The Kendal May to September 2015 monthly reports show specific instances, 37 in total, of violations of licenced PM 10 limit values.

Transcription

1 Assessment of Eskom Coal Fired Power Stations for compliance with their 1 April 2015 AELs over the period 1 April 2015 to 31 March 2016; and ranking of their pollutant and CO 2 emission intensities Eugene Cairncross 19 May

2 Contents 1 Summary Recommendations Introduction The reporting requirements of the MES Methods used to analyse the available data to estimate monthly average pollutant s10 6 Results Compliance with AEL values PM 10, SO 2, NOx and CO 2 intensities Addendum 1: Estimated monthly average stack s compared with AEL values A. Camden monthly average stack s B. Duhva monthly average stack s C. Grootvlei monthly average stack s D. Hendrina monthly average s E. Kendal monthly average s F. Komati monthly average s G. Kriel monthly average s H. Lethabo monthly average s I. Majuba monthly average s J. Matimba monthly average s K. Matla monthly average s L. Tutuka monthly average s Addendum 2: Eskom s current average emission intensities Addendum 3: Analysis of Kendal monthly reports for PM 10 exceedences: May to September Summary: Analysis of Kendal monthly reports Graphs of PM 10 daily during non-compliant months Unit 1: Non-compliant: 1 May and 18 July Unit Unit Unit 4: Unit Unit

3 List of Tables Table 1: Comparison of monthly average pollutant s with AEL values, and flagging of power stations that may have been non-compliant with the values during 2015/ Table 2: Eskom CFPSs pollutant intensities, sorted by PM 10 intensities, based on 2015/16 Annual Report data, highest to lowest Table 3: Eskom CFPSs pollutant intensities, sorted by SO 2 intensities, based on 2015/16 Annual Report data, highest to lowest Table 4: Eskom CFPSs pollutant intensities, sorted by NOx intensities, based on 2015/16 Annual Report data, highest to lowest Table 5: Eskom CFPSs pollutant intensities sorted by CO 2 intensities based on 2015/16 Annual Report data, highest to lowest Table 6: Camden monthly average s compared with AEL 24h average values Table 7: Duvha monthly average s compared with AEL 24h average values Table 8: Grootvlei monthly average s compared with AEL 24h average values Table 9: Hendrina monthly average stack s compared with AEL 24h average values Table 10: Kendal monthly average stack s compared with AEL 24h average values Table 11: Komati monthly average stack s compared with AEL 24h average values Table 12: Kriel monthly average stack s compared with AEL 24h average values Table 13: Lethabo monthly average stack s compared with AEL 24h average values Table 14: Majuba monthly average stack s compared with AEL 24h average values Table 15: Matimba monthly average stack s compared with AEL 24h average values. 19 Table 16: Matla monthly average stack s compared with AEL 24h average values Table 17: Tutuka monthly average stack s compared with AEL 24h average values

4 1 Summary Based on their estimated monthly average PM 10 s, seven (Duvha, Grootvlei, Kendal, Komati, Kriel, Matla, Lethabo) of 12 power stations were very likely or likely non-compliant with their PM 10 licensed values in 2015/16. In addition, Komati and Kriel are likely to have experienced noncompliance with their NOx s during this period. The Kendal May to September 2015 monthly reports show specific instances, 37 in total, of violations of licenced PM 10 values. (Addendum 3) The Duhva audit report notes that the AEL PM 10 of 100 mg/nm 3 was regularly exceeded during 2014/15. The Hendrina audit report noted one instance of non-compliance the south stack had not been calibrated at the time of the audit. The Komati audit report noted four non-compliances, including one instance of the PM 10 value being exceeded. The PM 10 emission intensities, expressed as tons of PM 10 emitted per, for Tutuka, Kriel, Grootvlei, Matla, Lethabo, Kendal and Komati all exceed 0.4 tons per compared with Eskom s overall average PM 10 emission intensity (relative particulate ) of 0.36, and compared to its planned 2015/16 relative particulate of The ratio of the average of the highest three PM 10 emitters to that of the lowest three PM 10 emitters is 10:1. Power station CO 2 emission intensities range from 925 to 1262 tons of CO 2 emitted per, or equivalently from to kg of CO 2 emitted per kwh sent out. Eskom s (2015/16) overall average CO 2 emission intensity is 1079 tons CO 2 per. The ratio of the average of the highest two CO 2 emitters to that of the lowest two is about 1.3:1. The SO 2 emission intensities range from 6.92 to 13.0 tons of SO 2 emitted per. The two most SO 2 polluting plants, Matimba and Hendrina, emit about 70% more SO 2 per than that of the least SO 2 polluting plants (ratio 1.7:1). The NOx emission intensities range from 2.34 to 6.17 tons of NOx emitted per. The ratio of the average of the highest two to that of the lowest two NOx emitters is about 2.6:1. The intensity of a process or facility, expressed as the mass of a pollutant emitted per unit of power sent out, is not regulated, but it is a useful measure of the relative impact of different plants or technologies with respect to climate change (GHG ) and air pollution (PM 10, SO 2, NOx ). The prioritisation of plants for decommissioning on the basis of their relative GHG (mainly CO 2 ) intensity may use GHG intensity as the main consideration. If near field (area close to the plant) and regional (area within several hundred kilometres of the plant) air pollution is the immediate consideration, then emission reduction through the installation of improved emission control technology, as envisaged in the new plant standards of the MES, is possible. A short to medium term (5 to 10 years) decommissioning schedule of Eskom s coal plants should include climate change impact, air pollution and other relevant factors in the decision matrix. 4

5 2 Recommendations Obtain reports of daily average s of PM 10, SO 2 and NOx, for each unit on each plant, for 2015/16, for Duvha, Grootvlei, Kendal, Komati, Kriel, Matla and Lethabo, in graphical format similar to that used in the Kendal monthly reports. (That is, the graphical format should be similar to the example below but with the x-axis extended to cover the full year.) Ascertain if the relevant local authorities have a management system in place for monitoring compliance with the MES values, and for responding to reports of non-compliance. Recommend to the NAQO that the reporting requirements of the MES (GN ) be elaborated on, clarified and standardised through the drafting of a guideline, including templates of the Annual Report and Monthly Reports, as apparently contemplated in paragraph 17 of the MES, in the form specified by the National Air Quality Officer. While paragraph 19 (of the MES) refers to and requires Annual Reports to be in the format for the internet-based National Atmospheric Emissions Inventory System, the MES and the NAEIS have two distinctly different air quality management objectives and functions, and the reporting requirements are significantly different. The responsibility of implementing the MES, including compliance monitoring and undertaking remedial action in the event of non-compliance, lies with the licensing authority (usually the local authority). The reporting functions and requirements of the MES should not be subordinated to those of the NAEIS. The reporting formats should therefore be, as far as possible, reconcilable with each other rather than identical. Amend all AELs to clarify reporting requirements, and to correct omissions (in current AELs) with respect to reporting requirements. The annual reports should include (but not be ed to) 5

6 graphs of daily average pollutant s similar in format to the Kendal monthly graphs but with the x-axis extended to cover the full year. 6

7 3 Introduction Following various PAIA requests to Eskom, the following documents are available for perusal and analysis: Annual Reports (ARs) for 1 April 2015 to 31 March 2016 for Eskom s 14 coal fired power stations (CFPSs) (Arnot, Camden, Grootvlei, Hendriena, Kendal, Komati, Kriel, Lethabo, Majuba, Mathimba, Matla, Medupi, Lethabo and Tutuka), the first cycle of annual reports submitted in terms of AELs issued 31 March The Arnot annual report (AR) could not be analysed as data are presented in graphical form only; the Medupi annual report covered the post-commissioning months only and included insufficient data for analysis; the Tutuka 2015/16 report covered the months April to September 2015 only. Kendal submitted May to September 2015 monthly reports, in addition to its annual report. AELs issued 31 March 2015, the first cycle of AELs issued under the MES. Compliance Audit Reports for 4 CFPSs (Camden, Duhva, Hendrina and Komati) PSs; 13 CFPSs (Grootvlei is the exception) submitted flue gas monitoring instrument calibration (for gas measurement) and/ or correlation (for PM 10 measurement) reports, but a review and analysis of these documents is not included in this assessment. This assessment focuses on compliance with Section 7.2 of each CFPS AEL, the point source maximum emission rates (for PM 10, SO 2 and NOx) under normal working conditions. [Note that although both the MES and the AELs describe the prescribed values (for PM 10, SO 2 and NOx) as maximum release rates and/ or maximum emission rates, the values are more properly described as maximum s. An emission rate or a release rate should have units of mass flow per unit time, for example, tonnes per day or kg per hour.] The pollutant maximum values ( values) are expressed in the MES and the AELs as daily average pollutant s, measured continuously in the discharge duct of each boiler unit, with units of mg/nm 3 under normal (that is, standardised) conditions of 10% O 2, 273 Kelvin and kpa pressure. Thus, in a typical power station consisting of 6 units, the from each unit are monitored continuously in the duct between the boiler unit and the stack. The continuously-measured pollutant s, averaged over a one day period, must be compliant with the relevant value. In some cases the licensed values may be different for different units of the power station. Assessment of day-to-day compliance with the values prescribed in Section 7.2 of the AELs should be a matter of comparing the continuously-measured s, averaged over 24 hours (midnight to midnight), with the prescribed maximum values. The Annual Reports unfortunately do 7

8 not include tables or graphs showing measured daily average s, for each pollutant, for the reporting period. A fairly elaborate procedure had therefore to be developed to estimate monthly average stack values using the annual report monthly emission rate (tonnes of pollutant per month) and aggregate monthly power sent out data, which provided that basis for indirectly assessing compliance with the prescribed values. 4 The reporting requirements of the MES Sections 17 and 18 of the MES refer specifically to reporting requirements, although these clauses should be read in the context (particularly the definitions) of the regulations as a whole. CFPSs are required to install continuous emission monitoring systems. The wording in the MES relevant to reporting requirements for CFPSs is as follows. Reporting Requirements (17) Notwithstanding the compliance time frames established in terms of paragraphs (8); (9); and (10), the Atmospheric Emission License holder shall submit an emission report in the form specified by the National Air Quality Officer to the Licensing Authority - (a) within one (1) year of the date of publication of this Notice; and (b) annually thereafter unless otherwise prescribed in the Atmospheric Emission License. (18) The report contemplated in paragraph (17) shall include- (a) The name, description and license reference number of the plant as reflected in the Atmospheric Emission License. (c) Where continuous emission monitoring is required for a listed activity, the report contemplated in paragraph (17) shall further include - (i) results of the spot measurements or correlation tests carried out to verify the accuracy of the continuous emission measurements; (ii) the most recent correlation tests; and (iii) the availability of the system as contemplated in (15)(b) in terms of the number of full hours per annum that valid results were obtained. (d) Following the compliance time frames established in terms of paragraphs (8); (9); and (10), an 8

9 explanation of all instances where minimum emission standards were exceeded and remediation measures and associated implementation plans aimed at ensuring that the accidences do not reoccur. The AEL template included in the AEL Manual for Licencing Authorities (DEA, 2010) includes the following reporting requirements for the Annual Report. Annual reporting The licence holder must complete and submit to the licensing authority an annual report. The report must include information for the year under review (i.e. annual year end of the company). The report must be submitted to the licensing authority not later than 60 (sixty) days after the end of each reporting period. The annual report must include, amongst others, the following items: (a) Pollutant trend; (b) Compliance audit report(s); (c) Major upgrades projects (i.e. abatement equipment or process equipment); and (d) Greenhouse gas. It would appear that, in the absence of more detailed reporting guidelines, licensing authorities have mainly simply repeated the above broad outlines of the required content of Annual Reports contained in the AEL template, without explicitly requiring the inclusion of a report of day-by-day compliance with the licensed values. The MES, in addition to the above list of items that should be included in the Annual Report, requires the annual report to include an explanation of all instances where minimum emission standards were exceeded and remediation measures and associated implementation plans aimed at ensuring that the accidences do not re-occur (s18 (d)). In this respect, the AELs are not compliant with the MES reporting requirements. The Annual Reports therefore currently report monthly trends in, generally interpreted as monthly aggregate (as tons of pollutant emitted per month) for each of the regulated pollutants, over the reporting year. The monthly trends cannot be used to determine whether or not the licensed values have been complied with or not, because monthly total emission rates cannot be disaggregated into daily average values and individual unit emission rates, and mass emission rates (tons per month or tons per day) cannot be interpreted as s without information on the corresponding gas volumetric flow rates. (The values are expressed as mass per unit volume, calculated at standard conditions, of gas flow. That is, as mg (of pollutant) per Nm 3 of gas.) 9

10 5 Methods used to analyse the available data to estimate monthly average pollutant s The flow rate of gas, measured under standard conditions, discharged from the stack is approximately proportional to the power sent out by each power station. Estimates of the ratio of stack flow rate to power sent out are calculated, for each power station, using 2012/13 data submitted by Eskom in support of their MES postponement applications. The Annual Reports include monthly total for each pollutant (PM 10, SO 2, NOx, CO 2 and in some cases N 2 O) as well as monthly GWh sent out (). The monthly values combined with the stack flow rate to power sent out (as ) ratio derived from the 2012/2013 data are used to estimate flow rates and thus corresponding monthly average pollutant s. The uncertainty in the estimated monthly average s has not been estimated, but it is probably +- 10%. The estimated monthly average pollutant s are compared with the 31 March 2015 AEL values. If monthly average pollutant s exceed the AEL values this indicates that at least several daily average values for the corresponding month were non-compliant with the AEL. Since the monthly average values are the average of 29 to 31 (the number of days per month) days values, even if the monthly average is less than the AEL value, this does not exclude the possibility that one or more daily average values is non-compliant with the AEL daily average values. The likelihood of one or more daily average values exceeding the value increases, the closer the monthly average value is to the AEL value, but it is also dependent on the day to day variability in the data. This day-to-day variability is different for each of the pollutants. It is greatest for PM 10 because the control of PM is primarily dependent on the efficient function of unit emission control devices electrostatic precipitators (ESPs) or fabric filters (FFs). If the PM emission control device malfunctions, PM emission rates may increase substantially over a short period of time. The day-to-day variation in SO 2 flue gas s are primarily due to variation in the sulphur content of coal, a factor that generally varies relatively slowly and within s, unless the source of coal and/ or the sulphur content of the coal change significantly. The variation in NOx s is due to both variations in coal composition ( fuel NOx ) and in combustion conditions ( thermal NOx ), the latter remaining relatively constant. Based on estimates of day-to-day variability, different criteria are used for each of the pollutants to assess possible non-compliance. In Table 1, monthly average PM 10 s that are more than 60% of the AEL are highlighted (in red font) as indicating that one or more non-compliances with AEL values may have or are likely to have occurred during corresponding months. (Monthly estimates 10

11 are presented in Addendum 1.) For both SO 2 and NOx, monthly average values greater than 75% of the AEL s value may indicate possible or likely non-compliance. The four available Compliance Audits, which are of variable quality, were perused for specific citations of non-compliance. While emission intensities, pollutant per unit of power sent out, are not regulated, they provide a measure of the relative climate change impact or relative dirtiness of each power station. The relative emission intensities may be used as additional criteria for prioritising the phasing out of individual power stations. Power stations are ranked by CO 2, PM 10, SO 2 and NOx intensities in Tables 2 to 5, and the calculated values are compared with Eskom s average values (Addendum 2) published in its 2015/16 Integrated Annual Report. 6 Results Compliance with AEL values Table 1: Comparison of monthly average pollutant s with AEL values, and flagging of power stations that may have been non-compliant with the values during 2015/16 31 March 2015 AELs Period April 2015 to March 2016 Rated PM 10 24h SO 2 24h NOx 24h Highest monthly average PM 10 Highest monthly average SO 2 Power station capacity (MW) value value value values values Arnot Insufficient data available Highest monthly average NOx values Camden less than 40 less than 1600 less than 850 Duhva less than 1600 less than 800 Grootvlei less than 1900 less than 750 Hendrina less than 30 less than 2100 less than 750 Kendal* * less than 1500 less than 710 Komati less than 1600 less than 1000 Kriel less than 1550 less than 1350 Lethabo less than 1000 less than 600 Majuba less than 20 less than 1050 less than 800 Matimba less than 20 less than 1600 less than 300 Matla / less than 1000 less than 760 Medupi** Insufficient data available Tutuka** / Insufficient data available Notes to Table 1: a) ** indicates a partial monthly data set. 11

12 b) The PM 10 values in red font indicate that one or more non-compliances with AEL values may have or are likely to have occurred during the corresponding month (monthly estimates are presented in Addendum 1). Thus Duvha, Grootvlei, Kendal, Komati, Kriel, Matla, Lethabo were very likely or likely non-compliant with their PM 10 licensed values in 2015/16. c) The NOx values in purple font indicate Komati and Kriel power stations may have or are likely to have experienced one or more non-compliances with AEL values during the corresponding month (monthly estimates are presented in Addendum 1). (The Kendal May to September 2015 monthly reports show specific instances, 37 in total, of violations of licensed PM 10 values. Details are presented in a separate report.) PM 10, SO 2, NOx and CO 2 intensities The intensity of a process or facility, expressed as the mass of a pollutant emitted per unit of power sent out, is not regulated, but it is a useful measure of the relative impact of different plants or technologies with respect to climate change (GHG ) or air pollution (PM 10, SO2, NOx). The prioritisation of plants for decommissioning on the basis of their relative GHG (mainly CO2) intensity may use GHG intensity as the main consideration. If near field (area close to the plant) and regional (area within several hundred kilometres of the plant) air pollution is the immediate consideration, then emission reduction through the installation of improved emission control technology, as envisaged in the new plant standards of the MES, is possible. A short to medium term (5 to 10 years) decommissioning schedule of Eskom s coal plants should consider both factors, climate change and air pollution impacts, as well as other relevant factors. Tables 2 to 5 present the intensities of Eskom s coal plants, ranked for each emission. Table 2: Eskom CFPSs pollutant intensities, sorted by PM 10 intensities, based on 2015/16 Annual Report data, highest to lowest Coal fired power station Rated capacity (MW) PM 10 SO 2 NOx CO 2 Tutuka Kriel Grootvlei Matla Lethabo Kendal Komati Duhva Camden Hendrina Matimba Majuba The monthly values used to calculate the annual values are listed in Addendum 1. 12

13 The ratio of the most polluting to the least polluting, by PM 10 intensity, of 1.053/ = 18:1 should be noted. The ratio of the highest three PM 10 emitters to the lowest three emitters is 10:1. Table 3: Eskom CFPSs pollutant intensities, sorted by SO 2 intensities, based on 2015/16 Annual Report data, highest to lowest Coal fired power station Rated capacity (MW) PM 10 SO 2 NOx CO 2 Matimba Hendrina Camden Grootvlei Komati Kriel Lethabo Duhva Kendal Majuba Matla In the absence of emission control technology, the differences in SO 2 emission intensities are mainly due to differences in coal sulphur content. The two most SO 2 polluting plants, Matimba and Hendrina, emit about 70% more SO 2 per than that of the least SO 2 polluting plants. Table 4: Eskom CFPSs pollutant intensities, sorted by NOx intensities, based on 2015/16 Annual Report data, highest to lowest Coal fired power station Rated capacity (MW) PM 10 SO 2 NOx CO 2 Kriel Komati Majuba Matla Camden Lethabo Hendrina Duhva Grootvlei Kendal Matimba

14 In the absence of post-combustion emission control systems, the differences in NOx emission intensities may be due to differences in burner control technologies and /or differences in the nitrogen content coal. The ratio of the highest two to the lowest two NOx emitters is about 2.6:1. Table 5: Eskom CFPSs pollutant intensities sorted by CO 2 intensities based on 2015/16 Annual Report data, highest to lowest Coal fired power station Rated capacity (MW) PM 10 SO 2 NOx CO 2 Grootvlei Komati Hendrina Camden Majuba Kriel Duhva Kendal Matla Tutuka Lethabo Matimba Differences in CO 2 emission intensities may be due to differences in coal quality and overall plant energy efficiency. The ratio of the highest two CO 2 emitters to the lowest two is about 1.3:1. 14

15 Addendum 1: Estimated monthly average stack s compared with AEL values A. Camden monthly average stack s Table 6: Camden monthly average s compared with AEL 24h average values Monthly average stack s Camden daily values Month PM 10 stack SO 2 stack NOx stack PM 10 AEL SO 2 AEL NOx AEL April May June July August September October November December January February March None of the Camden monthly average PM 10 values exceeds 60% of the daily average value, implying that the plant was very likely compliant with the licensed value throughout the period. B. Duhva monthly average stack s Table 7: Duvha monthly average s compared with AEL 24h average values Monthly average stack s Duhva daily values Month PM 10 stack SO 2 stack NOx stack PM 10 AEL SO 2 AEL NOx AEL April May June July August September October November December January February March

16 C. Grootvlei monthly average stack s Table 8: Grootvlei monthly average s compared with AEL 24h average values Month Monthly average stack s PM 10 stack SO 2 stack NOx stack Grootvlei daily values PM 10 AEL SO 2 AEL NOx AEL April May June July August September October November December January February March The two PM 10 monthly average values, highlighted in red, are more than 60% of the AEL value, implying that one or more of the daily average values during the corresponding month may have exceeded the licensed value (i.e. may have been non-compliant during these months). D. Hendrina monthly average s Table 9: Hendrina monthly average stack s compared with AEL 24h average values Month Monthly average stack s PM 10 stack SO 2 stack NOx stack Hendrina daily values PM 10 AEL SO 2 AEL NOx AEL April May June July August September October November December January February March None of the Hendrina monthly average PM 10 values exceeds 60% of the daily average value, implying that the plant was very likely compliant with the licensed value throughout the period. 16

17 E. Kendal monthly average s Table 10: Kendal monthly average stack s compared with AEL 24h average values Monthly average stack s Kendal daily values Month PM 10 stack SO 2 stack NOx stack PM 10 AEL SO 2 AEL NOx AEL April May June July August September October November December January February March F. Komati monthly average s Table 11: Komati monthly average stack s compared with AEL 24h average values Komati monthly average stack s Komati daily values Month PM 10 stack SO 2 stack NOx stack PM 10 AEL SO 2 AEL NOx AEL April May June July August September October November December January February March The April 2015 monthly average PM 10 substantially exceeds the daily average value, implying that PM 10 were non-compliant for many or most days during this month. In June 2015, January and March 2016, the monthly average PM 10 exceeded 60% of the daily average 17

18 value, implying that one or more of the daily average values during these months may have exceeded the licensed value (may have been non-compliant during this month). G. Kriel monthly average s Table 12: Kriel monthly average stack s compared with AEL 24h average values Month Monthly average stack s PM 10 stack SO 2 stack NOx stack Kriel daily values PM 10 AEL SO 2 AEL NOx AEL April May June July August September October November December January February March All Kriel monthly average PM 10 values are above 60% (5 above 100%) of the AEL daily value. H. Lethabo monthly average s Table 13: Lethabo monthly average stack s compared with AEL 24h average values Month Monthly average stack s Lethabo daily values PM 10 stack SO 2 stack NOx stack PM 10 AEL SO 2 AEL NOx AEL April May June July August September October November December January February March

19 I. Majuba monthly average s Table 14: Majuba monthly average stack s compared with AEL 24h average values Monthly average stack s Majuba daily values Month PM 10 stack SO 2 stack NOx stack PM 10 AEL SO 2 AEL NOx AEL April May June July August September October November December January February March J. Matimba monthly average s Table 15: Matimba monthly average stack s compared with AEL 24h average values Monthly average stack s Matimba daily values Month PM 10 stack SO 2 stack NOx stack PM 10 AEL SO 2 AEL NOx AEL April May June July August September October November December January February March

20 K. Matla monthly average s Table 16: Matla monthly average stack s compared with AEL 24h average values Monthly average stack s Matla daily values Month PM 10 stack SO 2 stack NOx stack PM 10 AEL SO 2 AEL NOx AEL April / May / June / July / August / September / October / November / December / January / February / March / L. Tutuka monthly average s Table 17: Tutuka monthly average stack s compared with AEL 24h average values Monthly average stack s Tutuka AEL values Month PM 10 stack SO 2 stack NOx stack PM 10 AEL NOx AEL SO 2 AEL Apr May June July Aug Sept Oct 2015 No data No data No data Nov 2015 No data No data No data Dec 2015 No data No data No data Jan 2016 No data No data No data Feb 2016 No data No data No data March 2016 No data No data No data

21 Addendum 2: Eskom s current average emission intensities These notes are intended to give context. Eskom 2015/16 Integrated Report, p123: Eskom s overall average particulate intensity (relative emission) for 2015/16 was 0.36 kg/mwh sent out, or 0.36 tons of PM 10 per. In 2015/16 Eskom s CFPSs sent out GWh. Therefore the intensities for 2015/16 for CO 2, SO 2 and NOx were respectively 1079 ( /199888), 8.50 ( /199888) and 4.47 (893000/199888). Eskom submitted the following planned reduction in relative PM target in its MES postponement applications. Estimating the 2015/16 value from the graph, it is approximately 0.33 kg/mwhso or

22 Addendum 3: Analysis of Kendal monthly reports for PM10 exceedences: May to September 2015 Summary: Analysis of Kendal monthly reports PM10: The value of 100 mg/nm3 applies to all units. Unit 1: Non-compliant: 1 May and 18 July. 2 days Unit 2: Compliant throughout the period. Unit 3: Offline. Unit 4: Non-compliant: 23 May, 4, 12 and 13 June, 25 July, 6 and 14 August, 2 and 31 September. 9 days Unit 5: Non-compliant: 23 May; 4, 5, 12, 19, 20, 23, 24, 25, 26 June; 2, 14, 15, 16, 17, 18, 22, 23 July; (offline 24 Aug to 12 Sept); 13 August. 19 days Unit 6: Non-compliant 1, 23, 24 29, 30 May; 23, 24 June; 2, 17, 31 July; 3, 5, 6, 17, 24, 29, 31 September 17 days Total number (all units) of days non-compliant with the 100 mg/nm3 value: 47. SO 2 (3500 mg/nm3 ); NOx (1100 mg/ Nm3 ): Compliant with both s, for all months. Graphs of PM10daily during non-compliant months Unit 1: Non-compliant: 1 May and 18 July. 22

23 Unit 2: Compliant throughout the period. Unit 3: Offline. Unit 4: Non-compliant: 23 May, 4, 12 and 13 June, 25 July, 6 and 14 August, 2 and 31 September. 23

24 24

25 Unit 5: Non-compliant: 23 May; 4, 5, 12, 19, 20, 23, 24, 25, 26 June; 2, 14, 15, 16, 17, 18, 22, 23 July; (offline 24 Aug to 12 Sept); 13 August 25

26 26

27 Unit 6: Non-compliant 1, 23, 24 29, 30 May; 23, 24 June; 2, 17, 31 July; 3, 5, 6, 17, 24, 29, 31 September 27

28 28

29 29