Prediction Of Breakthrough Curves Research Articles

Utilising date palm fibres as a permeable reactive barrier to remove methylene blue dye from groundwater: a batch and continuous adsorption study.

This study aimed to utilise cheap and abundantly available date palm fibre (DPF) wastes for the remediation of methylene blue (MLB) dye-contaminated groundwater. The DPF adsorbents were first prepared, followed by various characterisation analyses, including surface morphology, functional groups, and material structure. Subsequently, the DPF adsorbents were applied in the batch and continuous adsorption studies to assess the MLB dye removal from aqueous environments. The batch adsorption study achieved 98% maximum removal efficiency with a contact time, adsorbents dosage, initial pH, temperature, particle size, initial dye concentration, and agitation speed of 105min, 3g/L, 7.0, 45°C, 0.075mm, 50mg/L, and 150rpm, respectively. Langmuir was the best-fitted isotherm model depending on a higher correlation coefficient (R2 = 0.985), with a maximum monolayer dye adsorption capacity (qmax) of 54.204mg/g. Additionally, the second order was the best-fitted kinetic model (R2 = 0.990), indicating that MLB dye was removed through chemisorption. Besides, the positive enthalpy change (ΔH°) and negative Gibb's free energy (ΔG°) values verified the endothermic process and spontaneous adsorption. According to the impact analysis of initial dye concentrations and flow rates on the permeable reactive barrier (PRB) performance in the continuous adsorption study using the Thomas, Belter, and Yan models, the experimental results and predicted breakthrough curves reflected an excellent agreement (R2 ≥ 0.8767) and a sum of squared errors (SSE) ≤ 0.4834. In short, the results demonstrated DPF as an effective adsorbent material in PRB technology.

Dynamic behavior of a fixed-bed adsorption column for acid red removal using natural Algerian cactos

In this work, the ability of Natural Algerian Cactus (NAC) to adsorb Acid Red (AR) from aqueous solution was investigated in a fixed-bed column. Scanning Electron Microscopy (SEM) was employed to examine the surface of the biosorbent, Fourier Transform Infrared Spectroscopy (FTIR) was utilized to analyze the chemical composition of NAC, and Brunauer-Emmett-Teller (BET) analysis was conducted to determine the surface area of NAC. We examined the effects of flow rate and influent concentration of AR on the efficiency of the fixed bed. Additionally, we analyzed the sorption of AR using the Thomas and Yoon-Nelson models at various dye concentrations and flow rates to predict breakthrough curves and determine the characteristic parameters of the fixed bed sorption column. The applied models were found to describe the NAC column biosorption process. The obtained results show that the natural Algerian cactus is an efficient Acid Red biosorbent from its aqueous solutions. A sorption capacity of 39.51 mg/g was recorded for the highest concentration (100 mg/l) at a flow rate of 0.8 mL/min while for the highest flow rate (2 mL/min), the sorption capacity attained 30.27 mg/g for a concentration of 50 mg/l. However, the longest breakthrough time was noted for the lowest concentration and flow rate respectively (tb=21h and te=38h for 50mg/l and 0.8 mL/min).

Evaluation of a Volume-Averaged Species Transport Model with Micro-Macro Coupling for Breakthrough Curve Prediction.

In porous water filters, the transport and entrapment of contaminants can be modeled as a classic mass transport problem, which employs the conventional convection-dispersion equation to predict the transport of species existing in trace amounts. Using the volume-averaging method (VAM), the upscaling has revealed two possible macroscopic equations for predicting contaminant concentrations in the filters. The first equation is the classical convection-dispersion equation, which incorporates a total dispersion tensor. The second equation involves an additional transport coefficient, identified as the adsorption-induced vector. In this study, the aforementioned equations were solved in 1D for column tests using 3D unit cells. The simulated breakthrough curves (BTCs), using the proposed micro-macro-coupling-based VAM model, are compared with the direct numerical simulation (DNS) results based on BCC-type unit cells arranged one-after-another in a daisy chain manner, as well as with three previously reported experimental works, in which the functionalized zeolite and zero-valent iron fillings were used as an adsorbent to remove phosphorous and arsenic from water, respectively. The disagreement of VAM BTC predictions with DNS and experimental results reveals the need for an alternative closure formulation in VAM. Detailed investigations reveal time constraint violations in all the three cases, suggesting this as the main cause of VAM's failure.

Bayesian predictive modeling for gas purification using breakthrough curves

This study proposes a predictive model for assessing adsorber performance in gas purification processes, specifically targeting the removal of chemical warfare agents (CWAs) using breakthrough curve analysis. Conventional parameter estimation methods, such as Brunauer-Emmett-Teller analysis, encounter challenges due to the limited availability of kinetic and equilibrium data for CWAs. To overcome these challenges, we implement a Bayesian parametric inference method, facilitating direct parameter estimation from breakthrough curves. The model’s efficacy is confirmed by applying it to H2S purification in a fixed-bed setup, where predicted breakthrough curves aligned closely with previous experimental and numerical studies. Furthermore, the model is applied to sarin with ASZM-TEDA carbon, estimating key parameters that could not be assessed through conventional experimental techniques. The reconstructed breakthrough curves closely match actual measurements, highlighting the model’s accuracy and robustness. This study not only enhances filter performance prediction for CWAs but also offers a streamlined approach for evaluating gas purification technologies under limited experimental data conditions.

A Fixed-Bed Study on the Feedsorption of BSA Using PKC: Toward the Sustainable Agrosorption of Protein-rich Waste for Enhancing Low Nutritional-value Feed

Palm Kernel Cake (PKC), a by-product of Malaysia's palm oil industry, stands out as a sustainable and cost-efficient feed ingredient. However, its use is nutritionally limited for monogastric animals due to high fiber content and anti-nutritional factors. This groundbreaking study explores the innovative process of 'feedsorption'—a term we have coined under the broader concept of 'agrisorption'—to enhance the nutritional value of PKC. By adsorbing protein-rich agricultural, poultry, and farm animal wastes, represented by bovine serum albumin (BSA), we aim to elevate PKC's protein content. Through detailed experimentation involving varying bed depths (1 cm to 2 cm) and initial BSA concentrations (100 to 500 µg/mL) within fixed-bed columns, our findings reveal that increased bed depths significantly prolong breakthrough and exhaustion times, highlighting improved adsorption efficiency. Yet, depths beyond 1 cm pose a risk of clogging. Higher BSA concentrations were found to accelerate breakthrough, indicating a stronger driving force capable of overcoming mass transfer resistance. The Modified Dose Response (MDR) model outperformed the Thomas model in accurately predicting breakthrough curves across different conditions. This study not only confirms the feasibility of feedsorption to bolster the nutritional profile of low-quality feed using protein-rich waste but also introduces a promising avenue for enhancing sustainable livestock nutrition.

An extended colloid filtration theory for modeling Escherichia coli transport in 3-D fracture networks

Microbial transport in fractured carbonate rock using enhanced solutions is a significant and neglected research topic in the literature. We propose an extended colloid filtration theory (CFT) combined with a particle-tracking following streamlines (PTFS) model for the rapid prediction of breakthrough curves (BTCs) and plumes of pathogens in three-dimensional (3-D) discrete fracture networks (DFNs). We adapted CFT in porous media to pathogen transport in fractures containing Terra Rossa (soil) deposits. As an example of the model capability, a simulation was used to predict the 3-D motion field and Escherichia coli count in groundwater originating from the Forcatella managed aquifer recharge (MAR) Facility (Brindisi, Italy) using a DFN composed of 3,900 fractures. In arid regions, MAR facilities are significant for sustaining basic human needs, such as freshwater supply for drinking and crop production. The Markov chain Monte Carlo (MCMC) technique was applied to E. coli counts in the collected water samples to increase data representativeness. The pathogen transport coefficients were further supported by batch filtration tests carried out in the CNR/IRSA Laboratory (Bari, Italy). The mean E. coli attachment rate coefficient of 0.15 × 10−8 m2 d–1 (sticking efficiency = 1.1 × 10−8 m) resulted in a 2.1 log10 removal in 600 m of reclaimed water filtration. The simulation output visualized the E. coli 3-D pathways in groundwater and the positions of contaminated groundwater spring outflows on Forcatella Beach. The simulation results agreed with the mean MCMC output of E. coli concentrations in bathing water under unperturbed geochemical and environmental flow and transport conditions. However, results indicate that concentrations of pathogenic strains, parasites, and enteric viruses may enter the marine environment of MAR sites during flood periods.

Investigations of effective dispersion models for electroosmotic flow with rigid and free boundaries in a thin strip

Reactive flow and transport in porous media is topic of intense research since decades. Since dispersion is one of the key parameters in solute transport, its accurate modeling is essential to avoid wrong predictions of flow and transport behavior. In this research, we investigate novel effective dispersion models for reactive transport of electrically charged chemical species in a thin, potentially evolving strip taking into account Taylor–Aris and electroosmotic‐induced dispersion as well as their cross‐coupling effects. We prove positivity of the dispersion coefficient and the existence and uniqueness of strong solutions in the fixed geometry setting. Moreover, we numerically investigate scenarios for both the fixed and evolving geometry situation. The simulation results illustrate the possibility of separating charged species, such that the findings of this study can lead to a better understanding of mixing and separation processes of charged solutes and an improved prediction of breakthrough curves. Finally, we study the limits of vanishing channel width, precipitation layer thickness, and molecular diffusion. We show convergence of the solutions to the corresponding limit cases such as a hyperbolic model or the fixed geometry case. From these results, we can rate the impact of distinct dispersion mechanisms and evaluate the necessity of a detailed modeling for different parameter regimes.

Assessment of kinetic and statistical models for predicting breakthrough curves of bio-colloid transport through saturated porous media

The microbial contamination of groundwater and its prevention is a widespread concern in developing countries. The present study simulated the transportation and interception of bio-colloid, Escherichia coli in porous media experimentally using packed columns to address certain aspects of underexplored sorption potential and validated using several kinetic models. The breakthrough curves obtained through experiments are observed to be in good agreement with its prediction using kinetic models namely Thomas, Yoon-Nelson and Modified Dose-Response. The overall comparisons of R2 among all the three models suggest that the MDR model fits more perfectly to experimental results. The combined effect of independent factors (column depth, particle size and alumina content) on response factors (maximum relative concentration and time required to achieve peak concentration) was investigated by using Box-Behnken Design under Response Surface Methodology (RSM) to check statistical significancy of independent factors. The R2 values for both response factors are observed to be 0.94 and 0.99, indicating a very high correlation between predicted and actual values. The results obtained in the present study also confirms that the travel distance and particle size are the statistically significant parameters that efficiently impact on sorption of Escherichia coli during their transport whereas the alumina content also affects the sorption but is observed to be a statistically non-significant.

Efficient hybrid modeling and sorption model discovery for non-linear advection-diffusion-sorption systems: A systematic scientific machine learning approach

This study presents a systematic machine learning approach for creating efficient hybrid models and discovering sorption uptake models in non-linear advection-diffusion-sorption systems. It demonstrates an effective method to train these complex systems using gradient-based optimizers, adjoint sensitivity analysis, and JIT-compiled vector Jacobian products, combined with spatial discretization and adaptive integrators. Sparse and symbolic regression were employed to identify missing functions in the artificial neural network. The robustness of the proposed method was tested on an in-silico data set of noisy breakthrough curve observations of fixed-bed adsorption, resulting in a well-fitted hybrid model. The study successfully reconstructed sorption uptake kinetics using sparse and symbolic regression, and accurately predicted breakthrough curves using identified polynomials, highlighting the potential of the proposed framework for discovering sorption kinetic law structures.

Mass transfer limiting step determination and a new modified empirical model for Th(ІV) biosorption in a fixed-bed column

In this paper, a new empirical model for predicting breakthrough curves of Th(ІV) biosorption using Ca-pretreated Cystoseria indica alga in a fixed bed column was presented. Also, the comparison of the proposed model with the Thomas, Yoon-Nelson, and Clark models showed that the new model had better agreement with the experimental data. Furthermore, the optimum flow rate of 2.5 mL/min with a maximum biosorption capacity of 213.20 mg/g was obtained using the proposed new model. Also, four mass transfer models, including internal mass transfer, external mass transfer, biosorption on active sites, and triple resistance, were used to determine the rate-limiting step and mass transfer mechanism. The internal mass transfer was considered the mass transfer controlling step. In addition, the sensitivity analysis of the mass transfer coefficients (kf, ks, kad) again showed that the internal resistance is the governing mass transfer resistance to the Th(ІV) biosorption process. The biosorption of Th(IV) from a real effluent by Ca-pretreated Cystoseria indica alga in a fixed-bed column was investigated to validate the new empirical model. The results demonstrated that the new empirical model could describe biosorption behavior more accurately than the Thomas model.

Removal of Mn (II) from aqueous solution via biosorption technology for a drinking water treatment plant: From laboratory-scale tests to semi-industrial scale predictions

Biosorbents have drawn increasing attention for removing pollutants on lab-scale systems. Although biosorbents are ecofriendly, their potential on large-scale columns has not been assessed yet. This study evaluated the potential of sugarcane bagasse (SB) for the removal of Mn (II) from aqueous media. Firstly, batch experiments were conducted to establish the equilibrium data and predominant adsorption mechanisms. Secondly, a mathematical model was proposed to predict the breakthrough curves in a fixed-bed column. Alternatively, lab- and pilot-scale removal experiments were performed to validate the accuracy of the proposed model. Once the model was validated on both scales, it was employed to predict breakthrough curves for a semi-industrial column under actual conditions of a drinking water treatment plant. Langmuir isotherm (qmax: 2.82 mg/g) and Langmuir kinetic model (R2>0.98 and x2<0.01) reproduced the experimental equilibrium and kinetic behavior. The mathematical model successfully replicated the lab- and pilot-scale breakthrough data (R2>0.95 and x2<0.1). Therefore, its prediction on the semi-industrial scale was reliable. However, short breakthrough and exhaustion times were obtained on this column, which hinders SB from applying in actual raw water. The described methodology can be employed to assess the potential of other biosorbents for industrial applications.

Water Cleaning by a Continuous Fixed-Bed Column for Cr(VI) Eco-Adsorption with Green Adsorbent-Based Biomass: An Experimental Modeling Study

In this study, chromate adsorption onto red peanut skin (RPS) was investigated in a fixed-bed column; FTIR, PZC, SEM, DLS, and BET were used to evaluate its adsorption properties. The experiments were conducted to determine the effect of physical parameters, including the inlet initial Cr(VI) concentration (100–500 mg L−1), bed height (10–20 cm), and feed flow rate (13.59–23.45 mL min−1). They were carried out to predict breakthrough curves. The adsorption capacity coefficients were determined using the most widely used Bohart–Adams model. It was tested to fit experimental data, for a better understand the dynamic behavior, and for further optimize column performance. It was found that the Cr(VI) uptake decreases when increasing the flow rate and that high chromate concentration and bed height consequently increase the column’s life span. A high column adsorption capacity can be achieved with a higher Cr(VI) concentration due to the higher driving force. The results indicated that the Bohart–Adams model provides a good description (R2 &gt; 0.98) of the experimental data of the Cr(VI)’s removal from the aqueous solution on the RPS suggesting that the surface diffusion is the rate-limiting step in the continues adsorption process.. Breakthrough adsorption capacity is crucial for comparing RPS with other similar materials. Indeed, possible mechanisms have been suggested for illustrating adsorption onto RPS. The obtained results showed significant potential of 26.23 mg g−1 of RPS on Cr(VI) elimination at a natural pH of 5.35. Furthermore, this global investigation allowed for the design of a promising low-cost material for the future scale-up of cleaning wastewater polluted by metal and determine the properly conditions for operating column adsorption. This material provides an economical, efficient means of eliminating pollutants, thus meeting the main aims of the UN Sustainable Development Goals (UN SDGs).

Anomalous transport in a porous medium with randomly packed ellipse cavities

We investigate the transport of nonreactive tracers in a binary porous medium with randomly packed ellipse fluid-filled cavities. Anomalous transport features, such as early arrival time and long tailing, are observed due to the high contrast in medium properties and highly complex structure of fluid velocity. We use a particle tracking method to quantify transport features of the domain. Then, a continuous time random walk (CTRW) framework builds on tracer transitions in time and space to represent an upscaled model. We study the effect of several key parameters on the anomalous transport process. The parameters include the cavity aspect ratio, porous background permeability, and the Peclet (Pe) number. With the increase in Pe, a longer tailing and a larger residence time are observed, which presents a stronger anomalous feature. A similar situation corresponds to decreased porous medium permeability, which results in wider breakthrough curves. A longer tailing arises in the case of more elongated cavity of larger aspect ratio. The purely advective transport in the medium is investigated at Pe = ∞. This is considered a limit case for the anomalous behavior of the system. One can refer to this case as the most extended tail possible for each cavity arrangement. The widest breakthrough curves for a Pe = ∞ correspond to larger aspect ratios of the cavity and a lower permeable matrix. We show that the upscaled CTRW model can closely predict breakthrough curves in a binary medium with randomly distributed ellipse cavities. These findings give new insight into transport in vesicular and vuggy porous media.

Prediction of breakthrough curves for multicomponent adsorption in a fixed-bed column using logistic and Gompertz functions

In order to facilely predict the multicomponent breakthrough curves and avoid the complicated numerical solution, this work proposes two empirical models based on the logistic and Gompertz functions. The S-shaped and bell-like curves correspond to the adsorption and displacement processes respectively. The equilibrium loading of each component can be calculated by integration of the measured breakthrough curves. The effects of the model parameters on the breakthrough curves are investigated. The applicability of the two empirical models is validated by three binary and four ternary adsorption systems, including the gas–solid and liquid–solid adsorption. The residual plot and coefficient of determination ( R 2 ) are used to evaluate their fitting quality. The results indicate that the fitting curves agree well with the experimental data and all of the residuals are distributed randomly. The five model parameters ( k , τ , k *, τ * and c ) are easily obtained by the nonlinear regression. For example, the fitting results are k = 1.37 × 10 −2 min −1 , τ = 292 min, k * = 1.25 × 10 −2 min −1 , τ * = 453 min and c = 85.3 for adsorption of n -butyl acetate and p -xylene on granular activated carbon. On the whole, the Gompertz model is superior to the logistic model in terms of the fitting accuracy. The significance of this work is to provide a simple and practical method for prediction of the multicomponent breakthrough curves.

Breakthrough Curves Prediction of Selenite Adsorption on Chemically Modified Zeolite Using Boosted Decision Tree Algorithms for Water Treatment Applications

This work describes an experimental and machine learning approach for the prediction of selenite removal on chemically modified zeolite for water treatment. Breakthrough curves were constructed using iron-coated zeolite adsorbent and the adsorption behavior was evaluated as a function of an initial contaminant concentration as well as the ionic strength. An elevated selenium concentration in water threatens human health and aquatic life. The migration of this metalloid from the contaminated sites and the problems associated with its high releases into the water has become a major environmental concern. The mobility of this emerging metalloid in the contaminated water prompted the development of an efficient, cost-effective adsorbent for its removal. Selenite [Se(IV)] removal from aqueous solutions was studied in laboratory-scale continuous and packed-bed adsorption columns using iron-coated natural zeolite adsorbents. The proposed adsorbent combines iron oxide and natural zeolite’s ability to bind contaminants. Breakthrough curves were initially obtained under variable experimental conditions, including the change in the initial concentration of Se (IV), and the ionic strength of solutions. Investigating the effect of these parameters will enhance selenite mobility retardation in contaminated water. Continuous adsorption experiment findings will evaluate the efficiency of this economical and naturally-based adsorbent for selenite removal and fate in water. Multilinear and non-linear regressions approaches were utilized, yet low coefficients of determination values were respectively obtained. Then, a comparative analysis of five boosted regression tree algorithms for a selenite breakthrough curve prediction was performed. AdaBoost, Gradient boosting, XGBoost, LightGBM, and CatBoost models were analyzed using the experimental data of the packed-bed columns. The performance of these models for the breakthrough curve prediction under different operation conditions, such as initial selenite concentration and ionic strength, was discussed. The applicability of these models was evaluated using performance metrics (i.e., Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and coefficient of determination (R2). The CatBoost model provided the best fit for a breakthrough prediction with a coefficient of determination R2 equal to 99.57. The k-fold cross-validation technique and the statistical metrics verify this model’s accurateness. A feature importance assessment indicated that Se (IV) initial concentration was the most influential experimental variable, while the ionic strength had the least effect. This finding was consistent with the column transport results, which observed Se (IV) sorption dependency on its inlet concentration; simultaneously, the ionic strength effect was negligible. This work proposes implementing machine learning-based approaches for predicting water remediation-associated processes. The significance of this work was to provide an alternative method for investigating selenite adsorption behavior and predicting the breakthrough curves using a machine-based approach. This work also highlighted the importance of management practices of adsorption processes involved in water remediation.

The Dynamic Behaviour of a Binary Adsorbent in a Fixed Bed Column for the Removal of Pb2+ Ions from Contaminated Water Bodies

In the search for a technically efficient and abundant adsorbent in water treatment processes, a bio-composite adsorbent derived from agricultural wastes has been identified as a potential candidate. In this study, eggshells and sugarcane bagasse were combined in varied proportions (1:0, 1:3, 1:1, 3:1 and 0:1) and applied as biosorbents in a lab-scale adsorption column. The effect of bed depth (4–12 cm) of the biosorbents was investigated which enabled the prediction of breakthrough curves for the removal of Pb (II) ions. The life span of the column was extended by increasing the bed depth of the column. The binary adsorbent of 1:3 weight ratio of &lt;75 µm particle size showcased the highest removal efficiency of 91% at a bed depth of 12 cm. The mass transfer zone (MTZ) increased with increasing bed depth with a minor portion of the bed left unused, signifying that the process was highly efficient. The Thomas model constant, KTh, decreased with increasing bed depth with the maximum amount of Pb adsorbed being 28.27 mg/g. With the Yoon–Nelson model, KYN decreased with an increase in τ as the bed height increased. In this study, a novel approach was adopted where the proposed methodology enabled the use of a bio-composite adsorbent in heavy metal removal. The findings of this research will aid in the design and optimisation of the pilot-scale operation of environmentally friendly treatment options for metal laden effluent.

Adsorptive removal of Pb(II) using nanostructured γ-alumina in a packed bed adsorber: Simulation using gPROMS.

In this work, convective-dispersive and pore volume and surface diffusion models have been used to analyze Pb(II) adsorption from an aqueous solution over a nanostructured γ-alumina adsorbent in a packed bed adsorber. The models encompassing partial differential equation and a linear algebraic equation coupled with isotherm have been simulated in gPROMS using the backward finite difference approach. The predicted breakthrough curves of Pb(II) adsorption concerning flow rate, initial metal concentration, and bed height were matched with the experimental data. The accuracy of model predictions was analyzed through statistical measures such as coefficient of determination (R2), root mean square error, and chi-squared value. The simulation results also predicted the axial dispersion, distribution coefficient, mass transfer coefficient, pore volume, and surface diffusion coefficient, which are, otherwise, difficult to measure experimentally and, in turn, have been used to assess the mass transfer characteristics of continuous Pb(II) adsorption. Additionally, the values of breakthrough time, exhaustion time, adsorption column capacity, and mass transfer zone were determined as a function of flow rate, bed height, and initial metal concentration. Surface and pore volume diffusions (10-11-10-10 m2/s) apparently controlled the continuous adsorption process, with surface diffusion being dominant. The transport parameters evaluated in the current study could be beneficial for the large-scale Pb(II)/nanostructured γ-alumina adsorption system. As evident from the successful simulation, the developed gPROMS program can also be applied to other adsorbate/adsorbent systems with a slight modification concerning the operating parameters.

Entity aware sequence to sequence learning using LSTMs for estimation of groundwater contamination release history and transport parameters

Groundwater contaminant sources identification and parameter estimation using the simulation–optimization (S/O) approach require numerous runs of the computationally expensive simulation model through the optimization algorithm. The computational cost can be effectively reduced by using a surrogate model which can accurately approximate the simulation model. With the advent of deep learning, Long Short-Term Memory (LSTM) networks, which are suitable to learn sequential data, are being increasingly applied to regression problems involving time dependencies. However, for the simultaneous contaminant source identification and parameter estimation problem, the surrogate model requires to establish a relationship between release histories at the source locations to concentration measurements at observation points subject to given values of aquifer parameters. In this study, a novel deep neural network framework Entity Aware Sequence to sequence learning using Long Short-Term Memory (EAS-LSTM) is proposed as a surrogate model which takes both sequential i.e., release histories at the source locations at different stress periods and static variables i.e., transport parameters as inputs to predict breakthrough curves (BTCs) at observation points. The proposed surrogate model is applied to a heterogeneous field scale aquifer with 4 zones. The Mean Squared Error (MSE) using EAS-LSTM is 0.0033 ppm2 which is significantly better than that of Kriging and Support Vector Regression (SVR) which are 0.048 ppm2 and 1.302 ppm2 respectively. Comparison of EAS-LSTM model performance with Kriging and SVR based models demonstrates its higher accuracy. Further, the optimization algorithms for inverse modelling, the performances of Multiverse Optimizer (MVO), Grey wolf optimization (GWO) and Particle swarm optimization (PSO) are investigated. It is observed that the combination of EAS-LSTM and MVO provides better results in comparison to other surrogate simulation optimization (SSO) models.

A new approach for scaling up fixed-bed adsorption columns for aqueous systems: A case of antibiotic removal on natural adsorbent

The scaling up of adsorption columns is a crucial step toward the removal of emerging pollutants from domestic and industrial wastewaters. In this study, a fixed-bed column to remove sulfamethoxazole packed with sugarcane bagasse was scaled up from laboratory (DL = 2.2 cm) to pilot unit (DP = 4.4 cm) using a scaling factor (K = 2). In addition to the basic similarity rules for scaling, three new criteria were proposed for the mass adsorbent, flow rate, and bed volume. Then, three lab-scale tests at flow rate of 5 mL/min and bed heights of 15, 25, and 35 cm were transferred to the pilot-scale column at flow rate of 20 mL/min and bed heights of 30, 50, and 70 cm, respectively. The breakthrough curves and the fixed-bed parameters (residence time tR, saturation time ts, adsorption capacity qe, volume of solution treated Vef, and removal percentage %R) obtained in both scales were compared to define their effect with the increase of scale. Finally, a mechanistic model was proposed to predict the breakthrough curves in both columns. The results exhibited that the breakthrough curves in the pilot-scale prolonged in time with higher breakthrough and saturation times than the laboratory breakthrough data. Additionally, tR, ts, and Vef changed in function of the K value used: ts and tR doubled their value in the pilot column or tRP=KtRL; Vef was eight times higher in the pilot column than the lab-column or VefP=K3VefL; qe and %R remained constant in both scales; these results were corroborated with the predicted breakthrough curves. Besides, the mechanistic model predicted with great precision the breakthrough data in both scales (R2>0.9), which means that the model can be used confidently for scaling up purposes. This study demonstrated new criteria which can be easily applied to scale up adsorption columns whit results that showed a correlation between both scales.

Measurement and prediction of breakthrough curve of low-concentration toluene adsorbed by activated carbon

People spend most of their time indoors. The volatile organic compounds (VOCs) in indoor air represented by toluene inevitably affect people's health. Activated carbon can effectively reduce indoor VOCs pollution. The experiments in the previous studies were carried out at the concentration of tens or hundreds of ppm, which could not reflect the actual adsorption performance of activated carbon in the actual conditions. In this study, a method was proposed to predict breakthrough curve of toluene at ~1 ppm. The method was based on Wheeler-Jonas model, the equilibrium adsorption capacity was predicted by the Freundlich equation, and the adsorption rate constant was obtained by Yoon-Nelson equation. The predicted breakthrough curves were in good agreement with the experimental data in the range of 0.5~4.0 ppm. The adsorption rate constant was predicted using the relationship with inlet concentration of toluene. When the breakout fraction was 20-85%, the relative deviation was 5.54% and 4.08% at 0.5 and 1.0 ppm. This method can predict the adsorption performance and service life of activated carbon for toluene at ppb~ppm level.