Proceedings
of the
Symposium
on the
Current Status
of the
problem
and its control...
Thursday, April 26, 1973
33 W. 42nd Street,
New York, N.Y.
Sponsored by
REGION II
MIDDLE ATLANTIC CONSORTIUM ON
AIR POLLUTION (MACAP)
Funded by
Manpower Development Office
Environmental Protection Agency
Symposium Chairman
Elmar R. Altwicker, Ph.D.
Bio-Envir. Engineering Division
Rensselaer Polytechnic Institute
Troy, N.Y. 12181
518- 270-6554
-------
Proceedings of the
Symposium
on the Current Status of the
NO problem and its control
x
This Symposium was held on
Thursday, April 26, 1973
CUNY Graduate Center
33 W. 42nd Street
New York, New York
Sponsored by
REGION II
MIDDLE ATLANTIC CONSORTIUM ON
AIR POLLUTION
(MACAP)
Funded by
Manpower Development Office
Environmental Protection Agency
-------
CONTENTS Pa§e
Introduction by Elmar R. Altwicker, Symposium Chairman 1
Current Understanding of NO -Formation in Mobile and Stationary 2
Combustion Sources by Irving Classman, Center for Environmental
Studies, Princeton University.
NO -Control and Problems in Mobile Sources by William Balgord, 27
•y1
New York State Department of Environmental Conservation.
Technical and Economic Approaches of Power Plants towards NO^- 28
Control; Air Quality Benefits and Costs from Utility NO^-
Regulations by Peter C. Freudenthal, Consolidated Edison Company
of New York.
Control of NO -Emissions from and Control Equipment for Stationary 47
Sources; Future Approaches to Stationary Source Control of Stack
Gas NO by Aaron J. Teller, Teller Environmental Systems of New York.
x -
NO -Control by Wet Scrubbing by Robert S. Kapner, Engineering 53
Division, Cooper Union of New York.
EPA-View of Stationary and Mobile NO -Source Control by Conrad 101
Simon, Air Programs Branch, Environmental Protection Agency of
New York.
ii
-------
Introduction
This volume contains all but one of the papers that were presented at the
MACAP-symposium on NO -control. The problem of NO -control seemed - and still is -
X Js.
timely and was chosen as the subject of the first MACAP-symposium. In addition
to the technical information transmitted, this and future symposia are intended
to foster the interaction between control officials, industrial concerns, consul-
tants, and academicians on timely air pollution topics.
Two important topics - atmospheric chemistry and health effects of NOX -
were purposely excluded from the symposium, in order to permit greater treatment
in depth of the control aspects. It was therefore of considerable interest to the
chairman that the panel discussion - not included in these proceedings - elicited
a great deal of comment from the audience on standards and their relation to health
effects; another indicator of the breadth and complexity of the NO -problem.
^C
The organization of this symposium was greatly simplified by the able assist-
ance of John Bove, Paul DeCicco, Carl W. Kreitzberg, and Gerald Palevsky. All
but one contributor cooperated in making full length, referenced manuscripts
of their presentations available.
Elmar R. Altwicker
Symposium Chairman
Rensselaer Polytechnic Institute
Troy, New York 12181
1.
-------
CURR-EA'7 UNDERET/vl.uING OF NO FORMATION ^
Tfi MOI3ILE AND STATIONARY COMBUSTION SOURCES
by
2
Irvin Classman
Center for Environmental Studies
Princeton University, Princeton, N. J.
I. INTRODUCTION
Current interest in predicting NO „ emissions from
jTi.
mobile and stationary sources has led to the formulation
of various kinetic models for the formation of nitric
oxide (NO) in combustors. The formation of nitrogen
(NO-) from nitric oxide has received less attention;
however, there does appear to be general agreement as to
the main kine tic step in this conversion; simply NO + OH
-»• NO- + H. There are two principal sources of nitric oxide
in the combustion of conventional fuels: (1) oxidation of
atmospheric (molecular) nitrogen and (2) oxidation of
nitrogen containing compounds in the fuel (referred to as
fuel nitrogen).
Although the major interest of concern here is
the kinetic routes for the formation of NO, of greatest
concern from a practical point of view is the total NO
The author's work in kinetics is supported by the Air
Force Office of Scientific Research under Grand 69-1649
and the Environmental Protection Agency under Grant R-801194
2
Director and American Cyanamid Professor of Environmental
Studies.
This paper developed from the original framework given in
Reference 1.
2.
-------
emission from a particular device. Lest one thinks
kinetics are the dominant factor, it is best to ini-
tially discuss the NO emission problem in the context
X
of flames.
II. THE ROLE OF FLAME STRUCTURE
If one premixes a gaseous fuel with air so that
the system is completely homogeneous and if the mixture
ratio falls within the flammable range, then given an
ignition source a flame will propagate through the
• .u 4
mixture.
It is possible to calculate from simple thermo-
dynamics the maximum temperature reached in such flames
when one assumes that the conditions approach adiabatic
ones. Such calculations can be made with sufficient
precision such that the calculated adiabatic flame
temperature is very close to the flame temperature
measured experimentally. l.ideec , one can readily cal-
4
Under certain experimental conditions related to con-
finement, the flame, a subsonic wave (deflagration), will
undergo a series of transitions in which it will repeatedly
become a detonation wave (supersonic). The temperatures
and pressures generated by detonation waves are substan-
tially higher than a deflagration wave (flame), the amounts
of NO formed would be much greater as well. However,
detonations play no significant role in mobile and sta-
tionary power plant combustion processes and will not be
discussed further.
3.
-------
culate the adiabatic flame temperature as a function
of fuel/air ratio(2). The variation, which is similar
for all fuels, is shown on a relative basis in Figure 1,
in which the temperature is plotted as a function of <{>,
the equivalence ratio. The equivalence ratio is defined
as that ratio of the actual fuel/air ratio to the value
at the stoichiometric condition — the condition under
which there is precisely enough oxygen to burn all the
carbon atoms present to carbon dioxide and all the hydro-
gen atoms present to water vapor. Large values of
correspond to a (fuel) rich condition and small values to
a (fuel) lean condition. As a very simple explanation,
the temperature drops on the rich side because there is
insufficient oxygen present to burn all the carbon (or
carbon monoxide) and hydrogen to their fully oxidized
states which gives the greatest energy release. The
drop on the fuel lean side can be considered as a dilutive
effect by the excess oxygen.
At each temperature it is possible to calculate
the equilibrium concentration of NO given by the equation
II 1/2 N2 + 1/2 O ^?NO
from the equilibrium concentration of formation, K f
P r £ i
4.
-------
K = PNO _ XNO
P'f 1/2 1/2 1/2 1/2
PN P0 XN X0
IN U N U
where p is the partial pressure and x is the mole fraction
of the component in the complete combustion system. The
K f is readily found in thermodynamic tables or calculated
as a function of temperature from the relationship
AF° = -RT In K .p
Pf f
where AF° is the standard state free energy charge of the
(2)
equilibrium reaction of formation written above.
One should not infer from the manner in which the
equilibrium reaction is written that NO forms by the direct
reaction of N2 and O~. In fact as will be discussed in
detail later the route is through oxygen atom attack on
the nitrogen molecule.
Since K ,; is a function of temperature alone, it
P'T
is not surprising that the equilibrium concentration of
NO formed in homogeneous premixed fuel/air flames would
be expected to peak around stoichiometric just as the
temperature does, provided of course that time be allowed
for equilibrium condition to prevail.
It should be underscored that the utility of
making estimates of NO formation by an equilibrium calculation
5.
-------
exists only for premixed gaseous fuel/air flames. Most
practical combustion devices (other than laboratory
burners) utilize condensed phase fuels or gaseous fuel
jets under conditions which the fuel and air first mix
and then burn.5 Flames which exist under such conditions
are referred to as diffusion flames because the rate of
burning is controlled by rate of diffusion of fuel into
air and vice-versa. Figure 2a represents the situation.
When a liquid droplet burns the flame created releases
heat which diffuses back to the fuel to evaporate it,
then the fuel diffuses towards the flame where it is con-
sumed. The air is similarly heated and oxygen diffuses
towards the flame where it is consumed. Also shown in
Figure 2b is the burning of a gaseous fuel jet. In this
case the shape and position (length) of the flame is governed
by the fuel injection rate; nevertheless, once the flame
exists, fuel and oxygen diffuse towards the flame where
they are consumed. What is so very important is the fact
that it has been well established that the fuel and oxi-
(3
dizer flow towards each other in stoichiometric proportion
and because the reaction rates in this case are so much
faster than the diffusion rates, the reactions are con-
sumed in a very narrow zone. Since the flame is like a
* A large furnace injecting gaseous fuel or an aircraft
gas turbine burning natural gases are specific examples.
6.
-------
sink for fuel and oxygen, it is not difficult to accept
intuitively that they approach each other at a ratio of
rates equal to the stoichiometric index. Thus in a drop-
let burning situation where it is possible to specify an
overall mixture, two temperature zones exist — one, a
zone around each droplet (or group if they are close
together) and the other the final equilibrium temperature
corresponding to overall fuel-air mixture ratio. Thus
even though a system such as a diesel engine may have a
very lean fuel-air ratio, it will produce more NO than
expected because each droplet burns so that the flame
around it corresponds to the temperature at stoichiometric,
The consideration of the character of the flame
is the major one when concerned with the NO emissions.
J x
It is not surprising then that most research towards
reducin j NO emissions in devices such as aircraft gas
J^
turbines is directed toward prevarporizing the fuel and
premixing it with the air prior to injection into the com-
bustor can.
III. KINETIC ROUTES
(4)
Zeldovich was the first to propose that the
mechanism of NO formation can be represented by the
sequence of reactions
The units of all specific reaction rate constants, k,
values are cm, cal, °K, mole, sec.
7.
-------
4
III. 0 + N9 •*> NO + N k=1.4x!0exp -75800/RT
£t
2
IV. N + 02 -* NO + O k=6.4x!04exp -6280/RT
For any kinetic calculation the reverse reactions, although
not of prime importance, should be taken into account.
The specific reaction rate constants, k, are taken from
Reference 1. The constants for the reverse reaction can
be determined from these and the respective equilibrium
constants.
There has .appeared to be general agreement that
one could predict NO formation in flames based on the
Zeldovich mechanism plus the addition of the reaction
V. N + OH ->- NO 4- H k=2.8x!013
Since this reaction involves two reactive radicals whose
steady state concentrations will always be very low, it
would not normally be of significance. However, emis-
sions standards in NO have been set so low that it is
X
not wise to ignore the possible contribution from this
reaction. Results from several analytical studies of
NO formation in combustion processes have shown that
for most practical situations that radicals involving
N20 or NO and the recombination
VI. N + O + M (third body) •> NO + M
8.
-------
do not play a significant role in NO formation.
Considering that Reaction III must be very slow,
most investigators believed that the major portion of NO
forms in the post flame zone of the combustion process.
This zone is that area which exists after the major
energy release reactions have been completed and physi-
cally corresponds to the area past the luminous flame zone.
In describing Bunsen burners it is usually referred to as
the burned zone -- that zone which 'has the very faint
dull red emission.
The basis for this belief was reinforced by the
very interesting calcualtions of Martenay. Some of his
typical results are shown in Figure 3. The system he con-
sidered is in essence a constant temperature (and pressure)
flow system in which the fuel and air are considered to begin
reaction at some zero time. At t=10 sec the system can be
considered to have the composition as given by the left
hand ordinate scale. Thus the results would be applicable
to any fuel-air combustion which would have this initial
_7
(10 sec) composition. What the curves in Figure 3 show
so very nicely is that all the constituents which relate
to energy release have formed quite rapidly and appear to
reach their equilibrium values by at most 2x10 sec.
However, at this point the NO is less than one molar part
9.
-------
per million. Even at 10 2sec, the right hand ordinate,
the NO is a factor of 10 less than its equilibrium value,
which is specified on this ordinate as well. Thus it
would appear, from results such as these, possible to quench
the combustion reaction in steady flow devices prior to
NO formation and still obtain maximum energy release.
Indeed there is a great deal of current development work
whose objective is to establish this concept as a practical
scheme.
Perhaps it would be wise to comment here on
another development effort to reduce NO emissions. This
one concerns the injection of water into the combustion
zone. The addition of water has two effects -- one to
reduce the temperature and thus the oxygen radical .forma-
tion and ratio of Reaction I, and, two, to scavenge the
oxygen atoms by the route
VII. O + H20 -> 2 OH
Since hydroxyl radicals (OH), for all intents and purposes,
do not react with molecular nitrogen to form NO, the
scavenging could be effective. It is quite apparent that
the first effect is most dominant.
IV. "PROMPT" NO
In several recent NO formation studies, rates have
10 .
-------
been found to exceed that given by the post combustion
mechanism. Experimental measurements of NO concen-
trations made in the post flame zone gave values which
when extrapolated back to the flame position indicated
concentrations of NO in the flame that would not be
considered negligible in the context described earlier
and rates larger than those extrapolated from the Zeldo-
vich mechanism in which the equilibrium values of oxygen-
atom concentrations were used. Fenimore referred to
the NO formed in the flame as "prompt" NO and concluded
that reactions other than Reactions III-IV must be taking
place.
The question of whether the Zeldovich mechanism
and Reaction V completely accounts for the total NO forma-
tion from atmospheric nitrogen in combustion systems is
the major unsolved matter in thJs area. This point will
be discussed in the subsequent paragraphs.
To put the question in the proper context it is
(9)
best to discuss some related work. Bowman and Serry
for example, studied nitric oxide formation during shock-
induced methane combustion. The pressure and temperatures
created behind a shock give the gas a flow reactor character
in that reaction can be considered to take place at a
constant pressure in a constant temperature bath. This
11.
-------
Thus, it is argued that "prompt" NO comes about from a
large oxygen radical concentration in the flame zone
and that the intuitive concept that one could use equi-
librium oxygen radical concentrations to gain an estimate
of the initial NO formation is misleading.
Actually it should be realized that Bowman has
proven simply that the modified Zeldovich mechanism only
suffices for shock induced combustion and that it does
not appear necessary to include any other reactions. This
physical situation is different from flames in which there
is a very definite exponential rise of temperature from
the ambient condition to the flame temperature. Further
there are diffusion effects in flames which must be ac-
counted for in making kinetic estimates. Thus these
calculations necessary to make an exact analytical calcu-
lations of the kinetics in flames where species and
energy diffusive are involved are much too complex con-
sidering the number (*<34)of specific reaction steps'
involved in the overall combustion process.
Although there has been a tendency to accept the
oxygen radical overshoot as an explanation of the "prompt"
NO, recent work in Germany ' ' substantiates some of
Fenimore's early postulates and reopens the thinking with
respect to other reactions -- particularly in fuel rich
12.
-------
situation is achieved by using methane and oxygen in a
large excess of argon as the initial mixture. The con-
centration profiles measured were compared with detailed
kinetic calculations in which the time rates of change
of species concentrations and thermodynamic properties
during the reaction were determined by numerically inte-
grating the coupled reaction kinetics, state and energy
equations. The experimental and analytical results are
depicted in Figure 4. There are several interesting
and important observations to be made from this figure.
The initial rate of NO formation exceeds the rate later
in the reaction -- for both lean and rich mixture. The
slope at any point in the curve is an indication of rate
of formation of NO. However, the NO attains its equili-
brium value much more rapidly in rich mixtures than in
lean micutres. But, most importantly, in regard to the
major r^int being discussed in this part of this paper,
the experimentally measured concentrations agree with
those of the analytical calculations. Bowman and Seery
used only Reactions III-IV in their analytical scheme
and Bowman, in essence, has concluded that the prompt
NO could arise from the large overshoot of oxygen radical
concentration over its equilibrium which was found. This
same overshoot is very graphically depicted in Figure 3.
13.
-------
flames. Just and his coworkers^14 ' made, extensive
experimental determinations of NO formation in various
hydrocarbon flames. From these results they determined
the "prompt" NO. They show that for the hydrocarbon-air
system the "prompt" NO increases with equivalence ratio,
peaks at about an equivalence ratio of 1.4 and then
drops off sharply. Perhaps the most conclusive experi-
ments of Just, et al., were those in which they
controlled the temperature of. various propane-oxygen
flames by adding nitrogen diluent. They then plotted
"prompt" NO as a function of temperature for various
equivalence ratios. These results are shown in Figure 5
taken from Just, et al. In this figure A is the equiva-
lent ratio on the basis fuel to oxygen. Just makes
one very interesting speculation from Figure 5 and these
will be reinterpreted by this writer. The plots in
Figure 5 seem to indicate that at low temperatures the
slopes of the correlating lines are small and that at
high temperature the slopes are large. The change in
slope is most clearly seen for X = 1.1 and 1.2. If one,
as a first approximation, interprets the slope as an
activation energy (EA) or being determined by the E of
a particular reaction, then he can estimate that at low
temperatures EA = 12-16 kcal/mole and at high temperatures
14.
-------
E = 65 kcal/mole. This later value is of the same order
f\
of magnitude as that for Reaction III. The very interest-
ing speculative conclusion is that at high temperatures
the Zeldovich mechanism prevails and that at low tempera-
tures some other reaction mechanism is effected. Further,
the actual stoichiometry is not of total significance,
but the specific temperature which prevails is.
The German workers also measured HCN concentration
throughout their flames. They found that the HCN concen-
tration varied as a function of equivalence ratio, but
somewhat opposite to that of the "prompt" NO. The HCN
concentration remained at low levels until the equiva-
lence ratio which "prompt" NO dropped off sharply, then
the HCN concentration rose sharply. They conclude that
the HCN is an intermediate in the formation of NO and
when whatever is oxidizing the HCN disappears, NO no
longer forms and the concentration of HCN rises sharply.
Fenimore previously had postulated the possibility of
HCN formation in fuel rich flames from the step
VIII. CH + N9 ->- HCN + H
£*
V. FUEL NITROGEN
Very little specific information exists on the
kinetics of formation of NO from fuel nitrogen. The
existing data suggest that the conversion to NO from fuel
15.
-------
nitrogen is very rapid and occurs on a time scale compara-
ble to the combustion reaction. This fact is significant
in that it precludes the use of quenching to prevent NO
formation. One simply could not get the energy release
and stop the NO formation. The effect of fuel nitrogen
was dramaticallly demonstrated in the early experiments
of Matin and Berkman. ' They measured the NO emissions
from an oil-fired laboratory furnace using a base oil.
They repeated the experiment with 0.5% pyridine added to
the base oil. The pyridine additive increased the NO
emissions by a factor of five.
Various experiments in which fuel nitrogen was
introduced into flames by different techniques show that
the NO increases rapidly to values significantly larger
than its equilibrium value in the flame zone and decreases
in the post flame zone. ( 3/17) iowever, the rate of dis-
appearance in the post flame zone is relatively slow in
lean flame, but appears to be quite rapid in rich flames.
In lean flames containing fuel nitrogen it is very likely
that NO values greater than equilibrium will be found.
The destruction reactions for NO simply are very slow.
The steps one can think of are
0 + NO + N + 02
NO + NO + N20 + O
NO + RN •* •••-*• N
16.
-------
all of which are slow compared to the stay time of a fluid
element in a flame.
The general observations of NO formation from fuel
nitrogen are consistent with a mechanism in which the NO
is produced by the rapid reaction of an oxygen atom (or
hydroxyl radical) with the parent compound or with a nitrogen
containing fragment.
Bowman ' proposes that the partial equilibrium
concentration of nitric oxide is related to the fuel nitrogen
concentration by assuming the following reactions are
equilibrated:
IX fuel nitrogen = N~
X N2 + 2 0 = 2ND
The oxygen concentration used in Reaction X is the
partial equilibrium value calculated from
XI H + 02 = OH + 0
XII 0 + H2 = OH + H
XIII OH + H2 = H20 + H
Use of these reactions, together with the partial
equilibrium assumption, thus permits the calculation of
the maximum NO concentration in the flame zone without
knowledge of the detailed kinetics of the process by which
17.
-------
fuel nitrogen is converted to nitric oxide. It should
be emphasized that Reactions ix and X are not the kinetic
route in the fuel nitrogen case, but only a means of
calculating the maximum NO formed in the combustion zone
through an equilibrium technique. In calculating the
maximum NO in this situation, one is ignoring the reactions
which would remove NO. Recall these removal reactions are
thought to be very slow. Thus in conclusion, one can
generally assume that the NO formation from flames in
which the fuel contains bound nitrogen is greater than
the equilibrium value calculated from the adiabatic flame
temperature for the particular fuel-air mixture ratio.
18.
-------
VI. REFERENCES
1. Bowman, C. T., "Kinetics of Nitric Oxide Formation in
Combustion Processes," paper presented at Fourteenth
(International) Symposium on Combustion, Penn State
University, 1972.
2. Classman, I. and Sawyer, R. F., "The Performance of
Chemical Propellants," Technivisio'n Services, Slough,
England, 1970.
3. Hottel, H. C. and Hawthorne, W. R., "Diffusion in Lami-
nar Flame Jets," Third Symposium (International) on
Combustion, Williams and Wilkens, Baltimore, 1949.
4. Zeldovich, Jr., "The Oxidation of Nitrogen in Combus-
tion and Explosions," Acta Physicochemica URSS 21,
577 (1946).
5. Martenay, P. J., "Analytical Study of the Kinetics of
Formation of Nitrogen Oxide in Hydrocarbon-Air Combus-
tion," Comb. Sci. and Tech. , !L, 461 (1970).
6. Fenimore, C. P., "Formation of Nitric Oxide in Premixed
Hydrocarbon Flames," Thirteenth Symposium (International)
on Combustion, Combustion Institute, Pittsburgh, Pa.,
1971, p. 373.
7. Livesey, J. B., Roberts, A. L. and Williams, A., "The
Formation of Oxides of Nitrogen in Some Oxy-Propane
Flames," Comb. Sci. and Tech., 4_, 9 (1971).
8. Bowman, C. T., "Investigation of Nitric Oxide Formation
Kinetics in Combustion Processes: The Hydrogen-Oxide-
Nitrogen Reaction," Comb. Sci. and Tech., 3_, 37 (1971).
9. Bowman, C. T. and Seery, D. J., "Emissions from Con-
tinuous Combustion Systems," (W. Cornelius and W. G.
Agnew, Eds.) p. 123, Plenum, N.Y., 1972.
10. Thompson, D., Brown, T. D. and Beer, J. M., "The Forma-
tion of Oxides of Nitrogen in a Combustion System,"
Paper presented at the 70th National A.I.Ch.E. Meeting,
Atlantic City, N. J., Aug. 1971.
11. Sarofin, A. F. and Pohl, J., "Kinetics of NO formation
in a Premixed Laminar Methane Flame," Fourteenth
Symposium (International) on Combustion, Combustion
Institute, Pittsburgh, Pa. (in print).
19.
-------
12. Iverach, D. Basden, K. S. and Kern, N. Y., "Forma-
tion of Nitric Oxide in P'uel-Lean and Fuel-Rich
Flames," Fourteenth Symposium (International) on
Combustion, Combustion Institute, Pittsburgh, Pa.,
(in print).
13. Fenimore, C. P. and Jones, G. W., "Oxidation of Ammonia
in Flames," J. Phys. Chem., 65, 298 (1961).
14. Bachmaier, F., Eberius, K. H. and Just, Th. , "The
Formation of Nitric Oxide and the Detection of HCN
in Premixed Hydrocarbon-Air Flames at 1 Atmosphere,"
Comb. Sci. and Tech. 1_> Nos. 1 and 2 (in press) (1973).
15. Eberius, K. H. and Just, Th., "NO Formation in Rich
Flames: A Study of the Influence of the Hydrocarbon
Structure," in "Atmospheric Pollution by Aircraft,"
AGARD Conference Proceedings (in press) (1973). AGARD/
NATO, APO New York 09777.
16. Martin, G. B. and Berkau, E. K., "An Investigation of
the Conversion of Various Fuel Nitrogen Compounds to
Nitrogen Oxides in Oil Combustion," Paper presented at
the 70th National A.I.Ch.E. Meeting, Atlantic City,
N. J., Aug. 1971.
17. Maclean, N. D. and Wagner, H. G., "The Microstructure
of the Reaction Zones of Ammonia-Oxygen and Hydrazine-
Decomposition Flames," Eleventh Symposium (International)
on Combustion, Combustion Institute, Pittsburgh, Pa.,
1967.
20.
-------
FIGURE CAPTIONS
Figure 1 The Adiabatic Equilibrium Flame Temperature
as a Function of Equivalence Ratio
Figure 2 The Structure of Diffusion Flames: a) Droplet,
b) Fuel Jet
Figure 3 Concentration Variation as a Function of
Time in a Reacting Hydrocarbon-Air Mixture
(Temperature in °K). (From Ref. 5)
Figure 4 Comparison of Measured and Calculated NO
Concentration Profiles for Combustion of
Lean and Rich CH4-02-N2~Ar Mixture Behind
Reflected Shock Waves. Initial Post-Shock
Conditions: T = 2960K, P = 3.2 atm. (From
Ref. 1)
Figure 5 "Prompt NO" as a Function of the Ternperc ture
in Various Mixture Strengths in Adiabatic
Propane-Synthetic Air Flames. (From Ref. 15)
21.
-------
<-—
Fig. 1
22.
-------
f
Fig. 2
e L
23.
-------
Fig. 3
24.
-------
eq.
1.2
O
s
oo
Z
2 0.8
u
Z
O
u
^- 0.4
O
Z
0.2
0.4 0.6
TIME - MSEC
0.8
1.0
Fig. 4
25.
-------
J -
100
80
GO
20
0
pprn NO
X--. 1.4
x= 1.2
*X=1.0
+ X = 1,1
o
•* ~
/t."
1.2
1f>00
2100
TEMPERATURE
:si
Fig. 5
26.
-------
NO - Control and Problems in
j£
Mobile Sources
William Balgord
No Abstract or Paper Submitted
27.
-------
Air Quality Benefits and Costs From
Utility NO Regulations
X
by
Peter C. Freudenthal, Ph.D.
Chief Air Quality Control Engineer
Consolidated Edison Company of New York, Inc,
Presented at Middle Atlantic Consortium
on Air Pollution NO Symposium,
New York City Apri$ 26, 1973
-------
List of Illustrations
Figure 1 Off-stoichiometric burner operation in a front wall fired utility
boiler. Lower burners are operated fuel rich, and upper burners,
fuel lean.UJ
Figure 2 Nitric oxide emissions from front wall fired boiler during normal
and off-stoichiometric firing as a function of load.
Figure 3 Nitric oxide emissions from tangentially fired boiler during normal
and off-stoichiometric firing as a function of load. (Oil fuel)
Figure 4 Nitric oxide emissions from tangentially fired boiler during normal
and off-stoichiometric firing as a function of load. (Gas fuel)
Figure 5 Annual average nitrogen oxides concentration (ppm) attributable to
Con Edison operations during 1972.
Figure 6 Diurnal cycle of hourly concentrations of NO, CO, and S0_ during
July-December 1971.
Figure 7 Photograph of plume from Ravenswood Generating Station rising over
polluted air of New York City during episode of 28 October 1966.
29.
-------
ABSTRACT
Con Eidson has conducted NO emission reduction tests which have
A
demonstrated the applicability of off-stoichiometric firing as a control
technique on certain boilers. Con Eidson's emissions produced a maximum of
0.004 ppra of NO at ground level in New York City, and if City emission
X
standards can be achieved, ground level concentrations are expected to be
reduced 0.0016 ppm below existing levels. The benefit from such an in-
significant improvement in air quality is questioned in light of the high
cost of the emission control.
30.
-------
INTRODUCTION
Much of the attention towards improving; nitrogen dioxide concentrations
in urban air has been directed towards control of emissions from large electric
generating stations. The most stringent of these regulations are the NO
X
emissions limitations contained in the New York City Air Pollution Code.
Following enactment of this code in November 1970, Con Eidson began a program
to determine the extent to which nitrogen oxide emissions from its boilers could
be reduced. This paper discusses the techniques of control which Con Edison has
applied, and evaluates the improvement in air quality which would be expected
if the New York City NO emissions standards are achieved.
X
Several boilers of the Con Edison system have been tested for nitrogen
oxides emissions reduction. Two of these boilers, because of their differing
design, provide examples of the available control techniques and the emission
reductions which may be achieved. These boilers are Astoria 20 which is a
front wall fired, 170 MW, Babcock & Wilcox boiler, originally coal fired but
now burning oil and natural gas, and Ravenswood 20, which is a tangentially
fired 385 MW Combustion Engineering furnace. This furnace is divided, and
therefore each section might be considered as 193 MW each, and size comparable
to the Astoria 20 boiler. Ravenswood 20 was originally designed for coal,
although coal burning facilities were never installed. This boiler has only
been operated on oil and gas.
EMISSION CONTROL TECHNIQUES
The method used for reducing nitric oxide formation was off-stoichiometric
burner operation. The concept of this method is that by reduction of flame
temperature below 3,000 F, there should be appreciable reduction of nitric
oxide formation. Off-stoichiometric firing involves manipulation of existing
burners, and generally does not require reworking of existing equipment. This
31.
-------
firing method (Figure 1) involves first operating the top of the furnace
lean (more excess air) and then at the same time the bottom section fuel
rich (less than theoretical air). Both the fuel rich and the fuel lean burners
result in lowering the flame temperature; and the lower temperature yields less
NO .
x
The actual burner pattern selected for maximum NO reduction will vary
X
from one boiler to another. Generally speaking, it is preferable to choose
the lower burners to be fuel rich, so any CO and smoke can be burned off in
the air-rich upper section of the furnace.
To accomplish this two-stage combustion firing, the fuel was turned off,
and air left full on to some of the burners in the top row. Fuel pressure was
then raised to compensate for the burners which were out of service, and the
air dampers of some of the bottom row of burners were then closed.
This procedure produced smoking in all cases when the burners were not in
good condition.
Final trimming of the air/fuel ratio was accomplished by adjustment of the
air only. By trying various combinations of burners in and out of service,
and by manipulating oil pressure and air, NO formation was minimized.
Unfortunately, this manipulation is extremely difficult to accomplish.
Other factors, such as output and steam temperature must be maintained, and
some low velocity coal burners converted to oil will not tolerate much reduction
in air before excessive smoking occurs.
EMISSION REDUCTION
The emission reduction achieved at Astoria 20, the 170 MW front wall fired
unit, is shown in Figure 2. The upper curve shows the NO formation during
X
normal operation expressed as a percent of the maximum NO level, as a function
X
of load. The lower curve shows the "best" off-stoichiometric results. Note
that smoking became unavoidable over 140 MW, although a 37% reduction in NO
x
32.
-------
operation was achieved up to nameplate rating, although not to peak. In order
to keep the furnace in balance during the series of experiments, burners were
turned off two at a time, and at opposite corners of the furnace.
Emission reduction curves for gas-only firing are presented in Figure 4.
During normal firing conditions, load on this boiler had to be restricted
to below 300 MW to avoid excessive temperatures. Off-stoichiometry alleviated
some of these steam-temperature problems, because when all of the top tier of
burners were operated "air only", the flame was effectively lowered within
the furnace.
Although substanital NO reduction were achieved furing test conditions,
X
the practical ability to reduce NO may be somewhat less. The best patterns
for NO reduction may not be compatible with long-term and safe operation of
X
the boiler. Some compromise in NO levels must be accepted rather than operate
X
with high CO or too close to a smoking condition.
AIR QUALITY
The success of this or any other emission control program can be measured
by the resulting improvement in ambient air quality. In order to assess the
improvement in air quality, the total impact of Con Edison's NO emissions on
X
New York City's environment was estimated through the use of a meteorological
(2)
diffusion model. This model, which was developed for the Environmental
Protection Agency's evaluation of Implementation Plans, calculates annual
average ground level concentrations, given an emission inventory and a set
of climatological data. The model had been tested for sulfur dioxide, and was
(3)
shown to be accurate within a factor of 2 for both point and area sources.
Because the modeling techniques for elevated point sources are less complex
and the emissions inventories are more accurate than for area sources, the
accuracy for point sources is probably much greater.
The total NO emissions from each Con Edison stack during 1972 including
X
gas turbines, were entered into the computer program, along with stack
33.
-------
temperatures, velocities, locations, heights, and diameters, and five year
climatological data from LaGuardia Airport. Calculated ground level N0x
concentrations attributable to Con Edison operations are shown in Figure 5.
The peak NO concentrations, which is calculated to occur in Queens, was
^ x
about 0.004 ppm. This peak concentration represents a mixture of both NO and
N0?, but primarily the former. Boiler emissions of 'oxides of nitrogen are
usually over 90% as NO, for which there is no air quality standard. It takes
minutes to hours for NO to oxidize to N0?, and during the time oxidation occurs,
there is additional dispersion, making the peak NO concentration from Con Edison
much less than indicated in Figure 5. The most conservative way to compare the
estimated concentrations with the air quality standard would be to assume that
these concentrations represent 100% NO . Assuming this, Con Edison contributed
only about 8% of the Federal primary and secondary N0? air quality standard
during 1972.
The significance of this concentration is demonstrated by comparison to air
quality measurements in New York City. NO is measured at five stations in
New York City - 121st Street, 45th Street, Canal Street, Astor Place, The
Brooklyn Public Library, and Morrisania, Bronx. Only one station reported a
full year's data in 1972, but two operated for 11 months, one for 8 months, and
two for 7 months. All stations reported average N0? concentrations of 0.03 ppm
or less, indicating that the primary and secondary air quality standard has been
achieved.
Even is Con Edison could totally cease NO emissions, it would be impossible
X
for epidemiologists to detect and improvement of public health - and this is
what emission control is all about.
The reason for the extremely low impact of Con Edison's NO emissions on
x
ambient air quality is the very high effective stack height of its plants. Most
34.
-------
of the stacks are 300-500 feet tall, and because of the concentration of heat
released at the stack exit, the plumes typically rise hundreds of feet above
the stacks. The result of this high effective stack height is that nearby
the stacks there is nearly zero impact at ground level. Downwind the plumes
disperse in both the vertical and horizontal directions, and finally the bottom
edge of the plumes reach the ground. But by then dispersion has reduced trace
gas concentrations to very low levels.
An example of the benefit of high effective stack height is presented in
Figure 6. The photograph was taken during and air pollution episode, because
during an episode, the air becomes so laden with particulates that the resultant
haze becomes photographable. The plume well above the inversion layer seen
from the Ravenswood Generating Station is observed dispersing. It is apparent
that the plume is not contributing to the polluted air near the ground.
Aerometric data analysis supports the hypothesis that Con Edison, because
of its tall stacks, has a very small impact on ambient NO concentrations.
X
Figure 7 shows the diurnal cycles of CO, S0_ and NO from July through December
1971, as measured at the New York State Department of Environmental Conservation
monitoring station at Welfare Island. Because diffusion characteristics of
these gases can be assumed proportional to each other when they come from a
common source, CO can serve as a tracer for NO emitted from automotive traffic,
and S09 is suited as a tracer for NO from fuel burning. Although these tracers
distinguish between stationary and transportation sources, they do not dis-
tinguish between on-site boilers used for space-heating and large boilers used
for electricity generation.
All three curves show an early morning peak concentration and a decline
during the late morning hours. This peak, which occurs when meteorological
dispersion conditions are at their worst, occurs at the hours when both traffic
is at a peak and S0_ emissions are at a peak, because of the early morning
35.
-------
demand for hot water, space-conditioning and electricity. In the early
afternoon, when traffic begins to rise (center shaded area), both CO and NO
rise at equivalent rates, but SO tends to remain constant. This rise in NO
and CO occurs in spite of this being the best time of day for meteorological
dispersion. In the late afternoon and early evening, S02 also rises as emissions
increase due to early evening demand for hot water, heat and electricity, and
meteorological dispersion worsens. After midnight (left and right shaded areas),
SO concentrations remain nearly constant, but CO and NO decrease rapidly as
traffic declines. The coincidence of the nighttime decline and early afternoon
rise of both CO and NO, which does not correspond to the S02 curve, strongly
suggests the relationship of NO ground level concentrations to automotive
X
emissions. These curves would tend to confirm the conclusion presented above
based upon meteorological modeling, that power generation is not a significant
source of ground level NO .
X
COST-BENEFIT
In November, 1970 New York City, assuming that an NO problem existed, enacted
some of the most stringent NO limits in the entire nation - almost twice as
x
restrictive as the Federal new source standards. These regulations, which were
aimed only at utility boilers, restrict NO emissions to 150 ppm in existing
X
plants and 100 ppm in new utility boilers. If Con Edison is able to achieve and
maintain these emission levels, the maximum annual average NO concentration
attributable to the utility will be reduced by 0.0016 ppm at the area of maximum
impact.
Although this reduction is insignificant in terms of health and welfare, it
will cost tens of millions of dollars. The modifications to the new 800 MW
Astoria 60 boiler, through which Con Edison is attempting to achieve the never-
before attained level of 100 ppm, cost over 14 million dollars. These costs will
36.
-------
be eventually borne by Con Edison's ratepayers.
At the same time that the City passed these costly NO emission restrictions,
X
it was curtailing its own environmental programs for lack of money - rat control,
removal of lead paint from walls in the ghettos, ragweed control. Unfortunately,
no statistics are available to compare the number of asthmatic attacks which
will be avoided through the multimillion dollar public health expenditures for
0.0016 ppm NO reduction from Con Edison boilers, versus the number of attacks
X
which might have been avoided had the ragweed control program, costing a few
thousand dollars per year, not been curtailed.
37.
-------
SUMMARY AND CONCLUSIONS
Nitrogen oxide emissions reductions have been achieved by means of
off-stoichiometric firing during carefully controlled test conditions on
several Con Edison boilers. It is unknown, however, whether boiler operation
can be maintained at levels mandated by the New York City Air Pollution Code.
The expected reduction of ambient NO will be approximately up to 0.0016 ppm
X
if Con Edison's emissions are reduced to legislated levels, but this change in
air quality will cost tens of millions of dollars. Because of the insignificance
of this reduction in terms of air quality improvement and health and welfare
benefits, it seems appropriate that local environmental control strategies be
reevaluated.
Acknowledgment
The author gratefully acknowledges the assistance of Messrs. George
Stegmann, Paul Giardina, and 0. G. Hanson of Con Edison's Mechanical Engineering
and Environmental Departments. The emission reduction tests were conducted under
Mr. Stegmann's supervision.
38.
-------
REFERENCES
(1) Seabrook, H. H. and B. P- Breen, A Practical Approach to NO Reduction
in Utility Boilers, presented at the American Power Conference, Chicago,
April 18-20, 1972.
(2) TRW Systems Group, Air Quality Display Model, prepared for Department of
Health Education and Welfare, Public Health Service, National Air Pollution
Control Administration, Washington, D. C., November, 1969.
(3) Freudenthal, P- C., P. A. Giardina, and K. Juris, Application of the Air
Quality Display Model to New York City, presented at the Fall Annual Meeting
of the American Geophysical Union, San Francisco, December 1971.
39.
-------
'OFF-STOICHIOMETRIC11 BURNER OPERATION
FIGURE i
-------
ASTORIA UNIT 20 - REDUCED NITRIC OXIDE OPERATION
OIL FUEL
gioo
2
o
o
5 80
60
40
20
o
oc
UJ
o_
-JJ-
BASELINE (NORMAL) OPERATION
70
OFF-STOICHIOMETRIC
OPERATION
A
UNACCEPTABLE
SMOKE
RECOMMENDED
FIRING PATTERN
LOW NITRIC OXIDE
FIRING PATTERN
DETERMINATION
40
60
80 100 120
UNIT LOAD, MW
140
160 180
FIGURE 2
-------
RAVENSWOOD UNIT 20 - REDUCED NITRIC OXIDE OPERATION
OIL FUEL
NORMAL OPERATION
100
80
60
x 40
LU
O
20
0
100
n
REDUCED NITRIC OXIDE OPERATION
200
300
400
LOAD - MW
FIGURE 3
-------
RAVENSWOOD UNIT 20-NITRIC OXIDE REDUCTIONS
WITH NATURAL GAS FUEL
- lOOr
O
9 80
X
O
60
40
20
BASELINE (NORMAL)
OPERATION
100
RECOMMENDED OPERATING
CURVE
O NORMAL 02
A MINIMUM 02
D MINIMUM NOxWITH
DESIGN STEAM
TEMPERATURES
200 300
LOAD- MW
400
FIGURE 4
-------
FIGURE 5
44.
-------
FIGURE 6
DIURNAL. CYCLE OF NO, SOi; AND CO
JULY- DECEMBER IS7I
(WELFARE-ISLAND, MONDAY THRU FRIDAY AVERAGE DATA)
IU
Q
X
o
Q.
cc a.
o _
UJ
X
O
o
U_
-J
Z>
CO
.050
.000
LJ
O
X
o
2:
o
o
ffj
cc
<
o
HOURS
45.
-------
FIGURE 7
46.
-------
FUTURE APPROACHES TO STATIONARY SOURCE
CONTROL OF STACK GAS NO
x
BY: A.J. Tellera)
Nitrogen oxides appear to present one of the most formidable
pollution control problems that we are encountering. They are
emitted by all thermal processes as well as specific chemical
operations such as nylon and nitric acid manufacture and electro-
plating. As a result, nitrogen oxides represent the universal
pollutant.
Early work by Chambers & Sherwood( , Denbigh and Prince( '
(2)
and Carberry shed some light on a confusing situation. Nitrogen
oxides do not behave simplistically in the recovery -process of
contact with liquid. Chambers and Sherwood proposed a gas phase
mist reaction. Denbigh and Prince found that the data indicated,
the rate limiting reaction involves N?0, but that the rate of
absorption, V, is related as follows:
V = Ka [N204J - C [N204] 1/4 [NO] 1/2
Carberry, in relative agreement with Denbigh and Prince establish-
ed that the key reaction involves N_0, and precludes a gas phase
reaction.
However, it is acknowledged by all investigators that the
formation of NO,., whose concentration is limited from thermodynamic
implication but whose equilibrium attainment is rapid, may very well
enhance the absorption of nitrogen oxides and may also contribute
to the gas phase reaction and misting observed by Chambers and
Sherwood.
a)
Mr. Teller's paper was presented by Mr. O'Neall of Teller Environmental
Systems Inc.
47.
-------
Dekker, Snoek, and Kramers(7) confirmed the N^ mechanism of
absorption in water or caustic solutions and indicated, in agreement with
Sherwood, that the system could be gas phase controlling at high concentration
nitrous gas systems, but also agreed with Denbeigh and Carberry that it could
be liquid phase controlling at low nitrous gas concentrations.
Koval and Peters (14') indicated that the mechanism of N0x absorption was
far more complex than indicated by previous investigators and included the
effect of an increasing inventory of HN02 during the absorption process. The
model developed and reasonably confirmed by experimental data are
-d (e N02)g = kc (N02)g" + kd (N0)g (N02)g - k£
dt
-d (NO) = k, (NO) (NO.) - k (HNO )2
g d g 2 g e 2.
dt
The major concern from the aspect of pollution control is that only in
limited circumstances is nitrogen dioxide (or equivalently nitrogen tetroxide)
of major concern. In the case of combustion gases emitted from power plants,
incineration, steel, glass, non ferrous metal manufacture, the sources of the
major quantities of NO emissions, the significant specie of the gases is
A.
nitric oxide.
Additionally the nitrogen oxides are often emitted in streams combined
with SO resulting in the additional conversion of N0_ to NO via reaction
£- v£
S02 + N02 = S03 + NO
Even were NO to be absorbed, NO is created by the approximate overall
stoichiometric reaction (Koval and Peters correction not applied)
3 N03 + H20 = 2 HNO + NO
Thus the pollution problem we encounter in products of combustion is to
48.
-------
a major degree the problem of nitric oxide control, the removal of a
relatively insoluble unreactive gas and to a minor degree the removal of
a soluble NC>2 which process releases NO equivalent to 1/3 of the noxious
material removed.
In the case of chemical processing, munitions destruction, and plating,
the pollution problem is related to nitrogen ioxide control, but here again
the creation of nitric oxide during the liquid gas absorption process presents
a formidable problem.
As established by numerous investigators and confirmed recently by Morrison,
Rinker, and Corcoran , the oxidation rate of nitric oxide by oxygen to
nitrogen dioxide is extremely slow, NO, in concentrations of 10 - 50 PPM in
oxidized by oxygen to 80% completion only after 150 - 500 minutes contact
time.
Since federal legislation is directed to achievement of emission levels
in the order of 100 PPM total NO , recovery or preventative methods beyond mere
X
absorption of N0« are indicated.
Preventative control is achievable in combustion processes and in plating
operations.
For example, the use of tangential firing and gas recycle in boiler operation
has resulted in the reduction of NO emissions from the range of 1000 PPM a
X
range of 50 - 150 PPM. These achievements have resulted in the establishment
of emission standards by EPA of 150 - 200 PPM for liquid fuel fired operations
and 400 - 500 PPM for coal fired operations.
49.
-------
/Q\
Performance of operating systems reported is as follows
FUEL FIRING CAPACITY - MW ^ ~ PPM
Gas Tangential 32°
10-15% Recirc.
40%
Gas Wall Fired
Off Stoichiometric 480 150
Comb. - Equivalence
Ration 1.3 - 1.4
Gas Wall Fired 750 120
Off Stoichiometric
Oil Wall Fired 480 185 - 210
Off Stoichiometric
Wall Fired 180 187
Off Stoichiometric
Coal Tangentially Fired 320 85
15% Recirculation
Tangentially Fired 125 200
A second method of suppressing NO emissions is the introduction of
(13)
urea into plating baths . It is reported by Kerns that the NO emission
was reduced to less than 1 PPM as compared with normal emissions of 8000 PPM
when the concentration of urea was 6 oz/gallon in a 40% nitric acid solution.
In the nitric acid industry, the major process application for reduction
of NO emissions is the catalytic reduction of the nitrogen oxides. Both
X
methane and hydrogen are used as reducing agents over a platinum catalyst. Both
spherical catalysts and honeycombed catalysts are used in the conversion of
nitrogen oxides to nitrogen. Federal standards, based on existing performance,
were established as 3 Ib NO /ton HNO , which is equivalent to 209 PPM. In actual
X J
plants tested, catalytic reduction using methane achieved a low level of 110 PPM
in the exhaust gas with the average plant emitting at the level of 310 PPM.
Where hydrogen is used as a reducing agent, the emissions range from 50 - 150 PPM
Where N0_ is a major component of the emissions such as potassium
50.
-------
permanganate has resulted in s uppression of nitric oxide formation.
has, via introduction of a high liquid turbulence, increased the overall rate
of absorption of N02 by 25%. The use of FeSO, solutions for absorption of NO
results in the creation of a coordinate-covalent adduct that permits total
absorption of N02 in non-oxidizing media.
Major activity, recently has been devoted to avoiding the use of liquid
gas contactors for NO recovery because of the equilibrium limitations. The
X
activity has been divided into three areas
1. Catalytic reduction of nitrogen oxides.
2. Adsorption on silica or molecular sieves.
3. Ab-adsorption on chromatographically active materials.
The major emphasis on the reduction of nitrogen oxides to nitrogen has
been on the study of copper based catalysts (1, 4, 16, 17, 18, 19). In all
cases the reaction is related to the reduction of NO compounds by reducing
X
agents, primarily CO, over a copper based catalyst.
Adsorption of NO on silica gel and molecular sieves has been report
X
ed
Problems of interference by water vapor in molecular sieve work has been
encountered. However, pilot plant results reported by EPA indicate a reduction
of NO from the 1000 PPM concentrations range to 50 PPM.
X
The third area, that is the ab-adsorption on chromatographically active
material, is in an early stage of development. Selective ab-adsorption on
TESISORBS are indicated. These are potentially capable of regeneration and
early studies to this and are in progress.
All three of the more recent approaches to NO control attempt to avoid the
X
inherent limitations of the liquid-gas systems where equilibrium causes the
production of a relatively insoluble NO or where the liquid capability as a
solvent of NO is limited. It appears that economics will eventually direct much
of our efforts to low-energy consuming control systems such as minimizing NO
X
formation, and that in recovery systems, solid gas systems will prevail.
51.
-------
BIBLIOGRAPHY
(1) A. Bauerle, ACS Div. Petrol. Chem., 1971, 16(2), E70-72
(2) J.J. Carberry, Chem. Eng. Sci, 9(4), 189-195, (1959)
(3) F.S. Chambers and T.K. Sherwood, JACS, 59, 316, (1937)
(4) H. Chien, ACS Div. Water, Air, Waste Chem, Gen Paper 1971,
11(1), 71 0 75
(5) J. Collins, U.S.P. 3674429
(6) K.G. Denbigh and A.J. Prince, Trans. For. Soc., 790-801, (1947)
(7) W.A. Dekker, E. Snoek, H. Kramers, Chem. Eng. Sci, 11, 61-71, (1959)
(8) EPA -- Fossil Fuel Fired Boilers (1972)
(9) EPA — Nitric Acid Plant Survey (1972)
(10) J.W. Fleming and G.L. Nosine, U.S. Army CRLR 439, (1955)
(11) S.N. Ganz, J. App. Chem. USSR, 30(5), 732, (1957)
(12) R. Kazakova, Inf. SGN - IPIAPPOS, No. 6, 229-49, (1971)
(13) B.A. Kerns, IEC Proc. Des. & Dev. , 4(3), 263-266, (1965)
(14) E.J. Koval and M.S. Peters, IEC, 52(12), 1011-1014, (1960)
(15) M.E. Morrison, R.G. Rinker, W.H. Corcoran, I. & EC Fund, 5(2) ,
(16) J. Sorensen, IEC Prod. Res. Dev., 11(4), 423-6, (1972)
(17) U.S.P. 3702236, 1972
(18) U.S.P. 3682585
(19) U.S.P. 3701823
52.
-------
NO - CONTROL BY WET SCRUBBING
x
by
Robert S. Kapner
Engineering Division
The Cooper Union
Prepared for
Symposium on the Current Status of the NO problem and its control
X
Middle Atlantic Consortium on Air Pollution (MACAP)
April 26, 1973 CUNY, GRADUATE CENTER, NY, NY
Elmar R. Altwicker, Ph.D., Chairman
-------
Introduction
On approaching the problem of the control of nitrogen oxide emissions
from stationary sources, one is confronted with several possible methods
that would appear to offer promise of providing a satisfactory technical and
economic solution. Among the methods that have been proposed are:
(1) Process and equipment modifications which reduce
NOX emissions by changing either equipment design
or operating procedure.
(2) Removal of nitrogen oxides after they have been pro-
duced in combustion or manufacturing processes by
a) reduction of NO to N2 and Oo
b) absorption by wet scrubbing
c) adsorption onto solid sorbents
Most popular, at the present time, are methods involving process
and equipment modifications such as two-stage combustion, flue gas
recirculation, fluid bed combustion, and tangential firing - all applicable
to fossil-fuel combustion systems (power generation, steam generation,
domestic and industrial heating, incineration) which is the single largest
contributor to nitrogen oxide emissions from stationary sources.
Part of the reason for the popularity of the process and equipment
modification approach is that it attacks the problem of NOX emissions
directly at the point of NOx formation - in the combustion chamber. To
some extent, also, part of the reason for the popularity of the modification
approach is the failure to quickly and unequivocally find a control system
among the removal methods which has proven to be completely satisfactory
54.
-------
for abating NOX emissions from stationary sources. The same pattern
can be observed in NOX abatement programs from moving sources, where
the focus of control has shifted somewhat from catalytic removal methods
to engine modifications which are aimed at preventing NOX formation.
Despite the popularity of modification methods, it is strongly felt
in some quarters, that removal systems still hold promise of providing
solutions to the NOX emission problem, from both moving and stationary
sources. It is the purpose of this paper to review proposed wet scrubbing
systems having potential for NOX abatement from stationary sources.
From one point of view, wet scrubbing systems for NOX control
can be grouped according to the type of scrubbing solution used, the
nature of the products formed upon absorbing NOX) the ease with which the
scrubbing solution can be regenerated, and the economic value of materials
recovered from scrubbing systems.
A classification of this type would include:
(1) Scrubbing solutions which upon absorbing NOX are consumed
or transformed into intermediates that are difficult to
regenerate and have low economic value. Examples include
urea, which is decomposed upon absorbing NOX, and sulfite
and bisulfite solutions which are converted to sulfates by
NOV absorption.
J"i.
(2) Scrubbing solutions which produce intermediates that can be
regenerated by chemical and thermal means along with
nitrogen salts that have some economic value. Examples
include alkaline scrubbing by magnesium and calcium oxides
to form nitrites which can be converted to nitrates by thermal
and chemical processes.
55.
-------
(3) Scrubbing solution which complex NOX gases upon absorption
and which can be regenerated by heating, the regeneration
step releasing NOX in concentrated form. Examples include
aqueous solutions of some inorganic salts, organic solvents
and organometallic materials.
This paper will explore only the latter two categories, and then
rather narrowly, by examining selective examples which show the greatest
potential for practical application.
Although NO is the principal constituent in the general designation
NOX, we shall have to treat at least three other nitrogen oxides which play
an important part in wet scrubbing absorption processes. In addition to
nitric oxide (NO) these are, nitrogen dioxide (NO2) and its equilibrium
dimer,dinitrogen dioxide (N2O4) and dinitrogen trioxide (N2Os) which,
although a true compound, can be considered to be an equimolar mixture
of NO and NO2.
The extent to which one or all of these nitrogen oxides is present in
a stack emission depends somewhat on the nature of the source, but there is
also a major dependency on the concentration level of the total nitrogen
oxides present and upon the physical state of the stack gas, particularly
its temperature. At ordinary temperatures, NO is readily oxidized to
NO2 in the presence of oxygen which is almost always available from excess
air used in combustion processes and in some chemical operations such as
nitric acid manufacturing. Where NO2 is easily produced by NO oxidation,
56.
-------
the" presence of the other oxides of nitrogen are automatically ensured,
and when NO2 is not easily formed it is sometimes necessary to arti-
ficially introduce it as a preliminary step in wet scrubbing systems.
NITROGEN OXIDE SCRUBBING SYSTEMS -
WATER, BASIC AND ACID SOLUTIONS
A description of control methods for the removal of nitrogen oxides
from stationary emission sources by wet scrubbing methods properly begins
with an examination of the more obvious chemical properties of nitrogen
oxides, particularly their solubility and reactivity with water and with
aqueous solutions of acids and bases. Quite a bit is known about these
systems and they have been studied with great interest because of the
importance of their industrial applications.
Wet Scrubbing With Water
Nitric oxide is the most stable and nonreactive of all the nitrogen
oxides present in stack emissions. It is only sparingly soluble in water,
about 1.5 times that of oxygen at ordinary temperatures. Although
Partington (1) reports that NO disproportionates in water to form N2O3 and
the rate is exceedingly slow and NO is generally considered nonreactive.
Nitric oxide is also inert to aqueous solutions of acids or bases and in
fact can be purified by bubbling through H2SO4 and KOH to remove water
57-
-------
and NO2 respectively (2).
The other oxides of nitrogen, however, are quite reactive with
water. The overall reaction of NO2 with water is given by
3NO2 + H2O = 2HNO3 + NO (D
This single equation is often used to represent the commercial
production of UNO3 although it is known to actually proceed in a number
of complicated steps. It is generally conceded that although NO2 and its
dimer N2O4 are soluble in water, the chemical absorption of the dimer is
of far greater importance than the physical absorption of NO2 especially
at high concentrations of NO2 in the gas phase. The route to HNO3 by
N2O4 absorption is described by equations (2) - (5):
N2°4 + H2° = HNO3 + HNO2 (2)
3HNO2 = HNO3 + 2NO + H2O (3)
2 NO + O2 = 2NO2 (4)
2 N02 = N204 (5)
The nitrous acid produced by NO absorption, equation (2), is
c* ft
decomposed to produce nitric acid and nitric oxide , equation (3). The
process of HNO3 production initiated by N2O4 absorption can continue
only at high concentrations of NO. The oxidation of NO to NO2 by oxygen
in air in the absorption column, equation (4), followed by dimerization of
58.
-------
N02 to N204, equation (5), are critically dependent on NO concentration.
Initially, this reaction sequence predominates, but as nitrogen is removed
from the gas phase and appears as nitric acid in the liquid, the concentra-
tion of NO becomes too low for reactions (4) and (5) to produce significant
amounts of N2O4. Instead, toward the end of the absorption process,
when NO2 concentration is relatively low, a second mechanism predominates.
This is the absorption of N2O3, formed from NO and NO2 in the gas phase,
to produce nitrous acid which decomposes to form nitric acid and nitric
oxide.
N2O3 + H2O = 2HNO2 (6)
3HNO2 = HNO3 + 2NO + H2O (3)
2NO + O2 - 2NO2 (4)
NO + NO2 = N2O3 (7)
Although the route to HNOg by NgOg absorption is most important at low
NO and NO9 concentrations, it must also occur at high concentrations
&
since NO is everywhere produced by the decomposition of HNO2 - however
formed. The overall scheme of nitrogen oxide absorption to produce
nitric acid is shown in Figure I as described by Hoftyzer and Kwanten (3).
The steps describing nitric acid production by N2o4, NOg and
N2 °3 absorption cannot continue indefinitely. When the concentration of
nitrogen oxides becomes too low to support reasonable reaction rates, the
59.
-------
remaining nitrogen oxides are discharged. The irony of the problem of
removing nitrogen oxides from discharged nitric acid absorption column
tail gas by wet scrubbing is only too apparent at this point. If the NOX
concentrations are too low to efficiently absorb in water, then wet scrubbing
with water would appear to hold little promise as an emission control
method. Although NOX removal at low concentrations is difficult, it can
be accomplished. For example, in some older nitric acid manufacturing-
processes the nitrogen oxides in the tail gas have been scrubbed clean by
contact with aqueous Na0CO9 although this method is not practiced in
£j *^
modern, high pressure nitric acid plants.
Wet Scrubbing With Alkaline Solutions^
That a carbonate solution can potentially absorb nitrogen oxides
effectively at concentrations too low for efficient absorption in water (or
dilute HNOg), introduces the next set of systems, alkaline and carbonate
wet scrubbing for controlling NOX stack emissions. The reaction schemes
here are quite similar to those described for scrubbing nitrogen oxides
with water and are shown in Figure 2.
A significant advantage is realized with alkaline or carbonate
scrubbing solutions in that the nitrite formed from N2O4 or N2O3 absorption
is stable in basic solution and does not decompose to form NO as does
its acid analog HNO2 in acid solution. This is the principal idea behind
the use of aqueous NaOH as the absorbing medium in sampling for atmos-
pheric NOX.
60.
-------
Because of the apparent potential for wet scrubbing low concentra-
tion nitrogen oxides in basic solutions, considerable effort has gone into
studies whose goal was finding an aqueous alkaline or carbonate system (or
mixture) which could effectively be used for cleaning stack gases as well
as the best conditions for performing such processes. The outcome of such
studies has been a large number of publications in the technical literature
on NOX scrubbing systems, a steady issuance of patents on alkaline and
carbonate scrubbing systems, an examination of existing wet scrubbing
processes for SOg control using alkaline materials for simultaneous control
of nitrogen oxides and, finally, more than a little confusion as to the prac-
tical performance of such methods.
Some idea of the range of studies found in the literature is shown in
Table 1 which represents a small but typical selection found. A search of
Chemical Abstracts for the period 1962 - 1972 revealed some 60 articles,
mostly Russian, on the subject of alkaline and carbonate scrubbing systems.
In addition there have been at least two major studies on wet scrubbing per-
formed for the U. S. Environmental Protection Agency during the last
two years (4,5). One of these studies is experimental (Chappell, ref. 5)
and is abstracted in Table 2.
One of the chief variables affecting the extent and rate of absorption
of NOX is the ratio of NO2/(NO + NO2), denoted by a in Tables 1 and 2.
Most of the literature on NO absorption examines the particular region
.A.
61.
-------
a - 0.5 which on a scale from zero (pure NO) to 1 (pure NO2), corresponds
to a mixture with an average composition of N2O3. The importance of this
particular concentration ratio of nitrogen oxides in wet scrubbing systems
deserves special attention.
It was noted previously that N2O4 and N2O3 were the active species
in NOX absorption, not NO. However, upon absorbing N2O4 or N2O3 in
water or an acid medium, nitrous acid is produced which then forms nitric
acid and NO. Thus, the ability to absorb nitrogen oxides depends upon the
ability to form N2O4 and N2O3 from NO. Even for scrubbing systems in
which NO is not produced, alkaline mediums, for example, it is necessary
that the NO in the gas to be scrubbed be converted to NO2 so that either
N2 04 or N2O3 can be formed. At high concentrations of NO and in the
presence of air, NO2 and its dimer N2O4 will readily result. Unfortunately,
the ability to form N2O4 under normal stack conditions is almost nil due
to the low concentration of NO in emission sources and the rather high
temperatures that accompany many stationery emissions. The effect of
temperature and concentration are treated in considerable detail elsewhere
(6, 7) and will be discussed only briefly here.
Consider temperature first. At temperatures below about 140°F
as much as 50% of the NO2 in a gas will be present in dimer form. As
temperature increases the equilibrium amount of dimer present will rapid-
ly diminish. Above about 280 F virtually no N2O4 will exist. In addition,
the oxidation rate of NO to NO2 decreases with increasing temperature.
62.
-------
Thus, for nitrogen oxides emitted at relatively high temperature, as
for example those found in combustion processes, the effect is to deny
the availability of N2
-------
What these considerations tell us is simply that absorption
process which rely on N2O4 must be discounted in favor of absorption
in which the active species is N2O3. In addition, stack gases that are
mainly composed of NO will have to be adjusted to an average composition
corresponding to N2O3 (viz. a = 0. 5) principally by adding NO2 and not
by relying on the oxidation of NO by O2 in the stack gas. This, then, is
the major reason for almost exclusively examining absorption processes for
NO at concentrations corresponding to N2 03.
There is an additional unfortunate effect associated with the need to
add NO? to a stack gas effluent consisting mainly of NO so that the mixture
to be scrubbed is essentially N0Oo. To illustrate this, suppose we have a
Z o
stack gas with lOOOppm of NO that we wish to scrub. We add enough NO2
(lOOOppm) to form lOOOppm of N2Qo and then scrub with, say, 90% efficiency.
The scrubber effluent will have lOOppm of NO and lOOppm of NO0 and in
&
terms of the original lOOOppm of NO we will have reduced the NO content
.X.
by only 80%. In order to achieve 90% reduction of the original NO we would
have to scrub with an efficiency of 95%. Thus, the need to add NO2 to the
stock gas for the purpose of producing a scrubbable mixture requires that
the scrubber operate at an efficiency considerably greater than would be
necessary based upon the original NOX content of the stack gas.
If there is anything at all fortunate about the need to add NO0 in order
£
to produce a scrubbable mixture, it is that the NO2 is readily available
64.
-------
from those processes which regenerate the scrubbing materials, as we
shall see later.
Chappell (5) tentatively concludes from his experimental study that
alkaline slurries and solutions (calcium, magnesium, zinc, sodium hydrox-
ides) are not very efficient for NOx wet scrubbing. Table 2 shows that
less than 30% NO2 was removed and generally smaller percsntage of NO
was scrubbed out of NOX mixtures that never exceeded a total concentration
of 0. 08% by volume. In another study, First and Viles (6) found higher
scrubbing efficiencies, greater than 90%, with water as the scrubbing
medium and virtually no change in absorption efficiency upon adding caustic
to the scrubbing liquor. The theoretical study by Lowell (4) produced some
evidence that CaO was a potentially suitable scrubbing medium for NOX
removal from stack gases.
Although the First and Viles study (see Table 3) was conducted at
considerably higher concentration levels than in Chappell's work, the
effect of concentration on scrubbing efficiency does not by itself serve to
explain the rather large discrepancy in NOx removal capacity between
the two works. One possible contribution to the difference in scrubbing
efficiency could be due to mechanical differences between the two sets of
experiments. Chappell used a simple fritted bubbler to introduce the NOX
bearing gas into a static absorbing medium while First used a sixteen
stage gas absorber in which the scrubbing medium was sprayed over the
gas mixture in each stage. The effect of scrubber design on absorption
65.
-------
capacity has been investigated by Strom (8) and Peters (9) in laboratory
scale equipment. Their studies show that absorption efficiency is clearly.
dependent upon scrubber type. Until the effect of contacting efficiency
due to equipment variation is carefully controlled, experimental studies
will continue to be difficult to reconcile. Drawing conclusions from results
reported in the general literature (Table 1) is also complicated by the
effect of different contacting devices.
One further result can be drawn from the data presented. Although
it was suggested that at low NC) concentrations the predominant absorption
A.
mechanism should involve N2 03, the data do not bear this out. In the
First and Viles study, when the inlet NO approximates an N2Og mixture
(runs 1 and 4 for which a [n - 0. 5) then the NO and NO2 absorbed are
approximately equal, leading to calculated values of a removed ~ °° 5<
However, when there is more NO2 than NO in the inlet gas (runs 2 and 3
for which a -n > 0. 5) then more NO2 is absorbed than is associated with
NO to form N2 Og. This is shown in the calculated values of a removed
for runs 2 and 3 for which a removed > °-5- Apparently NO2 absorption
(possibly as N2O4) proceeds even at low concentrations of NO2 for which it
was previously assumed little absorption would occur. This effect is even
more pronounced in the Chappell data (Table 2) where for a total of 10
absorptions in alkaline solution each involving a -m^. 0. 5, 7 runs resulted
in more NO2 absorbed than NO suggesting that another species in addition
to N2O3 was being absorbed.
66.
-------
Notwithstanding the difficulty in interpreting NOX absorption data,
it would be appropriate to examine two complete system descriptions for
NOX removal from stack gases by wet scrubbing processes. The first of
these is attributed to Schmidt and Weinrotter (10) who describe NO
X.
scrubbing for a nitric acid plant using aqueous suspensions of Mg(OH)2
or MgCC>3. A schematic representation of the process is shown in
Figure 3.
The scrubbing process essentially produces Mg(NO2)2 according to
the reaction:
N2O3 + Mg(OH)2 = Mg(NO2)2 + H2O (8)
Because the alkaline solution will not react with NO alone, the tail gas
effluent which is fed to the scrubber must be adjusted to N2O^ by the addi-
tion of NO2. The NO2 used to balance the NO in the tail gas is produced by
oxidation of NO which is formed by the thermal decomposition of Mg(NO2)2
after leaving the scrubber. The decomposer typically operates at temper-
atures in excess of 300°F and 4 atm. to produce a mixture of Mg (OH)2 and
Mg(N03)2:
3Mg(N02)2 + 2H20 = 2Mg(OH)2 + Mg(NO3)2 + 4 NO (9)
The NO produced in this decomposition stage is oxidized to NO2,part of which
is used to blend with NOx in the tail gas to produce N2O3 and the balance
returned to the HNO3 plant. The recovery of Mg(OH)2 is completed in the
precipitator ( ammoniator) where, by the addition of NH3, the Mg (NO3)2
67.
-------
is" converted to NH4NO3 and
Mg(N03)2 + 2NH3 + 2H2O = 2NH4NO3 + Mg(OH)2 (10)
By combining equations (8) - (10), Mg(OH)2 and all its intermediates
can be eliminated:
3N2°3 + 2NH3 + H2° = 2NH4NO3 + 4NO
-------
(calculated as N2O3). For a typical 120 ton/day nitric acid plant dis-
charging tail gas at the rate of 80, 000 SCFM (85°F, 92 psig) containing
0. 125% N2O3, then about 37 tons /day of NH4NO3 must be disposed of.
If the NOX in the tail gas is essentially NO and all the NO from the de-
composer is oxidized to NO2 and returned to the nitric acid plant (except
for enough NO2 to adjust the tail gas to N2O3) then equation (11) shows
that 2/3 of 37 or about 25 tons/day of NH4NO3 are produced. In either
case a very large amount of low value (about $48 /ton) waste product must
be accounted for.
There are no readily available cost data for the control of NOV
.A.
emissions in nitric acid plant tail gas. However, Bartok (10) has analyzed
the sensitivity of control costs for NO recovery as NH^NOs for a !0°° M
gas - fired power plant using the Mg (OH)0 scrubbing system. His results,
Cl
Figure 4, shows that the scrubbing system annual operating cost would be
completely paid for if a net credit back to the plant of about $60 /ton could
be realized by the sale of the waste
Some experimental data on NO removal from combustion gases using
A.
wet scrubbing with MgO is available from pilot studies using this scrubbing
medium for controlling particulate and SO2 emissions from the combustion
of pulverized coal (19). The observed efficiency for SO2 and particulate
removal was better than 99% for both materials in a floating bed absorber
and 95 - 98% for the pollutants in a venturi scrubber. The limited data
69.
-------
taken for NO removal showed that no reduction could be measured for
A.
NOX as produced in the combustion furnace. Assuming this was due to
the predominance of NO in the combustion flue gas, NO2 was injected
into the gas upstream of the scrubbers. Under these conditions the total
NOV at the scrubber exit increased significantly while simultaneously pro-
j£
ducing large quantities of magnesium sulfate in the scrubbing liquor.
These results can be explained by oxidation reactions between NO2 and
SOX in the gas and liquid phases. In the gas phase NO2 will oxidize SO2
to SO3 which will produce 112804 upon being absorbed in the MgO slurry.
In the liquid phase, NO will oxidize sulfite ions to sulfate. In both cases
L*
the NO2 will be reduced to NO and pass out of the scrubber unabsorbed.
A tentative solution to this problem was recommended in terms of a two-
step process in which SO0 is first removed by MgO scrubbing followed
&
by NO2 injection to produce N2O3 with NO in the combustion gas and
absorption in a Mg(OH) slurry.
£
Despite the lack of satisfactory operating information for full-scale
or pilot demonstration sized units, the basic concepts developed in the
Mg(OH)2 scrubbing system appear sound. Particularly attractive is the
flexibility enjoyed with respect to scrubbing materials; other substances
such as carbonates of magnesium and calcium, could readily substitute
for Mg(OH)2 depending on availability and price. The precipitation step,
shown here to produce NH4NO3 from NH3, could use as substitutes alka-
line CO2 to produce NaNOs or other alkaline salts. There is also a useful
70.
-------
flexibility in the operation of the process, particularly with respect to an
ability to treat stack gases having a large variation in NOX composition and
a possibility of removing a slip stream of NO2 from the oxidizer which
could be recycled to the source plant to produce more nitric acid.
Wet Scrubbing With Sulfuric Acid
The second scrubbing system to be described employs sulfuric acid
as the scrubbing medium. Interest in sulfuric acid as wet scrubbing
material is based on the well known chemistry of the chamber sulfuric
acid process and the important catalytic role played by nitrogen oxides in
sulfuric acid production.
The process sequence starts with the oxidation of sulfur to SC>2 and
ammonia to NC>2. Both gases are brought together in a tower where the
NO2 oxidizes the SC>2 to 803 while itself being reduced to NO. If the SO2
oxidation by NO2 is carried out in the presence of water and excess air,
sulfuric acid is formed from the SO3 and the NO is reoxidized to NOg.
Most of the acid was produced in large lead-lined chambers where
sufficient time was allowed for SO oxidation by NO2 and NO oxidation by
air, and enough water sprayed over the gases to allow the withdrawal of
62-66% H0SO,,, so-called chamber acid. NO and NO leaving the chambers
A (± A
were absorbed in strong H2SO4, about 80%, and at low temperatures, below
100°F to produce nitrosyl sulfuric acid.
71.
-------
NO + N02 (or N203) + 2H2SO4 = 2NOHSO4 + H2O (14)
which is soluble in sulfuric acid. Reaction (14) is reversible and
nitrosyl sulfuric acid will decompose if diluted below 80% H2SO4 and
heated. This is accomplished by pumping the strong H2SO4 containing
nitrosyl sulfuric acid back to the nead of the process, where it is contacted
with hot SO2 from the sulfur burner and diluted with chamber acid. The
regenerated NO and NO gases in contact with SO are lead to the chambers
£ <£
where the process is repeated.
An adaptation of the lead chamber process for the simultaneous
control of SO2 and NOX in flue gases from stationary combustion sources
has been developed by Tyco Laboratories under EPA sponsorship (11).
Although the Tyco process essentially employs the same basic chemistry
as the lead chamber process, significant changes in some methods were
required. These are explained below for the basic Tyco flowsheet shown
in Figure 5.
1. Nitrogen oxides, with an average composition of N?O3, are
recovered in the chamber process by absorption in concentrated (80%),
cold (about 60 F), sulfuric acid by forming nitrosyl sulfuric acid which
is soluble in cold, strong acid. In the Tyco process, nitrogen oxides with
an average adjusted composition as N2O3, is stripped from stack gases by
absorption in concentrated (80%), hot (about 250°F) sulfuric acid by forming
nitrosyl sulfuric acid which is soluble in hot, strong acid. The substitu-
72.
-------
tion of hot for cold sulfuric acid is due to the fact that cold concentrated
H2SO4 readily absorbs water vapor. If the water vapor present in flue
gases was absorbed along with NOX then the acid would become too dilute
to efficiently absorb NOX. By maintaining the scrubber exit at a tempera-
ture high enough to produce a partial pressure of water equal to its partial
pressure in the flue gas, there can be no net absorption of water vapor
and no dilution of the scrubber acid0 The scrubbed flue gas would carry
with it as much water vapor as was produced during combustion. Tyco
reports that N^Oo scrubbing efficiencies of 98% are generally possible
even though the absorption takes place at a much higher temperature than
was considered optimal in the lead chamber process.
2. In the lead chamber process, the nitrosyl sulfuric acid was
decomposed to form NO by dilution and heating, the heat being largely
supplied by the hot SO2 gases from the sulfur burners. This route is not
available in a stack gas cleaning operation and to provide the heat necessary
for a thermal regeneration is prohibitive. As an alternate, Tyco developed
a catalytic oxidation process operating at scrubber temperature which simul
taneously denitrates the sulfuric acid of nitrosyl sulfuric acid and oxidizes
the NO released to NO2. Tyco reports 99% recovery of N2O3 with their
catalytic process of the nitrogen oxides absorbed during high temperature
scrubbing.
73.
-------
3. The NO released from nitrosyl sulfuric acid in the lead
chamber process was mixed with air and SO2, the former to oxidize NO
to NO9 and the latter to be oxidized by the NO2 formed by air oxidations
^
of NO. In the Tyco process, some of the NO2 produced in the catalytic
decomposition of the nitrosyl sulfuric acid is injected into the stack gas
coming directly from the combustion source. Sufficient NO2 is blended
with the stack gas to oxidize the SO2 present to SO3 and to form the equi-
valent of N2O3 which is then absorbed along with SO3, in the high tempera-
ture scrubber. Part of the NO2 from the catalytic unit is absorbed with
water to produce nitric acid and NO which is air oxidized and blended with
flue gas. The net result of the overall Tyco process is the production of
sulfuric and nitric acids from stack gases containing NOX and SO2. An
economic estimate of the costs for an 800 MW power plant burning coal
containing 3. 5% sulfur showed that 27, 700 tons of HNOo (100%) and 252, 000
O
tons of H2SO. (100%) could be produced. Capital investment for control
equipment would be about $12 million and the process would be economically
self-supporting if HNO3 could be sold for $40 /ton (100%) and H2SO4 sold
for $10/ton (100%).
Bartok (10) points out that the need to oxidize SO9 with NO0 from
& £
recycled NO adds a huge burden to the absorption process just to reduce
the NO in the flue gas to a reasonable level. For example, if the flue gas
contains 2000 ppm of SO2 and 1000 ppm of NO then 5000 ppm of recycle
N02 are required to oxidize the SO2 and produce the N2O3 necessary for
74.
-------
absorption. Therefore, if a 90% reduction of the original NOX in the flue
gas is required then the scrubber must operate at 98. 3% efficiency on the
6000 ppm of N2O3 actually delivered to the scrubber inlet. Bartok's
estimate of control costs for a 1000 MW coal-fired power plant controlled
by the Tyco process is shown in Figure 6.
WET SCRUBBING WITH COMPLEX FORMING SUBSTANCES
We have examined thus far, scrubbing systems using aqueous solvents
which combine with nitrogen oxides and convert them to nitrates, nitrites,
and nitric acid. A completely different class of scrubbing agents, which are
potential solvents for nitrogen oxides, are those which form complexes
with NO and NO2 of the .charge transfer or coordination type.
Nitric oxide, which contains an odd election, is particularly sus-
ceptible to forming complexes with many metals and salts due to an easy
ability to lose or share its odd election. In general, such complexes,
generally classified as nitrosyl compounds, are readily decomposed,
sometimes simply by heat, to the original reactants without serious degrad-
ation. Two principal types of products are formed. The first containing
ionic species such as NOCl(cationic compounds) or species such as NaNO
(anionic compounds) both formed by simple election transfer. The second
group is composed of coordination compounds in which a pair of elections
is donated by NO to a central metal atom. Examples of complex nitrosyls
75.
-------
include (1) CuNOXg formed by contacting NO with concentrated copper
chloride or bromide solutions, (2) [CoNO(NH3)5 ] X2 formed by replacing
NH3 in hexammine cobaltous salts [Co(NH3)g] X2, where X is either a
halide or sulfate and (3) Co(CO)3 NO formed by replacing CO in cobalt
tetracarbonyl Co(CO)4_ The formation of nitrosyl complexes is described
in detail by Moeller (13).
Among the most important complexes formed are coordination
+n
compounds of the type M(NO)X Ay where M = Fe, Co, Ni, Mn and
A =C1, SO4. Ganz and Mamon (14, 15) have critically studied one of
these complexes, that formed by contacting NO with Fe SO^ solutions.
The solubility of NO in FeSO^ solutions appears to be the greatest of all
the possible metal chlorides and sulfates belonging to this group. NO
solubility in 20% FeSO4 is of the order of 1200 times the solubility of NO
in water at 20 C.
Although this system is considered a classical example of the
complexing ability of NO, it is less than an ideal choice for a practical
scrubbing system. The presence of oxygen along with NO in the gas to
be cleaned will readily oxidize the solute to ferric sulfate which does not
complex with NO. Also, any NO2 present that will dissolve in water and
form HN03 will oxidize the FeSO4 to Fe2
-------
In-the absence of oxidizers such as O2 and HNO3, ferrous sulfate can be
regenerated many times.
In addition to inorganic salts, .some purely organic materials and
organometallics have been suggested as possible candidates for complex
formation with NO. Mauryand Nahill (16) have patented the use of such
materials as dime thy Iformamide, dimethylether and dioxane for this
purpose. They find that complex formation can occur in nonaqueous solu-
tions of the complexing material but that the presence of controlled amounts
of water increases the overall absorptive capacity of the system. The
suggested reason for this is that the organic complexing material acts as
a transfer agent, first complexing the NO and then releasing NO to the
water phase where nitrous acid is formed. Other materials reported by
Maury and Nahill as having a significantly increased absorptive capacity
for nitrogen oxides compared to water includes tri - n-butylphosphate, tri-
phenylphosphate, and dimethylsulfoxide.
Although there are some obvious objections to using low molecular
weight organic complexing agents with vapor pressures that are sufficiently
high to cause significant losses of the complexing material at normal stack
temperatures, there has been at least one practical use for such materials
involving NOX recovery. In 1970, Nash (17) reported the use of 2-methoxy-
phenol (quaiacol) as an absorption agent for use in sampling bubblers for
collecting NO9 in ambient air. Collection efficiencies as high as 99% were
77.
-------
reported for quaiacol-NaOH absorbers compared with about 70% when
NaOH was used alone. Atmospheric ozone and SO2 oxidize quaiacol easily
and interfere with the absorption process which makes the material in-
appropriate for cleaning stack gases, but the use in air quality sampling
systems is satisfactory.
At the Cooper Union, we have been examining as yet another complexing
agent which shows promise as an NO scrubbing material. In 1962,
Silvestroni and Ceciarelli (18) reported that aqueous solutions of cobalto-
dihistidine were efficient NO absorbers. Although the testing of this
•material has been exploratory only, some tentative results are available.
Figure 7 shows the solubility of NO and NOg in cobaltodihistidine solution
at 20 C. Only measurements at very high NO pressures have been made
.X.
and the solubility behavior at low pressures is by extrapolation and is sketchy
and tentative. Of considerable interest, though, is the indication of a
possible affinity of the solution for both NO and NO9 absorption, a valuable
-------
is-an example of an NOx recovery process for a 200 ton/day nitric acid
plant. The tail gas from the nitric acid plant is passed through an
absorption column where the NOx gases are absorbed, the lean gas is
then returned to the plant for power recovery and discharged to the atmos-
phere. NOx picked up by the absorber scrubbing Hquor is stripped
and returned to the nitric acid plant while the regenerated scrubbing
liquor is recycled to the absorption tower. Assuming an arbitrary 88%
reduction of NOX in the plant tail gas, the recovered gases can produce an
additional 5 million Ibs. of 40°Be acid per year for a net value of about
$85, 000 after deducting the cost of manufacturing this amount of acid
($30/ton) from the selling price ($83/ton, 100% basis). No attempt has
been made to estimate the cost of recovery system with the spare infor-
mation now available. A comparable flow sheet for NO recovery from
.X.
combustion systems has not been attempted since there is no specific data
on the effect of SO2 in the stack gas on the NG>X process.
Conclusion
This review has attempted to indicate the current status of wet
scrubbing as a possible control method for the removal of nitrogen oxides
from stationary sources. In so doing, it has been necessary to highlight
the field, selecting from among many possible scrubbing systems which
have been suggested as having potential for NOX control. We have purpose
ly avoided the problem of absorption dynamics and the related area of
79.
-------
absorption equipment in favor of describing the physical chemistry of ab-
sorptive processes for NOX> illustrating where possible the application
of specific chemical systems to stack cleaning processes. It is hoped
that the examples used have served to show the rich potential for NOX
control that can result by the resourcefull combination of sound chemical
and engineering principals.
From the material presented, it should be quite evident that the
desired goal of NO control by wet scrubbing has not been achieved as yet.
Ji
In come cases proposed systems await demonstrations of their theoretical
ability to provide adequate control; in other cases more research is apparent
ly necessary to determine the potential for control. In either case it is
hoped that wet scrubbing methods will continue to be explored as a possible
method for controlling NOX emissions.
80.
-------
ACKNOWLEDGEMENT
The author wishes to express his appreciation to Mr. Stuart Roth,
Air Programs Division, Region II, U. S. Environmental Protection
Agency for his assistance and guidance with EPA publications and research
activities and to Mr. Robert Meier for his thorough effort in surveying
and organizing the technical literature for this report and particularly for
his patience and contributions to our many discussions on environmental
control.
81.
-------
References
1. Partington, J.R. , " A Text-Book of Inorganic Chemistry",
Macmillan and Co. , Limited, London, 1939.
2. Yost, Don M. and Horace Russell, Jr., "Systematic Inorganic
Chemistry", Prentice-Hall, Inc., New York, 1946.
3. Hoftyzer- P. J. andF.J.G. Kwanten, "Absorption of Nitrous
Gases", Chapter 5, Part B, p. 164 in "Processes for Air Pollution
Control", G. Nonhebel (ed. ), CRC Press, Cleveland, Ohio 1972.
4. Lowell, Philip S. , et. al., "A Theoretical Study of NOX Absorption
Using Aqueous Alkaline and Dry Sorbents", Radian Corporation,
NTIS PB 211035, December 31, 1971.
50 Chappell, Gilford A. , "Development of the Aqueous Processes for
Removing NOX from Flue Gases", Esso Research and Engineering,
NTIS, PB 212858, September 1972.
6. First, Melvin W. and Frederick J0 Viles, Jr. , "Cleaning of
Stack Gases Containing High Concentrations of Nitrogen Oxides",
JAPCA, 2_3 (No.3), 122 (1971).
7. Morrison, M0E., Rinker, R. G, and Corcoran, W. H. , "Rate
and Mechanism of Gas-Phase Oxidation of Parts-per-million
Concentrations of Nitric Oxide", Ind. Eng. Chem. Fundamentals,
5_ , 1975 (1966).
8. Strom, S, S., "The Absorption of NO2 with a Venturi Scrubber",
Paper 67C, 67th Annual Meeting A. I. Ch. E. , Atlanta, Georgia,
February 1970.
9. Peters, Max S. , Chemical Engineering, 200 (May, 1955).
10. Bartok, W, et. al. , "Systems Study of Nitrogen Oxide Control
Methods for Stationary Sources - Vol. II, Esso Research and
Engineering, NTIS PB 192789 , November 20, 1969.
82.
-------
11. Walitt, Arthur and Arnold Gruber, "A Process for the Manufacture
of Sulfuric and Nitric Acids from Waste Flue Gases", Second
International Clean Air Congress, Englund, H. M. and W. T. Beery
(ed. ), Academic Press, New York, 1971.
12. Schmidt, A. and F. Weinrotter, U. S. Patent 3,034,853, May 15,
1962.
13. Moeller, T. " Inorganic Chemistry", John Wiley and Sons,
New York, 1952.
14. Ganz, S. N. and L. I. Mamon, "Absorption of Nitric Oxide by
Ferrous Sulfate", Zhur. Priklad. Khun., 26^, 1005 (1953).
15. Ganz, S. N. and L. I. Mamon, "Kinetics of Film Absorption of
Nitric Oxide by Ferrous Sulfate", Zhur. Priklad. Khim. , 30,
391 (1957).
16. Maury, L.G. and Nahill, G.F., U.S. Patent 3,044,344, July 17,
1962.
17. Nash, T. , "An Efficient Absorbing Reagent for Nitrogen Dioxide",
Atmospheric Environment, £,_661 (1970).
18. Silvestroni, Paolo and Laura Ceciarelli, La Ricerca Scientifica,
2, 121 (1962).
19. Downs, W. , et. al., "Magnesia Base Wet Scrubbing of Pulverized
Coal Generated Flue Gas - Pilot Demonstration", Babcock and
Wilcox Company, NTIS PB-198074, 28 September 1970.
83.
-------
TABLE 1
NOx ABSORPTION STUDIES
oo
-p-
(SELECTED LITERATURE - 1962/1972)
(1)
REF.
1
2
3
4
5
6
7
8
TITLE: PRODUCT:
. HYDROXIDES (MOH)
Abs. of NOx by NaN02 3
vibrating layers
of NaOH
Abs. of NOx with Ca(N02,3)2
milk of lime
Abs. of NOx in NH4 N02.3
aqueous NH3
CARBONATES (MCOs)
Abs. of NOx by NaN02,3
vibrating layers
of NaCOs
Abs. of NOx (w. NaN02,3
Na2C03) ..in
large scale plate
col-
Abs. of NOx by Na2 NaN02»3
C03 solutions ...
Abs. of NOx with Ca(No2,3)2
limestone suspension
Abs. of NOx in Ca(N02,3)2
solid CaC03
EQUIPMENT/ NOx FEED a
METHOD CONC. % (N02/NOX) n (%)
Vibrating
film
Mech.
Scru bber
Closed coni- 0.3-8 — 70-54
cal abs. with
sprayer.
Vibrating .55-1.8 .2-. 6 87-98
film
12-perf. tray .3 .5 75
scrubber
2 packed cols.
ceramic rings
15 tray bubble
cap-foam layer
ABSORBENT EFFICIENCY, ^ AS FUNCTION
CONC. OF SYSTEM PARAMETERS:
"n- max., f f [N02 or 02]
0 30 g/1
f[ f f LCaC03]
or TCfoanO.
Salt recovered
by H20 rinse.
(1) References for Table 1 will be found at end of Table.
Continued ,,,
-------
TABLE 1 (continued)
00
REF.
9
10
11
12
13
14
15
16
TITLE:
Abs. of NOx by
(NH4)2C03 solutions
(Abs. of) .. NOx
with (NH4)2 C03 &
NH4HC03
Scrubber for
selective removal
of N02 fr. NOx
(w. Ag.2 C03+PTFE)
ALKALINE SOLNS.
Abs. of NOx in solns
of NaOH & Na2C03
Alkaline abs. of
NOx
Removal of NOx . . .
. by H202 solutions
NOx absorption in
alkaline solns.
SALTS
Abs. of .. NOx by
aq. solns. of KMn04
PRODUCT:
NH4N02.3
NH4N02.3
HN03
Ag N03
NaN02,3
MN02.3
-™o2,3
MN02.3
KN03
EQUIPMENT/ NOx FEED . a ABSORBENT EFFICIENCY, n AS FUNCTION
METHOD CONC. % (N02/NOX) n U) CONC. OF SYSTEM PARAMETERS :
Column « .5 10-1'5 -- ?if as [N203]f , LAbsbt.Jf
4.5-10.4 .5 31-67 .2-1.7 M
8. 1 ppm 99
Vibrating
Layer
Column w. .6-1.2 .3-. 6 to 90*° 110 g/1. Tjf as <¥.t, [N0x]4 , T|
centrifugal
sprayer
Foam forming .7-1.0 .2 90-93*** 1-3 g/1 ^tas[NO]t , T| ,
apparatus (absorbt.)t
.5-2.5 ^.5 ^(NaOH) > ^ (NaCOs)
Foam absorber ~ — 92 * ^.5
2 perf. plates
COMMENTS:
HC03/C03 varied
fr. 0-1
100% Ad2 C03 or
PTFE +110-70%)
Abs. "complete"
10" ^ T ^ 40*
Oxidation rate
increases w.
KMn04 addition &
as [H202]| to 5%
* At optimum
conditions.
Continued ...
-------
TABLE 1 (continued)
oo
REF.
17
18
19
20
TITLE:
Abs. of NOx by
solns. of NaC03
in presence of
recycled NaN02,3
Abs. of NOx by
solns. of
di substituted
ammonium phosphate
PRODUCT:
»a»2,3
EQUIPMENT/
METHOD
NOx FEED
CONC. %
(N02/NOX)
,„. ABSORBENT
^/a) CONC.
[Na2C03]
<3 g/1
[Salts] =
34-36%
EFFICIENCY, '
OF SYSTEM
''I fas [N20;
and as [N02~,
n AS FUNCTION
PARAMETERS
N03]f*
COMMENTS:
* *? (NOf)
NH4N02,3
phosphates
& H3P04
Abs. of NOx by NH4N03
NH4N03 solns. & & HN03 *
HN03
Removal of NO fr.
industrial gases
(w. various salts*)
3-10% .5-. 9
8-stage model 8-10 .9-. 95
^97
40% H3P04 Tjtas [N0x]t , °
-------
Reference for Table 1
1. Mirev, D. et. al. , "Absorption of Nitrogen Oxides by Vibrating
Layers of NaOH Solutions", Compt. Rend. Acad. Bulgare Sci.
_14 , 259-62 (1961). CA 56:66a.
2. Ganz, S. N. , et. al. , "Absorption of Nitrogen Oxides with Milk
of Lime in Mechanical Absorbers", Khim. i Khim. Tekhnol. 5,
No. 1, 155-9 (1962). CA 57: 3245e. ~
3. Pawlikowski, Stanislow Aniol, et. al., "Absorption of Nitrogen
Oxides in Aqueous Ammonia in Nozzles. I". Przemysl Chem.
42 (9), 490-4 (1963). CA 60: 6501c.
4. Mirev, D. , et. al., "Absorption of Nitrogen Oxides in a Vibrating
Layer of Sodium Carbonate Solution III", Compt. Rend. Acad.
Bulgare Sci. _14, 345-8 (1961). CA 56: 5795c .
5. Kuz'minykh, I. N. , "Absorption of Nitric Oxide from Nitrogenous Exhaust
Gases in Large-scale Plate Columns I. ", Tr. Mosk. Khim. -Tekhnol.
List. 1961, No. 33, 43-7. CA57:2022h.
6. Krustev, I. , "Absorption of Nitrogen Oxides by Sodium Carbonate
Solutions Under Industrial Conditions", Khim. Ind. (Sofia) 1968,
(4), 147-51. CA 69: 78766y.
7. Rodionov, A. I. , et. al. , "Absorption of Nitrogen Oxides With A
Suspension of Limestone", Tr. Mosk. Khim. - Tekhnol. Inst.
1963 (40), 74-7. CA 61: 6463a.
8. Blasiak, Eugeniusz and Jariczek, Witold, Pol. Patent 43, 284
"Absorption of Nitrogen Oxides from Dilute Mixtures, Aug. 30,
1960. CA57: 7075f.
9. Atroshchenko, V.I. and B. N. Gushchin, "Kinetics of Absorption of
Nitrogen Oxides by Ammonium Carbonate Solution", Zh. Prikl.
Khim. _39 (12), 2627-33 (1966). CA 66: 57341d.
10. Gushchin, B. N. and V. I. Stroshchenko, "Relative Reaction Rat es
for Nitrogen Oxides with Ammonium Carbonate Compounds in Solution
and in the Gaseous Phase", Izv. Vyssh. Ucheb., Zaved, Khim.
Tekhnol. 10(3), 314-18(1967). CA 67: 76605c.
87.
-------
11. Neti, Radhkrishna, et. al., "Scrubber for the Selective Removal
of Nitrogen Dioxide from a Nitrogen oxide containing Gas Flow",
Ger. Patent 2,209,877, Sep. 1972. CA 77: 156038u.
12. Mirev, D., et. al. "Absorption of Nitrogen Oxides in Solutions of
NaOH and Na2CO3" Izv. Inst. Obshcha Neorg. Khun. , Org. Khim.,
Bulgare Akad. Nauk. _8, 83-101 (1961). CA 57: 5752h.
13. Ganz, S. N. and I.E. Kuznetsov, "Alkali Absorption of Nitrogen
Oxides in a Column Provided with a Centrifugal Sprayer", Zh. Prikl.
Khim. _36_(8), 1693-7 (1963). CA bO: 7694h.
14. Kuznetsov, I.E., S. N. Ganz and N. P0 Shpak, "Removal of Nitrogen
Oxides irom Exhaust Gases by Aqueous Hydrogen Peroxide Solutions"
Dneproopetrovsk. Khim. -Tekhnol. Inst. 1968, No, 13, 77-81.
CA. 71: 73740s.
15. Krustev, Iv. "Kinetics and Mechanism of Absorption of Nitrogen
Oxides in Alkaline Solutions. 2, " Izv. Otd. Khim. Nauki, Bulg.
Akad. Nauk. 1970, 3 (2), 203-13. CA 74: 43869.
16. Kuznetsov, I.E., S. N. Ganz and R. S. Sokol., "Absorption of
Weakly Acidic Nitrogen Oxides by Aqueous Solutions of Potassium
Permanganate", Dnepropetrovsk. Khim. -Tekhnol, Inst. , 1967 ,
No. 8., 162-5. CA 70: 152046.
17. Krustev, Inv. , "Absorption of Nitrogen Oxides by Solutions Contain-
ing Sodium Carbonate, Sodium Nitrite, and Sodium Nitrate", Izv.
Otd. Khim. Nauki., Bulg. Akad. Nauk. 1968, 1 (1) 109-23.
CA 70: 107834x.
18. Pozin, M.E. , et. al., "Absorption of Nitrogen Oxides by Solutions of
Disubstituted Ammonium Phosphate", Massoobmeiinye Protsessy
Khim. Tekhnol. 1969, No. 4, 174. CA 73: 37016t.
19. Pozin, M.E., et. al. , "Absorption of Nitrogen Oxides by
Ammonium Nitrate Solutions. Ill", Massoobmennye Protsessy
Khim. Tekhnol. 1969, No. 4, 174-6. CA 73: 37025v.
20. Bresan, Giancarlo and Salvatore Gafa, "Removal of Nitric Oxide
from Industrial Gases", U.S. Patent 3, 635, 657, January 18, 1972.
88.
-------
TABLE 2
NOx SOLUBILITY STUDIES
oo
(FROM G.A. CHAPPELL, "DEVELOPMENT OF THE AQUEOUS PROCESSES FOR REMOVING N(
NTIS PB212858 - 1972)
(1) NO & N02
CLASS
WATER
ABSORBENT
Water
ALKALINE NaOH
AMINE
ACID
SALT/
ESTER
NaOH
NaOH
NaOH
Ca(OH) si.
Ca(OH)2 sat.
Mg (OH)2 si.
Mg (OH)2 sat.
ZnO si.
ZnO sat.
NH40H
2-aminoethanol
ii 11
i i
H2S04
HOAC
NH4C1
NH40AC
NH40 Citrate
NaOAc
COMP. pH 6.6
?7 f f (Na, Ca, Mg ..) @ const. pH
-F [2AE]
(1) Unless otherwise specified: T = "Std. 125°F. (120-130), aQ = 0.5 (.46 - .54),
(2) NO was generated during NO, absorption
(3) The "comparison" list shows results for different conditions in Chappell's report
[NOX]
650 - 800 ppm.
continued
-------
CLASS ABSORBENT
SULFITES'S BISULFITES
COMP.
% ABSORBED
NO
NO 2
Na2S03
Na2S03/NaOH
Na2S03
NH4HS03
+ NH4S04
NaS
CaS03 si.
CaS03 si.
CaSOs si.
CaSOs si.
Sat. 0.5
Sat/5N
2.5 m
1.0
3.6 m
1.9
2.3 m 12
8 g/l ?
I 7'2
I 8'° 1
T 12 '
/Std. 21
12
16
15
24
Low
33
35
35
" 32*48
100
100
100
100
100
100
60
66
60
100*4
TABLE 2 (CONT.)
NOx SOLUBILITY STUDIES
• NO & N02
COMPARISON
% ABSORBED
NO
NO 2
Same 0 70-85"F. 27 TOO
Same @cxe=.6,855 ppm 18 TOO
COMMENTS
1? = f IT)
Lso3=j
Same 00^=1 .,370 ppm
100
continued
-------
TABLE 2 (Cont.)
NOx SOLUBILITY STUDIES
NO, N02 & S02
[NOx]
860
690
700
950
700
680
830
680
700
650
650
650
490
725
a.
1
1
1
0.47
1
1
1
1
1
1
1
1
0
0.45
-Jh
9.1/7.2
9.4/3.7
5.9/5.9
6.3/6.3
11.2/6.5
11.2/6.4
8.7/7.8
9.3/7.9
9.3/6.3
7.5/6.2
7.2/6.3
_
-
% ABSORBED
NO N02
TOO
- 100+45*
94
22 100
63
56
58
56
63
46
46
6 100
0
8 37
(S02)
(TOO)
(TOO)
83
(97)
(100)
(100)
(100)
(100)
(100)
MOO)
(100)
(95)
'(70)
(70)
ALTERNATE CONDITION % ABSORBED
ABSORBENT [NOx] Ou LS02J pH NO N02 (S02)
CaS03sl.8g/l 330 1.0 500 7.6/5.0 - 60 92
Ca(OH)2 si. 10 g/1 740 1 0 11.2/6.6 19
Mg(OH)2 si. 680 0.5 0 8.9 6 23
ZnO si. 723 0.5 0 7.5 7 16
Na2S 2.3 m 700 0.5 - 12 Low 100
Urea 3.8 m 480 0 - 3
ABSORBENT
SULFITES
CaS03 si. 25 g/1
MgS03 5 g/1
NH4HS03 2m 7
(NH4)2S04 0.5 m j
(NH4)2S03 Mixt. **
ALKALINE
Ca(OH)2 si. 15 g/1
18 g/1
Mg(OH)2 si. 7.4 g/1
Mg(OH)2 si. 14 g/1
ZnO si. 10 g/1
CaCOs si. 10 g/1
CaC03 si. 4 g/1
MISCELLANEOUS
2.1 m
Urea 3.8 m
Urea 3.8 m
* MgSOs depleted by 02 in flue gas?
** Mixt. = 11.2% NH4HS03, 14.6% (NH4)2S03, 16.6% (NH4)2 S04 in H20.
-------
TABLE 3
NOX Absorption In NaOH Solutions
Data of First & Viles (Ref. 6)
SCRUBBER INLET - SCRUBBER EFFICIENCY, % REMOVED
Run NO %NOX
1 0.
2 9.
3 5.
4 32.
845
60
63
8
%NO2
0.460
5.91
4.87
19.0
%NO
0. 385
3.69
0.76
13.8
a in
0.54
0.615
0.865
0.58
NCL NO2
90. 3
90. 3
92.6
97. 3
94.
94.
96.
98.
4
9
0
1
NO
85.
82.
70.
96.
a Removed
4
B
2
1
0.
0.
0.
0.
57
65
90
58
-------
NO,
NO I «*.
GAS-BULK
GAS-FILM
INTERFACE
LIQUID-FILM
FIGURE 1
NOX ABSORPTION MODEL
HN0^2NOH20
LIQUID-BULK
(REF. 3)
-------
FIGURE 2
Alkaline Scrubbing
NO + MOH
N2O4 (or 2 NO2) + MOH
N2 O3 +2 MOH
salting out of NO
MNO2 + MNO3 +
2 MNO,
H20
Carbonate Scrubbing
NO + M2CO3
N2O4 (or 2NO2) + M2CO<
N2 + M2CO3
salting out of NO
MNO
2MN02
MNO
CO
* where M = Na , K ,
+ +2
,• i Ca . etc.
** solubility of NO is reduced below that of NO in water by the
presence of the soluble hydroxide or carbonate.
94.
-------
FIGURE 3
Mg(OH)2 SCRUBBING PROCESS
STACK GAS
FROM
HN03 PLANT
TREATED
STACK
GAS
TO HN03
PLANT <-
NO,
0
X
I
D
I
Z
E
R
t
AIR
NO
SCRUBBER
DECOMPOSER
Mg(OH)2
Mg(NOs)2
NH3
(RECYCLE)
Mg(N02)2
(RECYCLE)
AMMONIATOR
Mg(OH)2
NH4N03
SOLUTION
HZ)
PRECIPITATED
Mg(OH)2
95.
-------
FIGURE 4
SENSITIVITY OF CONTROL COSTS
TO BY-PRODUCT CREDIT
SINGLE SCRUBBER Mg (OH)2 PROCESS
WITH THERMAL DECOMPOSITION
(1000 MW GAS-FIRED POWER PLANT )
(REF. 10)
2400
2000
o 1600
o
o
•ea-
H 1200
O
O
I
80°
400
400
50
25
_ C2S_T_ _
PROFIT
O
P
W
>
0
w
K
x
O
O
10
20
30
40
50
60
70
NO3 PRODUCT CREDIT $/TON NET BACK TO PLANT
96.
-------
FIGURE 5
TYCO CATALYTIC CHAMBER PROCESS
Gas to Stack
250"F'.
7% H20
150 ppm NOX
Gas
NO
N02
S02
Reactor
Air - NO
Flue Gas
0.3% S02
0.6% NOX
H20
HNOs
Absorber
Product
HNOs
High
Temperature
Scrubber
Gas
H2S04
Air
Product
H2S04
.80% H2S04
80%
H2S04
+
HNS05
Catalytic
Stripper
80%
H2S04
97.
-------
2400
o 1200
O
u
800
400
FIGURE 6
SENSITIVITY OF CONTROL COSTS
TO BY-PRODUCT CREDITS
SULFURIC ACID SCRUBBING SYSTEM
(1000 MW COAL-FIRED POWER PLANT)
(REF. 10)
2000 _
1600 —
0 L_
400
($/Ton Net Back Product
Credit for 80% H2SO4)
CNJ
O
ra
cd
w
K
O
100
HNO3 PRODUCT CREDIT, $/TON NET BACK TO PLANT
98.
-------
FIGURE 7
800-
600
CD
E
LJ
V}
V)
Ul
I-
o:
400
200
/
/
/
/
/ /
/ /
//
//
/ /
//
'/
SOLUBILITY OF NO & NOx IN
AQUEOUS COBALTODIHISTIDINE
(3 20°C.
9 gms CoCl? + 31 gms Histidine
per liter solution
$ = N02
@ = NO
I
0
i
//
0
4
8
12
16
20
Gm NOx / Gm SOLUTION
99.
-------
o
o
SCRUBBED TAIL GAS
85°F., 92 psig
[NOxJ = 350 ppm
CN2J = 96.2 %
[02J = 3.1 %
[H20] = 0.7 %
RECYCLE TO
PROCESS
85°F., 92 psig
NOx 10 %
H20 90 %
WASTE HEAT BOILER
L>
EXPANDER
TURBINE
J
c
STRIPPER
• ft*
TFAM ""
^ •— J
/ 0
J
te
xr^x
V V
^^ -^x
COOLER
S^
ABSORBER
f/\ ^ \^
.j.
TAIL GAS FROM
HMO? TOWER
fti
85UF., 92 psig
[NOx] = 3 %
[N?] = 96.1
[02] = 3.0 %
[H20] = 0.6 %
ABSORDER - STRIPPER RECOVERY SYSTEM
-------
EPA-View of Stationary and Mobile NO -Source Control
by
Conrad Simon
Introduction
EPA's view of the NO problem has undergone considerable change
since the passage of the Clean Air Act of 1970. A complete description
of EPA's policy reflecting this change will be forthcoming within the
next month or so. The most significant effect of this change in outlook
was reflected in Administrator Ruckelshaus' testimony before the Senate
Sub-committee on Air and Water Pollution during the week of April 16, 1973.
The Administrator asked the Senate Committee to rescind the statutory re-
quirement under Section 202 (b) of the Clean Air Act that light duty
vehicles and engines manufactured during and after the 1976 model year achieve
a reduction of at least 90 per cent from the average emissions of oxides of
nitrogen actually measured from light duty vehicles manufactured during the
1971 model year. The 1971 emissions averaged 3.5 grams per mile and the 1976
standard is 0.4 grams per mile . The Administrator has also requested that
Congress grant him the power under section 202 (a) of the Act to set standards
as he sees fit. If this were granted it is likely that a standard would be
set significantly higher than the 0.4 grams per mile currently in effect for
1976. Applicable standards expressed in terms of the 1975 Federal CVS procedure
are given in Table I.
EPA is particularly concerned about changing the 1976 emissions standards
for oxides of nitrogen in motor vehicles because it no longer appears necessary
for the achievement of national ambient air quality standards and because it
represents a most difficult emissions control goal to achieve. In fact it has
been determined that attempts to achieve the emission standard for oxides of
nitrogen using a catalyst system could jeopardize achievement of the CO and
hydrocarbon standards while imposing a sizable fuel penalty of about 15%. If
given this authority, the Administrator would not change the standard immediately,
101.
-------
but would wait until new health effects data were obtained and evaluated. The
standard would then be set at a level that is adequate to protect against health
effects and to provide for maintenance of the national ambient air quality standard
by minimizing the impact of growth on existing ambient concentrations. In
setting new emission limitations to maintain ambient air quality standards
for nitrogen dioxide (N02) EPA would seek to obtain a balance in the requirements
imposed on stationary sources and mobile sources.
This revision in EPA's evaluation of the N02 problem is primarily the
result of the Agency1 re-examination of the air quality measurement method
used in various presentations to establish the status of the N0? pollution
problem. Some of these presentations were made before Congress prior to the
passage of the Clean Air Act of 1970 and some constituted the air quality data
used in the development of States' implementation plans. EPA is now convinced
that important portions of the data based used in these two activities were
unreliable. It should be stated clearly at this point that a national ambient
air quality standard has been established specifically for nitrogen dioxide
(N0?). No air quality standard has been established, nor is contemplated, for
nitric oxide (NO). Emission standards and control strategy, on the other hand,
have been expressed in terms of oxides of nitrogen in recognition of the fact
that the major source of N0_ associated with significant ambient concentrations
is NO contained in the atmosphere as a result of emissions from combustion
sources. Only a small and relatively insignificant amount of N09 is produced
directly and emitted by these sources.
102.
-------
Air Quality Data Base
For several years various investigators have argued that discrepancies
existed hetween air quality measurements of NO made with the Jacobs-Hochheiser
procedure used in the National Air Surveillance Network (NASN) and air quality
data obtained through various other sampling techniques, particularly the
Saltzman method. This argument climaxed on April 30, 1971, when EPA promulgated
the National Ambient Air Quality Standards. The standard analytical methodology
to be used to measure the ambient concentrations of each of the six pollutants for
which national standards were being set was also stipulated. These were referred
to as "reference methods". The reference method for NO was the Jacobs-Hochheiser
method. The regulation establishing this "reference method" would have to meet
certain criteria in order to demonstrate their equivalence to the reference method.
These criteria will be published in the near future.
When States attempted to develop implementation plans in 1971, they found
that air quality data for carbon monoxide, nitrogen dioxide and photochemical
oxidants were generally sparse. In response to this need EPA in the summer of 1971
conducted a special study of air quality levels for these pollutants in those
urban areas where data were sparse. This Summer Study using the Jacobs-Hochheiser
method became the source of a considerable amount of the ambient NO data used by
the States to develop their implementation plans.
Based on the results of the Summer Study and various other data available
from the NASN stations, state and local agency networks, it was determined by the
end of 1971 that 47 Air Quality Control Regions (AQCR) in 29 states should be class-
ified Priority I for N0.
103.
-------
Priority classifications of either I or III are assigned to AQCR's as a means of
indicating the degree of control of emissions of oxides of nitrogen that may be
necessary to provide for the achievement of the NO standard. Priority I is
assigned to those regions where air quality levels have equalled or exceeded
110 ug/m3 or 0.06 ppm as an annual arithmetic mean. Those with lower concentrations
are assigned a Priority III classification. The national ambient air quality
3
standard was set at 100 ug/m or 0.05 ppm as an annual arithmetic mean.
In order to determine the control strategy necessary to meet air quality
standards for NO in these regions, Federal regulations required the following
procedures:
1. Assume certain emission reductions that will result from the Federal new
motor vehicle emission standards.
2. Take credit for the impact of any transportation control measures
taken to achieve carbon monoxide and photochemical oxidant standards.
3. Impose emission limitations attainable with reasonably available control
technology on stationary sources.
4. If 1, 2, and 3 do not achieve sufficient NO emission reduction of
X
hydrocarbons as may be possible by reasonably available control technology.
(Hydrocarons are associated with the conversion of NO to NO in the atmosphere).
EPA has determined that certain control technology for existing sources of
oxides of nitrogen is reasonably available for combustion sources and nitric acid
plants. The achievable emissions limitations are listed in Table II and compared
with performance standards for new sources. In 1972 some States adopted
regulations containing emissions limitations similar to these. Where States failed
to adopt such emissions limitations EPA proposed them. In deference to the
existing controversy over the accuracy of air quality data for NO. EPA refrained
from making final promulgations.
104.
-------
Measurement techniques - Prior to 1970 there was one basic technique used
for the measurement of ambient concentrations of NO and NO . This technique
involved the Griess-Ilosvay reaction for N02 which was utilized in the
Saltzman and the Jacobs-Hochheiser methods.
Other techniques not amenable to ambient sampling were gas chromatography,
long-path infra-red spectroscopy and electrochemical oxidation or reduction.
Over the past 6 years EPA has fostered the development of the gas phase
chemiluminescence technique for application to air quality monitoring.
The Griess-Saltzman method was deemed the most suitable manual method
for measurement of NO- in the atmosphere by the Standardization Advisory
Committee, National Air Pollution Control Administration and the Intersociety
Committee on Manual Methods for Ambient Air Sampling and Analysis. In this
procedure NO containing air is bubbled through the Griess-Saltzman reagent
for a period of up to 30 minutes. The Saltzman reagent consists of, among
other things, sulfanilic acid and N-(l-naphthly)-ethylenediamine dihydrochloride.
NO. reacts with these compounds to form a diazo dye with a characteristic color
whose intensity is proportional to the amount of NO absorbed. This method is
3
usable for NO in air of 40 - 1500 ug/m (0.02 to 0.75 ppm). Using sodium
nitrite as a calibration standard, Saltzman found that 0.72 mole of nitrite
produced about the same color as 1 mole of NO gas. The method is fairly specific
for NO and no significant interferences commonly occur. SO for example shows
significant interference only at concentrations of about 30 times that of N0_.
The problem with the field application of this method lies in the need to
limit absorption time to about 30 minutes and the analysis time to one hour
after color development. In order to develop a procedure to meet the need for
sampling over periods as long as 24 hours, the Jacobs-Hochheiser method was
developed.
105.
-------
In an EPA modification of this procedure sodium hydroxide (.IN NaOH)
is used as an absorbing reagent. Sulfanilamide replaces the sulfanilic acid
of the Saltzman reagent and the acid is phosphoric acid rather than acetic.
Since some of the NO absorbed goes towards the formation of sodium nitrate,
on the average only 0.63 mole of nitrite was required to produce the same color
as 1 mole of NO gas. This stoichiometric factor has been the source of
considerable controversy in recent years with values reportedly ranging from
0.5 to 1.00 (for the equivalent quantity of nitrite ion).
All NASN measurements have used an NO - Nitrite stoichiometirc factor of
1. 0. That is to say, only nitrites and not nitrate ions were assumed to be
formed during the absorption of the gas from the ambient air.
It was also determined that the absorption efficiency of NO was highly
variable with NO concentration and was also dependent on the amount of NO
present in the air. On the basis of detailed studies made under laboratory
conditions the NASN system was assumed to have an efficiency of approximately
35%.
The Saltizman method was also adopted to continuous analyzers and was used
in EPA CAMP stations as well as stations operated by various State and local
agencies. Most of these analyzers were also used for measurement of NO by
first oxidizing all NO to NO and subjecting it to the process of diazotization.
The oxidation process reportedly has an efficiency ranging from 40% to 100%
(using potassium permanganate or dichromate or chromium trioxide).
Differences in the air quality data obtained from the Saltzman and Jacobs-
^ochheiser methods at the same sites led EPA to make a careful examination of the
reliability of its reference method and the accuracy of the air quality data
used for priority classification.
106.
-------
A report on the first phase of the EPA investigations and the findings was
made by Hauser and Shy in October 1972.
In this first phase, a test of the efficiency of the reference method
was made using nitrogen dioxide - air mixtures of varying concentrations
generated by the use of NC>2 permeation tubes. For each of the test atmospheres
generated, at least five simultaneous sample were collected and analyzed. The
results shown in Figure 1 indicate that the collection efficiency of the
reference method varies nonlinearly with NO concentration from about 15% at 740
33 ^
ug/m to about 70% at 20-30 ug/m . For an ambient level of about 120 ug/m the
previously assumed efficiency of 35% is valid. Above that level, the use of a
35% efficiency will underestimate the actual concentrations in ambient air. At
lower ambient concentrations, the reference method will result in erroneously
high estimates of concentration if a 35% efficiency is assumed.
The effect of the presence of nitric oxides (NO) on the reverence method
was also examined. The response of this method to various concentrations of
NO. with and without NO is given in Table III. A comparison of the expected and
apparent N0_ recovered shows positive interference from NO.
Since these problems with efficiency and interference were related to the
absorbing solution, EPA examined several alternate solutions for possible
adoption. Three different absorbing reagents were tested in the NASN network
and two were rejected because of physical problems even though collection
efficiencies were higher and more consistent than those of the Jacobs-Hochheiser
Method. Hauser, T. and Shy, C: Position Paper: NO Measurement. Environmental
X
Science and Technology, October 1972.
107.
-------
The sodium arsenite - sodium hydroxide reagent has been found most satisfactory
by EPA. This reagent (the method is called the Christie method) has been
tested at all 200 stations of the NASN network since December 1971, and has
recently replaced the Jacobs-Hochheiser method in routine sampling.
In the second phase of this investigation beginning August 1972, cherailu-
minescent equipment has been installed at selected urban sites in 41 of the
AQCR's classified Priority I and at all CAMP stations where instruments using
the continuous Saltzman methods were located.
The results of these measurements show that of the 47 AQCR's previously
classified as Priority I only two, Los Angeles and Chicago, will definitely
retain their former classification. Three other regions, Salt Lake City, Denver
and New York, are marginal. The remaining 42 AQCR's will all be reclassified
Priority III.
Chemiluminescent analysers will also be maintained in 17 of the 47 AQCR's
3
where annual average concentrations are expected to exceed 75 ug/m . Multiple
sites will be located in New York, Los Angeles ,and Chicago. The chemiluminescence
method has been adapted to measure both NO and NO . In this technique NO
undergoes a gas phase reaction with ozone to produce NO in an excited State.
The intensity of light emission from the reactor is proportional to the NO
content of the sample. To measure concentrations of NO , a catalyst is first
used to convert NO to NO.
The NO thus produced is then reacted with Ozone.
Effects of findings on Control Strategies for Stationary Sources in Implementation
Plans
With the exception of Los Angeles and Chicago no control strategy for
existing stationary sources over and above that which had been fully implemented
by the end of 1972 is required to maintain the NO standard in the absence of
further growth. Where EPA has proposed emission limitations on sources of
combustion for implementation plan purposes, these proposed regulations will be
withdrawn.
108.
-------
Where states and localities have acted to promulgate regulations imposing
emission limitations on these sources, EPA will work with the agencies in
rolling back these actions as desired. There are some cases however in
which states and localities have developed emissions standards prior to any
EPA requirement for the purpose of meeting a local air quality standard or
to reduce the potential for the production of photochemical oxidants. In
these cases the matter remains one for local consideration. New York City
is one of the latter. In the case of nitric acid plants or any other sources
of N02 as a primary emission, existing control regulations will be required
to stand.
In terms of control strategy to achieve the national air quality standard,
it is estimated that Chicago will be able to show achievement by 1975 through
the following measures:
1. Reductions attributable to the Federal new motor vehicle program
through 1974.
2. Reductions obtained through the conversion in fuel from the use of
coal to gas and oil.
3. Reduction in the formation of NO in the atmosphere through the
control of hydrocarbons.
A. Reductions obtained from the transportation control plan developed by
Chicago to acheive CO and hydrocarbon standards.
3
Los Angeles was estimated to have an annual mean of 180 ug/m . The State
of California already requires control of NO from motor vehicles. EPA
believes that transportation control programs developed to meet the CO and
photochemical oxidant standard existing stationary source controls and the
vehicle emission control program will permit the achievement of the NO- standard
in this AQCR.
109.
-------
Control Strategy to Provide for Maintenance and Growth
The requirement that implementation plans provide for the maintenance
of standards into the far future imposes a need to develop control strategies
which will adequately handle additional emissions associated with future
growth. It is a very difficult problem to determine where the most cost
effective reductions can be made whether in stationary or mobile sources,
in high level or low level emissions.
An examination of the estimated emissions of NO in the U.S. in Tables 4,
X
5 shows that more than 99% of the total emissions are derived from combustion
sources - more than 90% comes from the combustion of fossil fuels.
In the relatively high temperature conditions accompanying the combustion
of fossil fuels and wastes, only a comparatively small amount of N0» is formed.
In fact, the rate of oxidation of NO to NO decreases with increasing temperature.
It is estimated that at temperatures of 2000 F the NO formed is only 0.5% of
the NO . At the low concentrations of NO normally found in the atmosphere
X
3
(l.Oppm or 1200 ug/m ) the subsequent oxidation of NO to N0_ by direct reaction
with oxygen in ambient air occurs very slowly. The rate of conversion is pro-
portional to the square of the NO concentration leading to rapidly decreasing
conversion with dilution. The primary mechanism for the formation of NO
therefore, is the photochemical process involving reactive hydrocarbons.
As a result of this, the diurnal variations in NO, N09 and ozone (or total
oxidants) show the following pattern (Figure 2): An NO peak at approximately
7 am; an NO peak at approximately 10 am and an ozone/ oxidant peak around 12
noon to 3 pm. The pattern is similar for both normal and stagnation conditions.
Just as in the case of hydrocarbons and photochemical oxidants, it is the 6-9
am emissions of NO which are most important in the production of the N00 maximum.
There are also indications that the conversion from NO to NO is a function of
emission rate, meteorologically determined diffusion rate, and the reaction
110.
-------
rate, which is determined by chemistry and meteorology. Because of the lag
time tin peak N02 -production, total NO emissions are perhaps more important
than individual point emissions from specific sources in determining peak
N02 production. In built up areas like New York City, where daytime wind
speeds near the surface are considerably lower than those at rooftop, the
rate of dilution of emissions from motor vehicles is less rapid than the rate
of dilution of emissions from rooftop and elevated sources. Since 40% of
N0x emissions are estimated to be derived from mobile sources, these sources
must remain a primary source for emission reduction in future control strategy.
This would hold true even in New York City where it has been estimated that
93,000 out of 332,000 tons of NO (or 28%) are produced by mobile sources.
However, based on air quality measurements of SO (for which a reliable
inventory is available) it appears likely that emissions of NO from stationary
X
sources in New York City are considerable overestimated and that mobile sources
play a greater role in NO/NO emissions in New York City than currently estimated
X
EPA considers the existing schedule up through 1974 for control of NO
emissions from mobile sources through the Federal new motor vehicle program
essential for the maintenance of standards. Additional limitations are also
needed for 1976 and later model year cars, but not as stringent limitations
as the 90% reduction required by the Clean Air Act. The actual needs
have not yet been determined.
To assist in maintaining air quality standards, EPA will retain standards
of performance for NO emissions from new and modified stationary sources.
X
In addition, to this, future performance standards for sources of significant
NO emissions will include NO emissions limitations that reflect best available
x x
control technology.
111.
-------
Effect of New Data on the National Standard
These findings concerning the reliability of the Jacob-Hochheiser method
are important in respect to the determination of the national primary standard
for NO in which the Chattanooga School Children Study played an important role.
The results obtained in that study with a modification of the reference method
were compared with measurements made by the U.S. Army using continuous monitors
based on the Saltzman method. These monitors, running simultaneously within
0.4 miles of the air monitoring sites established in the EPA study area, showed
values of 0.099ppm and 0.087 ppm in the period November 1968 through April 1969
as compared with a value of 0.109 ppm at the EPA study site. In addition, the
individual measurements were adjusted on the basis of the true collection efficiency
curve. The new estimates for the critical Chattanooga site showed an increase of
11.3% in exposure. Moreover, the U.S. Army collected NO data in the area which
demonstrated that the NO/NO- ratio observed at Chattanooga would have had little
effect on the apparent collection efficiency of N0_.
These independent measurements of NO exposures at Chattanooga, the findings
on the efficiency of the Jacobs-Hochheiser method and the low level of possible NO
interference, clearly indicate that the application of the Jacobs-Hochheiser method
in the circumstances of the Chattanooga study did not significantly effect the choice
of the national air quality standards. Further health effects studies are being
conducted at CHESS sites using the Saltzman method and are expected to further
substantiate the current standard. EPA is seeking, within the next 18 to 24 months,
to establish the basis for a short term standard for NO .
^.iforcement Standards - The major emphasis by EPA over the next few years will be
in the area of standards and implementation plan enforcement. Federal Standards
for NO have been promulgated for new motor vehicles only. Achievement of these
X
emission standards will be determined through extensive Federally observed
certification procedures of vehicle engines. There is still some question whether
112.
-------
all vehicles will, in the future, be required to achieve the standards in-
dividually or whether averaging will be allowed. Certification requirements
include durability tests for engines to maintain performance over 50,000 miles.
Compliance by in-use vehicles will be monitored through inspection/maintenance
programs.
Only in Chicago and Los Angeles will credit have to be taken for transportation
control measures to demonstrate achievement of air quality standards. Enforcement
of State developed transportation control plans will be the primary responsibility
of the state. Where EPA is bringing direct actions to enforce an approved State
plan or Federally promulgated section of a State plan, the Agency will proceed
under Section 113 (a) (1) of the Act. Under such circumstances EPA is bound to
give a violator 30 days notice before issuing an order or bringing civil or
criminal action.
EPA could also seek criminal penalties against an individual violator under
Section 113 (c) (a) of the Act. For cases in which there is wide-spread violations
or cases in which a State has failed to enforce its strategies, it is EPA's current
policy to seek compliance by the State under Section 113 (a) (2) of the Act. Under
this section EPA can issue orders and bring civil actions against the Director of
state or city agencies charged with implementation of specific strategies.
For cases on which EPA might have to promulgate a plan the Agency will operate
on the theory that the State is an emission source in that a highway or other
publicly owned property on which motor vehicles operate is an emission source. On
this principle, EPA would enforce a transportation control regulation through the
issuance of an order to the responsible State or City agency rather than directly
against a motorist. In these cases no 30 day notice is required.
Enforcement of stationary source control requirements for existing sources
required to apply reasonably available control technology will be the primary
responsibility of the states. Sources covered by new source performance standards
will be regulated by EPA until the states obtain delegation of this responsibility
113.
-------
and in cases where states have emission limitation regulations will occur at
start-up of new sources and on an ad hoc basis in conjunction with states
thereafter.
114.
-------
TABLE I
FEDERAL LIGHT DUTY floioR VEHICLE FJUSSFON STANDARDS
(GM/MI)
MODEL YEAR HYDROCARBONS CARBON NJNOXIDE OXIDES OF NITROGEN
PRIOR TO CONTROLS (8,7) (37) (3,5)
3968-1969
(8,7)
50-100 CID -
101-140 CID -
OVER-140 CID -
CALIFORNIA
fkrioNAL
10,7
8,5
6,8
/I.I
4,1
3,0
3,0
3,0
0,9
1.5
Ml
0,41
(87)
66
57
43
28
28
28
28,0
28,0
9,0
15,0
3,4
3,4
1970
1971 4,1 28 (3,5)
1972
1973 3,0 28,0 3,1
1974 3,0 28,0 3,1
1975 CALIFORNIA 0,9 9,0 2,0
3,1
1976 0,41 3,4 0,4
1977 0,41 3,4 0,4
ALL STANDARDS FOR !!C AND CO ARE EXPRESSED IN TERMS OF THE 1975 FEDERAL
CVS TEST PROCEDURE, THE 1975 FEDERAL NOX STANDARD MAS BEEN PRESCRIBED PURSUATIT
TO SECTION 202 (A) OF THE ACT,
115.
-------
TABLE 2
[)QX EMISSION LIMITATIONS REFLECTING AVAILABLE TECHNOLOGY
SOURCE TYPE EXISTING SOURCE - APPENDIX B NEW SOURCES - f\!SPS
FUEL COMBUSTION
GASEOUS 0,2 (170)A 0,20 (170)
LIQUID 0,3 (230)A 0,30 (230)
SOLID 0,70 (525)
NITRIC ACID PLANTS 5,5 LB PER TON OF 100% ACID PRODUCED 3,0 LB PER TON
WOO) (220)
( ) = PPM ON A DRY BASIS AT 3% OICYGE1
A - REPRESENTS ABOUT A 50% REDUCTION IN EMISSIONS FROM UNCONTROLLED FUEL BIOJIKG
EQUIPMENT,
116.
-------
TABLE 3
Effect of M on
Method for N02
uq/m3
NfJ2
100
102
105
122
189
244
248
215
311
316
318
356
the Refer
NO
0
63
127
627
0
1205
1279
1242
0
111
332
1060
•ence
Ratio
HO/N02
0.0
0.6
1.2
5.1
0.0
4.9
5.2
5.8
0.0
0.4
1.1
3.0
Expected
M02
recovered recovered
39
39
38
36
29
24
23
26
20
20
20
18
Apparent
N02
38
38
52
57
29
45
55
50
17
30
33
44
117.
-------
TABLE
NO.
x
U, S, 19G3
(TONS/YR)
SOURCE
MOBILE FUEL COMBUSTION 'IOTOR VEHICLES
OTHER MDBILE SOURCES
STATIONARY FUEL COMBUSTION
SOLID HASTE
COAL WASTE
AGRICULTURAL
INDUSTRIAL PROCESSES
EMISSIONS
7,200,00:1
L010,000
9,980,nrn
556,071
190,000
L 533,000
200,01)
20,^9,000
118.
-------
TABLE
T./YR.
SOURCE PART. SO? CO 1 1C
TRANSPORTATION 1,0 1,0 77,5 14,7 11,2
FUEL COMBUSTION
(STATIONARY) 6,5 26,3 1,0 0,3 10,2
INDUSTRIAL PRXESSES 13,3 5,5 11,1 5,6 0,2
SOLID I/ASTE DISPOSAL 0,7 0,1 3,8 1,0 0,2
MISCELLANEOUS 5,2 0,1 6,5 5,0 0,2
1971 TOTAL 27,2 33,0 93,9 26,6 22,3
1970 TOTAL 25,4 33,9 147,2 34,7 22,7
1969 TOTAL 27.3 33,6 154,0 35,2 22.5
THE MAJOR CHANGES FROM 1959 - 1971 ARE DUE TO REVISED EMISSION FACTORS
FOR TRANSPORTATION, HC AND CO EMISSIONS AND A CONSIDERABLE REDUCTION IN
SOLID WASTE DISPOSED BY INCINERATION,
119.
-------
c
1-1
n>
.80h
.70
.60
o
M
.50
r1
PI
|sj M
? £ .40
.30
.20
.10
PPM .016 .047
I L_
30 90
.080
1
150
.11
i
_J
210
.14
l
270
.18
1
330
.21
1
390
.24
1
450
.27
1
510
.30
1
570
.34
1
630
.37
1
690
.40
1
750
Concentration of NO Sampled
-------
Average Daily 1-Hour Concentrations of Selected
Pollutants in Los Angeles, July 19, 1965
NO
2400 0300 0600
0900 1200 1500
TIME OF DAY
1800 2100 2400
Figure 2
121.
------- |