the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 08 Feb 2024
Towards a sector-specific CO/CO2 emission ratio: Satellite-based observation of CO release from steel production in Germany
Abstract. Global crude steel production is expected to continue to increase in the coming decades to meet the demands of the growing world population. Currently, the dominant steelmaking technology worldwide is the conventional highly CO2-intensive Blast Furnace – Basic Oxygen Furnace production route (also known as the Linz-Donawitz process) using iron ore as raw material and coke as a reducing agent. As a result, large quantities of special gases that are rich in carbon monoxide (CO) are by-products of the various stages of the steelmaking process. Given the challenges associated with satellite-based estimates of carbon dioxide (CO2) emissions at the scale of emitting installations due to significant background levels, co-emitted CO may serve as a valuable indicator of the carbon footprint of steel plants.
We show that regional CO release from steel production sites can be monitored from space using 5 years of measurements (2018–2022) from the TROPOspheric monitoring instrument (TROPOMI) on board the Sentinel-5 Precursor satellite benefiting from its relatively high spatial resolution and daily global coverage. We analyse all German steel plants with blast furnaces and basic oxygen furnaces and obtain associated CO emissions in the range of 50–400 kt yr-1 per site. A comparison with the respective CO2 emissions on the level of emitting installations available from emissions trading data of the European Union Emissions Trading System yields a linear relationship with a sector-specific CO/CO2 emission ratio for the analysed steelworks of 3.24 % [2.73–3.89; 1σ] suggesting the feasibility of using CO as a proxy for CO2 emissions from comparable steel production sites.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(2115 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2115 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-2709', Anonymous Referee #1, 29 Feb 2024
This manuscript by Schneising et al. focuses on estimating CO emissions from steel production facilities in Germany using the TROPOMI CO product. The CO abundance was retrieved by the WFMD algorithm, which was developed by the same group. CO emissions were estimated by rotating the level 2 data around the point sources to align wind direction (from ERA5) and then averaging a number of cross-sectional fluxes. Long-term CO emissions are estimated by aggregating daily emission CO estimates and then compared with independent CO emissions from Thru.de and correlated to CO2 emissions from EU ETS to get sector-specific CO/CO2 emission ratio. The manuscript is well-written, and the results are novel and clearly presented. My comments are minor and hopefully help further improve this manuscript.
- The second last paragraph of the introduction listed related literatures that used TROPOMI CO observations for industrial emission estimation. An important difference that is not mentioned explicitly is that this study uses the WFMD product, whereas others (Park et al. 2021, Tian et al. 2022, and Wu et al., 2022) used the official S5PCO retrieval. In this paragraph or in section 2, it might be helpful to compare the WFMD and official TROPOMI CO products, e.g., in terms of algorithm design, data coverage, and precision/biases, and highlight if/why the WFMD product is advantageous for achieving the objectives of this work.
- Figures 2-7 appear to be after rotation, so the tick labels on the horizontal/vertical axes should not be latitude (N/S in degrees) and longitude (W/E in degrees). The spatial coordinate should be projected to uniform scales (e.g., in km) before the rotation. Also in line 90, clearly define the “region of interest” for wind averaging.
- The uncertainty calculation for the daily and long-term emission estimates are mentioned very briefly near lines 108 and 127, mostly referring to a previous work estimating methane emissions. I suggest including more details on how the uncertainties are calculated, as those are important results of this work. Specifically, the satellite only measures clear days, would there be any systematic difference in steel-producing facilities' emissions between measurable/unmeasurable days? How about the diurnal pattern of steel production, i.e., how well does TROPOMI data collected in the early afternoon represent the true daily average? Another important missing information is how the uncertainty is defined, 1σ or 2σ. Both are used in the later analysis, and it's good to be consistent.
- Section 3.2, "Air quality assessment" do not seem to fit in the scope of this manuscript or contribute to the conclusion. The analysis also use very roughly estimated numbers (e.g., constant PBL height of 500 m). It is suggested to remove this section.
- Lines 205-208: is it possible to comment on the uncertainties of x-direction, i.e., the CO2 emissions from EU ETS?
- Lines 246-247: it might be more accurate to conclude that the systematic low bias (~20%, can report the exact number) relative to Thru.de exists for the estimates of all sites.
Citation: https://doi.org/10.5194/egusphere-2023-2709-RC1 -
RC2: 'Comment on egusphere-2023-2709', Anonymous Referee #2, 21 Mar 2024
Schneising et al. presents in this work a CO/CO2 ratio for the steel production sector in Germany, with CO emissions estimated from space and CO2 emissions reported at facility and country level. The text is well-written, and the work clearly presented. I recommend publication after the revision of the comments below have been made.
Main comments
While I understand the point of not including information on how the uncertainty is calculated and referring to the method paper by Schneising et al. (2020b), I think it requires some more discussion once the results are shown and, for example, on page 8 line 178, it is stated that “the derived uncertainties are generally realistic”. The uncertainty values are rather constant, but aren’t there different characteristics from the different areas causing different challenges to the emission rate estimation process that would be reflected in the uncertainty estimate?
Are CO emissions assumed to be a continuous source? How does this assumption affect the estimated emissions? Authors mention using wind history of 2 hours to apply filtering criteria, but what if the emissions have been happening for more than 2 hours before the satellite overpass? Also, the plume may have travelled further than the selected box downwind.
The day-to-day variability of emissions is not discussed. This is fine as yearly CO emissions is the value that is used for the ratio, but also because this reason I fail to understand why instead of computing annual average from daily emissions, the method has not been applied to the annual average of CO concentrations. Are those two estimates (annual from daily and annual) consistent with each other?
Air quality assessment: assuming a boundary layer height for all days for all locations needs to be proven valid, especially the wide variety of emission ranges found depending on the location.
The possibility of having other CO sources in the domain should be further discussed. How does it affect the results as you compare facility level CO2 emissions to domain-wide CO emissions?
Given the nature of CO2 emissions data used in this study, does that imply that this study only applies to Germany? Are there other areas where this type of granular and process specific CO2 data is available? This should be further discussed in the conclusions section.
In the introduction or in the conclusion section, I would recommend highlighting the implications and advantage of having a CO/CO2 ratio. Why one would be interested in such a number?
Specific/Technical comments
First paragraph page 2: explain further the time ranges in operation of these satellites to put the reader in context. Also, is there any relevant work with these satellites from the CO steel production perspective?
Page 2, line 39: how important it is in terms of CO emissions of the industry as compared to other sources?
Lines 46-51: long sentence, split into two for readability.
Line 250: “than the other less representative estimates”: please explain which you are referring to.
Citation: https://doi.org/10.5194/egusphere-2023-2709-RC2 -
AC1: 'Final response to referee comments', Oliver Schneising, 30 Apr 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2023-2709/egusphere-2023-2709-AC1-supplement.pdf
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-2709', Anonymous Referee #1, 29 Feb 2024
This manuscript by Schneising et al. focuses on estimating CO emissions from steel production facilities in Germany using the TROPOMI CO product. The CO abundance was retrieved by the WFMD algorithm, which was developed by the same group. CO emissions were estimated by rotating the level 2 data around the point sources to align wind direction (from ERA5) and then averaging a number of cross-sectional fluxes. Long-term CO emissions are estimated by aggregating daily emission CO estimates and then compared with independent CO emissions from Thru.de and correlated to CO2 emissions from EU ETS to get sector-specific CO/CO2 emission ratio. The manuscript is well-written, and the results are novel and clearly presented. My comments are minor and hopefully help further improve this manuscript.
- The second last paragraph of the introduction listed related literatures that used TROPOMI CO observations for industrial emission estimation. An important difference that is not mentioned explicitly is that this study uses the WFMD product, whereas others (Park et al. 2021, Tian et al. 2022, and Wu et al., 2022) used the official S5PCO retrieval. In this paragraph or in section 2, it might be helpful to compare the WFMD and official TROPOMI CO products, e.g., in terms of algorithm design, data coverage, and precision/biases, and highlight if/why the WFMD product is advantageous for achieving the objectives of this work.
- Figures 2-7 appear to be after rotation, so the tick labels on the horizontal/vertical axes should not be latitude (N/S in degrees) and longitude (W/E in degrees). The spatial coordinate should be projected to uniform scales (e.g., in km) before the rotation. Also in line 90, clearly define the “region of interest” for wind averaging.
- The uncertainty calculation for the daily and long-term emission estimates are mentioned very briefly near lines 108 and 127, mostly referring to a previous work estimating methane emissions. I suggest including more details on how the uncertainties are calculated, as those are important results of this work. Specifically, the satellite only measures clear days, would there be any systematic difference in steel-producing facilities' emissions between measurable/unmeasurable days? How about the diurnal pattern of steel production, i.e., how well does TROPOMI data collected in the early afternoon represent the true daily average? Another important missing information is how the uncertainty is defined, 1σ or 2σ. Both are used in the later analysis, and it's good to be consistent.
- Section 3.2, "Air quality assessment" do not seem to fit in the scope of this manuscript or contribute to the conclusion. The analysis also use very roughly estimated numbers (e.g., constant PBL height of 500 m). It is suggested to remove this section.
- Lines 205-208: is it possible to comment on the uncertainties of x-direction, i.e., the CO2 emissions from EU ETS?
- Lines 246-247: it might be more accurate to conclude that the systematic low bias (~20%, can report the exact number) relative to Thru.de exists for the estimates of all sites.
Citation: https://doi.org/10.5194/egusphere-2023-2709-RC1 -
RC2: 'Comment on egusphere-2023-2709', Anonymous Referee #2, 21 Mar 2024
Schneising et al. presents in this work a CO/CO2 ratio for the steel production sector in Germany, with CO emissions estimated from space and CO2 emissions reported at facility and country level. The text is well-written, and the work clearly presented. I recommend publication after the revision of the comments below have been made.
Main comments
While I understand the point of not including information on how the uncertainty is calculated and referring to the method paper by Schneising et al. (2020b), I think it requires some more discussion once the results are shown and, for example, on page 8 line 178, it is stated that “the derived uncertainties are generally realistic”. The uncertainty values are rather constant, but aren’t there different characteristics from the different areas causing different challenges to the emission rate estimation process that would be reflected in the uncertainty estimate?
Are CO emissions assumed to be a continuous source? How does this assumption affect the estimated emissions? Authors mention using wind history of 2 hours to apply filtering criteria, but what if the emissions have been happening for more than 2 hours before the satellite overpass? Also, the plume may have travelled further than the selected box downwind.
The day-to-day variability of emissions is not discussed. This is fine as yearly CO emissions is the value that is used for the ratio, but also because this reason I fail to understand why instead of computing annual average from daily emissions, the method has not been applied to the annual average of CO concentrations. Are those two estimates (annual from daily and annual) consistent with each other?
Air quality assessment: assuming a boundary layer height for all days for all locations needs to be proven valid, especially the wide variety of emission ranges found depending on the location.
The possibility of having other CO sources in the domain should be further discussed. How does it affect the results as you compare facility level CO2 emissions to domain-wide CO emissions?
Given the nature of CO2 emissions data used in this study, does that imply that this study only applies to Germany? Are there other areas where this type of granular and process specific CO2 data is available? This should be further discussed in the conclusions section.
In the introduction or in the conclusion section, I would recommend highlighting the implications and advantage of having a CO/CO2 ratio. Why one would be interested in such a number?
Specific/Technical comments
First paragraph page 2: explain further the time ranges in operation of these satellites to put the reader in context. Also, is there any relevant work with these satellites from the CO steel production perspective?
Page 2, line 39: how important it is in terms of CO emissions of the industry as compared to other sources?
Lines 46-51: long sentence, split into two for readability.
Line 250: “than the other less representative estimates”: please explain which you are referring to.
Citation: https://doi.org/10.5194/egusphere-2023-2709-RC2 -
AC1: 'Final response to referee comments', Oliver Schneising, 30 Apr 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2023-2709/egusphere-2023-2709-AC1-supplement.pdf
Peer review completion
Journal article(s) based on this preprint
Data sets
TROPOMI/WFMD XCH4 and XCO v1.8 O. Schneising https://www.iup.uni-bremen.de/carbon_ghg/products/tropomi_wfmd/
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 464 | 137 | 26 | 627 | 16 | 14 |
- HTML: 464
- PDF: 137
- XML: 26
- Total: 627
- BibTeX: 16
- EndNote: 14
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Oliver Schneising
Michael Buchwitz
Maximilian Reuter
Michael Weimer
Heinrich Bovensmann
John P. Burrows
Hartmut Bösch
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2115 KB) - Metadata XML