|
A Technical Guide for Wine Producers
|
|
RECENT ARTICLES | WYNBOER HOME
Composition of effluents from wineries in the Western and Northern Cape provinces (Part 1):
Seasonal variation and differences between wineries
Reckson Mulidzi1, Giel Laker2, John Wooldridge1, Lourens van Schoor3
1 ARC Infruitec-Nietvoorbij, Stellenbosch
INTRODUCTION
During the 2000 vintage season studies were conducted at a number of wineries to gain information concerning the nature and magnitude of risks associated with possible pollution hazards at South African wineries. Several papers outlining the legal requirements for wastewater disposal and the potential pollution hazards of winery wastewaters have been published in Wineland magazine; notably by Hayward et al. (2000) and Van Schoor (2000, 2001). These papers described measures that wineries could take to ensure compliance with the legal requirements concerning effluent disposal. Aspects of these early studies were further reported upon by Bezuidenhout et al. (2002) and Mulidzi et al. (2002).
This paper presents data on effluent composition at ten wineries in the Western and Northern Cape Provinces, aspects of which have been reported by Mulidzi (2001) and Van Schoor & Mulidzi (2001). The data presented here should be read in conjunction with the companion article, which concerns the environmental impacts of these effluents.
MATERIALS AND METHODS
Regions represented by the ten participating wineries were Paarl, Stellenbosch, Worcester, Robertson and the Olifants and Orange Rivers. A Berg River winery was also included. Effluent samples were collected at 4-week intervals from ten wineries over the period December 1999 to April 2000. Samples were also collected from some of these wineries during June and July 2000. Sampling and analysis were carried out as described by Mulidzi (2001). Parameters determined were chemical oxygen demand (COD), pH, sodium (Na), chloride (Cl) and potassium (K) concentrations, sodium adsorption ratio (SAR) and electrical conductivity (EC).
RESULTS AND DISCUSSION
COD levels
A characteristic of winery wastewaters is that they contain organic material. The amount of organic material present is most usefully expressed in terms of the amount of oxygen required to oxidise (break down) the organic material in a given volume of the waste water. The results are usually expressed in mg O2 per litre (mg/l).
In the winery waste waters sampled by Van Schoor & Mulidzi (2001), the lowest COD values tended to occur in the pre- (December, January) and post harvest (June, July) periods (Table 1). In the intervening period, COD values commonly exceeded the maximum acceptable value of 5 000 mg/l, sometimes on two or three consecutive sampling dates. Effluent volumes were not measured during the sampling program described here. Nevertheless, Van Schoor (2000) has shown that peak harvest periods tend to coincide with periods of peak COD, and also with peak periods of waste water generation. That COD and wastewater production both reach their maximum at the same time has serious implications. This is because the acceptable COD limit of 5 000 mg/l is only valid where less than 50 m3 of wastewater is produced each day. Where more than 50 m3 of waste water is produced per day the maximum level of permissible COD decreases to 400 mg/l (Bezuidenhout et al., 2002). A further restriction is that the 400 mg/l limit applies where no more that 500 m3 of effluent are generated per day. Rates of effluent production greater than 500 m3 per day would probably meet with even more stringent restrictions. Variation in the period of high COD values reflected local differences in harvest period in the individual areas concerned. A sequence obtained by deducting the average of the two lowest COD values from the two highest COD values for each winery suggests that whereas high COD’s are not commonly observed at Paarl 1 and Robertson 2, the problem is greater at Paarl 2, Paarl 3, and Berg River, greater still at Stellenbosch, Worcester and Robertson 1, very high at the Orange River winery and critical at Olifants River. Both the COD values reported here, and the magnitudes of the differences between wineries, are in agreement with observations reported by Bezuidenhout et al. (2002). Since only a small number of wineries were sampled in each region, these results must be regarded as tentative.
Table 1: (click table to enlarge in separate window)
It is improbable that the organic matter in the higher COD effluents can be broken down by aerobic decomposition in the soil, especially at the high rates of application that are likely. Where aerobic decomposition does not take place at a sufficiently rapid rate, the largely undecomposed residues are likely to move into the subsoil where anaerobic processes predominate. These generate unpleasant odours and a variety of incomplete breakdown products, all of which have the potential to severely pollute groundwater and any watercourses into which they may seep (Mulidzi, 2001).
Few wineries have management systems capable of addressing the problems posed by high levels of COD, probably because early literature created the false impression that aerobic decomposition in the soil is able to break down organic matter applied in effluents. Nevertheless, Bezuidenhout et al. (2002) reported progress at the wineries which they studied in 2000 and 2002.
pH
pH is a measure of the acidity or alkalinity of a solution. Aqueous solutions with pH’s below seven are considered acidic, while those with pH’s greater than seven are basic (alkaline). Since the minimum pH allowed by the South African Water Act (Act No.36 of 1998) is 6.0, whereas winery effluents are usually characterized by pH values below 5.5 (Levay, 1995, Mulidzi, 2001; Van Schoor & Mulidzi, 2001), it is apparent that low pH’s in winery waste water is a perennial problem. In the present study, although some pH’s were above 5.5, notably during the pre-harvest period (December and January), the main winemaking period was characterized by effluent pH’s that were frequently below 4.0 (Table 2). However, with the exception of a few isolated cases of minor reductions in soil pH levels below those of the control sites, there was little evidence of acidification of the soils at effluent disposal sites (Mulidzi, 2001). Indeed, some soil pH’s were higher at the effluent disposal sites than at the control sites. This suggests that amelioration was being carried out effectively. According to Hazell (1997) the most effective method of amelioration is to add calcium hydroxide slurry at the point where the effluent leaves the storage dam and enters the irrigation system.
Table 2: (click table to enlarge in separate window)
Sodium adsorption ratio (SAR) and sodium concentration
Effluents containing high concentrations of sodium (Na) may lead to dispersion and physical degradation of soils. This is because the cations that occupy the charge sites on the exchange complex rapidly come to resemble, in terms of species and of their relative proportions, those of the equilibrating solution (the effluent). In normal agricultural soils the exchange complex is dominated by the divalent cations calcium (Ca) and magnesium (Mg). These, with their ability to bond soil particles together, particularly where finely divided organic material is abundant, have a stabilizing effect on soil structure. However, where the divalent cations are replaced by Na, the stability is lost, mainly because sodium ions acquire thick hydration shells and have a mutually repulsive effect. This causes inter-particle linkages to break down, enabling the soil particles to separate and move freely and independently, a process known as deflocculation. However, when the soil dries, surface tension draws the particles back together. In the absence of the stabilizing effects of divalent cations and organic material, the particles are able to rotate, ultimately achieving a state where porosity is minimized and bulk density maximized. In the dry state such soils are characterized by low infiltration rates, poor gas exchange, high soil strengths and extreme resistance to root penetration. On wetting, strength may be retained until a threshold is reached, beyond which the soil suddenly flows like a viscous fluid. In general, the higher the percentage of exchangeable base cations constituted by Na, the greater the negative effect. Where effluent is applied to the land by an irrigation system, South African legislation stipulates that SAR values must not exceed 5.0.
With the exception of Paarl 2, where SAR values exceeded 5.0 on all sampling dates, and by Robertson 1, where very high SAR values were observed in the pre-harvest period (December and January, Table 3), the effluents were characterized by SAR values that were below 5.0 on most of the sampling dates. High pre-harvest SAR values in effluent can be related to the cleaning of tanks before the harvest starts. According to Glaetzer (1995) sodium mainly enters the wastewater stream via the use of cleaning chemicals such as caustic soda (sodium hydroxide) and sodium hypochlorite. These chemicals may also be used during the harvest period. SAR values at Paarl 3, Robertson 2 and Berg River were consistently low. In the case of Paarl 2, which was the only winery where the effluent SAR values were high at all sampling dates, it is possible that the high SAR values were due to the use of poor quality borehole water in the cellar.
Table 3: (click table to enlarge in separate window)
As anticipated, incidences of raised sodium concentrations in the effluents (data not shown) corresponded with the raised SAR values shown in Table 3, sodium concentration tending to increase with SAR. The main concern associated with high-sodium effluents is the risk that sodium will leach from the disposal site, causing wide-spread sodicity. As will be shown in the companion article, evidence of negative soil effects induced by excessively high soil sodium concentrations was apparent at all of the disposal sites. Whereas the wineries seemed to manage and control the potential soil acidity problems well, the management of sodium at the disposal sites continues to be a major problem.
Electrical conductivity (EC)
Since the ease with which a solution conducts an electrical current is related to its electrolyte content, and since salts are the most common electrolytes in soils and effluents, it is convenient to use EC as an indication of the amount of dissolved salt that is present. Clearly, at any given application rate, the rate at which salt will accumulate in a soil (the rate of salinization) will increase with the salt content of the effluent. Low and high salinity waters have EC’s of <25, >75 mS/m, respectively. The maximum acceptable EC for disposal purposes is 200 mS/m (Bezuidenhout et al., 2002). EC values in excess of 200 ms/m were observed at five wineries (Table 4). At Paarl 2 and Olifants River, high EC values were observed at all sampling dates. These wineries were also characterized by high SAR values. The effluents from these wineries must be regarded as hazardous from a salinization viewpoint, both on and off the disposal site. Whilst the high EC from Paarl 2 can be attributed to low quality borehole water, the source of the salinity at Olifants River is not known. Isolated occurrences of high EC values were also observed at Paarl 1, Robertson 2 and the Orange River. At five of the wineries, the EC values were below 200 mS/m at all sample dates. Excessively high effluent electrolyte concentrations were therefore not a major problem at half of the participating wineries, although all samples fell in the medium to high salinity brackets. In the study reported here, EC was not determined on the pre-winemaking effluent samples (December, January), when the use of sodium-containing cleaning materials might have been expected to lead to high EC values at most wineries. This will need to be studied further.
Table 4: (click table to enlarge in separate window)
Chloride
Like sodium, chloride is phytotoxic, even at fairly low concentrations. From Table 5 it is apparent that the effluents from seven of the ten participating wineries contained acceptable concentrations of chloride (i.e. below 200 mg/l) throughout the study period. However, in addition to having high SAR, ER and sodium concentrations, chloride was high at Paarl 2 at all sampling dates, probably because chloride was abundant in the borehole water. At Stellenbosch and Olifants River, very high chloride concentrations were observed, but at one date only. These were attributed to the use of cleaning materials.
Table 5: (click table to enlarge in separate window)
Potassium
Unlike sodium, potassium is not toxic to plants. Neither do high potassium concentrations have negative impacts on soil physical conditions. Although potassium is one of the three major plant nutrients, it is not usually regarded as hazardous from a eutrophication viewpoint (unlike phosphorus and nitrogen) unless rapid leaching leads to its entry into streams. For these reasons potassium is not usually regarded as an important factor from an environmental monitoring viewpoint. Nevertheless, excessive K concentrations can lead to nutritional imbalances in a variety of fruiting plants. For this reason, and for the purposes of this study, effluent K concentrations above 200 mg/l were tentatively classified as high. Relative to this figure, the potassium concentrations in eight of the ten participating wineries (the exceptions being Worcester and Robertson 2), were high for all or part of the sampling period (Table 6). The origin of the potassium is uncertain. Potassium is abundant in grape juice, and could conceivable become concentrated in the bottom of tanks through settling, from whence it could be flushed as effluent. Potassium is also a component of certain detergents. The concentrations of potassium at Olifants River at the February to June sampling dates nevertheless appear to be too high for these explanations to be viable.
Table 6: (click table to enlarge in separate window)
GENERAL
The study reported here was incomplete in that it did not include the full list of environmentally important parameters which Van Schoor (2000) considers necessary for a complete environmental audit. Parameters that were omitted included Total Dissolved Solids (TDS), Suspended Solids (SS), Biological Oxygen Demand (BOD), nitrogen, phosphate and sulphur. Bezuidenhout et al. (2002) included TDS and SS in their survey and report average values for these for several cellars. Soil analysis data from the present study (not shown) indicates that trace metals, like manganese, copper and zinc, should also be determined in winery effluents.
CONCLUSIONS AND RECOMMENDATIONS
None of the ten participating wineries produced effluents that complied with the required environmental standards on all of the sampling dates. High COD values are a perennial problem particularly where effluent volumes are large. Further research on the reduction of COD in winery effluent is urgently needed. High effluent acidity (low pH) is also common, but appears to be well managed during disposal at the majority of the wineries. High sodium and potassium levels are found periodically at most wineries. Salinity (high EC) and chloride problems was observed at a minority of the wineries.
As is to be expected from the wine making process, most of the effluent parameters show seasonal trends. These must be taken into consideration when effluent sampling and management strategies are developed. In general, COD and potassium concentrations may be expected to peak during harvesting and pressing. High sodium concentrations, on the other hand, are more likely to occur during the pre-harvest period, or in association with cleaning operations during the season. Effluent pH’s are also most likely to be close to neutrality in the pre-harvest period. EC and chloride did not show clear seasonal trends.
Wineries differ markedly in terms of effluent quality, and with regard to the nature and severity of the problems encountered in the handling and disposal of effluents. There is no single solution that can be applied in all cases. For this reason it is recommended that all wineries should commission environmental audits so that their specific problems can be identified and addressed in the most economical and effective manner. Mulidzi (2001) suggested that monitoring should be carried out on a monthly basis for the first year, whereas Van Schoor (2000) is of the opinion that an intensive monitoring programme should be followed for the first 2 or more years’, followed by routine monitoring thereafter. Regardless of the strategy followed, the data obtained from the samples will be easier to interpret if accurate records are kept so that the winery activities which correspond to each sampling date are known.
For further information contact Reckson Mulidzi at mulidzir@arc.agric.za.
REFERENCES
Bezuidenhout, S., Hayward, N., Lorenzen, L., Barnardt, N. & Trerise, M., 2002. Environmental performance of the SA wine industry - are we competitive? Wynboer Tegnies 153, 79 - 81.
|
|
Wynboer is incorporated in WineLand, magazine of the SA wine producers.
Visit our sister sites:
|
| UP |
COPYRIGHT (C) 2000 WineLand |