NovaSolver›Arsenic Removal by Iron Oxide Adsorption Simulator Back
Water Treatment
Arsenic Removal by Iron Oxide Adsorption Simulator
Size the iron-oxide adsorption process used to treat arsenic-contaminated groundwater in Bangladesh, India, the US Southwest and other regions. Apply Langmuir or Freundlich isotherms, factor in pH and competing phosphate, and read off the adsorbent dose, column diameter and annual operating cost needed to reach the WHO 10 μg/L guideline.
Parameters
Influent As concentration
μg/L
As in raw groundwater. WHO guideline = 10 μg/L
Target effluent As
μg/L
Treated-water target. WHO/EPA limit = 10 μg/L
Adsorbent
Sets qmax and Langmuir/Freundlich constants
Isotherm model
Langmuir = monolayer; Freundlich = heterogeneous
pH
As(V) optimum at pH 7-8.5; drops on either side
Competing PO₄³⁻
mg/L
Phosphate competes for the same =Fe-OH sites
Flow rate
m³/h
Plant throughput. Household ≈ 1 m³/h, municipal 50-500 m³/h
Results
—
Loading q_e (mg-As/g)
—
Adsorbent dose (g/L)
—
Daily use (kg/day)
—
Column diameter (m)
—
Influent As (μg/L)
—
Annual cost (USD)
—
Adsorption column schematic — As capture on Fe-oxide beads
Brown beads are iron-oxide adsorbent particles; green dots are As(V) molecules. Influent arsenic binds to the surface =Fe-OH sites, lowering the effluent concentration.
q_e: equilibrium loading [mg-As/g], C_e: equilibrium concentration [mg-As/L]. Langmuir saturates at qmax (monolayer); Freundlich is an empirical fit for heterogeneous surfaces.
Adsorbent dose D [g/L], column volume V_col [m³] at EBCT = 5 min, column diameter φ_col [m] assuming a 2 m bed.
Arsenic (As) removal — Freundlich/Langmuir isotherms on iron oxides
🙋
When I hear "arsenic" I think mystery novels and poison — does it really turn up in tap water?
🎓
Sadly, yes — it's one of the largest public-health issues in the world. Bangladesh, West Bengal, parts of Inner Mongolia, Arizona, New Mexico: in each of these regions the groundwater is naturally rich in arsenic, and tens of millions of people are chronically exposed. The WHO drinking-water guideline is 10 μg/L (0.01 mg/L), but contaminated wells routinely read 500 or 1000 μg/L. Even Japan and Europe see local exceedances around hot springs and old mining sites.
🙋
How can you possibly drop a concentration that low down to 10 μg/L? It's not something you can filter mechanically.
🎓
Good intuition — mechanical filtration is useless here. You have to grab the arsenic chemically with an adsorbent, and the workhorse is "iron oxide", specifically GFH (Granular Ferric Hydroxide). Its surface is studded with hydroxyl groups =Fe-OH, which exchange ligands with the arsenate ion HAsO4(2-) and bind it strongly. Try switching the adsorbent on the left: ferrihydrite has the highest qmax at about 12 mg/g because it has by far the largest specific surface area.
🙋
There are two isotherm models, Langmuir and Freundlich. What's the actual difference?
🎓
Langmuir is the classical "monolayer" model, q_e = qmax K Ce / (1 + K Ce). It assumes a uniform surface and saturates at qmax. Freundlich is empirical, q_e = K_F Ce^(1/n), and represents heterogeneous surfaces where strong and weak sites coexist. At the ppb levels of drinking-water arsenic, real GFH columns usually match Freundlich better, so it's the practical choice for operating data. Flip the toggle and you'll see the two curves diverge sharply on the chart above.
🙋
What is "competing phosphate"? Does it really matter for arsenic removal?
🎓
This is where most field designs fail. Phosphate ion PO4(3-) is almost a twin of arsenate AsO4(3-) — same group 15, same tetrahedral structure — and the two fight viciously for the same =Fe-OH sites. Just 0.5 mg/L of phosphate cuts the loading by 22%; 1 mg/L by 40%; 2 mg/L by 63%. Bangladesh groundwater is famously high in phosphate, which is why so many GFH systems break through earlier than the spec. Silica adds another moderate hit. Always measure PO4, Si and HCO3 in the raw water before you trust any sizing calculation. Drag the slider from 0 to 2 mg/L and watch q_e collapse.
🙋
The column diameter comes out automatically too — how is that decided?
🎓
It's set by the empty bed contact time (EBCT). Arsenate needs a few minutes to diffuse into a GFH grain and reach the =Fe-OH sites, so the industry standard is EBCT = 5 min. Take the flow rate, divide by 60 to get min units, multiply by 5 and you have the bed volume. Assume a 2 m bed depth and back out the diameter. At 50 m³/h that gives about 1.6 m. A household point-of-use system runs 30 cm diameter; a municipal plant often uses several parallel tanks of 2-3 m diameter.
Frequently Asked Questions
Under oxidizing conditions arsenic exists mostly as arsenate As(V) — the anionic species H2AsO4- and HAsO4(2-). Iron hydroxides such as GFH, goethite and ferrihydrite carry a high density of surface hydroxyl groups (=Fe-OH) that exchange ligands with arsenate, forming strong inner-sphere complexes. Practical capacities are about qmax = 10 mg-As/g for ferrihydrite and 6-8 mg-As/g for GFH, with high affinity even in the ppb range. Iron-oxide adsorption is therefore one of the most widely used technologies for meeting the WHO drinking-water guideline of 10 μg/L (0.01 mg/L).
The Langmuir isotherm q_e = qmax K Ce / (1 + K Ce) assumes uniform monolayer adsorption and plateaus at qmax. The Freundlich form q_e = K_F Ce^(1/n) is empirical, represents heterogeneous surfaces with a distribution of binding strengths, and often fits drinking-water data better at very low Ce. Full-scale GFH columns running in the μg/L range typically match Freundlich more accurately, but Langmuir is still convenient at the planning stage because it reads qmax directly. Switch between the two in the tool and check how the dose changes before committing to a design.
Phosphate PO4(3-) is chemically almost a twin of arsenate AsO4(3-): both are tetrahedral oxyanions of group-15 elements, and they compete directly for the same =Fe-OH surface sites on iron oxides. This tool models the interference as exp(-0.5 [PO4]), so 1 mg/L of PO4 lowers the working capacity to about 60%, and 2 mg/L cuts it to 37%. Sulfate, chloride and bicarbonate are far less aggressive; silica is moderate. Always measure phosphate in the raw water and correct the design capacity accordingly.
Granular ferric hydroxide (0.3-2 mm grain) is mass-transfer limited: arsenate ions need several minutes to diffuse into a particle and reach the =Fe-OH sites. Too short an EBCT lets water escape before equilibrium and slashes the apparent capacity. Real plants use EBCT = 3-7 min, and 5 min is the conservative design value. The tool sizes the column volume at EBCT = 5 min, assumes a 2 m bed depth and returns the required column diameter. At high flow rates, several columns are usually run in parallel to keep each diameter manageable.
Real-World Applications
Well-water treatment in Bangladesh and West Bengal: The Ganges Delta aquifer carries naturally high arsenic at 50-500 μg/L, exposing roughly 50 million people. UNICEF and WHO have funded distributed point-of-use GFH columns (20-30 cm diameter, 1 m tall) at the village or household scale. Quick sizing tools like this one set the initial adsorbent volume and replacement frequency. Wells with elevated phosphate (1-3 mg/L) break through far sooner than spec, so raw-water analysis is mandatory.
Small-system drinking water in the United States: When the USEPA tightened the arsenic MCL to 10 μg/L in 2006, hundreds of small utilities across Arizona, New Mexico and Southern California adopted GFH and TiO2-based adsorbents. Operating cost is dominated by media replacement and typically lands at 0.10-0.50 USD per m³ at moderate (50 μg/L) influent. The annual-cost estimate produced by this simulator is in the right ballpark for budgeting.
Mine-drainage and geothermal effluents: Acid mine drainage from gold and copper mines, and geothermal discharge from selected hot-spring regions, can run hundreds of μg/L of arsenic. Iron-oxide adsorption typically sits as the polishing step downstream of aeration and coagulation/sedimentation, which knock out competing Fe, Mn and Pb. When raw water contains Fe(II), in-situ oxidation can generate a fresh ferric hydroxide that co-precipitates the arsenic — a low-cost variant that is essentially adsorption on a freshly grown adsorbent.
Disposal of spent media: Spent GFH loaded with arsenic is an industrial waste that must be stabilised. Ferric hydroxide is generally insoluble enough to qualify for landfill disposal, but long-term re-release is a real concern. In the United States the spent media must pass the EPA TCLP leaching test (less than 5 mg/L As in the leachate). The annual adsorbent consumption produced by this simulator feeds directly into disposal-cost estimates.
Common Misconceptions and Pitfalls
The biggest trap is choosing an adsorbent on qmax alone. Ferrihydrite has the highest qmax (12 mg/g), but its fine grain causes high pressure drop, its surface area crystallises slowly into goethite (capacity loss), and it is difficult to regenerate. GFH has a lower qmax (6-8 mg/g) but uniform grain size, low pressure drop and several mature commercial products (Bayoxide E33, AdsorpAs and others), which is why it dominates real plants. Use this simulator for the first-pass comparison, but always confirm the isotherm and breakthrough behaviour with a pilot column on the actual raw water before procurement.
Second, designing on equilibrium alone. The calculation assumes equilibrium loading, but in a real column the mass-transfer zone is finite and only 50-80% of the bed reaches saturation before breakthrough. The "required dose" produced here is therefore a theoretical lower bound, and full-scale designs apply a 1.5-2× safety factor. Vendor data also reports replacement frequency in bed-volumes (BV) — typically 30,000-100,000 BV before breakthrough — which is the right metric to convert into annual media use once you have pilot data.
Finally, ignoring the difference between As(V) and As(III). This tool assumes oxidized As(V) (arsenate). Under reducing conditions, much of the arsenic is As(III) (arsenite), which is uncharged at neutral pH (H3AsO3) and binds 1/3 to 1/2 as strongly to iron oxides. Bangladesh wells often contain 50-80% As(III), so a pre-oxidation step — chlorine, permanganate or air — is essentially mandatory to convert As(III) to As(V) before adsorption. Always measure the As(V)/As(III) ratio in raw water before trusting any sizing tool, including this one.
How to Use
Enter influent arsenic concentration (μg/L) from your groundwater analysis—typical Bangladesh tube wells range 50–500 μg/L against WHO limit of 10 μg/L
Set target effluent arsenic (μg/L), usually 10 μg/L for drinking water compliance
Input solution pH (6.5–8.5 optimal for iron oxide adsorption) and competing phosphate concentration (mg/L) from water quality tests
Simulator calculates equilibrium loading q_e, required adsorbent dose (g/L), daily media consumption (kg/day), and packed-bed column diameter for 5–10 minute contact time
Review annual operating cost including iron oxide replacement and backwash disposal
Worked Example
Village groundwater in West Bengal: influent As = 285 μg/L, target = 10 μg/L, pH = 7.2, phosphate = 2.1 mg/L. Ferric hydroxide (Fe(OH)₃) adsorption at pH 7.2 achieves q_e ≈ 4.8 mg-As/g. Required dose = 60 g/L iron oxide. For 500 m³/day treatment plant, daily media use = 30 kg/day. Column diameter sized for 4 m/h superficial velocity = 0.41 m. Annual cost ≈ USD 18,400 including adsorbent replacement every 6 months and sludge handling at USD 120/tonne.
Practical Notes
Phosphate competition reduces As adsorption by 15–35%; above 3 mg/L phosphate, increase iron oxide dose by 20–30% to maintain target removal
Iron oxide exhaustion occurs after 2–4 months; implement breakthrough monitoring at 50 μg/L effluent to schedule replacement before regulatory violation
pH drop during adsorption reduces capacity; maintain influent pH ≥ 6.8 using lime dosing (0.5–1.0 g/L) for sustained performance in acidic aquifers
Backwash frequency every 2–3 days prevents head loss exceeding 1.5 m; use 15–20% bed expansion at 12 m/h for 5 minutes