Custom water cooling loop issues with distilled water

Hello,

I have an issue with my custom water cooling loop.
Since this is the first time I’m having an issue like this I wanted to ask what’s most likely the cause and what’s the recommended approach to solving it.

At first I thought it was some brownish dirt or something judging by the residue on top of the water reservoir but soon afterwards residue started appearing on the bottom so I suspect it’s a result of galvanic corrosion I guess? I can see tiny bits that seem to be metalic in the fluid.

All parts are copper/brass/Nickel-plated copper/EPDM soft tubing.
I’m using Mo-ra 600 which I didn’t clean beforehand.
In the loop I’m using off the shelf distilled water with nothing added.
I’ve ordered and have on hand 10 liters of DP Ultra.
My plan was to drain the whole loop wash out all the fittings and tubing.
Dismantle the CPU block and see if it’s clogged up. Clean it all up as much as I can and restart the loop with DP Ultra. Not sure what to do
with the radiator.


1 Like


One more image which I wasn’t allowed to post in first message due to restrictions for new accounts.

1 Like

Yup. Why a buffer is needed rather than using unbuffered distilled.

Loop cleaner. Mayhems Blitz is pretty much the go to here but there’s others like Alphacool WAZ and Aquacomputer ArctiClean. Thing’s to keep the chemistry compatible and mind the runtime and flushing requirements. DIY’s also an option but I think might be more than you want to get into.

3 Likes

Could also be algae or bacterial growth, distilled water by itself is not sufficient. Manufacturing always leaves flux and residue in radiators, and in the case of some chinese factories they don’t even use clean water when flushing them. Contaminated water has led to a great many CLC recalls after the coolant gunked up. But when watercooling, the constant heat provides an energy source so it takes very little of anything else for stuff to start growing even in a blacked out case with black tubing.

Given you are mixing copper, brass, and nickel some corrosion inhibitor would probably be good. But you absolutely must include a biocide or algaecide in your next coolant. I use a pond algaecide and that’s worked great for my loops over the years. I recommend against silver as that only adds an even more reactive metal to the loop.

I’d second what Lemma said, you’ll need some brand of nuke or cleaner to kill off anything growing in the rad/parts you won’t take apart, then flush everything before starting over with new coolant.

3 Likes

Defiantly need some form of anti-growth of any nature in your water, its nearly impossible to have a perfectly clean loop setting up, so distilled water will not stop algae and bacteria growth. Your probably going to have to dissaseble the loop and clean the blocks and refill(with something containing anti growth).

Good point. As of the most recent MSDS I can find (2016), DP Ultra’s 20-35% ethylene glycol (and <1% benzotriazole). So, depending on the mix you get, it might not even hit the 25% ethylene that’s commonly taken as a biostatic threshold.

1 Like

This is why it is a bad idea to use just distilled water as a coolant.

I’m, not 100% certain what that is, but the truth is that whenever water cooling you need a coolant with a growth inhibitor, a corrosion inhibitor and it is usually a good idea to have a surfactant as well.

Even a thoroughly cleaned loop with pure distilled water will eventually experience microbial growth. Not many microbes will make their way in, and there won’t be much for them to feed on, but over time it will build up.

Way back in the day people used to use distilled water with just a drop of copper sulfate solution (like PT Nuke). Since then we have learned that this tends to react with and corrode metals.

Then many switched to pure distilled water with just a silver “kill coil” in it, with the silver giving off silver ions that tended to be somewhat effective as a growth inhibitor, but that has some problems as well.

Even if you closely match the metals in a loop, you will never have an absolute 0V galvanic series. There will always be a small voltage difference that can cause corrosion. Also distilled water will actually rip ions off of the metals it contacts over time. So you really want something with a corrosion inhibitor in there too.

A surfactant (to eliminate the surface tension of the water) is very helpful too when you fill the loop to get rid of water bubbles and ensure better liquid contact with the micro-channels in your blocks. This is totally optional though.

Now, as for what to use? That’s where there are as many opinions as there are assholes.

When I built a new loop back in 2016 EK’s nickel plating corrosion/delamination issue was fresh in my memory. I really wanted to avoid EK all together as a result, but the only fullcover block I could find in stock for my Pascal Titan at the time was EK and nickel plated, so I decided that if EK was finger pointing at everyone and everything after that issue, my best choice would be to use their coolant, because then they can only blame themselves.

They sold something branded as Ekoolant at the time, which is what I used. And then when their Cryofuel came out I bought the concentrate and used that.

It worked well. My loops stayed clean and blocks came out years later looking squeaky clean with no corrosion.

The only downside was that it was more viscous than water, so it slowed flow rates a little bit.

Now with EK collapsing again, for my new build I am currently working on, I need something else.

I recently did a test build that is currently cooling my workstation. I wound up using Primochill’s concentrated Liquid Utopoia both because it was cheap, because I had it handy, and because reviews were decent. It appears to work well. It performs well and gets good flow rates, but I haven’t had it in long enough to know if it has any corrosion or growth issues.

I also have several bottles of XSPC Pure coolant. This appears closer to EK’s Cryofuel. I’m not sure if I want to use this or Liquid utopia in my new build. Time will tell.

The TLDR version is this:

1.) Absolutely do clean your components before use. Even if it is just a thorough flush. There is usually crap in the parts from the factory, and you don’t want that crap in your pumps and microfins in your blocks.

2.) Absolutely use something with a growth inhibitor, or your loop will gunk up. It won’t even take that long. yes, even if you thoroughly clean and use distilled water and try your best to not “infect” anything.

3.) Absolutely use something with a corrosion inhibitor, or over time (though usually more than two years) you will see some corrosion. Even if all of your metals are the same

4.) Ty to avoid “showy” coolants. Clear undyed coolants are the best, colored coolants are OK (but they usually put so much damn dye in there that it stains over time). Shimmery or opaque coolants all eventually gunk up and are best avoided.

As for what to do now?

You have gunk in there. It looks like growth. You are going to want to get it out, and it is not going to be easy. You are at the very least going to want to do several flush cycles through the loop, filling, flushing, draining, over and over again. If you are lucky this will get everything out.

You may want to alternate flushes with slight amounts of dish soap (like dawn) and diluted household vinegar to try to clear out the gunk. Only use a drop, maybe two of dish soap, or it will foam like crazy, and take several flushes until you get it all out.

Same with vinegar. Don’t use too strong vinegar (dilute it to like 1/4 or 1/5) and don’t leave it in there too long, as vinegar is an acid, and will start eating into your metal parts if left unchecked.

If you use any cleaner at all in there, flush and drain several times with distilled water after to make sure you get it all out, before refilling.

Unfortunately based on my experience, once you get gunk there is no way around needing to open parts up and scrub them out. I’m talking at the very least the blocks. Unless they have completely clear tops and and you can see the fins and confirm they are not gunked, you are going to want to take the tops off, and scrub the fins with a (clean new) toothbrush and some sort of cleaner. Vinegar can work (with the same corrosion warning as above), dish soap can work, ispropyl alcohol can work (but don’t get this on any acrylic parts or rubber seals, as it can cause them to crack)

Then lube the o-rings/seals, re-assemble the blocks, and put everything back together again.

Then re-assemble everything again. If you have a pressure tester now is a good time to test the blocks you opened and re-assembled. If not, run the loop for several hours without powering on the PC and monitor for leaks.

As for what to use next time? Try Primochill’s Liquid Utopia. It’s a little 15ml bottle that supposedly is good for up to a gallon of coolant, and it is low cost, small bottle and seems to work really well.

3 Likes

My guess is the nickel plating failed on one or more of your parts. Take a look at that inside of whatever you can see. I’ve had cheap nickel plating fail before, just not quite so spectacularly.

I’ve been using plain distilled water just like you have for the last 15-20 years without any major problems, so all this talk about biocides and inhibitor-whatevers is all well and good but not necessary in my experience.

One computer that I bought came with an AIO and that thing packed itself to the brim with corrosion until the pump failed, and I’m sure that thing was slam full of additives. So you’re correct to be wary of galvanic corrosion.

1 Like

Experts on the topic disagree.

Here is a Hardforum members heatkiller block that was run with just distilled water:

Here is mine, run for a similar amount of time but with a coolant (EK Cryofuel) that containes biocides and corrosion inhibitors:

And I was really sloppy about maintaining mine. You are supposed to replace the fluid every couple of years at least, and I went for like 5 years without flushing it.

3 Likes

^^ typo fixed. Or Mayhems Inhibitor+ and Hades+. Also Koolance 702 and 705, Alphacool Liquid Eco, and Aquatuning AT-Protect, which are premix. All six lack Utopia’s aluminum restriction (not that it matters here). Don’t think I’ve ever seen a post from someone using AT-Protect, though, so I’d be a bit cautious there.

Yeah. I see 50% more often than anything else, which is pretty high.

Depends on the cleaner. Alphacool’s instructions are single flush, for example.

Don’t have to. If the loop’s in good shape just topping up the inhibitor every year or two should be sufficient if it’s one which calls for that.

With reasonable ethylene glycol ratios I think multi-year zero should be feasible if the setup’s good. aBav did eight, for example, though the Heatkiller’s end state isn’t shown.

1 Like

Maybe I’ve been using a different kind of water other then distilled and used that label incorrectly, otherwise I’m not sure how your experience has been so different than mine.

It’d probably take an inventory of all three loops’ components, prep and setup, use, and locations. Maybe cutting open the rads for a look inside. But if distilled’s sold like it is here accidentally buying something else is pretty much impossible.

One corrosion factor’s the narrower the fin pitch on the blocks the more likely it seems to be nickel plating won’t be complete, but there’s not much data for decent blocks. For biocide the unsurprising pattern I’ve noticed in folks’ experience is the more careful the setup the less likely it is to be needed.

1 Like

That’s a good reminder; should have one of the original heatkiller cooling blocks still lying around, “uncleaned” since I swapped the motherhoard socket and the old block isn’t that great on AMx CPUs with chiplets.

Recently placed the order for 3 MoRa IV 600 radiators for the planned multi-room setup, unfortunately Watercool has a bit of issues with the actual shipping availability of the new series.

2 Likes

if you can get a pressure washer at low setting that can work well to blow shit out of rads, I do that to all my radiators and a lot of shit tends to come out when when you buy then new, also tried distilled water first time, and it was a disaster, everything went yellow, and sludge clogging the blocks, needed to do full tear down, and blast every block out, with a pressure washer… full day wasted on a 10 dollar bottleof proper coolant…

Be VERY careful with pressure. Some radiators specifically built for computer cooling can’t even handle 1 bar of pressure without permanently deforming which can lead to leaky fittings due to damaged threads.

(found that out myself with air pressure while looking for potential leaks)

1 Like

How to make battery acid in a lead acid battery.

Add distilled water.
(+)
Run current through it.
(+)
Distilled water becomes acidic.

If you only run distilled water in a cooling loop, with different metals, you’ve got both the potential for electrolysis and corrosion.

Please use a corrosion inhibitor/coolant concentrate to your distilled water loop. Use clear or clear UV style if need be.

Magnesium anode, opaque tubing and a cyclone filter . If you have clear tubing add a biocide. Note adding silver metal will cause corrosion to copper, nickel, aluminium , steel parts.

There be your problem.

Biocide is absolutely required with distilled water. A couple of drops is all it takes.

This topic was automatically closed 273 days after the last reply. New replies are no longer allowed.