Paper · MERC
A Bench-Scale Woodchip-Enhanced Bioelectrochemical Denitrification Remediation Wall for Simulating Nitrate-Contaminated Groundwater In Situ Treatment
Summary
What the paper reports, in its own words.
Abstract
Excessive nitrogen fertilizer use has resulted in growing nitrate contamination of groundwater. In this study, an in situ bioelectrochemical reactor (isBER) reinforced with woodchips was developed for the treatment of actual nitrate-contaminated groundwater. During the 75-day experiment, the denitrification performance, grid permeability, and microbial community structure were investigated under different flow rates and current densities. The reactor achieved a remarkable nitrate removal efficiency of 97.6% ± 0.4% and a rate of 2.09 ± 0.14 mg-N/(L·h). These results were obtained at a temperature of 18.5 ± 0.8 °C, a current density of 350 mA/m², and a flow rate of 10 cm/d. Notably, the reactor can adapt to a wide flow-rate range of 5~20 cm/d and the operation proceeded smoothly without any blockages. Furthermore, the cathode module demonstrated enrichment of hydrogen autotrophic denitrifying bacteria (Pseudomonas, Stenotrophomonas) and heterotrophic denitrifying bacteria (Brucella, Enterobacteriaceae). Conversely, the anode module exhibited relatively high enrichment levels of aerobic microorganisms and lignin-degrading bacteria (Cellvibrio). The research results can provide novel insights and technical support for in situ remediation of groundwater nitrate contamination.
AI summary
82% confidenceThis bench-scale study demonstrates a woodchip-enhanced bioelectrochemical reactor (isBER) for in situ nitrate remediation of groundwater at 18.5 °C. Operating at 350 mA/m² and 10 cm/d flow rate, the system achieved 97.6% NO₃⁻-N removal efficiency (2.09 mg-N/(L·h)) from 30 mg/L spiked groundwater. The reactor tolerated flow rates from 5–20 cm/d without blockage, with removal efficiency declining but absolute removal rate increasing at higher flows. Cathode components dominated denitrification; anodes enriched aerobic nitrifiers that consumed ammonia. Nitrite and ammonia accumulated transiently but remained low at optimal conditions.
Key insights
Mechanism Electrical stimulation at 350 mA/m² increased overall central-area removal rate from 0.73 to 1.20 mg-N/(L·h)—a 64% boost—by enabling cathode hydrogen autotrophy, anodic lignin decomposition, and elevated local temperature.
So what. Operators should apply current to low-temperature groundwater systems to overcome kinetic limitations; 350 mA/m² is a practical sweet spot for this geometry.
Trade-off Removal efficiency and removal rate trade off with flow rate: at 350 mA/m², efficiency dropped from 97.6% (5 cm/d) to 81.6% (20 cm/d), but rate rose from 1.20 to 4.16 mg-N/(L·h).
So what. Choose flow rate based on treatment goal: prioritize efficiency for low-concentration plumes; prioritize throughput for high-flux contamination.
Threshold Cathode component I removed 53.4% of NO₃⁻-N at 10 cm/d (stage 4), while cathode component II removed 97.6% cumulatively, indicating sequential removal along the flow path and sufficient residence time in the second cathode.
So what. Design in situ barriers with multiple cathode stages or sufficient electrode length to achieve deep polishing; single-pass removal is incomplete at practical flow rates.
Threshold NO₂⁻-N accumulation increased from stage 3 to 5 despite electrical stimulation, correlating with higher flow rates that shortened contact time, even though stage 3 showed lower nitrite than stage 2.
So what. Monitor nitrite breakthrough at high flow rates; if NO₂⁻-N exceeds drinking-water limits (0.1 mg/L), reduce flow or increase electrode area.
Performance Three-dimensional seepage in isBER achieved 2.09 mg-N/(L·h) at 18.5 °C, not significantly different (p=0.081) from prior 2D column work at 2.58 mg-N/(L·h) at 25 °C, suggesting geometry and mixing compensate for 6.5 °C temperature loss.
So what. 3D flow reactors may be more efficient than 2D columns for field deployment in cold climates; prioritize geometry optimization over heating.
Contradictions1
Tensions the paper itself points to. Each side quotes the passage it comes from.
Within the paper Nitrite accumulation behavior differs between current-off and current-on stages despite both having low removal efficiency.
One sideStage 2 (no current): NO₂⁻-N initially increased then decreased as biofilm stabilized and nitrate removal improved
“During the operation of stage 1, the NO₂⁻-N accumulation in W3, W4, and W5 initially increased and then decreased. This trend occurred because the nitrate reduction efficiency decreased with the consumption of small-molecule organic matter ”
Other sideStages 3–5 (350 mA/m²): NO₂⁻-N gradually increased despite electrical stimulation and higher removal rates
“Furthermore, the NO₂⁻-N accumulation in W3, W4, and W5 gradually increased at stage 3~5. This increase may be attributed to the higher flow rate, which led to shorter contact reaction times among microorganisms, organic matter, and nitrate.”
Authors’ reading. Authors attribute stage 1–2 nitrite dynamics to biofilm maturation and organic matter availability, but stage 3–5 nitrite rise to flow-rate-induced kinetic limitation. The contradiction is resolved by recognizing that stages 3–5 operate at higher flow rates (5→10→20 cm/d) than stage 2 (5 cm/d), so the mechanism is consistent: faster flow = shorter contact time = incomplete NO₂⁻ reduction. Stage 2 and stage 3 both use 5 cm/d, so the difference is current application, which accelerates NO₃⁻ reduction and lowers NO₃⁻ concentration, reducing inhibition of NO₂⁻ reductase.
Opportunity areas6
Specific gaps in this paper, with an experiment that would close each.
High N₂O production and accumulation were not measured or reported. Given that denitrification proceeds through NO₃⁻ → NO₂⁻ → NO → N₂O → N₂, and NO₂⁻ accumulation was observed at higher flow rates, N₂O may be a significant byproduct or even escape to atmosphere.
Figure 2 shows NO₂⁻-N accumulation increasing from stage 3 to 5 (higher flow rates), and text notes 'Microorganisms tend to reduce nitrate first to obtain more energy, resulting in NO₂⁻-N accumulation.' This incomplete reduction pathway suggests N₂O may form; the paper does not address it.
Try. Measure N₂O in gas phase (headspace or outlet gas) and dissolved N₂O in water at each sampling port (W1–W6) across all stages; quantify N₂O yield (mol N₂O per mol NO₃⁻ removed) as a function of flow rate and current density.
High Cathode potential and anode potential were not independently reported or varied. The paper applies a fixed current density (350 mA/m²) but does not measure or control electrode potentials, which are critical for determining whether hydrogen autotrophy, electrochemical reduction, or DNRA dominates.
Section 3.1 attributes removal to 'nitrate electrochemical reduction, hydrogen autotrophic denitrification, and DNRA,' but without potential data, the contribution of each pathway cannot be quantified. Text states 'electrochemical reduction of nitrate by the iron cathode' but provides no cathode potential.
Try. Measure cathode and anode potentials (vs. Ag/AgCl or SHE) at each stage; vary cathode potential from −0.6 to −1.2 V (vs. SHE) while holding current constant, and measure NO₃⁻ removal rate, NO₂⁻ accumulation, and NH₄⁺ production to deconvolute electrochemical vs. biological pathways.
Medium Dissolved oxygen (DO) in the reactor was not measured. The paper mentions that anodic oxidation generates dissolved oxygen and promotes aerobic nitrification in the anode, but DO concentration profiles are absent.
Text states 'The local aerobic microenvironment in the anode promoted the denitrification process of micro-aerobic microorganisms' and 'dissolved oxygen generated by anodic oxidation, which promoted the local enrichment of aerobic nitrifying bacteria in the anode components.' Without DO data, the spatial extent and magnitude of aerobic zones cannot be assessed.
Try. Install DO probes at W3, W4, W5 (and W6) and measure DO concentration and profiles across all stages; correlate DO with nitrification rate and ammonia consumption at the anode.
Medium Catholyte pH was not independently controlled or varied. The paper reports pH at sampling ports (7.88–7.97 at stage 4) but does not explore how pH affects NO₂⁻ reductase activity or DNRA vs. denitrification partitioning.
Text notes 'nitrite reductase activity is closely related to pH' and explains NO₂⁻ reduction via pH effects, but no pH-dependent experiments were performed. Inlet groundwater pH is 8.94; reactor pH drops to ~7.9, but the effect of further pH adjustment (e.g., 6.5–8.5) on nitrogen speciation is unknown.
Try. Vary inlet pH (or add buffer) to maintain cathode compartment pH at 6.5, 7.0, 7.5, 8.0, and 8.5 while holding current at 350 mA/m² and flow at 10 cm/d; measure NO₃⁻, NO₂⁻, and NH₄⁺ profiles to identify optimal pH for complete denitrification.
Medium C/N ratio was not systematically varied. The paper notes that C/N ratio drives DNRA vs. denitrification partitioning, but all experiments used the same woodchip loading (80.2 g anode, 430.7 g cathode) and inlet TOC (1.32 mg/L).
Text states 'DNRA, a process positively correlated with the C/N ratio' and discusses how woodchip organic matter release affects C/N, but no experiments tested different woodchip masses, types, or inlet organic carbon concentrations to optimize the trade-off between denitrification and ammonia recovery.
Try. Test woodchip loading at 50%, 100%, and 150% of baseline (anode: 40–120 g; cathode: 215–650 g) at 350 mA/m² and 10 cm/d; measure NO₃⁻, NO₂⁻, NH₄⁺, and TOC to identify C/N ratio that maximizes denitrification while minimizing NO₂⁻ and NH₄⁺ accumulation.
Medium Long-term stability and electrode/woodchip lifespan were not evaluated. The experiment ran only 75 days; the paper acknowledges that 'the service life of woodchips and anodes may become limiting factors for continuous operation' but provides no data.
Text states 'In practical applications, the service life of woodchips [42] and anodes [12] may become limiting factors for continuous operation, requiring further evaluation of their lifespan and scheduled replacement.' No degradation curves or replacement schedules are provided.
Try. Extend operation to 6–12 months; measure electrode mass loss, woodchip carbon content, and removal rate decline over time; establish replacement intervals and cost per unit nitrogen removed.
Generated by MESSAI extraction pipeline · review against source PDF
Extracted values
Every value pulled from the paper, with its source, so each number can be checked against the text.
n = 84 extracted values
Select a citation to see the quote, page and confidence behind a value.
28 of 82 extracted rows modelable
- 39 not canonicalised (no canonical slug)
- 14 no SI value
- 1 contradicts other papers
| Parameter | Value | Unit | Confidence | Source | Notes | Modelable |
|---|---|---|---|---|---|---|
| Coulombic efficiency | 97.6 | % | 80% | llm | ||
| Temperature | [object Object] | °C | 80% | llm | ||
| Ca2+ concentration in groundwater | 78 | mg/L | 95% | icp-aes | "78.00 mg/L of Ca2+" · substrate=actual_groundwater_from_beijing | Excluded: not canonicalised (no canonical slug) |
| Total Organic Carbon | 1.32 | mg/L | 95% | toc analyzer | "with a TOC concentration of 1.32 ± 0.24 mg/L" · substrate=nitrate | Modelable |
| Mn2+ concentration in groundwater | 0.41 | mg/L | 95% | icp-aes | "0.41 mg/L of Mn2+" · substrate=actual_groundwater_from_beijing | Excluded: not canonicalised (no canonical slug) |
| Mg2+ concentration in groundwater | 32.7 | mg/L | 95% | icp-aes | "32.72 mg/L of Mg2+" · substrate=actual_groundwater_from_beijing | Excluded: not canonicalised (no canonical slug) |
| Flow rate | 10 | cm/d | 95% | extracted | "and a flow rate of 10 cm/d" · condition_hint=optimal flow rate for stage 4 | Excluded: no SI value |
| Nitrate Removal Efficiency | 97.6 | % | 95% | water sampling and ultraviolet spectrophotometry | "The reactor achieved a remarkable nitrate removal efficiency of 97.6% ± 0.4%" · T=18.5°C, substrate=nitrate, steady_state | Modelable |
| Cl− concentration in groundwater | 31.9 | mg/L | 95% | ion chromatography | "31.91 mg/L of Cl−" · substrate=actual_groundwater_from_beijing | Excluded: not canonicalised (no canonical slug) |
| Medium porosity | 45.9 | % | 95% | calculation from reactor volume | "the medium porosity (ηe) was calculated to be 45.9%" · condition_hint=reactor design parameter | Excluded: not canonicalised (no canonical slug) |
| Fe2+ concentration in groundwater | 1.9 | mg/L | 95% | icp-aes | "1.90 mg/L of Fe2+" · substrate=actual_groundwater_from_beijing | Excluded: not canonicalised (no canonical slug) |
| SO42− concentration in groundwater | 63.4 | mg/L | 95% | ion chromatography | "63.44 mg/L of SO42−" · substrate=actual_groundwater_from_beijing | Excluded: not canonicalised (no canonical slug) |
| NO3−-N concentration in groundwater | 2.08 | mg/L | 95% | ultraviolet spectrophotometry | "The main components detected in the groundwater included 2.08 mg/L of NO3−-N" · substrate=actual_groundwater_from_beijing | Excluded: not canonicalised (no canonical slug) |
| pH | 8.94 | — | 95% | ph meter | "The pH of the nitrate-contaminated groundwater was measured to be 8.94 ± 0.08" · pH(an)=8.94, substrate=nitrate | Modelable |
| Current Density (Areal) | 350 | mA/m² | 95% | extracted | "a current density of 350 mA/m2" · condition_hint=optimal current density for stage 3-5 | Modelable |
| Spiked NO3−-N concentration | 30 | mg/L | 95% | extracted | "resulting in a finalNO3−-N concentration of 30 mg/L" · substrate=nitrate | Excluded: not canonicalised (no canonical slug) |
| Nitrate Removal Efficiency | 2.09 | mg-N/(L·h) | 95% | water sampling and ultraviolet spectrophotometry | "and a rate of 2.09 ± 0.14 mg-N/(L·h)" · T=18.5°C, substrate=nitrate, steady_state | Excluded: no SI value |
| Enterobacteriaceae relative abundance (cathode component I) | 2.3 | % | 90% | 16s rrna gene sequencing | "its relative abundance increases to 2.3% in cathode component I and 9.9% in cath…" · T=18.5°C, substrate=nitrate, steady_state | Excluded: not canonicalised (no canonical slug) |
| pH | 7.88 | — | 90% | ph meter | "with values of 7.94 ± 0.03, 7.97 ± 0.02, and 7.88 ± 0.10" · T=18.5°C, pH(an)=7.88, substrate=nitrate, steady_state | Modelable |
| Permeability coefficient (stage 3) | 1.72 | m/d | 90% | darcy's law calculation | "to 1.72 ± 0.47 m/d (stage 3)" · T=18.5°C, substrate=nitrate, steady_state | Excluded: not canonicalised (no canonical slug) |
| Brucella relative abundance (cathode component II) | 10.3 | % | 90% | 16s rrna gene sequencing | "in cathode components I and II, the relative abundances of Brucella are 4.8% and…" · T=18.5°C, substrate=nitrate, steady_state | Excluded: not canonicalised (no canonical slug) |
| Permeability coefficient (stage 1, day 1) | 3.68 | m/d | 90% | darcy's law calculation | "the permeability coefficient K decreased from 3.68 m/d on the first day" · T=18.5°C, substrate=nitrate, startup | Excluded: not canonicalised (no canonical slug) |
| Sobs index (original sludge) | 344 | — | 90% | 16s rrna gene sequencing | "The Sobs indices of cathode component I and II were lower than that of the origi…" · substrate=activated_sludge | Excluded: not canonicalised (no canonical slug) |
| Permeability coefficient (stage 2) | 1.41 | m/d | 90% | darcy's law calculation | "the average permeability coefficient of the denitrification grid increased from …" · T=18.5°C, substrate=nitrate, steady_state | Excluded: not canonicalised (no canonical slug) |
| Bacterial count in effluent at 10 cm/d (low) | 2,180 | CFU/mL | 90% | 3m petriflmtm aerobic count plate | "Effluent at 10 cm/d 2.18 × 103~2.74 × 103" · T=18.5°C, substrate=nitrate, steady_state | Excluded: not canonicalised (no canonical slug) |
Literature priors
Against the literature
The author’s reported value (▼) sits on top of the literature distribution from MESS-Parameters. Values outside the band are flagged as outliers.
Methodology
What was the system and how was it run
Methodology
Methodology
A bench-scale 3D seepage tank (1000 mm × 200 mm × 300 mm) was constructed with five compartments separated by perforated screens. The center compartment housed two parallel carbon-rod anodes (each wrapped in porous tube with 80.2 g woodchips) and two stainless-steel mesh cathodes (each in porous groove with 430.7 g woodchips). Actual nitrate-contaminated groundwater (spiked to 30 mg/L NO₃⁻-N) was pumped through at flow rates of 5–20 cm/d. Current density was varied from 0 to 350 mA/m² across five 15-day stages. Water samples were collected daily at six ports (W1–W6) and analyzed for NO₃⁻-N, NO₂⁻-N, NH₄⁺-N, TOC, pH, and major ions. Permeability was calculated using Darcy's law from pressure-head measurements. Microbial communities were characterized by 16S rRNA gene sequencing of woodchip biofilms.
System configuration
| Field | Value |
|---|---|
| reactor type | BES |
| chamber config | 2-chamber |
Biological parameters
| Field | Value |
|---|---|
| dominant species | ["Pseudomonas","Enterobacter"] |
Performance metrics
| Field | Value |
|---|---|
| coulombic efficiency | 97.6 |
Operational parameters
| Field | Value |
|---|---|
| temperature | {"min":0.8,"max":18.5,"optimal":0.8} |
Electrochemical data
| Field | Value |
|---|---|
| max current density | 350 |
Operating conditions
14 runs extracted| Run | T°C | pHan | pHcat | Substrate | [S]mg/L | HRTh | Mode | R_extΩ | V_appV | Inoculum |
|---|---|---|---|---|---|---|---|---|---|---|
| v2-0bbd0575e725ab14 | 18.5 | 7.94 | — | nitrate | 30 | — | steady_state | — | 0 | — |
| v2-2220b0cc9ffc4101 | — | — | — | activated_sludge | — | — | — | — | — | — |
| v2-39cfac0adf2fb21f | — | — | — | nitrate | 30 | — | — | — | — | — |
| v2-44bdd28383343892 | 18.5 | 7.88 | — | nitrate | 30 | — | steady_state | — | 0 | — |
| v2-450068f7096095f7 | — | — | — | actual_groundwater_from_beijing | — | — | — | — | — | — |
| v2-4a00bd7c5c5ddb19 | 18.5 | 7.97 | — | nitrate | 30 | — | steady_state | — | 0 | — |
| v2-548fbd0258411ed2 | — | 8.94 | — | nitrate | 30 | — | — | — | — | — |
| v2-8d11c48599bb97a9 | — | — | — | nitrate | — | — | — | — | — | — |
| v2-9241ec33fdf7e7ae | 18.5 | — | — | nitrate | 30 | — | startup | — | 0 | — |
| v2-a2bc64f12ea1967f | 18.5 | — | — | nitrate | 30 | — | steady_state | — | — | — |
| v2-a979d8447acf51d4 | 18.5 | 8.94 | — | — | 30 | — | — | — | — | — |
| v2-b85e66544a047e46 | 18.5 | — | — | nitrate | 30 | — | steady_state | — | 0 | — |
| v2-d41d8cd98f00b204 | — | — | — | — | — | — | — | — | — | — |
| v2-d94c80160e0bd73a | 18.5 | — | — | — | — | — | — | — | — | — |
Outcomes
What was measured and what worked
Key takeaways
Coulombic efficiency
electrochemistry97.6percent
Pipeline
Raw extractions, artifacts, quality signal
Extraction quality
honesty signalCalibrate your trust in the extracted content against these signals — they're how the pipeline rated its own output for this paper.
Parameter completeness
75%
Fraction of expected parameters extracted
Data completeness
80%
Overall extraction coverage across fields
AI confidence
82%
Model self-reported confidence
Mean parameter confidence
90%
Average confidence across PaperParameter rows surfaced above
review against source PDF before citing · scores 0–1
Chart explorer
Move the inputs this paper used and see how the predicted outcome shifts.
Explore model
The reactor in 3D, with flow you can switch on and off.
AI insights
Contradictions, gaps and opportunities found by comparing this paper with the corpus.
Compare similar papers
The nearest papers by system, materials and organisms, with their outcomes side by side.
Source PDF
The paper as published.
No 3D model is mapped to this paper yet. The values extracted from it are listed under Extracted values.
In this record
Organisms, parameters and materials this paper reports. Each links to its page.
Parameters · 1
Keywords
- Denitrification
- Environmental remediation
- Groundwater remediation
- Groundwater
- Nitrate
- Contamination
Cite
Chen Yang, Yiheng Cao, Chuanping Feng (2025). A Bench-Scale Woodchip-Enhanced Bioelectrochemical Denitrification Remediation Wall for Simulating Nitrate-Contaminated Groundwater In Situ Treatment. Water. https://doi.org/10.3390/w17111593
- Journal
- Water