Quality checks needed!
Dataflow
Ideas for quality checks on the ePRTR data
These checks should run before submitting the data under the LRTAP convention:
- Total lps emssions < national total
- All releases to compartment air for given year inculded in output.csv
- unit mistakes
- not empty: IDs, stack height, GNFR, coordinates
- at least one pollutant given for each lps
- coordinates inside German borders
- combination of coordinates unique
- Sanitize plant names
- TSP > PM10 >PM2.5 >BC
- PM > HM
- Verify script result with a manual filtering in the Excel files also provided by thru.de
- Use old 2017 database from thru.de to recreate the 2017 submission and check for differences
We also need to address the review findings
Sector | Aspect | Summary | Details |
---|---|---|---|
LPS | General | Improve consistency with the latest ePRTR reporting. | Finding 2020: The TERT notes that for the year 2015, emissions are reported for 36 facilities in the E-PRTR database (v18) which could not be found in the LPS submission using the National IDs provided. The TERT notes that the sum of these missing facilities is below the threshold of significance. During the review Germany indicated that deviations could result from data updates or post-reporting as well as installations that are reported in the E-PRTR but had not reported emissions of pollutants relevant for reporting under CLRTAP. Germany indicated that both the methodology and the timeliness of the data will be improved in the next LPS reporting. The TERT recommends that Germany includes these improvements as well as some transparent explanation of the possible reasons for differences in its IIR. |
LPS | QA/QC | Improve coordinates given, check for collisions | Finding 2020: The TERT notes that in 2015, the same pair of longitude and latitude coordinates was assigned to more than one differently named LPS in 10 case(s). The TERT notes that LPS reporting guidance specifies that differently named LPS cannot have the same latitude and longitude. During the review Germany highlighted that the LPS dataset is a direct re-publication of the E-PRTR and that its 'E-PRTR/ PRTR'-geodata were used. The TERT notes that Germany's IIR includes a brief description of the methods, data assumptions and methods used to incorporate the E-PRTR data into the LPS. The TERT recommends that Germany provides a more transparent description of the use of its E-PRTR in the NECD inventory including an explanation of the assumptions on geo-referencing of the LPS sites. |
LPS | QA/QC | Make sure each point source reported has unique key build from attributes | Finding 2020: The TERT notes that in 2015, 53 combination(s) of LPS name, GNFR code and stack height class were reported two or more times. The correct reporting format requires that each combination of LPS, GNFR and height class must appear only once. In its response to the review question, Germany indicated that it only delivers a copy of PRTR-data for the LPS. The TERT recommends that Germany ensures that it provides unique references to the LPS name in order to differentiate LPS in its future submissions. |
LPS | Completeness | Add missing pollutants PAHs, PCBs, PM2.5 | Finding 2020: The TERT notes that for the year 2015, no LPS emissions are reported for the following pollutant(s): PAHs, PCBs, PM2.5. During the review Germany indicated that it relies on the national E-PRTR database for LPS reporting. Since the pollutants above are not covered by the E-PRTR, they do therefore not appear in the submitted template. During the review Germany confirms the completeness of gridded data in including all emissions sources in the national totals. The TERT therefore does not consider this an issue of under- or over-estimation. The TERT recommends that Germany considers improving its distribution of emissions for these pollutants by estimating and or attributing point source estimates to facilities that are likely to be producing emissions. For example, Germany could investigate whether any of the LPS reporting PM10 can reasonably be used to estimate PM2.5, for example using the particle size distribution from the EMEP/EEA Guidebook. |
LPS | Accuracy | Check emission data for facility "Heyne & Penke Verpackungen GmbH" | Finding 2020: For the year 2015, total emissions reported for the LPSs with the corresponding E-PRTR National ID of 03-05-05050581570 (Heyne & Penke Verpackungen GmbH) were different to those reported for that facility for the same year in the E-PRTR database (v18). In response to a question raised during the review Germany stated that reason for deviation can be a data update, which seems a correction of a scale error (now it's one-tenth), and that the data actuality will be improved in the next LPS reporting. The TERT recommends that Germany correct the LPS emissions for this plant in the next LPS reporting. |
LPS | Transparency | Reallocate livestock emissions from GNFR L_AgriOther to K_AgriLivestock | Finding 2020: For GNFR code K_AgriLivestock in the LPS submission year 2015, the TERT noted that there is a lack of transparency. Germany reports NH3 emissions for NFR code(s) 3B3 Manure Management - Swine, 3B4gi Manure Management - Laying hens, 3B4gii Manure Management - Broilers, 3B4giii Manure Management - Turkeys in the national inventory but not for GNFR code K_AgriLivestock in the LPS submission. This does not relate to an over- or under-estimate of emissions. In response to a question raised during the NECD review 2019, Germany explained that emission from large pigs and poultry farms where allocated in GNFR L_AgriOther. Germany agree that these emissions should be mapped into GNFR category K_AgriLivestock. Germany confirm to provide a reallocation to GNFR L_AgriOther in future submission. The TERT recommends that Germany reallocate the NH3 emission from GNFR code L_AgriOther to GNFR code K_AgriLivestock, for the next submission 2021. |