Professional Documents
Culture Documents
To cite this article: C.E. Baukal & P.B. Eleazer (1998) Quantifying NOx for Industrial
Combustion Processes, Journal of the Air & Waste Management Association, 48:1, 52-58, DOI:
10.1080/10473289.1998.10463664
Article views: 80
ABSTRACT INTRODUCTION
The objectives of this paper are to (1) identify the prob- There is a wide range of industrial heating and melting
lems with many of the units that are used to report processes that generate NOX (nitric oxide, NO, and nitro-
and regulate NOX, (2) show how to properly correct gen dioxide, NO2) emissions. These processes include, for
NOX measurements for oxygen-enhanced combustion, example, copper and lead smelting, steel and aluminum
and (3) recommend a preferred type of NO X unit. The production, and glass and mineral wool manufacturing.
current variety of NO X units make comparisons diffi- The fuel of choice in the United States for these processes
Downloaded by [78.100.77.180] at 02:25 30 November 2015
cult and can cause considerable confusion. NOX may is natural gas, which typically contains a small amount
be measured on a wet or dry basis, but it is commonly of molecular N2 and no organically bound N2.1 Therefore,
reported on a dry basis. The reported NOX may differ prompt and thermal NOX are the important mechanisms,
from the actual measurements, which may be con- while fuel NOX is not a concern.
verted to a specific O2 basis level. Nearly all of the mea- There are at two types of problems with the units
sured NOX from industrial combustion systems is in that are commonly used to report and regulate NOX in
the form of NO, which is converted to NO2 in the at- those industrial combustion processes. The first concerns
mosphere. However, when given on a mass basis, the the variety of units that have been used, which makes
measured NO is commonly reported as NO2 for regu- comparisons difficult. The second concerns how to make
latory purposes, but may be reported as NO, NO2, or the proper corrections to a given unit, when oxygen or
simply NOX in technical papers. Some existing regula- oxygen-enriched air is used as the oxidizer. One purpose
tions may penalize combustion technologies with of this paper is to alert both the end users and the regula-
higher efficiencies and lower flue gas volumes, such tory agencies to the potential for confusion. Another pur-
as oxygen-enhanced combustion. Confusion may oc- pose is to suggest more uniform and consistent units for
cur when applying some of the “conventional” NOX reporting NOX.
units to oxygen-enhanced processes. A better unit is Most combustion processes use air as the oxidizer. Air
the mass of NOX generated per unit of production, contains approximately 21% O2 and 79% N2 by volume. It
which also incorporates the overall process efficiency has been recognized that replacing air with pure oxygen
into the emissions. That unit does not penalize more can significantly increase the performance of a combus-
efficient processes that may generate more NOX on a tion system.2 For example, pure oxygen has been used as
volume basis, but less NOX on a production basis. the oxidizer to enhance the performance of metal melt-
ing,3 glass making,4 and waste incineration.5 There are also
benefits to enriching air with oxygen, so that the O2 con-
centration is higher than 21%. Some of the common ben-
IMPLICATIONS efits of using oxygen-enhanced combustion include higher
This paper recommends changing the units of measure- productivity and thermal efficiency with lower exhaust gas
ment used to report and regulate NO X emissions. The volume and pollutant emissions. 2 Quantifying those emis-
plethora of units that are used (e.g., ppm, lb/106 Btu) makes sions will be considered here. It should be noted that NOX
it difficult to compare NOX among industries and technolo-
is the pollutant of specific interest in this paper; however,
gies. Some existing NOX regulations penalize technolo-
gies with higher thermal efficiencies that generate less NOX the discussion would apply to other pollutants emitted from
but seem to produce more NOX when reported in an inap- combustion processes, such as CO.
propriate type of unit. A more appropriate unit is based on
the mass of NO X generated per unit of industrial produc- NOx UNITS
tion (e.g., lb NOX /ton glass). This type of unit encourages
One source of confusion regarding NOX is the lack of a con-
technologies that increase production efficiency, which
sistent unit of measurement for industrial heating processes,
reduces NOX by reducing fuel consumption.
which makes it difficult to compare measurements and
52 Journal of the Air & Waste Management Association Volume 48 January 1998
Baukal and Eleazer
regulations. This confusion can be illustrated by studying should be reported as either “ppmvd” for a dry basis, or
the proceedings of a recent combustion conference. Thir- as “ppmvw” for a wet basis. Unfortunately, it is usually
teen different units were used in the various papers that left to the reader to assume that “ppm” actually means
were presented. These included uncorrected ppm,6 ppm “ppmvd.” For example, in an air-CH4fuel system, 100
at 0% O2,7 ppm at 3% O2,8 ppm at 6% O2,9 ppm at 7% ppmvd is equivalent to 83 ppmvw, when both are cor-
O2,10 ppm at 11% O2,11 ppm at 15% O2,12 g/kg CH4,13 mass rected to 3% O2. For other oxidizer compositions, there
NOX /mass dry fuel (where the fuels tested included hu- may be an even larger difference between the dry and
mus, rice straw, waste paper, wood and a wood blend),14 wet bases. Figure 1 shows that there is a substantial differ-
lb/ton glass,15 lb/MMBtu (106 Btu),16 lb/hr,17 and mg/Nm3 ence between those bases for intermediate oxidizer com-
at 3% O2.18 In most cases, it was not stated, but it was positions that occur in oxygen-enhanced combustion
assumed that “ppm” was by volume and not by mass, processes. Note that when the fuel is pure CH4 and the
since this is the normal convention. It was assumed that oxidizer is pure O2, no NOX would be generated since all
the volume NOX measurements were on a dry basis. It the N2 has been removed from the combustion process
also was not stated, but assumed, that total NOX, not just (assuming that there is no air leakage into the system).
NO, was reported. In some of the papers, multiple units
were used. For example, Wang et al. used both ppm at 3% Oxygen Correction
O2 and ppm at 6% O2.9 NOX measurements are commonly normalized to a specific
Downloaded by [78.100.77.180] at 02:25 30 November 2015
Volume 48 January 1998 Journal of the Air & Waste Management Association 53
Baukal and Eleazer
Downloaded by [78.100.77.180] at 02:25 30 November 2015
Figure 2. Corrected NOx (ppmv) vs. O2 basis (%) for a raw NOx Figure 3. NO2 (lb/106MM Btu) vs. NO2 (ppmvd at 3% O2), where the
measurement of 100 ppm at 2% O2, for an air-fuel system. oxidizer is air and the fuel is CH4.
54 Journal of the Air & Waste Management Association Volume 48 January 1998
Baukal and Eleazer
The above example shows that the preheat air case process efficiency in the emissions, it is difficult to
produces more NO X based on a gross fuel basis (lb NOx/ compare NOX values between different industries. It
106 Btu) but actually emits 12% less NOX into the en- also makes it difficult to compare NO X reduction tech-
vironment (lb NOX /unit of production) because of the nologies used in the same process, but at different lo-
increased fuel efficiency. If the NOX regulation was 1.0 cations. One location may be poorly run, while an-
lb NOx/106 Btu, the preheat air case would fail to meet other may be very efficient. Differences in NOX may
that regulation, even though it would actually emit be attributable to either the combustion system, or to
less NOX into the atmosphere. Simple efforts to increase the production process, or to both. This type of NOX
the fuel efficiency of a combustion process (e.g., mini- unit may also increase the number of regulations, since
mizing air leakage into a furnace) could produce more each would have to be specifically written for a par-
NOX (e.g., due to higher gas temperatures caused by ticular process. This could be very complicated for in-
less air dilution) per unit of fuel but may actually pro- dustries where the material being processed varies
duce less net NOX, due to the increased fuel efficiency. widely. Consider a municipal waste incinerator pro-
A preferred unit would be the mass of NOX gener- cessing trash. One day the trash may be very dry, so
ated per net unit of energy that goes into the process the processing rates would be higher. Another day the
(e.g., lb NOX /10 6 Btu-net). That type of unit would trash may be very wet, so the processing rates would
incorporate the efficiency of the process. That unit is be lower. For that type of a process, it may not make
Downloaded by [78.100.77.180] at 02:25 30 November 2015
more commonly applied as the mass of NO X gener- sense to normalize NOX regulations to a production
ated per unit of production, since the net energy into basis, or the measurements may have to be averaged
the process is directly related to the production rate. over a longer period of time to average out the daily
An example of this type of unit is the regulation for variations in the material feed composition.
nitric acid plants in 40 CFR 60.72,19 where the NOX
limit is 1.5 kg NOX (expressed as NO 2) per metric ton NOX IN OXYGEN-ENHANCED COMBUSTION
of 100% nitric acid. Another example is the regula- Another major source of confusion regarding units of
tion in the South Coast Air Quality Management Dis- measurement pertains to converting NOX measurements
trict in southern California for glass manufacturers, for an oxygen-enhanced combustion process. One prob-
which is in lb NO X per ton of glass produced. 24 lem with current regulations that are written on a volume
A NOX unit based on production has other ramifi- basis is how to correct the measurements for oxidizers other
cations. It is affected by the rest of the process. For than air. The correction in eq 1 assumes that any excess O2
example, consider the interruptions in a production comes from air. This may or may not be the case with oxy-
operation. Two assumptions will be made in this ex- gen-enhanced combustion. If the excess O2 measured in
ample. The first is that the interruption is not caused the flue gases came from an oxygen-enriched oxidizer, then
by the combustion system. The second is that the com- eq 1 should be modified as follows:
bustion system remains on during the interruption,
which is assumed to be of a relatively short duration. O2 − O2 BASIS
Possible causes for the interruption might be a short- ppm CORR = ppm MEAS OXID
O2 (2)
age of feed material, failure of the material handling OXID − O2 MEAS
system, or failure of some other equipment in the pro-
duction process. In this scenario, NOX based on a unit where ppmMEAS = measured pollutant concentration in flue
of production would increase as a result of downtime. gases (ppmvd), ppmCORR = pollutant concentration cor-
The combustion system would still be emitting NOX rected to a standard O2 basis (ppmvd), O2OXID = O2 concen-
at the same rate, even though nothing is being pro- tration in the oxidizer (vol. %), O2MEAS = measured O2 con-
cessed during the interruption. On the other hand, a centration in the flue gases (vol. %, dry basis), and O2BASIS
production system with very little downtime but high- = standard O2 basis (vol. %, dry basis).
NOX burners could actually produce less NOX than a Air leakage into an oxygen-enhanced combustion
poorly run production system with low-NOX burners. process further complicates this conversion. For ex-
Therefore, a NO X unit based on production gives in- ample, if the oxidizer supplied to the burners is pure
centive to have both low-NOX burners and a well-run O2 and there is significant air infiltration into the pro-
production system. A NOX unit like lb/10 6 Btu-gross cess, then the effective oxidizer is a combination of the
does not incorporate the efficiency of the heating pro- O2 to the burners and the infiltrated air. In that case, the
cess or the efficiency of the production process. O2 concentration in the oxidizer (O2OXID) would have to be
Bluestein notes that there are potential drawbacks calculated based on the flue gas volume flow rate or on the
to a NOX unit based on production.25 By including the amount of N2 in the flue gases.
Volume 48 January 1998 Journal of the Air & Waste Management Association 55
Baukal and Eleazer
conditions) per unit volume of CH4, and available heat (assuming a Figure 5. NOx (ppmvd at 0% O2) comparison when the oxidizer is air,
2000 °F exhaust gas temperature) vs. oxidizer composition, for a and pure O2 when the fuel is CH4.
stoichiometric CH4 flame.
From Figure 4, it may be seen that the exhaust volume Net Versus Gross Basis
per unit of fuel input is dramatically reduced as the O2 con- As shown in Figure 4, O2/CH4fuel systems are often sig-
centration in the oxidizer increases. It may also be seen nificantly more efficient than air-CH4fuel systems. Less
that the available heat increases as well. Available heat is fuel is required for a given unit of production. There-
defined as the gross heating value of the fuel less the en- fore, a better NOX unit would be the mass of NO X gen-
ergy carried out of the process by the hot exhaust gases. erated per unit mass of production.
Reducing the flue gas volume and increasing the thermal This concept can be illustrated with a unit like lb/
efficiency are two common reasons for using oxygen-en- 6
10 Btu-net. This unit is derived by combining the NOX/
hanced combustion.2 Both result from the removal of N2 106 Btu-gross curve in Figure 6 and the available heat
from the oxidizer and both have a dramatic impact on the curve in Figure 4. It is assumed that the base case is
relevance of some of the NOX units. air/CH 4fuel combustion. Then, as the O 2 concentra-
tion in the oxidizer increases, the thermal efficiency
Volume Versus Mass Basis increases. The gross and net Btu curves are shown in
Because the flue gas volume may be reduced by >90% when Figure 7. The net Btu curve is similar to a production
replacing air with pure oxygen, comparing air/fuel NOX to basis since production rates are directly proportional
O2/fuel NOX on a ppmv basis is not appropriate. For ex- to the amount of energy that goes into the product. It
ample, 200 ppmvd NOX from an O2/fuel system is actually is assumed that the heat losses from the system and
less NOX by mass than 100 ppmvd NOX from an air/fuel the heat transfer to the product would not vary sig-
system because of the vast differences in flue gas volumes. nificantly using different oxidizers. This figure shows
Figure 5 shows a comparison of NOX in ppmvd corrected that it can be very deceiving to compare NOX based
to 0% O2 for systems using air and pure O2. The ppmvd of only on the gross firing rate when the oxidizer com-
NOX using O2 as the oxidizer is equivalent to 8.52 times the positions are different. Therefore, the recommended
ppmvd of NOX using air as the oxidizer, when the fuel is unit is the mass of NOX per unit of production since
CH4. This is strictly due to the change in the volume of the the system efficiency is also included.
dry products of combustion as a function of the oxidizer.
As can be seen, care must be exercised when comparing CONCLUSIONS
NOX on a volume basis for systems using different oxidiz- This paper discussed the wide variety of methods that
ers. Hence, if it is assumed that air/fuel and O2/fuel systems are used for quantifying NOX. For a number of reasons,
have equal efficiencies (which they usually do not), it makes reporting NOX on a volume basis has many drawbacks.
more sense to compare the NOX on a mass basis. Figure 6 A better unit for NOX is on a mass basis normalized to
shows NOX in both mass and volume units, as a function the gross firing rate. This unit can be further improved
of the oxidizer composition. by normalizing the NOX to a unit production basis.
56 Journal of the Air & Waste Management Association Volume 48 January 1998
Baukal and Eleazer
Volume 48 January 1998 Journal of the Air & Waste Management Association 57
Baukal and Eleazer
16. Beale, F. In Proceedings of AFRC/JFRC Pacific Rim International Confer- 24. Alternative Control Techniques Document — NOX Emissions from Glass
ence on Environmental Control of Combustion Processes, 1994. Manufacturing; U.S. Environmental Protection Agency, Research Tri-
17. Bell, A.W.; Haythornthwaite, S.M.; Sanders, C.F. In Proceedings of AFRC/ angle Park, NC, 1994; EPA-453/R-94-037.
JFRC Pacific Rim International Conference on Environmental Control of 25. Bluestein, J. NOX Controls for Gas-Fired Industrial Boilers and Combus-
Combustion Processes, 1994. tion Equipment: A Survey of Current Practices; Gas Research Institute:
18. Sayre, A.N. In Proceedings of AFRC/JFRC Pacific Rim International Con- Chicago, IL, 1992; GRI-92/0374.
ference on Environmental Control of Combustion Processes, 1994.
19. U.S. Government, Protection of Environment; Government Institutes,
Inc.: Rockville, MD, 1994; Code of Federal Regulations Title 40, Part
60.
About the Authors
20. ANSI/ASME, Part 10: Flue and Exhaust Gas Analyses; American Society
of Mechanical Engineers: New York, NY, 1981; Performance Test Code Dr. Baukal is a combustion research engineer and Mr. Eleazer
PTC 19.10. is the manager of the Combustion Center-of-Excellence, both
21. Nitrogen Oxide Control for Stationary Combustion Sources; U.S. Environ-
mental Protection Agency. U.S. Government Printing Office: Wash- in the Global Applications Development department of Air
ington, DC, 1986; EPA/625/5-86/020. Products and Chemicals, Inc., 7201 Hamilton Blvd., Allen-
22. IHEA. Combustion Technology Manual, Fourth ed.; Industrial Heating town, PA 18195. Correspondence may be directed to Dr.
Equipment Association: Arlington, VA, 1988.
23. Control Techniques for Nitrogen Oxide Emissions from Stationary Sources, Baukal (e-mail: baukalce@apci.com).
Revised Second Edition; U.S. Environmental Protection Agency, Re-
search Triangle Park, NC, 1983; EPA-450/3-83-002.
Downloaded by [78.100.77.180] at 02:25 30 November 2015
58 Journal of the Air & Waste Management Association Volume 48 January 1998