Coking Acidic Wastewater: Coke Fines, Emulsified Oil and the Stripper Trap
Where the water comes from
Coking acidic wastewater is a highly challenging organic stream: toxic, and hard to biodegrade. It originates mainly from delayed coking, catalytic cracking and hydrocracking in petroleum refining, and also from coking, gas purification and the recovery and refining of chemical products, plus coal chemical operations such as high-temperature carbonisation, gas cooling, coal washing and wet quenching. The specific points are the fractionator overhead reflux drums and the small water feed and blow-down steam of the coke drums. Sour water from the overhead reflux drums is discharged continuously, which matters for sizing: this is not a batch problem.
Four characteristics that decide the process
- Fine coke at high load. The water carries a heavy load of coke fines of very fine particle size, which readily affects equipment and process, for example clogging the stripper.
- High oil, severe emulsification. The oil droplets are equally fine and severely emulsified. This emulsified oil markedly reduces stripping efficiency and liquid ammonia quality, and adds to treatment difficulty.
- Complex composition. High concentrations of sulphides, ammonia nitrogen, phenols, coke fines, emulsified oil and suspended solids, plus organics such as benzene and phenolic compounds, aldehydes, ketones, tar and its derivatives, and harmful substances such as cyanides, acids and alkalis, and volatile organic compounds.
- Low pH, in most operations strongly acidic. The corrosivity shapes both metallurgy and the choice between chemical and physical demulsification.
Inorganics include ammonia nitrogen, cyanides, sulphides and suspended solids such as coal ash and cinder. The organic fraction splits into benzene and phenolic compounds, aldehydes and ketones, and tar derivatives carrying polycyclic aromatic hydrocarbons, organic sulphur and nitrogen compounds.
The stripper failure nobody budgets for
If this water enters the stripper without pretreatment, the coke fines combine with slop oil under the action of the steam heat source to form a viscous gel. That gel blocks the valve trays. Treated water quality drops, stripping performance collapses, and in severe cases the whole unit has to be shut down and cleaned, which lengthens the production cycle and raises cost. The failure is gradual, so it is usually discovered when stripping efficiency has already gone.
Why biology struggles
The complex composition and the high content of recalcitrant organics make it hard for microorganisms to metabolise and degrade effectively, so a biological train is difficult to stabilise. Conventional methods also perform poorly here: the composition is too varied for a single conventional step to meet the required standards and remove the various pollutants effectively. Pretreatment has to change the physical state of the oil first, not merely the chemistry.
A purely physical route: CDFU and KHC
The SINOKLE process is patented and built around CDFU cyclonic dissolved-gas flotation plus coarse and fine filters, with the KHC high-efficiency coalescing oil separator as the second core unit. No chemicals are involved at the demulsification step, which is the point: the recovered slop oil stays clean enough to reprocess.
CDFU works as the front-end pretreatment unit. Through ultra-fine nano-bubble action it carries out preliminary demulsification of the emulsified oil and achieves rapid oil-water separation under fluctuating conditions, which matters because coking units rarely feed a steady stream.
KHC then finishes the job. The coalescing oil separator delivers purely physical demulsification that is effective and stable, and the recovered slop oil is of good quality with low water content and can be reprocessed, saving energy and delivering both environmental and economic benefit.
What operators actually gain
Operation is fully automatic, enclosed and pressurised, with no safety or environmental exposure. The train is mainly static equipment with a high degree of automation, little maintenance, long inspection intervals and good reliability. Because no chemicals are added, no scum or slop-oil residue is produced and there is no secondary pollution, so operating and secondary treatment costs stay low. The equipment occupies little space, separates oil and water thoroughly, resists shocks well and delivers stable results across the load swings a coking unit throws at it.
Commissioning sequence that avoids the usual surprises
Bring the train up on circulating water first, not on process fluid. Set the CDFU bubble field and the cyclone pressure to their design points before any oil is present, because emulsified feed makes it almost impossible to tell a badly tuned unit from a genuinely difficult water. Once the unit is stable, introduce the sour water at a fraction of design flow and raise it in steps, watching differential pressure across the coalescing element. The number to record at each step is not oil removal alone but the pressure drop trend, since that is what predicts whether the KHC element will foul in weeks or in months. Only after the trend is flat should the filters be commissioned, and they should be commissioned with the media in place, because a filter run without its real media teaches nothing about caking behaviour.
Where the process boundary sits
Two limits are worth stating. The train is a pretreatment and oil-removal system, not a substitute for the stripper: ammonia and sulphide recovery still happen downstream, and the value of this route is that it protects that downstream equipment. Second, the recovered slop oil has to have a route back to the crude or slop system; where a refinery has no such route, the economic case rests on sludge reduction and equipment protection alone, which is still usually enough to justify the retrofit.