Pin hole leaks in copper pipes

bnolanApril 3, 2008

My home is 3.5 years old. Recently myself and my neighbor (his house is 6 months older) began experiencing pin-hole leaks in copper elbows. We had the same builder, plumber, electrician for both homes. Copper elbows tested by manufacturer who blames poor installation (too much flux). Plumber blames electrician claiming that system wasn't properly grounded. We have each had about 8-10 leaks. Both started near hot water heater. We both have recirculating pumps (we switched them out from cast iron to stainless steel). I don't trust this system and think plumber should re-plumb. Don't know where to turn with everyone pointing fingers. Any thoughts? Thanks

Thank you for reporting this comment. Undo

its a moot point if your not on well water, but I had similar problems with getting pin hole leaks in my copper pipes at my last house. I'm sorry I can't recall the exact cause but it had to do with the content of my well water.

    Bookmark   April 3, 2008 at 4:20PM
Thank you for reporting this comment. Undo

The cause or causes of pinhole leaks in copper plumbing systems are not fully understood at this point in time. While 80% of pinhole problems have been associated with older copper installations, new installations, such as yours, can have problems too. Engineers tend to believe one cause is due to excessive water velocity in water pipes which may erode or work on weak points in the materials. Pinhole leaks are common in recirculation piping and fittings. A pinhole leak can develope in a span of copper where there are no joints. A maximum water velocity of 4 FPS (Feet Per Second) is suggested thou 7-8 FPS is common. Your leaks are at the elbows, I would assume the water velocity is high. Another suspected cause is high PH levels in the water and another suspect is soft water. Other thoughts are simply manufacturing methods or the composition of the metal itself. One thing is for sure, no one will admit to wrong doing until the causes are clarified.

If you supply the pump data including GPM and the pipe size (3/4" assumed) on the recirc line, I could determine the water velocity.


Dan Martyn

    Bookmark   April 3, 2008 at 7:17PM
Thank you for reporting this comment. Undo

If you are experiencing pin-hole leaks in copper plumbing, the number one cause is low pH. What is your pH? If it is below 5.5 then copper will suffer. I never heard of velocity as a cause for copper thinning.

An acid neutralizer would need to be added to any existing equipment. Get it checked.

Andy Christensen, CWS-II

    Bookmark   April 3, 2008 at 8:31PM
Thank you for reporting this comment. Undo

How do I go about getting the pH level? Do I need to buy a kit, hire an engineer or something else? I have looked at the water quality report online from my utility district (, but not sure what I'm looking for. As for the recirc. pump I'm not sure about the type and information from it, but will post it tomorrow. Thanks for all of your input.

    Bookmark   April 7, 2008 at 10:18AM
Thank you for reporting this comment. Undo

its probably cavitation from high water velocity.

    Bookmark   April 7, 2008 at 10:40AM
Thank you for reporting this comment. Undo

Pinhole leaks in copper pipes serviced by public utilities is a nationwide problem - Virginia Tech is / was studying the problem in conjunction with WSSC (Washington Suburban Sanitary Commission). Google pinhole leaks or any variation and you will get more info then you ever want. Two of the things we have done: learned how to repair the simple leaks & those repairs are done using the heavier copper pipes used for well water. We lived 15 years with water and never had a problem - moved 2 miles to a house on public water and we deal with the continual pinhole leaks. The leaks are self sealing for a while. If you suspect a pinhole - DO NOT scratch the greenish - crusty like area unless you are prepare to fix it immediately.

    Bookmark   April 7, 2008 at 2:02PM
Thank you for reporting this comment. Undo

Just curious - are the elbows green and crusty from oxidation? I "heard" that would indicate the joint wasn't wiped to remove excess flux after sweating.

    Bookmark   April 8, 2008 at 12:28PM
Thank you for reporting this comment. Undo

Cavitation is the result of attempting to pump or pressurize beyond that static pressure available on the inlet. If this was the case, and the pumps were way to big, the result would be destroyed impellers on your recirc pumps.

Although PH also plays a big part with copper (thus RO units are piped with plastic tubing), if the pumps/circs are oversized the velocity will wear the inner copper, especially the turns like stones in my kids rock tumbler.

    Bookmark   April 8, 2008 at 5:52PM
Thank you for reporting this comment. Undo

One of the most common cause of pinhole leaks is acidic water however, if you get your water from a municipal water source you can rule that problem out.

Under the provisions of the "Federal Safe Drinking Water Act" as passed in 1976 and amended in 1986, 91 + 96, all municipal water suppliers are required to continually monitor all the water they supply and to treat it as necessary to insure all water maintains a neutral PH.

It is true that some pinholes are a result of pipe wall erosion from excessive water velocity, but to even think that a recirculation pump is responsible for excessive velocity borders on ridiculous. Recirculation pumps are extremely small pumps generally ranging in the `1/10HP - 1/24HP range and they are rated at approximately 1gpm into a 10 -24' vertical static head whereas, Pipe wall erosion does not occur unless the velocity of flow exceeds 8-10fps. Typically the return line on a recirc loop is 1/2" pipe and at 1gpm in 1/2" pipe the velocity of flow is only 1.6fps. Even if the line was downsized to 1/4", at 1gpm the velocity would only be 2.91fps, well below the erosion threshold.

Many people see the pale green corrosion on copper pipe and immediately get concerned that the pipe is going to fail Such is not the case. The greenish color that you see on the copper is "patina" and although most people seeing it on copper pipes are put off by it, in the architecture or art world it is highly desired, because the patina is a natural protective layer on the surface of the copper. In fact, when they rebuilt the statue of liberty the new copper plating was intentionally treated with a mild acid, the same as used in flux, to cause the patina to form. That is what gives the statue of liberty her beautiful green color. You can also see it on the copper roofing of colonial buildings.

I would inspect the joints in question very closely looking for a black residue or a dull, mottled reddish appearance on the copper. That is a sign of the most common mistake made by novice plumbers and DIY'ers, excessive heat while soldering.

Standard lead free 95/5tin-antimony solder(ASTM alloy SB5) melts at 452-464degF. (Some manufacturers are now making alloy SB5 solder with melting points as low as 410degF.)

In the plumbing trade for special applications we also use ASTM alloy E (440-500degF) and Alloy HB (460-630degF).

Standard petroleum based flux begins to bubble at about 600degF and it melts at 700degF. If the joint is overheated the flux melts, then burns leaving a thick carbon residue. The solder gap between the inner wall of a fitting and the outer wall of the pipe increases slightly as the size of the pipe increases, ranging from 0.001" clearance for 1/2" pipe up to 0.010" for 12" copper pipe. If the joint is overheated the clearance between the fitting inner wall and the pipe exterior will be filled with burnt flux, thus preventing solder from properly flowing into the joint, then as time goes by the water in the pipe will erode away at the carbon and you are left with a pinhole leak at the joint.

    Bookmark   April 9, 2008 at 1:06AM
Thank you for reporting this comment. Undo

"It is true that some pinholes are a result of pipe wall erosion from excessive water velocity, but to even think that a recirculation pump is responsible for excessive velocity borders on ridiculous. "

back in the 80s i owned a fleet of 18 wheel trucks. a problem popped up where we would get pinholes in the cylinder sleeves after only about a year. these sleeves are 1/2 inch thick cast iron.
it turned that the holes were from cavitation. it is not the water velocity that causes it as much as the tiny air bubbles caused by the pump. as these tiny bubbles hit a surface such as an elbow they pop causing a tiny explosion that eventually eats away at the metal. the fix in this case was special additives in the water but that is not practical in a house situation.
i cant be sure this is what is causing these pinhole leaks in pipe elbows but it sure sounds similar.,_scas_and_the_proper_maintenance_of_diesel_engine_cooling_systems.htm

    Bookmark   April 9, 2008 at 10:36AM
Thank you for reporting this comment. Undo

okay Pup, guilty as charged on the reddish joints, once in a while, not all, next time I'll back off with the torch, in my post above, I had accessive ph level, was on well water at that time, pin holes were in straight runs, vertical and horizontal, never had one at a joint, as far as too much volocity, that's over my head. Do appricate the tip on too much heat, thanks

    Bookmark   April 9, 2008 at 5:03PM
Thank you for reporting this comment. Undo

Quote:"back in the 80s i owned a fleet of 18 wheel trucks. a problem popped up where we would get pinholes in the cylinder sleeves after only about a year. these sleeves are 1/2 inch thick cast iron.
it turned that the holes were from cavitation. it is not the water velocity that causes it as much as the tiny air bubbles caused by the pump. as these tiny bubbles hit a surface such as an elbow they pop causing a tiny explosion that eventually eats away at the metal"

Not only does that explanation totally defy the laws of physics, it also defies the laws of logic and plain common sense. First of all, ask yourself this question, where does the pump get a source of air to create those supposed air bubbles? Now think about a bubble for a moment. A bubble is a pocket of air surrounded by a liquid. In order for a bubble to "pop" it must first rise to the surface of the liquid where the entrapped air can be released into the atmosphere or an air pocket above the liquid in the container. It simply is a total impossibility for a bubble to "pop" while still under the surface of a liquid because there is no place for the air to go other than to maintain itself as a bubble.

Cavitation is not caused by air but rather it is caused by steam, and it is a much more violent process then simple air bubbles.

Most people believe that water boils at 212degF but that is not true, or rather, that is an incomplete statement. Properly is should say that water boils at 212degF @ standard atmospheric pressure at sea level. If the pressure is then increased or decreased there is a proportional change in the boiling temperature. By example, in a perfect vacuum water boils at 32degF and at 250psig water boils at 406degF.

When water flashes to steam the reaction is instantaneous as the water expands to 1x12 to the 3rd power, or in simple terms when water expands to steam the steam occupies 1728 times the original volume of the water. In an automotive cooling system the steel cylinder liners develop microscopic heat cracks and if as little as a molecule of water gets into one of those cracks and flashes to steam it has the same effect as if the crack had been packed with explosives. If allowed to continue it would quickly erode through the metal. In order to prevent this from happening years ago all automotive engines were equipped with thermostats that maintained the coolant temperature at 180degF however in recent years that auto industry has managed to increase the output horsepower of engines by increasing the operating temperature. The first step in increasing the operating temperature was to install radiator "pressure caps" to increase the operating pressure of the water, thereby increasing the boiling temperature at the same time. The next step was to replace standard anti-freeze with a summer winter anti-freeze/coolant solution that chemically alters the boiling temperature of the water. The anti-freeze/coolant solution also has a chemical that coats the surface of the hot inner parts of the engine to prevent water molecules from getting into the microscopic pores of the metal.

In a pump, if the pump is discharging water at a faster rate that what water is entering the pump the pump intake (volute pressure) drops. If the internal pressure in the pump drops below the boiling point of the water at the intake temperature some of the water will immediately flash to steam in the pump impellor cavity. The impellor of a pump cannot pump steam, therefore it is left spinning in a void and any free droplets of water that may be in that cavity will then fly around and hit the impellor vanes with the force of a shot blaster. If left unchecked this condition can severely pit the impellor surface or even totally destroy a pump impellor in a matter of minutes. When this occurs the simple solution is to turn the pump off and allow it to cool, thus condensing the steam and refilling the pump cavity with water.

Cavitation is a common problem on condensate return pumps in a boiler room. In this case we dare not turn the pump off because maintaining water level in a boiler is critical. The quick solution here is to grab a water hose and run cold water over the pump housing to cool it until the steam condenses and the pump is pumping again.

While these are immediate solution, they should also be considered a temporary solution. To achieve a permanent solution we must either lower the physical location of the pump, thus increasing the "Static Head" at the intake of the pump or we can increase the diameter of the feedline and intake to the pump to increase the volume of flow into the pump.

In the case of a re-circulating pump such as used on a domestic hot water system cavatation is simply not possible. A re-circulating pump is operating on a closed loop system, therefore as water is expelled from the pump it forces an equal amount of water on the opposite end of the loop to enter the pump, thus a re-circulation pump is not really a pump but rather a mechanical stirring device.

    Bookmark   April 11, 2008 at 12:12AM
Thank you for reporting this comment. Undo

There have been some interesting thoughts and opinions, but I have go go with turbulence and/or cavitation. I am just a sad plumber, but the engineers I work with and the ones who spec the jobs I work on do worry about excessive velocity and there is always the possibility that the problem is pipe that has not been deburred... THAT, from the Copper Development Association.

    Bookmark   April 15, 2008 at 3:46AM
Thank you for reporting this comment. Undo

True cavitation cannot occur in a pipe and assuming that the pipes were properly sized for the anticipated maximum volume of flow turbulence in the fittings is not a problem.

    Bookmark   April 15, 2008 at 2:16PM
Thank you for reporting this comment. Undo


Thanks for your technical explanations in this thread. They make sense to me.

I have an old, but modernized, Victorian house that has three 15 ft high floors and a basement, making the bottom of the water heater in the basement about 50-55 feet below the three bathrooms that are on the top floor (the house has 9 bathrooms and about 8000 square feet of living space).

The house has a large Grundfos recirculation pump/system. While this recirc system works great to provide nearly instant hot water to the many bathrooms, it has been blamed by the previous owner and plumbers for several hot water pipe leaks that have developed in years 8-12 since the house was completely re-plumbed with new copper pipes 13 years ago.

The recirc system return line from the third floor is 1/2 inch copper pipe to the Grundfos pump which is located near the water heater in the basement and feeds into the bottom of a 100 gallon water heater. The hot water feed line from this water heater to the house was a 1.5 inch copper line that runs vertically through the house, but is being downsized to a 1.25 inch copper line in the basement with a new water heater installation. Both the inlet and the outlet of the Grundfos pump itself are 1/2 copper pipe.

My question is this: do you think I should downsize the Grundfos recirc pump to reduce the flow rate in the recirc system? After several of the copper pipe leaks were repaired by the previous owner, the recirc system was put on a timer so that it would not run continuously, as the theory was that the water flow was eroding the copper pipes. The current Grundfos pump model is a UP26-96BF.

Thanks for any insight that you can provide. I don't really want to be chasing copper pipe leaks for the next 20 years.

    Bookmark   June 3, 2008 at 6:01AM
Thank you for reporting this comment. Undo

To add to all this are issues of corrosion.
Copper pipes build up a surface film that protects the copper from dissolving in the water.
This normally appears as a black or brownish (even yellow) mottling on the inside of the lines.

The pinhole leak problem greatly accelerated when the EPA issued new (lower) standards for total dissolved solids in water.
A number of water sources implemented the standards early as part of regular upgrades.
Following this from WSSC (Washington Suburban Sanitary Commission) pin hole leaks started appearing in relatively new installations.
Lead leaching from older lead pipes also increased in Washington, DC around the same time.
The 'fix' was to introduce orthophosphates into the water.

This allowed the formation of the protective layer in both copper and lead piping and appears to have decreased the incidence of pin hole leaks.

As alluded to above, very clean water is rather aggressive in dissolving things, like lead and copper pipes.
Anyone with a marine aquarium becomes concerned with copper content of the water very quickly.
It is lethal to invertebrates and not very kind to many other marine animals.
I used to supply water to a friend who had a new house from my 1950s era copper pipes to avoid the dissolved copper.
After about 5-6 years his water was 'clean' enough.

The very clean water used in semiconductor fabrication is moved around in plastic tubing with a metal flash coat to prevent oxygen from entering. it is so aggressive it can damage even high grade stainless steels.

While pH and high velocity can cause erosion of copper (and other materials) they are relatively rare occurrences in residential plumbing.

Check and see if the water supplier has altered the treatment method recently.

    Bookmark   June 3, 2008 at 8:04PM
Thank you for reporting this comment. Undo

The notion that the circulating pump is contributing to pipe wall erosion is totally ridiculous.

In copper pipe wall erosion occurs when the velocity of flow is in excess of 10feet per second and in some plastic pipes pipe wall erosion can occur when the velocity of flow exceeds 16 feet per second. In response to this both the International Residential Code and the Uniform Plumbing Code require us to size the pipes to insure that the velocity of flow under full load will not exceed 8FPS in copper or 12FPS in a plastic pipe.

Typically the pumps used for circulators are extremely small fractional horsepower pumps generally in the 1/10 to 12HP range, an in particular the Grundfos VP26-96BF mentioned above is a 1/12hp pump with a maximum flow rate of 1gpm into a 24ft head.

Consulting the copper pipe friction loss tables we find that a 1gpm flow rate will produce the following velocity of flow.

1gpm into a " pipe = 1.39ft/sec
1gpm into a ¾" pipe = 0.66ft/sec
1gpm into a 1" pipe = 0.39ft/sec
1gpm into a 1-1/4" pipe = 0.25ft/sec
1gpm into a 1-1/2" pipe = 0.18ft/sec

The bottom line, the circulator pump in question is simply not capable of creating enough velocity of flow to cause pipe wall erosion in copper pipe, on the other hand, when homeowners and plumbing wannabeeÂs arbitrarily add additional fixtures on an existing line without taking the velocity of flow into consideration there is a very high risk of pipe wall erosion.

    Bookmark   June 3, 2008 at 9:47PM
Thank you for reporting this comment. Undo

This thread has (rightly so) got me worried about my own pipe situation, I am in British Columbia and have a pressure regulator (which is a bonus I guess) but therefore also have a backflow preventer, again good, but no expansion tank - I first realised it might be an issue when I noticed lots of leaks with regards to my humidifier (especially with those awful saddle/pipe-piercing valve which I have chucked out in favour of a ball valve) and then noticing warm water coming from the cold tap.

I have no expansion tank and now understand it's a no-no but due to lack of finances and other issues (I want to do some plumbing reorganisation which will make putting the expansion tank in a big job, I want to put a whole-house water filter in too) I haven't acted on it yet and this thread has me thinking at the least I should get one of those toilet cistern filler valves that's made for the purpose, as I understand it, it allows the expansion to go to waste in the toilet tank, I hate wasting water but better that than a burst pipe not to mention it's bad for the t&p valve...

In the meantime I solved the humidifier issue with a ball valve and figure it's actually acting much like the toilet valve I described, as it uses a very small float so any pressure in the water line probably quite easily forces water through the humidifier inlet, and there's an overflow so it's all good and keeps the water fresh. I also replaced the cheap plastic water line with copper, I bought a fairly large coil of the bendable soft copper and I used the whole coil (leaving the bulk of it coiled in one spot) so it works something like a spring, hopefully relieving some of the pressure in the meantime.

Apropos of these recirculating pumps, I can see with several stories it might be an issue, but was thinking about what my ex-wife's grandfather, and old-school british plumber (ee buh goom) suggested for hot services, he said you should always run a line from the furthest hot tap back to the tank. I always assumed you would tee to the hot water outlet as I'd never seen provision for a 'return' but think I've seen where you could run a warm return on some north american tanks (correct me if I'm wrong)

If the hot pipes were properly lagged (I was appalled to see mine weren't) I would think this would help somewhat, the idea of the complication (and inherent risks, from this thread) of an energy-wasting pump just so we get hot water quicker, seems somewhat wasteful. So is running the tap before you use the water, although from a health point of view, we're supposed to do just that, to flush contaminants from the pipes....

    Bookmark   June 4, 2008 at 12:06AM
Thank you for reporting this comment. Undo

I barely know the first thing about plumbing, but I do know that here in Southern Utah, a lot of folks are using PEX. Our inspectors recommend it over copper. We're getting ready to plumb our addition, and I'm putting PEX in it. Quick, easy, and hopefully no leaky.

    Bookmark   June 4, 2008 at 1:47AM
Thank you for reporting this comment. Undo

MattM,, I know PEX sounds alluring but I'm wary of plastics, been through alkathene in Australia and NZ in the '70's when it came into vogue, and went. I agree there are a lot of nice and alluring things about it and the idea of making joints without solder (especially in tight remodel spaces) is alluring, but there are two things I worry about - first, plastic characteristics change with age, and they tend to get brittle. Second, is the liquid version of off-gassing, or leaching. Think of polycarbonate. A miracle plastic and very useful, but, as it turns out, not good for water bottles due to leaching. No doubt pex is different but they will probably know what's wrong with it. Copper, at least, we know about.

    Bookmark   June 5, 2008 at 1:01AM
Thank you for reporting this comment. Undo

We had a construction project, remodeling bathroom of 22yr old house, old copper piping. Had to replace manifold, used Shark bite connectors. A week later copper piping 8" below on same pipe sprung pin hole leak. What could have caused this?

    Bookmark   July 1, 2008 at 11:42AM
Thank you for reporting this comment. Undo


A Grundfos UP26-99BF (Bronze) is also capable of 20 GPM at 10' head (closed system). Not knowing all youre piping characteristics, it is possible the oversized circ is partially or all to blame for your leaks.


Houses in the south have been plumbed with PEX for over 15 years.

PEX is not much different than the jug that holds the milk that you drink. Go for it and use PEX, it is proven!

FYI: PEX Systems have long warranties, copper tube or fittings do not.

    Bookmark   July 1, 2008 at 5:16PM
Thank you for reporting this comment. Undo

lazypup's in the ballpark but I think his long discourse on cavitation is more tailored to an extreme version you'd see in a hot water system.

Cavitation occurs at a microscopic level. It happens at points in the pump volute and on the impeller where low pressure occurs (the trailing edge of the impeller is the best example). If the suction head required by the pump to operate is greater than what's available due to suction line throttling or friction losses, the low pressure becomes low enough to cause water to vaporize. It's technically "steam" but not like you think b/c it can happen even at room temperature. This can be exacerbated by pumping a hot fluid b/c it has a higher vapor pressure (i.e. it 'boils' easier).

Your pump can be sitting there pumping water pretty much normally and still be cavitating. Lazypup's description makes it sound like you suddenly get a pump full of steam, but that's not the case. The bubbles that are formed by cavitation immediately collapse back on themselves and that's what causes the pump's parts to erode, since the bubbles occur at surfaces in the pump.

It sounds like you're pumping marbles.

Lazypup's description sounds more like a pump that's air binding. For hot water, it can be an extreme version of cavitation. But what he describes sounds more to me like a situation where the pump is depending on the pumped fluid for some of its lubrication/cooling. If the fluid isn't there, then the pump can quickly overheat.

    Bookmark   July 2, 2008 at 11:28AM
Thank you for reporting this comment. Undo

Just as a matter of info...The phenom of this problem in an engine's cooling system is know as cavitation erosion. It only occurs in diesel engines for the most part and is due to the particular vibrations/harmonics of that engine design. And don't worry it does not occur in all diesels. As stated it is easily corrected by the addition of a coolant treatment.

    Bookmark   July 7, 2008 at 10:34AM
Thank you for reporting this comment. Undo

Cavitation-erosion occurs when the fluid experiences high momentary peaks of sufficiently lowered pressure. Not necessarily near boiling points. It can occur at the bend in a pipe if the velocity is high enough (for a given viscosity of the fluid), and at the trailing end of rapidly moving objects (e.g., propellors, turbine blades). If the fluid is corrosive (relative to the given material; e.g., copper in acidic water) the cavitation-erosion effect will be considerably amplified. I have pinholes in my copper pipes also. It looks like "simple" corrosion to me?? All at straight pipe sections?
P.S. I teach this stuff at an engineering college. I love mysteries.

    Bookmark   August 6, 2011 at 2:56PM
Sign Up to comment
More Discussions
Does reverse osmosis water corrode copper plumbing? The answer ..
The answer is YES. I did a very unscientific study...
Culligan vs. Kinetico for very hard water
We just moved into a 3 year old house. The previous...
New under mount sink - drain off-center?
I had a new granite countertop installed with an under...
De-winterize, re-winterize...costs?
My 24 yr old son is buying a 1950 home on a short sale...
Waste pipe uneven elbow leak
The gaskets are removed to show the situation in the...
People viewed this after searching for:
© 2015 Houzz Inc. Houzz® The new way to design your home™