What Are the Disadvantages of a Screw Compressor? – Screw Air Compressors Manufacturer

Screw compressors hold about 80% of the industrial compressed air market between 30 and 300 kW. CAGI data, PNEUROP surveys, facility walk-throughs across North America and Western Europe all point to the same figure. The technology earned it.

Part-Load Energy Waste

The DOE's Compressed Air Challenge program has been running since 1998. Over a thousand plant assessments. The headline number from their sourcebook, "Improving Compressed Air System Performance," has not moved in twenty years: systems waste 25% to 50% of input energy. Two full decades of awareness campaigns, utility rebates, VSD rollouts, audit programs, and the waste band is the same width it was when Clinton was in office. Screw compressors at part load are why.

Rotors sweep a fixed volume every revolution. No cylinder unloading, no valve lift adjustment, no clearance pocket. The displacement is what it is. When demand drops, a fixed-speed screw compressor can throttle the inlet or unload. That is the entire menu.

Load/unload deserves the most attention here because the installed base is overwhelmingly fixed-speed machines on load/unload, and will be for years, maybe decades, regardless of how many VSDs get sold. CAGI sheets now publish unloaded power. On a 110 kW Kaeser CSD, the unloaded draw is around 30 kW. Atlas Copco GA series, similar. Ingersoll Rand R-series, similar. Thirty kilowatts to spin the rotors at full speed against a closed inlet and push oil through the circuit and run the fan. Zero CFM out the other end.

When demand sits around 45% of rated capacity, and the receiver is the 500-liter tank that came with the package, and the pressure band is 0.5 bar between load and unload setpoints, the machine short-cycles. Twelve to twenty seconds loaded, similar unloaded. Average power: ~70 kW. Useful air output: ~50 kW equivalent. That 20 kW gap, sustained 6,000 hours at $0.10/kWh, costs $12,000 per year. Per machine.

Two smaller compressors on a sequencer, one base-load running full, one trim running full or off, would capture most of that $12,000. A sequencing controller costs $3,000 to $5,000. The payback is inside a year. Distributors do not push this solution because it doubles the quoting effort, doubles the commissioning, and introduces controls integration that may sit outside their comfort zone. So the oversized single box gets sold again.

VSD

VSD technology deserves its own section separate from the general part-load discussion because the gap between VSD marketing and VSD field performance has become a real problem in equipment selection.

Kaeser and Atlas Copco both shifted to permanent magnet motors in their VSD compressor lines. This was a legitimate engineering improvement. PM motors hold efficiency at reduced speed substantially better than induction motors. In the 50% to 80% speed range, the improvement is measurable and meaningful.

Below 40% speed, the picture changes. Three effects stack up. Internal leakage through rotor clearances is a volume-per-revolution phenomenon independent of speed. At full speed, slip might be 3% of displacement. At 30% speed, the same absolute leakage volume is 8% to 10% of the reduced displacement. The VSD's IGBT switching losses take another 3% to 5% of throughput. Motor efficiency drops, even with PM designs. The specific energy curve bends upward. The marketing brochure's straight line from zero to full load does not exist in thermodynamics.

Compressed Air Challenge assessments have documented case after case where VSD compressors were selected for peak demand on systems that spend most of their hours at 35% to 45% of that peak. The VSD saved energy versus load/unload at the same demand point. A staged pair of fixed-speed machines would have saved more. The VSD got sold because it was one machine, one PO, one hookup. Staging was two of everything. This keeps happening.

It is worth being blunt: VSD is oversold in this industry. The technology is sound. The sizing and application are frequently wrong. The marketing charts are misleading in the lower third of the speed range, and the people making the purchase decisions are often relying on those charts because they do not have access to independent field performance data at low load points.

OEM Lock-In

Rotor profiles are proprietary. Atlas Copco does not publish rotor geometry. Kaeser trademarks their Sigma Profile and publishes marketing about it extensively without publishing the dimensional specifications that would allow independent manufacture.

The consequence: airend overhaul at 40,000 hours costs 55% to 60% of new-machine price from the OEM. The consistency of this ratio across Atlas Copco, Kaeser, Ingersoll Rand, and Sullair is worth noting without further comment.

Independent rebuild exists. Shops like Total Equipment Company in the U.S. can CMM-map rotor profiles and hit OEM-grade clearances. The quality spectrum among independents is enormous, from CMM-equipped precision operations down to shops working by feel. No certification body. No published standard. No accreditation.

The consumable pricing structure is where the lock-in generates the most cumulative revenue. Separator elements are the critical item. An OEM Kaeser separator for a CSD might cost $400 to $600. Aftermarket elements with identical external dimensions: $120. The difference in coalescing media grade, oleophobic treatment, and drain tube design between a good aftermarket element and a bad one is invisible to the buyer. A channeled element passes 30 ppm of oil while the dP gauge reads normal, because dP measures flow resistance, not filtration quality. There is no MERV-equivalent rating for compressor oil separators. None.

Over ten years at 6,000 hours, the consumable and service spend on a 75 kW box exceeds the original equipment cost. Compressor manufacturers know this. The initial sale opens a decade-long revenue stream.

Oil Vapor

Most online discussions of oil contamination in screw compressors stop at aerosol carryover. Fresh separator element, correct oil charge, 2 to 3 ppm carryover. Fine.

The vapor problem is different and is the reason oil-injected machines are hard-excluded from certain applications regardless of downstream filtration. Oil vapor pressure rises with temperature. Above 95°C bulk oil temperature, vapor-phase oil alone can approach ISO 8573-1 Class 1 limits. Coalescent filters do not remove vapor-phase molecules. Activated carbon adsorbers remove some, with finite and temperature-dependent capacity.

ISO 8573-1 Class 0 applications, increasingly specified in pharma and semiconductor, require agreement between user and supplier on allowable contaminant levels. In practice, Class 0 means oil-free compression. Oil-free screw compressors serve these applications at higher capital cost, lower efficiency, and heavier maintenance. Kaeser's DSG versus their CSD at equivalent capacity shows the premium clearly in both price and service requirements.

Rotor Wear and the Invisible Efficiency Tax

New airends: 0.04 to 0.06 mm tip clearance, depending on manufacturer and model. Those clearances open with hours. Oil-borne particles, even below the 10-micron filtration cutoff, erode rotor surfaces at tip speeds above 50 m/s. The sealing edges go first.

The control system compensates. Pressure holds setpoint. Power stays constant. Output drops. The compressor does not alarm. Nothing changes on the HMI. The efficiency loss, maybe 1% to 2% per year, accumulates to 10% to 15% by 30,000 to 40,000 hours. Detection requires calibrated flow measurement at delivery versus power input. The DOE assessment teams do this. Most plants do not, ever, unless an outside auditor comes in.

Oil-free machines degrade through coating wear instead of particle erosion. PTFE coatings are thin, 0.1 to 0.3 mm. Timing gear backlash increases with wear, causes rotor phasing variation, accelerates coating contact stress. The whole system is coupled. When the coating fails, rotor damage follows within days or weeks.

Heat, Noise, Condensate, Maintenance

These four topics get compressed together because they are well-covered elsewhere online and the marginal value of repeating the same information at length is low.

Heat: 96 kW of thermal load per 100 kW of electrical input, into the compressor room. Inlet temperature rises if ventilation is inadequate. Every 4°C costs about 1.3% of FAD at constant power. The bigger damage is to the lubricant. Oxidation rate doubles per 10°C above rated temperature. Varnish from degraded oil fouls the thermostatic valve, the cooler, and the separator element in a self-reinforcing sequence. A room that was fine during winter commissioning can push a compressor into varnish trouble by midsummer if the louvers are undersized or blocked.

Noise: tonal at the lobe passing frequency, 200 Hz on a four-lobe male rotor at 3,000 RPM. ISO 1996 penalizes tonal content up to 6 dB. A 75 dB(A) screw compressor can be subjectively equivalent to an 81 dB(A) broadband source. Discharge pulsations excite piping resonances at acoustic length multiples. Pulsation dampeners solve this and should have been in the original system design.

Condensate: hazardous waste. Fifty to a hundred liters per day from a 100 kW machine at moderate humidity. Oil-water separation, cartridge replacement, waste oil disposal. Permanent cost of oil-injected operation.

Maintenance: fewer events than a recip, higher cost per event. Airend overhaul takes weeks, requires factory specs or CMM capability, costs half a new machine. A recip valve job takes a millwright a day with hand tools. Screw compressor maintenance is infrequent and expensive. Recip maintenance is frequent and cheap. Calling the screw compressor "low maintenance" counts events and ignores invoices.

— end —