Reading Oil Content in Acidic Emulsified Water: Where the Standard Method Slips

Oil content data is never just a number on a certificate. It decides how a treatment train is tuned, whether a permit condition holds, and whether a discharge can be defended in front of a regulator. In China the reference route is the infrared spectrophotometric method set out in HJ 637-2018, "Water Quality — Determination of Petroleum Oil, Animal and Vegetable Oils — Infrared Spectrophotometric Method". An extractant such as carbon tetrachloride or tetrachloroethylene lifts the oily material out of the sample, the extract is dehydrated and dried, and an infrared oil analyzer reads absorbance at a chosen wavelength.

That works well on clean, neutral samples. Petrochemical plants, metal processing shops and fermentation plants rarely supply clean, neutral samples. What they send to the bench is acidic, and it is emulsified to a degree that the method was never really designed around.

Four bottlenecks that appear on the bench

Extraction efficiency falls away. Low pH changes the form in which oily substances exist and weakens the affinity between extractant and oil droplets. In a strong emulsion the droplets sit in the micron range, sometimes smaller, wrapped by surfactants or solid particles in a firm interfacial film that keeps solvent and oil apart.

The emulsion layer resists breaking. The textbook tools are demulsifiers such as aluminum sulfate, acidification with HCl, or centrifugation. On a genuinely stable emulsion they underperform: after extraction the oil-water-solvent boundary is blurred and a persistent middle layer remains, which loses oil and makes quantitative transfer awkward.

Accuracy and precision both suffer. Incomplete extraction plus entrainment losses inside the emulsion layer pull measured concentrations below the true value, and the spread between parallel samples widens.

Operators take risks to compensate. Reaching for more strong acid or repeated extractions raises exposure to hazardous solvent and the chance of chemical burns, and it multiplies hazardous waste.

Break the emulsion before touching the extractant

There is no need to abandon the standard principle. What pays off is strengthening the front end. For an unknown or strongly emulsified sample, run a small-scale trial in a separatory funnel: add an appropriate amount of NaCl first so the salting-out effect disrupts the electric double layer, then drip HCl or H2SO4 slowly to pH ≤2 while gently shaking and watching the interface. Stepping the pH gradually rather than jumping to a target helps locate the best demulsification point.

Physical assistance complements the chemistry. Refrigerated standing at 4°C for several hours suits some systems; a 40-50°C water bath suits others, though volatilisation needs watching. Under controlled temperature and time, a short low-frequency ultrasonic treatment uses cavitation to tear open emulsion droplets and interfacial films. Ultrasonic power and duration have to be tuned — push too hard and the sample simply re-emulsifies into something finer than before.

Demulsifier selection is its own screening exercise. Cationic and non-ionic polymer products are worth testing on streams such as metalworking fluid wastewater, but every candidate must pass a spike recovery check to confirm it introduces no infrared interference and does not adsorb or wrap oil into a negative bias.

Refine how the extraction itself is run

Replace one large-volume extraction with several small ones. Using 10-15 mL of fresh extractant each time and repeating 2-3 times on the same water sample, then combining the extracts, lifts overall efficiency — particularly where the partition coefficient is unfavourable.

Shake by observation rather than by stopwatch: vigorous shaking for 30 seconds, standing for layer separation and inspection, then more shaking, adjusting total contact time according to how the emulsion layer shrinks. If a layer still appears, take one or more of three routes: filter through glass wool or an anhydrous sodium sulfate column (adsorption loss verified beforehand); draw the layer off with a microsyringe or thin tube and add a little anhydrous ethanol or a stronger polar solvent such as dichloromethane for secondary demulsification, combining it with the main extract; or transfer the whole layer to a centrifuge tube and, after high-speed centrifugation, take the clear lower extractant phase.

Calibrate against the matrix, not against pure water

Calibration curves built on clean water quietly encode an error. Prepare standards in blank water adjusted to the same acidity and dosed with the same demulsifier as the unknowns, so the matrix effect is offset before the instrument ever sees the sample.

For each batch, or at least each difficult sample type, run a whole-process spike recovery with spiking points spanning the expected concentration range. A recovery in the 80-120% band is the practical test of whether the pre-treatment held up, and it can be used to correct results when it did not. Method blanks matter too: emulsified samples leave residue, so vessels need soaking in dedicated detergent and solvent, and blank frequency should rise to keep the background low and stable.

What the change delivers in practice

Take an acidic metalworking wastewater at pH=3.5, severely emulsified. Following the original procedure strictly, spike recovery of the target oil fraction came out around 65%. Applying salting out plus gradient acidification plus ultrasound assistance for demulsification, combined with three small-volume extractions, moved recovery to about 95%. The relative standard deviation went from over 15% down to within 5%.

That gap matters. A result sitting 30 points low does not merely misdescribe one sample; it misleads every decision taken on top of it.

Work on this front is part of what SINOKLE does when it validates treatment performance on acidic coking and metalworking streams, where emulsion stability is exactly the variable that makes or breaks a measurement.

Turning the method into a routine

Three habits make the improvement stick. First, write it down: for each recurring difficult water type, fix the demulsifier type, dosage, degree of acidification and number of extractions through experiment, and codify the result as an internal standard operating procedure. Second, train judgement rather than steps — operators who can read an emulsification phenomenon and adapt are worth more than operators following a fixed recipe, and their demulsification "recipes" should be recorded against each water type to build a knowledge base. Third, talk to instrument suppliers about automated or semi-automated pre-treatment hardware with better interference resistance and lower detection limits.

The determining variables are demulsification and complete extraction; everything else follows from them. (Any specific chemical or operation described above belongs behind normal laboratory safety rules, and method changes need verification under the laboratory quality management system.)