Here's the thing: I'm a quality compliance manager in building products, and I review radiant heating installations before they get covered with drywall. This spring alone, I checked about 40 manifold setups—new ones and reworks where the pressure test had already failed. The checklist below is the one I run on an Uponor ProPEX system. It works for a single-zone bathroom floor and for a whole-house layout with an Uponor 4 port manifold.
Before we get into it: if you landed here looking for a halter top, a scally cap, or how to get rid of fleas in house, this is not that post. I'm a quality inspector, not a fashion or pest specialist. But if you're installing PEX-a and want to avoid the mistakes I see every month, keep reading.
Seven steps. Most people skip Step 5. More on that later.
PEX is PEX, right? No. PEX-a, PEX-b, and PEX-c are different crosslinking processes, and they behave differently with expansion fittings. The ProPEX system is PEX-a with a cold-expansion method: you expand the tubing and the ring, insert the fitting, and let it contract. It's a solid system, but only if the tube, ring, and fitting are all rated for the same method.
Check the markings. The system is sometimes written as 'Uponor pro pex' and sometimes as 'Uponor ProPEX'—same expansion system. A genuine tube should have the Uponor name and the applicable PEX standard reference printed on it. The expansion fitting method is covered under ASTM F1960, so the print line should make sense to someone who works with PEX. If you see a no-name coil with a printed PEX and no standard, stop. I don't care what a bargain supplier told you.
I'm not going to tell you PEX-a is the only sensible pipe. To be fair, PEX-b and crimp systems have done plenty of good work. But the expansion method depends on the material having enough shape memory, and that's not a place for a surprise substitution.
This is boring, and it's where quality problems show up. In our Q1 2024 quality audit, about 14% of the failed pressure tests came down to expansion rings that weren't seated evenly. The installers had used the correct parts and the correct tool. They just rushed the insertion and didn't check all sides.
The ring should sit in the groove on the fitting with a consistent shoulder all the way around. If one side looks deeper, or the ring looks cocked, cut it off and redo it. That costs you one coupling and ten minutes. Covering it, failing the pressure test, and draining the system costs you a lot more.
Honestly, I've never fully understood why installers resist this step. My best guess is that the expansion tool makes the connection feel foolproof. It's not.
Some installers test the tubing before connecting the manifold, then assume the manifold connections are fine. That's optimistic. The manifold is where most field leaks appear, especially when it's an Uponor 4 port manifold with multiple loops tucked into a small cabinet.
The typical procedure is to fill the system, bleed the air, then bring it up to the test pressure required by local code and by Uponor's current documentation. I usually run 80 psi for 30 minutes and watch for a drop. But my number doesn't matter. What matters is that you use a test gauge you trust, not the one that's been rattling around in a truck since 2019.
If the pressure drops, find the leak before you open a wall. Pressure testing is not a negotiation.
Here's a mistake I see often: someone orders an Uponor 4 port manifold because they have four rooms, but the system is actually one big zone or three zones plus a spare port. That's fine if it's intentional. It's a problem if nobody checked the flow.
The manifold doesn't create zones. The valves and controls do. Before you mount it, confirm each loop's length and flow requirement against the heat loss calculation. A four-port manifold can carry four short loops, two longer loops, or three loops with one spare port—as long as the total flow stays within the manifold's rated range.
I also look for flow balancing. If the manifold doesn't have flow meters or balancing valves, you're setting up for temperature imbalance. Not every Uponor manifold includes them, so read the spec sheet. This is a quality control thing, not a sales pitch.
PEX-a expands and contracts more than copper or steel. If you clip it tightly at every stud with no allowance for movement, the pipe will still try to move, and the stress can concentrate at the manifold connection. I've seen a fitting pass the pressure test and then start dripping six months later. The issue wasn't the fitting. It was over-anchoring.
Use long-radius bends where the tubing changes direction. Let straight runs move a little. Avoid hard 90-degree bends right at the manifold outlet. The manufacturer's design manual shows the minimum bend radius and allowable lengths—use that, not the 'it is flexible, so it will be fine' approach.
This is the step I check first on a rework call. Almost nobody budgets time for it.
If there are no labels, I don't care how neat the work looks. You'll be guessing later. Label each valve at the manifold and the room end. Use a label maker or a handwritten tag—whatever will survive in a warm mechanical room.
Take a photo of the completed manifold, including the flow meters and any adjustment valves, before the cabinet door goes on. Upload it to the project folder. When someone calls about a cold floor in the middle of winter, that photo is worth more than a service call.
I wish I had tracked how many times labels saved a service visit. I can say anecdotally it's a lot. I'd guess at least a quarter of the homes I review have unlabeled loops in the manifold cabinet.
Here's something vendors won't tell you: the first number you see can leave out the parts that make the system work. A low quote might include the tubing and the manifold but not the expansion rings, the balancing valves, the cabinet, or the freight. That's not necessarily malicious—it's itemized quoting. But if the total seems low, ask what's not included.
I've learned to ask 'what's NOT included' before 'what's the price.' The vendor who lists all fees upfront—even if the total looks higher—usually costs less in the end.
One client saved $200 by buying a lookalike manifold online. It didn't match the ProPEX connection pattern, and the rework set them back about $1,100. The budget choice looked smart until the pressure test. Net loss: $900 and a week of schedule.
I don't care if the manifold is perfectly centered in the cabinet. I care that there's enough clearance to operate a balancing valve. I don't care if every wire tie is perfectly aligned. I care that the tubing isn't kinked at the manifold outlet. I don't care about a manufacturer logo facing the camera. I do care that the photo shows the balancing settings.
This isn't a beauty contest. It's a system that has to survive thermal cycles, water pressure, and a service technician who wasn't there on day one.
I don't have hard data on how many manifold failures are caused by poor pipe support versus bad parts. Based on years of reviews, my sense is that the parts are rarely the problem. It's the details around them. Run the checklist, take the photo, and keep the documentation. That's what separates an install I'll sign off on from one I'll make someone redo.
Share this article:
Leave a Comment