Sterilizer condensate in palm oil mills carries recoverable crude oil and boiler-grade heat. A well‑designed collection, separation, and heat‑recovery system turns that waste into yield and lower fuel bills.
Industry: Palm_Oil | Process: Sterilization
Sterilizer condensate—mill operators call it “white water”—flows hot and heavy out of palm oil sterilizers, often near saturation at 100–130 °C, and it’s anything but benign waste. Industry data put volumes on the order of 0.1–0.35 m³ of condensate per ton of fresh fruit bunches, or FFB (harvested palm fruit spikes), processed (patents.google.com). In one case, a continuous sterilizer serving a 60 t/h mill produced roughly 12 t/h (12 m³/h) of condensate—about 20% by mass of the FFB feed (www.researchgate.net).
The flow is oily. Studies show “oil‑rich” pulses during sterilization can generate 0.02–0.07 m³ of oily condensate per ton of FFB at pressures ~0.2–0.3 bar, with the remainder “oil‑poor” at 0.08–0.28 m³/t (patents.google.com). In practice, oil losses of ~1–3% of FFB weight into the condensate are observed when sterilization is overlong (ejournal.ft.unsri.ac.id) (patents.google.com). Left untreated, this drains yield and risks breaching Indonesia’s oil/grease limits for treated effluent (PermenLHK No. 5/2021) (www.aquajaya.co.id).
Sterilizer condensate volumes and oil
Reported condensate generation spans 0.1–0.35 m³ per ton of FFB (patents.google.com). One continuous unit at 60 t/h yielded ~12 t/h (12 m³/h), roughly 20% of the FFB mass (www.researchgate.net). Within a cycle, “oil‑rich” bursts (0.02–0.07 m³/t at ~0.2–0.3 bar) and “oil‑poor” portions (0.08–0.28 m³/t) are routinely observed (patents.google.com). When sterilization runs too long, ~1–3% of FFB weight can be lost as oil into this stream (ejournal.ft.unsri.ac.id) (patents.google.com).
Collection tanks and flash drums
Collection starts with dedicated piping and traps from each sterilizer to a central pressurized condensate receiver or atmospheric collection tank. Hot steam/condensate vents should be routed via steam traps or flash tanks to recover condensate, and outlet piping must handle both liquid and flashing steam. A small flash drum immediately downstream of the sterilizer allows bulk depressurization: steam flashes off to atmosphere or a vacuum pump, while hot condensate flows by gravity into a cooler collection tank.
All sterilizers tie into a common condensate header; gas‑tight valves on that header—as in patented designs—can time‑isolate “oil‑rich” versus “oil‑poor” phases based on the pressure profile (patents.google.com). Tanks are best heated‑jacketed or insulated to maintain ~70–90 °C (lower oil viscosity eases separation), and a surge/equalization tank smooths batchy flows (sterilizers cycle in batches). From the collection tank, condensate can feed a pretreatment train—packages of screens, skimmers, and first‑stage devices are grouped under primary separation systems.
For scale, typical flows span a wide range. At 100 t/d FFB, condensate is ~20 m³/d; at 3,000 t/d (~125 t/h), condensate is ~750–1,050 m³/d (assuming 0.1–0.35 m³/t; actual values vary with fruit moisture and sterilizer design). At 100 tph (2,400 t/d), condensate is 240–840 t/d (m³/d) at 0.1–0.35 m³/t FFB (patents.google.com). At 60 tph (1,440 t/d), it is 144–504 t/d (m³/d), with ~12 t/h observed in one case (www.researchgate.net). At 10 tph (240 t/d), it is 24–84 t/d (m³/d), with small mills seeing 2–8 m³/h. Assumes condensate density ~1000 kg/m³; volumes include both oil‑rich and oil‑poor portions (patents.google.com).
Oil–water separation pretreatment
The line into pretreatment should first use a coarse screen or mesh to catch fibers and solids; many mills specify an inline automatic screen for debris >1 mm. Downstream, a gravity clarifier skims floating oil as it rises; a baffled tank with weirs provides residence time before water overflows to the next stage, a configuration consistent with a clarifier in oily service.
To accelerate coalescence, inclined plates are common: a coalescing plate settler shortens rise distance for droplets. In hot, mildly corrosive service, a 316L unit such as a stainless plate settler suits long service life. Design rules of thumb in mills target 10–15 minutes of hold time at average flow to let 10–20 µm droplets coalesce and float. “White water clarifiers” (also called “API separators” or “skimmers”) are standard jargon.
Benchmarks indicate that well‑designed coalescing separators recover the vast majority of free oil: >80–90% removal of recoverable oil has been cited in palm service (www.huading-separator.com). For polishing, disc‑stack centrifuges—Huading’s PAND nozzle separator is one example—target the residual “third layer” of free oil after decanting, i.e., the last traces that escape gravity (www.huading-separator.com). Similar results are reported when skimming is followed by a coalescing cartridge filter stage.
Overall, capturing free oil from sterilizer condensate materially lifts crude oil yield. Even a 1% oil loss in condensate—common under average conditions (ejournal.ft.unsri.ac.id)—represents a non‑trivial revenue loss; recovering it can increase CPO output tens of tons per day in a large mill. After skimming, clarified condensate may still carry fine droplets and dissolved organics, so further treatment (e.g., flotation or biological pond) is typical before discharge. Dissolved‑air flotation is a common route to enhance oil removal in this phase; a packaged DAF system fits the role. Biological polishing to trim BOD/COD (biological/chemical oxygen demand) is also routine, including activated‑sludge basins; a conventional activated sludge train is one option for palm oil mill effluent (POME).
Heat recovery to boiler feedwater
Because sterilizer condensate exits at high temperature (~80–120 °C in many systems), it is a ready source of low‑pressure heat. A common strategy is simple: pass the hot condensate through a shell‑and‑tube or plate heat exchanger to warm incoming boiler feedwater. Pre‑heating from ambient to 60–80 °C trims boiler firing duty immediately.
Quantitatively, for a mill processing 1,000 t FFB/day (≈42 t/h), condensate might be ~100–250 m³/h at 80–90 °C. Cooling this stream to 30 °C releases roughly m·c·ΔT ≈ (100,000 kg/h·4.18 kJ/kg·K·50 K) ≈ 20,900 kW‑hours per hour (≈20.9 MJ/s, or 20.9 MW). Using this to heat feedwater from 30→60 °C (ΔT=30 K) at ~66,000 kg/h yields about 8.3 MW of heat transfer. In energy terms, that’s ~72 MWh per day (≈260 GJ/day) conserved. Even with exchanger inefficiencies and intermittent flow, recovery is substantial; one study estimated that recovering sterilizer exhaust steam could cut mill energy consumption by ~20% (www.researchgate.net). Palm oil mill case studies also show that 70% of sterilization steam is wasted without recovery (www.researchgate.net).
Case studies back the savings. Retrofitting an oleochemical plant’s boiler with waste‑heat recovery cut natural gas use by ~17% (www.mdpi.com). In sterilizers specifically, German patents describe passing condensate through a coil to yield ~100 °C hot water, then returning cooled condensate to the boiler make‑up (patents.google.com) (patents.google.com). In these designs the condensate cools to ~60–70 °C and is pumped back to feedwater tanks (patents.google.com).
In practice, a plate heat exchanger sized for ~1 t/h of condensate can lift feedwater by ~30–50 K. Given typical boiler efficiencies (~80–85%), recovering even 50% of condensate heat can cut boiler fuel needs by ~10–15% overall. Combined with flue‑gas heat recovery, the savings can be double‑digit percentage.
Integration, compliance, and payback
The economics stack up. Recovering oil and heat adds value: even a modest 1% increase in oil yield translates to ~5–10 t of CPO per day in a mid‑sized mill. At US$800/t CPO, recapturing 10 t/d is ~$8,000/d extra revenue. On the energy side, an oil‑fired boiler might save 5–10% of fuel with condensate heat exchange. If fuel costs $40/MWh and recovery supplies 2 MW thermal continuously (~48 MWh/d), that is ~$1,920/day saved. Payback on a small separator or exchanger (capex ~$50–100k) can thus be on the order of months.
Regulators are watching. Indonesia’s KLHK enforces strict effluent criteria; under PermenLHK No. 5/2021, mills must treat wastewater to low oil/grease levels before discharge (www.aquajaya.co.id). Better upstream handling reduces pollution and enables byproduct recovery—Biosargapark (2024) cites compost, biochar, and livestock feed as examples (www.aquajaya.co.id). Modern practice in Indonesian mills is moving toward integration: manufacturers now offer combined hot‑water tanks and oil‑skimming units, and guidelines stress “zero oil discharge” from sterilizers.
Summary design takeaways
A practical system does three things. First, collect all sterilizer drain water in a heat‑retaining tank sized for batch flows (with flash, traps, and a header that can time‑split “oil‑rich” and “oil‑poor” phases). Second, use oil–water separation—gravity and coalescence, upgraded by centrifuges where needed—to recover free oil; in palm service, >80–90% recovery of recoverable oil is reported (www.huading-separator.com). Third, route cooled clarified water to heat exchangers to preheat boiler feedwater; recovery schemes have been shown to trim boiler energy by ~20% in sterilizer contexts (www.researchgate.net) and ~17% in an oleochemical boiler retrofitting case (www.mdpi.com). Doing so recovers value (oil) and energy while aligning with Indonesian discharge standards (www.aquajaya.co.id).
Sources: condensate quantities (0.1–0.35 m³/t) and oil fractions are documented in patents and surveys (patents.google.com) (www.researchgate.net); without recovery, ~70% of sterilization steam is wasted (www.researchgate.net); separators like Huading’s PAND recover the residual free‑oil layer (www.huading-separator.com); and an MDPI case study reports ~17% fuel savings from boiler heat‑recovery in a palm plant (www.mdpi.com). Indonesia’s PermenLHK 5/2021 underscores the need for upstream oil–water separation (www.aquajaya.co.id).
