Journal of Contaminant Hydrology 225 (2019) 103502

Contents lists available at ScienceDirect

Journal of Contaminant Hydrology

journal homepage: www.elsevier.com/locate/jconhyd

Acid mine drainage sources and hydrogeochemistry at the Yatani mine, T

Yamagata, Japan: A geochemical and isotopic study
⁎
Shingo Tomiyamaa,b, , Toshifumi Igarashia, Carlito Baltazar Tabelina, Pawit Tangviroona,
Hiroyuki Iic
a
Division of Sustainable Resources Engineering, Faculty of Engineering, Hokkaido University, Kita 13 Nishi 8, Sapporo 060-8628, Japan
b
Mitsubishi Materials Corporation, 3-2-3, Marunouchi, Chiyoda-ku, Tokyo 100-8117, Japan
c
Graduate School of Systems Engineering, Wakayama University, 930 Sakaedani, Wakayama 640-8510, Japan

A R T I C LE I N FO A B S T R A C T

Keywords: This paper describes the geochemistry of groundwater and its flow system in the closed Yatani mine in southern
Abandoned and closed mine Yamagata Prefecture, Japan. The mine is located in a sulfide deposit containing pyrite and has been generating
Acid mine drainage (AMD) acid mine drainage (AMD). The study was intended to elucidate the formation of AMD and its flow patterns using
Groundwater recharge geological, hydrological, geochemical, and isotopic techniques. The results indicate that AMD is formed by the
Geochemical modeling
interaction of groundwater with sulfide minerals, sand slime, and tailings back-filled into excavated mine areas.
Groundwater recharge areas were identified on the mountain slope at an elevation of ~900 m. The formation of
AMD in the drifts and shaft was more extensive than that in the deeper drainage levels. Principal component
analysis was applied to the hydrogeochemical data to identify the causes of AMD formation. The first, second,
and third principal components reveal that the increased ion concentrations in mine drainage are a result of
water–mineral reactions in excavated mine areas, the contribution of groundwater in deep reductive environ-
ments, and isotopic fractionation during precipitation, respectively. A promising method of reducing AMD
formation is to prevent contact between dissolved oxygen and sulfide minerals by increasing the drainage level
or by filling the shallow underground excavated area with cementitious materials.

1. Introduction subsidiary aid cost of ~70 billion JPY (~770 million USD) (Kawabe,
2017). Water treatment is not a sustainable solution owing to its con-
There are > 5500 closed or abandoned non-ferrous metal mines in tinuing high cost. Land surface remediation and wetland treatment
Japan, many of which produce acid mine drainage (AMD; Ueda and have been implemented in Japan (Tomiyama et al., 2016a; Ogino,
Masuda, 2005; Herrera et al., 2007a, 2007b). AMD is not only acidic 2017; Okimoto et al., 2017), but again the cost is unsustainable. Now, a
but may also contain toxic contaminants such as Cu, Pb, Zn, and As, and “green remediation of abandoned mines” program is to be implemented
is a serious environmental problem in many active, closed, and aban- with the goal of returning mine landscapes to their pre-mining state,
doned mines (Younger, 2001; Gault et al., 2005; Molson et al., 2005; while also reducing treatment costs with a variety of techniques in-
Boularbah et al., 2006; Lim et al., 2008; Hien et al., 2012; Hierro et al., cluding the sealing of mining levels and wetland treatment (Nagai,
2014; Skierszkan et al., 2016; Tabelin et al., 2012a, 2012b; Zhao et al., 2018). An understanding of the origin and flow of groundwater in
2017). abandoned mines is essential for the development of countermeasures
AMD typically forms in old mine workings (mining levels, drifts and such as back-filling and sealing of surfaces in recharge areas.
shafts), waste-rock dumps, and tailings storage facilities where sulfide Groundwater flow modeling, based on the monitoring of changes in
minerals such as pyrite are exposed to oxidizing conditions (Singer and geochemistry of water percolating into mines, has provided useful in-
Stumm, 1970; Lowson, 1982; Tabelin et al., 2017a, 2017b, 2017c, formation in the development of an AMD mitigation strategy
2017d). The problem is commonly managed by neutralization (Kalin (Tomiyama et al., 2010a). Previous studies have characterized geo-
et al., 2006), precipitating heavy metals with basic materials such as chemical and isotopic compositions of AMD in Japan (Iwatsuki and
limestone, lime, or caustic soda. This method has been effectively used Yoshida, 1999; Okumura, 2003; Mahara et al., 2006; Tomiyama et al.,
in many closed or abandoned mine sites over the last 40 years, at a total 2010b; Yamaguchi et al., 2015; Tomiyama et al., 2016b) and other

⁎
Corresponding author at: Division of Sustainable Resources Engineering, Faculty of Engineering, Hokkaido University, Kita 13 Nishi 8, Sapporo 060-8628, Japan.
E-mail address: tomiyama@mmc.co.jp (S. Tomiyama).

https://doi.org/10.1016/j.jconhyd.2019.103502
Received 16 January 2019; Received in revised form 28 March 2019; Accepted 15 May 2019
Available online 16 May 2019
0169-7722/ © 2019 Elsevier B.V. All rights reserved.
S. Tomiyama, et al. Journal of Contaminant Hydrology 225 (2019) 103502

countries (Wunsch et al., 1999; Hazen et al., 2002; Sracek et al., 2004; and streams traversing the mining area. All samples were stored in 1 L
Casiot et al., 2005; Lim et al., 2008; Gammons et al., 2010; Pabst et al., polypropylene bottles before filtering through 0.22 μm membrane fil-
2018; Ramasamy et al., 2018). Further information could be provided ters. Temperature, pH, electrical conductivity (EC), and oxidation–re-
by studies of the origin and hydrogeochemistry of AMD in mining areas duction potential (ORP) were measured on-site using a thermo-pH
where water samples can be collected from inside mining levels and meter (KS-701, Shindengen Electric Manufacturing Co. Ltd., Japan), an
surface sites. EC meter (B-173, Horiba Ltd., Japan), and an ORP meter (RM-12P,
The Yatani mine area in central Honshu Island, Japan, was selected DKK-TOA Co., Japan), respectively. The ORP meter was equipped with
for this study. The area includes a closed non-ferrous sulfide mine that an Ag–AgCl electrode, and measured values were converted to standard
had been producing galena, sphalerite, native Au, and argentite. The redox potentials (Eh) by adding a temperature-dependent constant for
objectives of the study were to evaluate the geological and hydrological the electrode solution. The AMD flux on 9 November 2008, when mine
characteristics of the mine, to identify important minerals, and to de- drainage samples were taken, was 2.58 m3 min−1. There was no rainfall
termine their roles in the formation of mine drainage (AMD and neutral on 10 November 2008 when river water was sampled; precipitation
mine drainage). Groundwater flow systems in the mine area were during the 30 days before sampling total 110 mm (Fig. 3).
evaluated through water quality and isotopic studies interpreted by All sample analyses were performed at the Analytical Center of
principal component analysis (PCA). Mitsubishi Materials Techno Corporation. Major anions (Cl− and
SO42−) were analyzed by ion chromatography using Dionex DX-120
2. Materials and methods and IonPac As14 resin columns (Nippon Dionex K.K., Japan). Major
cations (Na+, K+, Ca2+, and Mg2+), total Fe, and total Mn were ana-
2.1. Geology and history of the Yatani mine lyzed by either inductively coupled plasma mass spectrometry (ICP–MS;
Agilent 7700×, Agilent Technologies Japan Ltd., Japan) or inductively
The Yatani mine area is located 28 km southwest of Yonezawa City, coupled plasma atomic emission spectroscopy (ICP–AES; ICAP-575,
within the Bandai–Asahi National Park (Fig. 1). It is situated in the Thermo Fisher Scientific K.K., Japan). Alkalinity was determined by
green tuff area of northeast Japan, and includes a granitic basement titration with 0.01 M H2SO4 to pH 4.8. Charge balances were
overlain by Neogene volcanic and sedimentary rocks. Surface mining within ± 5%, indicating that all major cations and anions were ade-
began in 1870 after the discovery of Au ore outcrops. When these were quately quantified. Analytical uncertainties of ICP–MS, ICP–AES, and
exhausted, mining continued underground for several decades. Taihei ion chromatography results were all 2%–5%.
Mining Co. Ltd. acquired mining rights to the area in 1952 and began Oxygen isotopic ratios were determined by the CO2 equilibration
mining Pb and Zn ores. By 1978 the monthly crude ore output was > method (Epstein and Mayeda, 1953) and hydrogen isotopic ratios by
10,000 tons with average Pb and Zn grades of 2% and 4.4%, respec- the method of Coleman et al. (1982). δ18O (1000((18O/16O)sample/
tively. Decreasing metal prices and the appreciating Japanese yen (18O/16O)VSMOW) − 1), ‰) and δD (1000((2H/1H)sample/
eventually made the operation unprofitable, and it closed in 1988 (Sato (2H/1H)VSMOW) − 1), ‰) values were calculated relative to Vienna
et al., 1978). Standard Mean Ocean Water (VSMOW) with analytical precisions of
The Yatani mineral deposit is of hydrothermal origin, formed within 0.1‰ and 1‰, respectively.
green tuff host rock. It includes two types of veins: one rich in Pb and
Zn, and the other containing Au and Ag. The PbeZn ore deposits con- 2.3. Principal component analysis
stitute the main veins of the mine, reaching a maximum width of ~10 m
and closely following NE–SW- and E–W-trending faults in the area. The Principal component analysis (PCA) is a multivariate statistical
PbeZn veins comprise mainly quartz, galena, sphalerite, pyrite, calcite, technique used to reduce complex high-dimensional data sets by con-
and chlorite and sometimes include native Au and argentite. The verting observations into sets of values or “principal components” (PC)
AueAg ore veins contain quartz, native Au, argentite, pyrite, and rho- related to original system components (Abdi and Williams, 2010). The
dochrosite. In both types of veins, ore minerals exhibit striped textures, approach is ideal for geochemical surveys because each sample is de-
indicating that quartz crystallized first, followed by precipitation of scribed by multiple factors such as pH, Eh, EC, and ion concentrations.
galena and zincblende, and finally by deposition of native Au and ar- PCA has been applied in many studies comparing the qualities of
gentite (Taniguchi, 1969). groundwater, spring water, and river water, and also in the classifica-
Two mining methods were employed at the mine, involving the cut- tion of hot-spring waters (Masumoto and Hibiya, 1996; Takamatsu,
and-fill method in deeper parts using sand, slime, and tailings as filling 1986; Mahlknecht et al., 2003).
materials that did not consolidate with cement; and the shrinkage- Here, PCA was performed using the multivariate analysis software
stoping method in shallower parts. A mineral concentration system Pirouette LE ver. 3.11 (GL Sciences Inc., Japan), and it was applied to
with its own power substation and repair facilities was constructed four variables measured on site (temperature, pH, EC, and Eh), the
underground. The adit mouth of the drainage level (level –2 L) is lo- concentrations of seven major ions (Na+, K+, Ca2+, Mg2+, Cl−,
cated at elevations of 626 and 2280 m from the shaft. The tunnel ex- HCO3−, and SO42−), and the isotopic ratios δD and δ18O. For nor-
tends from the bottom level (−10 L) to the top (+2 L) level with a malization of variables, logarithms were used for ion concentrations,
height difference of 600 m and runs E–W for ~2.5 km. Old mine and measured values for the four factors measured on site, and isotopic
workings deeper than the drainage level (level –2 L) are kept sub- ratios were the antilogs of the measured values.
merged, and AMD from the drainage level is presently treated by
neutralization. 3. Results and discussion

2.2. Water sample collection and analysis 3.1. Geochemical properties of water samples

Water samples were collected on 9 and 10 November 2008, from The geochemical and isotopic properties of water samples collected
locations shown in Figs. 1 and 2. In the old mine workings (mining from the Yatani mine area are summarized in Table 1 and their corre-
levels, drifts, and a shaft), samples were collected from the shaft and sponding water chemistry displayed in Stiff (Figs. 1 and 2) and Piper
other prominent parts of mine workings where significant discharge of diagrams (Fig. 4).
groundwater was observed, for example, from cracks in drainage level
walls, and fallen water from the back of the drift and the top of the 3.1.1. River water
shaft. Samples were also collected from several points in major rivers The pH, EC, Eh, and temperature of river water samples are in the

2
S. Tomiyama, et al. Journal of Contaminant Hydrology 225 (2019) 103502

Fig. 1. Location of water sampling points, with corresponding water sample properties indicated by Stiff diagrams.

3
S. Tomiyama, et al. Journal of Contaminant Hydrology 225 (2019) 103502

NNW SSE

Vertical
shaft
E.L.(m) M-4 M-5 M-6 M-7 M-8 M-9 M-10
900 M-2 M-3 Yatanizawa River
Adit Yatanizawa River
800 mouth
700
600 Drainage level (–2L) Concentrator
0 M-11500 m Mining level (0L)
㻯㼘

Legend
M-1
M-11 (drift) M-12 (drift)
Stiff diagram
Cations
Cations Anions
Anions Cations Anions
meq L-1 meq L-1
-2
2 -1
1 00 11 22 20
-2 10
-1 0 10
1 20
2

Na+K Cl Na+K Cl
Ca HCO3 Ca HCO3 M-13 M-14
Mg SO4 Mg SO4 (Shaft) (Shaft)

Fig. 2. Stiff diagrams of mine drainage samples from the drainage level (−2 L), including samples collected from the drift and shaft (note that the concentrations in
shaded Stiff diagrams are 10 times higher than those in unshaded diagrams).

(samples M-1 to M-9) were < 2.5 and 4.2 mg L−1, respectively, while
those in samples from around veins, drift, and the shaft (samples M-10
to M-14) were ~80 and ~50 times higher, respectively.

3.2. Hydrogen and oxygen isotopic ratios

The δD and δ18O values of water samples are in the ranges of

−67.1‰ to −55.6‰ and − 10.8‰ to −8.7‰, respectively, with re-
lationships illustrated in Fig. 6. Isotopic data plot close to the meteoric
water line (Craig, 1961; Dansgaard, 1964), and are very similar to those
of river water and rainwater collected in Yamagata Prefecture (Mizota
and Kusakabe, 1994). The Tengusawa River water samples from higher
Fig. 3. Relationships between precipitation, total AMD flux, and pH measured elevations (~930 m; samples R-9 and R-10) have lower δ18O and δD
at the neutralization treatment plant. values, suggesting that the relationship between elevation and isotopic
ratios is likely due to isotopic fractionation during precipitation
ranges of 6.0–8.9, 1.8–9.1 mS m−1, +353 to +579 mV, and (Dansgaard, 1964; Lawrence and White, 1991; Mizota and Kusakabe,
4.6 °C–7.9 °C, respectively. Concentrations of major ions vary with lo- 1994). The ranges of δ18O and δD values in mine drainage approximate
cation, with Ca2+, Na+, and HCO3– being dominant downstream in the those of river water, barring sample R-14. Rainwater in mine drainage
Yatanizawa River and the Tengusawa River (Na+ − HCO3– or is temporally decoupled from that in river water, so isotopic ratios in
Ca2+ − HCO3– type waters (samples R-1, R-2, R-9, R-10, R-11, R-12, R- river water may not display significant seasonal variations, with ratios
13, and R-14), while SO42− predominates in samples (Ca2+ − SO42− in mine drainage effectively being averaged over influx and under-
type water; samples R-3, R-4, R-5, R-6, R-7, and R-8) collected upstream ground flow. The δ18O and δD values of mine drainage also vary with
where mineralized zones and veins are located (Sato et al., 1978). location (Fig. 7). Isotopic ratios of drainage from the drift and shaft
(samples M-10 to M-4) are lower than those of samples from the drai-
nage level (samples M-1 to M-9). δ18O and δD values of mine drainage
3.1.2. Mine drainage
from the drift and shaft are similar to those of river water from higher
The pH, EC, Eh, and temperature of mine drainage are in the ranges
elevations (samples R-9 and R-10; Fig. 6). Drainage from the mine
of 2.9–7.8, 16.5–381 mS m−1, +304 to +656 mV, and 11.6 °C–18.7 °C,
drainage level may thus originate in the Yatanizawa River and its tri-
respectively. Mine drainage samples are Ca2+–SO42− type waters, ex-
butaries, whereas that from the drift and shaft may originate in river
cept for sample M-1 (from near the adit mouth), which is a Ca2+–HCO3–
water at higher elevations (> 900 m).
type water. Mine drainage samples M-1 to M-14 from old mine work-
ings has lower pH and higher EC values than river water, indicating
significant impact of AMD on groundwater quality. EC, pH, and major 3.3. Principal component analysis
ion concentrations of mine drainage also varies with location, with EC
values increasing with decreasing pH from adit mouth to veins and PCA results are shown in Table 2. Three principal components (PC1,
shaft (Fig. 5). Drainage sample M-12, from the drift, has the highest EC PC2, and PC3) were selected, which together account for > 82% of the
(381 mS m−1) and lowest pH (2.9). Sample M-11, from the shaft, has total variability in the dataset. This indicates that the water quality is
high EC (235 mS m−1) and a pH of 3.1. A similar trend was observed dominated by these three factors. The first principal component (PC1)
with concentrations of total Fe and Mn in mine drainage samples. Total has high loadings for Ca2+, Mg2+, and SO42− with values of > 0.35,
Fe and Mn concentrations in samples collected near the adit mouth suggesting that the increase in ion concentrations in mine drainage may

4
S. Tomiyama, et al. Journal of Contaminant Hydrology 225 (2019) 103502

Table 1
Geochemical properties of water samples.
Sample type Sample pH EC Eh Temp Concentration (mg L−1) Stable isotope ratios (‰)
(mS m−1) (mV) (°C)
− −
Na+
K +
Ca 2+
Mg 2+
Cl HCO3 SO4 2−
T-Fe T-Mn δD δ18O

River water R-1 8.9 2.6 353 7.9 2.2 0.3 2.5 0.5 1.9 7.8 4.3 < 0.1 < 0.1 −61.4 −10.1
R-2 8.5 9.1 392 6.2 2.3 0.6 18 1.3 2.1 35 23 < 0.1 < 0.1 −61.7 −10.1
R-3 8.1 3.7 412 5.0 1.6 0.7 4.4 0.7 1.8 4.4 11 < 0.1 0.2 −64.2 −10.4
R-4 6.0 4.0 525 7.1 1.8 1.9 3.6 0.7 2.1 0.7 15 < 0.1 0.1 −60.9 −10.2
R-5 6.4 4.9 472 6.5 2.7 1.0 6.8 0.6 1.7 9.2 16 < 0.1 0.2 −60.6 −10.3
R-6 6.9 6.2 414 4.6 3.2 0.6 9.9 0.9 1.9 14 21 0.1 < 0.1 −61.8 −9.7
R-7 6.7 3.5 579 5.6 2.7 0.8 4.4 0.4 1.6 7.4 11 < 0.1 < 0.1 −64.1 −10.3
R-8 6.7 4.2 507 7.3 2.0 1.5 4.8 1.0 2.1 6.6 13 < 0.1 < 0.1 −67.1 −10.5
R-9 6.7 2.1 498 5.5 2.5 0.5 2.1 0.3 1.6 11 1.7 < 0.1 < 0.1 −69.4 −10.8
R-10 6.7 1.8 497 6.4 2.3 1.0 1.0 0.3 1.7 8.9 1.0 < 0.1 < 0.1 −67.7 −10.7
R-11 6.7 2.0 514 6.8 2.4 0.5 1.7 0.4 1.9 8.6 1.8 0.5 0.1 −64.1 −10.1
R-12 6.9 3.9 511 6.1 3.2 0.6 5.4 0.7 1.8 22 4.4 0.2 < 0.1 −69.4 −10.6
R-13 6.9 3.6 515 6.5 2.3 0.7 5.4 1.1 1.8 13 10 < 0.1 < 0.1 −65.9 −10.4
R-14 7.1 5.2 517 6.2 2.6 0.8 7.6 1.5 2.1 18 11 < 0.1 < 0.1 −55.6 −8.7
R-15 6.9 4.6 512 6.4 3.2 0.7 6.4 0.5 1.7 14 11 < 0.1 < 0.1 −65.2 −10.5
Mine drainage from drainage level M-1 7.6 16.5 512 11.6 5.5 0.7 29 1.8 2.4 70 32 < 0.1 < 0.1 −63.8 −10.2
(−2 L) M-2 7.4 22.6 516 13.5 6.1 1.0 43 2.2 2.3 78 62 0.2 1.9 −63.9 −10.2
M-3 7.4 29.2 518 12.6 5.5 1.3 59 2.4 2.2 93 88 0.1 0.4 −64.1 −10.4
M-4 7.4 32.6 480 12.8 6.2 1.1 66 2.0 2.2 54 130 < 0.1 0.3 −63.2 −10.0
M-5 7.4 37.2 478 12.0 3.3 1.3 73 5.6 2.2 57 160 2.5 4.2 −64.6 −10.6
M-6 7.8 70.7 477 16.6 6.8 1.0 66 1.6 2.1 53 130 < 0.1 1.6 −65.0 −10.6
M-7 7.6 73.6 439 17.4 18 1.3 120 3.6 2.3 110 260 < 0.1 0.5 −66.1 −10.6
M-8 7.1 97.2 304 17.1 17 1.7 210 8.9 2.3 260 380 < 0.1 2.1 −63.7 −9.9
M-9 7.0 37.5 455 17.2 10 2.0 230 8.4 2.1 220 400 < 0.1 < 0.1 −63.4 −10.4
M-10 5.4 70.7 459 18.3 8.3 1.8 340 17 2.2 230 770 22 33 −61.1 −10.0
Mine drainage from drifts M-11 2.9 381 476 17.8 5.4 4.3 290 85 2.1 0 3200 210 210 −67.1 −10.7
M-12 3.1 235 452 17.7 6.3 3.3 170 45 2.2 0 1500 47 170 −64.9 −10.7
Mine drainege from shaft M-13 6.7 32.8 432 16.8 15 0.4 39 1.0 2.0 61 80 1.4 0.3 −66.7 −10.8
M-14 5.6 127 656 18.7 14 2.7 210 19 1.9 34 810 89 62 −66.1 −10.8

11 500
pH
9 EC 400

EC (mS·m–1 )
7 300
pH

5 200

3 100

1 0
500 1000 1500 2000 2500
Distance from adit mouth (m)

Fig. 5. Spatial variations in pH and EC of mine drainage samples along the mine
drainage level.

et al., 1987; Tabelin et al., 2017a, 2017b). The high Ca2+ loading re-
lates to reactions of calcite (CaCO3) associated with AueAg veins and
PbeZn veins (Taniguchi, 1969), and possibly to interactions of AMD
with concrete used in the construction of the underground concentrator
and its supporting facilities. The significant Mg2+ loading relates to
reactions of AMD with the mineral chlorite (Mg, Fe, Al)6(Al,
Si)4O10(OH)8 associated with PbeZn veins (Taniguchi, 1969) or with
cementitious materials containing significant amounts of Mg2+.
For the second principal component (PC2), there are high loadings
for pH and HCO3−, with values of 0.40 and 0.52, respectively, as ex-
pected because of the well-established inverse relationship between pH
and HCO3– concentration (Eang et al., 2018a, 2018b).
Fig. 4. Piper diagram showing the hydrochemical facies of river water, mine
The third principal component (PC3) has high loadings for H and O
drainage, and AMD samples collected from the Yatani mine area. isotopic ratios (> 0.60), possibly reflecting the strong correlation be-
tween them in water samples and the effect of isotopic fractionation
during precipitation (Figs. 6 and 7).
be attributed to water–mineral interactions within old mine workings.
The sum of the contributions of PC1, PC2, and PC3 is 82% (Table 2),
The high SO42− loading in PC1 could be explained by the oxidation of
indicating that the geochemistry of surface water and mine drainage are
pyrite (FeS2), a common sulfide mineral in metal deposits and altered
satisfactorily explained by the three PCs.
rocks. Pyrite oxidation is facilitated by Fe3+ formed through the oxi-
Scatter plots of PC1 vs. PC2 and PC1 vs. PC3 are shown in Fig. 8a
dation of Fe2+ by dissolved oxygen (Singer and Stumm, 1970; Moses

5
S. Tomiyama, et al. Journal of Contaminant Hydrology 225 (2019) 103502

12 > M-11. This shows that the source of mine drainage in drifts (M-
11 and M-12) did not come mainly from the shaft (M-13 and M-14). It is
possible that a large portion of the water from M-11 and M-12 origi-
nated from the deeper reductive groundwater. This is consistent with
the observed flow pattern in which M-11 and M-12 represent water
flowing from the back of the drifts, M-13 was the water dripping from
the top of the shaft, and M-14 was spring water from the shaft.
According to the principal component scores of PC1 vs PC3
(Fig. 8b), two groups of samples can be classified. The group of higher
PC1 scores belongs to river water from the elevation of < 900 m and
mine drainage from the drainage level. On the other hand, the other
group represents river water from the elevation of > 900 m and mine
drainage samples from the drifts and the shaft. These results indicate
that the main source of mine drainage from the drainage level is from
the Yatanizawa River while mine drainage from the drifts and shaft
were derived from surface water with higher elevation than that of the
Yatanizawa River. In addition, the board distribution of PC1 with
narrow PC3 scores on shift samples indicates the progress of water–-
mineral interactions from the same source of water, along the flow from
the top to bottom in the shift, as represented by an arrow in Fig. 8b.

3.4. Concept of groundwater flow and discharge

Fig. 6. Relationship between δD and δ18O values of water samples, relative to
the meteoric water line. A conceptual model of drainage flow was devised on the basis of
results described in Sections 3.1–3.3, as shown in Fig. 9. Groundwater
-9.5 -59 infiltrating from the Yatanizawa River and its tributaries above the
δ18O
18O mine drainage level, and from the mountain slope (elevation > 900 m)
δD
D north of the mining area, reacts with sulfide minerals, sand slime and
-10.0 -62
tailings back-filled into excavated mine areas releasing SO42−, Ca2+,
18 O (‰)

D (‰)

and Mg2+ before flowing out of the drainage level. In the PCA study,
-10.5 -65 PC1 is related to the degree of water–rock interaction, PC2 to the
contribution of groundwater from the deep reductive environment, and
PC3 to correlation between H and O isotopic ratios and isotopic frac-
-11.0 -68
500 1000 1500 2000 2500
tionation during precipitation. The total AMD flux measured at the
neutralization treatment plant increased during the snowmelt season of
Distance from adit mouth (m)
March to May, with the pH decreasing over that time (Fig. 3). This
Fig. 7. Spatial distribution of δD and δ18O values of mine drainage samples suggests that the groundwater flux, infiltrating from the mountain slope
along the drainage level. to the excavated area and reacting with sulfide minerals, increased
during the snowmelt season. It follows that a possible method of re-
Table 2 ducing AMD formation would be to prevent contact between dissolved
Factor loadings and contributions of principal components (PCs). oxygen and sulfide minerals by changing the drainage level from the
current −2 level to the 0 level, or by filling the shallow underground
Variables PC1 PC2 PC3
excavated area with cementitious materials.
pH −0.26 0.40 −0.01
EC 0.33 −0.19 −0.12 4. Conclusions
Eh 0.12 −0.46 −0.03
Temperature 0.36 0.23 −0.10
Chemical composition Na+ 0.29 0.33 −0.12
The findings of this geological, hydrological, geochemical, and
K+ 0.31 −0.21 0.12 isotopic study of acid mine drainage at the Yatani mine are as follows.
Ca2+ 0.35 0.26 0.06
Mg2+ 0.39 0.04 0.08 (1) AMD is formed through interactions between groundwater and
Cl− 0.24 0.12 0.34
sulfide minerals, sand slime, and tailings back-filled into excavated
SO42− 0.38 0.11 0.12
HCO3− −0.12 0.52 −0.02 mine areas.
Stable δD −0.06 −0.08 0.66 (2) The groundwater recharge area for AMD formation is located on the
isotopes δ18O −0.11 0.09 0.60 mountain slope at altitudes of ~900 m.
Rate of contribution (%) 46.3 23.1 13.2 (3) AMD formation in the drifts and shaft is more pronounced than in
the mine drainage level.
(4) Based on a conceptual model of groundwater flow at the mine,
and b, respectively. Two different groups of data are clearly observed in
AMD formation could be reduced by preventing contact between
both plots, belonging to the water samples collected from the river and
dissolved oxygen and sulfide minerals by increase the drainage
old mine workings, respectively. The river water samples are on the
level, or by filling shallow underground excavated areas with ce-
negative side of PC1 whereas the mine drainage samples have positive
mentitious materials.
PC1 loading values. This indicates the possibility of interactions be-
tween groundwater and sulfide minerals, sand slime, and tailings back-
Acknowledgments
filled into excavated mine areas.
The PC2 coefficients of mine drainage samples range from positive
The authors thank the editor and anonymous reviewers for their
to negative values in the following order: M-13 > M-14 > M-
constructive comments that improved this manuscript. We also thank

6
S. Tomiyama, et al. Journal of Contaminant Hydrology 225 (2019) 103502

Fig. 8. PCA principal component scores of (a) PC2 vs. PC1 and (b) PC3 vs. PC1.

Bird's-eye view of
ground surface
Watershed
N

Groundwater flow towards

the excavated areas

Watershed Excavated areas

Tengusawa River

Drift

Yatanizawa
Drift River
Shaft Adit mouth

Drainage level
(–2L level)
Cross-section

SSE NNW

Recharge area of Recharge area of

the drifts and shaft the drainage tunnel

Water table Shaft Groundwater flow towards E.L. (m)

the drainage level 1000
Adit mouth

Drainage level (–2L level) 500

Alteration zone
0
Groundwater flow towards
2 km 1 0
the excavated areas

Fig. 9. Conceptual model of groundwater flow and discharge.

7
S. Tomiyama, et al. Journal of Contaminant Hydrology 225 (2019) 103502

the staff of Mitsubishi Materials Corporation, Eco-Management Molson, J.W., Fala, O., Aubertin, M., Bussière, B., 2005. Numerical simulations of pyrite
Corporation, and Mitsubishi Materials Techno Corporation for their oxidation and acid mine drainage in unsaturated waste rock piles. J. Contam. Hydrol.
78, 343–371.
help, advice, and cooperation during this study. Moses, C.O., Nordstrom, D.K., Herman, J.S., Mills, A.L., 1987. Aqueous pyrite oxidation
by dissolved oxygen and ferric iron. Geochim. Cosmochim. Acta 51, 1561–1571.
References Nagai, Y., 2018. Action for the mountains of origin recurrence of closed mines which
utilized green remediation. In: Program and Abstracts, MMIJ Tokyo, pp. 36 (in
Japanese).
Abdi, H., Williams, L.J., 2010. Principal component analysis. Wiley Interdiscipl. Rev. 2 Ogino, T., 2017. Water purification by constructed wetland. In: Program and Abstracts,
(4), 433–459. MMIJ and Earth 2017 Sapporo, pp. 89 (in Japanese).
Boularbah, A., Schwartz, C., Bitton, G., Morel, J.L., 2006. Heavy metal contamination Okimoto, N., Yamagata, S., Mizukoshi, H., Sugimoto, H., Sakakibara, N., 2017.
from mining sites in South Morocco: I. use of a biotest to assess metal toxicity of Application of a soil erosion prevention method in the abandoned mine site by a
tailings and soils. Chemosphere 63, 802–810. greening technology with cement and wood chips, “Chipcrete”. In: Program and
Casiot, C., Lebrun, S., Morin, G., Bruneel, O., Personné, J.C., Elbaz-Poulichet, F., 2005. Abstracts, MMIJ and Earth 2017 Sapporo, pp. 91–92 (in Japanese).
Sorption and redox processes controlling arsenic fate and transport in a stream im- Okumura, M., 2003. Hydrogeochemical study of the 3M level adit drainage at the old
pacted by acid mine drainage. Sci. Total Environ. 347, 122–130. Matsuo mine. Resour. Geol. 53, 173–182 (in Japanese with English abstract).
Coleman, M.L., Shepherd, T.J., Durham, J.J., Rouse, J.E., Moore, G.R., 1982. Reduction of Pabst, T., Bussière, B., Aubertina, M., Molson, J., 2018. Comparative performance of
water with zinc for hydrogen isotope analysis. Anal. Chem. 54, 993–995. cover systems to prevent acid mine drainage from pre-oxidized tailings: a numerical
Craig, H., 1961. Standard for reporting concentrations of deuterium and oxygen-18 in hydro-geochemical assessment. J. Contam. Hydrol. 214, 39–53.
natural waters. Science 133, 1833–1834. Ramasamy, M., Power, C., Mkandawire, M., 2018. Numerical prediction of the long-term
Dansgaard, W., 1964. Stable isotopes in precipitation. Tellus 16, 436–438. evolution of acid mine drainage at a waste rock pile site remediated with an HDPE-
Eang, K.E., Igarashi, T., Fujinaga, R., Kondo, M., Tabelin, C.B., 2018a. Groundwater lined cover system. J. Contam. Hydrol. 216, 10–26.
monitoring of an open-pit limestone quarry: groundwater characteristics, evolution Sato, N., Takatori, I., Yanagisawa, S., 1978. Geology and ore deposits of the Yatani mine,
and their connections to rock slopes. Environ. Monit. Assess. 190, 193. Yamagata prefecture, Japan, with special reference to some suggestions on explora-
Eang, K.E., Igarashi, T., Kondo, M., Nakatani, T., Tabelin, C.B., Fujinaga, R., 2018b. tion. Min. Geol. 28, 177–190 (in Japanese with English abstract).
Groundwater monitoring of an open-pit limestone quarry: water–rock interaction and Singer, P.C., Stumm, W., 1970. Acidic mine drainage: the rate-determining step. Science
mixing estimation within the rock layers by geochemical and statistical analyses. Int. 167, 1121–1123.
J. Min. Sci. Technol. 28, 849–857. Skierszkan, E.K., Mayer, K.U., Weis, D., Beckie, R.D., 2016. Molybdenum and zinc stable
Epstein, S., Mayeda, T., 1953. Variation of the 18O content of waters from natural waters. isotope variation in mining waste rock drainage and waste rock at the Antamina
Geochim. Cosmochim. Acta 4, 213–224. mine, Peru. Sci. Total Environ. 550, 103–113.
Gammons, C.H., Duaime, T.E., Parker, S.R., Poulson, S.R., Kennelly, P., 2010. Sracek, O., Choquette, M., Gélinas, P., Lefebvre, R., Nicholson, R.V., 2004. Geochemical
Geochemistry and stable isotope investigation of acid mine drainage associated with characterization of acid mine drainage from a waste rock pile, mine Doyon, Quebec,
abandoned coal mines in Central Montana, USA. Chem. Geol. 269, 100–112. Canada. J. Contam. Hydrol. 69, 45–71.
Gault, A.G., Cooke, D.R., Townsend, A.T., Charnock, J.M., Polya, D.A., 2005. Mechanisms Tabelin, C.B., Igarashi, T., Takahashi, R., 2012a. Mobilization and speciation of arsenic
of arsenic attenuation in acid mine drainage from mount Bischoff, western Tasmania. from hydrothermally altered rock in laboratory column experiments under ambient
Sci. Total Environ. 345, 219–228. conditions. Appl. Geochem. 27, 326–342.
Hazen, J.M., Milliams, M.W., Stover, B., Wireman, M., 2002. Characterization of acid Tabelin, C.B., Igarashi, T., Yoneda, T., 2012b. Mobilization and speciation of arsenic from
mine drainage using a combination of hydrometric, chemical and isotopic analyses, hydrothermally altered rock containing calcite and pyrite under anoxic conditions.
Mary murphy mine, Colorado. Environ. Geochem. Health 24, 1–22. Appl. Geochem. 27, 2300–2314.
Herrera, P., Uchiyama, H., Igarashi, T., Asakura, K., Ochi, Y., Ishizuka, F., Kawada, S., Tabelin, C.B., Sasaki, R., Igarashi, T., Park, I., Tamoto, S., Arima, T., Ito, M., Hiroyoshi, N.,
2007a. Acid mine drainage treatment through a two-step neutralization ferrite-for- 2017a. Simultaneous leaching of arsenite, arsenate, selenite and selenate, and their
mation process in northern Japan: physical and chemical characterization of the migration in tunnel-excavated sedimentary rocks: I. column experiments under in-
sludge. Miner. Eng. 20, 1309–1314. termittent and unsaturated flow. Chemosphere 186, 558–569.
Herrera, P., Uchiyama, H., Igarashi, T., Asakura, K., Ochi, Y., Iyatomi, N., Nagae, S., Tabelin, C.B., Veerawattananun, S., Ito, M., Hiroyoshi, N., Igarashi, T., 2017b. Pyrite
2007b. Treatment of acid mine drainage through a ferrite formation process in oxidation in the presence of hematite and alumina: I. batch leaching experiments and
Central Hokkaido, Japan: evaluation of dissolved silica and aluminium interference kinetic modeling calculations. Sci. Total Environ. 580, 687–698.
in ferrite formation. Miner. Eng. 20, 1255–1260. Tabelin, C.B., Veerawattananun, S., Ito, M., Hiroyoshi, N., Igarashi, T., 2017c. Pyrite
Hien, N.T.T., Yoneda, M., Matsui, A., Hai, H.T., Pho, N.V., Quang, N.H., 2012. oxidation in the presence of hematite and alumina: II. Effects on the cathodic and
Environmental contamination of arsenic and heavy metals around Cho Dien lead and anodic half-cell reactions. Sci. Total Environ. 581–582, 126–135.
zinc mine, Vietnam. J. Water Environ. Technol. 10, 253–265. Tabelin, C.B., Sasaki, R., Igarashi, T., Park, I., Tamoto, S., Arima, T., Ito, M., Hiroyoshi, N.,
Hierro, A., Olías, M., Cánovas, C.R., Martín, J.E., Bolivar, J.P., 2014. Trace metal parti- 2017d. Simultaneous leaching of arsenite, arsenate, selenite and selenate, and their
tioning over a tidal cycle in an estuary affected by acid mine drainage (Tinto estuary, migration in tunnel-excavated sedimentary rocks: II. Kinetic and reactive transport
SW Spain). Sci. Total Environ. 18–28, 497–498. modeling. Chemosphere 188, 444–454.
Iwatsuki, T., Yoshida, H., 1999. Groundwater chemistry and fracture mineralogy in the Takamatsu, N., 1986. Classification of coastal type saline springs by principal component
basement granitic rock in the Tono uranium mine area, Gifu prefecture, Japan: analysis. J. Hot Spring Sci. 36, 158–166 (in Japanese with English abstract).
groundwater composition eh evolution analysis by fracture filling minerals. Taniguchi, H., 1969. Geology and ore deposits of the Yatani mine, with special reference
Geochem. J. 33, 19–32. to the gold and silver veins. Min. Geol. 19, 113–121 (in Japanese with English ab-
Kalin, M., Fyson, A., Wheeler, W.N., 2006. The chemistry of conventional and alternative stract).
treatment systems for the neutralization of acid mine drainage. Sci. Total Environ. Tomiyama, S., Ueda, A., Ii, H., Nakamura, Y., Koizumi, Y., Saito, K., 2010a. Sources and
366, 395–408. flow system of groundwater in the Hosokura mine, Miyagi prefecture, using geo-
Kawabe, N., 2017. The significance of the soil environment, aqueous environment and chemical method and numerical simulation. J. MMIJ 126, 31–37 (in Japanese with
planting trees measure in mine pollution control. In: Program and Abstracts, MMIJ English abstract).
and Earth 2017 Sapporo, pp. 89–90 (in Japanese). Tomiyama, S., Ii, H., Koizumi, Y., Metugi, H., 2010b. Modeling of groundwater recharge
Lawrence, J.R., White, S.D., 1991. A tribute to Samuel Epstein. In: Taylor Jr.H.P., O'Neil, and discharge in Tomitaka mine, Miyazaki prefecture. J. Groundwater Hyd. 52,
J.R., Kaplan, I.R. (Eds.), The Geochemical Society. vol. 3. Special Publication, New 261–274 (in Japanese with English abstract).
York, pp. 169–185. Tomiyama, S., Takahashi, K., Yamagata, S., Igarashi, T., 2016a. Study the effect of sealing
Lim, H.S., Lee, J.S., Chon, H.T., Sager, M., 2008. Heavy metal contamination and health ground surface on reduction of mine drainage. In: Program and Abstracts, MMIJ
risk assessment in the vicinity of the abandoned Songcheon au–Ag mine in Korea. J. Morioka, pp. 130 (in Japanese).
Geochem. Explor. 96, 223–230. Tomiyama, S., Igarashi, T., Ii, H., Takano, H., 2016b. Sources and flow system of
Lowson, R.T., 1982. Aqueous oxidation of pyrite by molecular oxygen. Chem. Rev. 82, groundwater in the Shimokawa mine, North Hokkaido, using geochemical method
461–497. and numerical simulation. J. MMIJ 132, 80–88 (in Japanese with English abstract).
Mahara, Y., Nakata, E., Oyama, T., Miyakawa, K., Igarashi, T., Ichihara, Y., Matsumoto, Ueda, H., Masuda, N., 2005. An analysis of mine drainage treatment cost and the tech-
K., 2006. Proposal for the methods to characterize fossil seawater: distribution of nical development to prevent mine pollution. J. MMIJ 121, 323–329 (in Japanese
anions, cations and stable isotopes, and estimation on the groundwater residence with English abstract).
time by measuring 36Cl at the Taiheiyou coal mine. J. Groundwater Hyd. 48, 17–33 Wunsch, D.R., Dinger, J.S., Graham, D.R., 1999. Predicting groundwater movement in
(in Japanese with English abstract). large mine spoil areas in the Appalachian plateau. Int. J. Coal Geol. 41, 73–106.
Mahlknecht, J., Steinich, B., Navarro de Léon, I., 2003. Groundwater chemistry and mass Yamaguchi, K., Tomiyama, S., Metugi, H., Ii, H., Ueda, A., 2015. Flow and geochemical
transfers in the independence aquifer, central Mexico, by using multivariate statistics modeling of drainage from Tomitaka mine, Miyazaki, Japan. J. Environ. Sci. 36,
and mass-balance models. Environ. Geol. 45, 781–795. 130–143.
Masumoto, K., Hibiya, K., 1996. Evaluation of regional groundwater flow in rock mass by Younger, P.L., 2001. Mine water pollution in Scotland: nature, extent and preventative
geochemical survey: a case study at tunneling site. J. Japan Soc. Eng. Geol. 36, 70–78 strategies. Sci. Total Environ. 265, 309–326.
(in Japanese with English abstract). Zhao, Q., Guo, F., Zhang, Y., Ma, S., Jia, X., Meng, W., 2017. How sulfate-rich mine
Mizota, C., Kusakabe, M., 1994. Spatial distribution of δD–δ18O values of surface and drainage affected aquatic ecosystem degradation in northeastern China, and potential
shallow groundwaters from Japan, South Korea and East China. Geochem. J. 28, ecological risk. Sci. Total Environ. 609, 1093–1102.
387–410.