Professional Documents
Culture Documents
Article Bryant CO2 Emissions 2015
Article Bryant CO2 Emissions 2015
A natural gas burner has been used as a precise and accurate source for generating large quantities of carbon dioxide (CO2)
to evaluate emissions measurements at near-industrial scale. Two methods for determining carbon dioxide emissions from
stationary sources are considered here: predicting emissions based on fuel consumption measurements—predicted emissions
measurements, and direct measurement of emissions quantities in the flue gas—direct emissions measurements. Uncertainty for
the predicted emissions measurement was estimated at less than 1%. Uncertainty estimates for the direct emissions measurement
of carbon dioxide were on the order of ±4%. The relative difference between the direct emissions measurements and the predicted
emissions measurements was within the range of the measurement uncertainty, therefore demonstrating good agreement. The
study demonstrates how independent methods are used to validate source emissions measurements, while also demonstrating how
a fire research facility can be used as a precision test-bed to evaluate and improve carbon dioxide emissions measurements from
stationary sources.
Implications: Fossil-fuel-consuming stationary sources such as electric power plants and industrial facilities account for more
than half of the CO2 emissions in the United States. Therefore, accurate emissions measurements from these sources are critical
for evaluating efforts to reduce greenhouse gas emissions. This study demonstrates how a surrogate for a stationary source, a fire
research facility, can be used to evaluate the accuracy of measurements of CO2 emissions.
863
Journal of the Air & Waste Management Association, 65(7):863–870, 2015. This article not subject to U.S. copyright law. ISSN: 1096-2247 print
DOI: 10.1080/10962247.2015.1031294 Submitted November 17, 2014; final version submitted February 3, 2015; accepted March 10, 2015.
Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/uawm.
864 Bryant et al. / Journal of the Air & Waste Management Association 65 (2015) 863–870
Nondimensional
Relative Standard Sensitivity Percent
Measurement Component, xi Value Uncertainty, u(xi)/xi Coefficient, si Contribution, %
Therefore, the equation represents the CO2 emissions com- than ±0.017 for the predicted emissions measurements.
puted from fuel consumption and composition—the predicted Improvements in the flow meter calibration and temperature
emissions measurements. Table 1 demonstrates nominal values and pressure measurements reduced the relative expanded
for the input measurements of eq 1. The burner was operated uncertainty to nominally ±0.010 or less. The largest compo-
with fires of 2 MW or less for this investigation due to the need nents of uncertainty were the fuel carbon content and the
to conduct the velocity traversing experiments, which required volume flow rate measurement. Exhaust stream measurements
the burner to run for extended periods. Using lower heat of CO2, CO (carbon monoxide), and O2 (oxygen) were per-
release rates limited the radiant heat exposure to the surround- formed to verify the burner conversion efficiency and plume
ing environment. When operating at full capacity, the natural capture assumptions. Measureable amounts of CO were not
gas fire generates approximately 0.5 kg/sec of CO2. The burner detected in the flue gas; therefore, complete carbon conversion
can be used to simulate steady-state and transient combustion was assumed (ηb = 1) and the detection limit of the measure-
processes from a moderate size stationary source such as an ment was used to estimate the uncertainty. A similar methodol-
industrial plant. ogy was used in a previous study of compartment fires to
An uncertainty analysis was performed to estimate the com- estimate combustion efficiency (Bundy et al., 2007). A detailed
bined uncertainty of the predicted emissions measurements of discussion of the uncertainty analysis for the predicted emis-
CO2. Assuming that the input measurements for eq 1 were sions measurements can be found in a previous publication
mutually independent, the following equation was applied to (Borthwick and Bundy, 2011). Only data for experiments
estimate the combined relative uncertainty: with complete capture of the fire plume by the canopy exhaust
vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi hood were included in this study.
u N
uc ðyÞ u X uðxi Þ 2
¼t s2i (2) Direct emissions measurements
y i¼1
xi
Large canopy exhaust hoods were used to capture the
The standard uncertainty, u(xi), for each input measurement, xi, combustion products from the burner. The canopy hoods direct
used to compute the predicted CO2 emissions (y ¼ m_ CO2 ;p ) is the flow into the exhaust ducts that run along the roof of the
listed in Table 1. The nondimensional sensitivity coefficient, facility and were instrumented for measuring gas temperature,
given as, velocity, and volume fraction of selected combustion products.
The maximum exhaust flow capacity is approximately
@y xi 50 kg/sec of air and the operating pressure in the duct was
si ¼ (3)
@xi y slightly below atmospheric.
Mean flow velocity in the exhaust duct was determined
is also listed in the table to reflect the weighting applied to the from a collection of point velocity measurements conducted
standard uncertainty of each component. Estimates of the rela- by traversing two S probes, equipped with thermocouples,
tive expanded uncertainty (twice the relative standard uncer- across a section of the exhaust duct.1 The exhaust duct,
tainty for a 95% confidence interval) were nominally better shown in Figure 3, runs horizontally along the roof of the
1
Note: Certain commercial entities, equipment, or materials are identified in this document in order to describe an experimental procedure or concept adequately.
Such identification does not imply recommendation or endorsement by the National Institute of Standards and Technology, nor does it imply that the entities,
materials or equipment are necessarily the best available for the purpose.
866 Bryant et al. / Journal of the Air & Waste Management Association 65 (2015) 863–870
Figure 4. Schematic of the measurement section of the 1.5-m exhaust duct. Bulk flow was computed from a series of point velocity measurements made by
traversing S probes across the duct. Gas samples flowed continuously from the sampling tee to the gas analyzers.
Bryant et al. / Journal of the Air & Waste Management Association 65 (2015) 863–870 867
Table 2. Example of an uncertainty budget for direct emissions measurements of CO2 in the exhaust duct of the LFRL
Nondimensional Percent
Relative standard sensitivity contribution,
Measurement component, xi Value uncertainty, u(xi)/xi coefficient, si %
Exhaust gas mean flow velocity, Vexh (m/sec) 20.91 0.0056 1.0 9.9
Exhaust duct diameter, d (m) 1.504 0.0079 2.0 77.6
Exhaust gas mean density, ρexh (kg/m3) 1.047 0.0034 1.0 3.6
CO2 net volume fraction—dry basis, XCO2 ;net;dry (m3/m3) 0.001819 0.0053 1.0 8.9
Exhaust gas H2O volume fraction, XH2 O;exh (m3/m3) 0.007947 0.0031 0.05 0
Exhaust gas molecular weight, Mexh (kg/kmol) 28.7734 0.0001 −1.0 0
CO2 molecular weight, MCO2 (kg/kmol) 44.0095 0.0000 1.0 0
Direct CO2 emissions, m_ CO2 ;d (g/sec) 107.3 0.0179 (0.0358) Standard (Expanded) Uncertainty
Note: The CO2 was generated using a 2-MW natural gas fire.
Conclusions
Fossil-fuel-burning stationary sources have accounted for
over half of all CO2 emissions in the United States and there-
fore play a significant role in the accuracy of greenhouse gas
reporting. Using a fire research facility as a near-industrial-
scale surrogate for a stationary source, two primary methods
for determining CO2 emissions, predicted emissions from fuel
Figure 9. Data quality check of exhaust duct gas sampling measurements
consumption measurements and direct stack measurements,
(direct) and fuel supply gas composition measurements (predicted). Fuel have been compared.
Factors computed from measurements are compared with the default value A natural gas fire, issuing from a well-characterized burner
for natural gas. and gas supply system, served as a precision source of CO2.
870 Bryant et al. / Journal of the Air & Waste Management Association 65 (2015) 863–870
Predicted CO2 emissions, computed from the fuel consumption Bryant, R., O. Sanni, E. Moore, M. Bundy, and A. Johnson. 2014. An
measurements, were demonstrated with an expanded uncer- uncertainty analysis of mean flow velocity measurements used to quantify
emissions from stationary sources. J. Air Waste Manage. Assoc. 64:679–
tainty of ±1% or less. Direct measurements of CO2 emissions
689. doi:10.1080/10962247.2014.881437
from the exhaust duct were demonstrated with an expanded Bryant, R.A., T.J. Ohlemiller, E.L. Johnsson, A. Hamins, B.S. Grove, W.F.
uncertainty of ±4% or less. The relative difference between Guthrie, A. Maranghides, and G.W. Mulholland. 2003. The NIST 3
pairs of predicted and direct emissions measurements was Megawatt Quantitative Heat Release Rate Facility. NIST Special Publication
generally less than the estimated measurement uncertainty, 1007. Gaithersburg, MD: National Institute of Standards and Technology.
therefore demonstrating good agreement. Fuel Factor values Bundy, M., A. Hamins, E.L. Johnsson, K.C. Sung, K. Gwon, and D.B. Lenhert.
computed from the gas composition measurements at the fuel 2007. Measurements of Heat and Combustion Products in Reduced-Scale
Ventilation-Limited Compartment Fires. NIST Technical Note 1483.
supply and in the exhaust duct compared well with the default
Gaithersburg, MD: National Institute of Standards and Technology.
value for natural gas, hence confirming the quality of the Quick, J.C. 2014. Carbon dioxide emission tallies for 210 U.S. coal-fired power
measurements. plants: A comparison of two accounting methods. J. Air Waste Manage.
This study demonstrates how the principle of the conserva- Assoc. 64:73–79. doi:10.1080/10962247.2013.833146
tion of mass and independent measurement methods are used Shigehara, R.T., R.M. Neulicht, W.S. Smith, and J.W. Peeler. 1978. Summary
to provide a cross-validation of CO2 emissions at a stationary of F Factor Methods for Determining Emissions from Combustion Sources.
source. The study also introduces and demonstrates the concept EPA-450/2-78-042a. Washington, DC: U.S. Environmental Protection
Agency.
of a precision test-bed for the purpose of evaluating and U.S. Environmental Protection Agency (EPA). 2000. Determination of Sulfur
improving greenhouse gas emissions measurements from sta- Dioxide Removal Efficiency and Particulate Matter, Sulfur Dioxide, and
tionary sources. Nitrogen Oxide Emission Rates. EPA Method 19. Washington, DC: U.S.
Environmental Protection Agency.
U.S. Environmental Protection Agency (EPA). 2007. Determination of Stack
Gas Velocity and Volumetric Flow Rate with Two-Dimensional Probes. EPA
Acknowledgment Method 2G. Washington, DC: U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency (EPA). 2008. Direct Emissions from
The authors gratefully acknowledge the technical and engi- Stationary Combustion Sources. EPA 430-K-08-003. Washington, DC: U.S.
neering support provided by Marco Fernandez, Laurean Environmental Protection Agency.
DeLauter, Doris Rinehart, and Anthony Chakalis, data acquisi- U.S. Environmental Protection Agency (EPA). 2011. Gas Analysis for the
Determination of Emission Rate Correction Factor or Excess Air. EPA
tion support provided by Artur Chernovsky, and data analysis
Method 3B. Washington, DC: U.S. Environmental Protection Agency.
support provided by R. Paul Borthwick. We are also grateful U.S. Environmental Protection Agency (EPA). 2013. Inventory of U.S.
for the technical guidance provided by Anthony Hamins and Greenhouse Gas Emissions and Sinks: 1990–2011. EPA 430-R-13-001.
Jiann Yang. Research support by the NIST Office of Special Washington, DC: U.S. Environmental Protection Agency.
Programs—Greenhouse Gas and Climate Science Measure U.S. National Archives and Records Administration. 2014. Code of Federal
ments, James Whetstone Program Manager—is gratefully Regulations: Title 40 Part 75, Continuous Emission Monitoring, Appendix F.
acknowledged.
About the Authors
Rodney Bryant and Matthew Bundy are mechanical engineers at the National
References Institute of Standards and Technology, Fire Research Division, in Gaithersburg,
Ackerman, K.V., and E.T. Sundquist. 2008. Comparison of two US power-plant MD.
carbon dioxide emissions data sets. Environ. Sci. Technol. 42:5688–5693.
doi:10.1021/es800221q Ruowen Zong is an associate professor at the State Key Laboratory of Fire
Borthwick, R., and M. Bundy. Quantification of a precision point source for Science, University of Science and Technology of China, and was a visiting
generating carbon dioxide emissions. Paper presented at EPRI CEM User guest researcher at the National Institute of Standards and Technology, during
Group Conference and Exhibit, Chicago, IL, June 8, 2011. the execution of this study.