Stainless steel house water filter, leaks past o-ring

[Spinnor]

Bought a stainless steel whole house water filter for a very good price ($100), turns out too good. On inspection I suspected it would leak so rather then installing it into the house water system I pressure tested it. There was a small leak that stopped after a minute or so. Relieve the pressure and repressureize and the leak came back and stopped as before after a minute or so. I am guessing the o-ring slowly deforms to fill the void in the shell. An image of the unit follows,

The tall can shaped part has a vertical seam weld on the inside that was ground smooth where the o-ring touches but done with a grinding stone with a significantly smaller radius then the inner radius of the shell. I am pretty sure that is where the leak is coming from.

The store I bought it from are willing to take my unit back and replace it with one that hopefully does not leak but I am sure there is a cleaver fix that involves some type of epoxy. The following is how I thought of fixing it where the inside weld of the shell was ground down a bit too much I would apply just the right amount of epoxy, not too much not too little. Then I would cover the epoxy with a piece of rather thickish clear tape and smooth out the epoxy with the straight edge of a credit card. If the tape didn't work the same idea but using a small thin piece of oiled aluminum might do the trick.

I will probably end up shipping it back but if you had to make the repair with a limited tool set what might you do?

Any suggestions for the type of epoxy to use? We have slightly acidic water.

Thanks for any thoughts.

 

[BillTre]

The epoxy seems kind of irreversible if it doesn't work out right.

What kind of rubber is your O-ring?
If it is fairly compressible, I would try an similar sized O-ring with a slightly large cross-section.

Alternatively, I would try putting silicon grease on it before inserting the O-ring. Silicon grease is often used with certain kinds of O-rings.

Or I would try using silicon aquarium sealer before inserting the O-ring. I would then clamp it in and give it time to set-up before pressurizing. Using aquarium sealer as opposed to other silicons will avoid the preservatives and anti-fungal compounds in many kinds of silicon sealers. If it does not work, the silicon should be relatively easy to remove.

 

[marcusl]

Whatever compound you put in there will eventually fail and you (or the next owner of your house) will have a problem. Send it back for one that works right at the start.

 

[Spinnor]

What kind of rubber is your O-ring?
If it is fairly compressible, I would try an similar sized O-ring with a slightly large cross-section.

Don't know the composition only that it is black. I did grease up the o-ring before the assembly. The o-ring fits its groove pretty tightly so the next larger diameter o-ring is probably not an option.

Thank you.

 

[Spinnor]

Whatever compound you put in there will eventually fail and you (or the next owner of your house) will have a problem.

You are probably right. Interesting material science question why such bonds work at first and over time fail.

Thank you.

 

[Rive]

Epoxy, silicon grease, silicon sealer... This thing would be installed into the drink water system, right?
I too would prefer just replacing it. That's guaranteed clean. At least, supposed to be.

Ps.: just one of my old mate comes to my mind when he was so proud about himself that he found such a good solder, easy to work with as an amateur for his pipes in the kitchen (copper stuff)... Guess what it was?

 

[256bits]

Let the water run a bit and it should have been fine for the old mate.

 

[256bits]

You are probably right. Interesting material science question why such bonds work at first and over time fail.

Thank you.

Temperature expansion between dissimilar materials is one reason.

I did the epoxy thing for an old Jetta that broke a fitting for the water system ( well I broke it getting it off to get at the thermostat ). It lasted a while, but did begin to leak slowly after about 3 months.

 

[Rive]

We had no qualification to tinker with Pb (Pb containing solder) in a (drink)water system. The piping were replaced by a professional who could do the soldering properly with the right type of solder.

 

[Spinnor]

Guess what it was?

Temperature expansion between dissimilar materials is one reason.

I did the epoxy thing for an old Jetta that broke a fitting for the water system ( well I broke it getting it off to get at the thermostat ). It lasted a while, but did begin to leak slowly after about 3 months.

How about wax as a filler? I think it might just have enough compressive strength to withstand deformation under 50 psi water pressure squeezing the o-ring against its surface and it does not seem like the kind of material to chip off of a surface it is adhered to, maybe because it is relatively soft? Mind you I am just curious what fix might work short or even long term using simple tools. I can see epoxy failing over time but wonder if some other material might work, melted plastic maybe? Maybe the epoxy fails because it is too hard, hello material scientists out there?

I guess I should just test the wax idea, the store has yet to return my email regarding the return of the defective unit.

Thank you.

 

[Spinnor]

Wax repair has been holding for 5 hours. I wonder if the wax will slowly deform under pressure over time?

 

[Asymptotic]

Does the filter or packaging have a manufacturing date?

In my experience, even when the filter housing is correctly manufactured, the o-ring permanently flattens after 2 to 3 years of assembly and must be replaced.

 

[Spinnor]

The housing just has the number 304 which I guess is a type of stainless steel. The o-ring was shipped separately so it was virgin material, I should have ordered extra of them but was not thinking. The issue appears to be a slight void on the inside shell of the cylindrical housing where a weld was improperly ground down with a rather small radius grinding stone. My wax addition into the void solved the problem and it has held pressure since yesterday but I do not know the long term compressive properties of wax, will it "flow" under pressure over time or will the bond between the wax and steel change over time. Still waiting for an email from the seller. There must be a whole field of science that deals with the bonding of one material to another and how to make that bonding better.

Thank you.

 

[BillTre]

There are a huge variety of different waxes with differing properties of hardness vs. temperature. Some waxes stay soft and malleable, others get hard and brittle.

Silicon sealer, on the other hand, will stay soft and malleable for long periods of time.

 

[Spinnor]

Silicon sealer, on the other hand, will stay soft and malleable for long periods of time.

Is that the garden variety silicone that one might use to seal the seam of a bathtub?

Thank you.

 

[BillTre]

Is that the garden variety silicone that one might use to seal the seam of a bathtub?

Not really. It would be either garden variety aquarium sealer, or "pure" silicon sealer with no preservatives (such as anti-fungal chemicals). Most silicon sealers for home use have additional chemicals you might not want to consume.

Similarly, chemicals can be added to other materials (greases or rubbers) you would probably not want to have in contact with your drinking water.
I have used a bioassay to screen through lots of materials for those that are safe for use in fish water systems.

 

[Tom.G]

Looking at an enlargement of the photo, that o-ring certainly does not look new. There are small-scale pits and grooves in it and embedded particles (probably sand). Make sure too, that the o-ring groove (seat) is free of debris and fairly smooth on all three surfaces.

If the seller does not come thru with a unit in better condition, I suggest replacing the o-ring with a new one. Contact the filter manufacturer if possible to either purchase a few or find out what the o-ring material is.

If no luck that way, UHMW (ultra-high molecular weight polyethylene) shows pretty good acid resistance. However the hardness of the o-ring material is also important, so that might not be the best choice.

Moss Rubber has a downloadable Chemical Resistance Chart at
www.mossrubber.com/pdfs/Chem_Res.pdf
A Google search turns up some additional charts.

Cheers,
Tom

 

[Asymptotic]

Looking at an enlargement of the photo, that o-ring certainly does not look new. There are small-scale pits and grooves in it and embedded particles (probably sand). Make sure too, that the o-ring groove (seat) is free of debris and fairly smooth on all three surfaces.

If the seller does not come thru with a unit in better condition, I suggest replacing the o-ring with a new one. Contact the filter manufacturer if possible to either purchase a few or find out what the o-ring material is.

If no luck that way, UHMW (ultra-high molecular weight polyethylene) shows pretty good acid resistance. However the hardness of the o-ring material is also important, so that might not be the best choice.

Moss Rubber has a downloadable Chemical Resistance Chart at
www.mossrubber.com/pdfs/Chem_Res.pdf
A Google search turns up some additional charts.

Cheers,
Tom

If you refer to the photo in post #12, it doesn't represent @Spinnor's filter, but one in an industrial application I had serviced. It had begun to leak due to o-ring aging and debris inclusions, and was presented to illustrate the tell-tale "flattening" effect.

Good point about cleaning the o-ring seat. In conjunction, clean the filter bowl where it comes in contact with the o-ring as debris often settles out there. Also, clean the male housing, and female tightening ring threads, and lightly wet them with a silicone grease (along the lines of Dow Corning #111) before reassembly.

By way of anecdote, my duties were primarily industrial electronics, but one day I observed a fellow maintenance tech whaling away at the tightening ring on one of these because "it wouldn't seal". It took hammering to remove the ring, too. Best I can reckon is the o-ring had been going bad for a while, and they were in the practice of over-tightening it to create a metal-to-metal seal between the top of the bowl and housing body (steel along the upper circumference of the bowl had "nippled"). After replacing the o-ring, and cleaning the bowl and threads the leak was gone, the ring fairly spun on, and could be tightened and loosened by hand.

For the problem @Spinnor has (provided the seller doesn't come through with a replacement) filling in the weld seam divot with a suitable silicone sealer as suggested elsewhere in the thread sounds good.