Industrial wastewater: choosing the treatment train

A municipal designer can begin from textbook per-capita loads. An industrial designer who does the same is guessing. Refinery effluent, brewery wastewater, tank-farm ballast water and a dairy CIP stream have almost nothing in common except that each will ruin a treatment plant designed for one of the others. Every sound industrial scheme I have reviewed started the same way: measure the streams, then segregate them.

Characterise, then segregate

The first engineering decision is not a technology; it is drainage. Oily process water, sanitary sewage, contaminated storm water and clean cooling blowdown belong in separate systems, because mixing them converts a small difficult stream into a large difficult stream. At the Humber refinery effluent extension the entire process strategy rested on keeping the biologically treatable flows apart from surges the biology could not survive.

• Flow and load balance across production cycles — a week of composite sampling minimum; a campaign plant needs a full campaign.
• COD/BOD ratio: above roughly 2.5, expect a significant non-biodegradable fraction and plan a physical-chemical stage.
• Inhibitors and toxics: sulphide, phenols, heavy metals, quaternary ammonium biocides — each caps what a biological stage can accept.
• Temperature and salinity — Gulf industrial effluents routinely arrive at 40 °C and brackish strength, which changes oxygen transfer, settling and membrane choice.

The standard building blocks

Stage	Typical technologies	What it removes
Primary / oil	API separator, CPI plates, hydrocyclones	Free oil, gross solids
Physical-chemical	Coagulation, DAF, precipitation	Emulsified oil, colloids, metals, phosphate
Biological core	Activated sludge, MBBR, MBR, anaerobic (UASB/EGSB)	Soluble BOD/COD, ammonia
Polishing	Sand/UF filtration, GAC, ion exchange	Residual solids, trace organics
Advanced	Advanced oxidation, RO, evaporation	Hard COD, TDS — reuse and ZLD duties

Two selections carry most of the risk. The biological core must match the effluent's strength and variability: high-strength, warm, soluble streams favour anaerobic reactors that produce biogas rather than consume aeration power; dilute variable streams favour robust aerobic systems with equalisation ahead of them. And the polishing stage must be selected against the permit, not the brochure — a plant that meets 95 per cent of its consent 95 per cent of the time has failed.

Reuse changes the arithmetic

In the Gulf the discharge permit is rarely the whole story: water scarcity prices treated effluent as a resource. Once a membrane polishing stage exists, the marginal cost of reuse-grade water is small, and cooling-tower make-up or irrigation demand inside the fence frequently pays for it. The right question at concept stage is therefore not “how do we dispose of this” but “how much of it do we want back” — the answer moves the whole train.