Limestone-Based Hybrid Passive Treatment for Copper-Rich Acid Mine Drainage: From Laboratory to Field

Abstract

Acid mine drainage (AMD) is an environmental concern that needs to be addressed by some mining industries because of its high concentrations of metals and acidity that destroy affected ecosystems. Its formation typically persists beyond the operating life of a mine site. Its management is even more challenging for sites that are abandoned without rehabilitation. In this study, a legacy copper–gold mine located in Sto. Niño, Tublay, Benguet, Philippines, generating a copper- and manganese-rich AMD (Cu, maximum 17.2 mg/L; Mn, maximum 2.90 mg/L) at pH 4.59 (minimum) was investigated. With its remote location inhabited by the indigenous people local community (IPLC), a novel limestone-based hybrid passive treatment system that combines a limestone leach bed (LLB) and a controlled modular packed bed reactor (CMPB) has been developed from the laboratory and successfully deployed in the field while investigating the effective hydraulic retention time (HRT), particle size, and redox conditions (oxic and anoxic) in removing Cu and Mn and increasing pH. Laboratory-scale and pilot-scale systems using simulated and actual AMD, respectively, revealed that a 15 h HRT and both oxic and anoxic conditions were effective in treating the AMD. Considering these results and unsteady conditions of the stream in the legacy mine, a hybrid multi-stage limestone leach bed and packed bed were deployed having variable particle size (5 mm to 100 mm) and HRT. Regular monitoring of the system showed the effective removal of Cu (88.5%) and Mn (66.83%) as well as the increase of pH (6.26), addressing the threat of AMD in the area. Improvement of the lifespan of the system needs to be addressed, as issues of Cu-armoring were observed, resulting in reduced performance over time. Nonetheless, the study presents a novel technique in implementing passive treatment systems beyond the typical treatment trains reported in the literature.

1. Introduction

Legacy mines, also referred to as abandoned or inactive mines, are sites where mining activities have occurred previously without any reclamation or rehabilitation made [1]. It becomes a problem as water and oxygen react with exposed mined sulfidic minerals, such as pyrite and chalcopyrite, which generate acid mine drainage (AMD)—water streams that are acidic and contain elevated concentration of metals [2,3,4,5,6,7,8]. Without treatment, AMD threatens nearby ecosystems and human health [9,10,11], impacting beyond the operational life of the mine site, which can persist for hundreds of years [12,13].

Thousands of legacy mine sites have been identified globally, the majority of which are found in the United States, Canada, Australia, South Africa, and China [14,15]. In the Philippines, a legacy mine site in Sto. Niño, Brgy. Ambassador, Tublay, Benguet, inhabited by a community of indigenous peoples (IPs), was identified by the Mines and Geoscience Bureau (MGB) as a legacy mine. The copper–gold mine was abandoned in the 1980s and repurposed primarily for agriculture [1,12,15]. Daily agricultural and domestic water demands are mainly sourced from rainfall, springs, and surface waters. However, a copper (Cu)- and manganese (Mn)-rich AMD were observed in the area—a geochemistry that is rarely reported in the literature that threatens the nearby community [16,17].

Treatment strategies for AMD are either using managed active systems (requiring constant energy, manpower, chemicals, and high capital costs) or passive systems (relying on natural processes mediated by microorganisms and plants, or low-cost neutralizing agents), depending on the limitations of the site to be treated [18]. Given the challenges in legacy mines, such as in Sto. Niño, passive treatment strategies are deemed to be a more appropriate approach due to minimum maintenance and operational costs [19,20,21]. Multiple passive treatment systems that are mostly limestone-based, such as limestone leach beds [4], open limestone channels [22], and anoxic limestone drains [23], have been studied widely and have been deployed in different configurations. This includes being used in parallel or in series with wetlands and bioreactors as treatment trains, or being mixed with other treatment media [20,24,25,26]. Although the geochemistry of the AMD in these studies was mostly rich in Fe, the reported effectiveness of these treatment systems showed that they can greatly reduce metal concentration by 80% while increasing the pH [21,27]. However, long-term performance limitations, such as armoring [28,29,30], clogging [30,31], and passivation [32], arise in treatment systems deployed in Fe-rich AMD, which requires periodic maintenance or replacement of media or substrate. Ultimately, this affects the longevity and cost of a deployed treatment system over time.

To address the problem of the Cu- and Mn-rich AMD in the legacy mine in Sto. Niño, a limestone-based hybrid passive treatment co-designed with the IP local community (IPLC) was developed. The hybrid system is a combination of a limestone leach bed (LLB) and a controlled modular packed bed (CMPB) configuration. Considering the unique geochemistry of AMD identified in the site, the development of the treatment system was divided into three phases: (i) laboratory-scale setup, (ii) pilot-scale setup, and (iii) field-scale setup. IPLCs were involved throughout the development, especially in the deployment of the pilot- and field-scale setups, to ensure their cultural sensitivities were taken into consideration in the design. Limestone was chosen for the development of the passive treatment systems due to its (i) established effectiveness elsewhere in AMD treatment schemes, (ii) the local availability of limestone, and (iii) strong support from the IPLC during the consultation process.

The paper is structured as follows: Section 2 discusses the study area, i.e., the Sto. Niño Mines in Brgy. Ambassador, Tublay, Benguet, Philippines. It also discusses the methodology for each phase of the development, investigating multiple parameters, such as particle size, hydraulic retention time (HRT), and redox conditions, and how a combined process LLB and CMPB was designed and deployed at a field scale. Section 3 presents the results for each of the phases, detailing the geochemistry and performance of the treatment system. Related parameters, specifically pH, Cu, and Mn, were compared against environmental standards and regulations in the Philippines [33,34]. Section 4 provides the key findings, conclusions, and way forward on the developed hybrid passive treatment system.

2. Materials and Methods

2.1. Case Study: Identification of Acid Mine Drainage in Legacy Mine in Sto. Niño, Tublay, Benguet, and Development of the Limestone-Based Hybrid Passive Treatment System

The legacy copper mine located in Sto. Niño, Tublay, Benguet, Philippines, has been the study area of the Bio+Mine Project [12]. The project aims to develop a holistic, site-specific rehabilitation strategy that integrates local indigenous peoples’ knowledge and the technical expertise of academic research [35]. As part of the baseline characterization of the area, sampling campaigns were conducted in different bodies of water in terms of different metals, anions, and in situ parameters. The assessment of Balboa et al. [36] was used to identify a site with a copper-rich AMD, as shown in Figure 1.

The extent of the occurrence of AMD can be observed directly as blue staining of the rocks downstream, shown in Figure 2. This staining clearly indicates high concentration of Cu—an occurrence that is not widely reported for AMD except in Iran [37] and Brazil [38]. More typical AMD is macroscopically visible by way of red to orange stains and sediments due to elevated Fe concentrations [5,21,39].

Given the extensive farming activity in the area and limited water resources for domestic consumption, it is crucial that an intervention is implemented to treat the AMD and limit its contamination and spread beyond the study area.

Considering the remoteness of the area, limited manpower, cost, and cultural sensitivities, a limestone-based passive treatment system was deemed to be the most appropriate option to be implemented. This involved three phases: (i) laboratory-scale setup, (ii) pilot-scale setup, and (iii) field-scale setup. Free, prior, and informed consent were obtained from the local community before any engagement was done with the IPLC, as outlined by Alonzo et al. [12]. Limestone was deemed to be acceptable upon consultation with the IPLC due to its availability. The limestone used was then characterized via X-ray fluorescence spectroscopy (XRF; Horiba Mesa 50, Horiba Ltd., Kyoto, Japan) and X-ray diffraction (XRD; Shimadzu LabX-6000, Shimadzu Corporation, Kyoto, Japan) to provide a baseline for the treatment system and potential sources of additional contamination to the AMD. The methodology of each phase of development wherein the IPLC is involved is shown in Figure 3 and is discussed in the following sections.

2.2. Laboratory-Scale Setup

A column test was conducted to determine the effective HRT and particle size of the limestone-based passive treatment. Three HRTs were tested at 1 h, 8 h, and 15 h. Limestone used for the experiment was collected from a local supplier in Benguet, Philippines, at a particle size ranging from 25 mm to 50 mm. A schematic diagram of the setup is shown in Figure 4 below.

Based on multiple sampling campaigns conducted in the study area, the following geochemical properties summarized in

Table 1. Average and extreme values (i.e., lowest pH and/or highest metal concentrations) of geochemical properties of AMD identified in Sto. Niño, Tublay, Benguet, compared against environmental effluent standards in the Philippines.

Parameter	Average Values	Extreme Values	Effluent Limits [33,34]
pH	4.76 ± 0.16	4.57	6.00 to 9.50
Metals (mg/L)
Al	2.39 ± 1.04	3.64	-
Ca	71.22 ± 11.67	85.46	-
Cu	8.90 ± 2.53	10.75	1.00
Fe	0.11 ± 0.06	0.18	7.50
Mg	16.41 ± 6.61	24.04	-
Mn	1.53 ± 0.37	2.06	2.00
Ni	0.01 ± 0.11	0.23	1.00
Zn	0.75 ± 0.17	0.97	4.00
Si	-	17.94	-
SO42−	292.03 ± 173.37	333.00	550

were used as a basis for the preparation of the synthetic AMD to be used for the laboratory scale setup. Extreme values of each parameter, in terms of contamination potential, were used to prepare the synthetic AMD to make sure that the treatment system developed is robust, similar to the methodologies in the literature [20,40,41,42]. These values were also compared to current effluent regulations in the Philippines, specifically for the Class C waters, as defined in the Philippine Department of Environment and Natural Resources Administrative Order (DAO) 2016-08 and DAO 2021-19 [33,34].

The column experiment was conducted continuously for 35 days. Fresh simulated AMD was prepared each day to ensure that the desired concentration was maintained. Samples were collected at the completion of each HRT. In situ parameters, such as pH, ORP, and EC, were measured. At least 250 mL of samples for each HRT were collected and filtered using a 0.45 μm polyethersulfone (PES) syringe filter (Cobetter^®, SFMPES-4525, Hangzhou, China) and stored in an acid-washed polypropylene (PP) bottle (Kartell, 1627, Milan, Italy). The samples were acidified with concentrated nitric acid (ACS; 69.0%–70.0%, JTBaker^®, Thailand) for storage and preservation prior to the analysis of Cu and Mn concentration via atomic absorption spectroscopy (AAS) (Shimadzu, AA-6300, Kyoto, Japan). Methods outlined in the Standard Methods for the Examination of Water and Wastewater (SMEWW) were followed [43].

2.3. Pilot-Scale Setup

Considering the results from the column tests, a pilot-scale setup was deployed in the study area in Sto. Niño, Tublay, Benguet, to treat the actual AMD. Two redox conditions of the treatment were tested: anoxic and oxic. Oxic conditions were achieved by leaving the LLB exposed to the atmosphere. On the other hand, the anoxic environment was reached by laying a high density polyethylene (HDPE) liner on top of the LLB and covering it with soil compost. A schematic diagram of the design of the setup is illustrated in Figure 5 below.

Using an HRT of at least 15 h and a limestone particle size of ~5 mm, the treatment setup ran continuously for 100 days. In situ parameters, such as pH and conductivity, were measured every day, while samples for measuring the Cu and Mn concentrations were collected every 2 weeks. All of the samples were filtered using a 0.45 μm PES syringe filter (Cobetter^® SFMPES-4525) and stored in an acid-washed HDPE bottle and acidified with concentrated nitric acid (ACS; 69.0%–70.0%, JTBaker^®) before being analyzed for metal concentration via AAS, following the protocols outlined in the SMEWW [38].

2.4. Field-Scale Setup

Following the results from the laboratory-scale and pilot-scale setups, a field-scale treatment setup was designed for implementation in the study area. Multiple site assessments and surveys, such as flow rate and profile measurements, were conducted to be used as a basis for designing the treatment system. A 3D model of the site was generated through light detection and ranging (LiDAR) scans using an Apple iPad Pro 2022. Doing so allowed the analysis and decisions about the design for the project site to be done remotely. A snapshot of the 3D model is illustrated in Figure 6, illustrating the flow path of the AMD identified.

Considering all the parameters measured, an engineering design of the limestone-based hybrid passive treatment system was developed. The system has two treatment strategies: (1) limestone leach bed (LLB) treatment and (2) controlled modular packed bed (CMPB) treatment. The final and actual implementation of the treatment strategy is illustrated in Figure 7 below.

The LLB treatment involved creating temporary storage areas filled with limestone to increase the retention time. Limestone deployed in these storages has particle sizes ranging from 50 mm to 100 mm. Two areas were selected as storage areas, labeled Storage 1 and Storage 2. These areas have been identified to have suitable characteristics regarding their volume capacity and minimal seepages.

Simultaneously, a CMPB was installed at each of the storage areas. The CMPB is a tubular reactor packed with fine limestone particles with sizes ranging from 5 mm to 25 mm in diameter. The flow inside the packed bed is controlled by valves located at the end of the pipe. Compared to the LLB treatment, a smaller volume was treated, but the quality of the treatment was expected to be higher, as long effective HRT and higher surface area were implemented. A total of six (6) tubular packed bed reactors were installed, two (2) of which were laid out in Storage 1 and the other four (4) installed in Storage 2. Effective HRTs were measured for each of the systems installed. Limestones ranging from 50 mm to 100 mm were also laid down throughout the length of the stream to ensure treatment and contact with the AMD.

HRTs were found by first estimating the effective volume summarized in

Table 2. Effective volume of LLB and CMPB in the limestone-based passive treatment system.

Treatment System	Effective Volume (m3)
Storage 1
LLB	0.364
CMPB	0.0472
Storage 2
LLB	0.561
CMPB	0.0944

. Flow rates were measured in each sampling campaign to determine the actual HRT.

Aside from budget constraints, the site only allowed two storage areas where the LLB and CMPB were deployed without significant changes in the study area. This is in consideration of the cultural sensitivities of the IPLC living in the legacy mine. That is, the treatment system should involve minimal engineering work without masonry. Detailed information about the area and the specification of the deployed treatment system can be found in Supplementary Data S1, while the LiDAR scan of the deployed system can be found in Supplementary Data S2.

Seven (7) monitoring points were identified in the project site, representing the source, midstream, and effluent, as summarized in Figure 7. Different parameters were measured at each point to assess the performance of the treatment system. pH, conductivity, and ORP were measured in situ. Samples were also collected for analysis of metal concentrations following the SMEWW [43]. A summary of the parameters and equipment used is shown in

Table 3. Parameters monitored for the passive treatment system.

Parameter	Method of Analysis/Equipment Used
pH	Electrode; Hach, HQ4300 and PHC 101 Probe, Loveland, CO, USA
Conductivity	Electrode; Hach, HQ4300 and CDC 401 Probe, Loveland, CO, USA
ORP	Electrode; Hach, HQ4300 and MTC 101 Probe, Loveland, CO, USA
Metals (Cu, Mn)	Atomic Absorption Spectroscopy; Shimadzu, AA-6300, Kyoto, Japan

.

Among the parameters being measured, pH, Cu, and Mn were the most critical for monitoring, as these went beyond the standards based on previous sampling campaigns conducted in the study area. Samples were collected every three weeks, with five samplings being conducted in total.

Given the combined treatment strategy implemented in the area, the flow path of the water can be divided into two. A visualization of the hydraulic balance is illustrated in Figure 8.

From the source (1), the water flows along the stream and splits off between the LLB and the CMPB of Storage 1. The treated water exits the LLB at (3) and the CMPB at (2). Both effluents from the first LLB and CMPB combine as they flow along the stream and meet at (4). The flow splits again between the LLB and CMPB of Storage 2 and exits at (6) and (5), respectively. Both effluents meet and mix with the river from upstream at (7). Note that less than 0.5% of the total flow rate passes through the CMPB, making LLB the major treatment for the deployed system.

2.5. Performance Assessment

Performance assessment of the treatment system developed was quantified using the following equation:

$$\%Removal={\displaystyle \frac{x_i-x_f}{x_i}}\times 100$$

In this equation, x is the value of the parameter being assessed. The subscripts i and f are the initial and final values of the parameter, respectively.

3. Results and Discussion

3.1. Limestone Characterization

The XRD pattern of limestone used for the experiments in Figure 9 shows strong peaks of calcite (CaCO₃). XRF analysis in

Table 4. XRF semi-quantitative analysis of limestone sample.

Element	Concentration (wt %)
Ca	91.30
Fe	3.02
Si	3.01
Al	1.14
Cl	0.48
Ti	0.27
Others (Cu, K, Mn, Zn, Rh)	0.78
Total	100.00

also reveals that the sample is mostly composed of Ca—further supporting that it is mainly composed of calcite. Other studies, such as those by Pocaan et al. [20], Turingan et al. [41], and Fathy et al. [44], show other phases, such as aragonite (CaCO₃) and dolomite (CaMg(CO₃)₂), as a component of the limestone.

Although the only phase that is detected in the XRD is calcite, the presence of Fe, Al, Si, and trace amounts of Cu, Mn, and Zn suggests that other phases may be present in the limestone. These include aluminosilicates that may occur on the surface of the limestone or are amorphous in nature. Moreover, some of these elements were not included in the monitoring, as they were not part of regulatory standards for water quality, except for Cu and Mn.

3.2. Determination of HRT from the Laboratory-Scale Setup

Noticeable trends in AMD treatment were observed in three different HRTs during the monitoring of the laboratory-scale setup in terms of in situ parameters, as plotted in Figure 10. The treatment system had introduced sufficient alkalinity to the simulated AMD to increase its pH. Among the HRTs evaluated, the 15 h HRT demonstrated the most substantial increase in alkalinity and pH of 35 mg/L CaCO₃ and 7.17, respectively, on Day 3, which meets the effluent standards. Similarly, the 8 h HRT also reached its peak alkalinity and pH of 26 mg/L CaCO₃ and 6.49, respectively. However, the 1 h HRT was not able to reach the minimum pH level of 6.00 even at an extended period, which implies that the HRT is insufficient to treat the AMD.

The results plotted in Figure 11 suggest that dissolved Cu and Mn were removed by the system, decreasing its concentration by at most 97.6% and 36.4%, respectively. However, only the 15 h HRT was able to decrease the concentration below the effluent standards. On the other hand, the Mn concentration of the effluent increased starting at Day 7 for all HRTs, implying redissolution of Mn into the effluent. This coincides with the pH starting to decrease from a peak pH 7.0 to approximately pH 5.0 to pH 6.0, where most metals increase in solubility [46,47]. Such observation requires careful management, especially in its implementation in pilot- and field-scale treatment systems. That is, the pollutants that are initially immobilized by the system may redissolve and disperse again to the environment, defeating the purpose of implementing the treatment.

Previous studies have reported the difficulty of the removal of Mn in both neutral and acid mine drainages [48], specifically for Mn (II). It has been reported that the removal of Fe and other competing ions is required before Mn is removed from the system [21]. However, it is important to note that the geochemistry of the AMD in Sto. Niño is mainly Cu, which has not yet been widely reported.

Precipitates were observed to form in three HRTs, covering the surface of the limestone, as shown in Figure 12. Blue gel-like precipitates were observed to form on the surfaces of the limestone, indicating that the dissolved metals have precipitated as amorphous sulfate, carbonate, or hydroxides [2,49,50]. Possibly, the phase may be in the form of aurichalcite or hydrozincite, which are known to change color between green and blue when dried out.

Characterization of these gel-like precipitates via SEM-EDS (JEOL, JSM IT500HR/LA, Tokyo, Japan) in Figure 13 shows clumped spherical particles that were composed of Mn, Fe, Zn, and Cu. Given the observed increase in pH of AMD, a neutralization reaction likely occurred, which promotes the precipitation and co-precipitation of these metals. Similar observations were noted by a number of studies involving limestone as media for AMD treatment, suggesting the same mechanism involved in the immobilization of pollutants [20,25,40,41,42]. Webb and Sasowsky [51] also pointed out that calcite in the limestone is primarily dissolved into carbonic acid. Further reaction of the calcite with the H⁺ present in the AMD promotes the increase of the pH, making HCO₃⁻ (bicarbonate) ions the dominant species and providing alkalinity to the system.

Further characterization of the precipitates via XRF (

Table 5. XRF semi-quantitative analysis of precipitates.

Element	1 h HRT	8 h HRT	15 h HRT
	in wt%
Cu	32.5	42.2	49.5
Si	28.24	23.3	20.1
Al	21.0	16.2	12.7
Fe	5.80	8.97	10.8
S	5.03	2.52	1.46
Ca	4.14	3.97	2.13
Cl	1.83	1.32	0.78
Zn	0.65	1.20	1.95
Mn	0.07	0.12	0.20
Others	0.79	0.24	0.41

) revealed the presence of Si, Ca, Zn, and S, beyond the elements detected in the EDS, indicating that co-precipitation occurred as the primary mechanism. Characterization of the precipitates was also performed via XRD; however, no peaks were observed, suggesting its amorphous state.

3.3. Comparison of Anoxic and Oxic Conditions in a Pilot-Scale Treatment

Observing that the most effective HRT for the laboratory-scale treatment was 15 h, the setup was scaled up to a capacity of 60 L/day for a pilot test using this parameter. The particle size of the limestone used was decreased to a particle size ranging from 5 mm to 25 mm to increase the surface area of the media for neutralization of AMD. This was done to increase the operating life of the limestone treatment setup and improve the reactivity of the system to remove pollutants [52,53]. Additionally, an anoxic and oxic treatment system were tested to compare and determine its performance, especially for copper-rich AMD.

As illustrated in Figure 14, fluctuations in the concentration of Cu and Mn in the AMD were observed, reaching a maximum of 13.75 mg/L and 2.44 mg/L, respectively, which greatly exceed the Philippine General Effluent Standards (GES) set by the Department of Environment and Natural Resources (DENR) [33,34]. Moreover, pH of the AMD was consistently below the standards, reaching a minimum of 4.87. However, both anoxic and oxic systems were effective in decreasing the concentrations of Cu and Mn and increasing the pH to a peak of 7.50, thus within the GES throughout the observation period. Similar to the laboratory-scale setup, the precipitates formed in the system were also gel-like, suggesting that a similar mechanism occurred in the removal of Cu and Mn.

The oxic and anoxic conditions do not have any significant difference in terms of removal for the targeted parameters, indicating that the treatment system can be applied in either condition. This finding allows flexibility in designing the field-scale treatment implementation, making LLB or anoxic limestone drains a feasible strategy. Being a Cu-rich AMD, the literature suggests that it precipitates either in its oxide/hydroxide or sulfide form [54]. However, the limited concentration of SO₄²⁻ ions in the AMD, as observed through multiple samplings in the site, suggests that the precipitates formed were in their oxide/hydroxide form.

3.4. Implementation of Field-Scale System: Treatment Performance and Challenges

Using the data and observations from both the laboratory-scale and pilot-scale treatment setups, a field-scale treatment setup was designed and deployed in the area. Providing insights on (i) the effective particle size for the removal of Cu and Mn, particularly at 25–50 mm and less than 5 mm, (ii) the effective HRT at greater than 15 h, and (iii) similar treatment performance between oxic and anoxic conditions, the design was scaled up accordingly to treat the whole stream of the AMD identified in Sto. Niño mines. However, due to the challenging terrain, limited land area, and minimal manpower in the area, a combination of LLB and CMPB was deployed in the system to address the following issues identified in the study area:

• Limited land area—Achieving an HRT of greater than 15 h requires a large land area, which the site does not allow due to it being situated on a slope. Using a CMPB allowed for the stream to be stored longer and be in contact with the limestone.
• High flow rate of the stream—The AMD stream in the area can reach more than 3 L/s, implying that the system developed must handle at least 162 m³ to achieve the desired HRT. Thus, including an LLB in the treatment system allowed the stream to be in contact with the limestone and neutralize the AMD.
• Minimal access to manpower and energy—Given that the limestone required further processing to reach the desired particle size, both 50–100 mm and 5–25 mm limestones were used in the LLB and CMPB, respectively. Additionally, the implementation of the passive treatment allowed for less frequent maintenance compared to its active counterpart.

A detailed design of the hybrid passive treatment system can be found in Supplementary Data S1 and its LiDAR data in Supplementary Data S2.

The hybrid passive treatment system was observed over 78 days. This includes the number of days after the system was installed, weather conditions before the samples were collected, and the flow rate observed at the influent, as summarized in

Table 6. Weather conditions and observed flow rates during sampling in Sto. Niño, Tublay, Benguet.

Time, Days	Weather Condition Prior to Sampling	AMD Source/LLB (L/s)	Storage 1 CMPB (L/s)	Storage 2 CMPB (L/s)
0	Clear, no precipitation observed	2.43	-	-
15	Clear, no precipitation observed	3.07	9.78 × 10−3	9.02 × 10−3
39	Scattered rain	-	-	-
59	Heavy thunderstorm	3.69	5.81 × 10−3	16.17 × 10−3
78	Heavy thunderstorm	(overflowing)	8.22 × 10−3	30.31 × 10−3

. The treatment system dealt with conditions ranging from low flow to high flow. Thus, it was exposed to environments representative of the dry and wet seasons.

Measurements of the parameters are plotted in Figure 15, Figure 16 and Figure 17 for pH, Cu, and Mn, respectively. Each figure shows the actual measurements of the parameters across the treatment system over time. As the system involved a parallel treatment through an LLB and CMPB, each measurement was plotted separately to show any difference in performance.

Plotted in Figure 15 and Figure 16, Cu and Mn concentrations in the AMD showed variability in its concentration. Higher concentrations were observed at high flow rates compared to low flow rates. Additionally, the pH of the water, as plotted in Figure 17, also increased during high flow rates, reaching 5.0 from 4.5 during low flow rates. These observations coincided with the weather conditions observed prior to sampling, as summarized in Table 6, suggesting that increased precipitation worsens the condition of the AMD.

Given this situation, the treatment system needed to accommodate a highly variable quality of AMD. Figure 15 and Figure 16 show removal rates for Cu and Mn from the stream of 88.5% and 66.8%, respectively. Moreover, considering the Storage 1 and Storage 2 as separate systems, majority of the treatment occurred at Storage 1, with minimal reduction at Storage 2. Additionally, SP7 largely helped in the overall efficiency in the treatment of the AMD, diluting the contaminants further. However, the concentration at the effluent point was still beyond the GES, suggesting that further treatment must be deployed at the end of the system for polishing. Additionally, the difference in the CMPB treatment was that it performed significantly better than the LLB. The CMPB’s effectiveness was maintained during the first 15 days of deployment, with a minimal difference from the LLB thereafter, suggesting that the system needs replacement of the limestone.

A large dip in the concentration of Cu and Mn and increase in the pH was observed mostly at SP7 throughout the monitoring period. This implies that the treatment was largely driven by the dilution of the river upstream after several days of deployment and not by the hybrid passive treatment system. The limited lifespan of the system can be attributed to the Cu-armoring of the limestone, as observed in Figure 18. The composition of this blue stain was confirmed via XRF analysis (

Table 7. XRF semi-quantitative analysis of blue precipitates on limestone after 78 days of deployment.

Element	wt%
Cu	65.43
Si	18.43
Al	10.76
Ca	1.75
Fe	1.38
Zn	1.15
Others (S, K, Mn)	1.10
Total	100.00

) to be mainly composed of Cu, with the presence of other elements, such as Si, Al, Ca, Fe, Mn, S, and Zn—similar to the composition of the precipitates observed in the laboratory-scale setup. This suggests a similar mechanism occurring in the field-scale system except the for the redissolution of Mn successfully immobilizing these metals. However, this observed armoring decreases the effectiveness of the treatment system over time, which needs to be dealt with.

Given this limitation, maintenance of the system is required to ensure that it is able to treat and remove the pollutant of concern. Based on the literature, armoring may be resolved through flushing, increased flow rates, or other mechanical agitation to remove the precipitates on the surface of the limestone [21,29,55]. However, doing so would defeat the purpose of immobilizing Cu and Mn by spreading them in the environment. These mechanical agitation strategies may be done, but additional strategies, such as a constructed wetland with hyperaccumulators [56,57], must be added at the end of the treatment to deal with precipitates that will be flushed out.

4. Conclusions

This study presented the process of co-designing a passive treatment system with IPLCs, and the deployment of a limestone-based hybrid passive treatment system designed for a Cu- and Mn-rich AMD identified in a legacy mine in Sto. Niño, Barangay (Brgy.) Ambassador, Tublay, Benguet, Philippines. Laboratory- and pilot-scale treatment systems were conducted that identified an effective HRT of beyond 15 h, particle size ranging from 5 mm to 100 mm, and redox conditions of oxic and anoxic environments for the treatment of the Cu-rich AMD. A novel approach of combining an LLB and CMPB was deployed in the study area, which addressed the issues of limited land area, high flow rate of the AMD stream, and minimal access to manpower and energy. The field-scale passive treatment system decreased Cu and Mn concentrations by 88.5% and 66.8%, respectively, while increasing the pH to 6.26 to be within the GES.

Nonetheless, improvement of the treatment system is needed, especially in increasing the limited lifespan and minimizing Cu-armoring. This could be done by integrating local indigenous knowledge, such as using additional locally available neutralizing materials that can further improve the quality of AMD. Moreover, the system may be expanded by introducing additional treatment units, such as a constructed wetland, wherein local hyperaccumulators are used to uptake metals. The system may also be scaled up by capturing the entire stream, which can further mitigate the spread of AMD in the area.

The development process that the study presented for the limestone-based passive treatment system—from laboratory to field—provided necessary identification of operating parameters. The consistent and constant consultation with the IPLC at each phase of the development provided consideration and resolution of any social and cultural concerns in implementing new treatment technologies, especially when engineering works are involved. This study provides a blueprint for implementing passive treatment systems for AMD in legacy mines, wherein considerations go beyond technical and environmental concerns through engagement with local stakeholders.