The HVAC Specification Gap: Why Your Project Budget Keeps Breaking (And What Actually Causes It)

You budgeted for equipment. The overruns came from everywhere else.

I've been managing our HVAC procurement budget for about six years now. The equipment line items? Those are usually fine. We get quotes, we compare, we negotiate. Standard stuff.

The problem is everything around the equipment.

After tracking roughly $340,000 in cumulative HVAC spending across those six years — and documenting every change order, every rush shipment, every "we missed something in the spec" email — I've come to a conclusion that took me way too long to see clearly: the biggest cost driver isn't the air handling unit or the thermostat. It's the gap between what the spec says and what the job actually needs.

That gap has cost us, on average, about 14% above quoted project budgets. Every year. On every project.

The problem isn't the price you see. It's the specs you don't question.

Here's what I mean. When a project comes in — say, a commercial fit-out or a system replacement — we get a specification document. Maybe it's from an engineer, maybe it's adapted from an old project. It lists equipment: so many outdoor units, so many indoor units, ductwork requirements, control systems, thermostats.

We quote it. We buy it. We install it.

And then, somewhere around week three, someone discovers that the spec called for a mitsubishi electric indoor unit model that was discontinued. Or the ductwork sizing doesn't match the actual ceiling void. Or the control system uses a different communication protocol than the rest of the building.

Now we're scrambling. And scrambling costs money — rush shipping, expediting fees, labor overtime, temporary equipment rentals. I've seen a single "minor" spec mismatch add $11,000 to a project that was supposed to be $80,000.

(I really should build that spec audit checklist. Mental note: do it before the next project cycle.)

The deeper issue: nobody owns the specification

I used to think the problem was bad specs. Specific errors, specific omissions. Fix the errors, fix the problem, right?

But after going back through our project records — about 40 projects over six years — I noticed something uncomfortable. The specs weren't always "wrong." They were just... disconnected.

The engineer specified equipment. The contractor interpreted the spec. The procurement team ordered against it. The installer worked from a slightly different version. And the mitsubishi electric thermostat compatibility question — whether the controls would integrate with the building's existing BMS — that fell into the gap between everyone's responsibility.

Nobody owned the whole picture.

I'm not sure why this happens so consistently. My best guess is that the industry rewards speed over completeness. Everyone assumes someone else already checked that detail. And when you're working across multiple vendors — an equipment supplier, a controls specialist, a ductwork fabricator — the handoff points are where things quietly fall apart.

It took me about four years and roughly 200 purchase orders to actually understand this. Early on, I blamed vendors for "not flagging issues." Then I blamed engineers for "incomplete specs." Now I think the responsibility is shared, and that's exactly why it keeps happening. When everyone owns a piece, no one owns the whole.

What it actually costs

Let's talk numbers, because "spec gaps" sound abstract until you see the invoices.

Across the projects I've tracked, here's where the unexpected costs landed:

• Expediting and rush shipping: 31% of total overruns
• Change orders and rework: 27%
• Temporary equipment rentals: 18%
• Labor overtime: 14%
• Compatibility fixes (controls, ductwork, power): 10%

Those rush shipments are usually the worst. An air handling unit that's normally 6-8 weeks out becomes 2 weeks if you pay a 40-60% premium. And when you're mid-project with a crew standing around, that premium is almost always worth paying. But it shouldn't have been necessary in the first place.

The ductwork private label situation is especially frustrating. We've had projects where the spec called for custom-fabricated ductwork, and we waited weeks for delivery — only to find that a standard configuration with minor modifications would've worked fine. That's not a supply chain problem. That's a specification problem.

An honest admission: I still don't have this fully solved

I've tried a few things. We now require a pre-purchase spec review with the contractor and the controls vendor on the same call — that alone caught three potential issues on our last project. We've also started asking for compatibility documentation for every mitsubishi-electric product we specify, especially around thermostat and control integration.

But I'd be lying if I said we've eliminated the problem. HVAC systems private label complexities, bespoke ductwork, integrated controls — these aren't commodity purchases. There's no single "correct" approach that works for every building.

I've never fully understood why some projects sail through with zero issues while others hit every possible snag. My honest guess is it comes down to how much time someone spent thinking about the edges — the connections between systems, not just the systems themselves.

What actually helps

I'm not going to give you a ten-step process, because I don't think the answer is a process. I think it's a mindset shift in how the specification gets built.

Three things have moved the needle for us:

1. Build the spec around the whole system, not the equipment list. An air handling unit specification guide that only covers the AHU itself misses the ductwork interface, the control integration, the power requirements, and the commissioning sequence. We now require a one-page "interface summary" that explicitly names every connection point and who owns it.
2. Verify compatibility before you order, not after. This sounds obvious. It isn't. When you're ordering mitsubishi electric indoor units alongside third-party controls, "compatible" can mean three different things depending on who you ask. Get it in writing.
3. Budget a 10% spec contingency — and track what consumes it. If you don't, the overruns silently eat your margin, and you have no data to prevent the next one. We started tracking this in Q2 2024, and it's been illuminating.

The efficiency argument here isn't about fancy software or digital tools. It's about closing the information gap between specification and execution. Every hour spent on that front end saves somewhere between three and five hours on the back end — based on our own project data, anyway.

High-efficiency systems and well-engineered equipment only deliver their promised performance when the specification actually reflects reality. The equipment won't save you if the spec gap eats the savings before the system ever gets commissioned.

That's the part I wish I'd understood six years ago.