What a Drop of Oil Can Tell You
Oil analysis is the blood test of the machinery world. A 4-ounce sample drawn from a gearbox or hydraulic system carries information about lubricant condition, contamination levels, and active wear patterns inside the machine. Done properly, it’s one of the most cost-effective predictive maintenance technologies available. A single oil sample costs $25-50 to analyze. The information it provides can prevent failures costing tens or hundreds of thousands of dollars.
The challenge isn’t the technology. It’s getting consistent, representative samples and knowing what to do with the results.
Three Categories of Oil Analysis Tests
Every oil analysis test falls into one of three categories. Understanding these categories helps you select the right test slate for each application.
Fluid Condition Tests
These tell you about the lubricant itself — is it still doing its job?
- Viscosity — The single most important lubricant property. A 10% change from the new oil baseline warrants investigation. A 20% change means the oil needs replacement regardless of service interval. Test at 40°C per ASTM D445.
- Total Acid Number (TAN) — Measures acidic degradation products. Rising TAN indicates oxidation. For mineral oils, consider changeout when TAN increases by 0.5-1.0 mg KOH/g over new oil values. Synthetic oils tolerate higher TAN values before performance degrades.
- Total Base Number (TBN) — Relevant for engine oils. Measures remaining alkalinity reserve. When TBN drops to 50% of new oil value or reaches 50% of the TAN value, it’s time for a change.
- Oxidation — Measured by FTIR spectroscopy in absorbance/cm. Track trending rather than absolute values since baselines vary by oil type.
- Water content — Karl Fischer testing per ASTM D6304 for precise measurement. For most industrial oils, keep water below 200 ppm. Turbine oils should stay below 100 ppm. Hydraulic systems with servo valves need less than 100 ppm.
Contamination Tests
These identify what’s getting into the oil that shouldn’t be there.
- Particle count — ISO 4406 cleanliness code is the standard reporting format. A code of 18/16/13 means different things for a splash-lubricated gearbox versus a hydraulic system running servo valves. Know your target cleanliness levels for each application.
- Moisture — Already mentioned under fluid condition, but moisture is both a contaminant and a degradation driver. It accelerates oxidation, promotes corrosion, and reduces film strength.
- Silicon — Elevated silicon usually means dirt ingestion. Check breathers, seals, and fill practices. Some silicone-based defoamants can also contribute to silicon readings.
- Fuel dilution — Relevant for engines. Gas chromatography per ASTM D3524. More than 2-3% fuel dilution reduces viscosity enough to cause accelerated wear.
Wear Debris Analysis
These tell you what’s wearing inside the machine and how fast.
- Elemental spectroscopy (ICP/OES) — Identifies wear metals in parts per million. Iron, copper, lead, tin, aluminum, chromium — each metal maps to specific components. Iron in a gearbox points to gear teeth and bearings. Copper in a hydraulic system suggests pump wear or cooler tube erosion.
- Analytical ferrography — Separates wear particles magnetically and examines them under a microscope. Reveals particle size, shape, color, and composition. The most informative wear test available but also the most expensive and time-consuming.
- Particle quantifier (PQ) index — Measures total ferrous debris including large particles that ICP spectroscopy misses. ICP becomes less effective for particles above 8-10 microns. PQ catches what ICP misses.
Getting a Good Sample: The Most Critical Step
Bad samples produce bad data. Bad data produces bad decisions. This is where most oil analysis programs fall apart.
Sampling Location
Always sample from a live, turbulent zone — downstream of components, upstream of filters. A sample from the sump bottom collects settled debris that doesn’t represent what’s circulating through the machine. A sample from downstream of a filter shows you clean oil, not system condition.
Install dedicated sample valves wherever possible. Minimally tapped ports with ball valves and pitot tubes work well. The Wiggins-style quick-connect sampling port has become an industry standard for good reason — it’s clean, fast, and repeatable.
Sampling Procedure
- Flush the sample valve with at least 5-10x the dead volume of the port and line before collecting.
- Use clean sample bottles provided by your lab. Don’t reuse bottles.
- Fill the bottle to the indicated line — not more, not less. Labs calibrate their instruments for specific volumes.
- Cap immediately after filling to prevent airborne contamination.
- Label completely. Unit ID, sample point, date, oil type, hours on oil, hours on machine. Missing data means the lab can’t give you meaningful comparisons.
- Ship promptly. Samples sitting on a shelf for weeks degrade, especially if the environment is warm.
Sampling Frequency
Monthly sampling works for most critical systems. Quarterly is adequate for non-critical applications with stable histories. Increase frequency when you see abnormal trends or after maintenance events that involved opening the system.
Interpreting Results: Context Matters More Than Numbers
A wear metals result of 45 ppm iron is meaningless without context. What’s the machine? What’s the oil sump volume? How many hours on this oil charge? What’s the historical trend?
Small sumps concentrate wear metals faster. A 5-gallon gearbox will show higher ppm values than a 500-gallon turbine oil reservoir at the same wear rate because the debris is diluted into a smaller volume. Rate-of-change calculations (ppm per hour or ppm per 1000 miles) normalize this.
Trending is your most powerful tool. Plot each parameter over time. Look for slope changes, step changes, and cyclical patterns. A gradual upward trend in iron over 12 months is normal wear. A sudden doubling of iron over one sample interval is an active failure in progress.
Application-Specific Guidelines
Gearboxes
Focus on iron (gears, bearings), particle count, and water. Industrial gear oils should maintain ISO cleanliness of 19/17/14 or better. Watch for copper if the gearbox has bronze thrust washers. Phosphorus and zinc depletion indicates additive consumption — the oil is working hard to protect the surfaces.
Hydraulic Systems
Cleanliness is paramount. Target ISO 16/14/11 for proportional valve systems. Servo valve systems need 15/13/10 or cleaner. Water is the enemy — it causes valve spool corrosion and erratic operation at surprisingly low concentrations. Track copper for pump wear (vane and piston pumps have copper alloy components).
Compressors
Rotary screw compressors ingest atmospheric contaminants. Expect higher silicon levels than enclosed systems. Watch TAN closely — compressor oils oxidize faster due to heat and air exposure. Synthetic compressor oils (PAO or PAG) handle the thermal stress better but cost more upfront.
Building Your Oil Analysis Program
Start by identifying which machines justify oil analysis. Anything with an oil sump over 2 gallons and a replacement cost or downtime impact that matters is a candidate. Calculate the cost per sample versus the cost of failure. The math almost always favors monitoring.
Select a reputable lab. Polaris, Eurofins TestOil, and Bureau Veritas are among the established players. Ask about turnaround time — results that take three weeks to come back aren’t helping you catch fast-moving failures. Look for labs offering 24-48 hour routine turnaround with emergency same-day capability.
Train your samplers. A 30-minute hands-on session covering proper procedure and common mistakes pays for itself immediately. Auditing sample quality for the first few months catches bad habits early.
The payoff from a disciplined oil analysis program compounds over time. As your historical database grows, trending becomes more powerful, false alarms decrease, and your confidence in maintenance decisions improves measurably.