I coordinate emergency work for compressed air systems. Not planned maintenance—the calls when a line is already down and someone needs an answer in hours. After dozens of rush jobs, one lesson stands out: when a plant asks whether an Atlas Copco G15 or ZT90 is in stock, stock is rarely the real problem.
At 2:30 on a Friday in March 2024, a maintenance manager called about a failed compressor. He needed an Atlas Copco ZT90. Did we have one? Could it be on site before the night shift? He had already called two other suppliers and was checking stock while he spoke. What he hadn’t done was check why the old unit had stopped.
I don’t say that to be difficult. I say it because the same pattern repeats on nearly every emergency call I take.
The request that sounds like inventory management
“Do you have one in stock?” is usually the first sentence out of the caller’s mouth. It’s also the most misleading sentence in the conversation.
From the outside, it looks like a logistics problem. Find the compressor, put it on a truck, install it before production loses more money. The reality is different. A compressor model number describes a family, not a specific machine. A ZT90 might be air-cooled or water-cooled, built for one voltage or another, with or without an integrated dryer, with a standard controller or something else. A G15 can vary in the same way. The word “stock” can’t capture that.
“ZT90” is a starting point, not a specification. If I tell someone yes, there is a ZT90 in stock, without checking the full order code and duty, I am guessing. Guessing is expensive.
Why the model number pulls us off track
In my first year doing this kind of work, I assumed the biggest factor was matching the old nameplate. Get the same model, get the same performance, done. That assumption failed me more than once.
A G15 and a ZT90 sit in very different parts of the Atlas Copco range. They are not two sizes of the same product. They can have different motors, different cooling systems, different pressure capabilities, and different air quality assumptions. But the deeper issue is this: the model on the old nameplate may have been chosen when the production line was different. If it was wrong from the beginning, buying the same model again simply reproduces the original mistake.
In urgent situations, people don’t want to hear that. They want a fast answer. But I have arrived at plants where the compressor that failed had been undersized for years. It ran hot. It ran almost continuously. It cycled off only when someone shut the line down. Replacing it with an identical Atlas Copco model because that model is available in stock does nothing except schedule the next failure.
The bigger issue: failures are rarely sudden
I used to assume breakdowns were random. That’s how they feel on the plant floor: the machine was running, then it stopped. Then I started asking for the controller’s event log before quoting a replacement. It was rarely empty.
There were high-temperature warnings weeks before the trip. There were service reminders that had been ignored. There was a slow pressure drop that operators had adjusted around. The machine had been sending signals for a long time. The “emergency” was postponed decision-making, not a surprise.
That’s uncomfortable to hear, especially when a production manager is standing next to a dead compressor. But it matters. If you don’t fix the reason the old unit failed, the new unit—even a perfect one—will be installed into the same bad situation.
What stock-based decisions actually cost
On the surface, the costs of an emergency are easy to see: overtime labour, express freight, a rental unit, lost production. Those are visible costs. The hidden ones are worse.
If the “in stock” machine matches only the model number and not the system’s actual requirement, the line can restart and still have problems. Maybe the compressor cycles too often. Maybe pressure sags when two lines start at the same time. Maybe the air purity doesn’t meet the process requirement. Then the plant has paid for a compressor and a quality incident.
In food, pharma, and electronics environments, the question isn’t just oil-free or not. It’s what ISO 8573-1 purity class the process needs. A stock-list search can find a machine. It cannot tell you whether that machine is suitable for the application.
The fix I’ve landed on: verify before you search
After too many expensive lessons, I now ask for four things before I quote an emergency compressor:
- The controller’s event history, not just the current alarm.
- A photo of the nameplate, not a model number from an old invoice.
- The required duty—pressure, flow, ambient conditions, and air purity class.
- The full configuration of any available machine—cooling, voltage, controller, dryer, and order code.
That list isn’t bureaucracy. It takes a few minutes, and it usually separates a correct decision from a costly one. Five minutes of verification often beats five days of correction.
Stock is not evil. It matters. But an Atlas Copco ZT90 or G15 sitting in a warehouse is just metal and electronics until it matches the application. The same machine that solves one plant’s emergency can create a second emergency somewhere else.
So the next time someone asks if you have an Atlas Copco G15 or ZT90 in stock, pause. Ask why the old one failed. Ask what the process actually needs. Then ask about the truck.
A compressor can be delivered fast. A wrong compressor, installed fast, costs much more than the downtime you were trying to save.