A rotary screw compressor comes in three flavors: oil-injected (the standard workhorse, what's running in maybe 90% of compressor rooms I've ever stepped into), oil-free (Class 0 certified, much more expensive), and water-injected (a niche middle ground from Atlas Copco and a few others).

The decision sounds technical but it really isn't. It comes down to a single question: does compressed air make direct contact with your product? If yes, you need oil-free at the point of contact. If no, oil-injected with proper filtration is the right call. Most plants spend tens of thousands extra on oil-free machines they didn't actually need, because vendors push oil-free as the default and the marketing is one-sided.

This page covers what each technology actually does, when each is the right call, and what the cost difference looks like over a 10-year life.

The three main options

Oil-injected (the standard workhorse)

The standard. Oil is injected into the screw element during compression for cooling and sealing, then separated back out before the air leaves the machine. Tiny traces of oil carry through (typically 2-5 mg/m³), which are removed downstream by a coalescing filter and an activated carbon filter if you want very clean air.

Why it dominates the market:

  • Cheaper to buy (often 30-50% less than the equivalent oil-free machine)
  • More energy efficient (better SFC, which adds up enormously over 10 years)
  • Simpler service, lower maintenance cost
  • Reliable, proven design, mature service ecosystem and parts availability

Standard examples: Atlas Copco GA, Ingersoll Rand R-Series, Kaeser SK / SX / CSDX, Almig, Chicago Pneumatic CPC. Every reputable industrial brand has an oil-injected line as their core product.

Oil-free (Class 0)

No oil in the compression chamber. Two tight-tolerance precision-machined rotors run dry, separated by a synchronizing gear so they never touch each other. The air leaving the machine carries zero oil, certified to ISO 8573-1 Class 0.

Costs much more to buy, costs more to maintain (those precision rotors don't tolerate neglect), and uses more energy than the oil-injected equivalent. The rotors run hot without oil to cool them, so the machine needs bigger intercoolers and aftercoolers working harder. A 75 kW oil-free screw runs €60,000 to €80,000. The oil-injected equivalent at the same kW runs €30,000 to €40,000.

Examples: Atlas Copco ZR / ZT, Ingersoll Rand Sierra, Kaeser CSG / DSG.

When you actually need it: a small list, covered below.

Water-injected (Atlas Copco AQ and similar)

A middle ground that vendors don't push much. Water instead of oil gets injected into the compression chamber for cooling and sealing. You get oil-free air without the cost and complexity of a true dry-screw machine. Atlas Copco's AQ series is the best-known, with a few smaller competitors.

Niche, but worth knowing it exists. Where it fits: you need certified oil-free air, the duty is continuous, and you'd rather not deal with the maintenance overhead of a dry-screw. Where it doesn't fit: small installations (AQ is mostly large-frame), or anywhere water handling on site is awkward.

When you NEED oil-free

Strip away the marketing and there's exactly one question that determines this:

Does compressed air make direct contact with the product, the inside of product packaging, or anything that will subsequently contact the product?

If yes, you need oil-free at the point of contact. Class 0 certified, no exceptions.

If no, standard plant air with good filtration is fine.

Examples of "yes":

  • Bottle-blowing air going into an empty bottle that's about to be filled with food, beverage, or pharma liquid
  • Conveying air directly contacting open product (flour, sugar, granules)
  • Mixing air bubbled through a liquid product
  • Bag inflation around food before sealing
  • Sterile pharmaceutical fill, breathing air, tablet coating
  • High-end semiconductor manufacturing
  • Specific food categories where customer audits require certified Class 0 regardless of risk analysis (some baby food, some dairy, some high-end olive oil bottling)

Examples of "no" (where oil-injected with filtration is fine):

  • Pneumatic actuators (cylinders, valves, grippers)
  • Conveyor drives and stoppers
  • Packaging equipment operating SEALED packaging
  • Plant air for cleaning floors and equipment surfaces
  • Pneumatic transport of dry materials in fully sealed conduits

That's the decision. Most factory air, even in food and pharma plants, falls in the "no" category. Often under 10% of total plant air actually contacts product.

When oil-injected + filtration is fine

Most of the time. Including in industries where vendors will tell you it isn't.

I'm going to take an unpopular position here because it's the right one for most readers: most food production doesn't need an oil-free compressor. Walk into any food plant and you'll see oil-injected machines running. Talk to a vendor selling a new system and they'll insist you need oil-free. Both can't be right, and the marketing one isn't.

What I've actually seen in 20 years of servicing food production plants:

  • Plastic packaging plant making food containers (cups, trays): three oil-injected screws plus a fourth added later, all oil-injected, no oil-free anywhere on site. Running 24/7. Air drove molding, robots, conveyors, packaging. Ran clean for years.
  • Cheese maker (Belgian smeerkaas, soft melted cheese in tubs): two small oil-injected piston compressors with food-grade compressor oil. That was the entire setup.
  • Candy factory: two big oil-injected screws around 100 kW each, standard food-grade lubricant. Decades of production, no contamination issues that traced back to compressed air.
  • Pharmaceutical plant (the smart install): an oil-injected screw for the bulk of plant air, plus a small scroll compressor for the air that actually contacted product. Two completely separate systems, two completely different price points.

The pattern is unmistakable. Most food production uses oil-injected. Some uses food-grade oil. The smart pharma installs are mixed.

The food-grade oil middle ground

A category vendors don't talk about much: oil-injected compressors loaded with food-grade lubricant (typically NSF H1 certified for incidental food contact) instead of standard mineral oil. The oil costs 2-3x standard, but the compressor itself is the standard oil-injected machine at the standard oil-injected price.

When this makes sense:

  • Food handling with a small, statistically rare possibility of trace air-to-product contact
  • Plants where regulators or customer audits ask "what oil is in your compressor?" and "food-grade NSF H1" is an acceptable answer
  • Belt-and-braces, when you're nominally in "no" territory but want to be safe

Bakeries, dairies, breweries (for packaging air), meat processors, cheese makers. A quiet middle ground that doesn't show up in vendor brochures because the oil-free sale is higher margin.

The smart mixed installation

The pharmaceutical example above had it right. Most of the plant ran on standard oil-injected screw air. The small portion that contacted product directly was fed by a small scroll compressor sized exactly for that need. Two separate distribution systems, no cross-connection.

This is the move for any food or pharma plant with some product-contact air needs:

  1. Size your bulk plant air with a standard oil-injected screw (or two for redundancy)
  2. Size your product-contact air separately. Often small enough to be a scroll compressor
  3. Run them on separate distribution piping, no cross-connection
  4. Validate the product-contact air system properly. Treat the rest as plant air

The savings versus running everything as oil-free can be tens of thousands on day one, and continue every year of the install's life.

The cost difference: purchase, energy, maintenance

Rough numbers for a 75 kW machine, European prices, 2026:

Item Oil-injected Oil-free Difference
Purchase price €30,000 to €40,000 €60,000 to €80,000 about 2x
Annual energy (€0.20/kWh, 6,000 hrs/yr) ~€90,000 ~€100,000 +10% (oil-free runs less efficient)
Annual maintenance €1,500 to €2,500 €3,500 to €6,000 2 to 3x
10-year TCO ~€940,000 ~€1,080,000 about €140,000 more for oil-free

The purchase-price difference (around €40,000 on a 75 kW machine) looks dramatic. The 10-year TCO difference (around €140,000) is where the real money is, because oil-free machines run less efficient AND cost more to maintain over their whole life.

If you genuinely need oil-free, this cost is worth paying because the alternative is product contamination risk. If you don't actually need it (which is most installs), it's a six-figure mistake.

Where to next