The should-cost standard for OPEX.
Every operating-cost line has a theoretical floor. We compute it, publish the reference, and defend it, starting with compression.
Peers tell you your rank. The floor tells you the truth.
Benchmarking against peers tells you where you rank inside a distribution of other people's spending. It cannot tell you what the line should cost, because everyone in the sample may be overpaying together. Should-cost starts from the other end: the vendor's own economics, equipment cost, target return, operating cost, built up into the theoretical bottom of the line. It is the same arithmetic the vendor's CFO runs to price you; we just run it in reverse.
The gap between that floor and what you actually pay is recoverable money, and it is a number nobody sells. RSMeans became the reference that construction contracts are written against by publishing exactly this kind of floor. That is the position we are building for oil and gas operating cost, one line at a time.
Two denominators, kept distinct
The word denominator does two jobs in this business, and the method depends on never letting them blur.
The market denominator
Installed compression horsepower per basin. How much iron is out there, who owns it, who contracts it.
A market total.
The production denominator
BOE, the bottom of the LOE per BOE ratio. Compression drives runtime, runtime drives production, and production divides your costs.
The bottom half of a cost ratio.
Whenever we publish either, we say which one you are looking at.
The one line with a hand on both ends of the fraction
Compression alone is too narrow to be a company, and all of OPEX is too broad to be credible from a standing start. Compression earns the lead because it is the one operating line that pulls both levers of LOE per BOE at once. It is a large cost line, so the numerator moves when you fix it. And it is the biggest driver of runtime and uptime, so production moves too: better compression performance grows the denominator even at flat cost. Chemicals touch the numerator. Water touches the numerator. Compression is the only line with a hand on both ends of the fraction, which is why the standard starts here.
How the census is built
There is no auditable total for the whole installed base, and small wellhead units and VRUs below 200 HP may never enter the public record at all, so a census that claimed to include everything would be guessing. A precisely scoped number can be checked. In Appalachia, where disclosure runs thickest, we check it: a public midstream company at 182,800 HP itemized by system, another at 128,286 HP across 24 stations, a public operator above 800,000 HP, a compression provider around 850,000 HP in the basin. Landing inside a defensible band with every delta explained is what proof-of-methodology means here.
Every mature industry has its reference. Ours never did.
RSMeans
A published cost reference so standard that bids are written against it, with a services layer on top (Gordian) that captures the savings the benchmark reveals.
The contributory model
Every carrier's data sharpens a benchmark every carrier buys.
The census + the method
Never had either one. The census plus the should-cost method is that reference, built compression-first, with the contributory piece (Ground Truth) already running.
The census is the method running at full automation.
The OPEX Desk is the same method with people attached: a retained seat that defends any operating line against its floor, builds custom spend views, and fuses our data with yours. If the census is the reference, the Desk is the crew that argues your side of it.
Talk to The OPEX Desk