membrane-pretreatment

What this skill does

# RO/NF Pretreatment

Reference for what feeds the membrane train, with KPI targets, chemistry, and instrumentation. The 80 % rule: most RO/NF/EDI problems originate upstream.

Convention markers:
- `⚠️` — likely bug or incorrect assumption in calling code
- `💡` — enhancement opportunity (new alert, new monitoring point, new dosing logic)

---

## §1 — Pretreatment Chain Overview

### 1.1 SWRO open intake

```
Sea intake (screen 5–25 mm)
  → Travelling band / drum screen (~1–3 mm)
  → In-line coagulation (FeCl3 0.5–5 mg/L as Fe) + flash mix (G ≈ 700–1000 s⁻¹, 30–60 s)
  → Flocculation tank (G ≈ 30–80 s⁻¹, 15–30 min) [optional with inline coag]
  → DAF (HABs/algae/oil) OR direct media filtration
  → Dual / Multimedia gravity filter (5–15 m/h) OR pressure MMF (15–25 m/h)
  → [Optional] UF 0.02 µm for SDI₁₅ ≤ 2
  → Cartridge filter 5 µm absolute (ΔP fresh < 0.3 bar; change at 1.0–1.5 bar)
  → Antiscalant injection
  → SBS dechlorination (only if Cl2 used pre-filter)
  → Acid dosing (H2SO4 or HCl) for LSI control
  → HPP → SWRO 1st pass
```

### 1.2 SWRO beach well (subsurface intake)

```
Beach well / infiltration gallery (natural aquifer filtration)
  → Cartridge filter 5 µm
  → Antiscalant injection
  → SBS only if shock-chlorination at wellhead
  → Acid dosing
  → HPP → SWRO
```

Subsurface intakes naturally deliver SDI₁₅ < 3 in most geologies → skip coagulation and media filtration. Yield (m³/h per well) and risk of Fe/Mn intrusion from anoxic strata are the practical limits.

### 1.3 BWRO well (saline groundwater)

```
Well pump → aeration / oxidation (Cl2, KMnO4, or O2) if Fe/Mn > 0.05 mg/L
  → Manganese greensand / pyrolusite filter (Fe/Mn removal)
  → MMF polishing
  → Cartridge filter 5 µm
  → SBS (if Cl2 residual present)
  → Antiscalant
  → Acid (for high alkalinity / LSI)
  → BWRO
```

### 1.4 BWRO surface water

Same architecture as SWRO open intake but lower TDS, lower turbidity peaks. pH control more critical (CaCO3 dominant scaling vs Mg-sulfate).

### 1.5 Tertiary reuse → BWRO

```
Biological secondary effluent → MBR or tertiary UF/MF
  → break tank
  → cartridge filter 5 µm
  → DBNPA continuous low-dose (1 mg/L) optional
  → antiscalant (low-P or P-free)
  → BWRO 1st pass
```

MBR/tertiary UF typically produces SDI₁₅ 0.5–2.0; secondary effluent may exceed 3 in some months.

---

## §2 — Pretreatment KPI Targets (pre-RO)

| Parameter | SWRO open | SWRO beach well | BWRO well | BWRO surface | Tertiary reuse |
|---|---|---|---|---|---|
| SDI₁₅ target | ≤ 3 | ≤ 2 | ≤ 3 | ≤ 3 | ≤ 3 |
| SDI₁₅ max | 5 | 3 | 4 | 5 | 5 |
| SDI₁₅ with UF | ≤ 2 | n/a | ≤ 2 | ≤ 2 | ≤ 2 |
| Turbidity (NTU) target | < 0.1 | < 0.1 | < 0.1 | < 0.1 | < 0.2 |
| Turbidity (NTU) max | 1.0 | 0.5 | 1.0 | 1.0 | 1.0 |
| Free Cl₂ pre-membrane | 0 ppm | 0 ppm | 0 ppm | 0 ppm | 0 ppm |
| ORP post-dechlor | < 200 mV | n/a | < 200 mV | < 200 mV | < 200 mV |
| AOC (µg C/L) | < 50 excellent / < 100 acceptable | < 50 | < 50 | < 100 | < 100 |
| BGP (µg/L) | < 70 | < 70 | n/a | n/a | n/a |
| Iron (mg/L) | < 0.05 | < 0.05 | < 0.05 | < 0.05 | < 0.1 |
| Manganese (mg/L) | < 0.02 | < 0.02 | < 0.02 | < 0.02 | < 0.05 |
| Aluminum residual (mg/L) | < 0.05 | < 0.05 | < 0.05 | < 0.05 | < 0.05 |
| Operating pH | 6.5–7.5 | 6.5–7.5 | 5.5–7 | 5.5–7.5 | 6–7.5 |
| Cartridge ΔP fresh | < 0.3 bar | < 0.3 bar | < 0.3 bar | < 0.3 bar | < 0.3 bar |
| Cartridge ΔP change-out | 0.7–1.0 bar | 0.7–1.0 bar | 0.7–1.0 bar | 0.7–1.0 bar | 0.7–1.0 bar |

⚠️ SDI₁₅ > 5 voids most OEM warranties. SDI₁₅ ≤ 5 is the absolute hard limit.

AOC threshold from Vrouwenvelder / Weinrich biofouling correlation studies. Raw seawater commonly 30–400 µg C/L (Tampa Bay 360 ± 180 µg/L; Monterey 30 ± 20 µg/L).

⚠️ Heavy metals (Fe³⁺, Mn²⁺, Cu²⁺, Co²⁺) catalyse SBS → oxidant conversion under O₂. Keep trace metals low and minimize SBS overdose (see §5.4).

---

## §3 — Coagulation / Flocculation Chemistry

### 3.1 Coagulant comparison

| Coagulant | Formula | Dose (as product) | Dose (as metal) | Optimal pH | Notes |
|---|---|---|---|---|---|
| Ferric chloride | FeCl₃·6H₂O | 5–40 mg/L | 1–10 mg/L Fe | 5.0–8.5 (best 5–7) | Wide pH; dense floc; residual Fe³⁺ accepted < 0.05 ppm; preferred for SWRO |
| Ferrous sulfate | FeSO₄·7H₂O | 10–50 mg/L | 2–10 mg/L Fe | 8.5–11 + Cl₂ oxidize Fe²⁺→Fe³⁺ | Cheap; needs alkaline conditions |
| Alum | Al₂(SO₄)₃·14–18H₂O | 5–60 mg/L | 0.4–5 mg/L Al | 6.0–7.5 (narrow) | Al residual → AlPO₄/Al(OH)₃ fouling on RO; avoid for RO when possible |
| PAC | Al_n(OH)_m Cl_{3n-m} | 5–30 mg/L | 0.5–3 mg/L Al | 5.5–9 (wide) | Lower residual Al than alum; less pH depression; efficient at cold T |

### 3.2 Polymer flocculant aids

- Type: anionic / cationic / nonionic polyacrylamide (PAM); also tannin-based natural polymers
- Dose 0.05–1 mg/L. Under-dose → no benefit; over-dose → carryover, RO fouling
- Cationic PAM acts as primary coagulant in low-turbidity surface waters

⚠️ Cationic polymer + anionic antiscalant → precipitate → RO fouling. Always jar-test the combination before deployment.

### 3.3 Jar test methodology

1. **Setup**: 6 paddle jars × 1 L; coagulant in dilute solution (1–10 % w/v)
2. **Flash mix**: 100–200 rpm × 30–60 s (G ≈ 700–1 000 s⁻¹)
3. **Flocculation**: 30 rpm × 15–30 min (G ≈ 30–80 s⁻¹, tapered if possible)
4. **Settling**: 30 min undisturbed
5. **Sampling**: supernatant at fixed depth; measure turbidity, TOC, SDI, residual coagulant metal, pH

G value:
```
G = sqrt(P / (μ · V))     # P = power (W), μ = dynamic viscosity (Pa·s), V = volume (m³)
```

### 3.4 Pros/cons for downstream RO

| Aspect | Iron-based (FeCl₃) | Aluminum-based (alum, PAC) |
|---|---|---|
| Residual carryover risk | Fe³⁺ acceptable up to 0.05 ppm | Al³⁺-phosphate fouling, severe at > 0.05 ppm |
| Floc strength | Dense, settle easily | Lighter |
| pH window | Wide | Narrow (alum); wider (PAC) |
| Cold water | Good | Good (PAC) / poor (alum) |
| TOC removal | Excellent | Moderate |
| SWRO compatibility | Preferred | Avoid alum; PAC OK with caution |

---

## §4 — Antiscalant Selection and Dose Math

### 4.1 Chemistry families

| Family | Examples | Best for | Typical dose | Notes |
|---|---|---|---|---|
| Phosphonate | HEDP, PBTC, ATMP, DTPMP, BHMTPMP | CaCO₃, CaSO₄, BaSO₄, SrSO₄, CaF₂ | 2–5 mg/L | Workhorse; P discharge constraints |
| Polyacrylate / acrylic homo & copolymer | PAA, AA-MA | BaSO₄, SrSO₄, suspended solids dispersion | 1–5 mg/L | P-free; lower Ca tolerance |
| Sulfonated copolymer | AA-AMPS, AA-AMPS-HPA terpolymer | High Ca, CaPO₄, ZnCO₃ | 2–5 mg/L | High Ca + alkaline tolerance |
| Maleic-based | MA-AA, polymaleic | CaCO₃, CaSO₄ broad | 2–5 mg/L | P-free; used in P-restricted discharges |
| Polymeric silica dispersant | PEG, PEGD, PVA, PAMAM, PEI | SiO₂, colloidal silica | 1–5 mg/L | For SiO₂ > 100 ppm in concentrate |
| Dendrimer / latest-gen | Polyaspartate, dendrimer-based | Broad multi-species | 1–4 mg/L | Biodegradable; premium |

### 4.2 Selection by scaling species

| Scaling species | Antiscalant of choice |
|---|---|
| CaCO₃ | Phosphonate (HEDP, PBTC) OR maleic-based (P-free) |
| CaSO₄ (gypsum) | Phosphonate + polyacrylate blend |
| BaSO₄, SrSO₄ | Polyacrylate or AA-AMPS (phosphonates can co-precipitate with Ba) |
| CaF₂ | Phosphonate |
| SiO₂ amorphous | Polymeric silica inhibitor (PEG/PEGD) + pH adjustment |
| Ca₃(PO₄)₂ | AA-AMPS terpolymer — **NOT** phosphonate |
| Fe / Mn fouling | Specialised dispersant blend |

### 4.3 Dose calculation

Three approaches:
1. **Vendor projection software** (preferred): DuPont WAVE, Hydranautics IMSDesign, Toray DS-Design, Veolia Winflows, Avista AdvisorCi, Genesys Genesys Member. Software computes LSI / S&DSI / IP-Ksp per stage and recommends dose.
2. **LSI-driven empirical**: target concentrate LSI ≤ +1.8 (conservative) or ≤ +2.5 (premium antiscalant). Typical dose 2–5 mg/L product.
3. **Jar-tested matrix**: when projection unavailable.

### 4.4 Compatibility traps

⚠️ Cationic biocide + anionic antiscalant → precipitate.
⚠️ Phosphonate + Ba²⁺ → BaSO₄ co-prec