Editorial reference
VSD vs Fixed-Speed Rotary Screw Compressors: When Each Wins
The short answer: A variable-speed drive (VSD) rotary screw compressor wins on duty profiles where average load is below roughly 70 percent of nameplate capacity. A fixed-speed compressor with a smart unload control wins above roughly 85 percent. Between those bands, the decision turns on the VSD price premium (typically 20-30 percent), the cost of an inverter rebuild over the machine’s life, and how well the fixed-speed package manages unload running. Read the duty profile, not the nameplate, before signing.
How each control type actually works
Fixed-speed (load/unload, modulation, or variable-displacement)
A fixed-speed rotary screw runs the airend at a constant rotational speed driven by a constant-speed AC motor. When system pressure reaches the cut-out setpoint, the inlet valve closes and the compressor enters unload: the airend continues to spin but no longer compresses air. Modern packages use blowdown valves to depressurize the sump, after which the unloaded motor draws roughly 25-40 percent of full-load power to spin the empty airend and fan. When pressure falls to the cut-in setpoint, the inlet valve reopens and full-load operation resumes.
This is simple, robust, and proven over decades. The economic weakness is unload running. On a duty cycle where average demand is 50 percent of compressor capacity, a fixed-speed unit spends roughly half its time unloaded, and that idle time is still drawing significant power without producing air.
Variable-speed drive
A VSD package replaces the across-the-line motor starter with an inverter that varies motor frequency and voltage to spin the airend faster or slower as system demand changes. At 60 percent demand, the airend turns at roughly 60 percent of rated speed, the motor draws roughly 60 percent of full-load amps, and the compressor delivers 60 percent of its rated SCFM at the setpoint pressure without ever cycling.
The thermodynamic gain is twofold. First, no time is spent unloaded at 25-40 percent idle power. Second, slower airend speed means lower slip and lower discharge temperature, both of which marginally improve specific power. The cost is the inverter (typically 20-30 percent of package price), the EMC filtering hardware, and a more complex failure mode profile.
The duty-cycle break-even
The break-even between fixed and variable speed depends on average load over a representative week, not peak load. The chart below summarizes the typical result for a 100 HP class compressor at $0.12 per kWh, 6,000 hours per year, comparing a 30 percent VSD price premium against the savings from eliminating unload running.
| Average load | Fixed-speed effective specific power | VSD effective specific power | Annual energy delta | Payback on premium |
|---|---|---|---|---|
| 40 percent | ~31 kW/100 CFM | ~18 kW/100 CFM | $11,000 / yr | 1.5-2 years |
| 60 percent | ~25 kW/100 CFM | ~18.5 kW/100 CFM | $5,200 / yr | 3-4 years |
| 75 percent | ~21 kW/100 CFM | ~19 kW/100 CFM | $1,800 / yr | 8-12 years |
| 90 percent | ~18.5 kW/100 CFM | ~19 kW/100 CFM | -$400 / yr (VSD slightly worse) | Never |
At 90 percent average load and above, the VSD’s inverter losses (typically 2-4 percent of motor input) outweigh any savings from variable-speed operation, because the compressor is essentially running flat-out the whole shift. A well-controlled fixed-speed package at that duty has very little unload time to begin with, so there is no idle penalty for the VSD to eliminate.
The cost premium and what it buys
A 50 HP VSD package typically lists 20-30 percent above an equivalent fixed-speed unit from the same manufacturer. The premium covers the inverter, line and motor reactors, EMC filtering, the VSD-rated motor (usually inverter-duty insulation, sealed bearings, and an external blower for low-speed cooling), and the more elaborate controller. For a 100 HP unit the percentage premium is similar, although the absolute number is larger.
Maintenance differs as well. Fixed-speed units have one principal mechanical wear point (the airend) plus the motor and the inlet valve assembly. VSD units add the inverter (capacitor life, IGBT/SiC module life, fan life on the drive heatsink), the external blower on a VSD-rated motor, and somewhat more demanding power-quality requirements (line-side harmonics, neutral current). Field service technicians report inverter rebuilds or replacements at roughly 8-12 years in clean industrial service, sooner in dusty or high-humidity environments.
When fixed-speed wins (and why operators over-spec VSDs)
The market has shifted aggressively toward VSD packages over the past decade, and not every plant should follow. Specifically, fixed-speed with a smart unload controller wins economically when:
- Average load over a representative shift is above 85 percent. Examples: 24/7 process plants with constant demand, continuous-bake bottling lines, pneumatic conveying for a continuous-process feedstock.
- The plant has multiple compressors in a sequenced base-trim configuration. The base unit runs flat-out and a smaller trim unit follows the variation; the base unit has no reason to be VSD because it never modulates.
- Power quality is poor enough that the harmonic distortion from large inverters violates IEEE 519 limits without expensive mitigation.
- The maintenance team has no inverter-trained technicians and the nearest service depot is far away. A fixed-speed motor failure is roughly an order of magnitude cheaper to diagnose and repair in the field than an inverter failure.
The most expensive mistake here is putting a VSD on the base machine in a base-trim system. The base never modulates, the VSD adds 2-4 percent inverter loss against flat operation, and the premium never repays. The base should be fixed-speed; the trim should be VSD if demand varies.
When VSD wins
- Single-compressor plants with workday-only operation (16/5 schedule with morning ramp and afternoon ramp-down).
- Plants with seasonal demand swings or multi-product lines where the demand profile changes week to week.
- Light-industrial shops where the compressor sits at 30-50 percent average load and the operator would otherwise tolerate frequent cycling on a fixed-speed unit.
- New installations where the buyer can specify and commission the VSD-rated motor and EMC filtering as a package, rather than retrofitting.
Real-world examples from the catalog
The catalog spans both control types across overlapping HP bands.
- Quincy QGS-15: 15 HP fixed-speed at 55.9 SCFM / 125 PSI. The QGS line is Quincy’s fixed-speed workhorse for shop-floor service where load is steady.
- Atlas Copco GA 75 VSD+: a 75 HP VSD package designed for facilities with strongly varying air demand. The VSD+ controller and the iPM (interior permanent magnet) motor are Atlas Copco’s higher-efficiency drivetrain.
- Ingersoll Rand R110i: a 100-125 HP class screw available in both fixed-speed and Nirvana VSD variants. The fixed-speed configuration is favored on base duty in multi-machine plants; the Nirvana variant lives as a trim machine.
- Kaeser CSD 75: a 100 HP screw at 461 SCFM / 110 PSI in fixed-speed configuration. Kaeser’s Sigma Control 2 includes built-in dual-mode (load/unload and modulation) logic that narrows the unload-running penalty on duty cycles in the 60-85 percent band.
Sequencing trumps single-machine control choice
A plant with two 50 HP fixed-speed units and a Sigma-style sequencing controller will often outperform a single 100 HP VSD on the same demand profile, because sequencing rotates which compressor is base and which is trim, both units stay in their high-efficiency operating bands, and a single failure does not stop production. For more on how to think about staging and sequencing, see the rotary screw archive and the foundational comparison rotary screw vs reciprocating air compressor.
How to read your duty profile before buying
The single most consequential pre-purchase data collection is a one-week run-time log from the existing compressor. The numbers that matter:
- Loaded hours / total run-time hours. Most modern controllers (Atlas Copco Elektronikon, Kaeser Sigma Control 2, Ingersoll Rand Xe-145M, Quincy ICONS) log this natively. A ratio below 0.70 says VSD will likely pay; above 0.85 says fixed-speed.
- Cycle count. Number of load-unload transitions per hour. Excessive cycling on a fixed-speed unit (more than 6 per hour at the airend) indicates undersized storage or oversized capacity; either condition argues for either more receiver volume or VSD.
- Peak demand vs average demand. If peak demand is 1.8× average for sustained minutes, a single-VSD strategy will turn down at average load but still need to support peak. Multi-machine base-trim or substantially larger receiver capacity helps.
- Discharge pressure during load. If the line pressure during the loaded phase consistently exceeds the cut-out by 2-3 PSI, the controller is hunting and the deadband should be widened before any control-strategy decision is made.
For the volumetric definitions behind these readings, see SCFM, and for the duty-cycle terminology, see duty cycle in the glossary.
Operating-cost worked example: 100 HP class, 60 percent average load
Worked dollars on a 100 HP class compressor, 6,000 run-time hours per year, $0.12 per kWh blended rate including demand, 60 percent average demand:
| Line item | Fixed-speed | VSD |
|---|---|---|
| Full-load specific power | 18 kW / 100 SCFM | 18.5 kW / 100 SCFM (+inverter loss) |
| Effective specific power at 60 percent load | ~25 kW / 100 SCFM (unload penalty) | ~18.5 kW / 100 SCFM (no unload) |
| Annual energy at 285 SCFM delivered (avg) | 427,500 kWh | 316,400 kWh |
| Annual energy cost | $51,300 | $37,970 |
| Annual savings (VSD over fixed) | $13,330 | |
| VSD premium (purchase) | ~$14,000 – $20,000 | |
| Simple payback | 1.1 – 1.5 years |
That payback collapses fast at high duty (under 1 year above 6,000 hours at 50 percent average), and stretches out (above 5 years) when the average load climbs above 80 percent or run-time falls below 3,000 hours per year. Use your own utility rate, your own price quotes, and your own duty data.
Common questions
What is the typical specific power for a modern 100 HP rotary screw?
At 100 PSI and 100 percent load, modern oil-flooded rotary screws come in around 17-20 kW per 100 SCFM measured at the package terminals, with single-stage units in the upper end of that range and two-stage units in the lower end. Specific power is a CAGI data-sheet metric; insist on it during procurement.
Does a VSD let me run at lower pressure to save energy?
Lowering plant pressure saves roughly 0.5 percent of compressor energy per PSI of reduction, regardless of control type. A VSD makes it easier to operate at a tight pressure band around the demand floor, but the savings come from the pressure reduction itself, not from the inverter.
What is the lower turndown limit of a typical VSD package?
Most industrial VSD rotary screws hold continuous operation down to roughly 20-25 percent of rated capacity. Below that, the airend’s internal leakage and the motor cooling requirements stop the savings curve from sloping further. Demand below the turndown limit either cycles the machine or requires a smaller trim unit.
Are permanent-magnet motors worth the premium over induction in a VSD package?
For applications that spend significant time below 50 percent load, yes. The iPM rotor avoids slip losses that an induction motor pays at every load point, and the differential is most visible at 30-60 percent. Above 75 percent load the gap closes. Permanent-magnet motors also tolerate the wide speed range of a screw compressor better than induction at very low speed.
If you want to dig further on the controls side
Where the worked example above points to fixed-speed with smart unload controls as the right answer, there are two ways to get those controls. The aftermarket path is a purpose-built unloader controller — a small device that sits between the motor contactor and the unloading valve and drops the motor sooner when demand stops. We cover one implementation in a sister project. The DIY path is a small PID or threshold loop on a PLC or microcontroller, watching discharge pressure or motor current and intervening on the unload contactor directly. Both routes work; the hardware path is faster to deploy and arrives pre-tuned, the DIY path is cheaper if you have controls engineers in-house and can absorb the tuning cycle.
Compressor ControllerAsk the editorial team
Have a question this guide didn’t cover?
Send it here. Questions go directly to the editorial team — they’re reviewed privately and folded into a future revision of this guide. Leave an email if you’d like a personal reply.