United States
Environmental Protection
Agency
Air and Energy Engineering
Research Laboratory
Research Triangle Park, NC 27711
Research and Development
EPA/600/SR-92/115 December 1992
EPA Project Summary
Demonstration of Sorbent
Injection Technology on a Wall-
fired Utility Boiler (Edgewater
LIMB Demonstration)
P.S. Nolan, T.W. Becker, P.O. Rodemeyer, and E. J. Prodesky
This document summarizes results
of the full-scale demonstration of Lime-
stone Injection Multistage Burner (LIMB)
technology conducted on the coal-fired,
105 MWe, Unit 4 boiler at Ohio Edison's
Edgewater Station. Developed as a
technology aimed at moderate levels
of sulfur dioxide (SO2) and nitrogen ox-
ides (NOX) emissions control for rela-
tively low-cost retrofit applications on
older plants, LIMB operation at a cal-
cium to sulfur (Ca/S) molar ratio of 2.0
is shown to be capable of achieving
approximately 55 to 72% SO2 removal
at capital costs significantly lower than
wet flue gas desulfurization systems
for units under 300 MWe and equal or
lower operating costs. The removal
depends upon the specific sorbent in
use and the degree of flue gas
humidification employed. Sorbents
tested include a commercial calcitic hy-
drated lime, both with and without a
small amount of calcium lignosulfon-
ate, a material added to improve reac-
tivity. Results are presented for
humidification of the flue gas both to
an 11°C approach to the saturation tem-
perature (where an increase of 10% [ab-
solute] in SO2 removal efficiency is ob-
tained), and for minimal humidification
(where water is added solely to main-
tain adequate electrostatic precipitator
[ESP] performance). The performance
of the DRB-XCL™ low-NO^ burners, with
an overall average emission of 206 ng/
J (0.48 lb/106 Btu), is also presented.
The document also discusses the im-
pact of LIMB technology on boiler and
plant operations. The effects are re-
lated primarily to the increased quan-
tity of particulate matter that must flow
through the boiler, ESP, and ash han-
dling equipment. The need for effec-
tive sootblowing is seen as the single
most important requirement when con-
sidering application of the technology:
without it, the increased particulate
loading can decrease heat transfer effi-
ciency between sootblowing cycles.
The other effects discussed in some
detail result from the quicklime compo-
nent of the ash and the precautions
that must be taken to avoid the poten-
tial for deposits at wet/dry interfaces
and for high pH conditions associated
with ash handling.
Finally, the application and econom-
ics of the technology are discussed in
terms of a hypothetical LIMB system
on a 150 MWe boiler. This forms the
basis for a comparison of the capital
and operating costs of the technology
with those of a conventional limestone
wet scrubbing system.
This Project Summary was developed
by EPA's Air and Energy Engineering
Research Laboratory, Research Tri-
angle Park, NC, to announce key find-
ings of the research project that is fully
documented in a separate report of the
same title (see Project Report ordering
information at back).
Introduction
With the growing concern about acid
rain and the improved understanding of
its origins, an active program has been
underway in the U.S. for many years to
develop and utilize technology for acid
Printed on Recycled Paper
-------
rafn control. The focus of this effort has
been on control of sulfur dioxide and ni-
trogen oxides, (SO2 and NOX), from coal-
fired utility boilers, since they represent
the major source of these acid rain pre-
cursors. National point source SO and
NOX emission standards were established
for utility boilers in the 1970s; however,
these standards apply only to newly con-
structed boilers. There exist today many
coal-fired boilers built before establishment
of these regulations. Many of these units
have 20 to 30 years of remaining useful
life, and the new acid rain legislation re-
cently enacted by Congress will require
some level of control for older units.
The SOa emission standards for new
boilers are based on the use of flue gas
scrubbing technology, an option that may
not be technically or economically fea-
sible for existing boilers. Space for locat-
ing the scrubber or disposing of its waste
may not be available and/or the high capi-
tal cost may not be justified for a boiler
With a relatively short remaining lifetime.
As a result, there is a need for low capital
cost, easily retrofitted, control technolo-
gies. The proposed acid rain legislation
suggests that such technologies, which
do not achieve the high removal efficien-
cies of scrubbers, can still be important in
meeting overall removal goals for existing
boilers.
One technology which could play a role
in a utility's compliance with new acid rain
legislation is LIMB (Limestone Injection
Multistage Burners), which is based on
the Injection of a dry sorbent into the boiler
for direct capture of SO in the flue gas,
combined with the use of low-NO burners
for reduction of NOX emissions. l!lMB is a
control technology for coal-fired boilers
which requires low capital investment and
is easily retrofitted to existing boilers. Al-
though it achieves removal efficiencies
lower than flue gas scrubbers, the cost
per unit of SO., removed is much lower
than that for scrubbers. Research and
development work by the U.S. EPA and
others in the late 1970s and early 1980s
established the technical foundation for
the design and operation of a LIMB sys-
tem on a coal-fired boiler.
In 1983 the LIMB process was consid-
ered sufficiently developed to plan a full-
scale demonstration of the technology on
an operating utility boiler. EPA formally
Initiated the demonstration in 1984 with
matching co-funding by the Ohio Coal De-
velopment Office. Babcock & Wilcox
(B&W) was selected as the primary con-
tractor and Ohio Edison's Edgewater Sta-
tion in Lorain, Ohio, was selected as the
host site; both of these organizations pro-
vided funding support. The system was
designed for Edgewater's 105 MWe Unit
4 boiler, a coal-fired unit burning a nomi-
nal 3% sulfur Ohio coal.
The LIMB system was started up in
July 1987 after a thorough review of the
research and development data base and
subsequent design and installation. Initial
operation showed that LIMB was highly
effective in removing SO2, but that the
system could be operated for only a few
hours at a time due to the inability of the
electrostatic precipitator (ESP) to control
particulate emissions. This problem was
diagnosed as being due to the inordinately
high electrical resistivity of the LIMB ash.
This situation made long-term, continuous
operation impossible due to the need to
maintain particulate emissions within the
mandated opacity limit of 20% and 43 ng/
J. The Edgewater LIMB system was op-
erated for several weeks intermittently in
a series of 2 to 5 hour runs to gather SO2
removal data; then the system was shut
down until the particulate removal prob-
lem could be resolved. A maintenance
shutdown of the Edgewater boiler had al-
ready been scheduled at this time, so little
project time was actually lost.
At about this time B&W's dry scrubber
experience and dry sorbent injection work
by Consolidation Coal indicated that
humidification of a flue gas stream in the
presence of an active sorbent resulted in
additional activation of the sorbent and
enhanced SO removal from the gas
stream. For LlMB, humidification offered
the promise of increasing overall SO2 re-
moval and sorbent utilization at a rela-
tively small increase in cost. Flue gas
humidification was also shown to solve
the LIMB particulate problem, since mois-
ture in the gas reduces the electrical re-
sistivity of the ash, reduces the volume of
the gas, and improves other electrical con-
ditions. With the promise of increased SO2
removal beyond what was originally pro-
jected and resolution of the ESP operat-
ing problems, a developmental program
was initiated to design and install a first-
of-a-kind flue gas humidification system
for Edgewater.
A bypass arrangement was used at
Edgewater due to the uncertain reliability
of the humidifier at the time. Humidification
in existing ductwork is envisioned for a
commercial system, if site-specific limita-
tions and design parameters permit suit-
able residence time.
The system started up again in Sep-
tember 1988 and operated until June 1989
under a variety of test conditions, includ-
ing some relatively long, steady-state, op-
erating periods. The remaining portions
of this document give results of the tests
performed, and describe operating experi-
ence with the Edgewater LIMB system.
Results and Discussion of SO2,
NOX, and Particulate Emission '
Control
The original project goal for SO2 cap-
ture, based on extensive bench and pilot
scale tests, was 50% using sorbent at a
calcium/sulfur (Ca/S) molar ratio of 2.0.
During the initial test runs of LIMB in Sep-
tember 1987, SO2 removal was approxi-
mately 55 to 60% at a Ca/S of 2.0. With
start-up of the humidifier in 1988, extended
LIMB run times could be realized because
of improved ESP operation, and the sys-
tem could be adjusted and optimized. The
humidifier added a potential complicating
factor in the determination of LIMB SO
removal, since SO could also be cap-
tured in the humidifier itself. It was soon
confirmed, however, that this removal oc-
curred only at close approach to the adia-
batic saturation temperature of the flue
gas (about 52°C). Data taken while the
humidifier was operated at high tempera-
tures, typically about 135°C where ESP
operation improved but no SO2 removal
occurred, were used to determine SO
removal in the furnace. Later tests with
humidifier outlet gas temperatures ap-
proaching the saturation point were used
to define the incremental removal in the
humidifier. However, unlike the 1987 tri-
als when tests were limited to only a few
hours, humidification stabilized ESP op-
eration and thus permitted continuous op-
eration and rigorous test periods with no
significant changes in operating conditions
extended to as long as 15 hours.
The September 1987 tests included
some runs using commercial, hydrated cal-
citic lime treated with calcium lignosulfon-
ate, an additive that appeared to improve
SO2 removal in earlier bench scale stud-
ies. Additional tests with more extensive
use (1518 metric tons over two periods of
10 and 25 days each) of the modified
sorbent were included in the 1988-89 dem-
onstration. Tests were conducted with
commercial hydrated calcitic lime both at
"minimal humidification" and at an 11°C
approach to saturation. The former term
covers results obtained both before the
humidifier was installed and when the hu-
midifier was used just to maintain a suit-
able flue gas temperature and humidity
for effectiye electrostatic precipitation, nor-
mally about 135°C. Corresponding tests
were performed using lignosulfonate-doped
sorbent. The data indicate that the modi-
fied sorbent increases SO2 removal from
55 to 63% with minimal humidification,
-------
and from 65 to 72% at close approach to
saturation, when operating at a Ca/S of
2.0. The increase is believed to be due to
a combination of the finer size and the
nonagglomerating properties of the modi-
fied sorbent, as well as the sorbent's abil-
ity to resist sintering.
SO2 Capture in the Furnace
SO removal is determined by measur-
ing SO at the stack and comparing this
level with an uncontrolled calculated SO2
level based on the coal sulfur content.
Typically the SO2 concentration falls rap-
idly when sorbent is first injected into the
boiler, and then continues to decrease
gradually to a steady state value within
about 30 minutes. This incremental in-
crease in SO2 removal is thought to be
due to supplemental reaction of sorbent
that settles out on boiler surfaces. After
this initial acclimation of the boiler, the
SO2 levels remain relatively unaffected by
sootblowing and other boiler operations
since normally there is a continual re-
newal of the deposits throughout such ac-
tivities.
Since pilot scale tests had indicated
that SO2 removal was particularly depen-
dent upon the temperature at the injection
point, limited parametric optimization tests
were conducted to quantify the effects at
full scale. The variables, all of which are
related to the temperature, included injec-
tion at different elevations in the furnace,
momentum flux ratio (essentially injection
velocity and furnace penetration at a given
load), the angle of injection (nozzle tilt),
and boiler load. The results show that
these parameters have little effect on SO2
removal in the Edgewater boiler over most
of the ranges tested, indicating that the
system is very insensitive to minor
changes, if the initial design parameters
enable near-optimum operation.
The only factors that influenced SO2
removal significantly were those related to
sorbent injection at the low end of the
1260 to 871 °C sulfation temperature win-
dow or with a gross reduction in injection
velocity. These conditions occurred when
sorbent was injected at the upper eleva-
tion in the furnace, and during periods
when the boiler was at minimum load and/
or when the top row of burners were off.
Under such conditions the estimated tem-
perature at the injection point was 1038°C
or lower and the increased residence time
did not appear to be sufficient to compen-
sate for the difference.
At one point outside of the test period,
the booster air was turned off entirely,
leaving the lime transport air as the sole
means of conveying the lime into the fur-
nace. A dramatic loss in efficiency re-
sulted, indicating that there is a point at
which the momentum flux ratio becomes
important. The condition was beyond the
range of control on the booster air flow,
however, and could not be readily tested.
Injection into higher temperature zones
was attempted in the form of tilting the
injection nozzles at elevation 181 down
15° from horizontal. No appreciable
change in SO2 capture was apparent.
While this was not expected to lead to a
dramatic difference, it leads one to sus-
pect that elevation 181, where the tem-
perature is about 1260°C, was very close
to an optimum injection point for this boiler.
Injection ports could have been located at
a lower elevation where temperatures are
higher; however, recirculation in the lower
furnace would probably have resulted in
sintering and decreased efficiency.
SO2 Removal by Humidification
The Edgewater project demonstrated
that LIMB SO2 removal could be enhanced
by humidification of the flue gas. Most of
the tests conducted at close approach to
the flue gas saturation temperature were
limited to a period of 8 to 12 hours so that
the baseline removal could be certified.
The longest continuous run with no inter-
ruption was 29 hours. Subsequently, how-
ever, the humidifier has been operated
continuously for as long as 11 days at an
11 to 14°C approach temperature as part
of the Coolside process tests under the
U.S. Department of Energy-sponsored
LIMB Demonstration Project Extension.
Humidification to an 11 °C approach tem-
perature increases SO2 removal from 55
to 65% at a Ca/S of 2.0. The tests run at
22 and 33°C approach temperatures sug-
gest that the enhancement is lost some-
where slightly above a 33°C approach.
However, the few data points obtained
are not considered sufficient to define a
truly representative curve. Pilot work and
dry scrubber experience predict an expo-
nential increase in SO2 removal when ap-
proaching the saturation temperature.
NOX Emission Control
In order to reduce NOX emissions, the
existing circular burners were replaced with
B&W's DRB-XCL™ burners as part of the
demonstration project. Discrete baseline
tests with the older burners had shown
emission levels of about 301 and 387 ng/
J at 63 and 97 kg/sec main steam flow
(mid and peak load conditions, respec-
tively). The DRB-XCL™ burners do not
appear to be as sensitive to load condi-
tions. An overall average NOX emission
level of 206 ng/J was obtained over the
full range of operating conditions during
the January - June 1989 period. Aver-
ages of the weighted 24-hour and 30-day
rolling averages of 202 and 211 ng/J were
calculated for the periods of January 8 to
26 and February 21 to April 22, 1989,
respectively.
The DRB-XCL™ burners operate at
about 1.0 to 1.2 kPa pressure drop, in
comparison to the 0.5 to 0.7 kPa required
by the circular burners. The unburned
carbon content averaged 1.54 wt % for
four LIMB ash samples collected
isokinetically while sorbent was being in-
jected at a Ca/S of 2.0 and NOX emissions
were about 206 ng/J. This would equate
to about 4.6 wt % in an ash undiluted by
reaction products and excess sorbent. It
was also noted that there appeared to be
no interactive effects between sorbent in-
jection and NOX reduction.
Particulate Emission Control
An early concern of the LIMB demon-
stration was the ability of the Edgewater
ESP to handle the two- to threefold in-
crease in particulate loading that accom-
panied sorbent injection. While the load-
ing and the finer size of the particulate are
issues that must be addressed, the pre-
liminary tests in 1987 dramatically indi-
cated that the extremely high electrical
resistivity of unconditioned LIMB ash was
by far the most immediate impediment to
continuous operation. As noted earlier,
the high resistivity, in particular, precluded
continuous operation beyond several hours
as the opacity climbed toward the 20%
limit at that time. Fortunately, the humidi-
fier installed for the 1988-89 operations
proved to be an effective means of condi-
tioning the ash and restoring the efficiency
of the ESP to near normal levels.
Although the Edgewater humidifier was
designed to accommodate close approach
to saturation for SO2 removal purposes, it
was soon found that adding even modest
amounts of water (lowering the flue gas
temperature to about 135°C) was suffi-
cient to maintain opacity generally in the 1
to 7% range throughout the demonstra-
tion. Specific ESP tests included mea-
surement of resistivity in situ, operating
voltage and current, and inlet and outlet
gas flow, mass loading, and particle size
distribution. These intensive tests were
conducted during the week of May 22,
1989, and cover operation of the ESP
with both three and five of the six fields in
service while the unit was operating at 75
MWe, due to operational concerns be-
yond the control of the project. Neverthe-
less, the conditions were such that the
humidifier was operating at the full design
-------
flow since repairs to various sources of air
In-leakage in the boiler had not been fully
successful.
For the tests, the ESP inlet temperature
was maintained at 135°C except for one
test with an inlet temperature of 74°C.
The results of these tests, coupled with
those from associated laboratory tests and
mathematical modelling, indicate that
humidiflcation to 135°C lowered resistivity
to 2.3 x 10" ohm-cm, permitted continu-
ous operation at acceptable opacity lev-
els, and resulted in paniculate emissions
on the order of 1.7 to 8.6 ng/J or less at
the conditions tested.
LIMB Commercial Design and
Economics
Two LIMB retrofit systems were de-
signed for a 150 MWe boiler to allow
comparison of the costs and impacts with
those of a wet flue gas desulfurization
(FGD) system. Previous economic stud-
ies had shown LIMB to have cost advan-
tages over wet FGD for units under 300
MWe, while wet FGD was more cost ef-
fective for larger units. A representative
unit firing 2% sulfur coal was used as a
basis. Complete costs of the two LIMB
system designs were estimated, and the
resultant operational changes analyzed.
The two systems were selected to cover
the range between the high cost when
humidification to an 11°C approach tem-
perature is used to obtain maximum sulfur
capture and the lower cost system where
minimal humidification only restores ESP
performance. The main difference be-
tween the systems is the inclusion of a
humidification chamber for the maximum
humidification case, while the minimal
humidification design included installation
of humidifier lances in existing flues.
Results of the analysis show that LIMB
system capital costs are in the range of
30 to 43% of a wet FGD system, and that
the costs per ton of SO2 removed are
lower than those of equivalent performance
Wet FGD systems. Additionally, the ad-
vantages of the LIMB systems include:
simplicity of operation and ease of main-
tenance, reduced outage time for installa-
tion, shorter lead time for material supply,
smaller space requirement, and lower re-
duction of plant electrical output
Operational Considerations
LIMB Effects on Boiler
Operation
The impact of LIMB technology on boiler
operation is related primarily to increased
particulate loading and depends heavily
on the adequacy of the sootblowing sys-
tem. With sorbent injection in the upper
furnace as was done at Edgewater, LIMB
technology results in a higher rate of ash
accumulation on tubes and other avail-
able surfaces, roughly in proportion to the
additional particulate matter. Virtually all
of the sorbent appeared to follow the gas
stream through the convective pass, rather
than being reported as bottom ash. Thus
injection at the elevation of the nose and
above avoids the potential for sintering
and slagging conditions that would prob-
ably have resulted had the sorbent been
introduced at a higher temperature and/or
a lower elevation. The four new steam
sootblowers added in the primary super-
heat area as part of the project were ef-
fective within their range, but did not make
up for limitations elsewhere.
The LIMB ash that did accumulate on
the tubes was found to be about as easily
dislodged by sootblowing as is normal coal
fly ash. The most notable difference was
that, instead of operating on a normal
once-per-shift basis, the sootblowers were
run about three times as frequently, effec-
tively in almost continuous cycles. As
one might expect, this strained the capac-
ity of the compressed air system at
Edgewater, to the point that all but the
four air heater blowers were converted to
a steam-driven system for the LIMB ex-
tension work to follow. This is not to say
that an air system would not work, only
that the capacity of the Edgewater system
could not keep up with the heavier duty
cycle.
The overall effect of ash accumulation
on the tubes was a decrease in heat trans-
fer and higher flue gas temperatures
throughout the boiler. Heat transfer rates
through the tubes appeared to return to
normal immediately after being blown, only
to begin the cycle once more. Because of
this and the relatively low capacity of the
air-driven sootblowers, the net result was
a flue gas temperature of about 177°C at
the air heater outlet, compared to a more
normal temperature of about 149°C under
steady state conditions. However, effec-
tive sootblowing is expected to -reduce
this temperature rise significantly.
The other source of energy loss neces-
sarily associated with LIMB is the excess
air used both for sorbent dispersion in the
furnace and water atomization in the hu-
midifier. At Edgewater, ambient air was
used for both, although preheated air,
recirculated flue gas, or even mechani-
cally assisted devices could be used for
sorbent dispersion in another installation.
While the data have not been reduced in
terms of a specific energy loss, the
"booster air fan" arrangement at Edgewater
used about 3.78 kg/sec of air to feed 1.51
kg/sec of sorbent into 126 kg/sec of flue
gas. Humidification to an 11°C approach
to saturation required about 2.52kg/sec of
air to atomize 5.54kg/sec of water. The
atomization system was designed conser-
vatively with an air compressor motor rated
at 876 kW, to ensure the production of the
fine droplet size required. Close approach
to saturation at Edgewater also required
flue gas reheat for plume buoyancy. A
steam coil reheater with the capability of
raising the flue gas temperature 22°C was
installed for this purpose at the outlet of
the ESP, just before the induced draft fan.
In applications where only minimal
humidification would be required for ESP
operation, the energy penalty associated
with high levels of humidification and re-
heat would be significantly lower.
Humidifier Operation
The humidification system installed at
Edgewater proved to be relatively prob-
lem-free and achieved the purposes for
which it was designed. Early mathemati-
cal and plastic flow modelling and quarter-
scale simulation tests, all conducted in
the early design phase, provided mean-
ingful insight into the flow patterns and
evaporation processes involved. For the
most part, the wet/dry interface problems
were limited to relatively small, manage-
able areas. The most significant problem
was a buildup of material near some of
the atomizers. The material would build
to several centimeters in length and then
fall to the floor of the horizontally config-
ured chamber where material would accu-
mulate. A modified airfoil lance assembly
for the atomizer array was designed and
installed in 3 of the 22 lance locations for
the last 3 months of the program. These
modified assemblies had a greatly reduced
tendency for deposit formation, and the
complete atomizer array was replaced be-
fore continuing with the Coolside process
tests.
When the humidifier was operated at
outlet temperatures 22°C and above the
saturation temperature, essentially no dif-
ficulties were encountered. Moreover, the
presence of unreacted quicklime, CaO, in
the ash undoubtedly contributed to favor-
able conditions by virtue of its exothermic
reaction with water. Although the quantity
of water vapor present in the flue gas as a
result of combustion was more than suffi-
cient for total rehydration of the CaO, the
gas/solid reaction kinetics appeared to be
quite slow when the humidifier was oper-
ated considerably above the saturation
temperature. Under this operating condi-
-------
tion, much steam evolved when mixing
LIMB ash with liquid water just prior to
disposal (described in more detail later).
Notably less steaming occurred when the
humidifier was operated at close approach
to saturation, suggesting that, whatever
the mechanism(s), the excess CaO has a
strong affinity for liquid water.
Predictably, deposits formed in the cen-
ter area of the humidifier outlet turning
vanes after extended operation precisely
at an 11°C approach to saturation. The
deposits built to a thickness of about 10cm
and then sloughed off onto the floor.
These deposits are not considered to be
a problem because the chamber was be-
ing tested at the extreme limit of its de-
sign and several remedial alternatives
could be used in a commercial system.
These include not only operation at a
slightly higher temperature, but also inten-
tional capture of large, unevaporated drop-
lets with periodic deposit collection and
removal through a hopper system. If site-
specific spatial limitations permit, a verti-
cal, down-flow configuration with a hopper
for solids collection would be preferred for
a commercial humidifier required to oper-
ate at close approach temperatures.
Ash Handling and Disposal
LIMB technology also has significant im-
pacts on the power plant's ash handling
and disposal practices. The most obvious
of these is the sheer quantity of material
that must be processed. As implied ear-
lier, operation at a Ca/S of 2.0 with a 10%
ash, high sulfur coal almost triples the
rate at which ash must be collected by the
ESP and then transferred to the ash silo.
As long as lines were free of obstructions
and the transfer equipment was in good
working order, the increased quantity gen-
erally did not pose a problem since nor-
mal demand and the basic design of the
ash handling system could accommodate
such amounts. By the same token, it was
important to react to upsets quickly since
there was less time available to make
required repairs.
Beyond the quantity of ash itself, the
other major considerations regarding LIMB
ash handling and disposal stemmed from
the quicklime component and the pozzo-
lanic properties of the ash. At Edgewater,
ash is pneumatically conveyed from the
ESP hoppers into the ash storage silo.
Even during the relatively short tests in
1987, LIMB ash tended to bridge over at
the wet/dry interface near the aspirating
water jets used to create the vacuum. A
device designed to ram through the de-
posit periodically was successfully de-
signed and installed to overcome this
seemingly small, but critical, problem.
Greater difficulty was encountered due
to the amount of steam generated by the
water/quicklime reaction occurring in the
beds of the trucks used to haul ash to the
landfill. Water was mixed with dry LIMB
ash from the silo in a pug mill that dis-
charged directly into the truck waiting be-
low. As the level rose in the bed, the
steam that resulted soon made it impos-
sible for the operator to see how much
room remained. As noted earlier, this
was true especially when minimal
humidification was employed. Although
the problem had been anticipated and a
large fan had been mounted to blow or
suck the steam away, the severity of the
steaming was far beyond the fan's capa-
bility. After considering a number of alter-
natives including weighing and sonic tech-
niques, a system was devised that in-
volved the operator's lowering a freely
hanging thermocouple to the desired fill
level in the empty truck bed. The thermo-
couple then read ambient temperature up
until the time it would begin to sense the
hot ash. While not regarded as a perma-
nent "commercial" solution to the problem,
it proved to be an expedient, practical
remedy for the purposes of the demon-
stration. In time, a "reference" thermo-
couple was added in the discharge chute
of the pug mill, since some variation in
temperature did occur, typically in the 66
to 121°C range, depending on the stoichi-
ometry and degree of humidification.
For future commercial systems, a more
sophisticated design including those based
on gravimetric, sonic, or even spectropho-
tometric techniques is envisioned, if treat-
ment is restricted to an ash unloading
system such as exists at Edgewater. As-
suming that rehydration of the quicklime
component is the preferred method of ash
disposal, another extreme is possible
where there are pre-existing facilities, such
as a pug mill and radial stacker at the
disposal site, which can readily accommo-
date steam from the ash. The steam
emanating from the surface of the ash
presents no sustained problem in that it
subsides to faint wisps within about 15
minutes.
Before the steaming problems were
brought under control, they gave rise to a
further complication that had been under-
estimated. While underfilling the truck
was the norm, on a couple of occasions
overfills did occur that had to be picked
up with a front end loader. The area then
was washed down as a final cleanup mea-
sure. When this was done, the lime-rich
ash would raise the pH of water draining
toward the plant's ash pond. While provi-
sion had been made for sensing and neu-
tralizing high pH water that could result
from a failure in the pneumatic ash con-
veying system and from minor yard spills
of lime or ash, the amount that spilled
from overfilling the truck had not been
anticipated. Moreover, sumps that accu-
mulate significant quantities of lime or LIMB
ash due to low flow or agitator failure
could also produce undesirably high pH.
The solution to all of these related difficul-
ties was rerouting all of the potential
sources through the neutralization system.
Most of the LIMB ash from Edgewater
was placed in Ohio Edison's ash disposal
site, adjacent to the normal ash disposal
area. Despite apprehensions that the
cementitlous properties of the ash might
cause some difficulty with the bulldozers,
no significant problems were encountered,
probably because the pozzolanic reactions
did not proceed to any appreciable extent
at the relatively low water/ash ratios used
to produce a material that can be readily
dumped from the truck. In addition to the
routine LIMB ash disposal, approximately
140 truckloads were placed in two test
cells in an isolated area of the disposal
site for study over the next several years
under a related DOE-sponsored program.
Summary and Conclusions
The results of the LIMB demonstration
project on a full-scale, coal-fired utility
boiler show that the technology has met
or surpassed program goals with respect
to SOg and NO emission control. More-
over, in spite of early difficulty in control-
ling particulate emissions with the ESP,
the design and installation of a full-scale
humidification system appears to have
overcome the adverse aspects of the in-
creased quantity of fine, high resistivity
ash. Specifically, the use of a lignosulfon-
ate-doped hydrated calcitic lime, produced
in bulk by a commercial supplier, and op-
eration of the humidifier at an 11°C ap-
proach to saturation resulted in SO re-
moval efficiencies of up to 72% at a Ca/S
of 2.0, far in excess of the 50% program
goal. Even without the lignosulfonate ad-
ditive, up to 65% capture was possible
under the same conditions. Removal effi-
ciencies of about 62 and 55% were ob-
tained in the absence of humidification to
the close approach temperature for the
modified and unmodified sorbents, respec-
tively. The SO2 data have also been
reduced and presented in a form that char-
acterizes the removal over a range of
stoichiometric ratios. At the same time,
an overall average NOx emission level of
206 ng/J was achieved over months of
operation with the DRB-XCL™ burners in-
stalled to meet a goal of 215 ng/Jor less.
Particulate emission control was effectively
restored by modest levels of humidification
-------
as indicated by opacity measurements at
the ESP outlet. Particulate mass emis-
sion levels of 4.3 to 8.6ng/J were found
to be far below the 43ng/J goal, although
the tests had to be conducted at condi-
tions representing approximately 75% of
the full boiler load. This load limitation
was imposed because long-term loss of
refractory material had exposed support
beams in the vicint'ry of the nose arch to
unaccepatably high temperatures.
The designs of two 150 MWe LIMB
systems, one with only minimal
humidification to restore ESP efficiency
and another with the capability of operat-
ing at close approach to saturation, were
used as a basis for cost comparisons with
wet limestone FGD. Economic studies
have shown that the capital and operating
costs of LIMB technology compete favor-
ably for units up to about 300 MWe. Ad-
ditional benefits of LIMB include ease of
operation and maintenance, relatively
simple installation, smaller space require-
ment, and a lower reduction of the plant's
generation capacity.
The overall impact LIMB technology had
on boiler and plant operations at
Edgewater is particularly notable in a few
areas, all of which are considered man-
ageable when viewed from the perspec-
tive of further commercialization. The single
greatest impact comes simply from in-
creased particulate loading through the
boiler, ESP, and ash handling system.
Unless restored by effective sootblowing,
heat transfer efficiency in the boiler de-
creases in proportion to the amount of
sorbent injected. At higher Ca/S ratios
the limitations of the compressed air
sootblowing system at Edgewater soon
pointed toward the need for future com-
mercial systems to address sootblowing
capability more thoroughly than had origi-
nally been anticipated for the demonstra-
tion project. In comparison with these
effects, the impact of increased gas flow
from air used to disperse the sorbent and
to atomize water in the humidifier is thought
to be quite low, especially if only moder-
ate humidification to perhaps 135°C is re-
quired to restore ESP efficiency. While
the data indicate an overall energy pen-
alty of about 1.5% in boiler efficiency for
the combined impacts, it is strongly be-
lieved that this can be recovered with more
effective sootblowing.
Again because of the increased quan-
tity of material, LIMB technology makes
additional demands upon the ash han-
dling system that must be taken into ac-
count in the application for future com-
mercial systems. Beyond this, the chemi-
cal composition of the LIMB ash, and the
quicklime component in particular, should
be reviewed carefully in light of possible
deposit formation at wet/dry interfaces
where the pozzolanic properties of the
ash can be important. Likewise the lime
component of the ash is important with
regard to potentially high pH conditions in
any wetting operation that might be used
as part of the ash disposal process.
•U,S. Government Printing Office: 1993 — 750-071/60164
-------
-------
P. Nolan, T. Becker, P. Rodemeyer, andE. Prodeskyare with the Babcock& Wilcox
Co., Barbeton, OH 44203.
David G. Lachapelle is the EPA Project Officer (see below).
The complete report, entitled "Demostration ofSorbent Injection Technology on a
Wall-fired Utility Boiler (Edgewater LIMB Demonstration)," (Order No. PB92-201
136/AS; Cost: $35.00, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States
Environmental Protection Agency
Center for Environmental Research Information
Cincinnati, OH 45268
Official Business
Penalty for Private Use
$300
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
EPA/600/SR-92/115
------- |