Refining Wastewater in 2026: Where Conventional Trains Fail and What CDFU plus CDOF Changes

Integrated refining and petrochemical complexes expanded quickly in China, and the wastewater side of those complexes did not always expand with them. The result is a familiar arithmetic problem: more barrels processed, more effluent generated, and a discharge permit that gets tighter rather than looser. Domestic refining enterprises now produce more than 1 billion tons of wastewater a year, most of it carrying petroleum-derived pollutants, emulsified oil, suspended solids and a long list of substances nobody wants in a river.

In 2026 the national discharge standards were tightened again. Processes that were adequate a decade ago are now being asked to deliver a quality they were never designed for, which is why refining companies are looking less for a marginal upgrade and more for a different starting point.

Four Properties That Decide Whether a Process Can Work

The first is the degree of emulsification. Oil droplets in refining wastewater are fine and tightly stabilised, and conventional separation equipment sized on droplet rise velocity simply cannot catch them. The second is composition: refractory organics, benzene series compounds, phenols and other toxic substances arrive together, so any single mechanism has to work in the presence of everything else.

The third is variability. Production changes move both flow and quality sharply, sometimes within a single shift, and a process with a narrow operating window spends much of its life outside that window. The fourth is the standard itself. When the limit is tight, a process that averages a compliant result does not pass; it has to be compliant on its worst day.

Where Each Conventional Route Stalls

Flotation is the obvious first answer and the most common disappointment. Bubble size is in the millimetre range, which gives poor removal of emulsified oil, and the equipment is bulky enough to consume land a refinery would rather use for process units.

Chemical demulsification works, but it works by spending money continuously. Large reagent doses carry high operating cost, and the residues they leave behind are a secondary pollution problem that has to be solved later in the same plant. Biological treatment asks a great deal of its influent: it needs a quality it can sustain, it needs a long start-up, and it does not react gracefully to the swings described above. Membranes separate rather than destroy, so they concentrate the problem into a reject stream, and they foul — which converts a treatment cost into a maintenance cost with an availability penalty attached.

The CDFU Stage: Cyclonic Field, Micro-Nano Bubbles, Flotation

The first of the two SINOKLE processes, CDFU cyclonic dissolved air flotation, combines cyclonic separation with micro-nano bubbles and dissolved air flotation in one vertical vessel. Produced water enters through a lift pump and forms a stable flow field under cyclonic action. Floating oil and suspended solids rise quickly on the bubble swarm and are discharged automatically through liquid level control, while treated water leaves through the bottom outlet.

Two mechanisms are doing the work. The cyclonic field creates a centrifugal force that drives the dispersed phase to where it can be collected, and the micro-nano bubbles give a collision probability that millimetre bubbles cannot. Neither mechanism depends on a chemical, so the separation stays physical and the sludge problem does not grow.

The CDOF Stage: Ozone Oxidation and Flotation in One Unit

The second stage, CDOF, is an organic combination of advanced ozone catalytic oxidation with flotation. It targets what flotation cannot: refractory organics that keep the COD above the limit, colour, odour, and the biological load that makes downstream disinfection unreliable. Because oxidation and separation happen in the same vessel, oxidised matter does not have to be re-pumped to a second piece of equipment, and the ozone that is already dissolved is used for both duties.

The practical applications listed for this stage are advanced wastewater treatment, decomposition of refractory organics, decolourisation and deodorisation, and disinfection. In a refining context that reads as a polishing stage after pretreatment, or as the main treatment stage where the effluent is going to reuse rather than to a municipal sewer.

How the Two Stages Compare with What Came Before

The comparison that matters is not a single efficiency figure but four at once. Bubble size moves from the millimetre scale to the micro-nano scale, which is the enabling change. Separation efficiency moves from 60-70% to above 90%. Equipment volume falls by 50%, and operating energy consumption falls by 20%. And the secondary pollution question changes character entirely: chemical reagent residues are replaced by purely physical separation.

The rest of the advantage list is engineering rather than chemistry. The process is physical throughout, so there is no reagent to buy, store, dose or dispose of. Micro-nano bubbles raise the probability of bubble-droplet collision well above what coarse bubbles achieve. The cyclonic field supplies centrifugal force on top of that. The structure is compact, the design is modular so capacity can be assembled from parallel trains, and control is automatic, which matters more than it sounds when the alternative is a shift team manually adjusting valves on a variable feed.

Two Reference Projects

The first is a wastewater treatment project at a petrochemical enterprise with a treatment scale of 2,000 m³/d. Influent oil content runs between 500 and 2,000 mg/L and is severely emulsified. Effluent oil content comes out below 10 mg/L with suspended solids below 20 mg/L, consistently meeting the standard rather than oscillating around it.

The second is an oilfield produced water project at a treatment scale of 5,000 m³/d, using a CDFU plus filtration combined process. Effluent quality meets reuse standards, which turns produced water from a disposal liability into a resource the field can put back to work.

What This Means for a Refinery Choosing a Process

Two things are worth taking away. First, the emulsified fraction is the part of refining wastewater that decides whether a train works, and it is the part that coarse-bubble flotation handles worst. Second, the cost of a process is not its capital number but its reagent bill, its sludge bill and its availability record added together over a decade. A physical train with a smaller footprint and no reagent consumption tends to win that comparison even when its equipment price is higher.

With the dual carbon goals moving through the sector, refining enterprises are being asked to treat wastewater to a standard and to account for the energy and waste involved in getting there. Processes that separate physically and oxidise in situ answer both questions at once.