United States
Environmental Protection
Agency
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
Research and Development
EPA-600/S7-82-025a August 1982
Project Summary
Ammonium Sulfate and
Bisulfate Formation in Air
Preheaters
J. M. Burke and K. L Johnson
Nitrogen oxide (IMOX), emissions
from electric utility boilers may be
reduced by 80-90 percent, through
the application of pollution control
technology based on the selective
catalytic reduction of NOxwith ammo-
nia; however, some unreacted ammo-
nia may be emitted from the control
system. This study investigated the
potential impact of these ammonia
emissions on a combustion air pre-
heater, downstream of a selective
catalytic reduction system. Athermo-
dynamic analysis was conducted
which indicated that both ammonium
sulfate ((NH4)2804) and ammonium
bisulfate (NHUHSCM could form in the
intermediate and low temperature
zones of an air preheater and that
(NH4)2SO4 was the thermodynamically
favored reaction product. A kinetic
analysis of the NHa-SOs reactions was
conducted which showed that IMH4
HSOj is the first compound formed
under the time-temperature histories
in an air preheater. This indicates that
the reaction which forms NH4HSO4
from gaseous reactants is more rapid
than that which forms (NH4);>SO4. The
study identified five techniques for
minimizing the deposition of ammo-
nium sulfates in an air preheater. A
technical and economic evaluation of
each solution was conducted: the
results indicate that the use of avail-
able air preheater design options is the
optimum technique for minimizing
deposit formation.
This Project Summary was devel-
oped by EPA's Industrial Environmen-
tal Research Laboratory, Research
Triangle Park, NC. to announce key
findings of the research project that is
fully documented in,a-separate report
of the same title (see Project Report
ordering information at back).
Introduction
Currently, the principal methods of
controlling nitrogen oxide (NOx) emis-
sions from stationary combustion
sources are combustion modifications
and flue gas treatment. Combustion
modifications are employed on new
utility boilers and are generally capable
of reducing NOx emissions up to 50
percent. Flue gas treatment is a technol-
ogy which is used commercially in
Japan and it is being applied on a limited
basis in the U.S.
Several flue gas treatment processes
have been developed in Japan. Of these,
only selective catalytic reduction (SCR)
of NOx with ammonia has commercially
demonstrated the capability of limiting
NOx emissions from stationary sources
by 90 percent or more. Because SCR
can attain such high NOX reduction
efficiencies, it is receiving increased
attention in the U.S.
For a utility application of SCR, a
catalytic reactor is located between the
economizer and air preheater sections
of the boiler. At this point, ammonia,
injected into the flue gas upstream of
the catalyst, reacts with NOx on the
-------
catalyst surface to form elemental
nitrogen and water. The overall reac-
tions can be represented by:
4NO + 4NH3 + 02 - 4N2 + 6H20 (1)
2NO2 + 4NH3 + 02 - 3N2 + 6H20(2)
Although SCR systems have under-
gone extensive commercial develop-
ment in Japan, an unresolved issue is
that of ammonia emissions from the
process and the impact of these emis-
sions on equipment downstream of the
catalytic reactor. Such equipment can
include air preheaters and other pollu-
tion control equipment. This study
specifically addresses the impact of
ammonia emissions on the operation of
air preheaters downstream of SCR
systems.
Study Objectives and Approach
This study had three major objectives:
(1) to collect and interpret data on the
thermodynamics and kinetics of NH3-
S03 reactions; (2) to identify techniques
for minimizing the deposition of ammo-
nium sulfates in an air preheater; and (3)
to complete a technical and economic
evaluation of those techniques.
The approach used to meet the study
objectives began with a thermodynamic
analysis of possible NH3 reactions to
identify conditions which favor the
formation of ammonium sulfates.
Experimental data were then compared
to the expected equilibrium results and
a kinetic model was developed to predict
the formation and deposition of ammo-
nium sulfates.
Information was also collected on air
preheater design and operation. These
data were then used, in conjunction with
the results of the thermodynamic and
kinetic analyses, to identify techniques
for minimizing the deposition of ammo-
nium sulfates in an air preheater.
Finally, a technical and economic
evaluation of each technique was
completed.
Problem Definition
The problem of ammonium sulfate
deposition has several aspects which
must be defined before solutions can be
identified. The chemistry of the NH3-
S03 reactions must be examined and
the principal causes of deposit forma-
tion pinpointed. In addition, the impact
on operation and maintenance of the air
preheater and the environmental im-
pacts associated with (NH4)2S04/NH4
HSO4 deposition must be considered.
NH3-SO3 Chemistry
Experience at SCR installations has
demonstrated that NH3 and SO3 can
react and deposit in air preheaters.
Chemical analysis of the deposits
indicates that the principal reaction
products are (NH4)2S04 and NhUHSCU.
Other components in the deposits
include corrosion products (NH4-Fe-SO4
compounds) and fly ash (Ca, Si, and Al
compounds). The following discussion
identifies reactions which can occur in
an air preheater.1l 2l 3'
The NH3-S03 reactions which can
take place in an air preheater are
illustrated by:
2NH3 + S03 + H2O5£(NH4)2S04 (s) (3)
NH3 + S03+H20*:NH4HS04(I) (4)
These reactions can occur as a gas con-
taining NH3, S03, and H2O is cooled. For
a flue gas stream, the exact temperature
at which solid/liquid products begin to
form or "condense" depends on the
reaction product(s) formed and on the
concentration of NH3, S03, and (to a
lesser degree) H2O. In general, as the
concentrations of reactants in the gas
increase, the temperature at which
, deposits will form also increases.
Another reaction which occurs as the
gas is cooled is represented by:
H2O + S03^:H2S04(g) (5)
Formation of H2S04 lowers the
concentration of SO3 available for the
reaction with NH3 (Equations 3 and 4.)
This, inturn, reduces the temperatureat
which either (NHUfeSCu or NH4HSO4 will
form. Of course, NH3 can react with
H2SO4 as illustrated by:
25O-1
200-
1
750-
700.
2NH3 + H2SO45j:(NH4)2SO4 (s) (6)
NH3+H2S04^NH4HSO4(D (7)
However, these reactions occur at a
lower temperature than those of NH3
with S03. As a result the formation
temperatures for the reactions with
H2SO4 define a lower limit for formation
of (NH4)2S04 and NH4HSO4, while the
formation temperatures for the reac-
tions with S03 define an upper limit.
Based on evaluation of NH3-SO3
reactions, the compounds which can
form in an air preheater are (NH4)2SO4
and NH4HSO4. If only (NH4)2S04 forms,
it will exist as a solid. This compound
does not melt; it decomposes. If
NH4HSO4 or both compounds form, the
deposit may be either solid or liquid,
depending on the deposit composition
and the temperature as shown in Figure
1. 3'4
Ammonium Sulfates and
Air Preheater Design
Early experience with corrosion and
plugging of air preheaters resultedfrom
condensation of H2SO4 vapor on the
metal surfaces in the preheater. The
deposition of NH3-SO3 compounds is
somewhat analogous to the H2SO<
condensation problem. Liquid NH4HSO<
or a solution of (NH4)2S04 and NH4HSO4
can condense and deposit in the air
preheater. Formation of solid (NH4)2S04
should not present a problem since the
solid should not adhere to the surface as
A - Solid (NH 4)2804 and
B - Solid NH4HSO4 and
Liquid Solution of
(NH4)2S04 and
NH4HSO4
C - Solid
-------
readily as a liquid. However, liquid
(NH4>2S04/NH4HSO4 deposits can
present problems. They tend to collect
fly ash particles and to react with both
the fly ash and the metal surface in the
air preheater to form a solid deposit.
Deposition of ammonium sulfates
presents a more significant problem
than H2SO4 deposition. To understand
why, it is necessary to examine the
basics of air preheater operation. Most
utility boilers employ regenerative air
preheaters in which heat is absorbed
from the flue gas by metal heat transfer
elements. These metal elements are
then exposed to the combustion air
where they release the heat absorbed
from the flue gas. Subsequently, the
elements are re-exposed to the flue gas,
and the cycle is repeated.
During the heat transfer cycle in a
regenerative preheater, the temperature
of the metal heat transfer elements
changes continuously as they are
alternately exposed to flue gas and
combustion air. Figure 2 is a typical
temperature profile for a regenerative
air preheater. The center line represents
the average metal temperature as a
function of depth through the preheater.
The left- and right-hand lines represent
the low and high metal temperatures,
respectively.6.
Some features of regenerative air
preheaters which are used to minimize
the impacts of H2S04 condensation/
plugging include soot blowers and water
washing equipment. Soot blowing has
been used as an effective technique for
controlling and minimizing deposits.
Soot blowers direct a high pressure
stream of either steam or air onto the
heat transfer elements where deposits
can accumulate. This dislodges the
deposits which are then entrained by
the flue gas.
Water washing is usually required to
supplement soot blowing of the air
preheater. Typically, washing is
restricted to boiler outages. However,
the frequency of water washing is
ultimately determined by the pressure
drop across the air preheater. Once the
pressure drop increases beyond a
certain level, washing is required.
The temperature profile shown in
Figure 2 indicates some additional
design features used to minimize air
preheater plugging and corrosion. As
shown, the fluctuation of metal temper-
atures in the extreme cold end of the air
preheater is less than that in the higher
temperature zones. This tends to limit
H2S04 condensation to the cold end and
is due to the design of the heat transfer
150 ~
125 -
100 -
1 "-
50 -
25 -H
Low Temperature Zone
1
50
I I I
100 150 200
Temperature, °C
250
300
350
Figure 2.
Temperature profile of heat transfer elements in a regenerative air
preheater.
elements in the cold end. The cold-end
heat transfer elements are constructed
of heavy-gauge material which allows
the elements to corrode and still retain
the strength, to withstand the soot
blower blast. In addition, the configura-
tion of the elements is different,
consisting of spaced, flat sheets
oriented parallel to the gas flow. This
permits better penetration of the soot
blower jet into the cold end. 5' 6 These
differences result in the cold-end
elements having a higher heat capacity
and lower heat transfer efficiency
which in turn limits temperature
fluctuations.
As previously discussed, the deposi-
tion of ammonium sulfates is analogous
to H2SC>4 condensation in an air
preheater. However, ammonium sulfates
condense at higher temperatures than
does HjSO-t, presenting significant
problems. Figure 3 illustrates the
relationship between typical H2S04(I),
(NH4)2S04, and NH4HSO4 initial forma-
tion temperatures and the air preheater
temperature profile from Figure 2. As
shown, H2SO4 can form in about 35
percent of the preheater but only the
extreme cold-end metal temperatures
are always below the acid dewpoint. On
the other hand, (NH4)2SO4 and
NH4HS04 can form in about 50 percent
of the air preheater. In this case, some
portion of the heat transfer elements in
the intermediate zone of the preheater
are always below the formation
temperatures for ammonium sulfates.
3
-------
150 -i
125 -
100 -
I
75 ~
•s
50 -
25-
50
Figure 3.
Temperature, °C
NH4HSO4. and HzSO^I) formation temperatures in an
air preheater.
The fact that both (NH4)2S04 and
NH4HS04 can form and deposit in the
intermediate temperature zone of an air
preheater presents several problems:
(1) the heat transfer elements in the
intermediate temperature zone are
especially susceptible to corrosion since
they are typically manufactured from
light-gauge carbon steel; (2) soot
blowing equipment is not as effective in
removing NHa-SOa deposits as it is in
removing H^O* deposits (the blowers
are at the ends of the preheater, while
NHs-SOs deposits can form in the
center); and (3) since soot blowing will
probably not be effective in controlling
deposit formation in the intermediate
zone, water washing may be required
more frequently (this could require
either forced boiler outages or periodic
reductions in boiler load to permit
washing and thereby maintain accept-
able air preheater performance).
Available data from installations in
Japan indicate that air preheater
plugging problems do occur. In some
cases, soot blowing is ineffective and
more frequent water washing is
required.1' ' 3
Impact of Deposit Formation
on Air Preheater Performance
Two major aspects of air preheater
performance can be affected by the
deposition of ammonium sulfates
thermal efficiency and pressure drop
Thermal efficiency, a measure of the
heat transferred from the flue gas to the
combustion air, is significant since £
decline in efficiency (evidenced by ar
increase in flue gas temperature at the
air preheater exit) decreases boilei
efficiency.
The impact of (NH4)2SC>2/NH4HSO-
deposition on the thermal efficiency 01
the air preheater is not expected to be
significant. The results of tests
conducted by the Electric Powei
Development Company (EPDC) in Japar
showed no evidence of a decline in the
thermal efficiency of the preheater.7 Ir
addition, another study reported thai
the presence of soot and fly ash deposits
can improve the thermal efficiency ol
the air preheater. This is a result of the
deposits actually increasing the heat
capacity of the air preheater.8
Pressure drop through the air
preheater is also an important aspect ol
preheater performance. An increase ir
pressure drop through the preheatei
can cause a slight decline in the therma
efficiency of the boiler. In addition, if the
increase is too large, the fans may noi
be able to maintain full flow rates anc
the boiler would have to operate al
reduced load or be shut down to permh
washing of the preheater.
Japanese experience has shown thai
(NH4)2SO4/NH4HSO4 deposition car
have an adverse impact on the pressure
drop through the preheater. In some
cases the air preheater pressure drof
has increased despite the use of soo
blowing, and more frequent watei
washing of the preheater has beer
required.
Environmental Considerations
Ammonia emitted from SCR systems
can affect operation of downstream
equipment and may also have some
environmental impacts. The principa
environmental impact which is expectec
to result from deposition of (NH4)2SCv
NH4HS04 will be associated with the
water stream which is used to wash
deposits from the preheater. This
stream can contain^issolved, NH4, SO42
Fe+z/3, and other compounds whicr
may be present in fly ash. The actua
composition of the wash water will be
similar to the composition of water f rorr
conventional air preheater washing
The principal difference is the presence
of NHj and possibly higher Fe concen
trations than normal.
-------
Air preheater wash water is classified
as a metal cleaning waste; as such, it
must meet the discharge limits which
are shown in Table I.9 This waste
stream can be treated separately or it
may be combined with other metal
cleaning wastes in a single treatment
facility. The type of treatment employed
is site specific.
Typically, metal cleaning wastes
contain copper and iron; depending on
the cleaning process used, significant
amounts of NH3 also may be present.
Some typical boiler cleaning chemicals
include ammoniated citric acid,
ammoniated EDTA, and ammonical
sodium bromate. Use of these chemicals
can result in NH3 concentrations of 700
- 5200 mg/l in the waste stream.9
Analysis of wash water from an air
preheater downstream of an SCR
system indicated an average NH3
concentration of 8 mg/l. (As a
maximum, the NH3 concentrations in air
preheater wash water should be less
than 100 mg/l.) These concentrations
are less than those expected in typical
metal cleaning wastes; as a result, the
most significant impact of NH3 in the air
preheater wash water will be to
increase the volume of metal cleaning
waste water which must be treated.
Thermodynamic and Kinetic
Analyses of NH3-SO3
Reactions
One objective of this study was to
quantify the factors which influence the
formation of (NH4)2S04/NH4HS04
deposits. To achieve this objective, two
types of analyses were conducted: (1)
the thermodynamics of NH3 reactions
were evaluated to identify which
reactions can occur in the preheater;
and (2) a kinetic analysis was conducted
to identify the factors which control the
rate of (NH4)2SO4/NH4HSO4 formation
in the preheater.
Thermodynamics
The formation of both (NH4)2S04 and
NH4HSO4 are temperature dependent
reactions which proceed as a gas
containing NH3, SO3, and H2O is cooled.
The temperature at which these
reactions begin to occur will depend
on the concentrations of reactants in
the gas phase and on the product
formed.
To quantify the relationship between
reactant concentrations, products
formed, and temperature, a thermo-
dynamic analysis of the (NhU^SO* and
NH4HS04 formation reactions was
conducted. This analysis employed
thermodymanic principles to estimate
the equilibrium concentrations of both
reactants and products. The results of
the analysis do not imply that the
reactions which can occur will proceed
at a rapid or even measurable rate. They
do, however, identify reactions which
will not proceed under certain
circumstances.
The only significant NH3 reactions
which can occur in an air preheater are
those which form (NH4)2S04 and
NH4HS04. In addition, H2S04 can form
from the reaction of SO3 and H2O. The
thermodynamic analysis conducted as
part of this study considered only these
reactions.
Some typical results of the thermo-
dynamic analysis are present in Figure 4
which shows the fractional extent of
reaction (E) as a function of temperature
for an initial S03 concentration of 10
ppm and various initial NH3 concentra-
tions. Fractional extent of reaction is
defined here as the fraction of the
stoichiometrically limiting species that
can react for a given temperature and
inlet NHa/S03 concentrations as shown
in Table 2.
A major result of the thermodynamic
analysis is that (NH4)2S04 is the
principal product at equilibrium for all
the cases examined. This can be seen in
Figure 4 which shows initial (NH4)2S04
formation temperatures that are 20 -
40°C higher than the NH4HSO4
formation temperatures. The reason for
this is that, for a given temperature, the
change in free energy for (NH4)2S04
formation is greater than the change in
free energy for NH4HS04 formation.
Another result of the thermodynamic
analysis is that both (NH4)2SO4 and
NH4HSO4form in a narrowtemperature
range. The temperature drop required to
permit 90 percent reaction of NH3 and
S03 is approximately 30°C and
approximately 20°C for 80 percent
reaction. This is significant in terms of
air preheater operation. It means that
most deposition could be limited to a
small region of the preheater.
Kinetics
The results of the thermodynamic
analysis identify what reactions are
possible at a given temperature, but
they do not provide any information on
the rate at which the reactions occur.
Rate data can only be obtained by
experimentation and subsequent
analysis of the experimental results. As
part of this study, Jumpei Ando
conducted laboratory experiments on
NH3-S03 reactions and supplied the
results of these experiments to Radian.
A kinetic analysis of Ando's experi-
mental results was conducted which
identified possible rate limiting steps in
the formation of ammonium sulfates in
a heat exchanger. A model of
(NH4)2S04/NH4HS04 formation and
deposition was then developed and
applied to Ando's experiments.
The three phenomena which can limit
the rate of (NH4)2S04 and/or NH4HSO4
formation in a heat exchanger are:
• Chemical Reaction Rate.
• Heat Transfer Rate.
• Mass Transfer Rate.
The model developed as part of this
study incorporated several assumptions.
First, NH4HS04 was assumed to be the
only compound found in the heat
Table 1. Discharge Limits for Metal Cleaning Wastes
Emission Limit for New Sources, mg/l
Stream Pollutant
Maximum
Average
Total Suspended Solids
Oil and Grease
Copper (total)
Iron (total)
100
20
1
1
30
15
1
1
Table 2. Fractional Extent of Reaction Defined as a Function of Reaction Product
and NHa/SOa Mole Ratio
NH3/SO3 Ratio
Bisulfate Formation
Sulfate Formation
> 1
< 1
>2
<2
1 / - initial, f - final, and Yz - mole fraction of species a.
5
-------
exchanger. This assumption was made
because analysis of deposits in Ando's
experimental apparatus showed an
NHa/SOs mole ratio of 1.1 (i.e., the
deposits were 90 percent NH4HSC>4).
Second, the rate of NhUHSCU formation
was assumed to be very rapid, such that
the gas was at equilibrium with respect
to NHUHSC^ formation in all areas of the
heat exchanger. This means that
NhUHSO* can form as either a deposit
on the heat exchanger surface or as an
aerosol in the gas and the rate of
aerosol/deposit formation is a function
of the rate of heat and mass transfer in
the heat exchanger.
A model based on these assumptions
was applied to Ando's experimental
heat exchanger resulting in a prediction
of deposit and aerosol formation in the
exchanger. Figure 5 presents typical
model results and compares those
results with some of Ando's experi-
mental data. The figure shows the
fractional amount of NH3 and H2SO4
converted to NmHSO4 as a function of
distance in the experimental heat
exchanger. The solid line in the figure is
a smooth curve drawn through the data
points from Ando's experimental
results. The dashed lines represent the
theoretical predictions from the model.
The lower dashed line represents only
the amount of NH4HSO4 predicted to
form at the heat transfer surface. The
upper dashed line represents the sum of
the condensate and aerosol predicted to
form.
The modeling results in Figure 5
appear similar to those obtained in the
laboratory experiments, although- the
exact fate of the aerosols cannot be
predicted. It does appear, however, that
the assumptions made in developing
the model are valid. At an initial
NHa/SOs mole ratio of 1.0, the principal
reaction product appears to be
NhUHSCX, and the reaction which forms
this product is very rapid.
The fact that NhUHSCU is the principal
product formed is in apparent conflict
with the thermodynamic predictions
which indicate that (NH4)2S04 should be
the only compound which forms,
regardless of the initial NHs/SOa
stoichiometric ratio. Apparently, the
reaction which forms (NhU^SC^ directly
from gas phase reactants is slow
relative to both the rate at which the gas
is cooled from the sulfate to the
bisulfate formation temperature and the
rate of the reaction which produces
NH4HSO4. As a consequence, the gas is
cooled below both the (NKUbSCU and
Uj
1.0
0.9
0.8
0.7
0.6-
0.5-
0.4-
0.3-
0.2-
0.1 -
0
KEY
(NHthSO* Formation
NH4HSO4 Formation
ppm at NH3 inlet
O 700
D 50
A 30
O 10
230 220 270 200 190 180 170 160 150 140 130 120
Temperature, °C
Figure 4. Thermodynamic equilibria for ammonium sulfate and ammonium
bisulfate with 10 ppm SOs at inlet.
KEY
Q Ando's Laboratory Resulti
Theoretical Predictions
J2 Q>
2 >5 0.10 -
Condensate + aerosol
Figure 5.
O 10 20 30 40 50 60 7O 80 90 100 110 12
Distance, cm
Predicted and actual W/4//SO4 formation for 600 Ncm/sec gas
with 200 ppm each NH3 and SOa, and 160°C oil bath.
NH4HS04 formation temperatures
before appreciable quantities of
(NH4)2SO4 can form. At this point, the
formation of NHUHSC^ predominates
due to its more rapid reaction rate. This
is not to imply that no (NhUJzSCU will
form in a preheater. On the contrary,
significant quantities of (NhUfeSC^ will
form in the presence of excess NHs.
In additional experiments using' the
laboratory-scale heat exchanger, Ando
found that the composition of deposits
-------
are related to the stoichiometric ratio of
the reactants. Based on Ando'sdata and
the results of the modeling work, it
appears that the formation of ammonium
sulfates proceeds via the reaction path
shown in the following equations:
S03(g) + H20(g)5tH2S04(g) (8)
NH3(g) + H2S04(g)3±:NH4HS04(l) (9)
aNH4HS04(l) + bNH3(g);s-
(a-b)NH4HS04-(bMNH4)2SO4(l) (10)
Sulfuric acid vapor is the first product
formed, and this reaction is essentially
complete at the NimHSO* formation
temperature. NH3 then reacts with
H2S04 to form liquid NhUHSCU which
can further react with NH3 as shown in
Equation 10. The compound shown on
the right side of Equation 10 represents
a liquid solution of NH4HS04 and
(NH4)2S04.
In summary, the primary factors
controlling NH4HS04 and subsequent
(NH4)2SC>4 formation are the concentra-
tions of the gaseous reactants and the
system temperatures. For given
reactant concentrations, there is a
specific temperature above which
NH4HS04 will not form. The phenome-
non of NH4HS04 deposition is more
complex. The relative rates of heat and
mass transfer must be considered to
predict the amount of deposits.
Techniques for Minimizing
Deposition of
(NH4)2SO4/NH4HSO4
Solutions to the (NH4)2SO4/NH4HS04
deposition problem should minimize or
eliminate plugging and corrosion of the
air preheater. This can be done in two
ways: (1) modification of air preheater
design and/or operation, and (2)
modification of SGR system design and/
or operation. The following discussion
presents specific techniques for
minimizing the problem.
Modification of Air Preheater
Design/Operation
Four techniques identified for
minimizing the impacts of deposit
formation require modification of air
preheater design or operation:
• Use of Available Air Preheater
Design Options.
• Heat Cleaning of the Air Preheater.
• Flue Gas Recirculation for Selective
Formation of (NH4)2S04.
• Increased Air Preheater Operating
Temperature.
Each technique is discussed in detail
below.
Use of Available Air Preheater
Design Options
One way to limit the impact of
(NH4>2SO4/NH4HSO4 deposit formation
is to use available options when
specifying an air preheater design.
Specifically, this includes:
• Both hot- and cold-end soot
blowers.
• Increased soot blowing frequency.
• Increased soot blowing steam
pressure.
• Provisions for in-service water
washing of the preheater.
• Use of corrosion resistant material
in both the intermediate and low
temperature zones of the preheater.
• Use of combined intermediate and
low temperature heat transfer
elements.
Employing these options minimizes
any impact associated with (NH4>2SO4/
NH4HSO4 deposition. Use of combined
heat transfer elements, increased soot
blowing intensity, and in-service
washing should minimize plugging of
the preheater, and the use of corrosion
resistant material should permit reli-
able air preheater operation.
Each option will require some change
in air preheater design or operation.
Unfortunately, very little experience is
available, and the effectiveness of the
air preheater design options in limiting
deposits is uncertain. However, EPDC
has recently awarded a contract for two
Ljunstrom heat exchangers to be
installed downstream of an SCR
system. These air preheaters incor-
porate most of the modifications
recommended here; based on pilot-
scale tests, EPDC expects them to
operate without problems.2
Heat Cleaning of the
Air Preheater
The fact that (NH4)2SO4/NH4HSO4
deposits will decompose at elevated
temperatures provides the basis for a
second solution to the deposition
problem.10 This solution, termed "heat
cleaning," requires that the operating
temperature in the preheater be
elevated to the point where (NH4)2SO4
and NH4HS04 will rapidly decompose
(350 - 450°C). Periodic cleaning using
this technique should help prevent
deposit buildup.
Use of the heat cleaning technique
has several drawbacks: (1) elevation of
preheater temperatures will result in a
temporary decrease in boiler efficiency;
(2) the flue gas exiting the preheater
may need to be cooled to prevent
damaging downstream equipment due
to the high temperature of the gas
(cooling the gas may also be required to
recondense the (NH4)2SO4/NH4HS04
so these compounds can be collected in
particulate control equipment); (3) if
temperature gradients greater than
those encountered during normal
operation of the preheater occur during
heat cleaning, differential expansion
can damage the preheater; and (4) heat
cleaning will effectively remove
deposits of NH4-Fe-S04 compounds and
NH4-fly ash-SO4 compounds which can
form in the preheater.
Use of heat cleaning to minimize air
preheater deposits requires that several
modifications be incorporated in the air
preheater design. In particular, these
modifications must include provisions
for increasing the cold-end metal
temperature. Examination of air pre-
heater design characteristics indicates
that the most promising technique of
increasing the cold-end metal tempera-
ture involves eliminating the flow of air
to the preheater while reducing the flue
gas flow rate. This will result in raising
the preheater temperature to the flue
gas temperature with a minimum of
problems. Since no air flows through
the preheater during this period,
differential expansion of the preheater
wheel should be minimized. Also, since
operation at a reduced flue gas flow will
minimize the decline in boiler efficiency,
it may not be necessary to cool the flue
gas exiting the preheater with dilution
air.
Note that this technique requires the
boiler to have more than one preheater.
Also, no data are available which
indicate if heat cleaning will work or
how effective it will be. This solution is
based strictly on engineering judgement
and requires experimental work to
substantiate its feasibility. Technically,
this solution is considered possible, but
not proven.
Flue Gas Recirculation for
Selective Formation of
(IMH4)2S04
The results of the thermodynamic
analysis have identified several useful
facts concerning the formation of
ammonium sulfates. First, (NH4)2S04
forms at higher temperatures than
NH4HSO4 and exists in pure form only as
a solid. Second, at 20°C below the initial
formation temperature, the (NH4)2SO4
formation reaction can proceed to
approximately 80 percent completion.
-------
By modification of the air preheater
design, it may be possible to exploit
these facts to selectively form (NH4)2S04
and thereby minimize deposit formation.
One potential technique for selectively
forming (NH4)2SO4 employs flue gas
recirculation from downstream of the
air preheater to cool hot flue gas in a
reaction chamber. Temperatures are
controlled in the chamber so that
(NH4)2SO4 (but not NH4HS04) is formed.
The solid (NH4)2SO4 should then pass
through the preheater and be collected
by paniculate control equipment.
Two preheater designs can be used
with flue gas recirculation: split and
single. The split design divides a single
air preheater into two preheaters in
series, separated by a reaction chamber.
The first air preheater is operated such
that the cold-end metal temperature is
maintained above the (NH4)2S04 forma-
tion temperature. Cooled flue gas is
recirculated from downstream of the
second preheater and injected into the
reaction chamber. The recirculation
rate is controlled to cool the flue gas in
the reaction chamber below the forma-
tion temperature of (NH4)2SO4 but not
that of NH4HSO4. This should result in
the formation of solid (NH4)2S04 only.
An alternative to the spl it a ir preheater
design is a single air preheater in which
flue gas is recirculated from the cold-
end to a reaction chamber upstream of
the hot-end. The recirculation rate is
controlled to maintain the flue gas
temperature entering the preheater
below the formation temperature of
(NH4)2S04 but above that of NH4HS04. In
principal, this technique is identical to
use of a split air preheater. However, in
practice, use of a single preheater
would require recirculation of more flue
gas and a larger preheater.
Flue gas recirculation was identified
as a possible solution to the deposition
problem, based strictly on the thermo-
dynamic analysis of (NH4)2S04/
NH4HSO4 formation. However, there
are factors which make the feasibility of
this solution uncertain. First, there are
no data on the rate at which (NH4)2SO4
forms from gas-phase reactants. All
experimental data, including analysis of
air preheater deposits, indicate that
NH4HSO4 is the first product to form
from gaseous reactants. The results of
Radian's kinetic analysis confirms that
the NH4HSO4 formation reaction occurs
very rapidly, but the (NH4)2S04 formation
rate could not be determined. Therefore,
it may be impossible to selectively form
(NH4)2S04 with a realistically sized
reaction chamber residence time (i.e., 1
to 2 sec).
A second problem associated with
flue gas recirculation is one of process
control. Changes in fuel sulfur content
or excess air to the boiler may change
the S03 concentration, while changes
in flue gas flowrate and NOX concentra-
tion can change the concentration of
NH3 which is emitted from the reactor. If
the air preheater is designed for one
range of NH3-S03 concentrations and a
significant change in these concentra-
tions occurs, it will be difficult to control
the system to limit (NH4)2S04 formation
to the reaction chamber and to prevent
NH4HS04 formation. The principal
reason for this is that both NH3 and SOa
must be measured continuously to
adjust the recirculation rate to the
reaction chamber and thus control
formation of (NH4)2S04. Unfortunately,
both NH3 and S03 are difficult to
measure continuously.
The problems associated with use of
flue gas recirculation represent signifi-
cant technical obstacles. No data are
available to indicate whether it is
possible to selectively produce (NH^^CU
In addition, the control problems
associated with changing NH3-S03
concentrations will be difficult to
resolve. Therefore, the technical feasi-
bility of using flue gas recirculation to
prevent deposit formation is considered
low.
Increased Air Preheater
Operating Temperature
The most direct way to minimize the
deposition of ammonium sulfates is to
increase the air preheater operating
temperature above that at which
NH4HS04 forms. This should minimize
plugging and corrosion since any
(NH4)2SO4 and/or deposits which form,
will be non-corrosive. These deposits
will occur in the extreme cold-end of the
preheater and should be easily removed
by soot blowing.
Modification of the air preheater
design to permit operation at a higher
temperature should be relatively simple.
The principal change required is a
reduction in thermal efficiency of the
preheater. Efficiency can be reduced by
reducing the size of the preheater so the
cold-end temperature is above that of
NH4HSO4 formation.
The major impact of increasing the
cold-end metal temperature will be to
decrease boiler efficiency. In addition,
there may be some impacts on down-
stream equipment and possibly an
environmental impact due to gas-phase
NH3 and SO3 emissions. For these
reasons, it may be necessary to cool the
flue gas exiting the preheater below the
NH4HS04 formation temperature.
Modification of SCR
Design/ Operation
A second approach to solving the
(NH4)2S04/NH4HSO4 deposition problem
is to modify the SCR system design
and/or operation. The intent of
modifying the SCR system is to reduce
the NH3 emissions from the reactor,
thereby reducing the formation of
ammonium sulfates in the preheater.
There is basically one way to modify the
SCR system so that the NH3 emissions
are reduced. Additional catalyst can be
used and the NH3 injection rate can be
lowered while maintaining the desired
NOX removal level, thus reducing the
NH3 emissions.
The design of an SCR system is
influenced by a trade-off between the
quantity of catalyst in the reactor, the
NH3 injection rate, and the NOx removal
efficiency of the system. By increasing
the quantity of catalyst and simulta-
neously reducing the NH3 injection rate,
the NH3 emissions can be reduced while
maintaining a constant NO* removal
efficiency. This relationship is quantified
in Table 3. As shown, NH3 emissions for
the base case are 30 ppm. The use of
about 30 percent additional catalyst
(Case 2) can reduce NH3 emissions to
about 10 ppm while 70 percent more
catalyst (Case 3) should effectively
eliminate NH3 emissions.
The most significant impact of
increasing the amount of catalyst in the
reactor will be to increase the capital
investment and operating costs for the
SCR system.
Table 3. Effect of Increased Catalyst on
Case Base
Relative Catalyst Amount 1.0
NH3/NO* for 90% Removal 1.0
NHs Emissions, ppm 30
NH3 Emissions3' 11
7
1.10
0.98
20
2
1.31
0.95
10
3
1.71
0.92
0
8
-------
Estimated Costs for Proposed
Solutions to the Deposition
Problem
Cost estimates were prepared to
determine the incremental capital and
first year annualized costs for each
solution identified by this study. The
basis for these estimates is defined in
Table 4. Incremental costs were
determined for each modification
required to implement a proposed
solution using the flowrates, tempera-
tures, and other characteristics of the
system defined in Table 4. The following
discussion presents the results of the
cost estimates, along with a specific
definition of the modifications included
in making those estimates.
Air Preheater Design Option
Costs
Table 5 compares design options
added to the preheater to minimize
deposit formation with an unmodified
preheater design. Table 6 contains cost
estimates for the preheater design
options. As shown, soot blowing costs
will increase significantly due to an
increase in both the number of soot
blowers and the soot blowing frequency.
Also, material costs will increase
significantly due to the corrosion
resistant material used in the inter-
mediate temperature zone of the
preheater.
Air Preheater Heat
Cleaning Costs
Table 7 gives estimated incremental
capital and first year annualized costs
for air preheater heat cleaning. The
costs associated directly with heat
cleaning include fuel costs which result
from raising the flue gas exit
temperature from 150 to 350°C for 30
minutes per day at 10 percent of the
maximum flue gas flow rate.
Flue Gas Recirculation Costs
Table 8 gives estimated costs for the
two flue gas recirculation options
defined in this study. The capital costs
shown include the incremental costs for
an additional or larger air preheater,
recirculation fans, ducts for flue gas
recirculation, and a reaction chamber.
Included in the annualized cost
estimates is a heat credit which results
from assuming that flue gas temperature
at the air preheater exit can be reduced
from 150 to 115°C due to neutralization
of H2S04 in the gas.
Table 4. Basis for Estimating Incremental Costs of Proposed Solution
Parameter
Base Value
Boiler Characteristics
• Size
• Thermal Efficiency
• Number of Air Preheaters
• Flue Gas Flowrate
• Heat Rate
• Operating Factor
Flue Gas Characteristics
• Reactor Inlet
—/VOx concentration
—NHz concentration
—SOa concentration
—temperature
• Reactor Outlet
—/VOx concentration
—NH3 concentration
—S03 concentration
• Air Preheater Outlet
—/VOx concentration
—NH3 concentration
—SOa concentration
—NHtHSO* concentration
—temperature
SCR Characteristics
• Relative Quantity of Catalyst
• /VOX Removal Efficiency
• NHa/NO* Mole Ratio
Air Preheater Characteristics
• Type/Size
• Number of Soot Blowers
• Soot Blowing Frequency
• Soot Blowing Steam Pressure
Cost Characteristics
• Year
• Fuel Cost
• Air Preheater Capital Cost
• Type of Installation
• Capital Recovery Factor
500 MWe
88%
2
620 kg/sec (82.0OO Ib/min)
9.5 MJ/kWh (90OO Btu/kWh)
7000 hrs/yr
350 ppm
350 ppm
10 ppm
350°C
35 ppm
30 ppm
10 ppm
35 ppm
20 ppm
0 ppm
10 ppm equivalent
150°C
1.0
90%
1.0
Ljungstrom Tri-Sector/Size 31
1 / air preheater (cold-end)
3/day
1.48 MPa (200 psigj
1979 Capital/1980 annualized3
$2.37/GJ
$2,600,000
New
14.6% of capital investment
^Annualized costs include annual operating and maintenance costs (including fuel}
plus capital-related charges such as depreciation, return-on-investment, and
interest-on-debt.
Table 5. A Comparison of Modified Preheater with Basic Preheater
Design Specifications
Specification
Basic
Design
Modified
Design
Number of Soot Blowers
Soot Blowing Frequency
Soot Blowing Steam Pressure
Materials of Construction
- intermediate temp zone
- low temp zone
1 - Cold end only 3 Cold end
3 Hot end
6/day
1.82 MPa
3/day
1.48 MPa
Light gauge
carbon steel
Heavy gauge
carbon steel
Combined intermediate
and low temperature zones
304 stainless steel
-------
Increased Air Preheater
Operating Temperature Costs
Table 9 shows incremental first year
annualized costs for increasing the
preheater operating temperature above
the NH4HSO4 and (NH4)2SO4 formation
temperatures. For this study, no change
in capital costs was considered since no
a'dditional capital expenditures are
required. The annualized costs given in
Table 9 are based on increased fuel
costs which result from raising the flue
gas exit temperature to 230°C and
250°C for prevention of NhUHSO* and
formation, respectively.
Increased Catalyst Costs
Table 10givesthe incremental capital
and annualized costs incurred to reduce
NH3 emissions to 1 0 ppm by increasing
the quantity of catalyst in the reactor.
The estimates of Table 10 include
incremental capital charges for the
initial catalyst charge, higher reactor
and annual costs for catalyst replace-
ment, and a credit for reduced NH3
ammonia consumption.
Conclusions
The major conclusions of this study
are:
• NH3 leakage from SCR reactors can
be a problem for an air preheater
downstream of the reactor. How-
ever, operating experience indicates
that NH3 concentrations below 10
ppm at the air preheater entrance
do not result in serious deposition
problems. This is probably due to
the fact that at low NH3 concentra-
tions, ammonium sulfates form in
the cold end of the preheater where
soot blowing equipment effectively
removes deposits.
• The effects of the deposition problem
are limited to plugging of the
preheater and corrosion of pre-
heater materials in the intermediate
temperature zone. The ability of the
preheater to transfer heat should
not be significantly impaired by
deposit buildup. In addition, normal
corrosion in the extreme cold end of
/he preheater will be reduced due to
neutralization of SOs-HzSC^ by
NH3.
• No significant environmental
problems will result from washing
deposits from the air preheater. The
NH3 levels in the preheater wash
water will be lower than those
typically encountered in power
plant metal cleaning wastes. A
TableG.
Option
Estimated Capital and First Year Annualized Costs for Air Preheater
Design Options
Incremental Capital
Costs. $1000's
Incremental Annualized
Costs, $10OO's
Soot Blowing Modifications
In -Service Washing
Corrosion Resistant Material
TOTAL
158
0
1376
1534
233
18
201
452
Table 7. Estimated Capital and First Year Annualized Costs for Air Preheater
Heat Cleaning
Option
Incremental Capital Incremental Annualized
Costs, $1000's Costs, $1000's
Heat Cleaning
Corrosion Resistant Material
TOTAL
370
1376
1746
70
201
271
Table8.
Method
Estimated Incremental Capital and First Year Annualized Costs for
Flue Gas Recirculation
Incremental Capital
Costs, $1000's
Incremental Annualized
Costs, $10OO's
Single Preheater
Split Preheater
9690
6223
857
(198}
Table9.
Option
Incremental First Year Annualized Costs for Increasing Air Preheater
Operating Temperature
Cold End Temperature
°C
Annualized Costs
$1000's
Prevent NHJiSO* Formation
Formation
230°C
250°C
3098
4270
Table 10. Estimated Incremental Capital and First Year Annualized Costs for an
Increased Catalyst Charge
Catalyst Life
years
Capital Costs
$1000's
Annualized Costs
$10OO's
1
2
6054
6054
3640
2248
slight increase in waste treatment
costs may result due to the increased
volume of wash water. The magni-
tude of this cost increase is site
specific and depends on the method
of waste water treatment employed.
The problems associated with
deposition of (NhUJaSCVNhUHSCU
can be minimized or eliminated by
several techniques. A relative
technical and economic ranking of
these techniques is given in Table
11. As shown, solutions with the
lower technical feasibility have the
lower costs, while the solutions
with the higher technical feasibility
incur higher costs.
Based on the results in Table 11, it
appears that use of available air
preheater design options is the
optimum solution to the deposition
problem both technically and eco-
nomically. However, the solutions
with low technical feasibility could
result in lower annualized costs
and thus merit further investigation.
10
-------
Table 12 gives the annualized costs
for various solutions to the deposi-
tion problem as a percentage of the
annual revenue requirements for
SCR. As shown the cost impact of
the solutions ranges from a 1.6
percent reduction to a 30 percent
increase in annualized costs.
References
1. Ando, Jumpei. Ammonium Bisulfate
Problem with NOX Reduction by
Ammonia. Private communication.
March 1979.
2. Jones, G.D. Selective Catalytic
Reduction and NOX Control in Japan.
EPA-600/7-81-030 (NTIS PB81-
191116), March 1981.
3. Ando, Jumpei. NOX Abatement for
Stationary Sources in Japan. EPA-
600/7-79-205 (NTIS PB80-113673),
August 1979.
4. Castellan, G.W. Physical Chemistry.
Addison-Wesley Publishing Com-
pany, Inc. 1971.
5. MacDuff, E.J. and N.D. Clark.
"Ljungstrom Air Preheater Design
and Operation." Combustion, pp. 24-
30. March 1976.
6. Campbell, H.H. CE Air Preheater.
Personal communication with J.M.
Burke. June 25, 1980.
7. Nakabayashi, Y. and K. Mouri. Test of
NH3/SOX Compound Deposit Pro-
blems on Air Preheater at Coal-Fir-
ing Boiler. Electric Power Develop-
ment Co., Ltd., Thermal Power De-
partment. Tokyo, Japan. 1977.
8. Karlsson, J. and S. Holm. "Heat
Transfer and Fluid Resistances in
Ljungstrom Regenerative-Type Air
Preheaters." Transactions of ASME;
65. pp. 61-72. 1943.
9. U.S. EPA, Office of Water and Waste
Management. Development Docu-
ment for Proposed Effluent Limita-
tions Guidelines, New Source Per-
formance Standards and Pretreat-
ment Standards for the Steam
Electric Point Source Category. EPA-
440/1-80-029b (NTIS PB81 -
119075), September 1980.
10. Kiyoura, R. and K. Urano. "Mecha-
nism, Kinetics, and Equilibrium of
Thermal Decomposition of Ammo-
nium Sulfate." Industrial Engineer-
ing Chemistry, Vol. 9, No. 4. pp. 489-
494. 1970.
11. Maxwell, J.D., et al. "Preliminary
Economic Analysis of NOx Flue Gas
Treatment Processes." EPA-600/7-
80-021 (NTIS PB80-176456);
February 1980.
Table 1 1 .
Technical and Economic Ranking of Proposed Solutions to the
Deposition Problem
Solution
Air Preheater Design Options
Heat Cleaning
Flue Gas Recirculation
- Single Preheater
- Split Preheater
Increased Cold-End Metal Temperature
Increased Catalyst/ Decreased NHs/NO*
Estimated Incremental
Relative Technical First Year Annualized Costs
Feasibility* 1000's
Intermediate
Low
Low
Low
High
Intermediate
452
271
857
1198)
3,098
3,668*'c
*These are somewhat subjective and based on engineering judgment,
B7/7/s cost is based on reducing NHs emissions to approximately 10 ppm.
cAssumes a 1-year catalyst life (current vendor guarantee for coal-fired applications}.
Table 12. Estimated Increase in SCR Annualized Costs for Proposed Solutions to
the (NH^iSO^/NHd-ISOA Deposition Problem
Solution Increase (Decrease} in SCR Costs, %
Air Preheater Design Options
Heat Cleaning
Flue Gas Recirculation
- Single Preheater
- Split Preheater
Increased Cold-End Metal Temperature
Increased Catalyst/Decreased NH3/NO*
3.7
2.2
7.0
(1.6)
25.4
30.0
J. M. Burke and K. L Johnson are with Radian Corporation, Austin, TX 78766.
J. David Mobley is the EPA Project Officer (see below).
The complete report, entitled "Ammonium Sulfate and Bisulfate Formation in
Air Preheaters," (Order No. PB 82-237 025; Cost: $21.00, 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:
Industrial Environmental Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
11
-------
«
o
5'
3
0)
^
O
•(*
en
M
O)
oo
00 <0
ni c
C5
m
m > -o m
5111.
i 3 (D -i
8Sg.§
oi 03
3 (D
0)
------- |