Heavy Fuel Oil Plant Effluent: Why One Flotation Stage Never Holds at 2000 mg/L

Industrial oily wastewater has one universally acknowledged hard case, and it is the effluent from heavy fuel oil power stations. High oil content, heavy emulsification, awkward composition, and a discharge limit that has to hold every day rather than on a good day. Conventional processes struggle to meet it consistently at a defensible cost. The project SINOKLE delivered at a heavy fuel oil power station in Bangladesh for BR powergen limited is worth studying precisely because it shows a principle-level design surviving contact with a real site.

Read the droplet-size distribution before selecting anything

The influent here carries 2000mg/L of oil and the effluent target is no more than 10mg/L. A removal requirement spanning that range cannot honestly be met by any single-stage unit, and the reason sits in the size distribution of the droplets.

In high-concentration oily wastewater at 2000mg/L, free oil, meaning droplets larger than 100μm, usually accounts for less than 30% of the total. A large share of the oil exists as dispersed oil at 10-100μm and emulsified oil below 10μm, that is, droplets under <10μm. The design parameters of a single-stage flotation unit, whether bubble size, residence time or surface loading, can only be optimised for a relatively narrow droplet size band. Make the bubbles extremely fine to catch emulsified oil and free oil removal efficiency drops; prioritise free oil instead and a large amount of emulsified oil escapes.

Two stages in series is therefore not redundancy. It is a way of treating the droplet size distribution in segments: stage one takes large droplets at high throughput, stage two takes small droplets at high precision.

Stage one CDFU: cyclonic centrifugation and large bubbles for coarse separation

The first-stage CDFU has one core task: bring the oil content down rapidly from the 2000mg/L level to 300-500mg/L.

The dominant mechanism at this stage is cyclonic centrifugal separation. Wastewater enters the CDFU tangentially at velocity and creates a high-speed swirling field inside the unit. Within the centrifugal field the difference in the forces acting on oil and water is amplified: oil is slightly less dense than water, and under centrifugal force its tendency to converge toward the centre is several orders of magnitude stronger than under quiescent conditions. This process needs no bubbles at all and relies purely on fluid mechanics, which makes it highly effective on free oil and on the larger dispersed droplets.

The first-stage CDFU does release bubbles as well, but they tend to be medium to large, and combined with the turbulent mixing of the swirl they raise the probability of collision between bubbles and oil droplets substantially. Note what this stage is chasing: not precision, but throughput. In the high oil concentration range, remove the bulk of the oil first and free up concentration headroom for the polishing steps downstream.

Stage two CDFU: ultra-fine bubbles and coalescence for polishing

The second-stage CDFU works in a completely different range. It faces the residual finely dispersed oil and emulsified oil left at concentrations of a few hundred milligrams per litre, and its technical focus shifts accordingly, to ultra-fine bubble generation and coalescence demulsification.

Ultra-fine bubbles are tens of microns in diameter, sometimes smaller. According to Stokes’ law and bubble-to-oil-droplet collision models, the smaller the bubble the larger its specific surface area and the longer it stays in the water, and the higher its probability of contacting a fine oil droplet. Conventional DAF units typically produce bubbles on the order of 50-100μm, whereas the ultra-fine bubbles generated here reach a considerably smaller order of magnitude, and that difference is decisive for capturing emulsified droplets below 10μm in size.

Coalescence demulsification covers the other failure mode: droplets too stable for bubbles to adhere to. Emulsified oil persists because a layer of surface-active substances or a charge layer on the droplet surface forms a barrier. Coalescence demulsification alters the surface properties so that fine droplets can merge and grow when they collide, and once they reach a certain size bubbles can capture them effectively.

It is worth pointing out that coalescence demulsification is a physical process and does not rely on chemical agents. Its significance for process economics and hazardous waste reduction is considerable, and it is the reason the operating cost of this chain stays flat rather than tracking oil price and sludge disposal fees.

KFM active filter media: from tens of mg/L down to single digits

Second-stage CDFU effluent oil content can normally be held to a few tens of milligrams per litre. Getting from tens down to single digits requires deep filtration, and that is the last line of defence.

The core of the KFM active filter media filter is not simple mechanical sieving. If it relied on sieving, achieving micron-level filtration precision would produce a very large pressure drop and require very frequent backwashing. Instead it relies on the surface adsorption and depth retention offered by the active filter media.

The media has a high specific surface area and a particular surface oleophilicity. As water carrying trace oil droplets flows through the bed, the droplets are adsorbed onto the media surface by van der Waals forces and hydrophobic interaction, gradually forming an oil film around the media particles. Because the driving force here is physical adsorption rather than mechanical interception, high-precision oil removal is achieved at a relatively low pressure drop.

KFM also runs in depth filtration mode rather than surface filtration: the entire depth of the bed takes part in retention, giving it a far higher dirt-holding capacity than surface filtration cartridges. Backwash intervals are long and routine operation stays simple, which suits continuous operation on an industrial site with a small crew.

The design philosophy worth copying

Looking back over the whole chain: stage one CDFU for coarse separation dominated by cyclonic centrifugation, then stage two CDFU for polishing dominated by ultra-fine bubbles and coalescence demulsification, then KFM for deep adsorption and retention by active filter media.

The philosophy can be summed up as apply force in stages, reduce the load step by step, and give priority to physical methods. Do not try to solve every problem inside a single piece of equipment; let each stage do what it does best. And favour physical mechanisms over chemical agents, because chemistry is a permanent operating cost and a permanent sludge stream.

The actual operating data from the Bangladesh project, influent 2000mg/L, effluent <10mg/L, and stable operation to this day, is the best endorsement of that approach.