Skip to content
The archive

The field history
behind the answer.

When the Engine tells you what probably went wrong, that ranking comes from somewhere. It comes from here: a curated library of real engine failure records, built from what techs actually found and actually did about it.

It is not a search index over scanned PDFs, and it is not a pile of forum threads. Every record is parsed into fields, so the system can count across decades of them and tell you how often a cause has genuinely explained a symptom on your make and model.

The order it trusts

Your manual comes first.

This ordering matters more than anything else on this page. A wrong torque or clearance can destroy an engine or hurt someone, so the archive never outranks the document that came with your machine.

First

The manuals you uploaded

Your own OEM manuals, bulletins and shop procedures are the source of record for any figure. If the figure is in there, that is the answer, cited by page.

Alongside

The archive

Our troubleshooting guides and solved field reports for that make and model lead on likely causes. A figure only the archive prints closes with a line telling you to verify it against your OEM manual.

Then

Web research

The Engine also searches published sources, which can back up the first two but never override them. A figure only stands when two independent sources agree on it or the maker’s own document prints it. It is labelled as industry research, with the same verify line.

Otherwise

It says it could not verify it

If the research does not hold up, you get no number. The Engine says it could not find verified sources, then gives you the procedure where it has one and the next best step.

What is in it

Structured, not searched.

A field report is messy by nature. Dozens of techs, decades apart, no two writing it up the same way. Turning that into something you can count on means parsing every one into the same shape.

Manufacturer aliases collapse to one name. Failures map to a shared schema. A write-up from 2003 and one from last month describe the same failure the same way, which is the only reason aggregating across them tells you anything true.

Make and model

Which machine it happened on, normalized so a 2003 write-up and a 2024 one name it the same way.

Component

What actually failed, mapped to a shared schema rather than whatever the tech called it that day.

Operating hour

Where in the machine’s life it went, which is what turns a pile of failures into a pattern.

The repair

What was done about it, in the order it was done.

The tech’s own notes

What they wrote down afterward. Often the most useful part, and the part no data sheet has.

What it covers

Engine data first, and we will tell you where it thins out.

The depth is in the families that dominate the field in northeastern BC and the compression business generally. Outside those, the archive has less to say, so the Engine leans on your manuals and web research, and the answer says when a figure came from research.

Where the records are deepest

CATWaukeshaArielAjaxSolarWhite Superior

Compressor packages, gensets, and the industrial engines that drive them. If you run something outside this and want to know whether we can help, ask us rather than guessing from this page.

Your documents never become part of this.

The manuals you upload stay in your own workspace. They are not added to the archive, not pooled with other customers, and not used to train models. The archive grows from our own curation, not from your uploads.

We also do not hold or redistribute OEM manuals. You bring your own copy and we make it findable inside your workspace, which is a different thing from us handing it to anyone else.

The honest test is a fault you already solved.

Give the Engine a job you already know the answer to and watch what it cites. That tells you more about the archive than any page describing it.

Can't find it? We'll source it

Tell us what you need. We will find it and send you a real quote.

Ask us to find it