Ask a Turkmen drilling supervisor for the word for "blowout preventer" and he'll say "preventer." In Russian. Because that's what's stenciled on the equipment, that's what the manual says, that's what the Russian-speaking foreman shouted during the last kick. The Turkmen term exists in a dictionary somewhere. Nobody on the pad has ever used it.
This is the gap I live in. The field speaks Russian and English. The document — the tender, the HSE policy, the contract addendum — has to be Turkmen, because a ministry or a court or a state company requires it. My job is to build a bridge between a working reality and a paper requirement, and the two don't share a vocabulary.
Where the terms actually come from
Turkmen oil terminology didn't grow organically alongside the industry the way English did. English got "christmas tree" and "mud" and "kill line" from decades of roughnecks naming things on the job. Turkmen got most of its technical layer imported — first through Russian during the Soviet years, then patched with English after independence. So you have three strata sitting on top of each other, and they don't agree.
Take "wellhead." You'll find a coined Turkmen compound in a glossary. You'll find the Russian ustye skvazhiny in every field log. You'll find "wellhead" untranslated in the English engineering package the whole project runs on. All three describe the same steel. When I write the Turkmen document, which one does the reader recognize?
Here's the thing nobody wants to admit: the "correct" Turkmen coinage is sometimes the least useful choice. If the engineer who reads my translation has to stop and mentally back-translate my pure-Turkmen term into the Russian he actually thinks in, I've added friction, not clarity. I've written something technically Turkmen and functionally foreign.
The decision isn't mine to make alone
This is why I stopped treating terminology in oil and gas as a solo call. On a legal document I'll defend a purist choice, because the register demands it and a court reads formally. On a technical procedure that a crew actually uses, I lean toward the term the crew will recognize on sight, even if that means keeping the Russian loan or the English word in parentheses.
But I don't guess which document is which. I ask. Who reads this, and where does it end up? A tender submission going to a state committee gets one treatment. A lockout-tagout procedure going to the wellsite gets another. Same subject, opposite instincts — and the PM who can tell me the destination is worth more than any glossary.
When the client can't tell me, I hedge in the smart way: Turkmen term first, source term in parentheses on first use. guýy agzy (устье скважины / wellhead). It's ugly. It's also the only honest thing to do when you know the reader lives in three languages at once and you don't get to choose which one is loaded in his head that day.
Build the memory or repeat the argument forever
The part that makes this expensive is that every project reinvents the wheel. Agency A's glossary says one thing for "casing." Agency B's TM, built by a translator who's now unreachable, says another. Neither cites a source. Both are locked in. And I'm the guy who has to pick between two anonymous decisions made by strangers, on a deadline.
So I keep my own termbase, and I keep it with provenance. Not just term-plus-target. Term, target, source, date, and a note on register — is this the field word or the paper word? When a PM sends me a reference glossary, the first question I have isn't "is it good," it's "where did this come from and who approved it." A client-approved term I disagree with still wins, because consistency with the client's other documents beats my private preference. But I want it in writing that it was their call. The comma that shuts down a rig has a cousin: the term that gets you excluded from a tender because it didn't match the reference package. Both are avoidable, and both come back on the translator.
One more thing, because it matters for pricing. Agencies quote oil and gas the same per-word as marketing copy and then wonder why the good translators pass. The word count is not the work. The work is the twenty minutes I spend deciding whether skvazhina becomes guýy or stays put, times every recurring term, times the fact that being wrong here isn't a stylistic ding — it's a document that gets rejected or, worse, a procedure a crew misreads. If you're paying flat per-word for this, you're paying for a dictionary lookup. You're getting one.
The field will keep speaking Russian. The state will keep asking for Turkmen. Somebody has to stand in that gap and make a call that a court and a roughneck can both live with. That's not a lookup. That's the job.