Master controls, built-in sequencing or cloud controls: how to choose

Every compressor has its own controller, and on its own it does a good job of holding its own pressure. Put three or four of them on the same header and each still works alone. None of them knows what the others are doing.

What goes wrong without system controls

The usual way to share the load without system controls is a pressure cascade: each compressor gets a slightly lower setpoint than the one before it, so they start one after another as the pressure falls. It works, but it costs energy:

• The pressure runs high. The band has to be wide enough to fit every compressor's setpoints, and the top of it is often well above what the plant needs. Each 2 psi costs about 1% more compressor power.
• Too many compressors run. Units start on a short pressure dip and then run part-loaded or unloaded, because nothing tells them they are no longer needed.
• More than one unit trims. Two or three compressors all running part-loaded at the same time cost far more than one doing the trimming while the rest run fully loaded or stay off.

Good system control does three things: it holds one narrow pressure band, runs only the compressors the demand needs, and lets one compressor, the trim unit, take up the swings. Usually that is the one best at running part-loaded, often a variable speed unit.

The three kinds of system control

1. Built-in sequencing

Many compressors can be networked through their own controllers, usually within one brand. One controller leads and the others follow.

• Does well: low cost, little extra hardware, simple to set up and use. Often the right answer for two to four compressors of the same make.
• Falls short: mixed brands and older units may not join in, or only through add-on modules. The logic is fixed by the maker, and changes or repairs go through them.

2. A dedicated master controller

A separate controller runs the whole compressor room. It may be a packaged master control from a compressor or controls maker, or logic written into the plant's own PLC or control system.

• Does well: handles mixed brands, ages and types; can use storage, flow and pressure readings from around the plant; runs locally, with no outside connection needed. Built on open standards, it can be supported by more than one integrator.
• Falls short: costs more up front, and it needs someone who understands both the controller and the compressed air system to set it up and keep it tuned. A badly set master control can do worse than a good cascade.

3. Cloud-connected controls

A local controller or gateway connects to the compressors and sends their data to a remote service, which adds monitoring, reports and, in some systems, the control decisions themselves.

• Does well: remote visibility, alarms on a phone, energy reports, and one view across several sites. Updates arrive without a site visit.
• Falls short: usually a subscription for as long as you use it. The logic is the vendor's, and your data lives on their servers. Most importantly, the compressors have to keep running properly if the internet connection drops.

Questions to ask before you buy

• What happens if the network or internet goes down? The compressors should fall back to safe, local control on their own, without a pressure dip that stops production.
• Can it run every compressor you have? Every brand, age and type, including the variable speed unit and the old one kept as a backup.
• Who can change a setpoint, and who can find a fault at 2 a.m.? If the answer is "only the vendor", count that into the cost.
• Can you get your data out? In a plain format you can open without the vendor's software.
• What will it cost over ten years? Not just the purchase: installation, commissioning, service contracts, subscriptions, upgrades and training.
• Is the rest of the system ready? Controls can't fix storage that is too small, a trim unit too small to cover the gap between the other units' steps, or pressure lost in dryers, filters and piping before the air reaches the plant.

Start by measuring how the compressors run now

The right choice depends on how the system behaves today, not on a feature list. A week of logged pressure and compressor power, including a weekend, shows:

• how often more compressors run than the demand needs, and how many run part-loaded at the same time;
• whether the trim unit can cover the swings, or whether there are demands no combination of compressors fits well;
• how wide the pressure band really is, and how much pressure is lost on the way to the plant;
• whether storage is enough to let a compressor stop instead of running unloaded.

That same week of data becomes the baseline to measure the new controls against. Without it, no one can say what they saved.

See it in a real report: which units should run at each demand

There is no single best kind of control. Built-in sequencing suits a small fleet of one make, a master controller suits a mixed or complex system with people to look after it, and cloud controls suit plants that value remote visibility and are comfortable with a subscription. Whichever you choose, choose it from measured data, and measure again after it is in.

Source: U.S. Department of Energy, Improving Compressed Air System Performance: A Sourcebook for Industry, compressor and system controls. Pressure rule of thumb as used in our other articles.