WhatsApp
betapramestiasia

Inside the quiet tank that makes fat-splitting work: design, dwell time, and the emulsion threat

  • beta-pramesti-asia
  • industry-oleochemical
  • process-fat-splitting-hydrolysis

Inside the quiet tank that makes fat-splitting work: design, dwell time, and the emulsion threat

Oleochemical fat-splitting lives or dies on a deceptively simple box: a horizontal decanter that cleanly peels fatty acids from glycerol. Its success depends on calm hydraulics, 5–10 minutes of residence, and stopping emulsions before they start.

Industry: Oleochemical | Process: Fat_Splitting_(Hydrolysis)

After hydrolysis, the mix of free fatty acids and glycerol/water is routed to a gravity settler built to separate a light oil layer from a dense aqueous layer. In practice, horizontal cylindrical gravity‑drainer vessels are common, with a feed zone to calm turbulence, a quiet settling section, and a controlled draw‑off at the outlet (vdoc.pub).

The design must buy time—typically 5–10 minutes—so droplets can coalesce and migrate by gravity. But too much shear at the inlet or surfactant carryover can spin that interface into a stubborn emulsion that stalls the entire split (fr.scribd.com | petrowiki.spe.org).

Horizontal settler layout and internals

Shell’s design guidelines for a horizontal two‑phase settler call for three compartments: an inlet compartment (with feed inlet device), a settling zone, and an outlet compartment with separate outlets or weirs for each phase (vdoc.pub). The inlet includes flow‑calming internals—such as an elbowed or slotted pipe feed with perforated “calming” baffles—to distribute feed evenly and minimize shear (vdoc.pub).

Shell recommends the inlet compartment length be at least 0.45×D (D = tank diameter) to allow the feed to decelerate before entering the settling zone (vdoc.pub). In the settling zone, drops coalesce and stratify into an upper oil (fatty acid) phase and a lower aqueous glycerol phase; vertical stile or horizontal scoop weirs control the interface and withdraw each layer (vdoc.pub). If one phase is very minor (e.g., glycerol ≲5% of flow), a “boot” configuration or simple weir may suffice for the small phase, whereas a full two‑phase separation is used when both phases are substantial (vdoc.pub). When the liquid interface is indistinct, design standards even allow double‑weir arrangements to stabilize level control (vdoc.pub).

To prevent re‑entrainment, designers add perforated baffles and vortex breakers on outlets, keeping flow calm where it matters most (vdoc.pub).

Residence time and velocity criteria

Vessel volume is sized to provide surge capacity and sufficient residence time. A widely used heuristic is 5–10 minutes of liquid hold‑up for coarse dispersions (coarse meaning droplet sizes roughly 100–300 µm, where µm = micrometers) (fr.scribd.com). The core principle: keep the continuous‑phase velocity below the terminal settling velocity of the dispersed droplets, commonly computed from Stokes’ law (the relationship governing settling of small spheres in a viscous fluid) (fr.scribd.com).

As a design basis, many sources assume ~150 µm droplets in a light oil; that yields a settling velocity on the order of 0.004 m/s (fr.scribd.com). In such a case, a 5–10 minute retention corresponds to roughly 1–2 m of settling distance, aligning with typical decanter heights of 2–3 m. The continuous‑phase velocity (flow/area) is then kept below ~0.004 m/s so most droplets reach the interface without carryover (fr.scribd.com | petrowiki.spe.org).

Why time matters, but not alone

Adequate residence time is crucial because gravitational coalescence is inherently slow. PetroWiki notes an exponential decay of droplet size with time: doubling hold‑up yields only incremental increases in coalesced droplet diameter (petrowiki.spe.org). Merely making the tank taller (more time) may not significantly improve separation if velocity is unchanged; widening the tank (lower upward velocity) better captures finer droplets (petrowiki.spe.org).

In practice, designers combine a base retention of ~10 minutes (for typical splits) with a safety factor for surge. Industrial practice indicates that with ~10 minutes hold‑up, more than 95% of dispersed glycerol can be removed under normal conditions. Empirical biodiesel data echo this: at ~40 °C and short hold‑up, gravity settling of glycerol from an oil‑rich phase reaches ~90–95% purity in ~10–15 minutes (at 1 g salt addition, g = gram) (pubmed.ncbi.nlm.nih.gov).

Emulsion formation at the interface

Interfacial emulsions arise from surface‑active impurities or vigorous mixing. Residual catalysts or neutralizing agents can form soaps (sodium/potassium salts of fatty acids), and unsaponifiables such as phospholipids and sterols act as emulsifiers; these amphiphiles adsorb at the oil/water interface and build elastic films, analogous to asphaltene/resin‑stabilized crude‑oil emulsions (www.sciencedirect.com). Fatty acids + soap + glycerol can even form structured lamellar gels; glycerol is known to stabilize 1:1 fatty‑acid/soap bilayers (pubmed.ncbi.nlm.nih.gov).

Mechanical shear at the inlet or pump discharge creates micron‑scale droplets that settle extremely slowly; without internal coalescers or gentle distribution, a recirculating vortex can re‑entrain separated phases. Design rules therefore prescribe perforated baffles and vortex breakers to suppress disturbances (vdoc.pub). Temperature and viscosity matter as well: cooler oil phases are more viscous and slow coalescence, while too‑hot conditions can flash vapor in glycerol; a moderate decanter temperature (e.g., 60–80 °C) is typically maintained to optimize densities and viscosities without volatilizing the aqueous phase.

Finely divided solids (e.g., coke, CaSO₄ from neutralization, silicates) can adsorb at the interface or onto droplets, forming Pickering emulsions that physically stabilize dispersions. Remedial measures include adding salt to the aqueous phase—1 g NaCl per 100 mL significantly reduced settling time by 100% with no loss of oil purity, though excess salt (5–10 g/100 mL) led to “miniemulsions” and poorer separation (pubmed.ncbi.nlm.nih.gov); heating; electrostatic coalescers; or chemical demulsifiers (petrowiki.spe.org). In practice, some plants pair a demulsifier program with accurate metering via a dosing pump and a formulated demulsifier to accelerate coalescence. Fundamentally, though, system design aims to minimize stirring at the interface and avoid emulsifiers: a calm settling section with gentle draw‑off helps prevent re‑entrainment. If a stubborn emulsion persists, operators may turn to decantation aids (coalescing plates, demulsifier salt) or even a dedicated centrifuge—recognizing that by design a basic decanter will only fully clear loose or weak emulsions (petrowiki.spe.org | petrowiki.spe.org).

Sources and references

Design heuristics and sizing rules from process engineering handbooks (e.g., Walas, Hooper) suggest ~5–10 minutes liquid hold‑up and ensure continuous‑phase velocity < droplet settling velocity (fr.scribd.com) (fr.scribd.com). Industry guidelines (Shell DEP 2008) specify horizontal settler layouts with inlet diverters and baffling to calm flow (vdoc.pub) (vdoc.pub). Academic and industry studies highlight the residence‑time/droplet‑size tradeoff (petrowiki.spe.org) (petrowiki.spe.org) and identify interfacial surfactants (asphaltenes/resins, soaps, glycerides) that stabilize emulsions (www.sciencedirect.com) (pubmed.ncbi.nlm.nih.gov). Empirical biodiesel reports illustrate strategies (NaCl addition) to accelerate glycerol separation by up to 100% (pubmed.ncbi.nlm.nih.gov). All figures and rules above reflect such data‑driven engineering guides.

References: see text citations (fr.scribd.com) (fr.scribd.com) (petrowiki.spe.org) (petrowiki.spe.org) (vdoc.pub) (vdoc.pub) (pubmed.ncbi.nlm.nih.gov) (www.sciencedirect.com) (pubmed.ncbi.nlm.nih.gov).