Oil sampling

Oil sampling is a workflow for wind gearboxes, hydraulic and yaw systems, and for transformers on any site. Aevy reads the laboratory report, files each result against the unit and the sample point it was drawn from, and raises the anomalies the lab flagged as work. The point is the trend: one report is a number, a run of them is a warning.

When to switch this on

When lab reports arrive as PDFs that get filed and never compared, so a wear metal climbing over four sampling rounds is only noticed on the round that fails. Ask your Aevy contact to switch it on and name the assets.

From a lab report to a trend

The dashed return is the whole value. A single iron reading means little; the same sample point rising across three rounds is a gearbox telling you something.

Aevy keeps what the lab actually measured rather than its summary line: the oil type and sample point, the sample and report dates, the laboratory and its accreditation, the physical chemistry (viscosity, water content, TAN and TBN, MPC, RPVOT), the particle counts to ISO 4406 or NAS 1638, and the wear metals in ppm, iron, copper, chromium, lead, tin, aluminium, silicon and the rest.

Set it up

Step	What happens	Who
1 Systems	The turbines or transformers exist as records, so a sample lands on the unit it was drawn from.	Aevy
2 Documents	The lab reports reach Aevy, by sync from your source. Include the older rounds: history is what makes a trend.	You
3 Agree the bar	Which anomalies are worth work, and whose limits apply. See below.	Together
4 Switch on	Aevy turns the workflow on for the assets you named and reads the rounds already on file.	Aevy
5 Review	Proposed anomalies wait under Suggestions, each naming the sample and the figure behind it.	You

Decide before it runs

Decision	What it changes	Where portfolios land
The bar	Which flags become work. A lab marks plenty of results as worth noting, and treating each as a job means somebody stops reading them.	Water over threshold, particle counts over the ISO limit, and a wear metal that has moved rather than one that is merely present. Additive depletion and oxidation noted but not raised.
Whose limits	Whether the lab’s own thresholds decide, or the OEM’s, where the two disagree.	The OEM’s where the gearbox is under warranty, since that is the number that will be argued about.
The start date	How far back Aevy reads. Here the answer leans harder toward everything than in other workflows.	Everything on file. A trend needs history, and a start date set at this year throws away the comparison that makes the workflow worth running.

Read the results

Anomalies are findings, so they sit alongside the rest of the open work on the asset. Group by system to see which unit is producing them, and open one to reach the report and the figure it was read from.

The reading that matters is per sample point over time. A single sample flagged for iron is worth a note. The same point flagged on consecutive rounds, with the number climbing, is the case for an endoscopy or an oil change, and the findings on that point are the evidence you make it with.

Because every figure is held as a value against a unit rather than left inside a PDF, you can ask for the trend instead of assembling it. Squirrel reads the same structured results the table does, so “show me iron across every gearbox sample on T3, oldest first” is a question rather than an afternoon with a spreadsheet. It is the fastest way to test whether one flagged sample is the start of something.

Limits

Aevy reports the lab’s figures and the lab’s flags. It does not re-interpret the chemistry, so a result the lab passed is not raised because the trend looks unhappy, and the judgment about what a rising metal means stays with your engineers.

Sample points have to be named in the report clearly enough to match a unit. A report that says only “gearbox oil” across a whole site gives you the site, not the turbine.

What’s next

• How workflows work: the set, and what each needs before it runs.
• Findings: the record each anomaly becomes.
• Preventive maintenance: the service visits that usually draw the samples.
• Systems and components: the units a sample point belongs to.