Sour Water Carrying 3,000 mg/L of Oil: Where Gravity Settling Fails and CDFU Kicks In

Refining sour water is one of those streams that looks manageable on a process diagram and turns hostile in the field. Composition shifts with the crude slate and the unit mix, and coking sour water sits at the harsh end: oil content can run between 500-3,000 mg/L, coke fines ride along at sizes mostly below 20 um, and the oil is heavily emulsified. Getting this water under control matters for three separate reasons, each capable of shutting a refinery down on its own.

Three ways untreated sour water hurts a plant

Production safety comes first. Emulsified oil and fine coke particles travel together and readily clog the sour water stripper, forcing unplanned shutdowns whose cost is measured in lost throughput rather than repair bills. Environmental and occupational exposure comes second: irritating gases carrying sulfides and ammonia nitrogen escape from open handling, polluting the surroundings and threatening employee health. Treatment difficulty comes third - conventional processes cannot hold discharge or reuse targets consistently, and every shortcut simply exports the problem to the wastewater plant downstream, overloading units sized for far milder feed.

A short history of failed fixes

Early practice leaned on gravity settling and cyclonic separation. Gravity settling needs 50-70 hours of retention time to do modest work, which means enormous tanks, a large footprint, and separation quality that still disappoints. Cyclonic processes improved throughput and compactness, but they handle emulsified oil poorly - droplets in the 0.1-20 um range resist centrifugal capture - and they clog easily while burning energy.

A later generation tried flotation coupled with inorganic ceramic membranes to chase higher precision. Precision improved, yet the economics collapsed: investment ran high, membranes fouled fast on oily feed, water recovery stayed low, and the process generated secondary pollution of its own. None of these routes solved the core problem, which is droplet size combined with fines.

The purely physical two-stage answer

SINOKLE rebuilt the flowsheet around CDFU (cyclonic dissolved gas flotation) supported by the KHC high-efficiency coalescing separator, producing a multi-stage train with no chemistry in it. The two units attack different ends of the droplet spectrum, and that division of labour is what makes the route work.

In the CDFU stage, cyclonic centrifugation and ultra-fine bubbles of 5-30 um operate together. The bubbles adsorb emulsified oil and coke fines with exceptional contact efficiency, and because no chemicals are added, nothing new enters the water. A single stage removes more than 90% of the oil load, which takes the pressure off everything downstream.

The KHC coalescing separator then finishes the job. Patented hydrophilic modified packing demulsifies the remaining fine droplets deeply and recovers crude oil of a quality that can return directly to refining. Separation precision reaches emulsified droplets of 0.1 um, and effluent oil content settles below 10 ppm once dissolved oil is excluded.

What the effluent and the balance sheet look like

Across the train, oil content falls below 100 mg/L and suspended solids removal climbs above 80%, with enough tolerance built in to ride out feed fluctuations without operator intervention. Because separation is purely physical, no oily sludge - classified as hazardous waste - is created, and every barrel of recovered oil carries economic value instead of disposal cost.

The hardware is compact by design: a skid occupies roughly 1/3 of the footprint a conventional process would claim, operates fully automatically in a closed loop with no gas leakage, and can be implemented online without tearing up the existing plant layout.

Why the chemistry-free choice pays twice

Choosing a physical route is usually framed as an environmental decision, but in sour water service the economics argue just as loudly. Every demulsifier and flocculant dosed into sour water ends up bound with oil in a sludge that must be manifested, transported and disposed of as hazardous waste, and that disposal line item recurs for as long as the plant runs. The physical train has no such ledger: the oil it captures is crude that re-enters the refinery's own accounting, and the only recurring costs are power and routine maintenance. Where water volumes are large, recovered oil alone has been enough to justify the equipment on payback alone, before any compliance benefit is counted.

Proven at two demanding refineries

Laboratory numbers only matter once they survive field duty. At Luoyang Petrochemical, a unit processes 15 m3/h of sour water with stable effluent oil below the specified standard. At the PetroChina Chad refinery, a second installation handles 20 m3/h under wholly different crude and climate conditions, with the same result. Together the two projects span the range of feed challenges refining sour water can present, from domestic heavy-oil operations to an African refinery running difficult local crude.

Why the combination holds up

The lesson from these deployments is that emulsified oil plus coke fines demands two matched mechanisms rather than one heroic unit operation. Flotation captures what centrifugation misses; coalescing packing converts what flotation cannot lift into recoverable oil. For refining and chemical enterprises balancing environmental compliance against cost, the SINOKLE train turns one of the plant's nastiest streams into a source of recovered crude - cutting costs and raising efficiency while the discharge numbers stay compliant.