DAF case studies and published performance data

Equipment suppliers publish case studies about their own projects, which is exactly the evidence a buyer cannot verify. This library does the opposite: it collects dissolved air flotation performance that has been measured, written up and published by someone else, with the source cited so you can check it.

What this is, and what it is not

Every figure on this page comes from a peer-reviewed paper, a research-council report or a published full-scale study, and every entry names its source. None of it is our own project work presented as a reference. Where a study reports poor performance, it is included on the same terms as one reporting excellent performance — the badly optimised plants are often the more useful reading.

Use these as evidence of what dissolved air flotation achieves on a given stream, and as a sanity check on any number a supplier quotes you. Do not use them as design values for your own plant: influent varies, and the only defensible design basis is a jar test on your effluent.

26Southern African plants surveyed
43–99%Reported TSS removal range
49–99.9%Reported FOG removal range
1986–2025Publication span

The evidence table

Stream and settingScaleReported influentReported removalChemistrySource
Algae-laden reservoir water
Lake Chivero, Harare, Zimbabwe
60 L pilot, 7 m/h, 300–400 kPaTurbidity 3.14–4.04 NTU
Chlorophyll-a 0.36–2.12 µg/L
pH 7.2–8.4
Turbidity −64%
Chlorophyll-a −95%
(Conventional settling on the same water: −53% and −52%)
Alum 70 mg/L, sulphuric acid to pH 6.5–6.9Water Practice & Technology 18(1) 102, 2023
Slaughterhouse effluent
Improved primary treatment
Full scaleNot fully stated in abstractFOG −95%
COD −80% across primary treatment
Coagulation and flocculation ahead of DAFWater Science & Technology 85(5) 1688, 2022
Poultry slaughterhouse
Operating strategy evaluation
Full scaleSlaughterhouse effluent, high FOG and proteinSS −43 ± 15%
FOG −49 ± 8%
24 mg/L as Al³⁺ plus 1.5 mg/L anionic polymerResources, Conservation and Recycling, 2008
Abattoir wastewater
Early full-scale evaluation
Laboratory and full scaleAbattoir effluentReported COD, SS and FOG reductions commonly 50–80% depending on air pressure and flocculantVaried by trialWater Research 20(1), 1986
Municipal effluent, oil and grease
Pressure and recirculation optimisation
Pilot, 1–5 bar, recirculation 1–20 minTS 800 mg/L
TSS 590 mg/L
Oil & grease 450 mg/L
BOD₅ 360 mg/L
COD 710 mg/L
O&G −84.4%
BOD₅ −88.9%
COD −88.7%
TSS −85%
Air only; Raschig-ring absorption columnWater (MDPI) 17(24) 3474, 2025
Palm oil mill effluent
Liquid–solid separation
Reported studyDigested POMETSS −90%
BOD −80%
560 kPa saturator pressureEnvironmental Technology Letters, 1988
Palm oil mill effluent
Flotation plus membrane bioreactor
Combined processPOMECOD −97%
TSS −93.9%
Turbidity −99.8%
Fats and oils −99.9%
Air flotation pretreatment ahead of anaerobic membrane stagePublished combined-process study, 2016
Southern African plants
Clarification and sludge thickening
Survey of 26 operating plantsRange of municipal and industrial dutiesDesign and operating parameters compiled across the sampleVariousWRC Report TT 60/93, Water Research Commission, 1993; Water Science & Technology 31(3–4) 203, 1995
Drinking water plant, South Africa
Recycling of DAF residue
Full scaleSurface waterFeasibility of recycling wastewater including DAF residue within the worksConventional potable coagulationWater SA, July 2019
Biodiesel wastewater
Acidification and coagulation
Reported studyBiodiesel process water, very high FOGSubstantially enhanced flotation efficiency after acidification ahead of coagulationAcid split then coagulantDesalination, 2011
Groundwater
TOC, colour and arsenic
Reported studyGroundwater with organic colour and arsenicRemoval of TOC, colour and arsenic assessed for flotation dutyCoagulation ahead of flotationApplied Sciences 15(20) 11255, 2025
Figures are as reported by the cited sources. Where an abstract does not state the influent, that cell says so rather than guessing. Percentage removals across different studies are not directly comparable — influent, coagulant, dose, pressure and recycle ratio all differ.

What the spread actually tells you

Put the poultry slaughterhouse study next to the improved-primary slaughterhouse study and the point of this library becomes clear. Both are full-scale, both are meat processing, both use flotation. One reports 49% FOG removal; the other reports 95%. The difference is not the technology.

  • Dose is the first variable. The poultry study ran at 24 mg/L as Al³⁺ with 1.5 mg/L of anionic polymer. That is a modest coagulant dose for a stream that heavily loaded, and the removals reflect it.
  • Coagulation pH is the second. Aluminium coagulants outside 5.5–9.0 form poor floc regardless of dose, and slaughterhouse streams swing with cleaning cycles.
  • Flocculation time and shear is the third. Floc that has not had 10–20 minutes of gentle mixing will shear apart at the recycle nozzle whatever the chemistry.
  • Air-to-solids ratio is the fourth, and it is the one most often left at whatever the commissioning engineer set on the day.

A supplier quoting you 95% FOG removal is not lying — that figure is published and achievable. Whether your plant reaches it depends on four variables that are set at design and defended in operation. See the design parameters reference and the troubleshooting guide.

Case studies in detail

Further detailed entries are being added as the underlying papers are read in full rather than summarised from abstracts. The evidence table above is updated as new published data appears.

The Southern African design guide

One reference deserves separate mention. In 1993 the South African Water Research Commission published A South African Design Guide for Dissolved Air Flotation (Report TT 60/93), built by combining a literature survey with a detailed survey of 26 dissolved air flotation plants then operating in Southern Africa, covering both water clarification and activated sludge thickening. It remains, three decades on, the most substantial body of African DAF operating data in the public domain, and it is available free of charge from the Water Research Commission.

Anyone specifying flotation for an African plant should read it. The companion papers in Water Science & Technology volume 31, issues 3–4 (1995) summarise the same work in journal form.

Frequently asked questions

Why do you publish other people’s case studies rather than your own?

Because a supplier’s own case study is the one piece of evidence a buyer cannot check. Published work has been through review, states its method, and can be read in full by anyone who wants to argue with it. If we later publish our own project data, it will be labelled clearly as ours and held to the same standard of stating the influent, the chemistry and the method.

Can I use these removal percentages to size my plant?

No, and no responsible supplier would let you. These figures tell you what flotation is capable of on a broadly similar stream. They do not tell you what your stream will do, because influent characteristics, coagulant response, temperature and pH all differ. Use them to sanity-check a supplier’s claim; use a jar test on your own effluent to size the plant.

Why is a study reporting only 43% solids removal included?

Because it is real, it is full-scale, and it is more instructive than the successes. It shows what an under-dosed flotation plant achieves, which is the outcome a buyer is actually at risk of. Selecting only the flattering results would make this page marketing rather than evidence.

Are there published DAF case studies from African industrial plants?

Fewer than there should be. The substantial African material is the Water Research Commission survey of 26 Southern African plants, the Harare algae study, and a small number of South African drinking-water papers. Most African industrial flotation installations are never written up, which is precisely the gap this library exists to narrow. If you operate one and are willing to have the data published, we would rather publish it than not.

How often is this page updated?

As new published data is found, and reviewed at least quarterly. Each entry carries its publication year so you can see how current the evidence is. Studies from the 1980s are included where they remain the best available data on a stream — flotation physics has not changed, though coagulant chemistry has.