For most of the modern history of commercial development, HVAC was a category that rewarded standardization. You calculated the load, specified equipment from a catalog, installed it, and moved on. The technology was mature, the products were interchangeable, and the primary variable was price. That market still exists, and it is growing at roughly the rate you would expect for a mature industrial category. But it is no longer where the commercial HVAC industry is making its money, and the most recent earnings calls from Trane, Johnson Controls, Honeywell, and Schneider Electric make the divergence impossible to miss. The market has split, and everything interesting is happening on the engineered side.

The clearest evidence of the bifurcation of the HVAC market is a single number from Trane. Applied bookings, which cover custom-engineered systems for buildings with complex thermal loads, jumped 130% in the second quarter. That marks the fourth consecutive quarter above 100% growth and a fourfold increase on a two-year stack. Trane’s overall organic revenue for the same period grew 9%. Standard commercial equipment is growing at single digits and engineered thermal systems are growing at triple digits.

Johnson Controls reported the same pattern in different terms, with organic revenue up 6% overall but growth “led by continued strength in applied HVAC. What is driving that divergence is a change in what buildings are being asked to do. Johnson Controls CEO Joakim Weidemanis put it plainly on the company’s most recent call. “This is the age of thermal management,” he said. “AI factories, advanced and biopharma manufacturing, large research hospitals and universities require thermal management solutions at unprecedented scale.” Every building type on that list has process loads rather than comfort loads. The mechanical system exists to serve equipment and operations, not to keep people at 72 degrees. That is a fundamentally different engineering problem, and it cannot be solved with catalog equipment.

Liquid cooling is the clearest expression of that shift. Trane has directed capital specifically toward liquid cooling and modular chiller plant innovation, and recently released a DC-focused reference design for data centers with Eaton. The technology delivers thermal management directly to heat-generating components rather than conditioning the air around them, which becomes necessary rather than optional at high power densities. Rack densities in data centers have climbed from roughly 6 kilowatts in 2020 to 27 kilowatts today. Air cooling stopped being viable somewhere in that progression, and the manufacturers built product lines to match. That is not an incremental product improvement. It is a different approach to the underlying physics of heat rejection.

Absorption chillers represent the same logic applied to the energy input rather than the delivery mechanism. Johnson Controls has been promoting them specifically as a way to reduce data center energy demand, with Weidemanis arguing that lower consumption expands where data centers can feasibly be sited. Absorption chillers use heat rather than electricity as their primary energy source, which means a facility that can source waste heat dramatically reduces its electrical load. In markets where grid interconnection is the binding constraint on whether a project happens at all, that difference is not an efficiency improvement. It is the difference between a viable site and a dead one.

The intelligence layer is where the engineered market is pulling furthest away from the commodity one. Trane CEO Dave Regnery noted that “chillers are getting smarter” through integrated control systems that let buildings predict and optimize efficiency in real time. Johnson Controls has been leaning on its building automation platform and optimization suite, and reported that its growth is driven substantially by the productivity improvements those systems enable.

Schneider Electric has been the most explicit about that reallocation. The company has built out its portfolio through partnerships and acquisitions including NVIDIA, ETAP, AVEVA, AiDASH, Cognite, and Kraken, extending into electrical design software, digital twins, and grid intelligence. None of those are equipment investments. They are bets that the durable value in building systems is migrating from what gets installed to the software that determines how it operates.

Honeywell has made a version of the same bet, with CEO Vimal Kapur indicating that building automation products will proliferate within a slimmed-down Honeywell as the company restructures around its highest-conviction businesses. A company actively narrowing its portfolio choosing to expand building automation says something about where it expects returns.

The important thing for commercial real estate is that this shift extends well beyond data centers. Regnery pointed out that of the 14 verticals Trane tracks in the Americas, 11 were up double digits year to date in orders, and in the second quarter all 14 were up over 20%. “The growth that we’re seeing right now is very broad-based,” he said. Life sciences, healthcare, higher education, and advanced manufacturing are buying the same engineered thermal solutions, because they face versions of the same problem. High process loads, expensive energy, and performance requirements that commodity equipment was never designed to meet.

That has consequences for how buildings get planned. Trane’s backlog reached a record $12.1 billion, up 70% year over year, with roughly $6 billion already slated for 2027 and beyond. Johnson Controls’ backlog hit $20 billion. Lead times on engineered systems are extending, and tariffs and input cost inflation are pushing prices up at the same time, with both Trane and Carrier flagging those headwinds even as demand surged. The engineered equipment is getting more expensive and harder to get simultaneously.

The practical implication is that mechanical decisions have moved upstream in the development process. A building’s thermal strategy used to be something you resolved during design development, working from a load calculation and a catalog. In the engineered market, it is a decision that determines what tenants a building can serve, whether a site is viable, and what the operating economics look like over a twenty-year hold. The manufacturers have reorganized their capital allocation around that reality. Their record backlogs suggest the market already understood it.

The post Commercial Buildings Don’t Need Cooling Anymore. They Need Thermal Engineering. appeared first on Propmodo.

​