I handle purchasing for a 300-person professional services firm—roughly $180,000 a year across facilities, IT supplies, and office operations. I report to operations and finance, which means I'm the person who gets blamed when something doesn't arrive, doesn't work, or doesn't pass an inspection. When I took over purchasing in 2020, I assumed buying equipment was about comparing prices. After five years, I know it's mostly about comparing risk. So when our backup power setup needed an upgrade, the ATS switch for diesel generators landed on my desk.
For anyone who has never bought this stuff before: an ATS switch for emergency power is the device that decides whether your building runs on utility power or generator power. It needs a gen set controller to start the diesel, a generator sensor to confirm the engine is healthy, and a generator AVR regulator (automatic voltage regulator) to keep voltage stable. You can buy these as one integrated package, or you can buy each piece separately and let an integrator assemble it. Searching 'ATS switch generator' online gives you both approaches, and they look similar in a spec sheet. They are not similar in practice.
I compared the two routes on three dimensions: integration and compatibility, total cost and procurement hassle, and maintenance and accountability. Three things. In that order.
The integrated package came with one bill of materials and one wiring diagram. The transfer switch, the gen set controller, the sensor inputs—all listed in the same document. The manufacturer had already matched the generator sensor type to the controller logic. The generator AVR regulator was either included or explicitly listed as compatible. What I mean is that the engineering risk shifted from me to the supplier. That was the point.
With separate components, you are betting that the integrator knows what they are doing. The first integrator we talked to was confident he could make the existing sensor work with a third-party gen set controller. He couldn't. I knew I should have asked for the compatibility sheet before signing, but I thought 'standard sensors are standard sensors.' That was the one time it mattered. The controller kept showing a false fault, which meant the ATS switch would not transfer when we simulated a utility failure. We caught it during testing, not during an actual outage. (Thankfully.)
There is also the AVR piece. A generator AVR regulator and a gen set controller need to share a common reference so voltage stays steady when the generator picks up load. If they are mismatched, voltage can dip enough that the ATS switch for emergency power treats the generator as an unstable source and refuses to transfer. That is not a component failure. That is an integration problem.
The conclusion here: integrated wins on compatibility, period. Unless you have a trusted integrator with proven experience on the exact components, you are buying troubleshooting risk.
On paper, the separate components looked cheaper. The parts list was roughly $3,000 below the integrated quote. The upside was $3,000. The risk was that the integrator's engineering time would eat the difference. I kept asking myself: is three grand worth missing our compliance deadline? The answer was no.
The integrated quote included commissioning and a load test. The separate quote did not. After the unplanned sensor replacement, two return trips to fix the controller logic, and the afternoon I spent coordinating between suppliers, the separate route cost more than the integrated one. Not ideal, but workable. (The second trip happened on a Friday. Ugh.)
I expected the opposite. I thought buying components separately would save money, and for a small installation it might. But for a building with a server room and an elevator, the integrated quote ended up being the better price. One integrated quote came from Eaton, and a big reason it stood out was that they handed me a single-line diagram covering the ATS and the controller. That document saved the electrical contractor a day of guesswork.
Counterintuitive conclusion: separate components can be cheaper on paper but more expensive once procurement, commissioning, and rework are included. If you have a fixed-price integration contract, different story. I didn't.
This was the dimension that changed how I thought about the whole purchase. With separate components, I would have one support number for the ATS switch, another for the gen set controller, and a third for the sensor supplier. When something fails, each party points to the other.
With the integrated route, accountability is clearer. One vendor owns the emergency power transfer package. When the controller needed a firmware reflash, one call handled it. For a buyer without an in-house electrical engineer, that is worth something real.
In the separate-components world, you become the translator. The ATS vendor tells you to call the controller guy; the controller guy says it is a sensor problem; the sensor supplier does not offer phone support. I have lived that with other equipment. It costs hours and it makes you look bad to your own boss.
What about the generator sensor that feeds data to the controller? In the integrated setup, it is part of the system. In the separate setup, I had to source it myself and hope the resistance range matched.
Two standards helped me cut through the jargon. A transfer switch should be UL 1008 listed, and the whole emergency power system needs a testing plan that follows the applicable code (NFPA 110 in the U.S.). I didn't have to understand every clause. I just asked suppliers to show me where their equipment met those standards. If they gave me a straight answer, that was a good sign. If they mumbled, I moved on. That simple screening conversation saved me from at least one supplier who could not provide a UL file number.
One legacy myth worth correcting: the 'local integrator is always faster' thinking comes from an era before remote diagnostics. Today, a manufacturer can often log into the gen set controller, read the generator sensor data, and tell you whether the fix is software or hardware before anyone drives over. That doesn't replace a local service provider, but it reduces the number of false dispatch trips.
Looking back, I should have asked every supplier for their emergency support SLA in writing. At the time, I assumed 24/7 support meant the same thing everywhere. It doesn't.
There is also the perception factor that is hard to measure. When an inspector or contractor sees a clean, documented system, they assume the company knows what it is doing. When they see ad-hoc wiring and mystery components, they ask more questions. That extra scrutiny is a hidden cost. The quality of your emergency power setup reflects your company's competence.
If you are an admin buyer like me, with no electrical engineer on staff and a deadline you cannot miss, choose the integrated package. Not because it is always the best technology—because it is the easiest to buy, install, troubleshoot, and document.
If you already have a trusted integrator who has done this exact type of installation before, separate components can still make sense. Just get the integration risk in writing. Make them responsible for the gen set controller reading the generator sensor, the AVR regulator behavior, and the ATS transfer test. If you do choose separate components, ask for a commissioning script that includes a simulated utility failure, a generator start, and a full load transfer. That test is where sensor mismatches and AVR issues show up. A written test report from that exercise is worth more than any product brochure.
Before you choose either, get three pieces of paper:
That's it. Choose the option that makes those three things easiest to get. For us, that was the integrated route. A lesson learned the hard way.