Artificial intelligence-based prediction model for the elemental occurrence form of tailings and mine wastes.

With the advent of the second industrial revolution, mining and metallurgical processes generate large volumes of tailings and mine wastes (TMW), which worsens global environmental pollution. Studying the occurrence of metal and metalloid elements in TMW is an effective approach to evaluating pollution linked to TMW. However, traditional laboratory-based measurements are complicated and time-consuming; thus, an empirical method is urgently needed that can rapidly and accurately determine elemental occurrence forms. In this study, a model combining Bayesian optimization and random forest (RF) approaches was proposed to predict TMW occurrence forms. To build the RF model, a dataset of 2376 samples was obtained, with mineral composition, elemental properties, and total concentration composition used as inputs and the percentage of occurrence forms as the model output. The correlation coefficient (R), coefficient of determination, mean absolute error, root mean squared error, and root mean squared logarithmic error metrics were used for model evaluation. After Bayesian optimization, the optimal RF model achieved accurate predictive performance, with R values of 0.99 and 0.965 on the training and test sets, respectively. The feature significance was analyzed using feature importance and Shapley additive explanatory values, which revealed that the electronegativity and total concentration of the elements were the two features with the greatest influence on the model output. As the electronegativity of an element increases, its corresponding residual fraction content gradually decreases. This is because the solubility typically increases with the solvent's polarity and electronegativity. Overall, this study proposes an RF model based on the nature of TMW that can rapidly and accurately predict the percentage values of metal and metalloid element occurrence forms in TMW. This method can minimize testing time requirements and help to assess TMW pollution risks, as well as further promote safe TMW management and recycling.

Evaluating the metal recovery potential of coal fly ash based on sequential extraction and machine learning.

Given the depletion of metal resources and the potential leaching of toxic elements from solid waste, secondary recovery of metal from solid waste is essential to achieve coordinated development of resources and the environment. In this study, hybrid models combining the gradient boosting decision tree and particle swarm optimization algorithm were constructed and compared based on two different datasets. Additionally, a new, quantitative evaluation index for metal recovery potential (MRP) was proposed. The results showed that the model constructed using more elemental properties could more accurately predict metal fractions in coal fly ash (CFA) with an R2 value of 0.88 achieved on the testing set. The MRP index revealed that the DAT sample had the greatest recovery potential (MRP = 43,311.70). Ca was easier to recover due to its high concentration and presence mostly in soluble fractions. Model post-analysis highlighted that the elemental properties and total concentrations generally exerted a greater influence on the metal fractions. The innovative evaluation strategy based on machine learning and sequential extraction presented in this work provides an important reference for maximizing metal recovery from CFA to achieve environmental and economic benefits with the goal of sustainable development.

Fluorides immobilization through calcium aluminate cement-based backfill: Accessing the detailed leaching characterization under torrential rainfall.

Urbanization and economic development have increased the demand for fertilizers to sustain food crop yields. Huge amounts of by-products, especially phosphogypsum (PG), are generated during the wet processing of rock phosphate to produce fertilizers. Chronic exposure to fluoride in phosphogypsum in groundwater as a result of the weathering of fluoride-containing waste poses a significant health risk to millions of people. We propose a method for using calcium aluminate cement (CAC) to remediate high fluoride contents in solid waste. Column leaching tests under harsh rainfall conditions confirmed the efficient fluoride immobilization capacity of a CAC binder. Although the fluoride concentrations in leachates during the first 1-2 days (1.25 mg/L) slightly exceeded the threshold of 1.00 mg/L, the concentrations over 3-28 days (ranging from 0.98 to 0.83 mg/L) consistently remained well within the acceptable range. Furthermore, our characterization and geochemical modeling revealed the fluoride retention mechanisms of CAC-stabilized PG under laboratory-simulated conditions of torrential rainfall. During leaching, physical encapsulation prevents fluoride from contacting leachate. However, an unfavorable pH value can cause the release of fluoride from the cement matrix, which is subsequently captured by aluminate hydrate through adsorption or co-precipitation. We quantified the carbon footprint of CAC for immobilizing 1 mg of fluoride in PG, obtaining a remarkably low value of 4.4 kg of CO2, in contrast to the emissions associated with the use of ordinary Portland cement (OPC). The findings suggest a unique opportunity for extensive PG remediation. This opportunity extends the horizons of achieving zero-waste emissions in the phosphorus industry and has practical significance in the context of reducing carbon emissions.

Ab initio calculations of CO2 adsorption on ß-C2S(100) and M3-C3S(001) surfaces: An exploration of early CO2 sequestration pathways.

Investigating CO2 sequestration in cement-based materials is significant for achieving carbon neutrality in the cement and concrete industries. The early CO2 sequestration pathways on cement-based materials are fundamental for CO2 sequestration, which is not clear. Towards this, the adsorption behavior of CO2 on ß-C2S(100) and M3-C3S(001) was investigated at the atomic level using density functional theory calculations, which were then compared with water adsorption results. The molecular adsorption configurations of CO2 on both ß-C2S(100) and M3-C3S(001) were tilted from their initial configurations due to the influence of surface Ca and O atoms. The CO2 adsorption energy on M3-C3S(001) and ß-C2S(100) were -0.458 eV and -0.426 eV, respectively, indicating adsorption on M3-C3S(001) was more energetically favorable. After CO2 adsorption, electrons were transferred from the surface to the CO2 molecule. Furthermore, the Ca-O bond orders of ß-C2S(100) and M3-C3S(001) after CO2 adsorption were maximally decreased by 2.79% and 6.99%, respectively. A more significant adsorption influence on surfaces was found for H2O, with more negative adsorption energy, more evident electron transfer, and a greater decrease in bond order. The CO2 adsorption on ß-C2S(100) and M3-C3S(001) were still spontaneous at 298 K and 1 atm. This study provides important theoretical insights into early CO2 sequestration at the atomic level, which has practical implications for the design of efficient CO2 sequestration technologies.

Utilisation of ultrasonic treatment to improve the soil amelioration property of coal fly ash.

Environment-friendly disposal of coal fly ash (CFA) is essential for sustainable development and cleaner production of electricity in thermal power plants. Although CFA has been employed for soil amelioration, direct application of CFA to soil may pose risks such as heavy metal contamination. This study investigated recycling of CFA through a novel method, which employs the ultrasonic treatment of CFA before its application. Physico-chemical properties of refuse dump soil and CFA were analysed. Subsequently, the effect of ultrasonic treatment on the physico-chemical properties of CFA was investigated. Different ultrasonic parameters (ultrasonic frequency, time interval, and temperature) were studied using response surface methodology. Finally, plant growth experiments were conducted to verify the feasibility of using ultrasonically treated CFA (UTCFA) for soil amelioration. The results show that untreated CFA cannot be used for soil amelioration due to its unsuitable high pH (10.20) and threatening concentrations of trace elements (6.80 mg/kg for Cadmium and 109.75 mg/kg for Arsenic). Ultrasonic treatment increases the soil amelioration properties of CFA by decreasing pH (to 8.50-9.20), decreasing concentrations of Cadmium and Arsenic (satisfying GB 15618-2018), and improving the water-holding capacity of CFA (reducing water loss). Plant indicators confirm the feasibility of using UTCFA for soil amelioration and suggest that the optimum UTCFA proportion is 20%. This study is a benchmark for the utilisation of ultrasonic treatment to improve the soil amelioration properties of CFA.

Lithium slag and fly ash-based binder for cemented fine tailings backfill.

This research work was an exploration of the feasibility of utilizing a lithium slag (LS) and fly ash (FA)-based binder for cemented fine tailings backfill (CFTB). Extensive experiments were conducted with different combinations of LS and ordinary Portland cement (OPC), along with FA as an additive. The unconfined compressive strength (UCS), micromorphology and slump values were analyzed. The results showed that (i) the LS and FA had a significant influence on the strength of binders. The OPC-LS-FA ratio of 2:1:1 appeared to be optimal with the highest strength and was referred as the LS and FA-based binder (LFB). (ii) The LFB significantly improved the UCS of the CFTB. The UCS values of CFTB specimens curing for 7,28 and 56 days reached 0.95 MPa,2.28 MPa and 3.37 MPa, respectively, with a 10 wt% content of LFB. The strength satisfied the strength requirement of backfill for supporting the surrounding rock of stopes in the Yinshan lead-zinc mine (0.8 MPa, 2.0 MPa, 3.0 MPa). (iii) The pore-filling effect of the secondary hydration products, which was mainly produced by LFB, played a significant role in the early stage (<7 days), while the pozzolanic activity worked mostly in the mid-long period (>28 days). (iv) The LFB reduced the slump value of CFTB slurry by 2.6%-9.4% compared with OPC when the mass concentration increased from 58% to 64%, which was acceptable to satisfy the requirements of better fluidity and less transportation resistance in the Yinshan lead-zinc mine. Therefore, the LFB could be utilized as an alternative cementitious material for CFTB, which also provides a safe and economical approach to recycle LS and FA in an underground mine.

Improved classification of soil As contamination at continental scale: Resolving class imbalances using machine learning approach.

The identification of arsenic (As)-contaminated areas is an important prerequisite for soil management and reclamation. Although previous studies have attempted to identify soil As contamination via machine learning (ML) methods combined with soil spectroscopy, they have ignored the rarity of As-contaminated soil samples, leading to an imbalanced learning problem. A novel ML framework was thus designed herein to solve the imbalance issue in identifying soil As contamination from soil visible and near-infrared spectra. Spectral preprocessing, imbalanced dataset resampling, and model comparisons were combined in the ML framework, and the optimal combination was selected based on the recall. In addition, Bayesian optimization was used to tune the model hyperparameters. The optimized model achieved recall, area under the curve, and balanced accuracy values of 0.83, 0.88, and 0.79, respectively, on the testing set. The recall was further improved to 0.87 with the threshold adjustment, indicating the model's excellent performance and generalization capability in classifying As-contaminated soil samples. The optimal model was applied to a global soil spectral dataset to predict areas at a high risk of soil As contamination on a global scale. The ML framework established in this study represents a milestone in the classification of soil As contamination and can serve as a valuable reference for contamination management in soil science.

A Deep Learning Approach for Chromium Detection and Characterization from Soil Hyperspectral Data.

High levels of chromium (Cr) in soil pose a significant threat to both humans and the environment. Laboratory-based chemical analysis methods for Cr are time consuming and expensive; thus, there is an urgent need for a more efficient method for detecting Cr in soil. In this study, a deep neural network (DNN) approach was applied to the Land Use and Cover Area frame Survey (LUCAS) dataset to develop a hyperspectral soil Cr content prediction model with good generalizability and accuracy. The optimal DNN model was constructed by optimizing the spectral preprocessing methods and DNN hyperparameters, which achieved good predictive performance for Cr detection, with a correlation coefficient value of 0.79 on the testing set. Four important hyperspectral bands with strong Cr sensitivity (400-439, 1364-1422, 1862-1934, and 2158-2499 nm) were identified by permutation importance and local interpretable model-agnostic explanations. Soil iron oxide and clay mineral content were found to be important factors influencing soil Cr content. The findings of this study provide a feasible method for rapidly determining soil Cr content from hyperspectral data, which can be further refined and applied to large-scale Cr detection in the future.

Mapping global distributions of clay-size minerals via soil properties and machine learning techniques.

Clay-size mineral is a vital ingredient of soil that influences various environment behaviors. It is crucial to establish a global distribution map of clay-size minerals to improve the recognition of environment variations. However, there is a huge gap of lacking some mineral contents in poorly accessible remote areas. In this work, machine learning (ML) approaches were conducted to predict the mineral contents and analyze their global abundance changes through the relationship between soil properties and mineral distributions. The average content of kaolinite, illite, smectite, vermiculite, chlorite, and feldspar were predicated to be 28.69 %, 22.30 %, 12.42 %, 5.43 %, 5.03 %, and 1.44 % respectively. Model interpretation showed that topsoil bulk density and drainage class were the most significant factors for predicting all six minerals. It could be seen from the feature importance analysis that bulk density notably reflected the distribution of 2:1 layered minerals more than that of 1:1 mineral. High drainage favored secondary minerals development, while low drainage was more benefited for primary minerals. Moreover, the content variation of different minerals aligned with the distribution of corresponding soil properties, which affirmed the accuracy of established models. This study proposed a new approach to predict mineral contents through soil properties, which filled a necessary step of understanding the geochemical cycles of soil-related processes.

The initial stages of cement hydration at the molecular level.

Cement hydration is crucial for the strength development of cement-based materials; however, the mechanism that underlies this complex reaction remains poorly understood at the molecular level. An in-depth understanding of cement hydration is required for the development of environmentally friendly cement and consequently the reduction of carbon emissions in the cement industry. Here, we use molecular dynamics simulations with a reactive force field to investigate the initial hydration processes of tricalcium silicate (C3S) and dicalcium silicate (C2S) up to 40 ns. Our simulations provide theoretical support for the rapid initial hydration of C3S compared to C2S at the molecular level. The dissolution pathways of calcium ions in C3S and C2S are revealed, showing that, two dissolution processes are required for the complete dissolution of calcium ions in C3S. Our findings promote the understanding of the calcium dissolution stage and serve as a valuable reference for the investigation of the initial cement hydration.

The rheological, mechanical and heavy metal leaching properties of cemented paste backfill under the influence of anionic polyacrylamide.

Anionic polyacrylamide (APAM) has widely been employed in backfill mining to accelerate the sedimentation of fine tailings particles and increase the concentration of tailings slurry. However, APAM inevitably remains in thickened tailings, leading to a nonnegligible influence on the rheological, mechanical, and heavy metal leaching properties of tailings-based cemented paste backfill (CPB). In an effort to solve these issues, the influences of APAM on CPB properties were examined in the present study. Experimental tests such as rheology, uniaxial compressive strength (UCS), toxicity leaching, and microscopy were conducted. The results showed that the presence of APAM first significantly increased the yield stress and viscosity of CPB slurry. APAM slightly improved the early UCS of CPB curing for 7 days but hindered the UCS development of samples cured for 28 days. Moreover, the presence of APAM restrained the hydration reaction, reduced the amounts of hydrated products, increased pore size, and loosed the microstructure of the test samples. Finally, the addition of APAM effectively reduced the leaching of Ag and As, while incremented that of Cu and slightly affected the leaching of Ba. In sum, these findings look promising for the safe production and environmental protection of the mining industry.

Influence of CaO on Physical and Environmental Properties of Granulated Copper Slag: Melting Behavior, Grindability and Leaching Behavior.

Due to its potential pozzolanic activity, granulated copper slag (GCS) has been proven to act as a supplementary cementitious material (SCM) after thermochemical modification with CaO. This modification method reduces cement consumption and CO2 emissions; however, the additional energy consumption and environmental properties are also not negligible. This paper aims to evaluate the economics and environmental properties of thermochemically modified GCS with CaO through the melting temperature, grindability, and heavy metal leaching characteristics. The X-ray fluorescence spectroscopy (XRF) results indicated that the composition of the modified GCS shifted to the field close to that of class C fly ash (FA-C) in the CaO-SiO2-Al2O3 ternary phase diagram, demonstrating higher pozzolanic activity. The test results on melting behavior and grindability revealed that adding CaO in amounts ranging from 5 wt% to 20 wt% decreased the melting temperature while increasing the BET surface area, thus significantly improving the thermochemical modification's economics. The unconfined compressive strength (UCS) of the cement paste blended with 20 wt% CaO added to the modified GCS after curing reached 17.3, 33.6, and 42.9 MPa after curing for 7, 28, and 90 d, respectively. It even exceeded that of Portland cement paste at 28 d and 90 d curings. The leaching results of blended cement proved that the heavy metal elements showed different trends with increased CaO content in modified GCS, but none exceeded the limit values. This paper provides a valuable reference for evaluating thermochemically modified GCS's economics and environmental properties for use as SCM.

The carbon uptake and mechanical property of cemented paste backfill carbonation curing for low concentration of CO2.

Large volumes of carbon dioxide are released during mining and ore resource development, and cemented paste backfill (CPB) materials are placed in the mined-out stopes where can be discharged from polluted air containing CO2. The construction of green mines and the goal of achieving carbon neutrality have become an inevitable choice for the mining industry to achieve the harmonious development of rational exploitation of resources and environmental protection. Against this background, to minimize the carbon emissions from the mining industry and promote the efficient utilization of CPB, this study investigated the carbon-uptake characteristics and mechanical property of CPB in underground mined-out stopes with 1.5 % concentration CO2. The results show that the carbonation curing (CC) increased the carbonation rate by nearly 4 times compared to natural curing, while the samples exhibited total carbonation within 28 days. This indicates that CO2 uptake could occur within the CPB. The CO2 was absorbed as calcium carbonate minerals, and each ton of CPB can ideally absorb about 78.4 kg CO2 and treat 2600 m3 of dirty air in the mined-out stopes. The increase in early uniaxial compressive strength (UCS) during CC required a higher cement concentrate, and the CC would retard the development of later compressive strength. Microstructure analysis indicated that the CC refined the pore structure and reduced the porosity of the CPB. It also affected the crystal growth and distribution of hydration and carbonation products, further influencing the difference in strength. In summary, CPB technology can potentially be useful during carbon uptake and may assist in mitigating carbon emissions from the mining industry and promoting environment friendly development.

Hydration and Mechanical Properties of Blended Cement with Copper Slag Pretreated by Thermochemical Modification.

The application of granulated copper slag (GCS) to partially replace cement is limited due to its low pozzolanic activity. In this paper, reconstituted granulated copper slag (RGCS) was obtained by adding alumina oxide (Al2O3) to liquid copper slag. Blended cement pastes were formulated by a partial substitute for ordinary Portland cement (OPC) with the RGCS (30 wt%). The pozzolanic activity, mechanical development, and the microstructure were characterized. The results show that 5-10 wt% Al2O3 contributes to the increase in magnetite precipitation in RGCS. The addition of Al2O3 alleviates the inhibition of C3S by RGCS and accelerates the dissociation of RGCS active molecules, thus increasing the exothermic rate and cumulative heat release of the blended cement pastes, which are the highest in the CSA10 paste with the highest Al2O3 content (10 wt%) in RGCS. The unconfined compressive strength (UCS) values of blended cement mortar with 10 wt% Al2O3 added to RGCS reach 27.3, 47.4, and 51.3 MPa after curing for 7, 28 and 90 d, respectively, which are the highest than other blended cement mortars, and even exceed that of OPC mortar at 90 d of curing. The pozzolanic activity of RGCS is enhanced with the increase in Al2O3 addition, as evidenced by more portlandite being consumed in the CSA10 paste, forming more C-S-H (II) gel with a higher Ca/Si ratio, and a more compact microstructure with fewer pores than other pastes. This work provided a novel, feasible, and clean way to enhance the pozzolanic activity of GCS when it was used as a supplementary cementitious material.

Concrete Strength Prediction Using Different Machine Learning Processes: Effect of Slag, Fly Ash and Superplasticizer.

Blast furnace slag (BFS) and fly ash (FA), as mining-associated solid wastes with good pozzolanic effects, can be combined with superplasticizer to prepare concrete with less cement utilization. Considering the important influence of strength on concrete design, random forest (RF) and particle swarm optimization (PSO) methods were combined to construct a prediction model and carry out hyper-parameter tuning in this study. Principal component analysis (PCA) was used to reduce the dimension of input features. The correlation coefficient (R), the explanatory variance score (EVS), the mean absolute error (MAE) and the mean square error (MSE) were used to evaluate the performance of the model. R = 0.954, EVS = 0.901, MAE = 3.746, and MSE = 27.535 of the optimal RF-PSO model on the testing set indicated the high generalization ability. After PCA dimensionality reduction, the R value decreased from 0.954 to 0.88, which was not necessary for the current dataset. Sensitivity analysis showed that cement was the most important feature, followed by water, superplasticizer, fine aggregate, BFS, coarse aggregate and FA, which was beneficial to the design of concrete schemes in practical projects. The method proposed in this study for estimation of the compressive strength of BFS-FA-superplasticizer concrete fills the research gap and has potential engineering application value.

Ab initio calculation of the adsorption of As, Cd, Cr, and Hg heavy metal atoms onto the illite(001) surface: Implications for soil pollution and reclamation.

Elucidating the mechanisms of heavy metal (HM) adsorption on clay minerals is key to solving HM pollution in soil. In this study, the adsorption of four HM atoms (As, Cd, Cr, and Hg) on the illite(001) surface was investigated using density functional theory calculations. Different adsorption configurations were investigated and the electronic properties (i.e., adsorption energy (Ead) and electron transfer) were analyzed. The Ead values of the four HM atoms on the illite(001) surface were found to be As > Cr > Cd > Hg. The Ead values for the most stable adsorption configurations of As, Cr, Cd, and Hg were -1.8554, -0.7982, -0.3358, and -0.2678 eV, respectively. The As atoms show effective chemisorption at all six adsorption sites, while Cd, Cr, and Hg atoms mainly exhibited physisorption. The hollow and top (O) sites were more favorable than the top (K) sites for the adsorption of HM atoms. The Gibbs free energy results show that the illite(001) surface was energetically favorable for the adsorption of As and Cr atoms under the influence of 298 K and 1 atm. After adsorption, there was a redistribution of positions and reconfiguration of the chemical bonding of the surface atoms, with a non-negligible influence around the upper surface atoms. Bader charge analysis shows electrons were transferred from the surface to the HM atoms, and a strong correlation between the valence electron variations and the adsorption energy was observed. HM atoms had a high electronic state overlap with the surface O atoms near the Fermi energy level, indicating that the surface O atoms, though not the topmost atoms around the surface, significantly influence HM adsorption. The above results show illite(001) preferentially adsorbed As among all four investigated HM atoms, indicating that soils containing a high proportion of illite might be more prone to As pollution.

Resistance Loss in Cemented Paste Backfill Pipelines: Effect of Inlet Velocity, Particle Mass Concentration, and Particle Size.

Cemented paste backfill (CPB), a technology placing the solid waste into mined-out stopes in the mine through pipeline transportation, has been widespread all over the world. The resistance loss is an important parameter for pipeline transport, which is significantly affected by the slurry characteristics. However, the coupling effect of inlet velocity (IV), particle mass concentration (PMC), and particle size (PS) has not been well evaluated and diagnosed. Hence, the CFD-based three-dimensional network simulation of CPB slurry flow in an L-shaped pipe at different combinations of the three parameters was developed using COMSOL Multiphysics software, and the findings were validated through a loop experiment. The results show that increasing IV and reducing PS will contribute to the homogeneity of the slurry in the pipeline, while the PMC presents little effect. The pipe resistance loss is positively correlated with IV and PMC and negatively correlated with PS. The sensitivity to the three parameters is IV > PS > PMC. In particular, the resistance loss is minimal at IV of 1.5 m/s, PMC of 72%, and PS of 1000 um. The calculation model of resistance loss regressed from simulation presented a high accuracy with an error of 8.1% compared with the test results. The findings would be important for the design of the CPB pipeline transportation and provide guidance in the selection of transfer slurry pumps, prepreparation of backfill slurry, and pipe blockage, which will improve the safety and economic level of a mine.

In Situ Remediation of Phosphogypsum with Water-Washing Pre-Treatment Using Cemented Paste Backfill: Rheology Behavior and Damage Evolution.

The accumulation of original phosphogypsum (OPG) has occupied considerable land resources, which have induced significant environmental problems worldwide. The OPG-based cemented paste backfill (OCPB) has been introduced as a promising solution. In this study, a water-washing pre-treatment was used to purify OPG, aiming to optimize the transport performance and mechanical properties of backfills. The overall results proved that in treated phosphogypsum-based cemented paste backfill (TCPB), the altered particle size distribution can alleviate the shear-thinning characteristic. The mechanical properties were significantly optimized, of which a maximum increase of 183% of stress value was observed. With more pronounced AE signals, the TCPB samples demonstrated better residual structures after the ultimate strength values but with more unstable cracks with high amplitude generated during loading. Principal component analysis confirmed the adverse effects of fluorine and phosphorus on the damage fractal dimensions. The most voluminous hydration products observed were amorphous CSH and ettringite. The interlocked stellate clusters may be associated with the residual structure and the after-peak AE events evident in TCPB, indicate that more significant stress should be applied to break the closely interlocked stitches. Ultimately, the essential findings in this experimental work can provide a scientific reference for efficient OPG recycling.

Particulate matter concentration from open-cut coal mines: A hybrid machine learning estimation.

Particulate matter (PM) emission is one of the leading environmental pollution issues associated with the coal mining industry. Before any control techniques can be employed, however, an accurate prediction of PM concentration is desired. Towards this end, this work aimed to provide an accurate estimation of PM concentration using a hybrid machine-learning technique. The proposed predictive model was based on the hybridazation of random forest (RF) model particle swarm optimization (PSO) for estimating PM concentration. The main objective of hybridazing the PSO was to tune the hyper-parameters of the RF model. The hybrid method was applied to PM data collected from an open-cut coal mine in northern China, the Haerwusu Coal Mine. The inputs selected were wind direction, wind speed, temperature, humidity, noise level and PM concentration at 5 min before. The outputs selected were the current concentration of PM2.5 (particles with an aerodynamic diameter smaller than 2.5 µm), PM10 (particles with an aerodynamic diameter smaller than 10 µm) and total suspended particulate (TSP). A detailed procedure for the implementation of the RF_PSO was presented and the predictive performance was analyzed. The results show that the RF_PSO could estimate PM concentration with a high degree of accuracy. The Pearson correlation coefficients among the average estimated and measured PM data were 0.91, 0.84 and 0.86 for the PM2.5, PM10 and TSP datasets, respectively. The relative importance analysis shows that the current PM concentration was mainly influenced by PM concentration at 5 min before, followed by humidity > temperature ≈ noise level > wind speed > wind direction. This study presents an efficient and accurate way to estimate PM concentration, which is fundamental to the assessment of the atmospheric quality risks emanating from open-cut mining and the design of dust removal techniques.

Flocculation-dewatering prediction of fine mineral tailings using a hybrid machine learning approach.

Polymer-assisted flocculation-dewatering of mineral processing tailings (MPT) is crucial for its environmental disposal. To reduce the number of laboratory experiments, this study proposes a novel and hybrid machine learning (ML) method for the prediction of the flocculation-dewatering performance. The proposed ML method utilizes principle component analysis (PCA) for the dimension-reduction of the input space. Then, ML prediction is performed using the combination of particle swarm optimisation (PSO) and adaptive neuro-fuzzy inference system (ANFIS). Monte Carlo simulations are used for the converged results. An experimental dataset of 102 data instances is prepared. 17 variables are chosen as inputs and the initial settling rate (ISR) is chosen as the output. Along with the raw dataset, two new datasets are prepared based on the cumulative sum of variance, namely PCA99 with 9 variables and PCA95 with 7 variables. The results show that Monte Carlo simulations need to be performed for over 100 times to reach the converged results. Based on the statistic indicators, it is found that the ML prediction on PCA99 and PCA95 is better than that on the raw dataset (average correlation coefficient is 0.85 for the raw dataset, 0.89 for the PCA99 dataset and 0.88 for the PCA95 dataset). Overall speaking, ML prediction has good prediction performance and it can be employed by the mine site to improve the efficiency and cost-effectiveness. This study presents a benchmark study for the prediction of ISR, which, with better consolidation and development, can become important tools for analysing and modelling flocculate-settling experiments.