Here’s my take: If you’re still specifying building controls based on what worked in 2020, you’re spending money you don’t need to spend and leaving efficiency on the table. I know that sounds like a strong statement, but I’ve seen the evidence firsthand in the quality audits I run. Industry has changed, and a lot of the “best practices” are hanging around by inertia.
I review roughly 200+ unique HVAC component spec sheets and delivery batches annually for a major commercial systems integrator. In Q1 2024 alone, I rejected 12% of first deliveries—not because they were broken, but because they were specified wrong for the current energy landscape. The specs were based on assumptions from 2019–2020 that just don't hold anymore.
The Big Shift No One Talks About
What most people don't realize is that the traditional hierarchy of “chiller efficiency first, controls second” has inverted for modern commercial buildings. A chiller from 2019 is still a perfectly good chiller. But the control strategy you’re using to run it? That’s where the waste is now.
I’ve seen a 5-year-old York chiller paired with a legacy pneumatic thermostat—like the Johnson Controls T-4002-201—and the building owner was convinced a new chiller was the only solution to rising energy costs. The surprise wasn't the chiller's performance. It was that retrofitting the control system alone dropped their cooling load by 18% without touching the mechanicals. The fundamentals of refrigeration haven't changed, but the execution of control has transformed.
Why “Standard” Thermostats Are a Liability Now
Look, pneumatic thermostats like the T-4002-201 are bulletproof. I still respect that. But here's something vendors won't tell you: the industry standard for commercial thermostat accuracy used to be +/- 2°F. That was considered fine. For a data center or a pharmaceutical storage area in 2025, that tolerance is a problem.
In our Q1 2024 quality audit, we ran a blind test comparing a standard pneumatic stat against a modern direct digital control (DDC) sensor in a test environment. 78% of our facilities team identified the DDC zone as “more comfortable” without knowing which was which. The cost delta on that upgrade? About $180 per zone on a 50-zone system. That’s a $9,000 total upgrade for measurably better perception and actual energy reduction. Compare that to the cost of a single chiller overhaul.
The Data Center Disruption
And then there’s data center cooling. Honestly, the rules have been rewritten here. If you’re still designing a data center cooling system without an AI-based control layer—like what we’re seeing with the 2025 Johnson Controls data center cooling AI systems—you’re playing a different game. The old best practice was to over-provision cooling and run redundant units at partial load. The new approach is predictive load balancing.
I still kick myself for not pushing for an AI overlay on a retrofit project in 2022. We went with a conventional sequenced chiller plant. The first year, we were trimming setpoints manually based on IT load projections. The control vendor’s “standard” algorithm was reactive—it responded to heat, it didn’t predict it. If I’d specified the predictive controls, I estimate we’d have saved about $22,000 in the first year alone in reduced fan energy and chiller cycling. That data center is now getting retrofitted anyway. The change in what the market expects has made some 2022-era designs obsolete.
The Machinery Behind the Controls
This isn’t just about thermostats for comfort cooling, either. We’re seeing the same evolution in industrial equipment. A Variable Frequency Drive (VFD) controlling a condenser fan or a process pump—the VFD hardware itself hasn't changed radically. But the communication protocol and the diagnostic feedback have. Five years ago, a VFD that could report “Overload Fault” was considered adequate. Now, the expectation is a VFD that sends a predictive maintenance alert before the bearing fails, tied into the building’s Metasys system.
When we specify equipment for a 50,000-unit annual order of components, the number one quality issue we flag now isn’t sheet metal tolerance. It’s the lack of BACnet or Modbus communication capability on a “dumb” piece of equipment. That upgrade costs pennies on the dollar compared to the labor of retrofitting a controller later. Every contract I write now includes “BACnet/IP communication” as a mandatory spec, not an option.
But What About My Existing Equipment?
I hear the pushback all the time: “I can’t rip out my 10-year-old chiller plant, my rooftop units, or my pneumatic system.” And you’re right—you shouldn’t. The fundamentals of the refrigeration cycle are the same. But don’t confuse mechanical longevity with operational efficiency.
The industry in 2025 isn’t forcing you to replace core machinery. What it is doing is making the case that the brain controlling that machinery—the thermostat, the control panel, the building automation front-end—is where the biggest ROI lives. Upgrading a control sequence or adding a modern sensor to an existing air handler can achieve results that would have required a full mechanical retrofit ten years ago.
So, my position stands: the best practice from 2020 is a starting point, not a target. The hardware you bought back then is likely still mechanically sound. But the way you control it? That’s probably costing you. The industry has evolved. It’s fine to respect the old equipment—I do. But ignoring how control technology has changed is a risk to your budget and your facility’s performance. The T-4002-201 is a classic. Just don’t ask it to do a 2025 job.
