Evaluation of Land Use Adaptation by Sequential Extraction of Soil Trace Elements at an Abandoned Gold and Copper Refinery Site in Northern Taiwan

Abstract

This study site is located at an abandoned factory of mining, smelting, and refining of gold and copper in north Taiwan for more than one hundred years. The present study used soil background investigation out of the site and the sequential extraction procedures for arsenic and copper to assess the reutilization potential of brownfields at the site. The upper limit of background concentration out of the site was 300 mg/kg for arsenic and 700 mg/kg for copper. The soil arsenic within the site was mainly in the immobile fraction, such as forms fixed by layer silicates, that were very low risk for environmental releases. The soil copper in the abandoned sedimentation basin, gold refinery, and copper refinery was in the mobile fractions such as acid extractable, reducible, and oxidizable forms with higher release risk; therefore, except merely those three zones in the entire site with higher risk for environmental releases of copper, the release risks of trace elements are quite low in the rest of the areas, and land reuse without contact with soil or plant non-edible plants is possible. Therefore, in response to public demand for opening part of the site to promote local tourism development, appropriate control and isolation measures can be implemented to prevent the toxic elements from affecting human health through soil ingestion, skin contact, and other exposure pathways. In terms of pollution control, reducing dust inhalation is also an option to efficiently reduce health risks to an acceptable level and achieve the goal of sustainable land use at the contaminated site.

1. Introduction

Although soils have self-cleaning capabilities such as adsorption, oxidation/reduction, and precipitation, they may still cause environmental hazards in case of overloading. Soil contamination caused by harmful substances from industrial activities may lead to restrictions on the reuse of contaminated land due to damage to human health and lower environmental quality. It is difficult to consider the hazardous characteristics of contaminated sites and take appropriate site management or remediation actions with limited funds while taking into account public expectations [1,2,3]. For example, there are about 500,000 abandoned mines in the United States, of which about 140,000 are registered as hardrock mines, and it is estimated that about 22,500 of them contain substances that may cause long-term effects on human health and environmental ecology after exposure. From 2008 to 2017, U.S. authorities spent an average of USD 287 million per year to address the physical safety and environmental hazards of these abandoned mines, for a total of USD 2.9 billion, with billions more estimated to be spent in the future [4]. In Europe, it is estimated that a total of six billion Euros will be spent annually to treat contaminated soils, most notably in abandoned mines [5]. Canada is planning to spend 2.2 billion Canadian dollars over 15 years starting in 2020 just to treat eight of the largest and most complex contaminated sites in the Yukon and Northwest Territories [6].

Shuinandong, a place with special gold mining culture whose history started with the discovery of gold mines at this site in northern Taiwan in 1893, is located in the famous “Jinguashi” gold-silver-copper mining area, the largest gold mine in Southeast Asia. Gold mining started in 1896. After enargite mines were discovered, a dry refinery was built in 1904. Official agencies took over the Shuinandong mine selection field and refinery (Shuinandong smelter) (Figure 1) and operated it until it was shut down in 1981. The overall geological environment of the site was influenced by natural background factors, resulting in high concentrations of heavy metals in the soil, and later, due to the influence of the century-old mining operation, the soil at the site contained significantly higher concentrations of heavy metals. Approximately 30 hectares were identified as being above the soil pollution control standards. However, these control standards are based on the results of the total analysis. This site is not only a soil contamination site with trace elements such as arsenic and copper but also a cultural asset; the segmentation of the whole area is very clear with the space layout during the World War II period. The facility layout is top-down arranged, sequentially with spaces such as mining areas as well as other community service facilities, religious facilities, and elementary school infrastructures to form a special top-down stepwise layout of refinery culture and history. Their layout is clear and in good order. The local government has respectively announced that Shuinandong smelter and Benshan sixth tunnel entry and ropeway system as historical architectures while Liandong copper refinery waste flute is a cultural landscape. Although this area has been closed for more than 40 years, local people still hope that through the development of tourism, such as mining sites, architectural ruins, and natural environment, the area can be provided with a different look.

In 2000, Taiwan Environmental Protection Agency (TEPA) enacted the Soil and Groundwater Pollution Remediation Act (SGWPR Act) [7] and established soil pollution control standards in 2001 to address and resolve soil and groundwater pollution caused by illegal discharges of pollutants. The main management framework of the Act focuses on the total quantity control of pollutants to stop the discharge of pollution and to treat the contaminated land, meaning that the soil is considered contaminated when the concentration of pollutants exceeds the regulatory control standard value, which may affect human health and environmental safety, and requires remediation operations.

Taiwan’s soil pollution management framework is different from the risk assessment-based thinking of European and American countries. For example, the U.S. Environmental Protection Agency (USEPA) provides regional screening levels for chemical contaminants at superfund sites (RSLs) for reference [8], in which after the site is investigated and the concentration of contaminants reaches these RSLs, the contamination hazard should be evaluated, the contaminant type and soil characteristics of the site must be clarified, and the exposure pathways and acceptable risks of the receptors should be integrated according to the land use and function after land remediation, and appropriate contamination improvement targets should be reviewed and proposed through risk assessment.

However, for more than 20 years after the implementation of the SGWPR Act, Taiwan has adopted a fixed value of soil contamination control standard as the only administrative benchmark, which restricts the remediation technology and subsequent land use of soil contaminated sites, and causes land stakeholders to be reluctant to invest in the site in anticipation of the high remediation cost. In other words, if the exposure risk of the site is low or the bioavailability concentration of contaminants is not high, even if the total concentration of contaminants is high, then it is possible to develop and utilize the site appropriately. Since this site is one of the few sites in Taiwan that is affected by both geological background factors and century-old mining operations, if the remediation target is set using the soil pollution control standards set by the current SGWPR Act, i.e., the total concentration must be reduced below the control standards, it is assessed that it is not practically possible to excavate all the soil affected by natural background factors, and the cost required is up to approximately USD 200 million.

In the abandoned mining area in Central Spiš, Slovakia, the soil remediation was performed from 1997 to 2015, but the trace element levels in the mining soils still exceeded the regulatory limits [9]. This case infers that it is hard to meet the regulatory limits through soil remediation due to the strong fixation of trace elements with background concentrations by soil solid phases in the mining area [10]. Additionally, in the case of remediation for the copper refinery in Anaconda Smelter Site, Anaconda, Deer Lodge County, MT, USA [11], the target of remediation is to avoid the exposure risk caused by contaminated soils and refinery wastes inhaled and eaten by a human.

As indicated by Kabata-Pendias [12], in considering ecological and human risks, the mobility and stability of trace elements instead of their total concentration should be taken into account with higher priority. Namely, measurements of the total concentration of trace elements cannot determine the precise risk to the environment [13]. The potential risks of trace elements in mining soils critically depend on their fractionation in the different phases of soils. Only soluble, exchangeable, and chelated species of trace elements in the soils are the labile fractions for the release into the environment [14].

A sequential extraction procedure can assess the actual influence of trace elements upon the environment; this procedure has the potential to assess the mobility and bioavailability of trace elements based on the operational definition of chemical reagent extraction sequence, extraction ability, and decremental solubility. A sequential extraction procedure can extract trace elements sequentially based on their dissolution difficulty and thus assess their environmental impact. No matter whether Aqua regia digestion or 0.1N HCl extraction analysis are used to determine the contamination potential of heavy metals in soils, they are improper for the natural background field of enargite mining. Sequential extraction techniques are commonly used to evaluate trace element fractionation in solid phases of soil [15,16,17]. A typical procedure of sequential extraction includes progressively stronger solvents to sequentially solubilize various element fractions from water or weak-acid soluble forms to elements bound in layer silicate structures [18].

Therefore, by focusing on the sustainable management of the contaminated sites, this paper aims to offer a scientifically based solution to the issue of evaluating the utilization adaptation of the abandoned mining site with preservation and cultural asset revitalization planning through risk management methods. To this end, a heavily contaminated area in Taiwan, for which there is limited remediation technology available under regulation, is used as a case study. More specifically, the paper illustrates the methodology by using (a) regional soil investigation and background concentration statically analysis; and (b) sequential extraction to prove the real risk of this area and promote proper risk management actions of the risk management techniques in this area. To request flexibility from the government under current Taiwan’s soil pollution management framework and facilitate the economic benefits of the local tourism industry.

2. Materials and Methods

2.1. Soil Sampling

According to the operation procedure of the US Environmental Protection Administration Exclusive Superfund for the contaminated site, the study site was divided into 15 zones, and thus 148 sampling points, 277 samples of surface (0–15 cm), and subsurface (15–30 cm) soils were obtained (Figure 2). For the total analysis of arsenic and copper in the air-dried soil samples, wavelength-dispersive X-ray fluorescence (XRF; Vanta™ VMR, Olympus, Center Valley, PA, USA) was used, 117 sub-samples were extracted of trace elements soluble in the Aqua regia method, 15 sub-samples in each zone were tested with sequential extraction procedures using the Wenzel procedure for As [19] and BCR procedure for Cu [16].

In order to respond to the local interested parties, which expect to open the land reutilization of Shuinandong smelter and the path within the waste flute area from north and south tunnel to Cyuanji Temple, additional 15 soil samples were collected and extracted of trace elements soluble in the Aqua regia method to assess their risks for environmental releases of As and Cu for the reutilization of brownfield (Figure 3). These samples were also tested with sequential extraction procedures using the Wenzel procedure for As [19] and the BCR procedure for Cu [16].

In order to evaluate the natural background concentrations of this area, this study also conducted soil surveys in the communities surrounding the site, 67 soil samples located in the communities, schools, mountains, and forests around the site without obvious human disturbance were collected. All samples were also extracted of arsenic and copper concentrations in the Aqua regia method.

2.2. Sequential Extraction Procedures

According to the report by Wenzel et al. [19], arsenic in soil solid phases is divided into five phase states, which include: non-specifically sorbed As (F1), specifically sorbed As (F2), amorphous hydrous oxide–bound As (F3), crystalline hydrous oxide–bound As (F4), and residual As (silicate-lattice fixed, F5). The extracts were filtered with a Whatman No. 42 filter, and their concentrations of as were determined using a flame atomic absorption spectrophotometer (FAAS) (Hitachi Z-2300; Hitachi, Tokyo, Japan) that was equipped with a flow-injection hydride generator and NaBH₄ to generate arsenic hydride (AsH₃) for measurement. The AsH₃ was continuously purged by argon into a heated quartz cell mounted in the light path of the FAAS, where as absorption was measured.

Regarding copper fractionation, the Joint Research Center of technology research and development for each member state of EU proposed the standardized sequential extraction procedure (BCR sequential extraction procedure), which divided cation metal elements such as cadmium, chromium, copper, nickel, lead, and zinc into three states such as exchangeable fraction (acid soluble, F1), reducible fraction (iron/manganese oxide bound, F2) as well as bound state of organic matter and sulfide (F3) [16]. The original BCR sequential extraction procedure was only extracted to separate three phase states from F1 to F3; additionally, strong acid digestions including HF and HNO₃ were used to extract the residual fraction (silicate-lattice fixed, F4) for this work. The extracts of all fractions were filtered with Whatman No. 42 filter, and their concentrations of Cu were determined by the FAAS.

2.3. Statistical Calculation

According to the soil inspection result at the study site, background concentration threshold is respectively calculated with descriptive statistics through inspection of outliers. Under factors such as allowable 5% of extreme values or occurrence of uncertainties, background threshold value (BTV) [20] in the area where this site is located is calculated with values including 95th percentile, maximal value of non-outliers in the box, and whisker plot and upper percentiles by ProUCL software by US Environmental Protection Administration [21], upper tolerance limit (UTL), upper prediction limit (UPL) and upper simultaneous limit (USL). Maximal and minimal values were calculated as the distribution range of background threshold value.

3. Results and Discussion

3.1. Regional Soils Background

The source of trace elements in the soils at this site was the geogenic background in the mining area or sintering by-products of smelting. It is supposed that the trace elemental fraction existed as metallic oxide or as a primitive crystal lattice state in the soils. In particular, the fractional distribution and solubility of trace elements were all different, so the total amount of trace elements in soils is not equal to hazardous amounts in the environment and organisms [12].

Under an acceptable 5% variability, the descriptive statistics of trace element background concentration (mg/kg) out of the abandoned factory are listed in

Table 1. The descriptive statistics of trace element background concentration (mg/kg) out of the abandoned factory (n = 67).

Item	Arsenic	Copper
Minimum	7.36	15.9
Maxima	924	1660
Median	74.5	244
Arithmetic average	136	328
Standard deviation	171	370
Skewness	2.51	2.38
Coefficient of variation	1.25	1.12
First quartiles Q1	36.2	101
Second quartiles Q2	74.5	244
Third quartiles Q3	150	356
Upper limit (Q3 + 1.5 × (Q3 − Q1))	321	739
Quantity of outliers	8	6

. Moreover, the background concentration upper limit is about 300 mg/kg for arsenic and 700 mg/kg for copper, and it thus proved the enargite mine background of this site (

Table 2. The estimated upper limits of trace element background concentration (mg/kg) out of the site.

Item	Arsenic	Copper
Distribution type	Lognormal distributed	Gamma distributed
P90 1	168	445
P95 1	226	547
P99 1	390	775
Upper limit of tolerance 2	305	697
Predicted upper limit 3	233	576
Synchronization upper limit 4	682	1190

¹ Upper percentile method. ² 95% UTL with 95% Coverage. ³ 95% UPL. ⁴ 95% USL.

).

3.2. Sequential Extraction Procedure of Arsenic

The impact on the environment from arsenic can be ranked based on the extraction difficulty in the order of F1 > F2 > F3 > F4 > F5. The soils sample were analyzed with the Wenzel sequential extraction procedure for arsenic, as shown in

Table 3. Chemical fractions (mg/kg) of As in the studied soils within the site; recovery was obtained by sum of As in each fraction divided by the total As content (F1: non-specifically-bound, F2: specifically-bound, F3: amorphous hydrous oxide-bound, F4: crystalline hydrous oxide-bound, and F5: residual phases).

Code	Depth (cm)	Total As	F1	F2	F3	F4	F5	Recovery (%)
A06	100–150	1448 ± 106	0.16 ± 0.001	48 ± 0.05	137 ± 1.47	913 ± 6.84	303 ± 13.0	97
B06	0–50	442 ± 32.3	0.28 ± 0.002	43.8 ± 0.04	134 ± 1.43	121 ± 0.91	202 ± 8.69	113
C02	0–15	977 ± 68.4	1.64 ± 0.01	98.8 ± 0.09	276 ± 2.95	207 ± 1.55	347 ± 14.9	95
D01	0–50	2180 ± 153	0.19 ± 0.001	106 ± 0.10	259 ± 2.77	735 ± 5.51	980 ± 42.1	95
E03	0–50	325 ± 22.8	0.26 ± 0.002	3.91 ± 0.004	23.3 ± 0.25	19.3 ± 0.14	322 ± 13.8	113
F04	0–50	140 ± 9.80	0.22 ± 0.001	15.6 ± 0.01	45.3 ± 0.48	36.6 ± 0.27	24.5 ± 1.05	87
G04	0–50	446 ± 31.2	3.86 ± 0.02	43.4 ± 0.04	59.2 ± 0.63	105 ± 0.79	175 ± 7.53	87
H06	0–5	27,100 ± 813	350 ± 2.08	1109 ± 1.05	11,934 ± 128	2904 ± 21.8	9681 ± 416	96
I02	0–30	1250 ± 37.5	3.04 ± 0.02	90.5 ± 0.09	451 ± 4.82	180 ± 1.35	599 ± 25.8	106
J06	0–15	1430 ± 34.3	0.78 ± 0.01	43.5 ± 0.04	209 ± 2.24	490 ± 3.67	656 ± 28.2	98
K31	0–5	3640 ± 116	75.8 ± 0.45	539 ± 0.51	593 ± 6.34	1197 ± 8.97	799 ± 34.4	88
L1-01	0–50	6350 ± 540	93.5 ± 0.56	1196 ± 1.14	1843 ± 19.7	1841 ± 13.8	1231 ± 53.0	98
L2-11	0–50	8496 ± 93.5	33.8 ± 0.20	1048 ± 1.00	1874 ± 20.0	3204 ± 24.0	2140 ± 92.0	98
M1-04	0–10	672 ± 7.39	1.55 ± 0.01	37.3 ± 0.04	334 ± 3.57	200 ± 1.50	133 ± 5.72	105
M2-01	50–100	410 ± 4.51	0.54 ± 0.003	31.8 ± 0.03	214 ± 2.29	136 ± 1.02	90.3 ± 3.88	115

. The analysis results were mainly phased states of F4 and F5 with quite a low possibility of environmental releases. For the phase states of F1 or F2, which have a higher impact on the environment, the concentration distribution proportion of arsenic is low. It indicates that the risk for environmental releases of arsenic at this site is quite low.

3.3. Sequential Extraction Procedure of Copper

With the analysis results through the BCR procedure, it was found that only the phase states of copper at a few points, such as the original sedimentation basin for copper, the original gold refinery, and the copper refinery among 15 zones, have a higher risk for environmental releases, where the proportion of the phase states F1 and F2 are higher (

Table 4. Chemical fractions (mg/kg) of Cu in the studied soils within the site; recovery was obtained by sum of Cu in each fraction divided by the total Cu content (F1: acid extractable, F2: reducible, F3: oxidizable, and F4: residual).

Code	Depth (cm)	Total Cu	F1	F2	F3	F4	Recovery (%)
A10	100–200	5093 ± 193	805 ± 10.3	1450 ± 23.6	688 ± 10.5	1619 ± 18.9	90
B03	50–100	6591 ± 250	3320 ± 42.3	2920 ± 47.5	506 ± 7.73	700 ± 8.19	113
C03	0–50	272 ± 0.82	63.5 ± 0.81	10.8 ± 0.18	36.8 ± 0.56	126 ± 1.47	87
D02	0–50	8212 ± 24.6	11.8 ± 0.15	13.4 ± 0.22	366 ± 5.59	6813 ± 79.7	88
E04	0–30	497 ± 1.49	83.0 ± 1.06	116 ± 1.89	129 ± 1.97	239 ± 2.80	114
F02	0–50	254 ± 0.76	29.5 ± 0.38	99.0 ± 1.61	54.4 ± 0.83	104 ± 1.22	113
G02	0–30	3710 ± 11.1	1050 ± 13.4	1100 ± 17.9	700 ± 10.7	613 ± 7.17	93
H02	0–50	32,236 ± 612	5920 ± 75.4	10,250 ± 167	3625 ± 55.3	10,000 ± 117	92
I04	0–50	1051 ± 20.0	253 ± 3.22	7.90 ± 0.13	506 ± 7.73	159 ± 1.86	88
J08	0–50	5555 ± 16.7	730 ± 9.30	69.5 ± 1.13	1581 ± 24.1	2175 ± 25.4	82
K08	0–5	2944 ± 26.5	203 ± 2.59	980 ± 15.9	619 ± 9.45	1494 ± 17.5	112
L1-07	0–50	911 ± 4.56	108 ± 1.38	500 ± 8.13	329 ± 5.02	121 ± 1.42	116
L2-08	0–50	581 ± 2.91	41.5 ± 0.53	16.0 ± 0.26	147 ± 2.24	263 ± 3.08	80
M1-01	0–10	177 ± 0.89	17.8 ± 0.23	15.8 ± 0.26	85 ± 1.30	79.4 ± 0.93	112
M2-02	50–100	1040 ± 5.20	123 ± 1.57	148 ± 2.41	316 ± 4.82	334 ± 3.91	89

). Although the total concentration of copper at point D02 in the original flotation plant area exceeded 8000 mg/kg, 94% of them are stable residual states (F4). The total concentration of copper at point B3 in the copper sedimentation basin was 6591 mg/kg, and 80% of copper is distributed in the phase states F1 and F2, with a higher risk for environmental releases. At point G02 in the original gold refinery and point H02 in the original copper refinery, there were over 50% of copper distributed in phase states F1 and F2, with a higher risk for environmental releases.

3.4. Assessment for the Areas Which Interested Parties Expect to Open

In 2007, the governmental authorities asked the public to participate in the “Selection of private investment construction & operation plan for cableway and light trolley near the mining area”. In 2009, the selection of public tender for a turnkey project for the construction of abandoned lands was completed by the authorities. Due to the announcement of soil heavy metal contamination at this site, the government informed the winner tenderer to terminate the contract. Before the announcement of the soil contamination site, the interested company of the polluted land had already put in about USD 3,500,000 for the soil investigation, planning, and soil remediation in the Shuinandong smelter in 2002. After the announcement of the soil contamination site, the interested company has successively put in about USD 3,000,000 to implement site closure and cleaning of waste residues between 2014–2017. Currently, risk monitoring and inspection are continuing. In July 2018, the Environmental Protection Agency of Taiwan announced this site as a site for remediation due to high contaminations of arsenic and copper regardless of their bioavailability in August 2018. However, according to the implementation plan and continuous monitoring assessment results of interested companies from 2013 to 2021, no immediate hazard to humans and the environment existed at this site.

In 2017 after two interested companies closed this site according to the control plan approved, although the contact risk by people is avoided, people still expect that this site could be used for cultural tourism purposes. The communities, which mainly consist of old miners in the site nearby and local elderlies such as “Alumni of Guashan elementary school”, “Action coalition for resuming miner rail”, and “Action coalition for walking back the road”, continue to request for opening Shuinandong smelter, 13 levels smelter and the path from north and south tunnels to Cyuanji Temple in the waste flute area through the ways of the petition, on-site investigation, and speech. Whether it is feasible is the focus of this study.

According to the assessment by two interested companies of the contaminated land from 2013 to 2021, the current measure of the closing site can certainly achieve risk control performance. In order to meet the expectations of folk, this study further used the sequential extraction procedure for 15 soil samples in the areas requested. The Wenzel sequential extraction procedure for arsenic is summarized in

Table 5. Chemical fractions (mg/kg) of As in the studied soils on the trail which interested parties expect to open; recovery was obtained by sum of As in each fraction divided by the total As content (F1: non-specifically-bound, F2: specifically-bound, F3: amorphous hydrous oxide-bound, F4: crystalline hydrous oxide-bound, and F5: residual phases).

Code	Depth (cm)	Total As	F1	F2	F3	F4	F5	Recovery (%)
TS01	0–15	2520 ± 73.0	2.21 ± 0.03	5.60 ± 0.08	1052 ± 11.6	1118 ± 5.59	112 ± 4.76	91
TS02	0–15	1060 ± 30.7	1.96 ± 0.02	4.99 ± 0.07	332 ± 3.65	501 ± 2.51	53.4 ± 2.27	84
TS03	0–15	393 ± 11.4	0.84 ± 0.01	4.66 ± 0.07	122 ± 1.34	190 ± 0.95	22.5 ± 0.96	86
TS04	0–15	1420 ± 41.2	0.55 ± 0.01	10.8 ± 0.16	739 ± 8.13	579 ± 2.90	76.4 ± 3.25	99
TS05	0–15	498 ± 14.4	1.19 ± 0.01	0.95 ± 0.01	240 ± 2.64	172 ± 0.86	0.23 ± 0.01	83
TS06	0–15	1190 ± 34.5	20.0 ± 0.24	10.1 ± 0.15	150 ± 1.65	889 ± 4.45	64.1 ± 2.72	95
TS07	0–15	38.9 ± 1.13	5.75 ± 0.07	0.75 ± 0.01	4.63 ± 0.05	14.7 ± 0.07	5.08 ± 0.22	80
TS08	0–15	2560 ± 74.2	157 ± 1.85	103 ± 1.50	302 ± 3.32	1493 ± 7.47	213 ± 9.05	89
TS09	0–15	3740 ± 108	45.1 ± 0.53	98.4 ± 1.44	525 ± 5.78	1191 ± 5.96	1114 ± 47.3	80
TS10	0–15	2800 ± 81.2	1.00 ± 0.01	4.66 ± 0.07	144 ± 1.58	1011 ± 5.06	1201 ± 51.0	84
TS11	0–15	122 ± 3.54	0.16 ± 0.002	0.68 ± 0.01	0.61 ± 0.01	124 ± 0.62	0.91 ± 0.04	103
TS13	0–15	494 ± 14.3	4.38 ± 0.05	41.6 ± 0.61	115 ± 1.27	200 ± 0.98	148 ± 6.29	103
TS14	0–15	597 ± 17.3	3.10 ± 0.37	10.8 ± 0.16	242 ± 2.67	103 ± 0.51	175 ± 7.44	89
TS15	0–15	2150 ± 62.4	16.7 ± 0.20	13.9 ± 0.20	338 ± 3.72	1140 ± 5.70	201 ± 8.54	80

; in most cases, arsenic was mainly in phase states F3, F4, and F5, with a low risk for environmental releases. The proportion of the phase states F1 or F2, which have a higher risk for environmental release, is low. Although the concentration of arsenic in the sample TS07 has 18.6% of distribution in the state F1, however, the total concentration of arsenic was 31 mg/kg, which is relatively low compared to the national control standard for soil As (60 mg/kg) in Taiwan. However, the total concentration of As in the sample TS08 on the path between two tunnels was 2560 mg/kg, but only 6.9% of total As existed in the F1. The risk of releasing arsenic into the environment from the rest of the samples is low (Table 5).

The fractionation of soil copper varied greatly in the 15 soil samples in the areas requested (

Table 6. Chemical fractions (mg/kg) of Cu in the studied soils on the trail which interested parties expect to open; recovery was obtained by sum of Cu in each fraction divided by the total Cu content (F1: acid extractable, F2: reducible, F3: oxidizable, and F4: residual).

Code	Depth (cm)	Total Cu	F1	F2	F3	F4	Recovery (%)
TS01	0–15	2820 ± 22.6	700 ± 8.89	625 ± 10.1	1188 ± 16.9	434 ± 3.91	104
TS02	0–15	2520 ± 20.2	625 ± 7.94	1080 ± 17.5	585 ± 8.31	244 ± 2.20	101
TS03	0–15	98.7 ± 0.79	0.50 ± 0.01	ND	64.4 ± 0.91	35.6 ± 0.32	102
TS04	0–15	1240 ± 9.92	186 ± 2.36	327 ± 5.30	493 ± 7.00	169 ± 1.52	95
TS05	0–15	888 ± 7.10	170 ± 2.16	195 ± 3.16	386 ± 5.48	124 ± 1.12	99
TS06	0–15	1400 ± 11.2	357 ± 4.53	520 ± 8.42	308 ± 4.37	214 ± 1.93	100
TS07	0–15	81.6 ± 0.65	8.50 ± 0.11	12.3 ± 0.20	39.9 ± 0.57	15.6 ± 0.14	94
TS08	0–15	1270 ± 10.2	155 ± 1.97	214 ± 3.47	493 ± 7.00	410 ± 3.69	100
TS09	0–15	1110 ± 8.88	21.8 ± 0.28	96.5 ± 1.56	491 ± 6.97	453 ± 4.08	96
TS10	0–15	1800 ± 14.4	71.5 ± 0.91	74.0 ± 1.20	819 ± 11.6	838 ± 7.54	100
TS11	0–15	233 ± 1.86	11.5 ± 0.15	39.8 ± 0.64	135 ± 1.92	48.0 ± 0.43	101
TS13	0–15	663 ± 5.30	11.7 ± 0.15	26.3 ± 0.43	394 ± 5.59	228 ± 2.05	100
TS14	0–15	489 ± 3.91	8.70 ± 0.11	8.70 ± 0.14	299 ± 4.25	164 ± 1.48	98
TS15	0–15	28,100 ± 225	15,700 ± 199	1360 ± 22.0	10,000 ± 142	731 ± 6.58	99

). For example, the samples TS03 on the roadside to area L2, TS08 on the path between two tunnels, TS09-TS11 in the south tunnel, TS13, and TS14 to Benshan sixth tunnel slope ropeway trail where over 70% of copper in the soils are in phase states of organic matter bound state (F3) and residual state (F4) with low risk for environmental releases. However, regarding the samples TS02 on the path from area J to area K, TS06 in the north tunnel, and TS15 at the stair from area G to area H, there were more than 60% of the copper phase states are in phase states F1 and F2 with high risk for environmental releases. The total concentration of copper in the sample TS15 was 28,100 mg/kg, and the sum of the phase states F1 and F2 was 61.4% distributed in phase states with a high risk for environmental releases.

3.5. Depiction of Soil Contaminant Risk in the Site

This site is located in a mining area, and Table 1 shows that the entire area should be influenced by the natural geological background of the mining area, resulting in high background concentrations of geochemical elements in the soil. In addition, the study site and its surrounding areas have been subject to gold and copper processing and smelting operations for nearly 100 years. Therefore, the soil concentrations of arsenic and copper (Table 2 and Table 3) obtained from the site investigation include both the “natural background concentrations” of the original geological environment and the “external human-induced concentrations” of various degrees due to the mining operations.

In the past, Taiwan has assessed the extent of contamination at pollution sites and set “Soil Pollution Remediation Goals”, by the Soil and Groundwater Pollution Remediation Act (SGWPR Act) [7], which uses the rigid “Soil Pollution Control Standards”, as a benchmark for assessment. In principle, any soil that exceeds the control standard is considered polluted soil and needs to be treated.

However, according to the 67 soil sampling locations outside the copper refinery that were not affected by the operation of the former copper refinery, including the community, schools, and forests near the site, the upper background concentration limits for soil arsenic and copper were about 300 and 700 mg/kg, respectively (Table 2), which are significantly higher than the current soil pollution control standards of 60 and 400 mg/kg for soil arsenic and copper in Taiwan. In this area, the natural background concentrations of arsenic and copper in the soil are generally higher than the soil pollution control standards, and it would be difficult to enforce them in practice if the current regulatory standards are used as the only assessment and management benchmark. Therefore, for the sustainable use of land and to ensure that the level of contamination at the site can be reasonably identified, this environmental background characteristic should be taken into consideration when assessing the level of contamination at the site or setting remediation and cleanup targets, and subsequent remediation operations should be conducted regarding the risk results.

Although the Taiwan Environmental Protection Agency (TEPA) currently stipulates in the SGWPR Act that if a contaminated site cannot be remediated to a concentration of pollutants below the soil pollution control standard due to geological conditions, a separate soil pollution remediation target can be proposed based on scientific health risk assessment results, and the contaminants at the site can be controlled to a level that will not cause human health hazards through engineering or administrative control and other risk management methods. This can be performed. However, in practice, due to the lack of supporting laws and regulations or administrative rules, the procedures are more complicated and lengthy, the assessment process has more variables (it is impossible to predict whether it will be passed smoothly), the assessment process may be more expensive than direct remediation operations, and there is pressure from outside public opinion (e.g., it is easy to question why it is allowed to set remediation targets higher than the control standards, and the public’s distrust of the assessment process, etc.). As a result, in the 20 years since the law was passed and implemented, there have been no successful cases of setting separate remediation targets using the risk assessment method. In order to accelerate the reuse of contaminated sites, TEPA has been accelerating the reuse of contaminated sites regarding the “brownfield” management approach, considering ways to allow contaminants to remain but control them through engineering means that may actually endanger human health or environmental exposure to reduce the degree of harmful effects of contaminants, which can significantly reduce the cost of cleaning up contaminants and accelerate the development and use of contaminated sites.

The feasibility assessment of the reuse of the contaminated brownfield site relies on the results of health risk assessment, but health risk assessment is mostly based on the total concentration of soil pollutants. In this study, the concentration distribution of different phases of contaminants was determined by sequential extraction (Table 3 and Table 4), which was used to assess the proportion of the actual impact on human health hazards, which can reasonably reflect the actual impact of contamination in the site. For example, except for a few sampling sites, the concentrations of F1 and F2 phases of arsenic and copper in the soil at most of the sampling sites accounted for less than 20% of the total concentrations, which can significantly reduce the problem that the current health risk assessment results are too conservative and make it easy to communicate with the residents and stakeholders around the site. The analysis of sequential extraction can classify the site into different levels of pollution hazards, providing decision-makers with a basis for subsequent risk control and priority for pollution remediation operations.

Based on the risk assessment results of this site and the principle of sustainable land use, the site can be opened to the public for recreation and leisure by using engineering methods to cover and green the contaminated soil and setting up a buffer zone to ensure that the public will not come into direct contact with the soil.

3.6. Risk Management Strategy of the Site

After an overall assessment, if we want to take into account the interests and needs of all parties and use limited administrative resources, and consider the complexity of the causes and distribution of contamination at the site, the management strategy should be to reduce the risk of the subsequent land use of the site and to propose an overall risk management strategy based on the differences in the degree of contamination in different areas of the site. The main risk management objectives are: to control pollution by blocking exposure pathways, such as soil ingestion and skin contact, and to reduce exposure channels of dust inhalation as much as possible. In order to reduce the health risk value to an acceptable level through risk management measures. The principles of risk management measures to be adopted are as follows.

Pollution source blocking: For substances with abnormally high pollutant concentrations or with high environmental release phases, it is recommended to remove a small amount of highly hazardous pollutants or to completely block them by applying cement paving on the surface to block the risk of pollution sources.

Pollution transmission pathway blocking: In areas within the site that have not yet been assessed for land development and utilization needs, the surface should be covered with 30 cm of clean soil to block the contaminated soil and avoid exposure to soil pollutants. The clean soil can be covered with straw mats or vegetation blankets after construction, and grass seeds can be sown to prevent soil erosion by rainwater so that a good vegetation stabilization cover can be formed after the natural succession of local herbs or woody plants. Plants with heavy metal tolerance can also be planted in consideration of local environmental suitability.

Receptor protection: If the site is to be opened as a walkway for the public to visit, it is recommended to limit the walking space for people by laying floor tiles to block bare soil or using elevated walkways off the ground to block contact with bare soil, and by installing railings and hedges as buffer zones to prevent people from entering the adjacent areas. For the tunnel section, structural safety, crack repair, water infiltration, and drainage facilities must be considered, along with the implementation of pollutant containment facilities and regular monitoring of tunnel safety.

4. Conclusions

Current soil conservation policies and legislation in Taiwan are based on a single concentration limit of contamination. However, it is hard to remediate all contaminated soils only by this limit. Alternatively, risk-based approaches may be a more useful decision-making tool to evaluate potential hazards that reflect on other factors (e.g., social, economic, political). This case provides a comprehensive approach applicable to national regulations guiding sustainable reclamation of post-mining sites. A scientific risk-based approach with background concentration identification and sequential extraction method to assess the actual hazard composition of the soil may be a good assessment tool in addition to the total concentration that has been the only basis for decision making. Moreover, it highlighted and prioritized local conditions and needs, such as cultural assets preservation and well-developed local tourism industry impact. It has been confirmed that the site is located in an area that has been influenced by natural background factors in the mining area. According to the sequential extraction analysis, appropriate control and isolation measures should be taken to reduce potential exposure hazards of copper when implementing future land reuse plans for the site. Summarily, this study offered a successful case to highlight the brownfield development approach by soil sequential extraction techniques in quantifying the trace element risk to human health under considering the heritage preservation of mining history and cultural asset revitalization planning.