Refiners are dialing in wash rate, mixing ΔP, and demulsifier dose to push effluent oil down to the tens of mg/L — with data to back every turn. The playbook: 5–10% wash, 0.5–2 bar shear, and 10–100 ppm chemistry, tuned to crude.
In crude desalting — the front door of the refinery that strips salts and water from incoming oil — small adjustments move big numbers. Lab work shows that flooding with 30% wash water (dilution water added upstream of the electrostatic coalescer) drives oil‑phase salinity down to ~2.8 mg/L, or a salt removal efficiency (SRE, share of salts taken out of the oil) of ≈99.46% (researchgate.net) (researchgate.net). Typical operations, however, hit >94–95% salt removal at just 3–10% wash — and that’s the heart of today’s optimization push.
The target isn’t only low chlorides in oil; it’s also clean water out. Properly run desalters deliver effluent brine (the separated aqueous phase) with <0.025% oil — often <250 ppm (parts per million) — according to SPE PetroWiki. In Indonesia, refinery effluent oil must be ≤50 mg/L (milligrams per liter) under MOE rules (researchgate.net). Hitting those numbers starts upstream, inside the desalter box.
Wash‑water rate and SRE trade‑offs
Wash fraction is directly tied to salt removal and oil–water separation. In the lab, 30% wash cut oil‑phase salinity to ~2.8 mg/L (SRE ≈99.46%) (researchgate.net) (researchgate.net), while industry‑typical 3–10% already yields >94–95% salt removal. One study reports that 5% wash at 52 °C with 0.5% demulsifier produced ~96.9% salt removal (researchgate.net).
Beyond ~10% wash, gains flatten: pushing to 30% adds only a few percent SRE but multiplies water usage. Practically, refiners run ~3–6% to meet chloride targets (e.g., <40 mg/L Cl⁻ in oil), then adjust for crude type. Industry guidance notes that “electrostatic coalescence is an increasing function of wash rate,” so heavy feeds may need more water or multi‑stage washing. The caution: excess wash expands brine volume and can increase oil carryover.
Mixing energy and pressure‑drop control
Mixing intensity governs droplet size before the electric field does its work (electrostatic coalescence, i.e., merging of water droplets under an electric field). The goal is enough turbulence to disperse wash into fine droplets for mass transfer, but not so much shear that the emulsion stabilizes. Mixing efficiency typically tops out at ~70–85% (petrowiki.spe.org).
Operators tune the mixing valve or static mixer ΔP (pressure drop across the mixer) after other parameters, with roughly 0.2–2 bar common depending on crude. Within this window, the “thin emulsion” responds to the downstream electric field. Too little ΔP leaves large water pockets and unwashed species; too much creates micron‑scale droplets, rag layers (a persistent interfacial emulsion layer), and slower breakup. Case evidence shows pushing ΔP beyond optimum produces finer emulsions and poorer separation. Adequate residence time and quiescent settling zones downstream further aid coalescence.
Demulsifier chemistry, dose and temperature
Demulsifiers (chemistries that weaken interfacial films so droplets merge) are central to breaking tight emulsions. Performance depends on type and concentration. In lab screening at 70 °C and 10 ppm, a fatty alcohol–ethoxylate delivered ~52% water separation and triethanolamine ~49% (scielo.br). Oil‑soluble products such as Basorol® E2032 reached ∼57% at 100 ppm (scielo.br).
Dose matters: increasing total demulsifier from 0.5% to 3% (vol%) with 5% water at 22 °C cut residual oil salinity from 442 to 48 mg/L (researchgate.net). An optimized blend — 60 ppm urea + 50 ppm triethanolamine + 40 ppm fatty alcohol ethoxylate — achieved ~98% water separation in a synthetic emulsion (scielo.br). Higher temperatures (to ~80 °C) accelerate film thinning; moderate brine salinity and near‑neutral pH (5–9) also favor breaking (scielo.br).
For asphaltene‑rich crudes, pairing multiple chemistries helps; adding an “asphaltene stabilizer” alongside a demulsifier improved separation in controlled studies (pubs.acs.org). In the field, demulsifiers are injected upstream — at feed or heaters — typically at 5–50 ppm, and even 100+ ppm for difficult cases. Plants often meter the chemical with an accurate dosing pump to hold ppm where effluent results demand it. Many operators standardize on an oilfield‑grade demulsifier, then refine type and dose through bottle tests or pilot skids; if oil‑in‑water in the brine trends above target, they raise dose or switch chemistries.
Integrated tuning and brine outcomes
A combined approach delivers clean separations. One Algerian lab study reported that 5% wash water + 0.5% demulsifier at ~52 °C drove salt removal to ~97% and produced oil with ~0.5% BS&W (basic sediment and water) (researchgate.net) (researchgate.net). In general practice, moderate wash (~3–5%) with adequate heat and ~20–100 ppm demulsifier meets salt specs, while the brine holds only trace oil; industry sources cite <0.025% oil (often <250 ppm) when desalters run well.
Regulatory anchors matter. In Indonesia, an oil ≤50 mg/L limit on refinery effluent (researchgate.net) sets a clear floor for brine polishing. Some sites add an upstream separations step, such as a compact deoiler, to trim residual oil before discharge or reuse. Where solids and dispersed oil must be handled together, plants frequently turn to a DAF unit downstream of the desalter water draw.
The cost signal is unambiguous. Increasing fresh‑water wash from 5% to 30% improved salt removal only a few percent but drove a nine‑fold jump in water use (researchgate.net) (researchgate.net). Once heating and basic mixing are set, fine‑tuning demulsifier type and ppm typically yields the biggest step‑down in oil carryover. In summary, maintaining 5–10% wash water (more for heavy crudes), setting mixing for ≈0.5–2 bar ΔP, and optimizing demulsifier at ~10–100 ppm can routinely achieve multi‑passion levels of separation, driving effluent oil down to only a few tens of mg/L.
Method notes and cited sources
The data points above are drawn from lab and industry documentation. For wash rate and SRE, see Sellami et al. (2016), which reports SRE ≈97% at 5% wash water and documents the diminishing returns between 5% and 30% (researchgate.net) (researchgate.net). For mixing performance and effluent expectations, see SPE PetroWiki (mixing efficiency ~70–85%) and its effluent oil note (<0.025% oil, often <250 ppm). On heavy crudes and wash‑rate effects on electrostatic coalescence, see JoinOil&Gas and its discussion of ΔP tuning.
For demulsifier chemistry and dose response, Hajivand & Vaziri (2015) provide quantitative separations at 10–100 ppm and temperature/pH effects (scielo.br) (link) (link). Sharma et al. (2024) document gains from pairing an asphaltene stabilizer with demulsifiers (pubs.acs.org). Indonesian benchmarks come from Helmy & Kardena (2015) (researchgate.net).
Sources: “Extensive studies and industry reports underlie these guidelines. For example, Sellami et al. (2016) found SRE ≈97% at 5% wash water (researchgate.net) (researchgate.net); survey monographs (SPE Petrowiki) report effluent oil <250 ppm in a well‑operated desalter (petrowiki.spe.org); and case documentation (JoinOil&Gas) explicitly notes that electrostatic coalescence improves with wash flow joinoilandgas.com. Empirical lab work with demulsifiers (Hajivand & Vaziri 2015) provides quantitative dose–response data (scielo.br) (scielo.br). These sources, together with Indonesian regulatory benchmarks (researchgate.net), form the basis for the parameter targets discussed above.”
References: Sellami et al. (2016) J. Pet. Environ. Biotechnol. 7:271 (researchgate.net) (researchgate.net); Hajivand & Vaziri (2015) Braz. J. Chem. Eng. 32(1):107–118 (scielo.br) (scielo.br); Helmy & Kardena (2015) J. Pet. Environ. Biotechnol. 6(5) (researchgate.net); Sharma et al. (2024) ACS Omega 9:12768 (pubs.acs.org); JoinOil&Gas (2024) “Emulsions, Oil Desalting, Dehydration Process”; Society of Petroleum Engineers (2025) PetroWiki: Desalting (petrowiki.spe.org).
