-
Who this checklist is for
-
Step 1: Lock down the operating conditions in writing
-
Step 2: Confirm the flange standard, not just "flanged"
-
Step 3: Decide swing vs. spring — deliberately, not by default
-
Step 4: Vet the manufacturer, not just the supplier
-
Step 5: Nail down installation orientation before the valve ships
-
Step 6: Compare total cost, not unit price
-
Things I still check every time
Who this checklist is for
I'm a procurement engineer. I've been handling industrial valve orders for nine years, and in that time I've personally made — and written down — eight significant mistakes totaling roughly $11,000 in wasted budget. Most of them were in swing check valve sourcing. That's why I keep this list.
If you're placing a swing check valve order and you want to not repeat my mistakes, work through these six steps in order. I'm not going to explain what a check valve does or give you a design history. You already know. This is about not screwing up the order.
Six steps: spec, connection, type, source, install, cost. Skip any of them and you'll probably still get a valve delivered. It just might not be the right one.
Step 1: Lock down the operating conditions in writing
In my first year — 2017 — I made the classic assumption error. I told a supplier we needed a "6-inch swing check valve for water service." That's not a spec. That's a wish.
The valve arrived. Six-inch was right. Water service was roughly right. But our line pressure was around 285 psi and the valve's rating was 200 psi WOG. Nobody caught it until installation day. We paid about $340 in return freight and ate a two-week delay while a replacement shipped.
What I should have sent was a one-page spec sheet: line size, pressure rating, temperature range, media, flow rate, horizontal or vertical orientation, and the applicable standard. For swing check valves, the relevant references are usually API 594 (valve design) and ASME B16.10 (face-to-face dimensions). If your supplier's engineer needs to ask you for any of those things, you haven't written the spec yet.
The check point: can a stranger read your spec and order the right valve without calling you? If no, keep writing.
Step 2: Confirm the flange standard, not just "flanged"
This one bit me in 2021. I ordered a batch of flanged swing check valves and specified "flanged, 4-inch." What I didn't specify was ANSI Class 150 vs. DIN PN16. The factory — a shop we'd used before, decent work — shipped DIN PN16 because that's their default for that line.
DIN and ANSI flanges don't bolt together. Not ever. Different bolt circle, different bolt count. I said "flanged." They heard "flanged, our standard." Result: eight valves, around $1,900 total, sat in our warehouse for six weeks while our engineering team tried to source adapters they ultimately wouldn't sign off on. We returned them.
Now every flanged swing check valve order includes: flange standard (ANSI/ASME, DIN, JIS, or BS), pressure class, facing type (RF, FF, RTJ), and bolt pattern. Five data points on one line of the spec. Takes thirty seconds to write. Has saved me four figures since.
Step 3: Decide swing vs. spring — deliberately, not by default
Everything I'd read starting out said spring check valves were the "better" choice because they close faster and reduce water hammer. In practice, for most of the orders I've handled — horizontal water lines, moderate flow, no rapid pump cycling — the swing check valve is the simpler, more maintainable choice, and I default to it unless the application says otherwise.
Here is how I actually think about swing check valve vs spring check valve now:
Swing: slower closure, relies on gravity and backflow, cheap to repair because you can usually replace the disc or seat, fewer parts to fail. Good for horizontal runs at low-to-moderate velocity. Can slam if flow reverses quickly.
Spring: faster closure, can install in any orientation, more parts to fail, and when the spring fatigues the valve behaves differently without any visible sign. Some spring designs also add cracking pressure to your system's head loss — which matters if your pump is sized close to its curve.
It's not a "which is better" question. It's a "which failure mode can your process tolerate" question. If a slam will damage downstream equipment, spring. If a spring failure would strand you with no replacement, swing. Decide on purpose, write the decision down, and move on.
Step 4: Vet the manufacturer, not just the supplier
This is the step that sourcing platforms tend to gloss over. Anyone can be a swing check valve supplier. Suppliers resell. Swing check valve manufacturers make the valve. Those are different businesses with different failure modes, and treating them as interchangeable cost me a project schedule in 2022.
I placed an order through a supplier who quoted a good price and promised three weeks. I never confirmed whether they were manufacturing or brokering. They were brokering. The actual factory was running behind, and my three-week lead time became nine. The valve was fine when it arrived. The schedule damage was not, and there was no language in the PO that gave me any recovery.
Now, before I order, I ask three questions: Is the factory you quote from your own or a partner's? Can you send a recent pressure test report from that factory? Who signs the certificate of conformance — your company or the factory's QA? A manufacturer answers all three without hesitation. A broker usually stumbles on question two.
The check point: if you can't name the factory, you can't inspect it, and if you can't inspect it, you're accepting lead-time risk that never showed up in the quote.
Step 5: Nail down installation orientation before the valve ships
Swing check valve installation seems obvious — flow arrow, horizontal pipe, done. Except "horizontal" isn't always horizontal, and the arrow doesn't cover everything.
I caught this one on a call in early 2025, luckily before a shipment went out. A junior engineer had spec'd a swing check valve for a line that ran mostly horizontal but turned vertical for the final two meters. A swing check valve depends on gravity to close fully. On a vertical section, with low or intermittent flow, it may not seat properly. Nobody had noticed because the drawing called the whole run "horizontal."
Three things I check before the order goes out now:
- Is the line truly horizontal at the valve location, or is there a slope or vertical section nearby? Swing valves generally want horizontal installation with the bonnet up.
- Is there clearance above the valve for disc swing and for maintenance? A valve crammed against a ceiling pipe is a valve you'll never service.
- Does the flow direction match the arrow on the valve body — and did you check the arrow against the drawing, not the box label?
I've seen a valve installed backwards because the shipping label had the arrow on the wrong side of the box. That particular mistake cost a Saturday.
Step 6: Compare total cost, not unit price
Last step, and the one that changed how I read quotes. A swing check valve's sticker price is maybe half its real cost. The rest lives in shipping, spares, downtime, and rework — which is the whole argument for total cost of ownership thinking, and I don't apologize for repeating it to my own team every quarter.
Here's a real comparison from 2023. Two quotes on a set of six-inch stainless flanged swing check valves. Quote A: $420 per valve, lead time six weeks, shipping quoted separately. Quote B: $510 per valve, lead time three weeks, shipping included, spare parts kit included.
On unit price, A wins by $90 per valve. Across eight valves, that's $720 "saved." But A's shipping came in at around $280, its spare kit was another $160, and it delivered a week later than promised — which pushed a hydrostatic test we'd scheduled. When I added it all up, B's all-in number was lower than A's. The "expensive" quote was the cheaper one.
What I include now in every valve TCO comparison: unit price, shipping and duties, spares and replacement parts, expected lead-time variance (I use the supplier's on-time rate if I have it, 90% as a default if I don't), installation labor, and expected repair interval. None of it is exotic. It's just honest.
Things I still check every time
A few loose ends that don't fit neatly into steps but keep saving me:
Confirm the test standard and ask for the actual report, not a "tested" checkbox. For most industrial valve orders, a hydrostatic shell test and a seat leak test to ISO 5208 (usually Rate A or B) are standard. If a manufacturer can't produce a report on request, that tells you something about the paperwork you'll get after delivery.
Ask for photos of the actual valve before it ships. Not the catalog shot. The one on the pallet. If the casting finish, the bolting, or the nameplate looks wrong, you now have a reason to hold the shipment instead of discovering it on install day.
And get the certificate of conformance signed before final payment. Not "we'll send it after." Before. If it needs a signature from the factory QA manager, that signature should exist when you release the wire. That one line has caught two material mix-ups for me since I started writing it into POs.
The pattern across all six steps is the same: the mistake almost never shows up at the price stage. It shows up two weeks after delivery, when you're the one explaining the delay. Write the spec, name the standard, check the factory, walk the install, and price the whole thing. The cheapest quote is rarely the cheapest order — and I say that as someone who has the invoices to prove it.