Desalter Wastewater: Coalescing Plus CDFU or Two-Stage CDFU? A Practical Selection Guide

Start with the influent, not the equipment

Refinery teams rarely argue about whether desalter effluent needs treating. The argument is about which route, and the answer mostly falls out of two measurements: how much oil is in the water, and how tightly it is emulsified. Get those two right and the configuration choice becomes close to mechanical. Skip them and you will size equipment against a guess.

SINOKLE builds its desalter offering around two patented units, and they do different jobs. The high-efficiency coalescing oil remover leans on the wetting and coalescing behaviour of special media. Fine emulsified droplets collide repeatedly as they pass through, merge, and eventually break out of the colloid and asphaltene interfacial film that was holding them in suspension. Its job is demulsification, and it is the piece that decides how well everything downstream performs when the feed turns ugly.

The CDFU cyclonic dissolved air flotation unit couples cyclonic separation with dissolved air flotation. Air is dissolved and released under sealed, pressurised conditions to produce a dense micro-bubble field that attaches to oil and lifts it out. Its job is advanced oil removal. Because the whole thing runs sealed and pressurised, there is no open tank, no vapour escaping, and no operator standing over it.

Two configurations, two duties

Scheme 1, coalescing oil remover plus CDFU

Built for high oil content and severe emulsification. Water enters the coalescing vessel first, where most of the free oil separates and the emulsion is broken, then moves on to CDFU for polishing. The two stages do different things and each one makes the other’s job easier, which is precisely where the shock tolerance comes from. When the feed spikes, the front end absorbs it and the back end still sees something it can handle.

Scheme 2, two CDFU stages in series

For lower and steadier oil loads. Water passes through two CDFU units one after the other, polishing in steps. The line is simpler and the footprint smaller, at the cost of some resilience when the feed swings. If you have watched the influent for a full month and it barely moves, this is the configuration that makes sense.

Parameters side by side

The figures below assume influent oil content of 0.5%-5%, which covers the range most units actually see.

• Applicable scenario: Scheme 1 takes high oil content with severe emulsification; Scheme 2 takes influent oil content 0.5%~5%.
• Configuration: coalescing oil remover + CDFU versus two-stage CDFU.
• Effluent oil content: less than 50mg/L on both.
• Oil removal rate: above 90% on both.
• Water content of recovered crude oil: less than 1% on both.
• Chemical dosing: none at any point, both routes are purely physical.
• Operating mode: sealed, pressurised, automatic.

Against the conventional options

Conventional processes are not worthless. Each has conditions where it is the sensible answer, and pretending otherwise does no one any favours. The point is what happens when emulsification is severe and the feed swings.

• Gravity settling: weak on emulsified oil, large footprint, low efficiency.
• Cyclonic oil removal: average emulsified-oil removal, sensitive to fluctuation.
• Membrane filtration: fairly strong but prone to clogging, with high membrane replacement cost.
• Coalescence + CDFU: strong emulsified-oil removal, no chemicals, low O&M cost, small footprint, high automation, and it can be installed and commissioned online.
• Oil recovery: crude oil water content under 1%, returned for refining, versus difficult or mediocre recovery elsewhere.

The maintenance side of the decision

Site teams weigh operation and maintenance at least as heavily as separation performance, and here the relevant features are cumulative rather than dramatic. No chemical dosing anywhere means no make-up, no storage and no reagent purchasing. Sealed pressurised operation means no odour and no spillage on the pad. Skid-mounted construction means the unit is positioned quickly after delivery and commissioned online while the desalter keeps running, which for a continuously operating refinery is often the deciding factor. Automation is high enough that routine inspection is most of the daily workload, and operating parameters regulate themselves.

None of this shows up in a performance table, which is why it gets missed in comparisons. But a route that needs no chemical logistics, no shutdown window and no dedicated operator changes the operating budget in ways a removal percentage cannot express.

How to choose

If the water is heavily emulsified and the oil content moves around, take Scheme 1. Coalescence breaks the emulsion, flotation finishes it, and the pair absorbs shock loads better than anything else in this space. If the influent stays inside 0.5%-5% with modest variation, Scheme 2 is enough and saves space.

Before committing, run a survey. Track peak oil content, degree of emulsification and the fluctuation band over a representative period, then put that next to the available area and the retrofit window you actually have. Decide at the same time where the recovered crude will go. Treating the treatment and the recovery as one account usually changes which scheme wins, and it is far cheaper to settle that question on paper than after the skid lands.