A customer called last spring ready to buy a 75 HP variable speed machine. Good shop, three shifts, and he had been told it would cut his air bill by a third. We asked him one question: how many hours a day does your current compressor run fully loaded?
Answer: basically all of them. Two shifts of steady production, a third shift of cleanup, and the machine almost never unloads.
We talked him into a fixed speed. He saved about eighteen thousand dollars on the purchase and he is not losing anything in energy, because a variable speed air compressor only pays you back when your demand actually varies. His does not.
That is the whole decision, and it gets buried under a lot of marketing. This article is about how to tell which side of it you are on before you spend the money.
What a VSD Actually Does
A fixed speed compressor runs its motor at one speed. When you need less air than it makes, it has to do something with the extra capacity, so it unloads, modulates, or stops. All of those waste energy to some degree. The expensive one is load/unload, where the machine keeps spinning while making no air.
That is the number worth remembering: a fixed speed compressor running unloaded still draws roughly 30 to 70 percent of its full load power depending on the design. You are paying most of the bill for none of the air.
A variable speed machine instead slows the motor down to match demand. Flow and power both come down in a roughly linear way, so at half demand you draw something close to half the power. There is no unloaded running state to pay for.
If your demand is flat, there is nothing to match, and the VSD's drive losses, typically a couple of percent, mean it is slightly worse than a fixed speed machine. If your demand swings all day, the savings are real and large.
The Honest Numbers
CAGI, the industry's trade association, puts typical energy savings from a variable speed machine at around 33 percent. An often cited study across 67 installations found average annual savings around $17,200 with payback landing in the two to five year range.
Treat those as what they are: averages across plants that had variable demand, which is the population that buys VSDs in the first place. Your number depends almost entirely on one thing.
| Your load profile | Realistic VSD savings | What to buy |
|---|---|---|
| Flat, loaded nearly all the time | Zero or slightly negative | Fixed speed, every time |
| Mostly steady with a modest swing | 5 to 15 percent | Fixed speed, or a VSD only if you are replacing anyway |
| Swings widely through the shift | 20 to 35 percent | VSD pays, usually in 2 to 4 years |
| Long periods at low demand, short peaks | 30 percent and up | VSD, or fixed speed plus a VSD trim machine |
| Weekends and nights at near zero | Large, but check the alternative | Sometimes a smaller fixed speed plus storage beats both |
Find Your Load Profile Before You Shop
You cannot answer this from memory and you should not answer it from a vendor's spreadsheet. Measure it.
The cheapest version costs nothing. Watch your existing machine for a shift and write down how much of the time it is loaded versus unloaded. Most controllers display loaded hours and total hours. If loaded hours are above about 85 percent of running hours, your demand is flat and you are done, buy fixed speed.
The better version is a data log over a full week, including a weekend, using a flow meter or an amp logger on the compressor. A week catches the shift changes, the lunch dip, the Saturday skeleton crew, and the Monday morning startup surge. Those are exactly the periods a VSD gets paid for, and exactly the periods nobody remembers correctly.
Our pieces on compressed air flow meters and what a compressed air audit is worth both cover how to get this data honestly.
The Turndown Trap
Here is a detail that gets left out of most sales conversations, and it has sunk real projects.
A variable speed compressor cannot slow down indefinitely. Below some minimum speed the airend does not get enough oil flow and cooling, so every machine has a floor, commonly somewhere around 20 to 30 percent of its rated capacity. Below that floor the machine does not keep slowing down. It stops and restarts, or it unloads, and you are back to the behavior you were trying to escape.
Why that matters: if you buy a 100 HP VSD for a plant whose demand sits at 15 percent most of the night, the machine spends the night cycling, not modulating. You paid the premium and got none of the benefit for a third of the clock.
Two rules come out of this:
- Do not oversize a VSD. The usual instinct to buy one size up is actively harmful here, because it pushes your normal operating point toward the bottom of the turndown range.
- Check the actual turndown range on the CAGI data sheet, not the brochure. Compare the minimum capacity figure against your real low demand. We walk through reading those sheets in how to read a CAGI data sheet.
The Setup That Beats a Single VSD
If you are running more than one machine, or you could be, there is a configuration that often outperforms buying one big variable speed compressor.
Run a fixed speed machine as your base load, sized to the demand you always have, and a smaller variable speed machine as the trim. The fixed speed runs flat out at its best efficiency point, which is where a fixed speed compressor is genuinely excellent, and the VSD handles the variation on top. You buy the drive only on the horsepower that actually needs to vary.
This usually costs less than one large VSD, gives you redundancy you did not have, and avoids the turndown trap entirely because the trim machine is sized for the swing rather than the total.
The catch is control. Two machines with independent pressure switches will fight each other and you will lose the savings. You need a proper sequencing scheme, which we cover in lead lag control for multiple air compressors.
Things That Save More Than a VSD
We sell variable speed compressors and we will happily quote you one. We would rather you do these first, because they are cheaper and the savings are more certain.
- Fix your leaks. A plant that has never been surveyed is commonly losing 20 to 30 percent of its output through leaks. That is the same order as the VSD savings, for the cost of a weekend and some fittings.
- Turn the pressure down. Every psi of discharge pressure is about half a percent in energy. Most shops run 15 to 20 psi above what they need because somebody chased a restriction with the pressure knob. See what PSI your compressor should be set to.
- Add storage. A bigger receiver lets a fixed speed machine run longer and cycle less, which recovers some of what a VSD would have given you for a lot less money.
- Kill the inappropriate uses. Open blow-off, cabinet cooling, and air motors running jobs an electric motor should do. Engineered nozzles alone often pay back in months.
- Recover the heat. Separate from the electricity question, but it is the largest single stream of recoverable energy in most shops. Heat recovery covers it.
If you do all five and your demand still swings hard, buy the VSD and it will pay. If you skip them and buy the VSD, you will have bought an efficient way to supply a leaking system.
Other Things a VSD Buys You
Energy is the headline but not the only argument, and two of these matter more than people expect.
- Soft starting. A VSD ramps the motor instead of slamming it across the line. That removes the inrush current, which can matter a great deal if your service is tight, if you are on a demand charge that punishes peaks, or if you are running on a generator.
- Stable pressure. A fixed speed machine on load/unload gives you a saw-tooth between cut-in and cut-out. A VSD holds a flat band. Flat pressure often lets you lower the setpoint a few psi, which is its own savings, and some processes genuinely care.
- Less cycling wear. Fewer starts means less stress on the motor and the airend.
Against that, the drive is another thing that can fail, it wants a clean and reasonably cool electrical environment, and a drive replacement out of warranty is not a small invoice. Dusty, hot compressor rooms are hard on drives. If that is your room, fix the room or weigh it in the decision.
Frequently Asked Questions
Is a variable speed air compressor worth it?
It depends on one thing: whether your demand varies. If your compressor runs loaded more than about 85 percent of its running hours, a VSD will not pay and a fixed speed machine is the better buy. If your demand swings through the shift or drops off at night and on weekends, savings in the 20 to 35 percent range are realistic and payback usually lands between two and five years.
How much energy does a VSD compressor actually save?
CAGI puts the typical figure around 33 percent, and a study of 67 installations averaged about $17,200 a year in savings. Those are averages from plants with variable demand. The right way to get your number is a week of logged data on your existing machine, not an industry average applied to your bill.
Can you oversize a variable speed compressor to leave room to grow?
Try not to. Every VSD has a minimum speed, usually around 20 to 30 percent of rated capacity, below which it stops modulating and starts cycling. Oversizing pushes your everyday operating point down toward that floor, which is exactly where the machine behaves like the fixed speed compressor you were trying to replace. Size it to your real demand and plan growth with a second machine.
Should I replace my fixed speed compressor with a VSD, or add one?
If the old machine has life left, adding a smaller VSD as a trim compressor alongside it is often the better economics. The fixed speed carries the base load at its best efficiency, the VSD handles the variation, and you get redundancy. You do need a real sequencing control so the two do not fight.
What should I do before buying a VSD?
Survey and fix your leaks, drop your pressure setpoint to the lowest number that runs your tools, add storage if you are short, and get rid of open blow-off. Those four are cheaper than a new compressor and the savings are more certain. Then measure your load profile for a week, and if it still swings, buy the VSD.
The Decision in One Paragraph
Log a week of data. If your machine is loaded nearly all the time, buy a fixed speed compressor and put the difference into leak repair and storage. If your demand swings, size a variable speed machine to your actual demand rather than one size up, check the turndown floor against your overnight low, and expect two to five years to payback. If you already have a working fixed speed machine, look hard at adding a smaller VSD as trim instead of replacing it.
Browse variable speed rotary screw air compressors, or the full rotary screw range. Send us a week of loaded and unloaded hours and we will tell you honestly which one you should be buying, including when that answer is the cheaper machine.
