[Home] [Journal papers]

SO2 REDUCTIONS AT A CHEMICAL DISSOLVING PULP MILL
Derek D Airey 

ABSTRACT

Sappi Saiccor is a producer of chemical dissolving pulp on the KwaZulu-Natal south coast. Sulphur and coal are two of the raw materials used in the manufacturing process and thus there is a great possibility that SO2 will be emitted into the environment. An SO2 monitoring network was initiated by Sappi Saiccor 14 years ago, and has been extended to the present size. 

The results show that progress has been made to reduce total SO2 emissions from the mill by installing or modifying equipment which has resulted in a reduction of the ambient SO2 concentration at the ambient monitoring stations. Various emission sources are identified using the SO2 monitoring system.

A major component in achieving the reductions in emissions is the use of a four – tier SO2 monitoring network. These tiers are: an electrochemical sensor network, stack monitors, an early warning system and atmospheric monitoring stations. The use of the HAWK dispersion model was used as a verification tool, together with the data from the atmospheric monitoring stations, when responding to complaints from the public. 

The paper deals with how each tier is used in ensuring that the mill strives to be compliant with legal requirements and that there is continual improvement in reducing emissions. Technical aspects of each of the tiers are explained as well as the recovery, collation, reporting and use of the data collected from the network.

INTRODUCTION

Sappi Saiccor, located on the bank of the Umkomaas River on the Kwa-Zulu Natal south coast, produces chemical dissolving pulp from plantation trees using the acid bisulphite process. Sulphur is burned in air and the resultant SO2 mixed with limestone or magnesium oxide slurry to produce the corresponding bisulphite cooking liquor. The liquor is loaded with wood chips and steam into large digesters and "cooked" for between 6 and 7 hours.

A large amount of steam and electricity is required by the mill which is produced by 6 coal fired boilers and a chemical recovery boiler. The emissions from the boilers go up two 64m high stacks which are permanently monitored on-line.

The bisulphite process is listed on the Second Schedule (# 15) of the Air Pollution Act (Act 45 of 1965) 1, together with power generating (# 29) and Certificates of Registration are required from DEAT for the mill to operate. With SO2 being used in the process, emissions into the surrounding environment occur and need to be monitored and ultimately reduced.

The challenge for efficient dispersion of any emission from any source is the surrounding topography – which in this case is a valley. The mill is surrounded on three sides by hills which are greater than 80 m high. The dispersion is exacerbated in winter when an inversion traps and concentrates the emissions.

Following an ambient emission survey, undertaken by consultants in 1986, an emission reduction programme was initiated. Since then, the major SO2 sources have been identified, sections of plant modified, scrubbers installed where appropriate and coal with a lower sulphur content purchased. The SO2 emissions have thus been reduced from 68 t/d to < 8 t/d since 1986.

Concomitant with these reductions, there has been a reduction in the SO2 concentration recorded by the ambient monitoring stations, which forms an integral part of the overall emissions reduction programme. At the same time there has been a reduction in the number of complaints from the public. These are received and dealt with speedily by environmental staff following a procedure whereby various information gathering devices are checked to confirm or reject the complaint. 

Fugitive emissions have played a part, primarily resulting in high SO2 levels within various areas of the mill. These have been targeted with the "sniffer" network of sensors. This network was initiated to comply with the Section 9 (e) of the Hazardous Chemical Substances Regulations of the OHS Act (# 85 of 1993) 2 and the results used to implement SO2 reduction plans. 

The reduction of SO2 emissions forms the basis of the Objectives and Targets of the ISO 14000 Environmental Management System (EMS).

MATERIALS & METHODS

Emission Reduction Programme
Following the consultants report in 1986 there have been regular modifications to the mill which have resulted in a reduction of the emissions from the site by 88 %. 

A variety of instruments have been used to reduce emissions, from purchasing grades of coal with low sulphur content, installing scrubbers in various plants and redirecting vents to enable the sulphur to be collected and reused.

The DEAT Certificate of Registration restricts total SO2 emissions from the Sappi Saiccor site to 12 t/d, this includes any emissions from the LignoTech SA plant.

Monitoring System
Sappi Saiccor has one of the largest SO2 monitoring systems in the country, consisting of eight ambient monitoring stations positioned strategically around the mill, seven early warning units within 2 km and over 110 "sniffers" located within the mill. The aim of the system, initiated in the early 1990's, was to help with the identification of sources and quantity of emissions.  

Ambient Monitoring Stations (AMS)
The AMS were placed at strategic localities surrounding the mill. The selection of the sites was based on the population size, risk, infrastructure and historical data such as a record of complaints from the public. Each station, consisting of a temperature controlled caravan and monitoring equipment, is set up to sample air quality and monitor wind conditions (speed and direction). The data is transmitted to the mill for display and archiving, as well as being logged on a data logger (PC) within the station.

Data stored in the PC's in the monitoring stations is down loaded every month end by the environmental officer for analysis. The file downloaded from the PC in the AMS is a Microsoft Access file and is converted into an Excel file. During the analysis, the following are determined for compliance to the DEAT registration certificate:

  • Instantaneous (5 minute average) SO2 readings above 600ppb - (Instantaneous excedences),
  • Hourly SO2 average above 300ppb - (Hourly excedences)
  • Daily SO2 average above 100ppb - (Daily excedences).

The following are also determined for compliance to Section 63 of the National Environmental Management: Air Quality Act (Act No. 39 of 2004) 3:

  • Instantaneous (10 minute average) SO2 readings above 191ppb - (Instantaneous excedences),
  • Daily SO2 average above 48ppb - (Daily excedences).

Those above together with the following

  • Monthly SO2 average
  • Monthly maximum, and
  • % Data Recovery, are calculated for inclusion in reports.

The instantaneous data (5 or 10 minute averages) is filtered to determine values over 600ppb or 191ppb respectively (instantaneous excedences). The hourly data and daily data are filtered to determine the hourly excedences (300ppb) and daily excedences (100ppb or 48ppb) respectively. The monthly average is calculated and the monthly maximum noted. The filtering is to remove any spurious data recorded as a result of a weekly span or calibration check conducted by our instrument technicians.

The above results are reported internally to senior management and externally to the Department of Environmental Affairs & Tourism and eThekwini Metro. The causes of the excedences and the corrective actions are discussed in the reports.

"Sniffers" (Electro-Chemical Sensors - ECS)
There are over 80 SO2 sensors installed at various places in the mill. This sniffer (ECS) network system replaced personal monitors, which used to be left with individuals or placed at certain strategic areas for a specific period of time, after which data would be downloaded and analyzed. Data from the sniffer's is transmitted to various control rooms and POCC (Plant Operations Co-ordination Centre) computer system where it is trended and alarmed. Fugitive emissions from various parts of the plant are thus monitored continuously, and any deviations above preset limits are alarmed in the respective control room and corrective action is taken.

The ECS network is also used when a worker reports to the Occupational Health Centre after having been exposed to a gaseous emission. The level recorded at the nearest ECS to where the worker was exposed, is noted by the OHC staff when administering treatment.

Complaints
There are three forms of complaint dealt with by Sappi Saiccor staff.

External complaints
Complaints from the public are usually received telephonically via Saiccor's toll free number. During office hours the calls are forwarded to the SHEQ department. After hours the duty security officer records the complainant's details and if an urgent response is required, will then advise one of the shift superintendents on duty. The superintendent will investigate the complaint, take corrective action and print the relevant trends and readings from the computer system in the POCC room. The shift superintendent will then advise the SHEQ manager. The duty security officer also e-mails the information to the SHEQ manager. The SHEQ manager will then use the shift superintendent's response together with the trends retrieved from the POCC computers to fully investigate the complaint and supply feedback to the complainant.

Written complaints.
Written complaints from the public are referred to the SHEQ manager for investigation. The SHEQ manager completes the required sections of the Incident Report form so that the complaint is correctly documented and closed out. The results of the investigation are sent to the General Manager who then replies in writing to the complainant.

Internal complaints.
Internal complaints from Saiccor personnel are normally phoned straight through to the SHEQ department. One of the staff will check details of the nearest ECS on the POCC computers. The relevant sections of the Incident Report form will be completed. The SHEQ manager will then distribute the completed report according to the distribution list on the form, and one copy is filed.

HAWK

Although not part of the AMS, the well-known HAWK atmospheric dispersion model played a role in verifying external complaints. Data was input into the computer programme continuously from the eight AMS (SO2 and Wind speed & direction) as well as from the UVW vane located near the centre of the mill and the stack SO2 analysers. Although not working at present, it is hoped to replace this with an ADMS in the near future.

RESULTS

Reductions Programme
The first modification to the mill following the consultant's report in 1986 was the installation of a line connecting the Hagglund tower vents to the MgO venturi's in 1988. SO2 is absorbed into the limestone slurry in the Hagglund towers and the "off" gases were passed directly into the coal boiler stacks. The modification allowed this excess SO2 to be utilised in the process. This resulted in an emission reduction of approximately 23 tons per day of SO2 from 68 t/d.

A move to using a coal with a lower sulphur content followed in 1990. This further reduced the mill's SO2 emissions by 6 tons per day. The MgO blowtank scrubber was commissioned in 1993 and this was followed by a scrubber on the washpits on the calcium side of the mill in 1995. Both these scrubbers resulted in an 18 t/d reduction in SO2.

In 1996 a 6th venturi scrubber was added to the MgO plant. In addition an effluent pipeline replaced the effluent channel and new pump house at the mill replaced the old pump house at the river mouth. These two measures resulted in a 6 t/d SO2 reduction as well as a reduction in the number of complaints from Umkomaas West.

Between 1997 and 1999, three further instigations brought the total SO2 emission level for the mill down to <12 tpd (see figure 1). One was the installation of a "low concentration gas collector system", which collects gasses from filter hoods and other sources, and passes them to the MgO recovery boiler where they are burnt. This system also reduces the VOCs from the MgO plant and as a result there was a reduction in the numbers of complaints. The others were the installation of a vacuum breaker and the purchase of an even better quality coal with a lower sulphur content.

so2_figure1
Figure 1 SO2 EMISSIONS REDUCTIONS SINCE 1986.

The coal fired boilers were fitted with scrubbers in 2003 and commissioned in 2005 reducing the SO2 emitted from the site to 7.7 tons per day.

With the completion of the expansion of the mill (Project Amakulu) the SO2 emissions are expected to be reduced further to 5 ton per day.

Ambient Monitoring Stations
Data from the eight AMS is used to collate the monthly reports. Some of the AMS have only been at the locations since 2000. Results of the monthly and yearly averages are presented for two of the eight AMS, "Drift" and "Umkomaas", which have data sets since May 1994 and April 1995 respectively.

Drift.
This AMS has been in three locations within the Drift area since 1994. It was initially located in the grounds of the Drift Primary School, not far from Saiccor's main entrance, for three years. It was then moved approximately 500m further from the mill. Finally at the end of 2000 it was moved approximately 100m further to one of Saiccor's properties, where it still is today.

so2_figure2
Figure 2 AVERAGE ANNUAL SO2 LEVELS AT DRIFT AND UMKOMAAS.

Annual averages show a reduction in the annual average SO2 levels recorded since 1994, with the greatest being between 1996 and 1997 followed by a continual decrease (Figure 2). The same data but month by month shows a typical result from a locality affected by summer ambient conditions – prevailing winds from North East and South West (Figure 3).

Umkomaas
This AMS was placed on top of the Umkomaas Spar in 1995, and in 2002 was moved approximately 300 m north of the Spar onto one of Saiccor's properties. Generally there is a downward trend in SO2 levels, but the years 1999, 2002, 2004 and 2006 show slight increases (Figure 2). This is most likely due to ambient conditions.

Monthly averages show the reverse of that shown by Drift in that the area is affected by winter conditions, with the highest levels between April and August (Figure 3).

so2_figure3
Figure 3 AVERAGE MONTHLY SO2 LEVELS AT DRIFT AND SPAR.

Sniffers (Electrochemical Sensors)
Data from all the sniffers is stored in the Information Management System (IMS) computer as 8 hour shift averages. Data provided by the computer programme is used to calculate monthly SO2 averages for all the sniffers. The number of shift TLV exceedences for each sniffer is also computed. The results are then trended and submitted to plant managers. The plant managers implement action plans to reduce SO2 gas emissions in their respective plants. Operation staff normally conduct their functions from a control room and are not exposed to these levels continuously during their shift.

One of the ISO 14000 objectives was to reduce TLV exceedences of SO2 to less than 10 % of shifts by December 2002. This objective was achieved in most plants. An example of a sniffer trend is shown in Figure 4. As the graph shows, there was a downward trend in SO2 average concentration as well as TLV exceedences, until December 2002. These results were achieved through a series of SO2 abatement projects. The graph shows the slackening off in vigilance in the area which resulted in a rise .

The MgO plant area showed the most improved reductions in terms of SO2 gas emissions initially, but this has decreased in the last two years. The plant manager for this area spent nearly R 2 million on seven projects to reduce SO2 emissions.

so2_figure4
Figure 4 MgO PLANT SO2 AVERAGE & TLV EXCEEDENCE TRENDS.

DISCUSSION

AMS have been placed in areas close to where members of the local community have complained.

The two AMS with the longest data sets, Drift and Spar, show a reduction in the level of SO2 recorded each year. The annual average for the Drift AMS shows a downward trend, which may be a result of the relocation of the AMS to three different localities, each further from the mill. To ascribe this decrease to extra dilution due to distance is refuted by the trend for the Spar AMS which shows the same general reduction in SO2 levels.

An anomaly are the results for 1999, 2002, 2004 and 2006 where at the Spar AMS a slight increase relative to the previous year was recorded. There were no specific incidents which occurred on site which would have resulted in such an increase.

The annual average SO2 for the Drift AMS fell below that for Spar in 1999 and has remained there since. The Drift AMS is approximately 850 m from the site and at about the same altitude as the top of the stacks, while the Umkomaas site is 3.5 kms away. The average SO2 levels for both sites since initiation are similar with the Umkomaas site being less than 1 ppb higher than Drift. The emissions from the stacks having more of an effect at a greater distance from the mill.

Monthly average SO2 results show the expected seasonal variations from two sites at right angles from each other in relation to the source, and prevailing wind patterns.

The data from the Umkomaas AMS does show (Figure 5) that there is an increase in the number of complaints in winter when the prevailing wind conditions transport the emissions from the mill towards Umkomaas, as well as an increase in the monthly SO2 level. However, there are few occasions when a complaint is received and the level of SO2 recorded at an AMS is near the Instantaneous Guideline level (191 ppb)

so2_figure5
Figure 5 COMPLAINTS AND AVERAGE SO2 LEVELS (PPB) AT THE UMKOMAAS AMS.

Every complaint received is thoroughly "researched" to determine the probable cause. As stated earlier, very few complaints are a result of a high SO2 level recorded at an AMS. The majority of complaints occur as a result of "factory odour", a combination of odours emitted from the stacks, and the by-products of cooking timber.

CONCLUSION

SO2 emissions from Sappi Saiccor have been reduced over the years as sources have been identified and capital expenditure approved to solve the problem. It is expected that these will be further reduced when the mill expansion (Project Amakulu) comes on line in 2009.

As with the majority of industries the human factor plays the largest role – training of all employees in the role they can play in identifying sources will ensure a healthy environment for all in and around Sappi Saiccor. 

REFERENCES

1 Atmospheric Pollution Prevention Act 45 of 1965, Second Schedule.
2 Occupational Health and Safety Act 85 of 1993.
3 National Environmental Management: Air Quality Act (Act No. 39 of 2004)

ACKNOWLEDGEMENTS

The authors would like to thank the Sappi Saiccor Management for permission to present this paper, as well as colleagues for their constructive criticism.

[Home] [Journal papers]