• No se han encontrado resultados

The design criteria recommended for acidity, metal and sulphate removal is based on the maximum loading rates at which there was not a significant decline in treatment performance (e.g. >≈98% metal removal and >≈99% acidity removal). This threshold was established during the second to last (week 16.0; 14.2≤θBGCR≤15.3) sampling event for acidity and metal removal from BGCRs containing mussel shells or a mixture of mussel shells and limestone (P2, S2, S3 and S4) as shown in Figures 4.11 and 4.12 where data points begin deviating from the 100% removal efficiency dashed line. Good treatment performance was achieved throughout for P1, but a decline in treatment performance was measured during the final sampling event (week 16.7). The decline in treatment performances occurred when increases in effluent DO (Figures D.7 and D.8) and Eh (Figure D.9) and declines in effluent pH and alkalinity (Figure 5.1) were measured.

Pilot and industrial-scale VFWs constructed in-situ have typically been trapezoidal-prism shaped basins (e.g. Gusek and Wildeman, 2002; Rose and Dietz, 2002; Ziemkiewicz et al., 2003; Rose, 2004; Skousen and Ziemkiewicz, 2005; Gusek et al., 2008; Reisman et al., 2008); therefore, it was most appropriate to assume the shape and treatment performance of trapezoidal-container reactors (S1-S4) for establishing design criteria rather than the cylindrical drums (P1-P3). It was also appropriate to assume that alkalinity amendments in the substrate compositions of future BGCR systems would consist of mussel shells and no limestone or NSD. Therefore, the most appropriate design criteria established based on results of this study were based on the results of BGCRs S2 and S3 (20 vol. % mussel shells) during the second to last (week 16.0) sampling event. This includes 66 g acidity (as CaCO3)/m2/day and 0.8 mol total metals/m3 substrate/day. Acidity and total metal removals averaged 99.1% and 98.4%, respectively. Better acidity and metal removal could potentially be obtained if the quantity of mussel shells in the substrate mixture was increased to 30 vol. %. Design criteria based on the treatment performance of P1 (30 vol. % mussel shells) in this study would be approximately 150 g

acidity (as CaCO3)/m2/day and 1.2 mol total metals/m3 substrate/day based on treatment performance during the final (week 16.7) sampling event. Acidity and total metal removals were 99.1% and 98.4%, respectively, the same as the average removals measured for S2 and S3 during the second to last (week 16.0) sampling event.

It should also be noted that BGCRs have a limited design life and treatment performance is dependent on a number of factors. Eventually, the reactive substrate media will be used up so treatment performance will eventually decline. The initial decline in treatment performance will likely occur due to the decomposition of the mussel shells. In a field application, such factors as variable AMD chemistry will influence treatment longevity. BGCRs should be designed to treat the greatest metal loading measured during monthly monitoring (such as that discussed in Chapter 2). Therefore, the BGCRs will nearly always be operating well below design criteria. Pilot-scale studies are recommended prior to designing and operating a full-scale system so that any short comings detected during field operation, such as less effective treatment performance, but not observed during laboratory treatability tests can be addressed (Gusek, 2002; Gusek, 2004; Wildeman et al., 2006). Pilot-scale studies and scaling-up challenges of BGCRs are discussed in detail in Chapter 7. Estimates pertaining to the design life and sizing of a full-scale BGCR treating Manchester Seep AMD in-situ

are discussed in Chapter 8 (Section 8.3).

A summary of design criteria and results of empirical studies evaluating the performance of VFWs and results of this study are summarised in Tables 4.9 (acidity removal) and 4.10 (metal removal). Overall, acidity and metal removal from the BGCRs containing mussel shells exceeded recommended design criteria used for similar VFWs employed in mine-water treatment incorporating limestone as the primary alkalinity amendment. Acidity design criteria was about two times greater in this study (66 g as CaCO3/m2/day) compared with design criteria recommended by Watzlaf et al. (2004) (25-30 g as CaCO3/m2/day) and Rose (2004) (35 g as CaCO3/m2/day). Wildeman et al. (2006) recommended design criterion of 0.3 mol total metal removal/m3 substrate/day for VFWs containing a mixture of organic materials and crushed limestone, which was approximately three times less than results of this study for S2 and S3 (>0.8 mol total metals/m3 substrate/day). The acidity design criteria recommended by Watzlaf et al. (2004) and the metal design criteria recommended by Wildeman et al. (2006) were comparable to those determined during this study for BGCR S1, which also incorporated limestone as an alkalinity amendment.

It may be plausible to design BGCRs to remove the majority of Al and rely on a subsequent aerobic treatment stage to remove Fe and residual Al in BGCR effluent. Sufficient effluent alkalinity and an adequate HRT in a subsequent pond and/or aerobic wetland would be essential for this to be feasibly implemented. Loading rates during the final (week 16.7) sampling event (average 132 g CaCO3/m2/day and 1.41 mol metals/m3/day for trapezoidal-shaped reactors and 155 g CaCO3/m2/day

and 1.24 mol metals/m3/day for drum-shaped reactors) are likely near the maximum loading threshold to confidently achieve this; however, this would not be plausible for S4. Effluent alkalinity from S4 was 5.15 mg/L as CaCO3, which is only sufficient to oxidise and precipitate 2.86 mg/L of Fe as a hydroxide based on stoichiometry discussed and empirically validated by Hedin (2008a). Average Al loading during the final (week 16.7) sampling event was 0.718 mol/m3/day.

Table 4.9: Design criteria established and empirical evaluations of acidity removal from VFWs treating AMD.

Study Description

Recommended Design Criteria and Empirical Study

Findings/Removal Rates (g as CaCO3/m2/day) Rose and Dietz (2002) Results from evaluation of 12 full-scale VFWs Removal rates between 25 and 50. Thomas and Romanek (2002a)

Laboratory-based column studies of VFWs incorporating 75 vol.% compost and 25 vol. % limestone

Average acidity removal was 87.8; average acidity feed rate was 57.8

Ziemkiewicz et al. (2003) Results from a comprehensive evaluation of RAPS Average acidity removal of 62.3 Proposed a non-Mn acidity design criterion of about 35 Rose (2004) Re-evaluation of earlier study (Rose and Dietz, 2002) Twice the acidity removal was

observe in systems

incorporating fine limestone Watzlaf et al (2004)

Evaluation of treatment performance of numerous full- scale VFWs

Design criteria of 25-30 recommended

>200 for two systems Five between 39 and 87 Eight between 2 and 17 Skousen and Ziemkiewicz

(2005)

Results from evaluation of 16 full-scale VFWs

One that did not remove acidity S1 (trapezoidal shaped; 12.5

vol.% limestone)

Design criterion established at 24

S2 and S3 (trapezoidal shaped; 20 vol. % mussel shells)

Design criterion established at 66

S4 (trapezoidal shaped; 12 vol. % mussel shells and 5 vol. % limestone)

Design criterion established at 66

P1 (drum shaped; 30 vol. % mussel shells)

Design criterion established at 153

This study

P2 (drum shaped; 12 vol. % mussel shells and 5 vol. % limestone)

Design criterion established at 80

Table 4.10: Design criteria established and empirical evaluations of metal removal from VFWs treating AMD.

Study Description

Recommended Design Criteria and Empirical Study

Findings/Removal Rates (mol/m3/day)

Wildeman et al. (2006) Based on empirical data of full-scale systems Recommended a design criterion of 0.3 S1 (trapezoidal shaped; 12.5

vol.% limestone)

Design criterion established at 0.3

S2 and S3 (trapezoidal shaped; 20 vol. % mussel shells)

Design criterion established at 0.8

S4 (trapezoidal shaped; 12 vol. % mussel shells and 5 vol. % limestone)

Design criterion established at 0.8

P1 (drum shaped; 30 vol. % mussel shells)

Design criterion established at 1.2

This study

P2 (drum shaped; 12 vol. % mussel shells and 5 vol. % limestone)

Design criterion established at 0.8

Iron and Al removal efficiencies reported from other VFW studies were variable and dependent on a number of factors (e.g. mine water chemistry, metal loading rates, system designs, environmental conditions, etc.). Rose and Dietz (2002) reported Fe removal efficiencies ranging from 14.3% to 96.8% (mean 65.2%; n=15 systems; Fe influent concentrations ranging from 16.0 to 208 mg/L (mean 69.6 mg/L) and Al removal efficiencies ranging from 25.0% to 100% (mean 59.0%; n=10 systems; Al influent concentrations ranging from 11.1 to 48.0 mg/L (mean 22.8 mg/L)) from full-scale VFWs. Gusek and Wildeman (2002) used crushed limestone in VFWs that achieved consistent Al concentrations <0.1 mg/L. Gusek et al. (2008) and Wildeman et al. (2006) demonstrated successful metal removal (average 96.1%) by a pilot-scale VFW treating AMD laden with Fe, Al, Cu, Zn, Cd, Co, Pb and Mn. Trumm et al. (2006; 2010) reported Fe and Al removal of 99% and 96%, respectively, from a small-scale VFW. Results of this study demonstrated substantial total metal removal (mean 99.0%) in VFWs incorporating mussel shells (P1, P2, S2, S3 and S4). Metal removal from BGCRs in this study (Figures 4.18 and 4.20) incorporating mussel shells (P1, P2, S2, S3 and S4) ranged from 94.2%-99.9% Fe and 99.5%-100% Al at loading rates up to approximately 0.8 mol total metals/m3/day. Effluent Fe concentrations were <1 mg/L at average total metal loading rates of 0.286 mol/m3/day and ranged from 2.28 to 5.34 mg/L for trapezoidal-shaped BGCRs S2, S3 and S4 up to loading rates of 0.827 mol total metals/m3/day. All systems incorporating mussel shells showed a significant decline in Fe removal at the highest metal and acidity loading rates tested with effluent concentrations 3.7-7.8 times greater than at the second highest metal and acidity loading rates. Effluent Al concentrations increased about an order of magnitude for all reactors at the second highest metal and acidity loading rates compared with lower loading rates (e.g. average of 0.036 mg/L increased to 0.210 mg/L). Aluminium concentrations increased an additional order of magnitude at the highest metal loading rates evaluated for S2 and S3 (3.10 and 2.09 mg/L, respectively) and an

additional two orders of magnitude for S4 (36.1 mg/L). Effluent Al concentrations for P1 and P2 were about two to three times greater at the highest metal and acidity loading rates (0.363 and 0.481 mg/L, respectively) compared with the second highest metal loading rates (0.122 and 0.246 mg/L, respectively).

The BGCRs operated during this study typically removed over 99% of Cu, Ni, Zn, Cd and Pb with effluent concentrations typically below laboratory PQLs. Effluent Cd concentrations remained below PQLs throughout this study for all BGCRs. Effluent Pb concentrations remained below practical quantitation limits (PQLs) except for the final sampling event (week 16.7) for S4 (0.0015 mg/L). Copper and Ni concentrations were below PQLs throughout the experiments for P1 and P2, but detectable at the highest metal loading rates tested for S2 (0.0015 mg/L Cu and 0.009 mg/L Ni), S3 (0.0007 mg/L Cu and 0.007 mg/L Ni) and S4 (0.0325 mg/L Cu and 0.084 mg/L Ni). Effluent Zn concentrations were relatively consistent throughout the experiments (median of 0.002 mg/L; range of <0.001-0.005 mg/L; Table 4.5) except at the highest loading rates tested for S2, S3 and S4 where Zn concentrations increased about an order of magnitude for S2 (0.04 mg/L) and S3 (0.025 mg/L) and about two orders of magnitude for S4 (0.628 mg/L). Reisman et al. (2008) reported successful treatment of Cu, Cd, Pb and Zn from a VFW with concentrations reduced by >96% on average but showed less success reducing Fe (average 1%) and Mn (average 12%) concentrations because of leaching from substrates. Figueroa et al. (2007) utilised a VFW as the second stage of a passive- treatment system and measured a reduction of Zn concentrations from a range of 45-55 mg/L to <0.1 mg/L. Trumm et al. (2006; 2010) reported 95% Ni and 99% Zn removal from a small-scale VFW.

Outline

Documento similar