|
Mapping, emissions and environmental and health assessment of chemical substances in artificial turf 8 Environmental assessment
A number of environmentally harmful substances were identified in the analyses of the liquids from the leaching tests. The tests thus identified a number of environmentally harmful substances either classified with R50, R51 or R52 and possibly in combination with R53 in leaching tests of either elastic infills, artificial turf or pads. Six substances were selected which occur in significant concentrations in the contact liquids from the leaching tests. For these, ecotoxicological data have been found for the assessment of the lowest effect level (predicted no-effect concentration (PNEC)) in the aquatic environment. The substances are as follows:
In this assessment, a reassessment was also made of data for calculating the PNEC for diethyl phthalate used in the Norwegian study, and it was also decided to add an environmental assessment of bis(2,2,6,6-tetramethyl-4-piperidinyl)sebacate. In the model scenario used, it is assessed that drainage water from artificial turf football pitches may leach in three different ways:
It is not possible to make a complete environmental assessment of all three scenarios within the framework of the present project. It is thus decided to assess the drainage of surplus drainage water to a watercourse nearby in accordance with TGD’s standard models, which were also used in a Norwegian environmental assessment (T. Källquist, 2005). At the same time, the potential contribution to drinking water contamination is assessed based on the concentrations of environmentally harmful and hazardous substances, and it is estimated whether there may be an increased impact from environmentally harmful substances on drainage water drained to a sewer. 8.1 Zinc and its saltsZinc, CAS no. 7440-66-6, is classified with: N;R50/53: Very toxic to aquatic organisms/May cause long-term adverse effects in the aquatic environment. The leaching of zinc will be in the form of zinc ions – probably as a zinc chloride solution, especially in the winter season. As the ecotoxicological effect of zinc is caused by the dissolved zinc ions, such data are not found in the IUCLID data set for the metal zinc (IUCLID data set zinc), and it was thus decided to find ecotoxicological data for zinc chloride, CAS no. 646-85-7. 8.1.1 Ecotoxicological dataTable 8.1 shows ecotoxicological data for fish, invertebrates and algae. Table 8.1 Ecotoxicological data for dissolved zinc
The solubility of zinc chloride is very high, the solubility being 4,320 g/l at 25°C. The bioaccumulation factor for dissolved zinc in algae is specified to be up to approx. BCF = 10,000, whereas the accumulation in most molluscs is up to 500, apart from molluscs living in sediment, e.g. crabs, with a BCF of up to 10,000 and oysters of up to 15,000 (IUCLID data set zinc). For fish, the bioaccumulation of dissolved zinc is < 500 (IUCLID data set zinc). 8.1.2 Estimate of no-effect concentrationWith an assessment factor of 100, the NOEC value for algae gives a no-effect concentration of PNEC = 0.5 µg/l. In the Norwegian environmental study (T. Källquist, 2005), PNECwater = 3.1 µg/l was used based on a better data basis in a draft EU risk assessment of environmental effects for zinc, and this value is thus used in the environmental assessment. 8.2 6PPD6PPD 1,4-Benzenediamine, N-(1,3-dimethylbutyl)-N'-phenyl-, CAS no. 793-24-8, with the advisory classification: N;R50/53: Very toxic to aquatic organisms/May cause long-term adverse effects in the aquatic environment. 8.2.1 Ecotoxicological dataTable 8.2 shows ecotoxicological data for fish, invertebrates and algae. Table 8.2 Ecotoxicological data for 6PPD
The distribution coefficient between octanol and water is log Kow = 5.4 (IUCLID data set 6PPD), for which reason bioaccumulation can be expected (risk at log Kow = 3). Biodegradation tests show that the substance is non-degradable. An aerobic biodegradation test with activated sludge thus resulted in 7.2% degradation after 32 days and an aerobic test with household sludge resulted in 13-40% after 28 days. In connection with the biodegradation tests, the EC50 specified for activated sludge is 450 mg/l for three hours’ exposure (IUCLID data set 6PPD). Chronic toxicity No data have been found for long-term tests. 8.2.2 Estimate of no-effect concentrationData for short-time exposure show that the substance is more toxic to fish with LC50 = 0.14 against 0.5-0.6 mg/l for invertebrates and algae. As data have been found for short-term tests for the three trophic levels, but not for long-term tests, an assessment factor of 1,000 and LC50 = 0.14 mg/l is used in accordance with TGD. It can thus be estimated that PNEC = 0.14 µg/l. 8.3 DicyclohexylamineDicyclohexylamine, CAS no. 101-83-7, with the advisory classification: N;R50/53: Very toxic to aquatic organisms/May cause long-term adverse effects in the aquatic environment. 8.3.1 Ecotoxicological dataTable 8.3 shows ecotoxicological data for fish and algae. Table 8.3 Ecotoxicological data for dicyclohexylamine
The distribution coefficient between octanol and water is log Kow = 3.5 (IUCLID data set dicyclohexylamine). The water solubility is 800 mg/l, 25°C. Biodegradation tests show that the substance is easily degradable. Aerobic biodegradation tests with household sludge thus resulted in > 90% degradation after 20 days at a concentration of 0.8 mg/l, and tests in activated sludge resulted in 76.9% degradation after 14 days at a substance concentration of 100 mg/l (IUCLID data set dicyclohexylamine). Chronic toxicity Data are only available for algal growth tests. 8.3.2 Estimate of no-effect concentrationThe data for ecotoxicity are inadequate as there are no data for invertebrates. In the growth test of Scenedesmus subspicatus algae shown in the table, an NOEC of 0.016 mg/l was specified with an LOEC of 0.031 mg/l, which may be used in the assessment of chronic effect on algae in accordance with TGD. As no data are available for invertebrates, a PNEC cannot immediately be estimated. To supplement the data basis, the related substance cyclohexylamine is reviewed. Table 8.4 Ecotoxicological data for cyclohexylamine
The substance shows short-term effects with LC50 for fish being at the same level as dicyclohexylamine. The short-term effects for invertebrates are at the same level as the data for fish, and the data for algae suggest that the toxicity of the two substance is more or less the same. As an estimate of the no-effect concentration PNEC for both substances, the lowest value for the algal growth tests is used and an assessment factor of 100, corresponding to PNEC = 0.2 ug/l. As the short-term effects for fish and invertebrates are approx. 50 mg/l, the expected factor for the no-effect concentration for these trophic levels will be 1,000 under LC/EC50, corresponding to 50 ug/l. The estimated PNEC thus provides adequate protection for fish and algae. 8.4 Diisobutyl phthalateDiisobutyl phthalate, CAS no. 84-69-5, with the advisory classification: N;R50/53: Very toxic to aquatic organisms/May cause long-term adverse effects in the aquatic environment. 8.4.1 Ecotoxicological dataTable 8.5 shows ecotoxicological data for fish, invertebrates and algae. Table 8.5 Ecotoxicological data for diisobutyl phthalate
The distribution coefficient between octanol and water is log Kow= 4.11 (IUCLID data set DIBP), for which reason bioaccumulation can be expected (risk at log Kow = 3). The bioaccumulation factors for cyprinus carpio are specified to be 780 based on model calculations. Biodegradation tests show that the substance is easily degradable. Two aerobic biodegradation tests with household sludge resulted in 79 and 94% degradation, respectively, after 28 days (IUCLID data set DIBP). Chronic toxicity For Daphnia magna, an NOEC of 1 mg/l was found in tests carried out over a 21-day period with LOEC = 3 mg/l (IUCLID data set DIBP). 8.4.2 Estimate of no-effect concentrationAlgae are almost 7.4 times more sensitive than invertebrates as EC50, 72-hour algae = 1 mg/l against EC50, 24-hour daphnia magma, for which reason PNECv, in accordance with TGD’s standard assessment factors, is estimated from the lowest short-time value for algae to be PNEC = 1µg/l by using an assessment factor of 1,000. 8.5 NonylphenolNonylphenol, CAS no. 25154-52-3, classified as: N;R50/53: Very toxic to aquatic organisms/May cause long-term adverse effects in the aquatic environment. 8.5.1 Ecotoxicological dataTable 8.6 shows ecotoxicological data for fish, invertebrates and algae. Table 8.6 Ecotoxicological data for nonylphenol
The distribution coefficient between octanol and water is log Kow= 4.48 (Rar nonylphenol), for which reason bioaccumulation can be expected (risk at values above log Kow = 3). In the reference, the bioaccumulation factor is specified to be 1280. Biodegradation tests show that the substance is not easily degradable. Two aerobic biodegradation tests with household sludge resulted in 10 and 19% degradation, respectively, after 10 days, as well as 53 and 62% degradation after 28 days (Rar nonylphenol). Chronic toxicity For fish and invertebrates, the NOEC from long-term tests is specified to be between 1 and 10 µg/l, whereas the NOEC value for algae is based on an EC10 value and put at 3.3 µg/l (Rar nonylphenol). 8.5.2 Estimate of no-effect concentrationThe estimated no-effect concentration for aquatic organisms, PNECv, is calculated to be 0.33 µg/l by using an assessment factor of 10 (Rar nonylphenol). 8.6 2,4-di-tert-butylphenol2,4-di-tert-butylphenol, CAS no. 96-76-4, with the advisory classification: N;R51/53: Toxic to aquatic organisms/May cause long-term adverse effects in the aquatic environment. 8.6.1 Ecotoxicological dataTable 8.7 shows ecotoxicological data for fish, invertebrates and algae. Table 8.7 Ecotoxicological data for 2,4-di-tert-butylphenol
The distribution coefficient between octanol and water is log Kow= 5.13 (IUCLID data set 2,4-di-tert-butylphenol), for which reason bioaccumulation can be expected. The water solubility is 12 mg/l, 20°C. Biodegradation tests show that the substance is not easily degradable. An aerobic biodegradation test with activated sludge and a substance concentration of 34.5 mg/l exhibited 2% degradation after 28 days (IUCLID data set 2,4-di-tert-butylphenol). Chronic toxicity No data have been found for long-term tests. 8.6.2 Estimate of no-effect concentrationAs data have only been found for fish, a PNEC cannot immediately be estimated. A search for data on related substances has thus been carried out. Table 8.8 shows data for the related substance 2,6-di-tert-butylphenol, CAS no. 128-39-2. Table 8.8 Ecotoxicological data for 2,6-di-tert-butylphenol
Table 8.9 shows data for phenol, CAS no. 108-95-2. Table 8.9 Ecotoxicological data for phenol
For phenol, a PNEC of 7.7 ug/l has been specified based on long-term effects for fry. As can be seen, in short-term tests 2,4-di-tert-butylphenol and the related substance 2,6-di-tert-butylphenol have LC50 values almost ten times lower than phenol for fish of the same species (leuciscus idus) and EC50 for invertebrates of the same species (daphnia magna). This may indicate a no-effect concentration which may be lower than for phenol. As is the case for phenol, it is assumed that algae are the least sensitive organisms to the substance 2,4-di-tert-butylphenol and the related substance 2,6-di-tert-butylphenol. In this case, the no-effect concentration can be estimated to be PNEC = 0.45/1,000 = 0.45 ug/l based on the lowest short-time effect and an assessment factor of 1000. 8.7 Diethyl phthalateThe data for PNEC for diethyl phthalate, CAS no. 884-66-2, have in the following been reassessed in relation to the assessment made in a Norwegian study of environmental effects from artificial turf pitches (T. Källquist, 2005). 8.7.1 Ecotoxicological dataThe following ecotoxicological data have been found:
8.7.2 Estimate of no-effect concentrationThe data suggest that fish in short-term tests are three to four times more sensitive than invertebrates and a factor of two more sensitive than algae. The NOEC from the long-term test can therefore not be used for invertebrates, you should use the lowest short-term effect LC50 for fish = 12 mg/l, thus resulting in an estimate of no-effect concentration PNEC = 12 µg/l (In the Norwegian study, a PNEC = 900 ug/l was estimated based on NOEC = EC10 for algae of 9 mg/l and an assessment factor of 10.) 8.8 Bis(2,2,6,6-tetramethyl-4-piperidyl)sebacateThe data for PNEC for Bis(2,2,6,6-tetramethyl-4-piperidyl)sebacate, CAS no. 52829-07-9, are in the following assessed on the basis of information in the supplier’s instructions for use (Instructions for use, Lowilite 77). The substance is a polymer stabiliser found in very high concentrations in liquid from leaching tests of an artificial turf product. 8.8.1 Ecotoxicological dataThe following ecotoxicological data have been found:
Distribution coefficient between octanol and water is log Kow= 0.35 (Instructions for use, Lowilite 77). 8.8.2 Estimate of no-effect concentrationOnly short-term data are available, showing that algae are the most sensitive organisms. Based on this, a no-effect concentration PNEC = 1.9/1000 = 1.9 µg/l can be estimated with an assessment factor of 1,000. 8.9 Estimate of effect of discharged substances8.9.1 Discharged substances in drainage waterBelow is a calculation of the amount of substances from rubber granules and artificial turf mat in drainage water for a one-year period, based on the assumption that the concentrations found from leaching tests are equilibrium concentrations which do not decrease over time. The calculation uses an area of a pitch of 8,000 m² and precipitation of 702 mm/year based on the Danish Meteorological Institute statistics 1966-1990 for East Jutland. The amount of artificial turf mat is calculated to be 24 tonnes based on an average weight of the samples received of 3 kg/m², whereas the weight of the rubber granules is put at 100 tonnes per pitch based on mapping information. The precipitation corresponds to a liquid-infill ratio of 0.15:1 for 24 hours as well as a ratio for artificial turf of 0.64:1 for 24 hours. This is much less per 24 hours than for the leaching test. Table 8.10 Worst case estimate of discharged amounts of substances from a football pitch with elastic infill per year
The table shows that less than 1% of the zinc content and up to 20% of the phthalate content leach during a year, assuming that the leaching occurs with the same efficiency during a one-year period as during a 24-hour period. As mentioned in the analysis section, it is assessed that the substance concentrations in the contact water are maximum concentrations which are probably somewhat higher than the concentrations that will be obtained in drainage water from a football pitch. In addition, the drainage water concentrations of phthalates will probably fall over time provided that the infills remain intact over time. Further studies are, however, required to confirm this. Table 8.11 Worst case estimate of discharged amounts from artificial turf from a football pitch per year
The table shows that a large part of the phthalate content and 158% of the nonylphenol leach during a one-year period if it is assumed that the leaching occurs with the same efficiency during a one-year period as during a 24-hour period. As mentioned in the analysis section, it is assessed that the substance concentrations in the contact water are maximum concentrations which are probably somewhat higher than the concentrations that will be obtained in drainage water from a football pitch. In addition, the drainage water concentrations of phthalates and nonylphenols will probably fall over time provided that the infills remain intact over time. Further studies are, however, required to confirm this. 8.9.2 Discharge of drainage water to watercoursesThe following model calculation investigates the effect of the discharge of selected substances to any watercourses near football pitches (local effects). This is particularly during heavy showers where the pitch’s drainage system cannot take all the rainwater. The calculation is made under the same assumptions as the Norwegian study (T. Källquist, 2005). It is thus assumed that the concentration in water from leaching at a water-sample ratio of 10:1 corresponds to the concentration in the drainage water. The concentration of the drainage water Ceff is assumed to be diluted 10 times (f) when drained to a catchwater drain/small watercourse as specified as default in TGD (Technical Guidance Document, 2003). The concentration in the water phase can be calculated to be: PECwater = Ceff /((1+Kpsusp x SUSPwater x 10-6)*f) where Kpsusp is the distribution coefficient between solid and water calculated from KOC with a content of organic substance in suspended material put at the default value c org,susp= 10%. This means that Kpsusp= 0.1* KOC. SUSPwater is the concentration of suspended material (mg/l) – the value is put at the default value of 15 mg/l. The dilution factor f = 10. Koc is calculated on the basis of Kow from the QSAR model for non-hydrophobic organic substances in TGD. Thus, log KOC = 0.81*log KOW + 0.1. PECsediment =Ksusp-water/RHOsuspxPECwater*1000 where the volume-based distribution coefficient of suspended material in relation to water Ksusp-water (m³/m³) is calculated using the model tool EUSES 2.0 with input of KOC and the selected QSAR model (EUSES 2.0). The no-effect value for sediment is calculated from the no-effect value for water with the following expression: PNECsediment = Ksusp-water/RHOsuspxPNECwater*1000. Finally, the following ratios are calculated: PECwater/PNECwater as well as PECsediment/PNECsediment (margins of safety) to assess whether there is a potential local environmental effect from discharge of the substances investigated. Table 8.12 shows a worst case calculation of the environmental effect of discharging drainage water from elastic infills to an adjacent watercourse, under the assumption that the drainage water concentrations of substances correspond to those found in contact water. Table 8.12 Calculation of discharge of leached substances from infills to watercourses
Table 8.12 Calculation of discharge of leached substances from elastic infills to watercourses, contd. As described in 8.9.1, the concentrations found in contact water are worst case scenarios as the concentration of leached substances must be expected to decrease over time just as the leaching method used is deemed to be more efficient with more liquid-solid contact than the actual situation on a football pitch. In this connection, it should be noted that a lysimeter test probably underestimates the leaching level as the physical impact from the players on a wet pitch (during rain) should increase the liquid-solid contact compared to a lysimeter where there is no mechanical impact on the pitch from the football players. Table 8.12 shows – based on the assumption that the leaching from elastic infills during use of a football pitch corresponds to the results from the leaching test – an effect on adjacent watercourses from zinc at a level corresponding to the result in the Norwegian study. Phthalates may also cause environmental effects with concentration levels in watercourses in the order of 10 times above the no-effect concentration. Similar concentration levels are observed for phenol, 2,4-bis(1,1-dimethylethyl)-. The amine compounds seem to be capable of causing significant effects with concentrations in the aquatic environment of around 1,000 times above the no-effect concentration. The 6PPD concentration is based on an uncertain determination of degradation products, but the calculation indicates that the concentration in the aquatic environment may be above the no-effect concentration. Table 8.13 shows a worst case calculation of the environmental effect of discharge of drainage water from artificial turf into an adjacent watercourse. Table 8.13 Calculation of discharge of leached substances from artificial turf to watercourses.
The concentrations specified in drainage water C eff from artificial turf are based on concentrations from the leaching tests on artificial turf no. 4 Table 8.13 Calculation of discharge of leached substances from artificial turf to watercourses, contd. As described in 8.9.1, the concentrations found are worst case scenarios as the concentration of leached substances must be expected to decrease over time just as the leaching method used, as mentioned above, is deemed to be more efficient with more liquid-solid contact than the actual situation on a football pitch. In this connection, it should be noted once again that a lysimeter test probably underestimates the leaching level as the physical impact from the players on a wet pitch (during rain), all other things being equal, should increase the liquid-solid contact compared to a lysimeter where there is no mechanical impact on the pitch from the football players. Table 8.13 shows – based on the assumption that the leaching from artificial turf during use of a football pitch corresponds to the results from the leaching test on artificial turf mats – that there may be a very large effect on adjacent watercourses from nonylphenol, where the concentration in water is 115 times above the no-effect concentration. Phthalates may also cause environmental effects as the sum of the phthalate concentrations in the water phase is approx. 16 times above the no-effect concentration. The substance bis(2,2,6,6-tetramethyl-4-piperidinyl)sebacate may also cause environmental effects for all types of artificial turf with a concentration in the water phase of approx. 10 times above the no-effect concentration for artificial turf no. 4, and seven times above the no-effect concentration for artificial turf mat no. 2 (calculated values). Artificial turf mat no. 7 does, however, cause a very intense effect, suggesting that the artificial turf mat is stabilised using an inappropriate water-soluble chemical. The leaching of the stabiliser is significant due to its solubility in water, resulting in a concentration in the water phase that exceeds the no-effect concentration 18,000 times. The results show that a number of environmentally harmful substances can be leached from both elastic infills and artificial turf having a potential environmental risk in the case of any spillover of drainage water to adjacent watercourses, but it is assessed that measurements are required of the substance concentrations under real conditions on the football pitches to be able to assess the risk. This is because significantly lower leaching values from football pitches were expected. The expectation of lower leaching values is based on the previously cited Swiss and French studies mentioned above. In such measurements, it is possible to benefit from the substances found and assessed in this study. 8.9.3 Risks for drinking waterThe substance concentrations in contact water have been compared to the requirements set out in the Executive Order on Drinking Water. Table 8.14 Drinking water requirements and concentrations found in contact water
As can be seen, concentrations of phthalates have been measured in contact water which are potentially significantly higher than those allowed in drinking water. One single measurement of the drainage water from a football pitch on Zealand also shows that the concentration in drainage water may exceed the drinking water levels set out and thus poses a potential risk in connection with the seeping of drainage water. The present leaching tests were carried out on new infills and artificial turf, and no comparable measurements of drainage water from newly-installed pitches are available. Analyses of drainage water from artificial turf pitches are required to be able to assess the real risk of groundwater contamination including an investigation of how the leaching of substances changes over time. It has not been possible to carry out analyses of drainage water within the financial framework of this project. 8.9.4 Significance of drainage of water to sewerThe leaching tests showed high concentrations of phthalates in the contact water, which may contribute to an impact from phthalates on sewage treatment plants. It is assessed that the leaching values will be highest at the beginning when a pitch is installed, and that the concentration in the drainage water from the pitch is probably somewhat lower than in the liquid from the contact tests due to a poorer liquid-solid contact. In (Status, 2003), the concentration of phthalates discharged to Danish sewage treatment plants is estimated to be approx. 25 µg/l (see Table 8.15). The results from the leaching of infills and artificial turf show that there may be a not insignificant contribution from drainage water from football pitches. The significance of the leached substances for the impact on sewage treatment plants is expected to be very dependent on local conditions. Table 8.15 Phthalates in waste water to sewage treatment plants
It is assessed that an analysis is required for phthalates in drainage water directly from football pitches to be able to assess the contribution of phthalates discharged to sewage treatment plants. An aspect of the monitoring should be to observe how quickly the concentration decreases following the installation of a new pitch. 8.9.5 ConclusionA number of environmentally harmful substances were found in the contact water from leaching tests on infills and artificial turf mats. Based on a comparison with foreign lysimeter tests, the leaching tests are expected to have exhibit significantly better liquid-solid contact. A leaching velocity higher than that obtained in lysimeter tests should thus be expected. The lysimeter tests are assessed to be more representative for the conditions on a football pitch. A worst case assessment has been made on any discharge of drainage water to watercourses based on the Danish leaching tests. Here, it is assessed that there may be environmental effects associated with infill tests for:
It should be noted that the ratio between the concentration and the no-effect concentration for the amines and 6PPD is a factor of almost 1,000. It is assessed that leaching from artificial turf mats may result in potential environmental effects for:
It should be noted that the ratio between the concentration and the no-effect concentration for nonylphenol is above 100 and for Bis-(2,2,6,6- tetramethyl-4-piperidinyl)sebacate almost 20,000. It is not known how much the leaching for the substances tested decreases under more realistic leaching conditions, e.g. lysimeter tests, but it is assessed that it cannot be ruled out that there may be an environmental effect for the substances for which a high value was observed in the ratio between the concentration and the no-effect concentration. To determine whether it is a real environmental effect requires lysimeter tests or measurements on drainage water from football pitches over a longer period of time (e.g. one year). The fact is that foreign results suggest a decreasing concentration over time (e.g. one year). It cannot, however, be ruled out that degradation of rubber may occur over a long period of time, which again increases the leaching values (e.g. a time-frame of 10-20 years). Such lysimeter tests or measurements on drainage water from football pitches over time are also required to be able to assess whether there is any risk to the drinking water. The substances which may pose a potential risk based on the requirements set out in the Executive Order on Drinking Water are assessed to be:
as the concentrations in the contact water from leaching tests are in the order of 20-800 times above the threshold values for drinking water. Here, it should be noted that the substances, to a varying degree, will be absorbed by the sand/clay layers which the drainage water passes. Zinc concentrations in percolate of significance for the drinking water quality have thus not been observed in foreign lysimeter tests. No data are, however, available for phthalates and nonylphenols under such realistic conditions. The assessment of the impact on waste water systems also requires more realistic lysimeter tests or measurements on drainage water from football pitches over time.
|