3.1 Physical and mineralogical properties of bottom ash
All bottom ashes showed a generally well graded grain size distribution with a d50 value of 4 mm (min. 2, max. 7 mm). Depending on treatment and metal recovery strategy, the maximum grain size ranged between 15 - 35 mm, whereby large grains were mostly identified as metallic residuals (Table 1).
The water content of deposited bottom ash is mainly controlled by weather conditions and showed seasonal variation with a mean value of 17.2 wt.% relative to the dry mass. The saturated water content of all bottom ashes showed a mean of 34.3 wt.% relative to the dry mass. Dry density of field samples was on average 1.50 Mg/m3 with a mean porosity of 50 vol.% (Table 1).
Higher compaction densities compared to the field tests were achieved with the proctor experiments with dry density ranging from 1.47 Mg/m3 up to 1.70 Mg/m3 with variable water contents. Highest compaction was reached by adding 11.4 wt.% (relative to the dry mass) to the bottom ash material. The laboratory tests showed mean porosity values of 32 vol.% and Ksat values of 7.3·10-7 m/s to 2.6·10-4 m/s (Table 1).
Mineralogical composition of the studied bottom ashes is reported in the appendix (Supplementary Table 2). Samples I – III showed high fractions of amorphous phases (61 - 75 wt.%). Minerals of the melilite group (i.e., akermanite and gehlenite) represented the main mineral phase with 9.0 - 9.7 wt.%. The SiO2-phases (i.e., quartz and cristobalite) together accounted for 6.0 - 8.2 wt.%. Augite and diopside were selected as representatives for pyroxenes, which showed together weight fractions ranging between 4.0 wt.% and 4.8 wt.%. Fe-oxides (i.e., magnetite, wuestite and magnesioferrite, up to 3.9 wt.%), carbonates (i.e., calcite, siderite, magnesite and vaterite, up to 4.8 wt.%) and feldspars (i.e., anorthite albite and orthoclase, up to 1.7 wt.%) were present in smaller weight fractions. Alteration products of iron bearing phases (i.e., goethite and hematite), as well as phases formed by water uptake of anhydrite (i.e., bassanite), were identified in larger fractions in wet extracted bottom ash samples (i.e., bottom ash of landfill I) and mixed samples which were dominated by wet extracted bottom ash (i.e., bottom ash of landfill II). Free metals (i.e., aluminium, copper, and iron) were detected in all samples but never showed values above 1 wt.%.
Dry extracted bottom ash sample (sample IV) allowed identification of the mineralogy of bottom ash directly after incineration (i.e., no water applied to the sample). Comparison of this sample with the samples I - III allowed identification of early mineral reactions, which were triggered by contact with water (during quenching after incineration for generation of wet extracted bottom ash or during wetting of dry extracted bottom ash for transportation to the landfill). Sample IV showed a lower amount of amorphous fraction (50.3 wt.%). SiO2-phases were dominant in the sample with a weight fraction of 15.4 wt.%, followed by carbonates (8.4 wt.%) and melilite (7.7 wt.%). Other silicates, such as wollastonite and perovskite were present in smaller weight fractions with various amounts. Compared to the landfill bottom ashes (samples I - III), clinker phases such as alite and belite were present in higher proportions. Furthermore, lime and portlandite were detected in the dry extracted bottom ash sample in small fractions (below 1 wt.%), while no landfill sample showed detectable weight fractions of these phases. Furthermore, salts (halite and sylvite, 1.2 wt.%) and anhydrite (1.9 wt.%) were found in quantitative weight fractions in the dry extracted bottom ash sample compared to the in the landfill bottom ash samples (I - III) (Supplementary Table 2).
3.2 Visualisation of landfill structure and water flow
A vertical cross section sketch of the upper 2.5 m of a bottom ash landfill is shown in Figure 1. It illustrates the internal structure of the bottom ash landfill and the typical patterns observed after dye infiltration (Supplementary Figure 1). In general, the landfill showed horizontally orientated layering due to the compaction with the roller. Grain destruction led to generation of finer grained layers, which represented top layers of a disposal sequence (Figure 1, Supplementary Figure 1). Horizontal textures were observed in the finer grained layers, which typically showed a thickness of 5 – 15 cm (Figure 1, Supplementary Figure 1). The finer grained sections tend to form hard, compact structures. The effect of compaction decreased with increasing depth, resulting in a gradual transformation to coarse-grained bottom ash layers. Upper lying, younger disposal sequences were separated with a sharp border. These underlying layers with grainsizes > 10 mm were not affected by compaction and maintained the internal structure initially formed during disposal. A reduced stability of the coarse-grained sections was noticed as material detached from these sections during the excavation process. With a thickness of 20 - 35 cm, the coarse-grained sequence showed a thickness more than twice the fine-grained sequence (Figure 1, Supplementary Figure 1). The larger grains formed a skeletal structure in the coarse-grained layers, with finer material filling up empty space between large grains and can easily be excavated due to decomposition of the skeletal structure.
3.3 Leachate composition during regular sampling
The regular sampling of the various leachates represented background concentration values that can be expected from bottom ash landfills (Table 2). With pH values ranging between 7.19 and 9.26, all leachate samples showed neutral to slightly alkaline pH conditions. EC, which acts as a proxy for solute content, showed also a variation across the different compartments, with values between 14.9 mS/cm and 28.4 mS/cm. Consequently, main ions (i.e., Na, K, Ca, Mg, Cl, SO4, HCO3 and NO3) showed large variations of concentrations as well. Figure 2 shows the relative concentration of major cations and anions (in % meq/L) of the investigated compartments (I, II1, II2, III1, III2, III3, and IV). Detailed concentration values of each compartment are listed in the appendix (Supplementary Table 3). Na and K, and often also Ca, represented the main cations in all leachate samples. The anionic composition varied with Cl and SO4 being present in various proportions to each other. During the regular sampling, it was noted that pH and EC increased with decreasing discharge during dry conditions (i.e., low precipitation season).
Potentially toxic elements (i.e., heavy metals) and other components with defined threshold values (e.g., DOC) (Swiss Confederation 2023a) only accounted for minor fractions of solute content. However, these elements are of interest in the context of limit assessment for discharge into sewage and later near surface systems. pH, Cu, Mo, and DOC were identified as critical parameters that potentially exceed the threshold values (Table 2). For instance, leachate samples of landfill IV showed elevated concentrations for all these parameters (Supplementary Table 3). Other trace metals such as Al, Cr, Fe, Pb, Sb, V and Zn showed concentrations below 1 mg/L (Al and Fe), 0.5 mg/L (Zn) and 0.2 mg/L (Cr, Pb, Sb and V), respectively. In case of Cr, Pb and Zn, all current restrictions were fulfilled (Table 2).
3.4 Leachate composition during heavy precipitation events
Figure 3 shows plots of Q, pH, EC, and selected element concentrations of leachate samples, which were taken during heavy precipitation events in December 2021 (event 1, 25.2 mm / 23 hours) and September 2022 (event 2, 55 mm / 38 hours) on landfill I. Table 4 shows the comparison of samples collected during both events at maximum discharge and concentrations from continuous monitoring at landfill I.
During both events, pH of the leachate showed a relative increase (initial pH: 7.39 (1) / 8.12 (2), max pH: 7.90 (1) / 8.76 (2)) with increasing discharge (initial Q: 13.4 L/min (1) / 14.9 L/min (2), max Q: 82.5 L/min (1) / 824 L/min (2)). The EC value decreased relatively strongly in both sampling series (initial EC;16.7 mS/cm (1) / 20.4 mS/cm (2), max EC: 10.0 mS/cm (1) / 6.35 mS/cm (2)), which is also visible in concentration values of main ions such as Na and Cl decreasing with increasing discharge. Mg showed different trends and remained stable during event 1, while a decrease was observed during event 2. Decreasing concentrations trends similar to those of the main ions are observed for Mo and B. DOC on the other hand, showed no significant effect, but remained stable. For Al, Fe, and V, increasing concentrations were noted during both events. In the case of Fe, concentration values above detection limit were measured specifically during the discharge increase, while no Fe was detectable prior to the precipitation events. During event 1, Cu and Cr concentrations showed no notable change, while Zn and Sb concentrations decreased, with Zn decreasing below the detection limit (Figure 3). On the other hand, during event 2, concentration values of these ions showed a clear increase with increasing discharge (Figure 3). Pb, which was normally below detection limit, was measured during the heavy precipitation event in September 2022.
3.5 Simulation of leachate generation in the landfill body
The samples leachates at each landfill represent results of individual interaction of water and bottom ash and is strongly controlled by environmental conditions. To gain a deeper insight into the interaction times (i.e., residence time, which varies as a function of water flow velocity) between bottom ash and water, leaching experiments and speciation modelling were combined. The eluate tests simulate approximate equilibrium conditions (i.e., long residence time) between water and bottom ash and the results were verified with thermodynamic speciation calculations. Table 4 shows measured values of EC and pH and selected concentration values of the leaching experiments. Overall, the samples showed similar pH, EC, and concentration values. In comparison to the average leachate samples of landfill I (Table 3), pH of the leaching solutions showed high values, while EC is significantly lower. Similar to landfill leachates, Na, K, and Ca were identified as dominant cations, while Cl and SO4 represented the main anions in the solutions. In context of trace elements and DOC, measured concentrations generally showed similar values in the experimental solutions compared to the landfill leachates. Al and DOC were identified as the only exception, showing higher concentrations with values around 17.0 mg/L and 16.5 mg/L, respectively.
The speciation modelling of the eluate tests with PHREEQC resulted in an aqueous solution, of which the pH and concentration values of Na, Ca, Mg, Cl, SO4, Al, and Fe are shown in Table 4. Halite (NaCl) and anhydrite (CaSO4) were assumed as the main source of Na, Ca, Cl and SO4 in the aqueous solutions of both the landfill leachates and the experiments. Since both phases were not quantitatively detected with the XRPD (Supplementary Table 2), they were added to adjust the Na- and Ca-concentration of the model to the measured concentrations. Saturation indices (SI) and their residual amount present after the interaction with the liquid are shown in the appendix (Supplementary Table 4).
The pH conditions in the modelled solution showed lower values compared to the experimental solutions. While Na and Ca concentrations of the models matched the measured values, Cl was slightly higher in the modelled solution, while SO4 showed lower concentrations in respect to the experimental solutions. Al concentrations in the experimental solutions were up to 12 times higher than the thermodynamic model predictions for the given conditions.
During the interaction, the thermodynamic modelling indicated, that anhydrite and halite were completely dissolved. Anorthite, calcite, diopside, and magnetite were partly dissolved (Supplementary Table 4). Gibbsite, goethite, and quartz were identified as net precipitating phases, as indicated by a mass gain (positive delta values). Other secondary phases such as ettringite and gypsum showed negative SI values, which indicated that these phases remained undersaturated at the present conditions and consequently did not precipitate.