 1979.
Grand Forks has also identified upcoming dry FGD related work. After
the startup of the Coyote plant, in spring 1981, Grand Forks will sample
the FGD system for particulate and SO removal efficiencies.
Another upcoming program has actually begun, but very little progress
has been made. This program involves column work and beaker tests with
samples of spent sorbent and fly ash products from the Leland Olds dry FGD
36
-------
pilot operations. Initial studies involve leaching and solubility studieH
and toxicant extraction of the untreated wastes. Future studies may include
similar testing of chemically fixed wastes.
Information Sources
Blythe, Gary. Telephone conversation with Stanley J. Selle, DOE, June
7, 1979.
Blythe, Gary. Telephone conversation with Harvey Ness, DOE, July 3,
1979.
4.1.7 KVB Incorporated, Bench Scale Tests. Late 1977 to Present
This bench scale study was funded by the Electric Power Research Insti-
tute, Inc. (EPRI) for the purpose of obtaining basic process data for a dry
injection/baghouse collection FGD system. The test work was performed
on KVB's bench-scale coal-fired combustor In Tustin, California. Six sodium-
based sorbents were tested over a wide range of operating conditions.
Process Description
The source of flue gas for the test work was the coal-fired KVB bench
scale combustor which produced a flue gas flow of approximately 725 scfm.
A baghouse was Installed downstream of the combustor for collecting fly ash
and spent sorbent. A heat exchanger was installed in the duct between the
combustor and baghouse in order to control the flue gas temperature at the
baghouse inlet. The duct was designed for sorbent injection at several
points between the boiler and baghouse, so that sorbent injection temperature
and residence time in the duct could be varied.
Parameters Investigated
The sodium-based sorbents investigated in this extensive test program
were:
1) commercial bicarbonate,
2) nahcolite,
3) Green River (Wyoming) trona,
4) Owens Lake (California) trona,
5) soda ash solution (sprayed into duct for evaporation), and
6) predecomposed nahcolite (heated to release C0£ and water) .
The following data summarizes the parameters and ranges investigated
in this test program:
Stoichiometric ratio 0 to 4:1
Injection temperature 550 to 800°F upstream of heat exchanger
230 to 320°F at baghouse inlet
37
-------
Particle size
35 to 400 mesh
Inlet SO- concentration
Baghouse air-to-cloth ratio
350 to 750 ppm
ft3/min
1 to 4
ft*
Sorbent feed method
Continuous
Batch
Semibatch (bag precoat)
Results
The final report of these results is due to be completed in Jate.
1979. Since the reporting of the results is currently in draft form, EPRI
was reluctant to discuss results in detail. However, the results were
described as being generally very encouraging as to the future of dry
injection as an FGD system.
Information Sources
Blythe, Gary. Telephone conversation with Navin Shah, EPRI Project
Director, July 17, 1979.
Shah, N.D., et al., "Application of Dry Sorbent Injection for SO and
Particulate Removal," paper presented at the EPA Symposium on Flue Gas
Desulfurization, Hollywood, Florida, November 11, 1977.
4.1,8 Carborundum, Dry Injection/Baghouse Collection Pilot on Stoker-Fired
Boiler. 1976 - Present
j>cppe
Carborundum (now a division of Kennecott Development Company) has tested
a dry injection/baghouse collection system using 100 acfm slipstream from a
small stoker-fired boiler near their Knoxville, Tennessee offices. A larger,
1000 acfm baghouse has been installed, and dry injection testing is continuing
at this scale. The 100 acfm unit was also equipped for spray dryer/baghouse
testing, and current plans are to equip the 1000 acfm unit with a spray dryer
as well.
Process Description
Dry injection/baghouse collection studies have been completed with
sodium bicarbonate, nahcolite, and ammonia sorbents. The smaller (100 acfm)
baghouse is an industrial size unit, with small bags. The larger (1000 acfm)
baghouse has 4 to 6 full size (11^-inch by 32-foot) bags. A small Bowen spray
dryer employing rotary atomization was used for spray dryer testing on the 100
acfm scale.
38
-------
Information Sources
Elythe, Gary. Telephone conversation with Don Boyd, Carborundum, May
16, 1979.
Blythe, Gary. Telephone conversation with Hank Ma jdesk!, Carborundum,
July 6, 1979.
Majdeski, H. M. Personal communication with Gary Blythe, July 17, 1979.
4.2 SPRAY DRYING/PARTICULATE COLLECTION
4.2.1 Atomics International (Rockwell) at Mohave Station. 1972
Under an agreement with Southern California Edison (SCE), Atomics
International conducted pilot plant test work of their (AI's) aqueous car-
bonate process (ACP). The test work was conducted at SCE's Mohave Generating
Station in the first half of 1972. Funding was provided by the WEST (Western
Energy Supply and Transmission) Association of utilities.
The major objective of the test program was to determine optimum opera-
ting conditions for the spray dryer scrubber using aqueous sodium carbonate
solutions for removal of S02 from flue gases. The operating results were also
to be used in designing a full-scale open-loop ACP process.
Process Description
In the open-loop (no sorbent regeneration) ACP process, the SO. is
removed by contacting the hot flue gas with an atomized solution of sodium
carbonate (Na^CX^). The dry product formed is a combination of sodium
sulfite, sodium sulfate, and unreacted reagent.
The Mohave test Installation included a 23-year old modified spray
dryer, 5 feet in diameter, a sodium carbonate feed system, and a multi-
cyclone solids collection system. A slipstream of flue gas, obtained down-
stream of the station ESPs was fed to the spray dryer. Provisions were made
for SO, spiking, and for testing at various temperatures. S02 concentrations
were continuously monitored.
Parameters Investigated
The test work was designed to investigate the following paramters with
respect to their effect on $©2 removal and system performance:
39
-------
•sorbent concentration,
•stoichiometry,
•Inlet flue gas temperature,
•flue gas flow rate,
•concentration of SC>2 in inlet gas, and
"recycle of products to feed.
One series of tests was conducted with the objective of demonstrating
effective scrubber performance under optimum design conditions.
Test Conditions
The flue gas flow rate varied from 1150 to 1375 scfm at an inlet temper-
ature ranging from 250 to 340°F. The inlet S(>2 concentration was usually
about 400 ppm, except for testing to determine the effects of higher S02
concentrations, in which case concentrations of up to 1465 ppm 502 were
used. Temperature drop across the spray dryer was usually between 120 and
160°F. However, when operation in the "wet" mode (>25 percent moisture in
the product) was required to meet particulate removal requirements in the
downstream cyclone, the temperature drop was 180°F (although the exit gas
was always maintained at least 25 °F above its dew point.) Inlet gas moisture
content was about 10 percent by volume. A liquid-to-gas ratio of 0.3 gal/1000
scf was employed in most tests. The weight percent of Na2CO_ in the feed
scOution varied between 4 and 5.5, except in the tests investigating the
effects of sorbent concentrations where it varied between 20 and 32 percent.
Results
The results of sorbent concentration testing with 400 ppm SO 2 flue gas
indicated that equivalent SO- removal could be achieved more efficiently with
low weight percent solutions. The most efficient 802 remova^ from 400 ppm
SO 2 flue gas at 300°F was obtained with a 4 percent weight Na2C03 solution.
This solution concentration gave a stoichiotnetric ratio (Na2C03 to 802) of
about 1. A lower temperature flue gas would require a more concentrated solution
to avoid saturating the exit gas. Higher inlet 802 concentrations would also
require a more concentrated feed solution for the same degree of removal.
A larger temperature drop over the spray dryer resulted in increased
removal efficiency. At a fixed feed rate, lowering the inlet gas temperature
also increased removal efficiency. In both cases, the increased vet contact
time enhanced removal.
At a constant feed rate of a 5.4 percent weight Na2C03 solution, in-
creasing the gas flow rate from 1150 to 1300 scfm resulted in a linear increase
in removal efficiency.
Sorbent utilization was found to Increase a« the inlet S02 concentration
increased. In some cases, utilization exceeded 100 percent. The authors
suggest that this may be the result of sodium bisulfite (NaHS03> formation.
40
-------
Recycled sulfate product salts did not Inhibit removal of SO.. Some
results indicated that sorbent utilization increased as the recycle amount
increased.
Reasonably consistent results showed two to three percent NO removal
under most conditions, although the accuracy of the analytical methods avail-
able was suspect.
In the series of tests designed to demonstrate system performance under
optimum conditions, 90 percent S02 removal efficiency was regularly achieved
using a 5.3 percent weight ^CO. solution at an L/G of 0.3 gal/1000 scf and
a stoichiometric ratio of 1.15. The total system pressure drop was between
9 and 11 in. 1^0. During these tests it was established that higher
atomizer wheel speeds Increased removal efficiencies. No mechanical or
maintenance problems were encountered during the testing. Additive
utilization approached 90 percent.
Conclusions and Comments
The spray dryer was shown to be an efficient method of contacting a
dilute Na2C03 solution with hot flue gas for SO, removal. For a fixed feed
rate, maximum removal efficiency occurred at lower Inlet gas temperatures.
Both additive utilization and removal efficiency were near 90 percent.
The Mohave facility was equipped with a multi-cyclone collector which
sometimes required the spray dryer to operate in the "wet" mode (• 25 percent
H20 in the product) to achieve the required particulate removal requirements.
Information Sources
Gehri, D.C. and J.D. Gylfe. Pilot Test of Atomics International Aqueous
Carbonate Process at Mohave Generating Station. Final Report AI-72-51,
Atomics International Division/Rockwell International, Canoga Park, CA.
September 1972.
4.2.2 Koyo Iron Works. Pilot-Unit. 1973
Scope
Performance characteristics of a spray drying process with NaOH and
Na_CO aerosols was studied by Koyo Iron Works in Japan. The tests were
carried out on a pilot-scale facility during 1973.
Process Description
Flue gas from a 3000 Ib steam/hr oil-fired boiler was passed through a
spray dryer for removal of SO by reaction with NaOH or Na-CO. droplets. A
two-fluid atomizer was used to disperse the liquid from the top of the dryer
into the flue gas. The gas, containing dry product solids, then passed through
41
-------
a multi-cyclone and an electrostatic precipitator to remove the solids. The
dry powder collected was a mixture of Na^SO , Na CO and Na.SO..
Parameters Investigated
The parameters investigated in the tests included:
•inlet sulfur dioxide concentration,
•temperature drop across the dryer, and
•stoichiometric ratio (ratio of NaOH fed to theoretical amount needed
to react with all incoming 802).
Test Conditions
The flue gas flow rate was about 850 scfm in all cases. The
temperature of the gas varied from 320 to 428°F. The fuel oil for the
boiler contained 1 to 3 percent sulfur, resulting in inlet S02 concentrations
of from 600 to 1700 ppm.
Results
Sulfur dioxide removal was reported to increase as the stoichiometric
ratio or inlet 862 concentration increased. An SC<2 removal of 89 percent was
reported for flue gas containing 1300 ppm S02, The stoichiometric ratio
required for this removal was 1.2. At the same inlet S02 level, but with
a stoichiometric ratio of 1.0, removal decreased to 84 percent. The outlet
flue gas was 278°F in both cases, resulting in a 80°F temperature drop over the
dryer.
At constant inlet Ras temperature, it was founrl rh«r sulfur dioxide
removal increased as the temperature drop across the dryer increased. This
reflects the increase in stoichiometric ratio that occurs when feed rate is
stepp3d up to obtain a lower outlet gas temperature (greater AT). It was
also observed that for a constant temperature drop over the dryer, a lower
inlet gas temperature resulted in greater removal efficiency. This is
attributed to the lower water evaporation rate at the lower temperature,
thus allowing more reaction time in the liquid droplet where the bulk of
the sorption reaction occurs. A summary of the magnitude of the temperature
effects is presented in Table 4-6.
TABLE 4-6. TEMPERATURE EFFECTS ON S02 REMOVAL - KOYO SPRAY DRYER10
SO 2 Removal
(%)
86
74
81
69
79
68
Inlet Gas
Temperature
<°F)
320
320
356
356
374
374
Temperature
Over Spray
(°F)
146
108
. 153
108
153
115
Drop
Dryer
42
-------
collection
The authors reported 99 percent collection efficiency for the ESP down-
stream from the multi-cyclone.
Conclusions and Comments
The sulfur dioxide removals and utilizations reported show spray drying
to be an efficient method of sorbent-gas contacting. The report states that
either sodium hydroxide or sodium carbonate may be used as the sorbent, but
it is not absolutely clear which was used for the data reported.
Information Sources
Isahaya, E.F., "A New FGD Process by a Spray Drying Method Using NaOH
Aerosols as the Absorbing Chemical," Staub Reinhaltung der Luft , (In
English), 33(4), April 1973.
4.2.3 Rockwell/Wheelabrator-Frye at Leland Olds Station. 1977-78
The primary objective of the Leland Olds pilot plant was to demonstrate
the applicability of Rockwell's open-loop Aqueous Carbonate Process (ACP) on
an existing lignite-fired boiler. The tests were conducted on Unit 2 at Basin
Electric's Leland Olds SLation in Stanton, Nl> during 1977 and 1978. The
results of the test work were used in design! n;:, a full-scale FC'.l) system for
the proposed Coyote Station of the consortium headed by Otter Tail Power
Company and to be located in Beulah, NU.
Sodium carbonate and trona were used as the scrubbing media for demon-
stration tests. Another series of tests was conducted to explore the
feasibility of less costly sorbents such as lime or limestone. Also, a
week-long endurance test (using trona) was conducted to demonstrate long
term reliability.
Process Description
The open-loop ACP includes a spray dryer for contacting the atomized
sorbent solution with hot SO -laden flue gas and a downstream solids
collection device to remove fly ash and entrained product solids. At
Leland Olds the collection device was a Wheelabrator-Frye two-compartment
fabric filter with a total filter area of 1098 ft . The bags were cleaned
by reverse air-shaking methods. The spray dryer was a 7-ft diameter Bowen
model equipped with a rotary atomizer. The clean flue gas vented to the stack
from the filter did not require reheat, nor did the gas exiting the spray
dryer as it was usually maintained about 40°F above the dew point.
43
-------
Tests were conducted on a 2 MW (6000 acfm) sidestream of flue gas
from a 440 MW cyclone boiler burning North Dakota lignite.
Scrubber feed was prepared by dissolving or slurrying sorbent in a
makeup tank from which it was pumped to the scrubber. In the open-loop
process the solids are discarded or sometimes used for recycle.
Parameters Investigated
Test subprograms included sodium carbonate tests, trona tests, hydrated
lime and limestone tests, and trona endurance tests. The parameters investi-
gated in each series are listed below.
1) Sodium Carbonate, Trona, and Hydrated Lime and Limestone Tests
•sorbent concentration
• inlet S02 concentration
•inlet temperature of flue gas
•temperature drop over the dryer
•flue gas flow rate (except for lime/limestone tests)
2) Endurance Tests with Trona
•inlet SO- concentration
•inlet temperature of flue gas
Test. Conditions
Table 4-7 lists test conditions for the various series.
TABLE 4-7. OPEN-LOOP ACP PILOT-PLANT TEST CONDITIONS -
ROCKWELL AT LELAND OLDS4
Sorbant Concentration Inlet SO, Cone.
Serlca (vt *) (pp.)
Sodiua Carbonate 4 to 22
Trona 7 to 12
Hydrated LiM 3, 5 4 10
and Llaeatone
Trona En- 8 to IB
durance Teeta
900
800
600
700
to 2300
to 1400
to 2000
to 1400
Flue Get rioo Dryer Inlet
•efn (*r>
1500, 2500 t 3100 260 to 350
1500 to 2500 250 to 350
2500 310 to 350
2400 to 2500 293 to 329
(with boiler
operation)
Dryer
•r
90 to
100 to
120 to
100 to
«a
170
170
170
130
a The difference in flue gao temperature between the spray dryer
inlet and outlet.
44
-------
Results
Although specific findings of this study are considered proprietary
information, overall the open-loop ACP was demonstrated to be an effective
S02 removal method. Trona and sodium carbonate were equally effective in
removing S02 with up to 85 percent of the S0? being removed in the spray dryer
and an additional 10 to 20 percent removal occurring on the fabric filter
cake. Sorbent utilizations ranged from 80 to 100 percent. In all cases
the product was a dry, free-flowing powder.
Hydrated lime was not as effective an absorbent as the sodium alkalis
on a once-through basis. Removals were in the 45 to 60 percent range and were
accompanied by low additive utilization. The single test in which lime-
stone was used gave poor S02 removal despite a high stoichiometric ratio.
The parameters that seemed to influence the reaction of S02 with
material collected on the filters were filter temperature, pressure drop,
gas flow rate, and sorbent utilization in the spray dryer. An increased
temperature drop across the spray dryer, caused by increasing liquid rate,
increased removal.
The endurance test with trona was successful. No degradation of SO,
removal or serious equipment malfunction was observed.
Conclusions and Comments
The open-loop ACP was shown to be a viable S0? removal process. The
additional removal occurring in the filter cake did not affect baghouse
performance.
As a result of this and subsequent test work at the Leland Olds pilot-
plant facility, Rockwell/Wheelabrator-Frye was awarded a contract to design
and build a full-scale open-loop ACP system at the proposed Coyote Station.
The system is to be completed in spring, 1981.
Information Sources
Dustin, D.F., Report of Coyote Pilot Plant Test Program, Test
Report, Rockwell International, Atomics International Division
Canoga Park, CA, Nov. 1977.
45
-------
4.2.4 Joy/Niro at Hoot Lake Station. 1977-1978
Scope
In late 1977 Joy Manufacturing and its Western Precipitation Division
entered into an exclusive agreement with Niro Atomizer to design and market
dry FGD systems. Niro had begun test work using a spray dryer for FGD
applications in 1974, on the 1000 to 3000 acfm scale, at their Copenhagen
facility. The tests investigated various alkaline sorbents such as lime,
limestone, and sodium carbonate. During the six-month period of November
1977 to April 1978 Joy/Niro designed, constructed, and operated a pilot
plant test facility at Otter Tail Power Company's Hoot Lake Station in
Fergus Falls, Minnesota. The purpose of this pilot unit was to demonstrate
the viability of the Joy/Niro process and to provide the design basis for a
bid on a commercial unit to be constructed at Basin Electric's Antelope
Valley Station. Additional test work was conducted during the period of
September to December 1978 to acquire data for preparation of a bid for a
dry FGD system for the Basin Electric Laramie River Station.
Process Description
Flue gas from a 53 MW CE boiler was supplied to the spray dryer at an
average rate of 20,000 acfm. The gas temperature was 3108F. The Joy/Niro
dry FGD process consists of a spray dryer absorber equipped with a Niro
atomizer and a Joy baghouse that was used for collection of solids from the
existing flue gas.
The process flow diagram for the Hoot Lake test facility is shown in
Figure 4-2. Hot flue gas entered through a roof gas disperser and contactc-d
the atomized slurry. Some of the dry product was collected at the conical
bottom of the spray dryer, the rest being swept out with the exiting flue gas
and collected, along with the fly ash present, in the downstream baghouse or
electrostatic precipitator (ESP). The quantity of slurry was controlled
using variable speed Moyno pumps to maintain the flue gas outlet temperature
at 30°F to 40°F above saturation.
The spray dryer used was 11.25 ft in diameter and was equipped with
a Niro atomizer coupled to a variable speed motor. The four-compartment
baghouse had a 9000 acfm capacity with bag cleaning by both reverse-air and
shaking methods. Both acrylic and fiberglass bags were tested with no
difference in performance observed. The ESP employed in some tests was a
single field unit capable of handling up to 5000 acfm of flue gas.
Parameters Investigated
The testing was conducted in two phases. First, a series of short
(10-12 hr) tests covering the expected operating ranges were performed.
The second phase consisted of a 100-hr endurance test conducted to
determine the ability of the system to meet maximum SC^ removal requirements
46
-------
head tank
flue (aa fro*
hot flit of alr_
prcheatar
flue gat
air preheater
* Q transf
O poa*
ansfer
cold air
Inlet
feed
n '
—-fa***-
to stack
\ A-otarjr
VatMlzer
reclrculatlng
Figure 4-2. Hoot Lake pilot plant flow sheet.
-------
during continuous operation.
The first phase of tests was conducted using various types of lime and
lime-slaking methods. Optimum hydration/lime-slaking methods were determined
for use in the 100-hr endurance test. The other parameters investigated
during the first phase were spray dryer inlet temperature, flue gas tamperature
drop across the spray dryer, SOo inlet concentration, flue gas flow rate,
sorbent concentration, atomizer speed and wheel configuration.
Additional short tests were conducted to investigate the Joy/Niro
solids recycle concept. Figure 4-3 depicts the recirculation flow scheme.
Tests using soda ash as the sorbent were also conducted. As a final
verification of process reliability, two types of upset condition testing
were performed to check 1) the control system and 2) the effects of low
spray dryer exit gas temperature on baghouse performance.
Test Conditions
The primary variables used in the first phase of test work are listed
below:
SO™ Inlet Concentration: 800, 1200, and 1600 ppm/v
Lime Slurry Concentration: 5, 10, 15, 20, and highest possible wt %
Temperature Difference Across the Spray Dryer: 70, 105, 140, and 175°F
These test conditions were specifically related to the proposed Antelope
Valley dry scrubbing system. Additional test work was conducted over a wide
range of conditions. S02 inlet concentration varied from 200 to 4000 ppm and
the stoichiometric ratio was varied from 0.9 to 3.0.
Results
4 Specific correlations between S02 removal and stoichiometric ratio at a
given S02 level were not available. The results presented in the literature
covered wide ranges of SO, levels but gave only general removal figures. For
a "once-through" system, the SOo removal was reported to be 90 percent at an
inlet S02 concentration from 200 to 2000 ppm with a stoichiometric ratio of
2.0 to 3.0, depending on the gas temperature drop across the dryer. Removal
versus stoichiometric ratio curves were presented for "typical" cases, but
they were not accompanied by S0 concentration levels.
48
-------
SLURRY
/Ts.
SPRAY
DRYER
FLUE GAS
CLEAN
QAS
BAQHOUSE
WASTE DISPOSAL
I -S MAKEUP
LIME SLURRY
70-1585-1
Figure 4-3. Flowsheet of pilot-plant operation
with partial solids recycle.3
49
-------
Because of the limited data available, only general trends can be
presented:
1) Additive utilization was significantly improved in tests where
partial recycle of solids was employed. Ninety percent removal
was obtained with stoichiometric ratios between 1.3 and 1.7.
2) Soda ash was found to be much more reactive than lime, with
stoichiometric ratios of 1.0 to 1.2 resulting in 90 percent S02
removal at inlet S02 concentrations ranging from 800 to 3000 ppm.
3) Increasing the temperature drop over the dryer at a constant
stoichiometric ratio was fo-nd to increase the SO- removal. For
an example of this trend see Table 4-8. If the outlet temperature
of the gas is taken to be 30°F above the adiabatic saturation
temperature of 130°F, a temperature drop of 155°F corresponds to a
dryer inlet flue gas temperature of 315°F, very near normal opera-
ting flue gas temperature at both the Hoot Lake Station and the
proposed Antelope Valley Station.
TABLE 4-8 THE RELATIONSHIP BETWEEN SPRAY DRYER TEMPERATURE DROP AND
SO- REMOVAL AT A CONSTANT STOICHIOMETRIC RATIO OF 2.5
JOY/NIRO AT HOOT LAKE3
Spray Dryer AT (°F)
Temperature Difference
155
125
105
S0« Removal
90
75
60
U)
The control system responded well to upset conditions, such as a large
temperature drop in the inlet gas to the dryer or operation at temperatures
low enough to result in a wet product. Baghouse performance was not affected
during periods of "wet" operation as the exiting gas remained unsaturated.
Particulate removal exceeded 99+ percent in all cases with both baghouse
and ESP collection.
During the mid-September to mid-December 1978 test period, coal from
the source to be used at the proposed Laramie River Station was shipped to
the Hoot Lake Station for use in several days of test runs. The Laramie
River ash was reported to be quite cementitious, but no operating problems
were reported. Some tests were conducted adding water treatment sludge
to the atomizer feed. This was found to be a suitable method of handling
for the sludge.
In limited test work with commercially available ground limestone the
best S02 removal obtained was 50 to 60 percent.
50
-------
Conclusions and Comments
The Joy/Niro dry FGD system performed well during pilot plant parametric
and endurance testing. No major problems were encountered with the equipment
or process chemistry.
Though no specific removal data were given, it appears that 85 percent
SO, removal was achieved with near theoretical amounts of lime when partial
recycle of solids was employed. The exact removal is, of course, dependent
on the inlet S02 concentration among other factors.
The literature also states that the "initial" stoichiometric ratios were
considered conservative due to large air leakages. It is not clear at what
point during the test period this problem was corrected.
It was also concluded that the method of lime slaking is an important
factor in determining the effectiveness of lime as a sorbent.
Sulfur dioxide removal Is limited by, among other factors, the amount
of slurry that can be supplied to the dryer. To avoid moisture condensation
in downstream particulate removal equipment or stack, the slurry delivery
rate must be maintained below levels that would cool the flue gas below
its adiabatic saturation temperature.
After the Hoot Lake Station test work was completed, Joy/Niro bid on
and was awarded a contract for a 440 MW commercial dry FGD system to be
built at Basin Electric's proposed Antelope Valley Station. The system to
be constructed will be a large-scale version of the pilot plant facility
described above with the only major change being the method of gas dispersion
into the spray dryer. See Section 5.12 for further details on this commercial
system.
Information Sources
Elythe, Gary. Meeting notes, meeting at Joy Manufacturing, June 14,
1979.
Davis, R.A., et al., "Dry SO Scrubbing at Antelope Valley Station,"
presented at the American Power Conference, April 25, 1979.
Felsvang, Karsten, "Results of Pilot Plant Operations for S0?
Absorption," presented at the Joy Western Precipitation Division
Seminar, Durango, CO, May 21, 1979.
Kaplan, Steven, "The Niro-Joy Spray Absorber Development Program - Pilot
Plant Description and Test Results," presented during Joy/Niro-
sponsored tour for executives of U.S. power industry, Copenhagen,
Sept. 23-30, 1978.
51
-------
4.2.5 Babcock and Wilcox. Pilot Scale Spray Dryer/ESP at Velva, 1978
B&W was one of four vendors which piloted a spray dryer-based dry FGD
system at Basin Electric. B&W began testing at the Basin Electric's
William J. Neal Station, near Velva, N.D., using technology developed by
Hitachi of Japan. This pilot plant eventually employed two-fluid nozzles
for atomization and a horizontal reactor chamber, whereas the Hitachi
process employed nozzle atomization in a vertical reactor. Gas flow out of
the Hitachi reactor was through a duct in the bottom of the vessel, which
tended to plug during testing. This led B&W to abandon the Hitachi config-
uration and instead adapt the steam atomized oil burner ("Y-jet" nozzle)
technology in a horizontal reactor. Figure 4-4 illustrates the Y-jet
nozzle configuration.
Process Description
The reactor at Velva was 30.5 ft long by 5,75 ft square, with one
atomizing nozzle and three hoppers located In the floor to handle solids
drop out. The gas capacity of the unit was 8000 acfm. While the Hitachi
reactor was tested with both a 16-bag Mikropul pulse jet baghouse and an
ESP collector, the Y-Jet configuration was tested only with an ESP collector.
Reagents tested included soda ash, pebble lime, hydrated lime, ammonia
addition with lime, and precipitated limestone. A ball mill slaker was
used for lime slaking. Reactor inlet and outlet gas temperatures, Inlet
SO. concentration, gas flow, and sorbent stoichiometry were also varied
during the test program.
The Y-jet configuration pilot unit began operation in June 1978. Future
studies include full parametric studies on a 1500 acfm pilot plant at the
company's research facilities in Ohio and a 120,000 acfm demonstration unit
treating flue gas from a 500 MW unit at a large western utility. Details
of these programs are discussed in Section 5.
Results
Because the results of this privately funded research effort are con-
sidered proprietary, no specific performance results are available. How-
ever, it was found that due to reagent costs, line was the preferred reagent.
Also, B&W found that they could operate more comfortably near adiabatic
saturation at the reactor outlet when using an ESP rather than a bag
collector for downstream particulate collection. The increase in reagent
utilization due to closer approach to saturation was found to be greater
than that resulting from SO. removal on a bag collector surface. Apparent
utilizations approaching or even exceeding 100% were found to be possible
during lime tests when the reactor outlet temperature closely approached
saturation. Utilizations greater than 100% may be due to the alkaline
species in the recycled fly ash reacting with the flue gas SO .
52
-------
ATOMIZER BODY
vn
u>
QUICK DETACHABLE
COUPLING
INNER TUBE
CAP
AIR OR STEAM
SLURRY
SPRAYER PLATE
Figure A-4. "Y-jet slurry atomizer."
11
-------
In test work using only recycled fly ash as a reactant, it was found
that fly ash reactivity was increased by recycling through the ball mill
slaker rather than by recycle directly as a slurry. Presumably, this
increase was at least partially caused by the reduction in particle size
resulting from treatment of the fly ash in the ball mill.
Information Sources
Blythe, Gary. Meeting notes, meeting at Babcock and Wilcox, June 28,
1979.
Janssen, Kent E. and Robert L. Eriksen, "Basin Electric's Involvement
With Dry Flue Gas Desulfurization," paper presented at the Fifth EPA
Symposium for Flue Gas Desulfurization, Las Vegas, Nevada, March 5-8, 1979.
Slack, A.V., "A.V. Slack FGD Report #62," January 1979, pp. 15-30.
4.2.6 Carborundum/DeLaval Spray Dryer Pilot Plant at Leland Olds. 1978
As one of the four process vendors invited to pilot a spray dryer-based
FGD system for bidding on the Coyote station, Carborundum participated in
pilot testing with a 15,000 acfm unit at the Basin Electric's Leland Olds
Station. The pilot unit employed a DeLaval spray dryer as a flue gas
contactor. A Carborundum baghouse was used as a particulate collection
device.
Process Description;
The pilot unit was a spray dryer baghouse system. Lime, soda ash, and
ammonia were tested as sorbents.
Results
The lime tests by Carborundum at Leland Olds are significant because
a wide range of spray dryer outlet temperatures were tested. Outlet
temperatures ranged from around 122°F (a 4°F approach to saturation) to
over 200°F (90°F approach to saturation). S0» removal varied from 55 percent
to 91 percent (stoichiometric ratio of about f, inlet SO. of 700 pptn).
Higher removals occurred at the lower dryer outlet temperatures.
Information Sources
Blythe, Gary. Telephone conversation with Don Boyd, Carborundum,
May 16, 1979.
Blythe, Gary. Telephone conversation with Hank Majdeski, Carborundum,
July 6, 1979.
Majdeski, H.M. Personal communication with Gary Blythe, July 17, 1979.
54
-------
A.2.7 Bechtel Power Company, Conceptual Spray Dryer/Baghouse andDry
Injection/Baghouse Study, August 1978 to Present
Bechtel is being funded by EPRI in a economic study comparing various
spray dryer-based and dry injection-based dry FGD systems with a wet limestone
scrubber. Initial work was begun in August 1978; a final report is expected
by the end of 1979. Most of the economic cases were for & hypothetical
500 MW, 85 percent SO. removal unit, although two 200 MW cases were calculated.
Four coals were considered, ranging from 0.5 to A.O percent sulfur. Trona
and nahcolite sorbents are being considered for the dry injection cases,
lime and soda ash sorbents are being considered for the spray dryer cases.
The study compares capital and operating costs of the dry systems to a
wet limestone scrubber. Also, the study considers the engineering aspects
of applying dry FGD, such as integration of the unit with the utility boiler.
Information Sources
Elythe, Gary. Telephone conversation with Navin Shah, EPRI, July 20,
1979.
4.3 OTHER RESEARCH
This area Includes test work with bench- or pilot-scale bed reactors
and combustion of a coal/limestone fuel.
4.3.1 FMC Corporation. Bench-Scale Fixed Bed. June 1970
During the period of June 1969 to July 1970, FMC Corporation undertook
a series of bench-scale screening tests designed to evaluate potential SO,,
sorbents. The study was funded by the National Air Pollution Control
Association (now EPA).
Process Description
The experimental apparatus was designed to evaluate the sorption
characteristics of a given material in less than an hour. The sample
holder was a vertically mounted stainless steel tube A inches long and
1/2 inch in diameter. The sorbent was supported on a sintered steel plate
near the bottom of the tube. Synthetic flue gas, prepared by blending
bottled gases, was passed through the fixed bed of sorbent.
55
-------
Parameters Investigated
Various types of sorbents were tested: sodium carbonates, impregnated
silica gels, impregnated fly ash, hydroxides, and sulfides.
Test Conditions
All experiments were carried out with one gram samples of sorbent. The
gas flow rate was about 0.6 ft /sec with an inlet SO concentration of 3500
ppm at 257°F. The "typical" dry gas composition was 82.7 percent N , 14.9%
CO- and 3.0% JO . The moisture content was 4 or 6 mole percent.
Results
Table 4-9 lists the sodium-based materials tested, the SO outlet
concentration after 25 minutes, the duration of the test run, and the
sorbent utilization at test completion. The sodium carbonates were
considerably more effective than any of the other materials tesU'd.
Conclusions and Comments
The results of this study indicate significant sorption of SO by
"fixed-bed" samples of sodium carbonate and sodium sesquicarbonate. Natural
soda ash and sesquicarbonates were superior to the commercial carbonates, both
in amount of SO, absorbed after a given time and in utilization of the sorbent.
Information Sources
Friedman, L.D., Applicability of_Inorganic Solids Other Than Oxides to
^he Development of New Processes for Removing SO? from Flue Gas, FMC
Corporation for the National Air Pollution Control Administration,
Contract No. CAA 22-69-02, Princeton, NJ. December 1970.
4.3.2 Nagoya Institute of Technology, Multi-Stage Bed. September 1973
In 1973, bench-scale studies were conducted to evaluate the use of soda
ash in removing SO- from air in a multi-stage bed. Also, the rate of S0?
sorption was investigated in bench-scale equipment. l
Process Description^
The active soda ash was added to the top stage of the bed and moved to
lower levels with rakes. The dimensions of the apparatus were not available.
The test BBS was air with controlled amounts of SO. and water vapor added
to achieve the desired composition.
56
-------
TABLE 4-9. RESULTS OF SCREENING TESTS ON SODIUM CARBONATE AND SODIUM
SESQUICARBONATE - FMC CORPORATION6
Reactant
SO
Inlet
(ppnO
S02 Outlet Cone.
After 25 Min.
(ppm)
Duration of
Test (minutes)
Stoichiometric
Utilization*
(percent)
FMC Soda Ash
Grade 50b
3500
2694
72
30
FMC Soda Ash
Grade 120
3500
2497
33
FMC Soda Ash
Grade 100
3500
2688
7A4
42
Precipitation 3500
CakeC
2120
91
78
Natural Ashd 3500
FMC Sodium 3500
Sesquicarbonate
1765
1879
70
746
64
81
FMC Anhydrous
Sescuicarbonate
3500
1688
97
91
aBased on moles of Na20 fed per mole S02 fed in the gas.
^Less porous than Grade 100 or 120.
clmpure sodium sesquicarbonate (impure Na2C03'NaHCOs»2HaO),
!*Calcined trona.
57
-------
Parameters Investigated
The tests were conducted to investigate the effects of the following
on SO removal:
•concentration of water vapor,
•particle size,
•temperature, and
•mole ratio of soda ash to inlet SO .
Test Conditions
Table 4-10 provides the experimental schedule. Inlet S02 concentration
varied from 1050 to 2460 ppm.
Results
Table 4-11 lists the results of the test work. Removals of 97+ percent
were reported for all conditions.
In other tests with a thermobalance, active soda ash was found to absorb
SO. at the rate of 1 mg/cm2-min below 170*C 038°F). Soda ash containing
up to 50 percent sodium sulfite absorbed SO. as fast as pure soda ash.
Wien more than 50 percent sulfite was present, the absorption rate fell off
slowly. Sulfite was oxidized to sulfate at temperatures above 428°F.
Conclusions and Comments
The study concluded that the multi-stage bed removal method was suitable
for small plants. Using the scale-up factors proposed by the researchers,
a 330 MW plant with 2000 ppm S02 flue gas would require about 14 desulfurizers,
each 30 feet in diameter.
Information Sources
Yaraada, Tamotsu, et al. Desulfurization of Combustion Exhaust by
Soda Ash. Nagoya Institute of Technology, Japan, 25 395-403, 1973~
-------
TABLE 4-10. EXPERIMENTAL SCHEDULE - NAGOYA MULTI-STAGE BED23
Parameter
A:
B:
C:
D:
Test Number
1
Water Vapor (%) 6
Particle Size (mesh) -50/+100
Temperature (°F) 308
Mole Ratio . 1.0
2
3 A
12 18 24
-100/+200 -200/+325 -325
592 482 572
1.2 2.0 2.5
TABLE 4- 11. EXPERIMENTAL RESULTS - NAGOYA MULTI-STAGE BED23
Parameters
A
1
1
1
1
2
2
2
2
3
3
3
3
4
4
4
4
B
1
2
3
4
2
1
3
4
3
4
1
2
4
1
2
3
C
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
D
1
3
4
2
4
2
1
3
2
4
3
1
3
1
2
4
Inlet S02 Cone.
(ppm)
1050
1550
1440
1440
1650
1800
2140
1540
1320
1740
1590
2460
1580
1760
1770
2310
Outlet S02 Cone.
(ppm)
18
11
21
17
20
20
14
28
18
11
35
23
10
44
38
20
-------
4.3.3 Stearns-Roger/Superior Oil, Fixed-Bed and Countercurrent Reactors.
1974
In December 1973, Superior Oil contracted with Stearns-Roger for the
design, construction, and management of a pilot facility of investigate the
feasibility of using nahcolite as a stack gas SO. removal agent. Bench-
scale tests using a fixed bed reactor were conducted to investigate reaction
kinetics. A pilot-scale countercurrent reactor was used to verify bench-
scale results and to provide data for a conceptual design and a hypothetical
500 MW power plant.
Bench-Scale Test Work
Process description—After passing through a filter to remove particu-
lates, the hot flue gas flowed downward through a thin bed (1-inch deep) of
screened and sized nahcolite or sodium bicarbonate. The flue gas flow was
large enough with respect to the thin bed so that SO. concentration was
essentially constant across the bed. Provisions were made for spiking the
gas with S07 and for heating it to the desired temperature. Portions of the
nahcolite bed were withdrawn periodically to yield conversion versus time
results.
Parameters investigated—The main parameters investigated were granule
size and reaction time. The grade of sorbent Commercial sodium bicarbonate
and nahcolite) and S02 concentration were also varied. The prime objective
was to correlate experimental data with mathematical models to detirmine
the rate controlling step of the reaction.
Test conditions—The nahcolite and sodium bicarbonate granules ranged
in size from 10 mesh to 1/2 inch. The SO. concentration was varied between
450 and 10,000 ppm.
Results—Comparison of conversion versus time experimental data and
predicted" results of mathematical models of reaction kinetics led to the
observation that the nahcolite-60. reaction is controlled by the diffusion
of SO. through the layer of sodium sulfate that builds up on the outer
surface of the particle. The experimental data correlated very well with
the ash diffusion model, whereas the gas-film diffusion and chemical reaction
controlled models were not confirmed by data.
Other results suggested that pretreatment of the nahcolite (i.e., by
heating to decompose it to the more porous Na.CO,) may improve its SO.
removal capabilities.
60
-------
Conclusions and Comments—The bench-scale study determined that the
NaHCCL/SO. reaction is controlled by the diffusion of SO. through the sodium
sulfate layer that builds up on the particle. This observation only applies
to the fairly large particle sizes used; the reaction may be controlled by
gas-film diffusion for smaller particles.
Pilot-Scale Test Work
Process description—The pilot-scale countercurrent reactor was 42 inches
in diameter. Hot flue gas flowed up through a slowly descending bed of
nahcolite or sodium bicarbonate. Since the reactor was insulated,
temperature drop was negligible. The solids residence time ranged from 100
to 200 hours. Nahcolite was added to maintain a constant stoichiometric
ratio, and spent solids were removed to maintain a constant bed level.
Temperature, pressure drop across the reactor, and inlet/outlet SO.
concentrations were continously monitored.
Paramters investigated—The test variables were:
• temperature,
•inlet SO. concentration,
• bed heignt, and
• flue gas flow rate.
Test conditions-Test conditions are given in Table 4-12.
Results—The results of the four tests conducted are given in Table
4-12. Sulfur dioxide removals of 80 to 84 percent were achieved with less
than stoichiometric amounts of sodium bicarbonate. Additive utilization
was about 97 percent. The SO. removal efficiency was lower in the nahcolite
test; 67 percent with 75 percent additive utilization. The investigators
pointed out that the purpose of the nahcolite test was to verify the
predictive ability of their mathematical model, not to optimize SO. removal
or additive utilization.
Pressure drops over the reactor ranged from 16 to 22 in. of H.O. No
results were presented as to the effect of varying temperature, bed height,
or flue gas flow, although the temperature was 20°F lower in the nahcolite
run. Nitrogen oxides removal was reported to be 42 percent In the test with
nahcolite, the only run for which such a determination was made.
Conclusions and comments—Although high SO. removals with good additive
utilization were achieved in "the countercurrent reactor, there remain certain
disadvantages to the process. The countercurrent reactor is a somewhat
mechanically complicated piece of equipment and might be prone to operating
and maintenance problems on a full-scale size. Also, the pressure drop over
the reactor is much greater than that for dry injection or spray dryer systems.
61
-------
TABLE 4-12. TEST RESULTS WITH PILOT-SCALE COUNTERCURRENT REACTOR20
Average particle
size (cm)
Bed depth (in.)
Flue gas flow
(acfm/ft2)
Temperature (°F)
Inlet S02 (ppm)
Test 1
Sodium Bicarbonate
0.43
36
100
300
810
Test 2
Sodium Bicarbonate
0.43
36
100
300
810
Test 3
Sodium Bicarbonate
0.43
36
100
300
2480
Test 4
Nahcolite
0.37
42
98
280
675
Stoichiometric
ratio
removal (%)
NOX removal (%)
Fly ash
removal (%)
Pressure drop
(in. H20)
Additive
utilization (%)
0.86
84
18
98%
0.86
84
^93
21
98%
0.82
80
16
97%
(150 ppm NOX)
0.89
67
42
98.6
22
75%
-------
Information Sources
Steams-Roger, Nahcolite Granule Scrubbing System Feasibility Study,
Vol. 1, for Superior Oil Company, Nov. 1974.
4.3.4 R. W.E. Tests in Germany
R. ViE., a German utility company has evaluated various calcium-based
alkali materials for removing S00 from flue gas by injecting them through
low NO burners in an existing 60 MVe lignite-fired boiler. R.V.E. has
been using limestone for about a year in this process to achieve compliance
with local air pollution regulations.
Process^Descript ion
This is a very simple process that takes limestone from storage,
pulverizes it along with coal, and injects the coal/limestone mixture into
the boiler through low NO burners. This is a retrofit installation that
uses the boiler's existinf burners. Process control is simple and straight-
forward; an instrument is used to monitor outlet S0? concentration which in
turn regulates the flow of limestone to the coal pulverizer. The coal fired
by the utility is a German brown coal which is similar to a low-grade U.S.
lignite. Its heating value is reported to vary from 4300-5000 BTU/lb with
a sulfur content of 0.4 to 0.7 percent.
Results
R.W.E. has apparently examined the use of three sorbent materials:
limestone, lime, and calcium sludge from water treating. Their
experience has shown that limestone is the best sorbent for their use
due to its superior handling characteristics ovur the other sorbents.
Results of their testing indicate that S02 removals of hO-90 percent
can be achieved with stoichiometric calcium to sulfur ratio of up to 3.
No time intervals were, however, given for these high removal rates so it
may not be known how effective this process will be in achieving high
S02 removals over long periods of time.
Over the last year, the system has operated well achieving compliance
with local air pollution regulations. However, due to the low sulfur
coal being burned, coupled with non-stringent SO control requirements,
the system has only had to achieve 25-50% SO- removal during this time.
The stoichiometric ratio required to achieve this level was reported to be
about 1. Capital investment costs (including modifications to the
particulate control equipment) were reported to be about $150,000 for
the 60 MWe system.
63
-------
Conclusions and Recommendations
Results of the test work described are promising and illustrate the
ability of this technique to remove SO. without plugging of boiler tube spaces.
The major unanswered question is the ability of this technique to achieve the
stringent removal levels required by U.S . air pollution regulations under
sustained periods of operation. Although it appears that the application
of this technique in the U.S. may be limited to only low sulfur Western
coals, this technique, if proven, appears to provide a very economical S0?
removal alternative.
Also, it should be noted that boiler tube spacings for German lignite-
fueled boilers are larger than those used in conventional coal-fired boilers.
In addition, the flow arrangement is such that the hotest portion of the
boiler is not at the "top" of the boiler. Consequently, coal-fired boilers
may need to be re-designed to avoid plugging if coal-limestone mixes are
to be used successfully.
Information Sources
Dickerman, J.C., Telephone Conversation with R.M. Statnick, U.S. EPA,
October 4, 1979.
64
-------
SECTION 5
CURRENT AND ON-GOING ACTIVITIES
This section summarizes pertinent activities of companies involved
in current and on-going developments for dry FGD systems. Included here
are background information on the companies, a summary of dry FGD related
research, a summary of commercial sales and marketing activities and a
description of the type(s) of dry FGD the company is marketing or deve-
loping. The current development activities of each of the companies
discussed in this section are summarized in Table 5-1.
The discussions of current and on-going development activities are
presented in alphabetical order by company name. For companies which have
been in the dry FGD business for several years, the information here includes
description of substantial research efforts and details of commercial sales.
For companies which have only recently entered this market, only a brief
discussion of their intentions may be given here. The results of most
current activities reported in this section are very preliminary and are
unpublished. Consequently, moat data were obtained through personal contact,
either by telephone or meetings held with process vendors. For this reason,
company representatives who may be contacted for additional information on
their dry FGD activities are provided.
5.1 BABCOCK AND WILCOX
Address: Babcock and Wilcox Babcock and Wilcox
20 South Van Buren Alliance Labs
Barberton, Ohio 44023 P. 0. Box 834
Alliance, Ohio 44601
Contact For More Information; Tom Hurst J. B. Doyle
Barberton Alliance Labs
(216) 753-4511 (216) 821-9110
Background
B&W was one of four companies which piloted spray dryer-based dry FGD
systems a,t Basin Electric in late 1977. B&W originally tested Japan's
Hitachi two-fluid nozzle atomization technology In a Hitachi vertical spray
dryer/reactor. When this concept proved inadequate for utility FGD applica-
tion, B&W developed a horizontal reactor using "Y-jet" steam-atomized nozzles
which were adapted from their standard oil burner technology. In this
aspect B&W has a rather novel approach to spray drying technology. Other
spray dryer-based dry FGD system vendors tend to use more standard spray dry-
ing technology with a. back-mix type reactor and two-fluid or rotary atomiza-
tion. B&W also takes a different approach in that they favor ESP collection
of the spent sorbent/flv ash mixture rather than baghouse collection which
appears to be favored by other vendors.
65
-------
TABLE 5-1. SUMMARY OF CURRENT R & D AND COMMERCIAL ACTIVITIES
STSTO6(»)/SCaU
LOCATIONS)'IMTE(»>
DcMonstration: 120.000 mcfm spray dryer/tiP. western utility: Scheduled for
start-up In late 1979.
••search: 1SOO acf* pilot «tt. Spray
dryer/cyclone systeu nolnly for parasc trie
testing and react 1m •echanli* Investiga-
tion.
Cossaerctal: Constructing spray dryer/CSF
Call-scale system (300 MO.
• 4 tf's Alliance Ohio Labs: Initial
test work has begun.
las la Electric 's Laranie River
Station. Wheat land. wro«ing: Start-up
scheduled for spring 4-r 1982.
Li**, soda ash. sodlus>-
based waste liquor.
Line with recycle of sol Ids
Slipstreams fross the spray dryer can be routed
to either an ESP or a baghouse.
This unit will provide data to better nderstani
the spray drying SOj reaction •echanlSB*.
Designed for 901 S<>2 removal, with a smlssBi
fttoichioswtry of 1.12. Designed for no recycle.
•uall/Anhydrc
Joint Venture
Pavelupaint: Till I 'oahj ilrn Is parttclpa- Initial work done at Anfaydro's
ting In privately f ilil pingisM to develop Copenkogon facilities. Future work
•pray dryer kasrit dry PCD systo*. Initial planned at Colorado Springs' Hsrtin
work cons-leted. Fsture plan* for testing Drake Station.
20.000 acfi dryer.
DuKiiist ration: WOO acfn dry inject Ion/ban- Colorado Springs' Martin Drake
hmisr cystoB includisg sorSent |ii • 111 ilnanl Station; Initial test work has been
and waite treat»eot studies. started.
Nahcolite, raw trona
(191 HaHCO,) asd upgraded
trona (922 MaHCO}).
lUbcolite. raw trona,
upgraded troni and liste.
Pending results of test uorfc suell upsets to
bid both spray dryer and dry injection systens.
Buell has bid a line spray dryer/baghouse industrial
syttan. but results have not been announced yet.
Originally undertaken to study fabric filter
ranoval of partlculates. Expanded to study dry
scrubbing and related waste disposal.
Coufeustion Engineering
(CE Power System)
Dcvelopnent: Installing a 20.000 acfu "dr> Northern States Power Sherbourne
absorber" followed by baghous* and ESP In County Unit ffl/Start-up date: July
parallel. 1979.
C-E also planning dry waste disposal study in
conjunction with pilot unit testlag.
DOE/Crand Forks Energy
Technology Center (CPETC)
Research: 200 acfn dr* Injectlon/bagbouse GTETC Laboratories! Tests
•ystcu. continuing through nost of 1979.
Hahcolite and trona.
GFETC plans to sasjple for SO2 and partlculate
resovsl efficiencies when Coyote plant *tarts up
(Spring 1961). Also planning waste disposal studies.
DOC/Morgantown Energy
Technology Center (JtETC)
Research: Laboratory studies on passing
huKldified fl»e «** thmvh fiud bed of
crushed Its**tone.
HTTC Labs/Prelinl»acy test work
-oapleted.
Ho further work is currently planned.
DOC/Pittsburgh Energy
Techno1 env Center (PETC)
Research: 500 Ibfbr coal-fired furnace PETC Laboratories/Tests will
being used to test dry injection of various continue through I98O.
sorbent s.
Sodiuu carbonate, sodIt-
bicarbonate, raw nfthcollte.
Testing is to continue to evaluate dry Injection
for high sulfur coals. Plans are to add a spray
dryer In •Id-1980.
Ecoltire Systevs, Inc.
Development: 10000 acfn sjobile spray dryer ' Operation in late 1979 at an » vet
baghouse pilot plant being constructed. unnj»ed utility.
Not Specified.
Pilot unit designed for flexibility to establish
design data base for dry FCD systeai for both utility
and Industrial syaeeaiB.
Energy and Pollution Controls, Inc. Development: Totally dry reactor ("slincer" Development work started in 1978.
distributes iry sorbent) combined with bag- rrt is continuing t«rvt work.
coal Mret! I - f- '?: ^tu'hr Nailer. Systa- is
f.T indtisrrttl applications.
Kydrated Line.
Strictly *n Industrial systeau
(continued)
-------
TABLE 5-1. (continued)
attell*--:. '.jnv.s Labs
fteseartb: Ccmtsisiicm »f coal '3i»«---^ IF* and BattcUe-t:o|w»t>us Labs
pellet ia pil^t-sctle spreader-stoker br-ilers (for 2S.OOO Ib st*am/hr boller>/oa-
and 25,000 lt> steamte industrial srre*ier- £"lnc tests - preliminary work on
Stofcer. ^O bhp scale has bvcn conducted to
date.
Tests were with Ca:S ratio of 7:1 la tke pellet.
Purther test work will investigate • 3:1 Cs:S pellet.
The coal Is a high sulfur Illinois coal (4.4! S>.
Energy and Envi
•esearch Corp.
Wsearcb: Coat>nstlon of a pvlvvrizad c«al'
Ismentonc nlztvrc In a low 9C bwrner system.
frevious tests condmcted oc the s4at 4* St-'ar
heat ispvt scale, work cootiamis* en "S.X*?
•tu/hr scale.
COC labs;
Test wrk irtll alto be coa4uete4
Btu/hr scale.
oai 10 Ml aW » «
Kerr Industries
itratlon cf 40.000 acfst dnr iB««ti?n/
textile flnlshiag »!•
Lfae, Liawstone
Bas>hooM fire has oelaTtd test start. Testls* •Lavs***
for March. 19W.
Joy/HIro Joint Veotare
CossMrelal: TW «at«l«e« Vail*? FCT
rystcsi (440 »'> Is snder rasctractin. T>«
process is s *Pr*? dT7*rKbaa>c«se srsteai.
Aetelop* Valley Station . Start
«F expected la the spring of 1992.
Lia» («ritfc partial recyel.)
ha* 2 bid. under *val«ttM. 1 la bean
and expects 5 to 6 bid reqeests by the eerf of 1919.
Joy/Hn> neaotlatlsc ie
provide a 100 JsV scale dsBoasirjtioa ««lt.
Sprar dryer follovW b? bacSc-nx -r IST
Xortbem States rower'* Uverslde
Sutlosi. Scheduled start-up la
lace 1900.
•ot Specified
1*0*1. Envirooatental rVodvcta B***lofweat : Stofcarr-firad boiler »roduc-
Division (formerly Carborvndta) lac 1000 acfsj of five fa*. Tost *rrt bclr«
caarfuctad on drr laijec cion/baab«voe . ?lanc
call for ia*tallls« 1000 acfn pilot sprav
dryer.
Dry isjectlofl studie
nave b««un. Snray drver to be 1
stalled In near future.
Pending results af pilot test msc*. ET» may offer
both dry injection and nprny dryer aysttsi
Iocs. EssjJnnerins
0eveJ»*SMnt: Work fcasun oa anray dryer
bseed PCD nvsten.
Ssdlna hydroxide. Sodltai Hear-
bcmte, Llaw.
Koch Is renortedly OM ye
PCD syaton. Pull-scale psrs
begun.
mkropul Corp.
Parnawtric studies: 1000 act* spray dryer/
poise-jet bscboune system.
nit, NJ lab«/on-S0io«.
Mkropul has no
PCD narket.
Hate plans tc enter vtility
Coamerclal: 40.000 acfn ladustrlal rrsten Strathnorc Paper Co.. Uoronoco.
(spray drver/pulse-jet baphoyf* beie« completed.) Start up in July.
M/A
Kesearch-CoCtrell Research •
Developnent
Development: 10,900 acfv pilot-scale
spray drver/baitbouse s»sttsi.
•Ig Brown unit (Te«a« Utilltlei)/
Pilot unit tests completed.
Teatad varioaa sorbenta
(Not Specified)
teacarcb-Cottrell la now In a poaitioe to begin
narket Inn utility and Industrial try PCD systssKs.
tockwell Internatlnnal/Uheeiabrator- Parametric: 4000-5000 acfn spray dryer/
Fry* Joint Venture baghouse ESP syst^n.
Declgfi: MXJO acfn neblle pilot «prav
dryer system.
Comnofwenlth Edison's Joliet Sta-
tlon. Tests begun nld-1979.
Operated at Northern States Power's
Sherbourne Station and PPaL's Jin
Brtdger Station.
lerrlal: The Coyote Station FCC Otter Tall rower Co. Coyote Sta-
systesj (&19 HW> Is under construct ten. The tlon at 8eul*h. W>. "Ittrt-up scheduled
process Is Rockvell's Open Loop kCT 'spray In the spring «f 19*1.
Mot Specified
Hot Specified
Lime. Soda Ash
Sodlun Carbonate
Mobil* unit is primarily for acajolrias; bid data, tf
Is also conduct tng vaste disposal s todies. KI/W ha*
2 blda under evaluation, 1 In-hove* and vxpccta 5 .-r
6 requests for bids by the end of this year.
Cnsnerrlil: ^-i.OOO acfm Induslrial s
for Cel*n«»»e Corr-. Sprav drver iwlse-t
baghruse ^vster.
Celanenc Cnrp, textile plant,
fuaberland. Mviryland. Start-up
ktheduled (or rarlv I9AO.
-------
Research
As mentioned above, initial test work by B&W at Basin Electricfs
William J. Neal station near Velva, North Dakota was with a Hitachi two-
fluid nozzle in a vertical reactor. When this configuration proved inade-
quate, B&W went to a modified "Y-jet" steam-atomized oil burner for slurry
atomization. Flue gas enters the reactor through registers or vanes, which
impart a spinning motion to the flue gas around the nozzle area. The
reactor itself is a box designed to give proper retention time. Some dried
sorbent drops into hoppers at the bottom of the horizontal reactor, while
the remainder is collected with fly ash in a precipitator or baghouse.
The pilot unit at Velva was rated at a nominal 8000 acfm gas flow.
Only ESP collection was tested with the "Y-jet" unit. The Hitachi reactor
had been tested with both ESP and baghouse collection, but B&W felt they
could safely operate with lower dryer outlet temperatures using an ESP col-
lector than with a baghouse collector. The increase in sorbent utilization
resulting from operating the spray dryer outlet nearer to saturation was
apparently greater than that corresponding to reaction on a bag surface.
Reagents tested included soda ash, pebble lime, hydrated lime, ammonia
addition with lime reagent, and precipitated limestone.
The pilot plant in Velva has been shut down and is currently being
dismantled. Research work is being continued with a 5 x 10 Btu/hr (1500
acfm) combustor using a single reactor with one nozzle. As well as doing full
parametric studies, B&W hopes to achieve an increased understanding of the
SO- removal reaction mechanisms in the spray dryer with this unit.
Commercial Status
Following the work at Neal Station, B&W successfully bid on the FGD
system for Basin Electricfs Laramie River Station Unit 3, a 500-MW unit to be
built near Wheatland, Wyoming. Engineering activities for the Laramie River
Station are on schedule with construction slated to begin soon. The unit
is scheduled to go into commercial operation in April, 1982.
The system at Laramie River will consist of four reactors each followed
by an electrostatic precipitator. Normal plant operation will call for the
use of three reactors, with one as a spare. Double louver dampers (1 set
isolation, 1 set control) regulate flow to the modules, with one set on each
reactor inlet and a downstream set on each ESP outlet. Air flow control to
individual "burners" in each reactor is set by vanes in the distribution
box. A first pass on the vane design for Laramie River has been model-
tested to give equal flows (within one percent) to the burners.
Each reactor will be equipped with 12 "Y-Jet" nozzles in three rows
of four nozzles. Reactor size was chosen to correspond to the size of the
ESP used. B&W precipitators of the same size and design as those on the
existing Laramie River Units 1 and 2 will be used. Units 1 and 2 use wet
limestone FGD systems.
68
-------
The turndown ratio for individual "burners" or "Y-jet" nozzles on the
Laramie River design is about 2:1 with the fixed registers and relatively
low nozzle AP employed. By using variable registers and higher design AP,
a greater individual turndown ratio could have been achieved. The nozzles
to be used at Laramie River will weigh approximately 64 pounds and will take
one to two men from 2 to 5 minutes to change.
The Laramie River design employs no recycle, even though the coal
contains a theoretical A.5:1 Ca/S ratio. The design calls for 90 percent
S0_ removal at maximum fuel sulfur and flow rate, with a maximum stoichio-
metry of 1.12 based on entering sulfur. The gases leaving the spray dryer will
be 10°F above saturation; 3 percent hot gas bypass will raise the dryer outlet
temperature 15°F before gases enter ESP. The Laramie River design calls for
30,000 Ib/hr of 150 psig steam with 50°F superheat for atomization. The capital
cost for this system was reported to be $49 million in a paper presented by Basin
Electric at EPA's Fifth Flue Gas Desulfurization Symposium, March, 1979 (Rei. 11).
A flow diagram of the Laramie River dry FGD system is shown in Figure 5-1.
Although a demonstration and not strictly a commercial FGD system, a
120,000-acfm reactor (about 40 MWe) is scheduled for start up at Pacific Power and
Light's Jim Bridget Station in late 1979. Tests have been delayed due to a
boiler outage. The reactor will treat a slipstream of flue gas from a 600-MW
boiler fired with a low sulfur Wyoming coal (about 450 ppm S02> and return the
treated gas to one of six ESPs. Slipstreams from this reactor will be
treated by a nominal 4000 to 5000 acfm baghouse and a 5000 to 6000 acfra
pilot ESP. This demonstration will use one six-nozzle reactor with the same
automatic control system proposed in the Laramie River design. It will
provide a check on the control logic for Laramie River. The system will have
no hot gas by-pass, as Laramie River will have. Baseline parametric studies
will be with lime reagent, but the utility is interested in testing ammonia
addition, soda ash, and an available waste sodium-based liquor. Basin Elec-
tric is also interested in comparing the results of the pilot ESP testing on
the spray dryer outlet with some original pilot ESP data used for the design of
the full-scale precipitators for the Laramie River Station. B&VJ will operate the
spray dryer and pilot ESP for 4 to 6 months for their baseline testing, but can
extend its operation if the utility funds additional testing. During base-
line testing, the target S0_ removal is the 85 percent required at Laramie
River. ^
Testing is proceeding on a 1500-acfm pilot unit at B&W's Alliance, Ohio
research facility. The unit, intended for parametric studies, uses a single
Y-jet nozzle in the reactor, with a cyclone for particulate collection. Plans
call for the addition of an ESP and a baghouse. Initial plans call for
several coals to be tested and Hue gas SO- concentrations of 500 ppm to
2000 or perhaps 3000 ppm. Lime reagent will be used, and test parameters
include inlet and outlet gas temperatures, sorbent stoichiometry, and
recycle schemes. A paste slaker will be used for fresh limes. Tests on low
sulfur lignite (0.5%S), low sulfur sub-bituminous (0.9XS), and low sulfur
bituminous (2.2%S) coals have been completed. Detailed results of romplot oil
test work are not available at this time, although an internal report IH
being prepared.
69
-------
1 FROM BOILER
REACTANT
HOLDING
TANK
MILL PRODUCT
TANK
TO DISPOSAL
Figure 5-1. Laramie River Station flow diagram.
-------
Earlier research work indicated that recycled flyash/spent sorbent mix-
tures were more reactive when mechanically ground to smaller particle size
than when reslurried directly. Consequently, some recycle work will be done
with the ball mill slaker from the now disassembled Velva pilot plant used
to reduce the particle size of recycled material.
One additional anticipated result of this test program is that B&W
expects to derive a better understanding of the S0~ reaction mechanisms In
the spray dryer. B&W feels that neither gas-liquid reaction where absorbed
gas phase SIX reacts with alkaline droplets, nor gas-solid reaction where
flue gas S02 adsorbs onto dried alkaline solids can account for all of the
S0_ removal. Currently, they expect to find that chemisorption of S02 across
several molecular layers of residual moisture on spray dried alkaline mater-
ial will account for a major portion of the S02 removal.
Continuing work will focus on validating the preliminary reaction mech-
anisms that have been formulated to date. Immediate plans arp to burn oil and
spike the flue gas with SCL, with a possibility of testing high sulfur fuels in
the near future. B&W also plans to add a baghouse downstream of the spray dryer
to compare SO2 and particulate removal efficiencies to those obtained with
the spray dryer/cyclone configuration. Longer range plans Include burning
different fuels and the investigation of the S(>2 removal capabilities of
virgin flyash in a spray dryer system.
B&W has submitted 5 bids, all currently under evaluation with awards
expected in the next few months. Also, 4 or 5 bids are being prepared for sub-
mission by the end of the year. B&W feels that the economic considerations
involved in selecting FGD systems for low sulfur Western coal applications
favor dry systems such as theirs.
Information Sources
Blythe, Gary. Meeting notes at B&W, Barberton, Ohio, June 28, 1979.
Janssen, K.E. and R.L. Eriksen. "Basin Electric's Involvement with
Dry Flue Gas Desulfulzation," paper presented at Fifth EPA Symposium on FGD,
Las Vegas, Nevada, March 5-8, 1979.
Kelly, M. E. Telephone conversation with John Doyle, B&W Alliance
Labs, October 11, 1979.
Kelly, M. E. Telephone conversation with Tom Hurst, B&W, October
18, 1979.
Slack, A. V. "Lime Scrubbing by 'Dry Processes'." A. V. Slack Report
#62, January 1979, pp. 15-30.
71
-------
5.2 BUELL/ANHYDRO AND EPA/BUELL
Address: Buell Emission Control Division
Envirotech Corporation
200 North Seventh Street
Lebanon, Pennsylvania 17042
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
Contact for More Information;
Dale A. Fulong
Research Projects - Senior Scientist, Research and
Development Department
(717) 272-2001
T. G. Brna
EPA Project Officer
(919) 541-2683
Commercial Activities - Mr. Lloyd Hemenway
Buell Emission Control Division
(717) 272-2001
Background
Buell, a division of Envirotech Corporation, has two dry FGD efforts
underway. One is a totally in-house development of a dry injection/baghouse
collection system which would use a Buell bag collector. In the other, Buell
is working with Anhydro, a Copenhagen-based spray dryer company which began
as a spin-off from Niro Atomizer, to develop a spray dryer/baghouse system.
Test work on both systems will be conducted at the City of Colorado Springs'
Martin Drake Station.
Research
EPA-funded test work on a 3000 acfm dry injection/baghouse system at
the City of Colorado Springs Martin Drake Station was due to start in late
October. The test work will investigate three sodium-based sorbents:
nahcolite, raw trona ore, and upgraded trona. The parametric testing, to be
conducted through December 1979, will include Investigation of various
stochiometric ratios (0.7, 1.0 and 1.5), inlet gas temperatures (325, 425
and 500°F), and inlet S02 concentrations (400 to 600 ppm). The fuel used
in the Martin Drake unit is a Colorado bituminous coal with a higher heating
value of about 12000 BTU/lb and a sulfur content of from 0.3 to 0.7%. The
nahcolite to be used in the tests has been obtained from a Bureau of Mines
pilot shaft sunk in a Colorado nahcolite/oil shale deposit near Denver. The
nahcolite from this mine will be made available for government and industry
dry FGD test work. This pilot shaft is currently one of the few available
sources of nahcolite. A final report on this dry injection work is expected in
early 1980.
72
-------
Buell is also conducting EPA-funded waste disposal studies. The major
process to be studied is sintering of the dry waste product, based on the
sinterna process (patented by Industrial Resources, Inc.). The waste
disposal studies are expected to last for about nine months and are being
performed by Battelle Memorial Institute (Columbus Laboratories) under a
subcontract from Buell.
In a parallel privately funded research program, Buell and Anhydro are
developing a dry FGD system with a spray dryer and baghouse. Initial work
with both lime and soda ash sorbents, was done at Anydro's Copenhagen facil-
ity with a 3000-acfm spray dryer. Future plans are to conduct EPA-funded
demonstration studies on a 20000 acfm spray dryer/baghouse system at the
Martin Drake Station. The spray dryer, supplied by Anhydro under an exclusive
technological agreement, is a 13-ft diameter tower equipped with a rotary
atomizer. Sorbents to be tested Include lime and the three sodium-based
sorbents used in the dry injection studies. Recycling of the spent
sorbent-flyash products will also be investigated. The spray dryer/baghouse
system is expected to start up in December 1979, and tests will be run for
6 months to a year. The primary purpose of the unit will be to obtain
design data for development of a full-scale spray dryer system.
Buell/Anhydro has submitted a detailed technical paper on their spray
dryer/baghouse system that will appear in the December 1979 issue of the
Air Pollution Control Association Journal. Buell also plans to set up a scale
model of a 20000 acfm lime-based spray dryer/baghouse system at the American
Chemical Society Chemical Show in New York, December 3-5, 1979.
Commercial Status
At this point Buell's top priority is the completion of test work on
the spray dryer and dry injection systems at Martin Drake. Buell has invited
several architect-engineer (A-E) firms to visit the facility in late January
and February of 1980. Buell plans to rely heavily on presentations to A-E's
and utilities in marketing their spray dryer system.
The Buell/Anhydro joint venture has submitted five budgetary prices to
utilities for spray drying systems. They have also submitted a budgetary
price for one industrial application and bid on another industrial applica-
tion.
Information Sources
Blythe, Gary. Meeting Notes at Buell, Lebanon, PA, June 29, 1979.
Kelly, Mary E. Telephone conversation with Dale Furlong, Buell
Envirotech R & D, October 11, 1979.
Kelly, Mary E. Telephone conversation with Lloyd Hemenway, Buell
Environtech, October 11, 1979.
73
-------
5.3 COMBUSTION ENGINEERING
Address; C-E Power Systems
Combustion Engineering, Inc.
1000 Prospect Hill Road
Windsor, Connecticut 06095
Contact For More Information; Kal W. Malki
System Design
Environmental Systems Division
(203) 688-1911
Background
Combustion Engineering has conducted several in-house studies related to
dry FGD. In 1972 they studied sludge drying as a dewatering method, and they
have an ongoing program studying char-ash drying in the C-E coal gasification
process. C-E has also studied an ammonium sulfate dry SO- scrubbing process
available from a licensor.
Regarding a lime-based dry FGD system, CE has gone through a literature
survey and a study of available pilot data. They have tested several atomiz-
ing devices, presumably for use in a spray dryer, and have completed a con-
ceptual design of a lime-based dry FGD system.
Research
As a result of the above, C-E has begun test work on a pilot unit con-
sisting of a 20,000 acfm C-E spray dryer equipped with a two-fluid nozzle
atomizer using air as the atomizing fluid. The spray dryer is followed by a
fabric filter and ESP in parallel. The pilot unit will be installed at the
Northern States Power Sherbourne County Unit //I, which burns Sarpy Creek
(Montana) coal (1.0 percent sulfur, 10 percent ash, 8000 Btu/lb). The pilot
unit will use lime as a sorbent. Test work will involve parametric studies
of temperatures, air-to-cloth ratios, L/G ratios, S02 levels, scrubber
velocity, and recycle of spent sorbent/fly ash mixtures. In conjunction with
this pilot program, C-E is also planning a dry scrubbing waste disposal study.
In the future C-E plans to continue pilot plant testing of their
spray dryer system although no definite plans were disclosed at this time.
C-E is currently preparing bids for two dry FGD systems for utility applica-
tions.
Information Sources
Blythe, Gary. Telephone conversation with K. W. Malki, C-E, June 6, 1979.
Malki, K. W. C-E System Design, personal communication with Gary
Blythe, July 11, 1979.
Kelly, M. E. Telephone conversation with Kal Malki, C-E System Design,
October 11, 1979.
74
-------
5.4 DOE/GRAND FORKS ENERGY TECHNOLOGY CENTER
Address: Grand Forks Energy Technology Center
Grand Forks, N.D. 58201
Contact For More Information; Harvey Ness
Grand Forks Energy Technology Center
(701) 795-8000
Background
Grand Forks Energy Technology Center (GFETC) has been conducting
research on dry injection systems on the 200 scfm scale, comparing nahcolite
and trona sorbents. Parameters being investigated include inlet S0? concen-
tration, inlet gas temperature, bag materials, air-to-cloth ratios,'and the
sequencing of sorbent addition and bag cleaning cycles. SO. removal effi-
ciencies of up to 90 percent have been achieved with nahcolite at 50 to 60
percent sorbent utilization, with the lower sorbent utilization observed at
a high air-to-cloth ratio.
Research
Dry injection test work on the 200 scfm scale is expected to be com-
pleted in early 1980. Tests to optimize performance with trona have been
completed. Test work to optimize performance with nahcolite is proceeding
after some delay in obtaining nahcolite. (A Bureau of Mines pilot shaft near
Denver has provided a new source of nahcolite.) SO, levels representative
of low-sulfur Western coals, up to 1500 ppm SO. on a dry basis, are being
investigated in the current test work. A final report Is expected by mid 1980.
Research conducted at GFETC has indicated that trona is somewhat less
reactive than nahcolite as an S0« sorbent. This diminished reactivity can be
explained in terms of specific surface area (m /g). At equivalent conditions
of temperature and time, GFETC studies have shown that nahcolite has a
significantly larger surface area (7.0 m /g vs 5.0 m /g at 500°F and 30
minutes activation time). The difference is greatest at high temperatures
and long activation times. The GFETC investigators suggest that it is possi-
ble that the use of a sorbent with a smaller particle size, injection at
elevated temperatures, and operation of the baghouse at the highest practical
temperatures would overcome the apparent lower reactivity of trona enough to
make it "a practical alternative to nahcolite" for use in dry injection/
baghouse systems.
GFETC is planning to expand their current dry injection program in the
near future. They are presently designing a 150 scfm baghouse that will be
dedicated to dry injection testa (the current baghouse Is also used for
partlculate characterization studies). The new baghouse will be a pulse-Jet
type. Plans are to test both nahcolite and trona sorbents while varying
parameters such as temperature, residence time (air-to-cloth ratio), and
75
-------
stoichiometry. Higher SO- levels, above 1500 ppm, may also be investigated.
Bench scale waste disposal studies have been completed and a final report is
being prepared.
Future Research Plans
In addition to dry injection work, Grand Forks will be involved with
the Rockwell-Wheelabrator Frye spray dryer/baghouse FGD system being con-
structed at Otter Tail Power's Coyote Station. GFETC will sample for
particulate and SO. removal efficiencies. In addition, they are currently
negotiating with Otter Tail Power to determine the ranges in which GFTEC will
be allowed to vary such parameters as temperature and sorbent feed rate to
characterize the FGD system. The Coyote station FGD system is scheduled to
start up in the spring of 1981. As a supplement to the proposed test work at
Coyote Station, Grand Forks is conducting conceptual and small scale laboratory
studies to investigate large-scale recovery of sodium from the spent sorbent/
fly ash mixture for reinjection into the system. The advantage of sodium
recovery is seen as two-fold: (1) it will help to stabilize the waste
products by reducing their soluble sodium content, and (2) reinjection of
recovered sodium will reduce fresh sorbent consumption, which would provide
considerable savings in operating costs.
As a final note on dry injection, there remain uncertainties in both
trona and nahcolite availability. In the case of trona, depletion allowance
complications may keep trona from being available in the quantities needed
for full-scale dry injection systems. The new Bureau of Mines pilot shaft is
currently the only available source of nahcolite.
Information Sources
Ness, H. M., Stanley Selle, DOE /GFETC and Oscar Manz, University of
North Dakota. Power Plant Flue Gas Desulfurization for Low-Rank Western
Coals, presented at the 1979 Lignite Symposium, Grand Forks, ND, May 30-31, 1979.
Blythe, Gary, Telephone conversation with Stanley J. Selle, DOE,
June 7, 1979.
Blythe, Gary. Telephone conversation with Harvey Ness, DOE, July 3,
1979.
Kelly, M. £. Telephone conversation with Harvey Ness, DOE, October 16,
1979.
76
-------
5.5 DOE/MORGANTOWN ENERGY TECHNOLOGY CENTER
Address; Morgantown Energy Technology Center
Morgantown, West Virginia 26505
Contact For More Information: Dr. Charter Steinspring
Exploratory Research Group
(304) 599-7546
Background
The use of powdered dry lime or limestone in dry injection/baghouse
systems has resulted in much lower removal efficiencies than in systems where
sodium-based sorbents, such as nahcollte and trona, have been used. However,
the waste solids from sodium alkali-based processes are water soluble and
pose potential disposal problems, whereas calcium-based product solids are
considerably more stable. Studies have been underway at Morgantown Energy
Technology Center (METC) to develop a "dry" limestone FGD process.
Research
A patented technique, involving the addition of water vapor to hot
flue gas (300°F) to increase the saturation temperature of the gas above a
critical minimum before it is passed over a bed of limestone chips to remove
the S02» was developed by Shale and Cross in 1976. Both laboratory kinetic
studies and bulk evaluation studies have been conducted. Figure 5-2 is a
flowsheet for the bulk evaluation tests for this "modified dry" limestone
process (MDLP).
Results of the kinetic studies show reaction rates equivalent to those
found in high temperature fluid bed processes Involving calcined limestone.
Other results of the kinetic studies showed that reaction occurred on the
limestone chip surf ace and that calcium sulfate was the major product.
In the bulk evaluation studies, simulated dry flue gas was heated to
280°F and passed through the saturator prior to entry into the bed of crushed
limestone. The gas entered the bed at a temperature of 150° to 160°F and a
space velocity of 500/hr. Tests were conducted using limestone beds 1-inch
in diameter and 9 inches deep and Jj-inch diameter by 4 inches deep. Limestone
chips were 1/16-inch by Jg-inch. Figure 5-3 shows results of bulk evaluation
studies with 1600 ppm inlet SO, in the gas at 150°F. At a saturation
temperature of 150°F, a S02 removal efficiency of greater than 90% was main-
tained for over 3 hours. At lower saturation temperatures (i.e., 100° to
110°F), removal efficiency decreased rapidly with time due to the formation
of a CaSO-j/CaSO^ layer on the pellet after an initial period of high removal.
Experiments at higher space velocity (4000/hr) have shown that SO-
removal efficiencies of greater than 90% can be achieved when the water vapor
saturation temperature of the gas, controlled by the addition of water in the
saturator, is no more than 30°F below the actual gas temperature entering
the limestone bed. Sorbent utilization is reported to be "less than 90%".
77
-------
GAS
SAMPLE
GAS
SAMPLE
1
HOOD
LIMESTONE
REACTOR
PR - Pressure Recorder
TI - Temperature Indicator
70-ISW-1
Figure 5-2 Flow sheet for bulk evaluation studies of
modified dry limestone process
-------
100
SO2 CONCENTRATION, 1,600 ppm
CRUSHED QREER LIMESTONE, 1/16" x 1/4"
SPACE VELOCITY, 500 v/v/hr.
GAS TEMPERATURE, 150° F
• SATURATED 100° F
£ SATURATED (•) 110° F, (b) 120° F
O SATURATED 150° F
4 5
TIME, HOURS
Figure 5-3 Effect of moisture on S02 removal in a
fixed limestone bed*
70-1591-1
79
-------
Current Status
Mr. Shale retired from METC in September, 1979. At the present time
no further work is being carried out on the "modified dry" limestone process,
although studies to investigate the limestone/SOj reaction mechanism are
being conducted.
The data from Mr. Shale's test work has been verified for accuracy, but
a complete interpretation of the results has not been conducted. Preliminary
economic analyses, based on both the kinetic and bulk evaluation studies
employing a counterflow moving-bed reactor, have shown the capital and operat-
ing costs for the MDLP to be greater than those for a conventional wet lime/
limestone process. The excess cost is due in part to the large pressure drop
characteristics of counter-flow moving-beds.
Information Sources
Shale, C. C. and G. W. Stewart, "A New Technique for Dry Removal of
SO", DOE/Morgantown Energy Technology Center, paper presented at Second
Symposium on the Transfer and Utilization of Particulate Control Technology.
Denver, CO, July, 1979.
Kelly, M.E. Telephone conversation with Charter Steinspring, DOE/METC,
October 18, 1979.
5.6 DOE/PITTSBURGH ENERGY TECHNOLOGY CENTER
Address Pittsburgh Energy Technology Center
4800 Forbes Road
Pittsburgh, Pennsylvania 15213
Contact For More information; Mr. Richard Dempski
Pittsburgh Energy Technology Center
(412) 675-5730
Background
DOE's Pittsburgh Energy Technology Center (PETC) has had a program
underway for several months to evaluate dry injection of various sorbents.
They have a 500 Ib/hr coal furnace that provides the flue gas for their
testing.
Research
PETC's initial testing has focused on moderate sulfur Pittsburgh seam
coals (1.5-2%S). System parameters that were evaluated include sorbent type,
stoichiometry, sorbent feed mechanism, and sorbent particle size. Four
sorbents are being tested: sodium carbonate, sodium bicarbonate, nahcolite,
and trona. Data from this test program is still under evaluation, but pre-
liminary results have shown greater than 90 percent SO- removal with sorbent
80
-------
stoichiometries of up to 2. Pressure drop through the baghouse was
reported to vary from about 10-14 inches of water.
Future Research
Future research plans are to continue evaluation of the system
for S02 removal from higher sulfur coals. PETC also has requested
bids for the construction of a spray dryer which they hop to
have installed by mid-1980. Testing of the spray dryer system will begin
the second half of 1980.
Information Sources
Dickerman, J. C. Telephone conversation with Richard Dempski, PETC
January 2, 1980.
5.7 ECOLAIRE
Address; Ecolaire Systems, Inc.
Two Country View Road
Great Valley Corporate Center
Malvern, PA 19355
Contact For More Information; Carl Newman
Vice-President of Engineering
(215) 648-8600
Terry McRae
Senior Vice President
(415) 676-6315
Background
Ecolaire Systems is part of the Ecolaire Corporation which was founded
in 1971. The Ecolaire Corporation includes several companies which have
been in the business of supplying equipment to the utility industry
for many years. These companies include the following:
Research
Ecolaire Systems has had no previous dry FGD research and develop-
ment program other than Industrial Clean Air (a subsidiary of Ecolaire
Corporation) experience with dry additives for enhancement of bag filter
performance. However, Ecolaire has constructed a 10,000 cfm mobile pilot
plant for demonstration of a spray dryer/baghouse FGD system. When
completed, and mobile plant can be trucked to any location on six
semitrailers and erected in a 60-ft by 60-ft area in approximately a two-
veek hookup time. The unit under construction uses an Ecolaire-modified
Niro atomizer spray dryer and an ICA four-section baghouse using 12-in.
diameter by 36-ft long bags. Design data for system bidding will come
from the mobile demonstration unit, illustrated in Figure 5-4,
81
-------
70-1592-1
Figure 5-4 Ecolaire's mobile demonstration dry FGD unit.
82
-------
The mobile unit has been set up at Nebraska Public Power District's
Gerrel Gentleman Unit 1. System start-up is scheduled for late 1979.
Ecolaire plans to conduct parametric and design optimization studies for up
to 4 months. They will be investigating both rotary and nozzle atomization.
The primary sorbent to be tested is lime and provisions have been made for
investigating recycle of spent sorbent/fly ash mixtures. Other parameters
that will be varied include inlet SO,
inlet gas temperature, temperature dl
concentration.
)2 concentration (up to 2000 ppm S02),
Irop over the spray dryer, and sorbent
Commercial Status
Ecolaire Systems will not bid on a utility or industrial system until
they have design data available from their mobile demonstration unit.
Current thinking for a commercial system would call for four or five spray
dryer modules, each with five rotary or nozzle atomizers. Primary control on
the system would be outlet gas temperature, which sets the water rate to the
scrubber. Sorbent concentration would be controlled based on Inlet and out-
let SO, concentrations, boiler load, etc. Perhaps a minicomputer would be
used to calculate required sorbent concentration based on these various in-
puts. The control system would be designed by Ecolaire Systems, using
purchased components.
Although the mobile demonstration unit uses a modified Niro spray dryer,
Ecolaire has no agreement with a spray dryer manufacturer for exclusive use
of their equipment. By not making direct ties with a spray dryer company,
Ecolaire feels they have more flexibility in providing the best system for an
individual application by choosing among any number of commercially offered
spray dryers.
Information Sources
Blythe, Gary. Meeting Notes, meeting at Ecolaire Systems, Inc.,
Malvern, PA., May 22, 1979.
Kelly, M.E. Telephone conversation with Carl Newman. Ecolaire.
November, 1979.
5.8 ENERGY AND POLLUTION CONTROLS, INC.
Address^ Energy and Pollution Controls, Inc.
Subsidiary of Flick-Reedy, Inc.
7N015 York Road
Bensenville, 111. 60106
Contact For More Information; Grant Hollett, Jr.
Energy and Pollution Controls
(312) 766-3400
Background
EPC is a recently formed subsidiary of Flick-Reedy, Inc. They have
developed a totally dry FGD reactor system for Industrial applications.
83
-------
Research
Development work on the EPC-designed reactor illustrated in Figure 5-5
was started in early 1978. This dry FGD reactor is intended for combination
with a baghouse to result in a totally dry FGD system for industrial applica-
tions. The reactor is designed for cyclonic flow of flue gas near a hydrau-
lically driven "slinger", which distributes dry hydrated lime in a direction
countercurrent to the flue gas flow. Hydrated lime is fed to the slinger
with a commercially available dry chemical feeder. Directly above the slinger
is an air-operated eductor which captures and recirculates the lighter
fraction of the partially spent sorbent. Below the reaction section of the
reactor, a conical expansion reduces the flue gas velocity to allow dropout
of heavy particulate matter before the gas flows to a bag collector.
Gas velocities in ductwork to and from the reactor are typically 8 to
12 ft/sec. Velocities inside the reactor vary from approximately 25 ft/sec,
in the cyclonic section down to approximately 5 ft/sec, in the expansion
section.
Development work on the reactor used flue gas from a coal-fired (1.6
million Btu/hr hot water boiler) burning a 3.3 percent sulfur Illionis coal.
Flue gas sulfur content varied from 1600 to 2300 ppm S02- Flue gas
temperatures in the reactor varied from 350°F to 500°F. but temperatures to
the downstream cartridge filtration device were limited to 350°F to protect
the NOMEX cartridge. The speed of the slinger was not specified, due to
patent considerations. Pressure drop across the reactor in all tests was
less than 0.5 inches of water. S02 removal efficiencies varied from 45
percent to 95 percent at 0.8 to 3.6 times the stoichiometric ratio. Table
5-2 summarizes the SO, removal results of dry reactor test work.
RECIRCULATOR
SLINGER
HYDRAULIC
MOTOR
EXPANSION
SECTION
Figure 5-3. Air pollution control (SO:) reactor for dry reagent
-------
TABLE 5-2 DRY REACTOR CHRONOLOGICAL TEST SUMMARY
AUGUST 1978 - OCTOBER 1978
Stoichiometric
ratio
Material
Lime
Lime
Lime
Lime
Lime
Lime
Lime
Lime
Lime
Lime
(hydrated)
(hydrated)
(hydrated)
(hydrated)
(hydrated)
(hydrated)
(hydrated)
(hydrated)
(hydrated)
(hydrated)
(a)
(b)
(c)
(d)
(e)
3
3
2.5
0.8
2
1.5
I
1.5
1.5
1.5
Inlet SO.
Concentration
(ppm)
1600
2000
2000
2000
1800
2300
2000
1700
2000
2000
Outlet SO^
Concentration
(ppm)
400
180
400
1100
400
750
700
200
1100
100
Removal
efficiency
(Z)
75
91
80
45
78
67
65
88
45
95
Notes:
a. Half was added at point approximately 5 ft upstream of reactor.
b. Half was added at point approximately 5 ft upstream of reactor. In-
spection revealed accumulative pile in duct at end of test.
c. After 10 minutes flow, efficiency rose to 95%.
d. Geometry of lower section of reactor was modified.
e. Slinger speed (RPM) varied from other tests.
Commercial Status
EPC will market the reactor for Industrial applications. They report
that capital and operating costs for SO. and particulate control using their
dry system will be approximately half those for a system using a wet scrubber,
on a dollars per ton of coal basis.
Preliminary design has been completed for construction of a commercial
model (25,000 acfm) for Kerr Industries in South Carolina. This unit will
be used in EPA sponsored tests at Kerr to be performed by Environmental
Testing, Inc. The tests are scheduled to begin March, 1980.
85
-------
Information Sources
Blythe, Gary. Telephone conversation with Grant Hollett, Jr., EPC,
June 4, 1979.
Hollett, Grant T., Jr. "Dry Removal of S02 Applications to Industrial
Coal Fired Boilers." Presented at APCA Convention, Cincinnati, Ohio,
June 25-28, 1979.
Kelly, M.E. Telephone conversations with Grant Hollett, Jr., EPC,
October 5 and October 26, 1979.
5.9 EPA/BATTELLE-COLUMBUS LABS
Addresses; Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Battelle-Columbus Labs
405 King Avenue
Columbus, OH 43201
Contact For More Information; J. H. Wasser
EPA, Project Officer
(919) 541-2476
Robert Giammar
Battelle
(614) 424-7701
Background Information;
As part of an EPA-funded program to evaluate industrial coal-fired
stoker boilers, Battelle developed a limestone/high sulfur coal pellet in an
attempt to control S0_ emissions. Initial tests were conducted in a model
20 bhp spreader-stoker boiler using a pellet with a Ca;S mole ratio of 7:1.
Preliminary results indicated that 70 to 80% retention of the available sulfur
in the fuel was achievable, however, the increased particulate emissions
resulted in an opacity increase from a normal 4 to 7 percent up to 22-25
percent. In an effort to reduce the increased particulate emissions that
result from firing the limestone/coal pellet, Battelle began development and
testing of a 3.5:1 Ca:S pellet.
Research
The newly developed pellet (Ca:S ratio of 3.5:1) has mechanical strength,
durability, and wea durability characteristics comparable to that of raw coal.
Laboratory tests in the 20 bhp model spreader stoker and in a fixed-bed reactor
have resulted in 60 to 80 percent retention of the available sulfure in the
fuel. A high sulfur Illinois coal was used for these tests. Laboratory tests
are continuing in preparation for a 1-day test to be conducted in late 1979 to
86
-------
evaluate firing the pellet in Battelle's 25000 Ib./hr steam boiler at the
Columbus lab facilities. The 1-day test will investigate combustion charac-
teristics and sulfur retention capabilities of the pellet with a Ca:S mole
ratio of 3.5:1 as well as determine the effects of the pellets on boiler
operation. With regard to boiler operation, one preliminary hypothesis is
that the limestone present in the pellet may result in higher ash fusion
temperatures, thus helping to reduce the clinker formation tendency of the
coal.
In other on-going work, bench-scale waste disposal studies conducted by
Battelle have shown that the ash produced from combustion of the coal/lime-
stone pellet does not exhibit the same acid-leaching characteristics as fly
ash from straight coal combustion. Battelle is also beginning reaction
mechanism studies to determine the details of the SO, removal reaction In
the stoker bed. Plans are to use scanning electron microscope techniques to
determine what compounds are being formed and to identify their structures.
Preliminary cost estimates have been prepared and have been confirmed
in further economic studies that Indicate the cost of producing the coal/
limestone pellet will be about $15/ton. This cost includes both capital
and operating costs for grinding and pelletizing and reagent costs for the
binder and limestone.
Future Research
Future work including a 30-day full scale demonstration test on a
75,000 Ib steam/hr boiler is currently under funding evaluation by the EPA.
The results of the present study will be evaluated to determine their poten-
tial impact on the nation's energy supply and pollution standards.
Information Sources
Giammar, Robert D., et al, "Evaluation of Emissions and Control Techno-
logy for Industrial Stoker Boilers," in Proceedings of the Third Stationary
Source Combustion Symposium, Vol. 1, EPA-600/7-79-050a, February, 1979.
Kelly, M. E. Telephone conversation with Robert Giammar, Battelle-
Columbus labs, October 10, 1979.
Kelly, M. E., Telephone conversation with J. H. Wasser, EPA, October 11,
1979.
5.10 EPA/ENERGY AND ENVIRONMENTAL RESEARCH CORP. (EERC)
Address; Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, N.C. 27711
Contact For More Information! Blair Martin
EPA Project Director
(919) 541-2235
87
-------
Background
EERC conducted a preliminary feasibility study as part of a program to
test EPA's concept of limestone injection,into a low-NO burner for S0_
removal. Initial test work at the 1 x 10 Btu/hr heat input rate was
promising, and work is continuing on a smaller (70,000 Btu/hr) scale.
Research
In this process, limestone is mixed and pulverized with coal prior
to combustion and fired through a B&W low-NO burner. Sulfur contained
in the fuel reacts with the limestone present to form calcium salts which
are collected with the fly ash emitted from the boiler.
The B&W low-NO burner system consists of two physically isolated
combustion zones: a £uel-rich water-cooled primary combustion furnace and
a secondary furnace where combustion products from the first furnace mix with
the air necessary to complete combustion. EERC has measured the retention of
sulfur resulting when limestone or trona is mixed with coal prior to com-
bustion in the burner. The effectiveness of alkali addition to this type of
burner is apparently related to the lower flame temperature resulting from
the two-stage combustion as compared to conventional combustion. The lower
tempe-at-un keeps the particles from approaching their melting temperatures,
which can result in a glazing of the surface of the reagent particles that
produce relatively unreactive particles.
EERC has noted that sulfur retention effectiveness is dependent upon
good mixing of the reagent with the coal. They have found that adding
reagent to the coal prior to passing the coal through the pulverizer
substantially improves SCL removal. The pulverizer, which is designed for
75 percent minus 200 mesh, is believed to promote good mixing between the
reagent and coal. EERC has also noted that sulfur retention can be greatly
influenced by combustion conditions in general.
The test work so far has been very preliminary in nature. Table 5-3
indicates the S02 removal which has been achieved.
Future Research
As a result of these Initially favorable performance data, EERC has
proposed a full parametric study of sulfur retention in three sequential
combustor sizes: 70 thousand Btu/hr, 10 million Btu/hr, and 50 million
Btu/hr. These proposals are currently under review by EPA. There have
apparently been no cost estimates yet made for this process due to its early
stages of development.
88
-------
TABLE 5-3. S02 REMOVAL IN THE EERC LOW-NOX COAL BURNER
Basis: Utah Low Sulfur Coal
Reagent/Sulfur Percent
Reagent Mole Ratio ?_9J3£yed
Limestone 1 53
2 73
3 88
Trona 2 41
4 80
89
-------
Information Sources
Blythe, Gary. Telephone conversation with Bill Nurick, EERC, July 25,
1979.
Jones, D. J. Personal communication with Ted Phillips, Pacific Power
& Light, March 1979.
Kelly, M. E. Telephone conversation with Blair Martin, EPA, October
1, 1979.
5.11 EPA/KERR INDUSTRIES
Address: Industrial Environmental Research Labs
Particulate Technology Branch
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
Contact For More Information; Jim Turner
EPA Project Officer
(919) 541-2925
Background
EPA is funding dry FGD test work at a Kerr Industries textile finishing
plant in South Carolina. The major portion of the test work will focus on
Energy and Pollution Control Inc.'s "dry pollution control reactor", (see
Section 5.9).
Research
A baghouse fire has delayed testing of this sytem. Test work is
scheduled to begin in March, 1980, to evaluate dry FGD system which will
treat 25000 acfm of flue gas (about 4 MW). The baghouses (modified reverse
air and pulse-jet) will be operated at air-to-cloth ratios of from 3 to 6
and both high and low inlet SO. concentrations will be tested. Lime and
limestone sorbents will be used.
Information Sources;
Blythe, Gary. Telephone conversation with Jim Turner, EPA, June 25, 1979.
Kelly, M. E. Telephone conversation with Jim Turner, EPA, November 7
1979.
90
-------
5.12 JOY/NIRO JOINT VENTURE
Address : Joy Industrial Equipment Company
Western Precipitation Division
4565 Colorado Boulevard
Los Angeles, California 90039
Niro Atomizer, Inc.
9165 Rumsey Road
Columbia, Maryland 21045
Contact For More Information; Jim Meyler
~ "" Joy/Western Precipitation Division
(213) 240-2300
Steven M. Kaplan
Niro Atomizer, Inc.
(301) 997-8700
Background
Niro supplies some 75 percent of the world's spray dryer needs. Spray
dryer applications include: instant milk, instant coffee, pvc, floor tile
ceramics, kaolin, dyestuffs, copper and nickel sulfide for smelting applica-
tions, and raw cements. The kaolin plants are significant because they in-
volve large (30-ft diameter) dryers that approach the size of those in utility
FGD applications. The smelting operations are significant because some use
flue gases from coal-fired generating stations for process heating and thus
operations follow boiler load as would a utility FGD spray dryer. Many of the
t plant installations also use coal-fired flue gases lor process heating.
Niro began test work using a spray dryer for FGD and HC1 removal appli-
cations in their Copenhagen research facility in 1974. In this test work
they made several hundred test runs at 1000 to 3000 acfm using lime, lime-
stone, sodium carbonate, and magnesium oxide as sorbents. As a result of this
test work, Niro sold a small (2000 to 3000 acfm) spray dryer to Fiat of Milan,
Italy, which used sodium carbonate to remove SO^ from flue gas. In November
1977, Niro entered into a 2-year agreement with Joy Manufacturing to develop
and market a spray dryer-based FGD system which employs baghouse or ESP
participate collection. The agreement has since been extended another 5
years.
Research
The Joy/Niro joint venture team was invited by Basin Electric to pilot
a spray dryer-based dry FGD system at their Antelope Valley Station. In
mid-November 1977, design work was begun on a pilot unit to be built at Otter
Tail Power's Hoot Lake station. The pilot unit included a 20,000-acfm spray
dryer followed by a 3000-acfm electrostatic preclpitator and a 9000-acfm
baghouse in parallel. The baghouse had four compartments and 60 one-foot
91
-------
diameter bags. The pilot unit was constructed from late November to mid-
February, and first operated from mid-February to April 24, 1978. During
this period sodium carbonate and lime sorbents were tested, and design data
were acquired for use in preparing a bid for the Basin Electric Antelope
Valley Station FGD system. Parametric studies of the effects of inlet and
outlet temperatures, stoichiometric ratio, recycle techniques, and parti-
culate collector type on S0» removal were included in this test work. A
100-hr demonstration run of the pilot unit at Antelope Valley design condi-
tions was also included.
' Niro and Joy returned to Hoot Lake in mid-September 1978 to acquire
data for preparing a bid on Basin Electric's Laramie River Station and to
verify data from the previous test work. During this mid-September to mid-
December 1978 test period,some 30 to 40 thousand data points on FGD perform-
ance were acquired. Also, during this period rail cars of Laramie River coal
were shipped to Hoot Lake to operate the 60-MW Unit 2 boiler for several
days at Laramie River conditions. The Laramie River ash was reported to be
quite cementitious, but no operating problems were incurred during this
period. Some tests were conducted adding water treatment sludge to the
atomizer feed, and this was found to be a suitable method of disposal for the
sludge. Some testing was conducted with commercially available ground lime-
stone as a sorbent with very limited success. Other testing included SO-
spiking up to a 4500 ppm flue gas concentration.
All current FGD research activities are being done on a new FGD pilot
unit at the Niro Copenhagen research facilities. After substantial testing
at Hoot Lake, Niro feels that parametric, research-oriented testing can best
be done at their facility. Site specifics, such as fly ash alkalinity effects,
can be evaluated by fly ash injection. The 5000 acfm test facility uses a
propane burner as the source of flue gas. Both reverse air and pulse Jet
baghouses can be tested. Provisions have been made for spiking with fly ash,
S02, SO., and steam to stimulate any flue gas condition.
The primary use of the Copenhagen facility is to obtain site specific
information on SO,, removal and fly ash reactivity for responding to bids.
In addition, Niro is also looking at the chemistry and reaction mechanism of
the SO. sorption reaction. Results of this test work are for internal use
only, but they may use these results to prepare papers for presentation at
upcoming symposia.
Commercial Status
Following the pilot program at Fergus Falls, the Joy/Niro joint venture
successfully bid on the FGD system for Basin Electric's Antelope Valley
Station Unit 1, the first of two 440-MW stations to go in near Beulah,
North Dakota. The Western Precipitation Division of Joy Industrial Equipment
Company was awarded the contract with Niro Atomizer Company as prime sub-
contractor. Construction of the Antelope Valley FGD system is reported to
be on schedule with start-up planned for the spring of 1982. Figure 5-6
depicts the Antelope Valley dry FGD system.
92
-------
VO
U»
Figure 5-6. Antelope Valley Station gas cleaning system.
-------
The Antelope Valley installation will use five 46-ft diameter dryer
nodules, although any four modules could handle the total flue gas volume.
Each module will be equipped with one atomizer with a direct drive motor.
There will be one spare atomizer for use in any of the five modules. The
sorbent used will be primarily lime slaked in a ball mill slaker, although
sludge from the station's primary water treatment plant and a portion of the
recycled sorbent/ash mixture will be added.
The S02 removal guarantee with only Unit 1 operating will be 62 percent
for average sulfur fuel (0.68 percent S) and 78 percent for maximum sulfur
fuel (1.22 percent S). After Unit 2 comes on line over a year later, the
removal guarantee will be 81 percent for average sulfur, and 89 percent for
maximum sulfur. The emission limitation set by the North Dakota State De-
partment of Health is 0.78 Ib S02 per million Btu.
The sorbent/fly ash mixture leaving the spray dryer will be collected
in a Western Precipitation baghouse. The baghouse will have 28 compartments
with a total of approximately 8000 fluorocarbon-coated fiberglass bags.
The bags will be 12 inches by 35 feet and will be cleaned by reverse air.
The gross air-to-cloth ratio under maximum flue gas flow conditions will be
2.19:1. The bag life guarantee for the Antelope Valley system is proprietary,
but the standard Western Precipitation bag life guarantee is two years.
The lime utilization is very nearly 100 percent. This high utilization
is possible because of the ability to utilize available alkalinity in the
fly ash through spent sorbent/fly ash recycle. Also, the spray dryer outlet
temperature will be controlled to near saturation. Bypass of up to six per-
cent of the total flue gas flow will be used to reheat the spray dryer outlet
to 185°F.
The Antelope Valley FGD system will use an innovative supervisory con-
trol scheme designed by Honeywell. The computer-controlled system will monitor
boiler load, inlet SO level, outlet SO. level, and inlet and outlet spray
dryer temperatures and adjust lime concentration in the spray dryer feed
for minimum lime consumption required to maintain the desired SO,, removal.
Basin Electric reported the capital cost of this system to be over $49 million
in a paper presented at EPA's Fifth Flue Gas Desulfurization Symposium,
March 1979 (Reference 11).
Waste disposal is not Included in the Joy/Niro responsibilities, Basin
will return the waste product-fly ash mixture to the mine in coal trucks for
underground disposal.
To date, the Antelope Valley Station is the only FGD system sold by
the Joy/Niro venture. However, they currently have two bids in evaluation,
one bid in-house, and expect five to six bid requests by the end of 1979.
Although it cannot be strictly considered a commercial system, Joy/Niro
has been awarded a contract to provide a 100-MW demonstration spray dryer/
particulate collection system to Northern States Power for their Riverside
-------
Station. The lime-based system would treat up to 660,000 acfm of flue gas in
a single 46-ft diameter spray dryer equipped with a rotary atomizer. The
utility would operate the unit and participate in the demonstration R&D
effort with Joy/Niro.
The flue gas to be treated is generated by burning an Eastern coal of
up to three percent sulfur or a Western low sulfur coal or a mix of the two.
The spray dryer/ESP system (using an existing ESP) will be completed during
the summer of 1980, with a baghouse to be completed about six months later.
This arrangement will permit comparison of ESP and baghouse particulate
collection and SO. removal on a large scale unit.
Joy feels the market for utility spray drying systems is favorable.
They foresee responding to five or six bids by the end of 1979. They feel
the market for industrial systems will develop after promulagation of
industrial boiler New Source Performance Standards.
Information Sources
Elythe, Gary. Meeting Notes, Joy, Los Angeles, CA. June 14, 1979.
Janssen, Kent and Robert L. Eriksen. "Basin Electric's Involvement
with Dry Flue Gas Desulfurization," paper presented at the Fifth EPA
Symposium on Flue Gas Desulfurization, Las Vegas, Nevada, March 5-8, 1979.
Dickerman, J. C. Telephone conversation with Jim Meyler, Joy,
October 26, 1979.
5.13 KENNECOTT DEVELOPMENT COMPANY (ENVIRONMENTAL PRODUCTS DIVISION)
(formerly Carborundum)
Address: Environmental Products Division
Kennecott Development Company
P. 0. Box 87
Knoxville, TN 37901
Contact For More Information! Hank Majdesk!
~~~ Manager SO. Project
(615) 693-7550
Background
As Carborundum, the Environmental Products Division (EPD) conducted
pilot-scale dry injection studies (1000 acfm) at their University of Tennessee
facility -and pilot unit tests on a 15000 acfm spray dryer/baghouse system at
Basin Electric's Leland Olds Station.
Research
The Environmental Products Division is continuing pilot-scale test work
95
-------
(1000 acfm) on both dry injection and spray dryer SO control systems at their
University of Tennessee facility. The U.T. unit is a small stoker-fired boiler
burning a coal with an average sulfur content of 0.5% and a higher heating
value of about 10000 Btu/lb.
Dry injection tests have been run with both sodium bicarbonate and
nahcolite. Temperatures of the entering flue gas ranged between 270 and 300°F,
while entering S0? concentrations ranged between 1000 and 4000 ppm. (Some
test work has been done at the 5000 ppm SO- level.) In tests using relatively
non-alkaline fly ash as the sorbent, 25% S02 removal was obtained. Results
of the major portion of the test work are considered proprietary.
The Environmental Products Division is planning to start up a 1000 acfm
spray dryer/baghouse system in late 1979. Sorbents to be tested include lime
and soda ash. Recycle of spent sorbent/fly ash mixtures and removal
efficiency at high inlet S02 concentrations will also be investigated.
Commercial Status
The Environmental Products Division plans to emphasize both utility and
industrial applications in marketing their spray dryer-based flue gas cleaning
systems. They are in the process of executing an exclusive agreement with a
spray dryer company with a demonstrated background in both rotary and nozzle
atomization. EPD will most likely employ rotary atomization in their lime-
based spray dryer systems. As Carborundum, EPD had an agreement with DeLaval
for supplying the spray dryers, however, this agreement has been terminated.
The Environmental Products group recently submitted a bid to Colorado
Ute Electric Asso. for a 450 MW flue gas cleaning system. Colorado Ute Is
considering only dry systems for this unit. In regard to the future imirki't
for utility and industrial dry FGD systems, the Environmental Products Divi-
sion sees spray dryer-based FGD systems being, at the very least, competitive
with wet systems for both low and high sulfur coal applications.
Information Source
Elythe, Gary. Telephone conversation with Don Boyd, Carborundum,
May 16, 1979.
Blythe,. Gary. Telephone conversation with H. M. Majdeski, Carborundum,
July 6, 1979.
Majdeski, H. M. Carborundum, personal correspondence with Gary Blythe,
July 17, 1979.
Kelly, M. E. Telephone conversation with H. Majdeski, EPD, Kennecott
Development Co., October 18, 1979.
96
-------
5.14 KOCH ENGINEERING
Address; Koch Engineering Co., Inc.
161 East 42nd Street
New York, NY 10017
Contact For More Information; Ahmed Akacem
(212) 682-5755
Summary of Activities
Koch began development of a "spray dryer based dry FGD system" in early
1979. In May 1979 Koch was reportedly about one year away from having a
marketable dry FGD system.
Koch Engineering has completed studies on their spray dryer system
verifying published efficiency data. They have achieved SO- removals com-
parable to literature values. (Neither sorbent type or test conditions used
to achieve these removal levels were specified).
Koch plans to begin marketing their spray dryer/baghouse system in
February 1980. No details on the marketing program were available due to
the somewhat unique nature of the approach they plan to take.
Koch is presently conducting in-house demonstration and design studies
on'the 3000 to 5000 acfm scale. No test results were provided. They also
plan to conduct larger scale pilot tests on an 8-ft diameter spray tower.
Sorbents to be investigated Include sodium hydroxide, sodium bicarbonate and
lime. Initially testing will be conducted with inlet S02 levels representa-
tive of low-sulfur coal combustion, with plans for future testing at higher
S02 levels.
Information Sources
Blythe, Gary. Telephone conversation with Ahmed Akacem, Koch Engineer-
ing, May 16, 1976.
Kelly, M. E. Telephone conversation with Ahmed Akacem, Koch Engineering,
October 19, 1979.
5.15 MIKROPUL
Address: Mikropul Corporation
10 Chatham Road
Summit, New Jersey 07901
Contact For More Information; Tom Reinauer
Vice-President of Engineering
(201) 273-6360
97
-------
Background
Mikropul is a division of U.S. Filter, which includes Ducon, the wet
scrubber manufacturer. Mikropul was originally in the pulverizing business
and was called Pulverizing Machinery Company. Mikropul's air pollution
equipment business was a result of the necessity to control pulverizing
emissions. Their first equipment was a fabric blow ring collector dating
back to the AO's. A version of their current continuous pulse jet baghouse was
introduced in 1957. Since then nearly 200 installations around the world
have been put into service with Mikropul Pulsejet bag collectors for recovery
of the spray dried product. In the early 1970's, Mikropul began work in the
pure dry scrubbing business, designing and installing equipment using dry
alumina to control fluoride emissions from aluminum smelters. In these
applications, fluid bed feeders are used to introduce fresh and recycled
alumina to a reactor to contact fluoride-containing flue gases. Some 70
percent of the fluoride removal occurs in the reactor, the remaining 30
percent occurs on the downstream bag collectors. Pulse jet bags, rather
than reverse air or shaker bags are employed, partly because they provide for
more turbulent gas/solid contact in the vicinity of the bag.
Their largest installation at Ormet Aluminum in Hannibal, Ohio treats
2.6 x 10 acfm, using 144 pulse jet bag modules containing some 55,000 bags
and around 1000 pulse jet valves. Six reactors are used, and some reactant
captured on the bags is recycled to improve overall sorbent utilization. ,
Total Mikropul installed dry scrubbing equipment worldwide treats 12 x 10
acfm. Mikropul has also sold dry systems which use dry hydrated lime to
remove fluorides from a gas stream in a glass manufacturing facility.
Research
A sister company to Mikropul, Filtrol, a refining catalyst manufacturer,
has participated with Mikropul in identifying possible dry FGD sorbents for
evaluation in pilot-or bench-scale dry injection/baghouse collection
facilities. They have tested zeolites, which unfortunately selectively
remove water before S0_. Of the common dry FGD sorbents, Mikropul has only
tested sodium bicarbonate: seventy percent SO removal was obtained at flue
gas containing 1600 to 2000 ppm S02 at 300°F.
Mikropul began work on a spray dryer-based FGD system in January 1978.
Bayliss Industries supplied the spray dryer technology. Besides lime
studies, Mikropul has evaluated zinc oxide wastes as an FGD sorbent. Over 50
percent SO. removal with a ZnO slurry feed to a spray dryer system was obtained,
Mikropul's current test facility at Summit NJ includes a small nozzle
atomization spray dryer which discharges to a small pulse jet bag collector.
Flue gas is supplied by a gas burner. Provisions are available to spike the
flue gas in the 1000 acfm unit with S02, moisture, and fly ash from the Mercer
generating station. Another U.S. Filter subsidiary, Drew Chemical, is working
with Mikropul to provide lime slurry additives. Drew Chemical has provided
polyelectrolyte additives which reduce lime slurry settling tendencies and im-
prove atomization in a spray dryer by acting as surfactants.
98
-------
As the result of their spray dryer work, Mikropul was awarded a
contract in late 1978 to provide a spray dryer/baghouse on an industrial
coal-fired boiler at Strathmore Paper Company, a subsidiary of Hammer-mill, Inc.,
at Woronoco, Mass. Their agreement with Strathmore allows Mikropul to
vary numerous parameters during a six-month test period in order to establish
a data base for future designs. The system is lime-based and the single spray
dryer will use multi-fluid nozzles for atomization. A pulse jet baghouse
will be used. Provisions for solids recycle will exist. Parameters to be
varied include reagent type, sulfur levels, slurry density and additives,
slaking techniques, atomization techniques, and baghouse parameters. In
addition, other collection techniques may be investigated. Table 5-4
summarizes the parameters to be varied and their ranges.
Commercial Status
As mentioned above, Mikropul has sold a lime-based spray dryer/baghouso
collection FGD system to Strathmore Paper at Woronoco, MA for use on an
Industrial boiler. This PC boiler, which generates 90,000 Ib of 675 psin
steam per hour, produces 40,000 acfm at 350 to 400°F by burning 2 to 2^
percent sulfur, 9 percent ash coal. The plant must meet a Massachusetts air
quality regulation of 0.55 Ib S02 per 10° Btu. This fuel source can be
augmented by up to 30 percent (by Btu content) #6 fuel oil. Mikropul's
guarantee calls for 75 percent SO, removal on 3 percent sulfur coal, with a
maximum stoichiometry on the order of 2.75. Pilot studies indicate 75 percent
removal can be achieved at up to 2000 ppm SO- inlet concentration at a
stoichiometry of around 1.2. Sorbent utilization has been on the order of
70 percent. Flue gas is withdrawn downstream of the boiler air heater at
350 to 400°F, but a provision is available to bypass hot flue gas if
necessary.
The Mikropul system started up in September, 1979. Mikropul personnel
are operating the system on a 24-hr, 7-day per week basis. They have
experienced some operating problems but are in the process of correcting
these by instituting design changes. Performance results are not publicly
available at this time. Mikropul will continue operation of the system for
optimization and establishment of a design data base. Further results may
be available in early 1980.
The design details of the FGD system are considered proprietary by
Mikropul, but the system can be generally described as a lime sorbent spray
dryer followed by a Mikro-Pulseair pulse Jet bag collector. Spent reactant
and fly ash disposal will be handled by the paper company. Figure 5-7
illustrates a general Mikropul dry FGD system similar to that at Strathmore
paper.
99
-------
TABLE 5-4 SPRAY DRYER TEST VARIABLES
STRATHMORE PAPER COMPANY
Parameter
Range of variation
Sorbent
Fuel Sulfur content (%)
Slurry density
Recycle
Slurry additives
Air-to-cloth ratio
Materials of construction
Disposal techniques
Atomization technique
Collection technique
Slaking technique
High calcium granular limes;.
High calcium pebble lime .
Dolomitic lime.
Eastern Coal
2 to 2 1/2 % (nominal)
4% (maximum at Mikropul Expense)
Fuel Oil
0.7 to 2%
10 wt % solids desired, up to 20% as needed.
From baghouse product, amount to be determined.
Various polyelectrolytes, varied concentrations
(on the order of 2 g/10 gal H^O).
3.2:1 to perhaps 7:1, as success dictates.
Various metal coupons to be placed in dryer,
ducts, and baghouse for periodic examination.
Evaluate suitability as feed to Portland cement
plant.
Determine best landfill treatment methods
(e.g., treatment in pug mill to wet and compact).
Steam and air atomization at various rates
May evaluate ESP collection on a slipstream.
Most tests will use baghouse.
Evaluate effect of slaking techniques on
reactivity of dolomitic lime,.which has shown
lower reactivity than high Ca limes (appar-
ently due to poor reactivity of commercially
slaked MgO content)
too
-------
MIKRO-PULSA/RE
COLLECTORS
BAYLJSS-MIKRO
REACTOR
Figure 5-7. Mikropul spray dryer/baghouse dry FGD system.
-------
A third party is involved in the Strathmore Paper project: Lucien J.
Luckel, an engineering constructor with an extensive background in industrial
boilers, is providing the expertise for dealing with industrial boiler clients,
as well as designing all interfaces and tie-ins with the boiler itself.
Mikropul has no Immediate plans to enter the utility boiler market, but may
in the future depending on the success of the Strathmore Paper unit. If
Mikropul should go into the utility business, the third party providing the
interface capabilities with utility clients would be another U.S. Filter
company, Resource Scientists of Tulsa, which has utility engineering and
construction experience.
In a utility design, a reverse air rather than a pulse jet bag
collector may be required. The break-even point for the two types is about
200,000 acfm, (applications where the pulse jet can operate at an air-to-
cloth ratio of 3.4 times that of a reverse air unit). Mikropul has installed
reverse air units at a Southern Colorado Utilities power plant and on several
industrial boilers.
Information Source
Blythe, Gary. Meeting Notes, meeting at Mikropul, Summit, NJ,
May 24, 1979.
Blythe, Gary. Telephone conversation with Tom Reinauer, Mikropul
Corp., October 25, 1979.
5.16 RES EARCH-COTTRELL
Address; Research-Cottrell Research and Development
P.O. Box 1500
Somerville, New Jersey 08876
Contact for More Information: Kishor N. Parikh
Project Manager
(201) 685-4879
Background
Previous FGD work by Research-Cottrell has focused on wet lime/
limestone systems. The only previous purely dry FGD work done by a Research-
Cottrell company has been that done by their wholly-owned subsidiary, KVB.
A spray dryer/baghouse pilot system at the Texas Utilities Big Brown unit
is the first major dry FGD effort by the Reeearch-Cottrell R&D group. This
spray dryer/baghouse pilot dry FGD program will evaluate several sorbents.
Research-Cottrell has an exclusive agreement with Komline-Sanderson for
use of their spray dryer in a dry FGD system.
Research
Research-Cottrell has completed pilot unit studies with lime at the
Texas Utilities Big Brown unit (10,000 acfm). Details of the test work are
102
-------
considered proprietary at this time. A final report on the test work should
be completed by December 1979 and Research-Cottrell may present the results at
a seminar or conference. Research-Cottrell's wholly owned subsidiary, KVB,
has completed an Electric Power Research Institute (EPR1) funded study on dry
injection with the results scheduled for publication in the near future.
Commercial Status
Now that pilot unit testing is completed, Research-Cottrell (RC) is
ready to offer a commercial spray dryer/baghouse system. They are confident
of scaling up design data from 10000 acfm because they have scaled up wet
system designs for up to 3000 MW worth of FGD from 5000-10,000 acfm pilot
data. For a utility size installation, Research Cottrell anticipates using
multiple atomizers per dryer vessel. The designs of the reagent feed system
and the slaker, fan, dampers, and other overall system components will be
based on Research-Cottrell's wet FGD system experience. Research-Cottrell has
in-house nozzle atomlzation experience, and if a sodium system is ever required,
nozzle atomization would probably be used. Rotary atomizers are believed
superior in a calcium system because they are less susceptible to pluggage
and erosion than nozzles. Plans are to operate the spray dryer outlet at
about 50°F above adiabatic saturation to avoid flue gas buoyancy and collector
bag problems.
Information Source
Blythe, Gary. Meeting at Research-Cottrell, Sommerville, NJ,
May 23, 1979.
Kelly, M.E. Telephone conversation with Kishor Parikh, Research-
Cottrell, October 18, 1979.
•
5.17 ROCKWELL INTERNATIONAL/WHEELABRATOR-FRYE JOINT VENTURE
Address; Environmental & Energy Systems Division
Rockwell International
Energy Systems Group
8000 DeSoto Avenue
Canoga Park, CA 91304
Air Pollution Control
Wheelabrator-Frye, Inc.
14920 S. Main Street
Gardena, CA 90248
Contact for More Information; Dr. Dennis C. Gehri
~~~~ Project Manager
ACP Research, Development and Applications
Rockwell International
(213) 341-1000
103
-------
Background
For the past eight years, Rockwell International has been developing
their sodium carbonate-based regenerative Aqueous Carbonate Process which
uses a spray dryer as a flue gas contactor. A simplification of the process
excludes regeneration and instead involves operation in an open loop
manner as a "throwaway" process. Employing the open loop portion of the
Aqueous Carbonate Process (ACP), Rockwell and Wheelabrator-Frye have Jointly
developed a two-stage dry scrubbing process where an alkaline solution or
slurry is introduced into a spray dryer contactor, and dry reaction products
and fly ash are collected by a fabric filter. Rockwell successfully demo-
strated the open loop portion of the ACP in a 7-ft diameter dryer at the
Southern California Edison Mohave Station in 1972. Flue gas was taken
downstream of the station's electrostatic precipitators and contacted with
a sodium carbonate solution in the spray dryer.
Wheelabrator-Frye piloted a dry injection/baghouae collection FGD
system at the Basin Electric Leland Olds station using nahcolite as a sorbent.
When it became evident that nahcolite supplies would not be available in
commercial quantities in the near future, the Rockwell open loop ACP spray
drying concept was combined with the Wheelabrator-Frye baghouse that provided
spent sorbent and fly ash collection. It was found that the S0_ removal by
the open loop portion of the Rockwell ACP was not as sensitive to sorbent type
as was dry injection; yet, it still retained the desired dry collection
feature. Initial testing was with sodium-based sorbents, although lime was
later found to be a suitable sorbent.
In all spray dryer-based FGD development work and In commercial sales,
Rockwell has used exclusively Bowen Engineering spray drying equipment.
Although the Rockwell-Wheelabrator Frye joint venture developed the two-
stage process, Bowen Engineering has an exclusive agreement to provide spray
drying equipment.
Research
The spray dryer/baghouse pilot unit at Leland Olds was dismantled in
September 1978 after 16 months of operation. During this period Rockwell/
Wheelabrator studied a variety of sorbents, with some sorbent recycling, at
S02 levels from 700 to 3000 ppm.
There are ongoing studies at Bowen's facilities in New Jersey using a
7-ft dryer and a Mikropul pulse-jet collector. Flue gas is from an auxiliary
burner with S02 and fly ash spiking provisions.
Rockwell has completed testing on their portable pilot plant unit at
Northern States Power's Sherbourne County station and at Pacific Power and
Light's Jim Bridger Station. This unit uses a 7-ft dryer with a pulse jet
bag collector and has provisions for warm gas bypass. At Jim Bridger, a
pilot ESP which was originally used to design the existing precipitators in
available to test a spray dryer/ESP system. Fortunately, this Flakt pilot
104
-------
ESP unit is about the same size as the portable test unit (approximately 5000
cfm).
Additional pilot unit testing has begun at Commonwealth Edison's Joliet
Station. The Joliet pilot unit is to be flexible enough for full parametric
studies. The unit uses a 7-ft, 4000 to 5000 acfm spray dryer, with provisions
for using either a reverse-air, off-line cleaning baghouse or a pulse jet,
continuous flow baghouse as well as provisions for an electrostatic precipitator.
The unit will have provisions for both warm and hot gas bypass and sorbent/fly
ash recycle. Most work will be done with lime sorbent, but some soda ash studies
may be made. The Joliet plant burns varying mixtures of four Western sub-
bituminous coals, 8000 to 9000 Btu/lb, 0.5 to 1 percent sulfur. Testing will
last 1 to 2 years.
Other test work being conducted by Rockwell includes bid support studies
on their spray dryer at their California facility and periodic design studies
on a 7-ft diameter dryer at Bowen labs in New Jersey. Rockwell is also nego-
tiating a site for testing lime-based system for high sulfur Eastern coal
applications.
Other Rockwell research includes studies in waste disposal. These
studies show that for nearby disposal (up to 1 mile), pneumatic conveying
may be the most economical method of transportation of waste material.
For longer distances, pelletizing or briquetting of wastes may be required to
permit open trucking. This operation could be accomplished for about $1.25
per ton (operating and capital). The resulting briquette may be unleachable
due to fusing. The only binder material required would be water.
Rockwell has also looked into reuse schemes. By adding a proportionate
amount of water to the waste material, it sets up resulting in load bearing
properties of around 500 Ib/in . Permeability of the product is less than
1 ft/yr (10~7 cm/sec). Some of the waste materials may be useful as a
concrete additive.
Commercial Activities
Following the research work at Leland Olds, the Rockwell/Wheelabrator-
Frye joint venture was awarded a turnkey subcontract for the furnishing,
fabrication, delivery, erection, and successful operation of a complete
emissions control system for the Coyote plant. The Coyote Station is a
410-MW lignite-fired unit to be located near Beulah, North Dakota. The
station is to be owned by a consortium of five North Dakota and Minnesota
power companies. Otter Tail Power Company is to operate the plant.
The 410-MW Coyote Station FGD system is designed for 70 percent S02
removal for all fuels. Guaranteed sorbent utilization is 80 percent, a
conservative value for this high utilization sodium-based system.
Design of the Coyote Station calls for an air preheater outlet
temperature of 285°F. The stack gas must exit at 185°F. These conditions
105
-------
are not optiumum for spray dryer performance; 185*F Is 50* to 60°F above
adiabatic saturation. The design temperature seems somewhat low as most
lignite-fired boilers experience air preheater exit temperatures of 325 to
350°F. Rockwell does not expect any ash alkalinity utilization in the
design since there is no ash recycle and reaction of solid phase fly ash
alkalinity with S02 is minimal.
The Coyote Station will use four 46-ft dryers, each with three 150-hp
atomizers (although at design conditions each will draw only 83hp). The
primary control variable of the system will be dryer outlet temperature.
Outlet dew point will also be, measured and used for setting approach to dew
point on the outlet. Dryer outlet and stack S02 will have a narrow range of
control on sorbent feed rate. In other words, if the outlet SO. level is
well below the control point, the SO- input to the sorbent feed control would
make a small decrease in the feed rate. If outlet SO- level was a primary
control point, the sorbent feed rate might be shut off completely until the
outlet S0? increased sufficiently.
Bag temperature will be limited by an internal bypass between the
inlet and outlet plenums on the baghouse. These bypass dampers will be set
to open if the baghouse inlet temperature exceeds a maximum set point. The
baghouse will use a synthetic fabric, such as Dacron, and will operate at a
net air-to-cloth ratio of 2.7 to 1. This involves a total of 28 compartments,
with two off for maintenance and two off for cleaning at a given time.
Rockwell will use no control valves in slurry service. All slurry feed
rates will be set by progressive cavity pumps (such as Moyno) with variable
speed drives. There will be no flue gas flow dampers or control valves. The
air distribution through the FGD system is controlled by careful design of
the ductwork. The dryer vessels are designed with three atomizers per
vessel. The atomizers are placed a standard distance from the vessel wall to
avoid wall wetting. No agglomeration problems on overlap of spray patterns
from the three atomizers are foreseen. In tests using three atomizers in a
7-ft dryer with extensive spray pattern overlap, virtually the same S0~
removal was measured as would be achieved using a single atomizer flowing the
same amount of sorbent. In l/16th-scale air flow testing of the Coyote duct-
work design, they found they were able to distribute equal flow to four dryers
within two percent and equal flow to three atomizers in a dryer to within
two percent. Bowen feels the three-atomizer approach is advantageous because
loss of an atomizer would not result in loss of a whole dryer module.
Although the use of multiple atomizers has not been demonstrated at the
500,000 acfra level, Bowen has demonstrated flowing 1/3 of the total design
gas flow (170,000 acfra) to a single atomizer.
The Coyote plant will use coal trucks to return waste sorbent/fly ash
to the mine. The material will be handled in the dry form, using dustless
loading equipment. Design of this equipment Is not within the Rockwell
battery limits, however.
106
-------
Construction on the Coyote plant is well underway. On site, foundations
and considerable structural steel are in place. Off site, fabrication is
proceeding on dryer vessels. The plant should start up by the spring of
1981. Start up will more than likely be with commercial soda ash. Rockwell
is currently looking into alternate sources of lower quality sodium products
for use as sorbent at Coyote. Sources from both the Green River, Wyoming
area and the Owens Lake, California area are being pursued. One source
would involve a material that is roughly 50 percent Na.CO. at a cost of
$13/ton. On a cost per ton of Na2C03 equivalent, this is less than half
the $60+/ton of commercial sodium carbonate. Of course, shipping costs
for the less pure product would be greater due to the greater quantities
involved. Soda ash will be stored on site at Coyote as a sodium carbonate
monohydrate slurry, with water circulated through the storage tanks. Feed
to the spray dryers will be saturated sodium carbonate monohydrate solution
from these tanks, which is diluted to achieve the desired spray dryer
temperature. Soda ash will be stored as a monohydrate slurry rather than
as the anhydrous solid because the monohydrate slurry has a bulk density
of 71 Ib/ft versus 55 Ib/ft for commercial dry soda ash.
The Rockwell/Wheelabrator-Frye joint venture has also sold an industrial
dry FGD system to Celanese Corporation. The Celanese project Involves treat-
ing flue gas from a stoker-fired industrial boiler in Cumberland, Maryland.
Currently, the boiler is slated to burn 1 to 2 percent sulfur coal but may go
to a 3 to 4 percent sulfur West Virginia coal. The flue gas flow rate is
65,000 acfm at 350°. The Rockwell system will use lime sorbent with no recycle
of sorbent/fly ash. S02 removal will be 70 to 80 percent for 1 to 2 percent coal,
higher for 3 to A percent coal. Sorbent utilization will be on the order of
70 to 80 percent. Particulate collection will be with a pulse jet, continuous
type collector. Lime slaking will be accomplished with a "pore tube"
slaker. Construction of this system is on schedule, with start-up planned for
January 1980.
As far as other commercial sales, Rockwell has two bids under evaluation
and one bid being prepared. Rockwell expects to bid on up to six more utility
dry FGD systems for Western low sulfur coal or lignite applications, by the end
of 1979.
Rockwell has a standard approach to a dry system design. First, deter-
mination is made whether a "standard" system will work. This involves
approximately a 40°F approach to dew point, no recycle, no gas bypass, and
preferably about 90 percent sorbent utilization. If this Is unattainable,
then warm gas bypass is tried taking a small amount of flue gas after the air
preheater and routing it around the spray dryer to reheat from a closer-than-
40°F approach to dewpolnt. If this is insufficient, hot gas bypass is in-
cluded taking flue gas upstream of the air preheater for reheat. There is a
penalty associated with hot gas as bypass approximating 0.5 percent of the total
fuel rate to the boiler if 5 percent of the gas flow is bypassed. Recycle,
considered to be a supplement to any of the above designs, can be used to
get an apparent sorbent utilization of greater than one. Recycle of fly
ash and partially spent sorbent can result in both improved dryer performance
107
-------
and improved filter performance for a combined synergistic effect. Rockwell
designs recycle equipment as an "add on" with a minimum of redundancy (minimum
first cost). Adding recycle provisions might add about two percent to the
total equipment cost.
Rockwell reports that for large Eastern applications, the cost of lime
reagent is a major factor in the viability of dry systems. In utility high
sulfur applications, the cost differential between lime in a dry system and
limestone in a wet system can mean sorbent savings of millions of dollars a
year for the wet limestone system. This puts the operating economics of a
dry system at a disadvantage for these applications. A scheme for using a
limestone sorbent in a dry system could greatly improve the economics.
Rockwell sees the application of dry FGD to Eastern applications as focaJ
point of future R&D efforts.
Information Sources
Blythe, Gary. Meeting notes with Rockwell International, Canoga
Park, CA, June 13, 1979.
Kelly, M. E. Telephone conversation with Dennis Gehri, Rockwell
International, October 18, 1979.
Janssen, Kent and Robert L. Eriksen. "Basin Electric's Involvement
with Dry Flue Gas Desulfurization," paper presented at the EPA Symposium on
Flue Gas Desulfurization, Las Vegas, Nevada, March 5-8, 1979.
Johnson, O.B., et al., "Coyote Station - First Commercial Dry FGD
System," paper presented at the 41st Annual Meeting American Power Conference,
Chicago, Illinois, April 23-25, 1979.
Moore, K. A., el al., "Dry FGD and Particulate Control Systems," paper
presented at the Fifth Annual EPA Flue Gas Desulfurization Symposium Las
Vegas, Nevada, March 5, 1979.
108
-------
REFERENCES
1. Bechtel Corporation. Evaluation of Dry Alkalis for Removing SO 2 from
Boiler Flue Gases. EPRI Final Report FP-207. Prepared for Electric
Power Research Institute, Palo Alto, CA. October 1976.
2. Chemical and Engineering News. 57(20) : 42. May 14, 1979.
3. Davis, R.A., et al. Dry SO, Scrubbing at Antelope Valley Station.
(Presented at the 41st Annual American Power Conference. Chicago, Illinois.
April 25, 1979.)
4. Dustin, D.F. Report of Coyote Pilot Plant Test Program. Test Report,
Rockwell International (Atomics International Division). Canoga Park,
CA. November,1977.
5. Felsvang, Karsten. Results of Pilot Plant Operations for S0?
Absorption. (Present/ed at the Joy Western Precipitation Division Seminar,
Durango, Colorado. May 21, 1979.)
6. Friedman, L.D. Applicability of Inorganic Solids Other Than Oxides to the
Development of New Processes for Removing SO. from Flue Gas. Contract
No. CAA 22-69-02. Prepared by FMC Corporation for the National Air
Pollution Control Administration, Princeton, NJ. December 1970.
7. Gehri, D.C. and J.D. Gylfe. Pilot Test of Atomics International
Aqueous Carbonate Process at Mohave Generating Station. Final Report
AI-72-51. Atomics International Division/Rockwell International, Canoga ,
Park, CA. September 1972.
8. Giammar, Robert D. et al. Evaluation of Emissions and Control Technology
for Industrial Stoker Boilers. In: Proceedings of the Third Stationary
Source Combustion Symposium, Volume I. EPA #600/7-79-050a. U.S.
Environmental Protection Agency, Research Triangle Park, NC. February 1979.
9. Hollett, Grant T. Dry Removal of SO, - Application to Industrial Coal-
fired Boilers. (Presented at the APCA Convention. Cincinnati, OH.
June 25-29, 1979.)
10. Isahaya, E.F. A New FGD Process by a Spray Drying Method Using
NaOH Aerosols as the Absorbing Chemical. Staub Reinhaltung der Luft in
English. 33:4. April 1973.
11. Janssen, K.E. and Robert L. Eriksen. Basin Electric's Involvement with
Dry Flue Gas Desulfurization. (Presented at the Fifth EPA Symposium on
Flue Gas Desulfurization. Las Vegas, Nevada. March 5-8, 1979.)
109
-------
12. Johnson, O.B., et al. Coyote Station - First Commercial Dry FGD System.
(Presented at the 41st Annual American Power Conference. Chicago,
Illinois. April 23-25, 1979.
13. Kaplan, S.M. and Karsten Felsvang. Spray Dryer Absorption of S09 from
Industrial Boiler Flue Gas. (Presented at the 86th National AICflE
Meeting. Houston, TX. April 1979.)
14. Kaplan, Steve. The Niro-Joy Spray Absorber Development Program:
Pilot-Plant Description and Test Results. (Presented during Joy/Niro
sponsored tour for executives of U.S. Power Industry. Copenhagen,
Denmark. September 23-30, 1978.)
15. Lui, Han and R. Chafee. Evaluation of Fabric Filters as a Chemical
Contactor for Control of S02 in Flue Gas. (Presented at the Air Pollution
Control Office Fabric Filter Symposium. Charleston, SC. March 1971.)
16. Moore, K.A. et al. Dry FGD and Particulate Control Systems. (Presented
at the Fifth EPA Symposium on Flue Gas Desulfurization. Las Vegas,
Nevada. March 5-8, 1979.)
17. Rivers, R.D., et al. The Role of Fabric Collectors in Removing SO..
(Presented at the First National Fabric Filter Alternatives Forum,
Denver, CO. July 1976.)
18. Shah, N.D., et al. Application of Dry Sorbent Injection for S0_ and
Particulate Removal. (Presented at the Fourth EPA Symposium on Flue
Gas Desulfurization. Hollywood, Florida. November 11, 1979.)
19. Shale, C.E. and G.W. Stewart. A New Technique for Dry Removal of SO .
(Presented at the Second Symposium on the Transfer and Utilization of
Particulate Control Technology. Denver, CO. July 1979.)
20. Slack, A.V. Lime Scrubbing by "Dry " Processes. A.V. Slack Report #62.
January 1979. p 15-30.
21. Steams-Roger. Nahcolite Granule Scrubbing System Feasibility Study
(Volume 1). Prepared for Superior Oil Company. Steams-Roger Corporation,
Denver, CO. November 1974.
22. Veazie, F.M. and W.H. Kielmeyer. Feasibility of Fabric Filter as Gas-
Solid Contactor to Control Gaseous Pollutants. Prepared for the National
Air Pollution Control Administration. Report No. APTD-0595. Owens-Corning
Fiberglass Corporation, Granville, OH. April 1970.
23. Wheelabrator-Frye. Non-confidential test data from nahcolite pilot
baghouse study. Leland Olds Station. Unpublished. March 1977.
24. Yamada, Tamotsu, et al. Desulfurization of Combustion Exhaust by Active
Soda Ash. Nagoya Kogya Daigaku Gakuho (Nagoya, Japan). 25: 395-403 1973.
110
-------
APPENDIX A
CONVERSION FACTORS
1 ft
1 short ton «
1 Ib
1 gaL
1000 cfm
1 gal/1000 ft3-
1 BTU
0.3048 meter
0.91 metric ton
0.454 kg
3.79 liters
0.5 m3/s
0.13 liters/m3
0.252 kcal
111
-------
TECHNICAL REPORT DATA
(Please rttd Iiuinictiont on the revtnt before completing)
1. REPORT NO.
EPA-600/7-80-030
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Survey of Dry SO2 Control Systems
6. REPORT DATE
February 1980
«. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
G.M.Blythe, J.C. Dicker man, and M.E.Kelly
B. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
1NE827
9. PERFORMING OROANIZATION NAME AND ADDRESS
Radian Corporation
P.O. Box 8837
Durham, North Carolina 27707
11. CONTRACT/GRANT NO.
68-02-2608, Task 71
12. SPONSORING AGENCY NAME AND ADDRESS
NO PERIOD COVERED
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD C<
Task Final; 4/79 - 1/80
14. SPONSORING AGENCY CODE
EPA/600/13
16 SUPPLEMENTARY NOTES iERL_RTP project officer is Theodore G. Brna, Mail Drop 61,
919/541-2683.
16. ABSTRACT
The report gives results of an assessment of the status of dry flue gas
desulfurization (FGD) processes in the U.S. for both industrial and utility applica-
tions. The assessment is based on reviews of past and current research, develop-
ment, and commercial activities. Systems covered include: (1) spray dryers with
either baghouse or electrostatic (ESP) particulate collectors, (2) dry injection of
alkaline material followed by baghouse or ESP collection of wastes, and (3) other
systems, such as coal-alkaline material feeds to a combustor and .passage of flue
gas through a fixed bed of alkaline material. A summary of dry FGD processes,
including key features of three types of dry systems and commercial systems, is
provided. Limited economic data are also presented. Conclusions and recommenda-
tions are given on the potential role EPA can take to advance the overall environ-
mental acceptability of dry FGD systems as viable SO2 control alternatives.
KIY WORD* AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COS AT i Ftold/Group
Pollution
Sulfur Oxides
Flue Gases
Desulfurization
Fabrics
Filtration
Electrostatic Precip-
itation
Dust
Aerosols
Alkalies
Pollution Control
Stationary Sources
Dry Processes
Baghouses
Fabric Fitters
Particulate
Alkaline Additives
19. SECURITY CLASS (ThU ReportJ
Unclassified
20. SECURITY CLASS (T*l*put*)
Unclassified
13B
07B 13H
21B 11G
07A,07D
11E
18. DISTRIBUTION STATEMENT
Release to Public
21.NO. OFPAOli
112
». PRICE
IPA Perm 22>(M (t-73)
112
------- |