Sand fliter vs Carbon fliter vs Ultrafiltration vs Reverse Osmosis — Best Uses Explained

Why "one-size-fits-all" fails

Pollutant morphology and scale mapping

The contaminants are of 6 orders of magnitude. Thus, sand filters are used to remove 10-100 µm particles. UF screens are used to remove 0.01-0.1 µm colloids. GAC is used to adsorb molecules less than 2 nm in size. RO is used to remove ions less than 1 nm in size. Thus, treating 5 µm particles with RO is of no value. Similarly, using sand filters to remove sodium ions is against the laws of physics.

The primary cause of underperformance for water treatment systems is misalignment. This was observed for RO membranes at Deiiang™ which foul in 90 days or less due to carryover of upstream floc, a problem that could be easily solved with a sand filter at a fraction of the cost.

Jason.peng insight: “In multiple lithium battery projects along the Yangtze River basin, I saw RO membranes fail prematurely — not because of high salinity, but because the front‑end sand filter had manual valves and no automatic backwash. Operators skipped backwash during night shifts, sending turbid water straight to the UF and RO. A simple automatic backwash valve would have saved tens of thousands in membrane replacement.”

Key water quality indicators

Fractionate COD into particulate, colloidal, and dissolved phases. Map turbidity, UV₂₅₄, conductivity, hardness, and SDI₁₅. A 120 mg/L COD may be 80% particulate (sand filter), 15% colloidal (UF), 5% dissolved (GAC).

Pretreatment failure chain — how upstream upsets cascade downstream

Treatment trains are only as strong as their weakest link. A seemingly minor disturbance in an early stage triggers a domino effect that accelerates damage to far more expensive assets downstream.

Two failure chains dominate field service calls: (1) turbidity escape from a poorly backwashed sand filter forces UF into accelerated chemical cleaning, which in turn shortens membrane life and risks fiber breakage — broken fibers then dump colloidal loads directly onto RO lead elements. (2) Exhausted GAC allows free chlorine to oxidize RO membrane polymer, causing irreversible salt rejection loss within 200–1,000 ppm‑hours of exposure. In both cases, the root cause is upstream — but the cost is felt at the most expensive component.

3-step selection method

1. Identify dominant phase: particulate, colloidal, dissolved organic, or ionic.

2. Set numeric targets: NTU <0.1, SDI₁₅ <3, conductivity <10 µS/cm.

3. Match technology: particulate → sand; colloidal + microbial → UF; dissolved organics + chlorine → GAC; ionic → RO.

Sand Filtration

Sand filters are very reliable for removing particulate matter. A properly designed sand filter can give an effluent of less than 1 NTU from a 15-30 NTU influent, which gives 85-95% removal of TSS. Deiiang™ engineer Jason.peng reports in his blog that in municipal wastewater treatment applications his sand filters run for 48-72 hours between backwash cycles.

Scenarios & design

Post-clarification floc capture, pre-filter for GAC/UF, and final polishing. Design at 5–12 m/h with d₁₀ 0.35–0.55 mm, UC <1.6. Backwash at 30–50 m/h consumes 2–4% of daily flow. Fine floc (<50 µm) shortens filter runs dramatically.

Deiiang™ Sand Filter Data

• TSS removal: 88–96%

• Backwash water: 2.2–4.5% of throughput

• Effluent turbidity: 0.5–2.0 NTU

Fig.1 — Sand filter depth filtration.

Granular Activated Carbon (GAC)

The use of activated carbon filter solves dissolved organics and chlorine. With 800–1,200 m²/g surface area, it adsorbs molecules that pass through physical barriers. Capacity is finite — a 2 m³ vessel at 10 min EBCT may last 15,000–25,000 bed volumes before chlorine breakthrough. NOM pre-loading reduces target adsorption by 30–50%.

Scenarios & lifecycle

Dechlorination (1–3 mg/L to <0.1 mg/L), DOC polishing (40–70% UV₂₅₄ reduction), odor removal, and biological activated carbon. EBCT: 8–15 min for chlorine, 15–30 min for DOC. Thermal reactivation restores 85–95% capacity with 5–10% mass loss per cycle.

Monitoring risks

Track ΔP and online UV₂₅₄. Excessive biofilm causes channeling and sloughing. HPC >10,000 CFU/mL warrants investigation.

Ultrafiltration (UF)

Innovative ultrafiltration (UF) applications are growing rapidly between granular filtration and reverse osmosis (RO) to ACHieve >4-log removal of bacteria and other microorganisms, and very low levels of turbidity (e.g. <0.1 NTU) and suspended solids (e.g. SDI15 <2.5) to serve as ideal pretreatment to RO. The SDI15 values obtained in various applications for electronics water reuse using Deiiang™ technology average 1.8–2.2.

Removal spectrum & fouling control

UF rejects colloids, bacteria, macromolecules >100 kDa. It passes salts, dissolved silica, and small organics. Operate at 40–70 LMH with TMP 0.3–1.5 bar. Backwash every 25–45 min; CEB every 12–48 h. CIP when TMP exceeds 1.5× baseline.

SDI stability is more critical than average. Fluctuations between 1.5 and 3.8 damage RO more than a steady 2.8.

Deiiang™ UF Snapshot

• Effluent turbidity: <0.08 NTU

• SDI₁₅: 1.5–2.5

• TMP ramp: 0.02–0.06 bar/week

• CIP interval: 45–90 days

Reverse Osmosis (RO)

The use of reverse osmosis (RO) targets ionic separation. Polyamide membranes reject 99.0–99.7% NaCl. Deiiang™ two-stage systems reduce conductivity from 1,200 µS/cm to 18–35 µS/cm at 75% recovery. RO generates concentrate at 4× feed salinity — 25 m³ per 100 m³ feed.

Critical conditions & concentrate

Pretreatment: SDI₁₅ <2.5, chlorine <0.1 mg/L, iron <0.05 mg/L. Antiscalant 2–5 mg/L. Recovery limited by least soluble salt. Concentrate options: sewer discharge, recycling, thermal ZLD ($3–8/m³), resource recovery. Without a viable pathway, RO is not feasible. For downstream polishing, EDI is often required to reach ultrapure water resistivity.

Deiiang™ RO Data

• Recovery: 72–78%

• Permeate: 12–35 µS/cm

• Flow decline: <8%/year

• CIP recovery: 92–98%

Selection Framework

Target-first approach

Discharge TSS 30 mg/L needs only sand; cooling tower <500 µS/cm may need RO; boiler <10 µS/cm demands RO. Map each parameter to its numeric threshold.

Quick Decision Tree (NTU & Conductivity)

Typical process trains

Four long-term costs

Water losses (3–8%), energy (0.8–1.5 kWh/m³ brackish RO), membrane/media replacement (RO every 3–5 yr), concentrate disposal (up to $5/m³). CAPEX is only 25–35% of 10-year TCO.

OPEX Comparison — Cost per 1,000 m³ Treated

Operating expenditure drives 65–75% of lifecycle cost. The table below compares typical consumable replacement intervals, energy intensity, and dominant cost drivers.

When NOT to install RO

Data & Visualization

Performance bars

Scale mapping & pretreatment thresholds

Pull Quote

"Select by contaminant morphology, not price. A well-designed sand filter protecting membranes delivers more value than a misapplied RO system with unmanageable concentrate."
— Jason.peng, Deiiang™

Deiiang™ Product Data & Methodology

Test standards

Turbidity iso 7027; SDI ASTM D4189; UV₂₅₄ SM 5910B; RO rejection ASTM D4516; GAC breakthrough ASTM D6586.

Deiiang™ Case Studies

Case 1: Lithium battery reuse (Yangtze River basin, high seasonal turbidity)

120 m³/h, target <5 µS/cm. Raw water from Yangtze tributary: TDS 850 mg/L, NTU 8–25 (monsoon spikes to 40+). Solution by Jason.peng: Coag→SF→GAC→UF→RO→EDI. Results: RO flow decline <6%/yr, SDI₁₅ 1.6–2.1, permeate 2–4 µS/cm. Concentrate to ZLD ponds.

Case 2: Municipal reuse polishing (East China, monsoon rainfall)

200 m³/h to Class A. Rainy season turbidity spikes 8→55 NTU in 2 h. Installed flow-paced coag + dual-media SF + pressurized UF. UF effluent <0.06 NTU, SDI₁₅ 1.5–2.0 year-round.

FAQs

Can UF ｒｅｐｌａｃｅ sand + carbon?

UF replaces sand for turbidity, but not GAC for dissolved organics or dechlorination.

Is carbon always before RO?

Only if oxidants present. Dechlorination is mandatory for polyamide membranes.

Safe SDI for RO?

SDI₁₅ <3 minimum, <2.5 recommended. Stability matters more than average.

GAC breakthrough planning?

Pilot test, online UV₂₅₄, plan regeneration logistics, budget 5–10% makeup carbon.

No concentrate pathway?

RO is infeasible. Explore sewer, high-recovery, reuse, ZLD, or non‑RO alternatives.

References

• AWWA B100 — Granular Filter Material

• ASTM D4189 — SDI Test Method

• ASTM D4516 — RO Performance Data

• ASTM D6586 — GAC Adsorption Prediction

• ISO 7027 — Turbidity

• WHO Guidelines

• GB/T 19249 — Reverse Osmosis Equipment

© 2026 Deiiang™ | Designer: Jason.peng | Sitemap