Two approaches have been used to assess the impact of municipal solid waste incineration bottom ash on lentic ecosystems, especially lake coastlines. One of the aims of this study was to complete the methodology for the assessment of waste ecocompatibility by assessing a scenario in which bottom ashes are reused as road embankment. A laboratory lysimeter was chosen to simulate the road embankment and produce the bottom ash leachate. The first approach was based on three bioassays. Results led to the following ranking of these bioassays based on organisms sensitivity, in descending order: algae Pseudokirchneriella subcapitata > duckweed Lemna minor > cladoceran Daphnia magna. At the same time, leachates were assessed with a 2 litre freshwater/sediment microcosm. All species were impaired. Toxicity effects increased with leachate concentration, from 1.56% to 8.0%. Comparison between bioassays and microcosm assays revealed that the representativeness is higher in the multispecies systems. Finally, bottom ashes have been assessed in a simplified risk assessment procedure. Predicted environmental concentration is close to the concentration that caused first effects in microcosms. Recommendations are made for the reuse of bottom ashes as road embankment.
Municipal solid waste incineration (MSWI) produces large amounts of bottom ash. In France, these secondary wastes represent 25 to 30% of the municipal solid waste, that is, more than 2.7 Mt y−1 (ADEME, 2002). This bottom ash is composed of heterogeneous wastes (slag, ferrous and non-ferrous metals, ceramics, glass, combusted and non-combusted organics) and consists of porous lightweight aggregates which contain high concentrations of metals and salts, particularly chlorides and sulphates (Wiles, 1996; Chandler et al., 1997).
In France, this combustion residue is used more and more in construction, particularly for roads, rather than being disposed of in landfills. This approach has been promoted by a law passed in 1992 concerning waste elimination. Indeed, since the year 2002, only final wastes can be placed in landfills. However, the MSWI bottom ash reuse depends on a French regulation that is based on only a few physico-chemical characteristics of wastes (MATE, 1994): soluble fraction, release of few elements (As, Pb, CrVI, Hg, Cd, O42−) and total organic carbon content. The impact of the waste deposit on the recipient environment is not taken into account (Perrodin et al., 2002). To achieve this goal, ADEME (French Agency for Environmental Management and Energy Conservation) launched a research program called ‘Waste Ecocompatibility’ in 1995. Ecocompatibility has been defined as a situation where the pollutant flux from waste disposed of, or used, is compatible with the environmental acceptance of the receiving environments. This program has led to a methodology for the measurement of the impact of waste storage or reuse scenarios (ADEME, 2002).
One of the aims of our study was to complete this methodology by investigating a new scenario: the impact of bottom ash reused as road embankment for lentic ecosystems. In this scenario, the road is built in a mountainous region. Wastes receive rain and run-off water leading to the production of leachates at the bottom of the embankment. Below the road-side, leachates run off on a grassland before reaching a lake coastline.
In most studies, the characterization of MSWI bottom ash and the assessment of its impact on the environment was made by using extraction procedures (batch, column or lysimeter). These approaches assess the possible release of pollutants by the solid waste and their possible transfers to aquatic and soil environments (Chandler et al., 1997; Meima and Comàns, 1999). In the last decade, single-species bioassays have been used to account for the biological impact of those materials (Ferrari et al., 1999; Lapa et al., 2002). Tests usually used are standard acute bioassays carried out with the bacterium Vibrio fisheri, the alga Pseudokirchneriella subcapitata and the water flea Daphnia magna.
We propose a new approach using a 2-l indoor microcosm assay. This freshwater microcosm protocol was proposed by Clément and Cadier (1998), validated by studying a copper-spiked artificial sediment and applied to the assessment of dredged materials (Clément et al., 2004). The microcosm consists of a glass beaker containing synthetic freshwater, a natural lacustrine sediment and pelagic and benthic organisms which could interact with each other and potentially influence, by their activities, the fate of pollutants and consequently effects on other organisms.
Materials and methods
MSWI bottom ash sampling and leachate production
Bottom ashes were collected in May 2001 on a weathering platform of a municipal solid waste incinerator situated in a mountainous region in France. The incinerator was a high capacity unit which can burn about 56 000 tons of domestic wastes per year. A laboratory lysimeter (1 m2, polyethylene) was used to produce bottom ash leachates. Two hundred kg of waste were deposited on a gravel bed and compacted in similar conditions to field road building conditions. Wastes were covered with sand to limit carbonation and to ensure a homogenous infiltration of demineralised water supplied day (5 l d−1). Rain water was not chosen because of its variability in composition and acidity. Leachates were collected at the bottom of the lysimeter and stored at 4°C. Assays were conducted on fractions which corresponded to different liquid/solid ratios: 0.5, 1.0 and 1.5, respectively. They were coded P1, P2 and P3.
Physico-chemical characterization of bottom ashes leachates
Leachate ion contents (Ca2+, Cl−, K+, Mg2+, NO3−, NH4+, PO43−) were determined using ionic chromatography (DIONEX, model DX-100, Sunnyvale, CA, USA) following standard procedures (AFNOR, 1999a). Metal contents (Cd, Cr, Cu, Ni, Pb, Zn) were determined by atomic absorption spectrometry (HITACHI, model Z-8 200, Tokyo, Japan) following AFNOR methodology (AFNOR, 1998a). Dissolved organic carbon (DOC) contents were analysed using an organic carbon analyser (PPM, model LABTOC®, Sevenoaks, UK) following standard methodology (AFNOR, 1997).
Four bioassays were conducted on P1, P2 and P3 leachates: bacteria bioluminescence inhibition test with V. fisheri (AFNOR, 1999b), algal growth inhibition test with P. subcapitata (AFNOR, 1998b), duckweed growth inhibition test with Lemna minor (AFNOR, 1996) and D. magna (neonates) immobilization test (ISO, 1989). Leachate pH was adjusted to 8.0 ± 0.5. Leachates were also tested with and without filtration (Whatman glass fiber, 1.2 μm, Clifton, CO, USA).
The microcosm consisted of glass cylindrical beaker (23 × 13 cm) containing a synthetic water (2 l) and a natural sediment (150 g). The synthetic freshwater was a modified OECD medium (OECD, 1993; Clément and Cadier, 1998) with moderate nutrient contents (50 μg P l−1, 654 μ g N l−1) and vitamin additions (75 μ g thiamine l−1, 1 μ g B12 l−1 and 0.75 μ g biotin l−1). The control sediment is a carbonated sediment, taken from Aiguebelette lake (Savoie, France).
On day (−7), natural sediment was sieved at 2 mm and 150 g were distributed in the glass beakers. One litre of modified OECD medium was gently added so as to limit sediment resuspension. In order to allow for particles to settle, the overlying water was aerated only after 24 h by gentle air bubbling (Pasteur pipettes) to keep the dissolved oxygen content close to the saturation level. The systems were left in these conditions for 7 d, during which a balance between water and sediment was reached. The temperature was kept at 20± 2°C.
On day (0) of the experiment, MSWI bottom ash leachates were diluted in modified OECD medium. One litre was added to the microcosms in order to obtain 3 final leachate concentrations: 1.56, 4.0 and 8.0%. Those concentrations were chosen based on single-species bioassays results. Nine replicates per treatment were carried out. The following organisms were introduced into the overlying water: 10 000 algal cells ml−1 of P. subcapitata, 3 colonies of two-frond L. minor, 10 neonates of D. magna (aged less than 24 h), 10 young Hyalella azteca (aged of 2 weeks on average) and 25 larvae of Chironomus riparius (aged 2 d after hatching). Algae and duckweeds were cultivated following AFNOR recommendations (AFNOR, 1996, 1998b). Daphnids, amphipods and chironomids were bred in as described by Clément et al. (2004). The microcosms were illuminated 16 h d−1 using 2000 lux daylight fluorescent tubes. Glass beakers were moved daily at random. Benthic organism food consisted of TetraMin; 10 mg d−1 was added to each system.
Temperature, conductivity, pH and dissolved oxygen were measured twice a week in the overlying water during the 30 d assay period. Table 1 summarizes measurement frequencies and methodologies used to monitor biological parameters.
On days 10, 21 and 30, 3 microcosms of each treatment were terminated in order to collect pelagic and benthic organisms and to carry out measurements on the abiotic components: water content, loss on ignition, sediment pH, sediment and pore water metal contents, overlying water and pore water ion contents and overlying water and pore water DOC content. Overlying water and pore water samples were filtered at 1.2 μ m (Whatman, glass fiber). Ion content, metal concentrations (Cd Cr, Cu, Ni, Pb, Zn) and DOC contents were analysed before and after filtration.
Data were analysed with Statview F-4.5. Significant differences between means were determined using Kruskal-Wallis and Mann-Whitney tests (p < 0.05).
Results and discussion
MSWI bottom ash leachate characterization
Leachate pHs were high, more than 10.0 units (Table 2) due to the presence of calcium hydroxides, Ca(OH)2, formed during the bottom ash extinction the incineration furnace (Meima et al., 2002). The pH decreased during leachate production because of the dissolution of calcium hydroxide and carbonatation processes (Meima et al., 2002).
Most of the DOC was extracted from the solid wastes at the beginning of the leaching procedure. The DOC content was high in P1 and decreased from P1 to P3 (Table 2). No organic pollutants were found in P1. Dissolved organic carbon was probably composed of humic and carboxylic acids and glycerol (Dugenest et al., 1999). This kind of organic material is biodegraded during bottom ash weathering (Dugenest et al., 1999) which could partly explain the DOC decrease observed during leachate production.
Leachate conductivities were high in relation to ion contents (Table 2) and decreased during leachate production. Most of the chlorides, sulphates, sodium and potassium salts were eluted during P1 production. Ions probably came from complete dissolution of alkaline ions (e.g., NaCl, CaCl2, KCl), gypsum (CaSO4, 2 H2O) and ettringite (3CaO, Al2O3, 3CaSO4, 32 H2O; Meima and Comans, 1999).
Metal content decreased from P1 to P3 (Table 2) except for Zn and Ni contents which were higher in P3. Chromium and Pb contents were low in all the leachates due to their low mobilities in MSWI bottom ash at pH near 10.0 (Meima and Comans, 1999). Cadmium was not detected. The high Cu concentrations measured are usual in that kind of liquid matrix (Chandler et al., 1997; Meima et al., 2002). Copper was probably eluted associated with DOC because of its high affinity for those elements (Meima and Comans, 1999).
Results of single-species bioassays
Leachate toxicity decreased during the leaching procedure (Table 3) in relation to ion and metal content decrease. The algae P. subcapitata was the most sensitive species followed by V. fisheri and L. minor. Other studies have found similar results with respect to the higher sensitivity of algae to MSWI bottom ash leachates (Ferrari et al., 1999; Lapa et al., 2002). In those studies, 72 h-EC50s ranged from 0.9 to 2.2% of leachates. Metals were probably responsible for this higher sensitivity (Ferrari et al., 1999).
Daphnids did not show signs of toxicity in terms of mobility and mortality to P1, P2 or P3, filtered or not (Table 3). This result is probably due to pH adjustment which modifies metal solubility and speciation (Chandler et al., 1997) and filtration, which induces pollutant removal. Lapa et al. (2002) reported a 48h EC50 of 7.4% of MSWI bottom ash leachate, without pH adjustment or filtration.
Physico-chemical conditions in microcosms assays
During the first hours of experimentation, overlying water pH values were higher in the contaminated microcosms. For example, initial pHs were 8.1 in controls, 9.0 at 8.0% in P1 and P2, and 8.5 in P3.This increase was linked to the concentration and the leachate tested. Within 72 to 96 h, pH values were similar to values in control systems owing to the buffer capacity of the systems. Therefore, leachate pH was not adjusted. After 20 d, overlying water pH progressively increased due to the photosynthetic activity of primary producers (algae and duckweeds).
Initial electrical conductivity of the overlying water was proportional to leachate concentration. For example, in P1 systems, conductivity values were: 290 μ S cm−1 in the control, 330 at 1.56% leachate, 400 at 4.0% leachate and 500 at 8.0% leachate. This increase was related to ionic content in leachates (chlorides, sulphates, sodium and potassium). Consequently, for each concentration, conductivity decreased from P1 to P3. During assays, conductivity progressively decreased partly due to ion assimilation by primary producers and partly due to ion precipitation and complexation. In pore waters, ion contents were similarly characterized.
In the overlying waters of controls, metals mean contents were: 4.4 μ g l−1 Cu, 0.2 μ g l−1 Cr, 0.9 μ g l−1 Ni and 3.8 μ g l−1 Pb. In the contaminated microcosms (P1, P2, P3), Cr, Ni and Pb contents were not different from the controls (Mann-Whitney, p > 0.05) whereas Cu contents were higher, even at the lowest concentration of 1.56% (Figure 1). Copper content decrease during assays was probably due to Cu binding on dissolved and particulate organic materials in water and sediment.
Dissolved organic carbon contents of overlying and pore waters were slightly higher in the contaminated microcosms than in the control systems. Concentrations were proportional to leachates concentrations. In the overlying water, the maximum was about 9.2 mg C l−1 at 8.0% of P1, after 10 d of experimentation, versus 5.3 mg C l−1 in the controls. In pore water, the maximum was about 102.5 mg C l−1 at 4.0% of P1, after 10 d of experimentation, versus 75.2 mg C l−1 in the controls. Dissolved organic content decrease during assays was probably related to the diffusion of organic materials from pore water to overlying water. In addition, DOC might have been degraded by micro-organisms in microcosms.
Impact on primary producers and zooplankton
From the beginning of the assays, MSWI bottom ash leachates had acute effects on D. magna survival (Mann-Whitney, p < 0.05). Their survival (Figure 2) was less than 10% in the following treatments: P1-4.0%, P1-8.0% and P2-8.0%. Those effects were proportional to leachates concentrations and decreased from P1 to P3. For most treatments (Pl, 1.56, 4.0 and 8.0%; P2, 4.0, 8.0%; and P3, 8.0%) after 10 d, Cu contents were higher (Figure 1) than the Cu 48 h EC50 for this species, which ranges between 25 and 32 Cu l−1 (Girling et al., 2000; Radix et al., 2000). Since Cu contents on day (0) were at least as high as on d 10, this metal was probably responsible for the effects observed on daphnid neonates. New organisms were introduced into the systems on d 3. Their survival was near 80% up to 20 d and then decreased progressively. At the highest concentration (8%), new organisms were introduced after 10 d of exposure in P1. The decline of food (algae) was probably responsible for daphnid mortality at the end of assays (Clément and Cadier, 1998).
Bottom ash leachates also had chronic effects on daphnids. The first broods were delayed at 8% in comparison to controls, 14 d against 10 d, on average (Mann-Whitney, p < 0.05). Number of neonates was significantly decreased at 4.0 and 8.0% of P2 and 8.0% of P3 (Mann-Whitney, p < 0.05; Figure 3). The 21d-NOEC is about 10 to 15 μ g Cu l−1 (Radix et al., 2000) which is below the overlying water concentrations of Cu in the following treatments, at the end of the assays: 25.9 μ g Cu l−1 for P2-4.0%; 47.1 μ g Cu l−1 for P2-8.0% and 19.4 μ g Cu l−1 for P3-8.0% (Figure 1). As a consequence, Cu was probably responsible for those chronic effects. Other metals were not expected to be toxic for daphnids at the concentrations measured in the water. Our results seem to be in contradiction with Ferrari et al. (1999) who demonstrated no effect of MSWI bottom ash leachates on daphnid reproduction in single-species tests. Daphnid exposure was probably higher in microcosms which could explain the observed effects. In microcosms, daphnids were exposed to pollutants through the overlying water but also by consuming algae on which pollutants where adsorbed and grazing on the sediment surface (Taylor et al., 1998).
Algae growth was impaired by leachates at the higher concentrations, 4.0 and 8.0%, of P1, P2 and P3. Figure 4 presents the example for the evolution of chlorophyll a contents in the overlying water during the P2 assay. The pollutants seem to have algistatic effects (Ward et al., 2002): algae densities were low in that treatment after 10 d of experimentation but rose after that period. Effects were probably due to Cu which could have direct effects on algae at the concentrations measured in systems (72-h EC50 for this species is about 47 μ g Cu l−1; Radix et al., 2000), and might also complex nitrates and phosphates which become unavailable for algal growth (Rai and Mallick, 1993).
The comparison of the evolution of algal growth and daphnid survival in controls underlies the trophic relationship between algae and daphnids. This relationship has been demonstrated by numerous studies (Taub, 1989; Rai and Mallick, 1993). Algae represent the unique source of food in microcosms for daphnids, the primary consumer in the 2-l microcosms. When daphnid survival and reproduction were low, algae were not controlled and risks of algal bloom increased. This is what was observed at 4.0 and 8.0% of P2 (Figure 4) and 8.0% of P3. When daphnid survival was high, they controlled algae population and then risks of algal bloom were negligible. This is what we observed in control systems (Figure 4) and at the concentration of 1.56% of P1, P2 (Figure 4) and P3.
Duckweeds were characterized by linear growth during the three assays in both control and contaminated microcosms. The maximum growth was about 120 fronds in controls. Growth was significantly different (Mann-Whitney, p < 0.05) from the control at the following treatments: P1, 4.0%; P2, 4.0%; P1, 8.0%; P2, 8.0%; and P3, 8.0%. Ion salts weren't responsible for those effects. Toxic concentrations of ions in bottom ash leachates were indeed much higher than those measured in overlying water of microcosms: 7d-EC50 of 500.0 mg NaCl l−1 (Buckley et al., 1996), 4d-IC50 of 930.0 mg Cl− l−1 (Wang, 1986) and 4d-IC50 of 1.0 g SO42− l−1 (Wang, 1986). However, Cu which exhibits herbicidal properties (Taub, 1989; Girling et al., 2000), was present in close to toxic concentrations for L. minor in microcosms: 14d-NOEC of 60 μ g Cu l−1 (Jenner et al., 1993), 20d-LOEC of 83 μ g Cu l−1 (Girling et al., 2000). In addition, some of the effects could be induced by the simultaneous presence of other potentially toxic metals.
Impact on benthic organisms
The amphipod survival in control systems was higher than 80% at the end of the three assays. At the highest concentration (8%), the leachate had lethal effects on H. azteca (Mann-Whitney, p < 0.05): 56.7% mean survival in P2 and 53.3% mean survival in P3. No survival and growth decrease were seen in the other treatments.
Lead, Ni and Cr contents in leachates were lower than known toxic concentrations for H. azteca. Mortality was, again, probably due to Cu content which was higher than accepted toxic values in overlying and pore water: 10d-LC50 of 43 μ g Cu l−1 (Kubitz et al., 1996), 28d-ERL of 5.3 μ g Cu l−1 and 28d-ERM of 19.8 μ g Cu l−1 (Ingersoll et al., 1996). As for duckweeds, effects may be partly due to the simultaneous presence of several pollutants.
The concentration of 8.0% leachates also caused significant mortality (Mann-Whitney, p < 0.05) to C. riparius larvae. Mortality was about 100% in P1,72% in P2, and 54% in P3. As well, at 4.0% of P2 leachate, survival was reduced by 23%. Survival and growth were not impaired in other treatments. Every surviving organism emerged. Copper toxicity levels were exceeded in pore water of microcosms in which effects were observed: 14d-ERL of 5.3 μ g Cu l−1 and 14d-ERM of 20.6 μ g Cu l−1 (Ingersoll et al., 1996). In pore waters, Ni, Pb and Cr contents were slightly lower than their toxic concentrations. These metals may have increased Cu effects by additivity or synergy.
Both approaches used in this study underlined the toxic potential of bottom ash leaches for lentic ecosystems. Biological effects were in agreement with the physico-chemical characteristics of leachates. The toxicity was proportional to leachate concentrations and decreased from P1 to P3 in relation to salt, metal and DOC content decreases. Leachate toxicity was mainly due to Cu which was present at higher concentrations than those causing toxic effects for most of the species tested.
Species sensitivities differed between bioassays and microcosm assays. Table 4 summarizes biological effects measured in microcosms. The alga P. subcapitata was more sensitive in bioassays. Daphnia magna was the most sensitive species in microcosms, followed by the algae P. subcapitata. Benthic organisms were less impaired. The results of both approaches are difficult to compare for algae sensitivity. Times of measurement are very different and the trophic relationship between daphnids and algae reduced the capability of microcosm to detect leachate impacts on algae at the lower concentration (1.56%). The differences of responses between bioassays and microcosms should not have been expected after the analysis of bioassay results. These differences could be partly explained by assay conditions. Leachates have been modified to respect standards (pH was adjusted and leachates were filtered) during bioassays but not during microcosm assays. D. magna and L. minor were, hence, less exposed to pollutants in bioassays. Furthermore, numerous routes of exposure are not represented in bioassays and this could explain differences in sensitivity. For example, in microcosms daphnids are exposed via pollutants in the overlying water, via pollutants sorbed on suspended organic matter, via ingestion of contaminated algae and via contacts with the sediment during grazing of settled-algae.
To assess risks of bottom ash reused as road embankment we considered an exceptional rain fall of 100 mm and used microcosm assay results. A ‘no effect concentration’ was not found in our study, but this concentration is probably close to 1.56% of leachates because, at this concentration, only daphnids were impaired at the beginning of the P1 assay. Probable no effect concentration (PNEC) is thus around 0.156% by applying a safety factor of 10, as usual for complex bioassays (Chapman et al., 1998). Our scenario assumes that the road embankment is built of 3600 tons of MSWI bottom ash (width: 2 m, length: 1 km, height: 1 m, MSWI bottom ash density, 1.8 t m−3). The rain fall leads to the production of 1700 m3 leachate, received by the coastal area of a lake of surface 0.1 km2 and depth 4 m, on average (400 000 m3). The predicted environmental concentration (PEC) is consequently 0.43% for this rain fall. The ratio PEC/PNEC is then about 3.0. Thus, risks for lentic ecosystems are not negligible (ratio > 1.0). Consequently, even if the scenario is maximized by comparison with real reuse conditions, recommendations could be made for this kind of situation. The MSWI bottom ashes could be weathered for several weeks before being used in road construction in order to stabilize most of the pollutants. The road embankment could be protected by a plant cover. Leachates that come from the road embankment could be collected in a basin. They could be partly treated before being discharged into aquatic ecosystem at a flow rate which could keep pollutant concentrations at non hazardous values.
These results indicate that the 2-l indoor microcosm assay has many advantages compared to bioassays in this kind of study. Leachates could be tested without any treatment as compared to single-species bioassays where standardisation dictates adjustment of the pH in many situations. The 2-l microcosm assay allows the simultaneous exposure of different species of different trophic levels. Organisms interact with each other and with the abiotic compartments and can, consequently, influence other species' responses. This was the case for daphnids and algae. The complexity of this assay also permits to assess pollutant distribution between abiotic and biotic compartments (sediment, pore water, overlying water and biota). Consequently, results obtained with the microcosm assays presented a higher degree of ecological relevance compared with those obtained with the single-species bioassays.
This work was supported by ADEME (French Agency for Environmental Management and Energy Conservation) and by the French Ministry of Public Works. The authors thank Marc Danjean, Martine Ghidini-Fatus and Thérèse Bastide for their technical and logistical assistance.