Refinery Wastewater Pretreatment: Why Horizontal-Flow Flotation Falls Short and What Replaces It
Refinery wastewater is a compounded problem: crude oil, sulfides, phenols, COD-bearing organics and other pollutants arrive together, under environmental requirements that tighten with every revision. The classical train - oil separation, flotation, biochemical treatment - was built for an earlier era of limits, and its pretreatment stage is where the age shows first.
Where the traditional pretreatment stage falls short
Horizontal-flow flotation generates bubbles by aeration to adsorb oil. It is simple in structure and cheap, and for lightly loaded streams it is adequate. The weaknesses are structural: bubble diameters run 50-100 μm, so contact area with pollutants is small and treatment efficiency sits at 60%-70%; tanks demand 30-60 minutes of residence time and a large footprint; scum accumulates easily and maintenance is frequent. Shallow (compact) flotation trims the bubble distribution and lifts efficiency to 75%-85%, but the equipment is complex, energy consumption is higher, and on high-concentration feed - oil content above 500 mg/L - performance stays unstable and misses current discharge standards.
Some refineries stack combinations such as inclined-plate settling tanks with walnut shell or fiber ball filtration, or settling tank plus centripetal flotation plus fiber ball or multi-media filtration and ceramic membrane. These met basic treatment needs for a time, but the defects compound: unstable performance, poor shock resistance, a tendency to form aged oil, complex process flow, large footprint, difficult operation and maintenance - and chemical dosing that may cause secondary pollution. Push feed concentration or flow fluctuation high enough, and effluent quality wanders out of compliance.
Bubble size deserves its own explanation, because it drives everything else. Flotation capture probability falls steeply as bubbles grow: a 50-100 μm bubble sweeps a large volume but collides inefficiently with fine oil droplets, while ultrafine bubbles attach to droplets and rise with them reliably. Shrinking the bubble shifts the separation from sweeping enough water long enough to contacting each droplet at least once - which is what collapses residence time from an hour to a few minutes and tank volume by the same ratio.
The CDFU rework
The SINOKLE CDFU cyclonic dissolved air flotation unit integrates cyclonic centrifugal separation, dissolved air flotation and ultrafine bubble technology. Single-stage oil removal runs above 90% with suspended solids removal above 85%, and effluent oil holds below 5 mg/L. Residence time is 1-5 minutes - 3-8 times more efficient than traditional flotation. Demulsification is purely physical, so no chemical dosing enters the train, oily sludge is eliminated at the source, and secondary pollution never starts. The skid-mounted structure occupies about 1/3 of traditional equipment, and IoT integration supports remote monitoring with fully automatic unattended operation.
Two field references
At a petrochemical company in Cangzhou, Hebei, the coalescing pressure deoiler plus CDFU cyclonic dissolved air flotation combination was applied to electro-desalting wastewater and achieved efficient water purification. At a petrochemical company in Xinjiang, the coalescing pressure deoiler plus two-stage CDFU configuration handled black water at a single-unit treatment capacity of 150 m³/h with stable, compliant effluent. Both projects ran as retrofits into operating plants - the modular design installs easily with a short construction period, which is precisely what upgrading an existing refinery requires.
Sludge handling is the hidden line item. Chemical dosing regimes create an oily chemical sludge that must be dewatered and hauled - a cost that survives every efficiency improvement upstream. Physical demulsification produces no such stream: separated oil leaves as recoverable product and solids follow the normal sludge path. Over a decade, that difference compounds into one of the larger items in the lifecycle cost comparison, and it is entirely independent of the effluent numbers.
Instrumentation completes the picture. Both field trains run unattended because the control loop reads inlet and outlet oil, differential pressure across the coalescing internals, and recycle gas pressure on the flotation side, then trims operating parameters within their design envelopes automatically. Alarms route to the plant DCS, so operators treat the skid as another unit in the water balance rather than a special project. That integration - mechanical simplicity plus standard signals - is what lets a small package carry a duty that used to justify a tank farm and a full-time crew.
Comparison, honestly framed
Against the traditional train, CDFU leads on three axes at once: treatment performance, operating cost and environmental profile. Purely physical demulsification keeps the process green and saves the chemical budget; the compact, intelligent package saves space and labor; the fast residence time shrinks everything between the desalter and the bioreactor. As environmental policies tighten and the refining industry upgrades, the technology is becoming a preferred pretreatment solution - not because any single number is dramatic, but because 90% single-stage removal at 1-5 minutes with no chemicals changes what the downstream stages have to survive.
Selection notes for retrofit planners
Three checks make the evaluation concrete. First, sample at the worst case - backwash windows and crude slate changes - not at design average. Second, verify the emulsion: ultrafine bubbles plus swirl handle emulsified feed that defeats large-bubble flotation, but the coalescing stage in front should be sized to the measured droplet spectrum. Third, check the plot plan: the reduced footprint usually means the new skid fits inside the old tank's shadow, which converts a capital project into a tie-in exercise. Run those three checks and the comparison against the incumbent flotation cell tends to settle itself.