DOCA sensor trials under heavy particulate loading in compressed air

The DOCA Project has been refining an optical sensor capable of detecting oil contamination across liquid, aerosol, and vapour phases in high-purity compressed air. As work moves from laboratory prototyping toward industrial demonstration, one of the most demanding validation activities has been stress-testing the device against streams heavily laden with dust, rust, desiccant fines, and ambient aerosols.

For Australian stakeholders evaluating the technology, these trials carry particular weight. Compressed-air networks stretching from Pilbara iron-ore operations to Melbourne's biopharmaceutical manufacturing clusters routinely deal with intake air carrying red-soil particles, pollens, and process-generated dust alongside any oil that may be present.

Why dust and oil confuse optical detection

Optical sensing relies on how light interacts with the target substance. Oil droplets scatter, absorb, and fluoresce at characteristic wavelengths, allowing the DOCA sensor to distinguish clean air from contaminated flow. Particulates, however, also scatter and absorb light, and many mineral dusts fluoresce in their own right. When a beam passes through air carrying both iron-oxide particles and compressor oil aerosols, the detector receives a mixed signal that can mask the oil signature or trigger false alarms.

The team therefore built the campaign around worst-case loading, deliberately exceeding the particulate concentrations typical of ISO 8573-1 Class 1 or Class 2 air. By pushing the sensor beyond its intended operating envelope, the researchers could map where measurement begins to degrade and how the onboard algorithms recover.

Building a realistic test bench

A dedicated rig at a partner laboratory reproduces field conditions. The bench combines a Class 0 oil-free compressor, a precision aerosol generator, and a controllable dust feeder loaded with sub-10-micron particles representative of Australian red soils and silica dusts. Calibrated reference instruments, including a condensation particle counter and a gas chromatograph, sit alongside the DOCA prototype.

Each cycle begins with a clean-air baseline, after which particulate loading is ramped in steps while oil concentration is held constant. The sequence repeats for compressor oil, hydraulic fluid, and synthetic lubricant to check whether spectral fingerprints stay separable under heavy optical noise. Engineers log response times, baseline drift, and recovery intervals throughout.

Signal processing and algorithmic compensation

Raw photodetector signals drift noticeably when particulate concentrations rise above roughly one million particles per cubic metre. The DOCA firmware applies a multi-stage compensation routine that subtracts a slowly varying baseline attributed to bulk scattering, then applies a wavelength-ratio test to distinguish particulate-only events from oil-bearing ones. Trials confirmed this compensation reduces apparent oil readings by more than ninety percent in dust-only streams, letting genuine oil events remain visible above the residual noise.

A second algorithm monitors the temporal pattern of scattering events. Oil aerosols produce smooth, sustained signals, while dust pulses are short and erratic. Pattern recognition, rather than simple amplitude thresholding, proved decisive when particulate loading reached its highest tested values.

Quantifying detection limits under load

Performance was judged against three metrics: minimum detectable oil concentration, response latency, and recovery time after a transient spike. Under loading typical of well-maintained industrial systems, the sensor held its nominal detection limit below 0.1 milligrams per cubic metre across all three oil phases. At the extremes of the test matrix, the limit rose modestly to about 0.3 milligrams per cubic metre, still well within pharmaceutical and medical-air specifications.

Response time stayed under eight seconds for aerosols and below thirty seconds for vapour detection, even at the highest dust concentration tested. Recovery was the parameter most affected by heavy loading, occasionally extending to several minutes when upstream filters became saturated.

Calibration stability across extended exposure

A persistent worry with optical instruments in dusty environments is gradual drift caused by contamination of optical windows. The DOCA sensor incorporates a self-cleaning airflow path and periodic reference-pulse checks, and the particulate trials verified long-term stability under realistic conditions. They ran for hundreds of hours of dust-laden flow at concentrations well above factory calibration points.

Baseline drift remained within specification across the campaign, and only minor adjustments were needed between cycles. The reference-pulse system flagged window fouling early, triggering automated purge cycles before accuracy was compromised. This matters for remote Australian sites where manual recalibration is logistically expensive, such as Pilbara processing plants or east-coast offshore-equipment service hubs.

Relevance to Australian operating environments

Several Australian sectors stand to benefit from validated performance under heavy particulate loading. Western Australia's resources sector relies on instrument-air systems that must tolerate dust ingress during cyclones and filter changes. Pharmaceutical manufacturers in Melbourne and Sydney operate under Therapeutic Goods Administration oversight and depend on compressed air that meets strict purity targets; any drift in a measurement device can trigger costly batch investigations.

Food and beverage producers in regions such as the Barossa Valley and Tasmania's dairy belt need assurance that QA instruments remain trustworthy in plants where flour, sugar, or milk powders add to the background load. Hospital medical-air systems in Brisbane, Perth, and Adelaide form another priority group, with Australian Standard AS 2896 mandating tight limits on oil and particulate contamination in medical breathing air and continuous monitoring becoming standard practice in newly commissioned facilities.

Moving from validation to deployment

The particulate-loading trials have given the consortium confidence to proceed toward pre-commercial demonstration. Next steps include long-duration field trials at industrial sites and integration of the sensor into compressor control loops for automatic filter management. Patent applications covering the compensation algorithms and the optical chamber geometry are progressing in parallel, protecting the validated performance as the technology approaches market entry.

If your facility runs compressed-air systems in dusty Australian conditions and you would like to discuss pilot deployments of the DOCA optical sensor, the project consortium can be reached through the contact page to arrange a conversation.