PERFORMANCE EVALUATION AND MACHINE LEARNING-ASSISTED PREDICTION OF A HYBRID BIOCHAR–ZEOLITE COLUMN REACTOR FOR TURBIDITY AND IRON REMOVAL FROM TROPICAL RIVER WATER

High turbidity and elevated dissolved iron are the two most persistent water quality problems facing tropical river communities in Kalimantan, Indonesia, where both parameters regularly exceed the limits set by the WHO. We investigated how well a continuous-flow hybrid biochar–zeolite column reactor could remove turbidity and dissolved iron at the same time, working through 100 controlled experimental runs using synthetic river water. Three operating variables were tested—influent turbidity (50, 150, and 300 NTU), hydraulic loading rate (0.3, 0.5, and 0.8 L min⁻¹)—and biochar-to-zeolite ratio (0.5:1, 1:1, and 2:1)—arranged in a factorial design. Iron adsorption equilibrium was described by Langmuir (qmax = 18.5 mg g⁻¹; KL = 0.62 L mg⁻¹; R² = 0.961) and Freundlich (Kf = 6.2 mg g⁻¹; n = 2.1; R² = 0.943) isotherms, and breakthrough dynamics were described by the Thomas kinetic model. Mean turbidity removal efficiency was 83.93 ± 5.10% (range: 74.52–93.18%) and mean iron removal efficiency was 72.06 ± 7.05% (range: 55.83–85.71%). Among all variables tested, hydraulic loading rate had by far the greatest influence, showing a strong inverse relationship with turbidity removal (r = −0.919) and iron removal (r = −0.848; both p < 0.001). We then built predictive models using multiple linear regression (MLR) and Random Forest regression (RFR, 200 estimators). The RFR model gave more accurate iron removal predictions (R² = 0.921, MAE = 1.89%), while MLR worked well enough for turbidity (R² = 0.859). Partial dependence plots and interaction heatmaps consistently pointed to 0.3 L min⁻¹and a 1:1 media ratio as the best overall operating condition. The results indicate that this hybrid system is a promising, low-cost pre-treatment option for decentralised rural water supply in tropical settings where both particulate and dissolved iron contamination occur together, pending confirmatory testing with natural river water containing dissolved organic matter and competing ions.

Authors

Institutions

Publication Details

Journal
Nigerian Journal of Technology
Published
2026-10-28
DOI
https://doi.org/10.67358/njt.2026.7023
Primary Topic
Adsorption and biosorption for pollutant removal
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

PERFORMANCE EVALUATION AND MACHINE LEARNING-ASSISTED PREDICTION OF A HYBRID BIOCHAR–ZEOLITE COLUMN REACTOR FOR TURBIDITY AND IRON REMOVAL FROM TROPICAL RIVER WATER

Lenie Marlinae, Tien Zubaidah
1 links
Nigerian Journal of Technology
Adsorption and biosorption for pollutant removal
article

PERFORMANCE EVALUATION AND MACHINE LEARNING-ASSISTED PREDICTION OF A HYBRID BIOCHAR–ZEOLITE COLUMN REACTOR FOR TURBIDITY AND IRON REMOVAL FROM TROPICAL RIVER WATER

Lenie Marlinae, Tien Zubaidah
article en

Abstract

High turbidity and elevated dissolved iron are the two most persistent water quality problems facing tropical river communities in Kalimantan, Indonesia, where both parameters regularly exceed the limits set by the WHO. We investigated how well a continuous-flow hybrid biochar–zeolite column reactor could remove turbidity and dissolved iron at the same time, working through 100 controlled experimental runs using synthetic river water. Three operating variables were tested—influent turbidity (50, 150, and 300 NTU), hydraulic loading rate (0.3, 0.5, and 0.8 L min⁻¹)—and biochar-to-zeolite ratio (0.5:1, 1:1, and 2:1)—arranged in a factorial design. Iron adsorption equilibrium was described by Langmuir (qmax = 18.5 mg g⁻¹; KL = 0.62 L mg⁻¹; R² = 0.961) and Freundlich (Kf = 6.2 mg g⁻¹; n = 2.1; R² = 0.943) isotherms, and breakthrough dynamics were described by the Thomas kinetic model. Mean turbidity removal efficiency was 83.93 ± 5.10% (range: 74.52–93.18%) and mean iron removal efficiency was 72.06 ± 7.05% (range: 55.83–85.71%). Among all variables tested, hydraulic loading rate had by far the greatest influence, showing a strong inverse relationship with turbidity removal (r = −0.919) and iron removal (r = −0.848; both p < 0.001). We then built predictive models using multiple linear regression (MLR) and Random Forest regression (RFR, 200 estimators). The RFR model gave more accurate iron removal predictions (R² = 0.921, MAE = 1.89%), while MLR worked well enough for turbidity (R² = 0.859). Partial dependence plots and interaction heatmaps consistently pointed to 0.3 L min⁻¹and a 1:1 media ratio as the best overall operating condition. The results indicate that this hybrid system is a promising, low-cost pre-treatment option for decentralised rural water supply in tropical settings where both particulate and dissolved iron contamination occur together, pending confirmatory testing with natural river water containing dissolved organic matter and competing ions.

Nigerian Journal of TechnologyVol. 45(3)
Lambung Mangkurat University (ID), Poltekkes Kemenkes Banjarmasin (ID)
Clean water and sanitation, Responsible consumption and production
Openalex Percentile: Top 25%
Adsorption and biosorption for pollutant removal
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon