Simulation of Contaminants Transport in Groundwater From Nickel Mining Waste Dump Southeast Sulawesi

The irregularities of nickel resource mining in Indonesia cause many serious environmental problems. Piles of leftover rocks on nickel mining waste dumps have the potential to be a source of heavy metal seepage into the water. This study was conducted to assess the impact of nickel mining in the Langgikima Subdistrict of the North Konawe Regency of Southeast Sulawesi Province. The focus is to assess the migration of hexavalent chromium (Cr 6+ ) and iron (Fe) using MT3DMS to model the transport of solutes. The study's goal was to identify cr 6+ and Fe concentrations in waste dumps and predict the spread of contaminants over a 20-year period of time. XRF (X-Ray Fluorescence) is done to determine the content of elements and minerals in rocks. Toxicity Characteristics Leaching Procedure (TCLP) is performed to estimate the concentration of Cr 6+ and Fe in waste dumps. AAS (Atomic Absorption Spectrophotometer) to nd out the content of Cr 6+ and Fe in surface water and land water samples. The results showed the highest concentrations of Cr 6+ of 0.0462 mg/L and Feat 0.8709 mg/L. Simulations without compacted clay coatings, Cr 6+ and Fe contaminants could be transported consecutively by 2.7 km and 2.42 km while simulations used compacted clay layers with a hydraulic conductivity of 1 × 10 −9 m/s of Cr 6+ and Fe contaminants could be transported consecutively by 0.412 km and 0.467 km. It can therefore be concluded that heavy metals in the remaining rock piles from the waste dump can be transported into groundwater, and the action of using a compacted layer of clay must be taken to prevent contaminant migration into groundwater.


Introduction
Southeast Sulawesi province is an area in Indonesia with nickel reserves of 97.4 billion Wmt, spread over an area of 480 thousand hectares. This potential causes the number of mining companies in Southeast Sulawesi to grow rapidly, especially in recent years. The nickel mining system applied in Indonesia is an open-pit mine. The irregularities of nickel resource mining in Indonesia cause many serious environmental problems. Piles of leftover rocks in nickel mining waste dumps have the potential to be a source of heavy metal pollutants such as Pb, Cd, Cr, Fe, Cu, Zn, and Ni (Ngkoimani & Chaerul, 2017).
Among these elements, cr has the strongest capacity to migrate under certain climatic conditions.
Overburden containing high cr2O3 and Fe compounds can be a source of hexavalent chromium and iron seepage in surface water and groundwater (Tiwary et al., 2005).
Hexavalent chromium is soluble and highly mobile rather than the trivalent form. Cr 6+ is highly carcinogenic and is known to be 100 -1000 times more toxic than Cr 3+ (trivalent chromium) (Linos et al., 2011). Cr 6+ is known to cause cancer of the lungs, nasal cavity, paranasal sinuses, and stomach, as well as the larynx (Linos et al., 2011). Cr 6+ is produced from the natural oxidation process of the cr 3+ carrier mineral. Cr species are affected by pH and Eh water (Henderson, 1994). The release of Cr 6+ into laterite groundwater is related to the oxidative reaction of Cr 3+ with Mn oxide (Equeenuddin & Pattnaik, 2020;Fandeur et al., 2009), while the iron is a microelement that is essential for the body that is needed in the formation of blood, namely in hemoglobin (Hb), but if receiving too much Fe intake will cause various health problems such as poisoning, where vomiting, diarrhea, intestinal damage, hemochromatosis, cirrhosis, liver cancer, diabetes, heart failure, arthritis, impotence, infertility, hypothyroidism, and chronic fatigue occurs (Agustina et al., 2021).
Analysis of dissolved Cr 6+ content of liquid waste resulting from the remaining nickel mining of PT. Vale Tbk, found cr 6+ concentration exceeded water quality standards (Marzuki, 2016). Water quality analysis results at nickel mines in New Caledonia show high concentrations of Cr 6+  been developed by Waterloo Hydrogeologic, Canada, which adopts different methods up to and combined with modern visualization technology. Visual Mod ow software has the ability to simulate more complex conditions, can provide more realistic results, and is widely used to simulate groundwater ow, evaluate groundwater resources, and predict the spread of contaminants with MT3DMS packages (Adhikari & Mal, 2021;Ma et al., 2012;Rahman et al., 2018). Along with the development of technology, Visual Mod ow can also be combined with a genetic algorithm (GA) in solving groundwater problems (Widodo et al., 2018).
The aim of the study was to assess the concentrations of Cr 6+ and Fe that can be escaped from leftover rocks in nickel mining waste dumps and conduct groundwater ow simulations using Visual Mod ow as well as MT3DMS codes used to evaluate Cr 6+ and Fe migrations in groundwater over 20 years across a variety of scenarios. This is the rst time that water ow modeling and contaminant transport (Cr 6+ and Fe) have been conducted at nickel mining areas in Southeast Sulawesi. The results of this study may have implications for planning the prevention or mitigation of groundwater pollution around nickel mining areas and provide input to the government on the determination of regulations in the design and layout of good waste dumps in nickel mining.

Study Area
The study area is located in the subwatershed of Langgikima Subdistrict, North Konawe Regency which covers an area of 56.25 km2 from 3°13'00'' -3°15'00'' South Latitude and 122°14'2.8'' -122°17'10'' East Longitude (Figure 1), where the Tobimeitta River ows northward while the Sari Mukti River ows westward. Average annual temperature -average 29.8° and average annual rainfall -averages 2404.9 mm. Waste dumps are located on the western part of the settlement with a general slope of high in the Southwest and low in the Northeast. The elevation is highest at 800 masl and the lowest is 25 masl. The slope in the west is quite steep, ranging from 25 -49% and revolves from 0 -2% in the East of the research area.

Soil analysis in waste dump
Overburden (OB) is collected from nickel mining waste dumps as many as 5 samples and one soil sample is taken from settlements. Soil samples (about 1 kg each) are collected in HDPE plastic and used in geochemical studies and TCLP studies (Toxicity characteristic leaching procedures)

Characteristics of overburden
Geochemical studies include the characteristics of residential and overburdened soils in nickel mining waste dumps. Soil samples that have been collected are dried into an electric oven for ± 105°C, then ground into a dish mill machFluorescenceine for ± 8 minutes to the size of 200 mesh. Soil samples that have been nely tested XRF (X-Ray Fluorescence Spectroscopy) using epsilon 4 tools to nd out the content of elements and minerals in the soil.

Toxicity characteristic leaching procedure (TCLP)
The TCLP study of the USEPA method (1990) was used to determine the potential for cr 6+ and fe entangler in soil samples. Sodium acetate 1 M is used as an extraction uid and the pH is maintained at 4.99. Then 100 grams of test samples are dissolved with 2 liters of extract solution according to the pH of the speci ed test sample. The solution is stored in a glass bottle and put into an extractor with a spin at 30 ± 2 rpm at a temperature of 25ºC for 18 hours. Then, the solution is ltered using Whatman class lter paper No. 42, the lter results are watered with HNO3 1: 6, then the metal is ready to be tested AAS (Atomic Absorption Spectrophotometer).

Hydrology Parameter
In groundwater modeling, it takes several important data that affect the result of models, one of which is a hydrological parameter. Hydrological parameters include evapotranspiration, run-off, and recharge.

Calculation of evaporation of thornthwaite method (Seiler & Gat, 2007):
Where, Etr is real evapotranspiration (mm/year); P is rainfall (mm/year); Tm is the average annual temperature (°C Where, Ro is run-off (cm/year); P is rainfall (mm/year); Tm is annual temperature (°C); A is watershed area (km2) Calculation of groundwater addition (recharge) using the Lerner formula in (Devy, 2019): Where, U is the addition of groundwater (mm / year); P is rainfall (mm/year); ETr is real evapotranspiration (mm/year); Ro is run-off (mm/year Where, Kxx, Kyy, Kzz is hydraulic conductivity in the direction of x, y and z (LT -1 ); h is potential head (L); W is charging or pumping per unit area (LT -1 ) Ss is speci c storage (LT -1 ) T is time (T).

Groundwater solute migration mathematical model
Transport of contaminants in groundwater is simulated using the CODE MT3DMS. In this model MT3DS is used to break down the transport of advection-dispersion contaminants from the groundwater ow model. Dissolved adsorption and chemical reactions of pollutants are not taken into account in this model given the complex reactions that may occur in the aquifer layer (Xie, 2021). Mathematical equations are used as follows: Where, C is the concentration of solution (M/L3); D is the dispersion coe cient (L2T); T is time (T); Vx is the average speed of linear groundwater (L/T).

Calibration model
After inputting data in groundwater modeling, it is continued with the model calibration, to match the calculated groundwater level with the measured. The value of groundwater recharge, boundary conditions, hydraulic parameters are adjusted within geological and hydrogeological limits. One of the criteria in matching the groundwater level is calculated and measured is by minimizing the value of RMS     Table 3.
Concentrations of Cr 6+ in surface water ranged from 0.0235 -0.0462 mg/L and Fe 0.6354 -0.8709 mg/L, the highest concentrations of Cr 6+ and Fe in surface water were found in sediment pond water samples (AP 2) of 0.0462 mg/L Cr 6+ and 0.8709 Fe. The high concentration of Cr 6+ is associated with the oxidation process of Cr 3+ to Cr 6+ (Equeenuddin & Pattnaik 2020). In water concentrations cr 6+ ranged from 0.0223 -0.0404 mg/L and Fe 0.1114 -0.8606 mg/L, The highest concentration is in the sample of drill wells (ATB 2) in the mine which is 0.0404 mg / L Cr6 + and 0.8608 mg / L Fe which indicates there has been pollution, according to Dokou (Dokou et al., 2018) the concentration of Cr6 + in the natural condition in ultrama c rocks ranges from 0.02 mg / L so that the water sample in the drill well (ATB 2), Twice as much as its natural condition, while fe concentration has been 2.8 times greater than the standard quality of Class I with a threshold of 0.3 mg / L. Conceptual models are a simpli cation of complex eld problems. Based on 52 drilling data, the research area rocks consist of weathering rocks (limonite and saprolite) and are based on ultrama c rocks. The limonite layer consists of clay and silt, which has a low permeability value so that it is considered an aquitard, saprolite layer is an aquifer in the research area consisting of a mixture of clay, silt, sand, gravel, and boulder, while ultrama c rock is the underlying layer of the model, the fracture part of this rock can pass water but with a small amount and is considered an aquitard. Based on hydrogeological conditions, aquifers of research areas include semi uncon ned aquifers.
Hydraulic boundary conditions that limit the modeling area, namely the groundwater divide boundary, are in the western to southwestern mountainous areas that make the mountains as a no ow boundary, while the timut model is limited by the Tobimeitta river and the northern part of the model is limited by the Sari Mukti river, both rivers are used as river boundary. Recharge limit in the research area is obtained from the calculation of hydrological parameters which is 899.23 mm / year. Figure 2 is a conceptual model of groundwater ow.

Groundwater Modeling
To understand the ow patterns of groundwater in the research area, modeling is carried out using Visual

Model Discretization
The model was built with easting coordinate dimensions along 7.5 km and Northing along 7.5 km with a total area of 56.25 km2. The model is made using a horizontal grid size of 50 × 50 meters and divided into 3 layers in a vertical direction with a maximum elevation of 820 meters and a minimum of -80 meters. Details of the model grid design can be seen in Figure 3. Property input data is described in Table  4, which is obtained from various references. Because there has been no detailed hydrogeological study in the research area, aquifer property data is obtained from several literatures. The water system in the model is assumed to be a porous medium, which has homogeneous and isotropic properties in a horizontal direction (Kx = Ky), while in the vertical direction the hydraulic conductivity value is assumed to be less than 1 exponent order than the horizontal hydraulic conductivity value.

Flow model calibration
The   year. From these results, it is proven that the addition of a compacted clay layer on the basis of the waste dump can suppress the plume spread distance. This is because the value of hydraulic conductivity plays a big role in reducing the distance of plume distribution in contaminant transport that only takes into account the advection-dispersion process (Torres, 2020).
But the Concentration of Cr 6+ of 0.038 mg / L and Fe of 0.75 mg / L was found in the Observation well 1 which is 50 m away from the waste dump in the 20th year, which shows plume can still penetrate the compacted clay layer, it is also due to the thin aquifer layer that ranges from 3-28 m so that in modeling the aquifer layer is quickly saturated with contaminants. Plume migration in Scenario-B did not reach the Observation 3 well which is 500 meters from the waste dump in the 20th year of the simulation.

Conclusions
Numerical simulations were used to model the ow and transport of contaminants at nickel mine sites. Model calibration is used to nd optimal parameters and minimize computer calculation errors with eld survey data so as to help us in predicting contaminant transport over the next 20 years. The results of TCLP testing showed cr 6+ and fe that can be separated from waste rocks in the waste dump are so small that the concentration of pollutant sources used is the result of testing water samples in sediment ponds that accommodate leaching results from waste dumps with the content of Cr 6+ 0.0462 mg / L which is still below the Class I quality standard while fe content has exceeded the Class I quality standard of 0.8709 mg / L.
The results of MT3DMS simulation for 20 years, revealed that Cr 6+ and Fe can migrate rapidly in the aquifer layer, but when the addition of a safety layer in the form of a compacted clay layer on the basis of a waste dump that has a conductivity value of 1 × 10 -9 m / s, resulting in the distance and speed of plume spread is shorter even the spread distance is not up to 500 meters from the waste dump and plume has not passed the mine boundary in the 20th year simulation, So as to create sustainable exploitation of mining resources in accordance with the conservation of southeast Sulawesi's natural environment, the author strongly encourages the mining industry to use a compacted clay layer at the base of the waste dump and pay attention to the layout and distance of waste dumps with settlements. z