Why Surabaya Is Not Baku: The 2026 Receiving-Water Story
Surabaya is not a closed basin like the Caspian, but the receiving-water and permit baseline still drives the 2026 design. The Brantas basin and the Kali Surabaya estuary set the envelope for any hyperscale or fab project on the east-Java coast, and a hyperscale discharge that cannot be met by the receiving works pushes the design toward on-site reuse or haul-off — the same closure logic the Baku 2026 guide applies to the Caspian.
Surabaya's municipal-sewer capacity is not sized for fab chemistry or for cooling-tower blowdown in the 200–1,000 m³/day band, and the receiving-water evidence from Kenjeran and the wider Surabaya coast (Bahri et al., JESSD 2020) shows that a hyperscale project cannot rely on municipal treatment as the design sink. KLHK PP No. 22/2021 and the standard AMDAL/UKL-UPL workflow are the 2026 regulatory anchors; the supplied research does not quote the numeric effluent limits, so each project must verify the current parameter set against the chosen receiving-water class before final sizing. A first-pass check the reader can run now: ask KLHK whether the proposed discharge temperature, TDS, and any trace metals meet the 2026 parameter set; if not, the design must shift to on-site reuse or zero-liquid discharge (ZLD). For a parallel tropical comparison, the Lyon 2026 data center CTBD guide walks through a similar receiving-water negotiation under a different regulatory regime.
Stand-Alone Data Hall vs Co-Located Fab: Two 2026 Design Envelopes
The envelope you design for depends entirely on whether a fab is co-located. A stand-alone Surabaya hyperscale data hall in 2026 sees cooling-tower blowdown at 4–6 cycles of concentration (COC) with TDS up to 2,000 ppm and 30–40 °C, plus air-handling-unit (AHU) condensate at <50 mg/L TDS; the baseline train is membrane bioreactor (MBR) plus two-pass reverse osmosis (RO) at 80–95% recovery, with mechanical vapor recompression (MVC) applied to the RO brine only when the discharge path is restricted (HydropureWater field data, 2026). A co-located semiconductor fab changes the problem qualitatively: HF-etch fluoride at 50–500 mg/L, tetramethylammonium hydroxide (TMAH) in the developer stream, chemical-mechanical planarization (CMP) nanoparticles (silica, ceria, alumina), and Cu up to 100 mg/L in untreated CMP effluent — the combined load forces full-stream ZLD as the 2026 default (HydropureWater field data, 2026, citing Lai & Lin 2004). The co-location boundary is the single most important 2026 design decision: a fab-hall hybrid cannot be designed to the stand-alone envelope and then upgraded, because HF, TMAH, and Cu-CMP cannot be blended down to the Surabaya receiving-water band. A 2026 design decision framework: if any HF-etch, developer, or CMP stream is on the same outfall as the cooling-tower blowdown, plan full-stream ZLD from the first mass balance; otherwise MBR + two-pass RO at 80–95% recovery is the operating envelope. The Pune 2026 semiconductor and data-hall guide applies the same decision rule under India's consent-to-operate framework.
The 2026 Process Train, Stage by Stage
The process train for a Surabaya 2026 site is a sequence of unit operations, each with a defined purpose. The reader should walk the stages in order because the cost table later in this article is sized against this exact stack (HydropureWater field data, 2026).
- Stage 1 — Segregation and equalization. A dedicated equalization (EQ) tank with 4–8 h hydraulic residence time (HRT) and online pH/conductivity dampens the 1–5 pH excursions that follow chiller trips; AHU condensate (<50 mg/L TDS) is segregated on its own line because glycol from coil leaks requires stripping rather than blending into the main RO loop.
- Stage 2 — DAF and multi-media filtration. A dissolved air flotation unit in the 4–300 m³/h class floats oils, biofilm, and metal-hydroxide floc, followed by a multi-media filter that drops the Silt Density Index (SDI) below 3 and protects the RO from Cu, Fe, and Zn fouling; for fab trains with Cu-CMP load, a hollow-fiber ultrafiltration (UF) pretreatment step upstream of DAF improves colloid removal and stabilizes downstream RO performance.
- Stage 3 — Softening and antiscalant dosing. A twin-tank industrial softener (1–45 T/h class) targets hardness <50 mg/L as CaCO₃ and silica <10 mg/L as SiO₂, with PLC-controlled antiscalant dosing tied to RO feed flow; on a Surabaya intake with seasonal silica swings, the softener setpoint needs to track intake quality rather than hold a fixed value.
- Stage 4 — MBR polishing (stand-alone halls only). A submerged PVDF MBR with 0.1 µm membranes delivers <1 NTU and <10 mg/L chemical oxygen demand (COD), allowing direct RO feed without media filtration; for fab streams, MBR is replaced by a dedicated HF-removal and metals-precipitation step upstream of UF.
- Stage 5 — Two-pass RO at 80–95% recovery. First pass runs at 150–250 psi (10–17 bar) for bulk salts; second pass polishes to TDS <200 mg/L and Cl⁻ <100 mg/L; above 95% recovery, silica scaling on the second-pass membranes drives clean-in-place (CIP) frequency up sharply, so 80–95% is the practical operating window. See the 2026 MBR membrane cleaning field guide for the CIP regime that keeps the train on this envelope.
- Stage 6 — MVC brine concentration. Mechanical vapor recompression at 25–40 kWh/m³ of brine concentrated; full-stream ZLD is over-specified for a stand-alone Surabaya hall, so reserve MVC for the RO brine stream when the discharge path is restricted, or treat it as a non-negotiable baseline for any co-located fab.
A chlorine dioxide generator upstream of the RO controls biofouling without the isothiazolinone residual load that shortens RO membrane life, and a side-stream filtration step at 1–5% of total circulation using 10–25 µm self-cleaning spiral units drops suspended solids to levels the RO can handle without pre-coat — the cheapest way to push COC higher and shrink the RO train. The two-pass RO system itself is specified against the two-pass industrial RO reference unit.
| Stage | Unit operation | Key 2026 setpoint | Why it matters on Java |
|---|---|---|---|
| 1 | EQ tank, AHU segregation | 4–8 h HRT | Dampens 1–5 pH excursions after chiller trips |
| 2 | DAF + multi-media + (UF if fab) | SDI <3 post-filter | Protects RO from Cu, Fe, Zn fouling |
| 3 | Twin-tank softener + antiscalant | Hardness <50 mg/L; SiO₂ <10 mg/L | Must track seasonal Surabaya silica |
| 4 | MBR (stand-alone) / HF + metals precipitation (fab) | <1 NTU, <10 mg/L COD | Replaces media filtration for sanitary load |
| 5 | Two-pass RO | 80–95% recovery; permeate <200 mg/L TDS, Cl⁻ <100 mg/L | 95% is the ceiling before silica CIP rises |
| 6 | MVC on RO brine | 25–40 kWh/m³ of brine | Reserve for restricted discharge or full-stream ZLD |
Tropical-Ambient Sizing: Adjusting the 2026 Mass Balance for Surabaya
The Baku blowdown formula B = E/(COC−1) still governs — 25% blowdown at 4 COC, 20% at 6 COC — but the evaporation rate E in Surabaya rises with the higher ambient wet-bulb, so the absolute blowdown flow at the same COC is larger than a temperate equivalent; the supplied research does not quote a numeric Surabaya wet-bulb, so each project must size E from its site-specific psychrometric data (HydropureWater field data, 2026). Surabaya intake silica can swing seasonally; the softener setpoint and the antiscalant dose must track the intake rather than hold a fixed value, or the second-pass RO will scale. Condenser lift penalty at tropical ambient means the cooling-tower fan and pump energy rises, which increases the operating cost of any on-site evaporation step and is one more reason to push COC as high as silica scaling allows. A 2026 design rule of thumb: hold first-pass RO recovery at 80–90% to keep silica below saturation on the second pass, dose antiscalant tied to RO feed flow, and use RO elements rated for high-silica feed to control replacement frequency (HydropureWater field data, 2026). The qualitative take-away is that any Surabaya mass balance carried over from a Caspian or temperate-European reference must be re-anchored against local psychrometrics before the COC and the RO train size are frozen.
2026 Cost Bands and Payback for a Surabaya Site
The 2026 cost band for a Surabaya site is a function of flow and chemistry envelope, not a function of the receiving-water class (HydropureWater field data, 2026). A small hall under 200 m³/day with package plant plus haul-off runs $150–300/m³/day installed. A mid-size 200–1,000 m³/day MBR + RO train runs $400–700/m³/day installed, which is where the typical 5–20 MW Surabaya hyperscale hall sits. A hyperscale or co-located fab with brine ZLD runs $800–1,200/m³/day installed. Avoided-discharge math at $5–15/kgal: 100 m³/day of untreated blowdown at the upper end is roughly USD 400/day, so an 80% recovery RO pays back inside ~24 months at hyperscale flow. The 2026 manufacturing water-reduction guide walks the same payback logic across other high-water-stress basins. OPEX lines that push cost up: membrane CIP frequency above 95% recovery on the second pass due to silica scaling; isothiazolinone biocide residuals shortening RO life, which is the case for a ClO₂ side-loop upstream of the RO rather than dosing in the tower; MVC electricity at 25–40 kWh/m³ of brine concentrated.
| 2026 envelope | Flow band | CAPEX ($/m³/day installed) | Dominant OPEX line |
|---|---|---|---|
| Small hall, package + haul-off | <200 m³/day | 150–300 | Brine haul-off, softener regeneration |
| Mid-size MBR + RO | 200–1,000 m³/day | 400–700 | RO CIP, antiscalant, ClO₂ residual |
| Hyperscale or co-located fab, ZLD | >1,000 m³/day | 800–1,200 | MVC electricity 25–40 kWh/m³, high-silica RO replacement |
2026 Permit and Supplier Checklist for a Surabaya Project
Run the KLHK/AMDAL pre-meeting first, not at submission: confirm the receiving-water class, the 2026 effluent parameter set, and the 80% reuse floor that the supplied research treats as a hard constraint by analogy with other high-water-stress basins (HydropureWater field data, 2026). Decide the envelope before sizing the equipment: stand-alone hall = MBR + two-pass RO at 80–95% recovery; co-located fab = MBR + two-pass RO + full-stream MVC ZLD as the non-negotiable baseline. Validate tropical-ambient sizing: confirm the site-specific wet-bulb, the seasonal silica profile of the Surabaya intake, and the condenser lift penalty before fixing the COC and the RO train size. Supplier selection criteria for 2026: a track record on both data-hall cooling-tower blowdown and semiconductor fab chemistry, factory-tested skids, RO/UF membrane elements rated for high-silica feed, and a service footprint that covers Java. Final 2026 sanity check: confirm the avoided-discharge payback inside ~24 months at the projected blowdown flow, and confirm that the second-pass RO recovery ceiling is held at 80–95% so silica scaling does not erode the OPEX gain. The receiving-water logic behind the permit step is the same one the Lyon 2026 data center CTBD guide applies to the Rhône corridor.
Frequently Asked Questions
What 2026 CAPEX band should a Surabaya hyperscale hall put in front of a procurement committee?
A mid-size 200–1,000 m³/day MBR + RO train runs $400–700/m³/day installed, which is where the typical 5–20 MW Surabaya hyperscale hall sits; a small hall under 200 m³/day with package plant plus haul-off runs $150–300/m³/day; hyperscale or co-located fab with brine ZLD runs $800–1,200/m³/day (HydropureWater field data, 2026). The actionable check is to lock the flow band first, then size the RO and the brine step separately, because the cost gap between the mid-size and ZLD bands is the single largest line-item swing on the BOQ.
How do I pick the right 2026 train envelope for a Java project that may host a fab later?
If any HF-etch fluoride (50–500 mg/L), CMP nanoparticles, or TMAH can plausibly co-occur in the wastewater envelope, plan MBR + two-pass RO + full-stream MVC ZLD from day one (HydropureWater field data, 2026, citing Lai & Lin 2004). The co-location boundary cannot be retrofitted because HF, TMAH, and Cu-CMP cannot be blended down to the Surabaya receiving-water band after the fact. The actionable check is to ask whether the fab process is on the same outfall as the cooling-tower blowdown; if the answer is yes, the envelope is ZLD regardless of current fab load.
What is the 2026 permit risk for a hyperscale project that plans to discharge to the Surabaya municipal sewer?
The receiving-water evidence from Kenjeran and the wider Surabaya coast (Bahri et al., JESSD 2020) shows why a hyperscale project cannot rely on municipal treatment as the design sink, and KLHK PP No. 22/2021 plus the AMDAL/UKL-UPL workflow are the 2026 regulatory anchors; the supplied research does not quote the numeric effluent limits, so the actionable check is to confirm the 2026 parameter set against the chosen receiving-water class before sizing the train. Treat municipal-sewer discharge as a non-starter above the small-hall threshold.
Which supplier selection criteria actually matter for a 2026 Surabaya data-hall or fab-hall project?
A track record on both data-hall cooling-tower blowdown and semiconductor fab chemistry, factory-tested skids, RO/UF membrane elements rated for high-silica feed, and a service footprint that covers Java (HydropureWater field data, 2026). The actionable check is to ask bidders for a reference list with at least one operational site on Java and one with fab-stream ZLD, because the two envelopes pull on different unit operations and a stand-alone-hall-only reference will not validate the fab upgrade path. For a parallel benchmark under a different regulator, the Mumbai 2026 semiconductor and data-hall guide sets out the same supplier criteria against the MPCB consent envelope.
