SULPHUR DIOXIDE MEASUREMENTS IN SOUTH DURBAN : THE CULMINATION OF 8 YEARS OF MONITORING

The South Durban industrial basin comprises residential areas in close proximity to industry. The South Durban Sulphur Dioxide Management System was formed in 1994 in response to concerns regarding SO2 concentrations in the area. ASteering Committee, consisting of representatives from the national and local authorities, industry and the local community, was formed to monitor ambient SO2 concentrations to thereby facilitate informed decision-making on management strategies for industrial pollution control in the South Durban region. ECOSERV (Pty) Ltd were appointed as project managers in November 1995 to oversee: • the effective storage of the data collected by the system; • day-to-day maintenance of the extensive and varied equipment comprising the system; and • regular reporting and analysis of the data. The monitoring network consisted of an array of continuous SO analysers, an O analyser, a NO analyser and weather stations. Data was relayed regularly via radio telemetry to a database server computer, which was in turn linked to a pollution dispersion model to enable real-time modeling of conditions. During the eight-year lifetime of the network, permanent monitoring stations were established at Wentworth, Southern Works,AECI andAthlone Park, while a mobile trailer containing a SO analyser was also deployed at various localities of concern within the region. Results obtained from the mobile station when at Settlers School suggested the need for further investigation and a semipermanent caravan was established by the Ethekwini (Durban) Municipality in 2000. The monitoring system was accredited by the South African National Accreditation System (SANAS) for SO monitoring in March 1999, which together with the monitoring system run by Ecoserv in Richards Bay, were the first to meet this quality standard in SouthAfrica. While the monitoring system latterly underwent various name changes, eventually being called the Durban Air Quality Forum, the functions and personnel essentially remained the same. The functions of the Steering Committee and monitoring network terminated at the end of 2003 as the Ethekwini Municipality embarked on a new and expanded monitoring network in terms of the South Durban Basin Multi-Point Plan. Results of continuous SO data collected at the Steering Committee monitoring stations from July 1996 to December 2003 are presented. A general improvement has been noted in the air quality in South Durban in terms of SO since the monitoring program was operational, which, while encouraging for air pollution managers, still indicates local areas of concern and the persistent problem of the high loading of SO in South Durban. The Municipality's 2 3 x

• the effective storage of the data collected by the system; • day-to-day maintenance of the extensive and varied equipment comprising the system; and • regular reporting and analysis of the data.
The monitoring network consisted of an array of continuous SO analysers, an O analyser, a NO analyser and weather stations.Data was relayed regularly via radio telemetry to a database server computer, which was in turn linked to a pollution dispersion model to enable real-time modeling of conditions.During the eight-year lifetime of the network, permanent monitoring stations were established at Wentworth, Southern Works, AECI and Athlone Park, while a mobile trailer containing a SO analyser was also deployed at various localities of concern within the region.Results obtained from the mobile station when at Settlers School suggested the need for further investigation and a semipermanent caravan was established by the Ethekwini (Durban) Municipality in 2000.The monitoring system was accredited by the South African National Accreditation System (SANAS) for SO monitoring in March 1999, which together with the monitoring system run by Ecoserv in Richards Bay, were the first to meet this quality standard in SouthAfrica.
While the monitoring system latterly underwent various name changes, eventually being called the Durban Air Quality Forum, the functions and personnel essentially remained the same.The monitoring network (refer Fig. 1) comprised: • Three permanent SO analysers, situated at Wentworth, Southern Works and AECI (and initially a permanent station atAthlone Park, discontinued at the end of 1999).
• A mobile SO monitoring trailer (deployed over various periods since 1999 at different locations of concern in South Durban).
• Settlers School, originally a mobile location from June to October 2000, was established as a semi-permanent monitoring station by the Ethekwini Municipality at the end of October 2000, including SO , CO, NO and particulate monitoring.
• Weather stations were established at Wentworth, Southern Wastewater Treatment Works (abbreviated to Southern Works), AECI, Sapref, Bluff and Island View.Measurements included wind speed, wind direction, temperature, solar radiation, rainfall, pressure and humidity.
• Ozone and NO analysers at the Wentworth station.
• The HAWK pollution dispersion model, situated at the Wentworth base station.
• A radio telemetry network relaying data from the remote stations to a base station at Wentworth.
• A structured query language database, collecting and processing all data, acting as the core of the wireless network and dial-in network facilities for data users.
The meteorological and air quality data were initially recorded as 3-minute averages, however the averaging time was changed to 5 minutes from January 2002 to facilitate The results relate only to the air sampled and care should be taken in extrapolating the results to surrounding areas.In terms of quality assurance all measurements allow for a maximum precision error of 15% and a zero tolerance of plus or minus 10 ppb.The South Durban monitoring system was accredited by the South African National Accreditation System (SANAS) for SO2 monitoring in March 1999, which together with the monitoring system run by Ecoserv in Richards Bay, were the first to meet this quality standard in South Africa.Accreditation is ongoing and is reevaluated annually.
The study area and monitoring stations in relation to industry (black font) and main residential areas (blue font) are indicated in Figure 1.The South Durban basin is separated from the sea by a Holocene dune cordon 80 to 100 m in height.This topography effectively prevents air exchange between the basin and sea at lower levels, particularly during stable atmospheric conditions.Winds blow predominantly from the NNE and SSW to SW (refer Fig. 2).Light nocturnal land breezes blow mainly from the NNW to west.Land breezes are more prevalent during winter, during which surface air temperature inversions are common.
The Wentworth measurement station is located approximately 70 m above msl (the level at which the temperature inversion often occurs during winter).The industrial areas of Jacobs and Mobeni are to the west; while the Engen refinery is to the south and the Sapref refinery and Mondi Plant are to the SSW and south-west of the station, respectively.The Southern Works and Settlers School stations are situated at 15 to 17 m above msl and between the Engen refinery (to the north-east and NNE) and the Sapref refinery and Mondi (south-west to SSW of the stations).The Southern Works station was moved some 800 m northwards during April 1998 and the demarcation on Figure 1 shows the station location from May 1998 to the end of December 2003.The AECI station was located near the Umbogintwini industrial complex, including the Tioxide and AECI industries to the west to WNW.The Athlone Park station was located approximately 100 m above msl on the seaward side of the Holocene dune cordon.The station was discontinued at the end of 1999.

GUIDELINES
The

ANNUAL & INTER-ANNUAL SO2 CONCENTRATIONS
Annual average SO for each of the permanent analyser stations is presented in Figure 3.

DAILY AVERAGE SO2 GUIDELINE EXCEEDANCES
Unfortunately the data averaging time of three minutes prior to 2002 does not allow for comparison of historic data to the new 10-minute average guideline, therefore only the coarser resolution daily averaging period can be applied to make comparison of compliance over the longer term.The number of exceedances of the revised DEAT/WHO daily average guideline of 48 ppb is presented in Figure 8.The initial year of data capture (1996) is not reflected as data was only valid for half of the year (July 1996 onwards).The number of 10-minute average guideline exceedances recorded at Wentworth during 2003 decreased compared to the previous year, while that at AECI was consistently low.
The number recorded at Settlers was consistent, however still unacceptably high, while there was a significant increase at Southern Works, following a particularly low number of exceedances during 2002.The source directions of the 10-minute average exceedances at Wentworth, Southern Works and Settlers School are provided in the frequency roses in Figure 9.The scale represents percentage of total exceedances per station.Note the variation in scale between stations and cognisance must be taken of the wind frequency pattern for the area (Fig. 2).
Figure 9 shows the differences in the locations of the monitoring stations relative to sources and the predominant winds.At Wentworth, most exceedances were related to light westerly to WSW winds, the most likely source of which corresponds to industry in the Jacobs and Mobeni areas.The majority of these exceedances were recorded during the early morning and were correlated with Clean Air Journal, Vol 14, No 1, August 2005 atmospheric stability and poor dispersion conditions.At Southern Works, 91% of 10-minute average exceedances were measured during moderate to fresh winds from the SSW.The most likely sources are Sapref and Mondi (Sasol Fibres was shut down in 2001).This situation is in striking contrast to that at the Settlers School, less than a kilometre to the east.The Settlers School station typically measured elevated SO during moderate to fresh northerly to NNE 2 winds, which correspond to the direction of the Engen refinery.In the case of both Southern Works and Settlers, the elevated ground level concentrations are most likely a result of stack down-drafting.In isolated cases light wind events in the morning have suggested downward fumigation of accumulated pollutants with the diurnal break-up of the temperature inversion.A great deal of work has been done understanding the causes of these effects and is reported separately.Clean Air Journal, Vol 14, No 1, August 2005 revised National 10-minute average guideline for SO .(December 2001).Sulphur dioxide was measured by means of continuous ultraviolet fluorescence analysers, which were regularly maintained and calibrated.The SO analysers are designated as equivalent methods in terms of the United States Code of Federal Regulations Volume 40 Part 53.Equivalent method numbers assigned to the various analysers used are: Dasibi 4108 (EQSA-1086-061), TECO 43A (EQSA-0486-060) and API 100A (EQSA-0495-100).

Figure 3 :
Figure 3: Annual average SO2 for each of the permanent analyser stations in South Durban, 1996 2003.

Figure 4
Figure 4 indicates the annual average SO at the permanent monitoring stations for the most recent annual period (2003) compared to the previous 5-year mean for the period 1998 to 2002.The mean for Settlers School is applicable for the period June 2000 to December 2002 (data capture only commenced during June 2000).Even though the annual average SO at Southern Works increased during 2003 compared to the previous year, comparison with the 5-year mean still shows an overall improvement in air quality at the station.At Wentworth a significant reduction in annual average SO is evident for 2003 compared to the 5-year mean for 1998 to 2002, indicating good improvement in the air quality at the station.Annual average SO at the AECI station was slightly lower than the 5-year mean, while the annual average at Settlers School was consistent with the previous two years.

Figure 4 : 2
Figure 4: 2003 annual average SO2 concentrations at South Durban monitoring stations compared to previous 5 year mean

Figure 5 :
Figure 5 : Long term trend in monthly average SO2 at Wentworth

Figure 6 :
Figure 6 : Long term trend in monthly average SO2 at Southern Works Clean Air Journal, Vol 14,No 1, August 2005

Figure 7 :
Figure 7 : Long term trend in monthly average SO2 at Settlers School

Figure 8 :
Figure 8: Number of exceedances of the revised DEAT/ WHO daily average guideline since 1997.

Figure 9 :
Figure 9: Source directions of 10 minute average exceedances

Figure 11 :
Figure 11: South Durban Multi point Plan monitoring station locations The functions of the Steering Committee and monitoring network terminated at the end of 2003 as the Ethekwini Municipality embarked on a new and expanded monitoring network in terms of the South Durban Basin Multi-Point Plan.
expanded network of 16 monitoring stations, which includes the monitoring of PM and oxides of nitrogen, in addition to SO , will provide a more detailed insight into pollution levels in the Durban area.This paper summarises the results obtained from continuous SO monitoring in South Durban from July 1996 to December 2003.The South Durban industrial basin has long been an area of concern with regard to air quality.Industrial development has occurred, with little or no consideration for atmospheric pollution or the cumulative effects on air quality (South Durban Basin Multi-point Plan, 2003).Poor atmospheric dispersion conditions, during winter and topographic channelling of pollution (Diab & Preston-Whyte, 1980) contribute to elevated pollution concentrations.The close proximity of residential areas to industry and major traffic routes has exacerbated the problem.There is a history of conflict between local communities and industry, which has intensified over time (D'Abreton, et.al.) and there is increasing concern over the impacts of air pollution on human health.

Table 1 : Recommended DEAT guidelines for SO2 (21/12/2002)
revised National Department of Environment and Tourism (DEAT) guidelines for SO (Government Notice No. 1387 of Government Gazette 22491, 21 December 2002), and which are equivalent to the World Health Organisation (WHO) guidelines, are given in Table 1.Unfortunately the data averaging time of 3 minutes prior to 2002 does not allow direct comparison of historic data to the2