Hydraulic Oil Analysis Explained: How to Read a Report Before It Causes Downtime

Most industrial hydraulic oil analysis packages cover six core areas. Each measures a different aspect of oil or system condition.

• ISO 4406 particle count - reported as three range codes, for example 19/17/14, covering particles at or above 4, 6 and 14 microns. This is what most OEM specifications are written against.

• Water content - usually Karl Fischer titration, reported in ppm. Free water is visible; dissolved water is not. A clear sight glass tells you very little.

• Viscosity at 40°C - drift beyond roughly ±10% of the nominal grade may indicate cross-contamination, oxidation or an incorrect top-up.

• Total Acid Number (TAN) - rises as the oil oxidises. Judged against the new-oil baseline, not a universal figure.

• Wear metals - iron, copper, chromium, aluminium, lead, tin. These come from components, not from the oil.

• Additive elements - falling zinc and phosphorus against baseline means the anti-wear package is being consumed.

Some packages add MPC varnish potential under ASTM D7843, which answers a different question again. We cover what the MPC varnish test measures separately.

Before any of it means anything, the sample has to be valid. Same point, same operating temperature, system running, same method every time. A sample drawn cold from a static reservoir tells you what has settled, not what is circulating  and results built from inconsistent samples describe the sampling, not the oil. If that consistency is hard to hold in-house, it is what a structured oil condition monitoring programme is for.

Here is a real before-and-after from Stadco's presses - the plant losing a £20,000 valve most weeks. The "after" sample was taken six weeks into continuous offline oil filtration.

Four findings from these results demonstrate why the complete report must be considered

The coarse number looked acceptable. At >14 µm the "before" sample was 26 particles per millilitre -  a range code of 12, unremarkable for a press circuit. The problem was at 4 µm, where 2,644 per millilitre puts the system at code 19. A system can be severely contaminated in the size range that destroys clearances while the number most people glance at looks broadly fine

Water was not driving the failures. It moved from 26 ppm to 20 ppm and was not a significant factor in this case. Worth saying, because water is the assumed culprit in a great many hydraulic investigations. The report tells you which problem you actually have, not the one you expected.

MPC was the real story. A varnish potential of 45.9 is severe. Varnish is what was sticking those valves -  soft, sub-micron oxidation product that no particle count captures and no standard inline filter removes. If MPC isn't on your test package, this is the failure mode you cannot see.

Range codes are logarithmic, so improvements compound. Moving 19 to 14 at 4 µm is five codes, each roughly a halving. The measured reduction was 98.8%. This is why chasing one code of improvement is worth more effort than it sounds.

One caveat that matters operationally: the oil was clean in six weeks, but the system was not. Historic deposits in the tank and pipework took approximately twelve months to lift and capture. Clean oil is fast. A clean system is a programme.

👉 Read the full case study, including independent laboratory analysis and the two years that followed without a single valve-related breakdown: Zero valve breakdowns in two years

Delta-Xero designs and manufactures offline fluid conditioning systems in Fareham, and configuration comes from the fluid, not from a product list. The contamination profile in your report determines cartridge selection, cartridge count and unit sizing, alongside fluid type, system volume, operating temperature and contamination level.

Filtration removes particulate, water, varnish and oxidation products, siloxane and salts down to 0.1 micron - well below the sizes ISO 4406 reports against. Because it runs offline in a kidney loop, it conditions the reservoir continuously without interrupting the primary circuit, and without relying on the machine's inline filters, which exist to protect components from catastrophic debris rather than to control contamination.

Sizing follows the asset. The DX1525 Series suits reservoirs up to 1,200 litres - individual CNC machines, presses, smaller power packs - where footprint is the constraint. The DX1000 Series is the modular option, running one to twelve cartridges in parallel for higher-volume or hazardous-area installations. At Stadco, presses ran between one and four cartridges each depending on oil volume and contamination burden, and cartridge intervals settled at eight to twelve months. A similar approach on a hydraulic baler at Biffa applied the same principle to a different duty cycle.

For the maintenance team, the practical outcome is fewer oil changes made on suspicion, contamination maintained at a defined target and clearer evidence for maintenance decisions.

If your latest hydraulic oil analysis shows a rising ISO particle count, water contamination, climbing TAN or repeated filter blockages, send us a sample. We will assess the contamination profile, review the likely root cause, and tell you whether the fluid should be filtered, monitored or replaced.