Minisplit Testing and Commissioning Standards
Like many here, I’ve learned a lot from Dana Dorsett about ductless systems. Here’s his suggested standard of testing/commissioning a minisplit: “Pressure test at > 500 psi for at least 1h, preferably 24h, followed by pumping it down to < 500 microns and verifying that it will stay that low for 1h minimum. They should also measure air flow at the coil, and verifying that delivered CFM matches design CFM.” See #11 here: Daikin Quaternity | Terry Love Plumbing Advice & Remodel DIY & Professional Forum.
I’ve talked to every Mitsubishi Diamond contractors in the Austin area. None test to this standard. Generally, they just: Pressure test at 450-475 psi w/nitrogren for 20-30 min, then pump it down to 500 microns for 30 min, check CFM. Beyond that, they just offer contractor jabber about using special flare fitting leak lock material and expensive wrenches.
How big of a deal is this? Is Dana’s standard unnecessarily high, or something worth continuing to dig for (or offer to pay a premium for)? And, more selfishly… in my own case, I just need a single-head, 2-ton ductless unit. So perhaps this isn’t as big an issue for me?
Thanks!
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
This is a matter of good/better/best practices. Time is money- 24 hour tests take time and multiple site visits. Most installers won't go that route on a simple 1-zone mini-split install, but they arguably should for multizone systems. But it's still fine if they're willing to warranty their work using less stringent testing, but get it in writing. A Diamond installer will almost always use the right tools & materials, and by installing many dozens of systems in a year will know how often shorter tests screw up using their own methods,
A 30 minute test on a Nylog sealed system will rarely leak, and a single 20-30 minute purge of dry nitrogen will usually get enough of the moisture (the true enemy of longevity) out of the system. DIYers and "Chuck with a truck" low volume installers cutting corners (when they don't really need to) are the biggest source of installer-related problems. With the luxury of time a DIYer can spend more time testing to increase confidence in the install.
On a recent mini-split replacement project, installing a single zone Daikin with less than 15' of refrigerant line the installer (one guy, working on his own) was in and out in under 3 hours. But the contractor is of long standing with a good reputation, and backs up their work. I'm not too worried, but in the unlikely event that it craps out (usually takes less than 2 years if there's much moisture in the system, even less time if it leaks) I know they're good for it.
Thank you so much for this reply. I have literally met people on the street who have learned from your posts on this topic. Can't thank you enough! But of course I have to ask a bonus question: Mitsubishi recommends to let the unit sit for 24h after commissioning, before turning it on. Most contractors have told me: "I've never had a problem with turning it on right after the install. If you want to do it this way, that's fine, but you'll be the one to turn it on."
Turning something on seems uh, pretty minor, but I am brand new to the world of expensive, sensitive, high quality minisplits. Worth worrying about?
(I should add... I'm a contractor myself, not used to delegating, and very used to seeing guys cut corners (to put it mildly... "it'll be fine", the golden rule of bad work), so I may be overly paranoid here!)
>"Mitsubishi recommends to let the unit sit for 24h after commissioning, before turning it on."
That would be news to me! I haven't been involved in a Mitsubishi install super-recently, but I have never heard of, nor can I come up with a realistic rationale as to how that 24 hour post-commissioning idle would make a difference. I've never read that in the installation manuals (not that I've read every page.)
I have heard recently from a mini-split installer that it's better to never turn them off, even during low/no load conditions. The rationale given was that the longer it sits idle, there is some separation between the R32 & R125 (the subcomponents of R410A), and separation of the lubricants. I suspected that to be mere armchair speculation on his part- perhaps a guess as to why one of his installs may have failed early. (When in doubt, blame it on the homeowner?)
>"...I am brand new to the world of expensive, sensitive, high quality minisplits."
In general these things are not very "...sensitive...", though "...expensive..." can often be the case. (Especially but not exclusively with Mitsubishi equipment, which has been experiencing a lot of supply chain issues, with many models backordered and experiencing long delivery times in my area. That makes anything currently in inventory more valuable/expensive.) Old school 1 or 2 stage heat pumps need to be leak tested and moisture purged with dry nitrogen prior to charging & commissioning too.
Bottom line- if it's already been powered up and tested, it should be ready to run.