Industries / Energy, water and waste / Pipelines and midstream / Compression services
Engines and motors
A compressor needs a driver. In the field it is usually a natural gas engine burning some of the gas it moves, and sometimes an electric motor.
5
Milestones achieved, from 1939 to 2024.
1
To watch for: announced, not yet confirmed outside.
1
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1939
A fast engine that failed
From 1931 Clark Bros. built a radial, high-speed engine-compressor, a concept promoted by Howell Cooper of Hope Natural Gas. It cost well over half a million dollars to develop, sold just eight units, and was abandoned in 1939 because "the units were simply too light for continuous operation at high speeds," writes compression historian Norm Shade.
Lessona faster, lighter machine needs materials that can take continuous duty.
Credit
Clark Bros. Company; Howell Cooper
1966
The small separable compressor
In the mid-1960s Jim Buchwald designed a small, high-speed piston compressor and founded Ariel Corporation in 1966 to make it. A separable compressor is a compressor that bolts to a separate engine or motor, so either can be swapped. Ariel calls itself the world's leading maker of separable reciprocating gas compressors; that claim is its own.
Credit
Jim Buchwald, Ariel Corporation
1991
A large lean-burn gas engine
Caterpillar introduced the G3600 gas engine for compression in 1991, a trade magazine reported on its 30th year. Lean-burn means the engine runs with extra air, which cuts nitrogen oxides. The G3608 version is rated at 2,750 horsepower at 1,000 rpm, per a Caterpillar dealer, and is sold for central gathering stations.
Credit
Caterpillar
2013
Catalysts required on engines
EPA amended its air toxics rule for engines so that existing rich-burn engines over 500 horsepower at smaller sites in populated areas must install a three-way catalyst (non-selective catalytic reduction, NSCR), and lean-burn engines an oxidation catalyst. Rich-burn engines run with just enough air to burn the fuel.
Credit
US EPA
2024
A bet on electric drive
Archrock agreed to buy Total Operations and Production Services for $983 million, about 580,000 horsepower of electric-driven compression, which its chief executive Brad Childers said "establishes our leadership position" in that segment. EPA had long noted that electric motors need an uninterrupted power supply.
Credit
Archrock
Watch for
A turn back to gas engines
Kodiak's chief executive Mickey McKee said that "an increasing amount of compression [is] moving away from electric due to access-to-grid challenges," with grid connection in some areas "outside of a decade to maybe never," and lead times over 180 weeks for large gas compression engines. Kodiak's 2026 proxy said about 40% of the roughly 150,000 new horsepower it placed in 2025 was electric.
Achieved whenannual filings from two or more firms show the electric share of new horsepower falling.
Claimed by
Kodiak Gas Services, March and May 2026
Open
01
Engine or motor, with the wait
An open cost model for one compressor site that compares a gas engine with an electric motor over 15 years, counting fuel gas burned, power price, equipment lead time, the wait for a grid connection and the cost of emissions.
The briefCredit · unclaimed
Open 01 · the brief
Engine or motor, with the wait
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
An open cost model for one compressor site that compares a gas engine with an electric motor over 15 years, counting fuel gas burned, power price, equipment lead time, the wait for a grid connection and the cost of emissions.
Why it is still open
Firms make this choice case by case, and Kodiak's chief executive has said the grid wait now decides it in some basins. No public model that includes the wait for connection was found.
Built with what exists
Python or a spreadsheet; US Energy Information Administration gas and industrial power prices; engine fuel rates from published Caterpillar data sheets; grid wait times from utility interconnection queue reports.
First step
Build the 15-year comparison for one 1,380 horsepower unit in the Permian with a 0-, 24- and 60-month grid wait.
Done looks like
The model published with inputs for three basins; for each, the grid wait in months at which gas and electric cost the same, with every input linked to a source.