house loses pressure when main is shut off

[ewmyers]

Hi,

While trying to figure out how 2 people in a small house could use so much water, I discovered that the plumbing in my house loses pressure. If I have no water running in the house, I have 80 pounds of pressure. If I then shut off the main to the house, pressure drops to 40 pounds within a minute or two. None of my faucets drip, and the toilet doesn't run. There is an additional gate valve shutoff at the back of the house that leads to a spigot and 2 sprinkler valves, but there is no leaking there either. I can't find any leaks in the plumbing under the house, which is all copper.

Is there something other than a leak that could cause this, or is this normal? Or is there just definitely a leak that I can't find? How much water would have to be lost through leakage to produce such a drop in pressure? Drops? Teaspoons? Cups? Gallons?

Thanks, this has got me totally stumped.
Eric

[LazyPup]

The water pressure drop that you noted is normal and is a result of the simple physical laws of nature.

The water in your house is pressurized from the municipal water main pressure. When you turn the water main off you are isolating your house from the source of pressure. At that moment all the water in the pipes is in a sealed container (the piping system) until a faucett is opened which releases the stored pressure in the piping. Water is not compressible, however their is a small amount of air trapped in the water hammer arrestors that is compressed to equal the supply water pressure. When the faucett is opened that trapped air will begin to expand, pushing the water out of the faucett until the air pressure drops to standard atmospheric pressure and the air is fully expanded. From that point their will be static head pressure at the faucett which is created by the weight of any residual water in the lines above that faucett. The static head pressure will be approximatly 1/2psi for each vertical foot of water above the faucett until all the water in the pipes above the faucett drains out.

[ewmyers]

Thank you LazyPup. The thing is, I see the drop in pressure without opening any faucets. Also, the house is from the 20's, did they use water arrestors back then? Well, the house is from the 20's, but the copper piping is newer than that, I'm pretty sure.

Thanks,
Eric

[mrcaptainbob]

I would shut off as many valves in the house as I can. Start when it's at the 80 psi and start at the valve furthest from the main. shut them all off nice and snug. Then shut the main. See if the pressure still drops. Hopefully not. If not, then open the first valve from the main and check for a drop. Continue on and observe after each valve is opened. You may find the branch of pipe where the leak is at. By the way....are you on city water or well? That toilet not leaking....verify that. It could be a slow trickle so as to not notice. Even if the waer level's not at the overflow tube, it could slow leak past the flush stopper a minor amount to cause the float to let the water in. The toilet valves and sink valves should all be part of the valves that are shut off.

[LazyPup]

The use of copper pipe in plumbing systems did not gain acceptance until the early 1940's. Due to the wartime requirements of copper to make brass munitions casings and the extreme amount of copper used in the MANHATTAN PROJECT to produce the Atomic Bomb copper was not available to the civillian market. (Note that copper was is such short supply that the 1943 penny was made of steel.

Copper finally came into common use the in plumbing trade for commercial buildings and high end residential during the post WWII building boom in the late 1940's and early 50's but did not gain wide acceptance until the late 1950's.

If your house is vintage 1920 either the water system was replumbed once during the 1950's or later, or more likely if it was originally a rural structure it did not originally have indoor plumbing.

All plumbing systems regardless of what material they are made of require some form of water hammer arrector/expansion chambers to compensate for water expansion when the water is heated in the water heater.

Most water hammer arrestors are field manufactered by the plumber during installation and are usually concealed inside a wall cavity out of sight. a field manufactured water hammer arrestor/expansion chamber is simply a 12" to 24" length of vertical pipe with a cap on the top end. As the water system is filled air is trapped in the vertical riser. Water is not compressible, whereas the air trapped in the vertical riser is compressible and will compress maintaining the air pressure equal to the water pressure. When the water expands in the water heater the air is further compressed to absorb the expansion volume.

With the exception of the 30 or 40 gallons of water stored in the water heater there is supprisingly very little water in the plumbing system in your house.

1 gallon of water occupies 231 cubic inches.

The typical water lines in your house are a 3/4" dismeter line from the water main to the water heater, then 1/2" diameter lines downstream from the heater.

To compute the volume of the lines you multiply the cross sectional area x the lenght of the line.

cross sectional area is Pi x the radiu squared thus:

3/4" diameter line is 3.14 x (.75 x .75) = 3.14 x 0.5625 = 1.77sq.in.

dividing 231cu.in (1 gallon of water) by 1.77 = 130 linear inches divided by 12 = 10.83' linear ft of pipe/gallon.

1/2" diameter line is 3.14 x (.5 x .5) = 3.14 x .25 = .785sq.in.

231 cu.in. divided by .785 = 294" divided by 12 = 24.5 linear ft/gallon

A two story victorian with one bath typically has about 30' of 3/4" and 250' or less of 1/2" pipe, therefore their is approximately 10 to 15 gallons of water in the lines at any given time.

The amount of water that would be required to leak out to reduce the pressure to static head would be the actual volume of the water hammer arrestors. Assuming one 24" arrestor on both the hot and cold water lines would then equal 24" of 1/2" line.

24" = 24 x .785 = 18.84cu.in.

231 divided by 18.84 = 12.26 therefore it would be 1/12 of a gallon.

That would mean the a leak of 12oz would reduce the pressure to zero + static head.

You stated in your post that the measured pressure drop was from 80psi to 40psi thus the amount of water lost would be equal to approximately 1/2 the volume of the water hammer arrestors or less. (6oz or less).

Even with all faucetts closed their are other automated demand loads that could account for that volume. (Toilet fill valve with a minor leak into the tank, heating system, dishwasher or laundry machine valves, laundry machine feed lines leaking at the washer box, ice maker etc etc.)

The next question is, given that water pressure guages are not normally installed on residential supply systems, how are you measuring the pressure?

Your original concern was how can two people use so much water? Can you tell us how much water you are consuming in a month and perhaps we can compute if your consumption is typical or perhaps suggest some conservation techniques you may want to consider.

[ewmyers]

This is exactly what I love about the internet You guys rock!

The house is on city water, and the water pressure was measured by a guy from the water district. He had a water gauge that attached to a spigot on the house side of the shutoff valve. (I have one shutoff valve at the street, then one where a line T's off to feed the rear of the property and one where the water enters the house, and finally another on the line out the back of the house to the spigot and the 2 sprinkler valves.) We verified that the house was isolated. What we did NOT do is close all the shutoffs inside the house. D'OH! Thanks mrcaptainbob, I will try that tomorrow after I purchase a pressure gauge.

I'm pretty certain the toilet doesn't leak. I used those dye strips in the tank, and hours later there is no hint of dye in the bowl. I replaced the old flush stopper about 3 years ago, so it shouldn't be in too bad of shape. The faucets in the kitchen and bathroom are both only 5 years old and definitely don't leak. The tub/shower faucet is 4 years old and also doesn't drip. The shutoffs for the washer are old, but didn't drip at all a year ago when I disconnected the washer during a laundry room remodel. Presumably there is some other sort of valve inside the washer, so perhaps the problem is there?

As for how much water we use, we are both really educated and observant in water use. We have a 4 year old Neptune washer and a 1 year old top of the line Maytag dishwasher. Neither uses much water, and we probably do 3 loads/week in each. No water cooler/ice maker in the fridge. We take short showers (water saving shower head), don't leave faucets running, etc. No pool or spa, no carwashing, small lawn, and anything else that needs water is on drip, except during the winter when all irrigation is shut off. We both grew up in the 70's drought (California) and are almost miserly about water.

So, at our least water intensive time of year, from Nov1 to Jan6, (66 days) our water bill was for 78,540 gallons, or almost 600 gallons/person/day. I'm not sure if I could use that much water if I tried. Incidentally, the water district guy tested to see if the low flow indicator was working, and it was. I keep suspecting that the water meter is broken [8D] but now I sort of wonder about the washer too.

LazyPup, thanks for all the info. I think my house may have been replumbed when it was connected to city sewer instead of the old septic tank, but I don't know when that was. Using your calculations, my house probably has about 3 gallons of water in the pipes - tiny house, one bathroom.

OK, its late for me here, I'm going to bed. I really appreciate your help, and I'll let you know what I find tomorrow!

-Eric

[LazyPup]

That water consumption definitely looks exhorbitant. Let us consider typical water consumptions for two persons:

A shower is typically 3gal/min. thus two people taking one ten minute shower daily equals 3 gal/min x 20 minutes = 60 gal/day

A old style toilet consumes 3.5gal/flush. Allowing 6 flushes per day/ per person = 12 x 3.5gal = 42gal/day

Bathroom lavatory (hand washing, shaving, hair and cosmetic) 1.5gal x 20 times daily = 30gal/day.

You stated you had two washing machines each doing three loads a week. a wasing machine typically uses 8 gal/cycle (Wash & rinse) thus it is 6 loads x 16gal/load =96gal/week /7 = 13.71gal/day.

Dishwashing 3 gal/cycle (wash & rinse) = 6 gal 3 times daily = 18 gal/day.

cooking and food prep = 10 gal/day.

These estimates would equal 173.71 gal/day.

In your post you stated that your biling for 66 days was 78,540 gallon. For a 66 day period the above estimates would equal 11,464.96 gallon in 66 days.

78,540 minus 11,464.96 = 67,075 gals or approximtely 1000gal/day.

dividing 1000 gallon by 8gal/cu.ft = 125cubic foot of water per day.

With all due respect, i think it a waste of time to search for a leak of that dimension in the house. If you had a leak of that dimension your question would more properly be,,how do i install a sump pump.

By contrast 1 gallon of water occupies 231 cu.in, therefore the loss is 231,000 cubic inches of water.

The square root of 231,000 = 480 inches / 12"/ft= 40ft.

Thus 1 inch of irrigation on a lawn area of 40' x 40' would equal 1000 gal. day.

[ewmyers]

OK, I got the pressure gauge and reran the test but with the various shutoffs closed. With all the shutoffs closed (except for the tub/shower because those shutoffs are in the crawlspace under the house)I don't lose any pressure, 80 ponds all the way. I lose 4 psi when I open the washing machine shutoff, and I lose about 20 psi over 2 minutes when I open the toilet shutoff. I don't seem to lose any pressure to the kitchen and bathroom sinks. So I lose pressure at the toilet, but I can't find any sign of a leak there?!?!

LazyPup, I agree that the house isn't the place where most of this water would be lost, I would hear it, see it, be wading it, or something if it was. However, if it was outside, there would be a permanent mudwallow even at the height of summer, and thats not the case either. I'm starting to go back to my thought that it may be the meter. I have one of the first meters they installed/retrofitted in my neighborhood, maybe there was some calibration they messed up?

We actually use less water than your calculation, I'm sure. Our shower head is 2gpm, I'd be very surprised if we used more than .5 gal per use of the bathroom sink, we only have 1 washing machine that does 3 loads per week, and 1 dishwasher that also does 3(maybe 4) loads per week. So that's more like 110 gal/day used and 1080 gal/day unaccounted for.

Thanks for your help, I appreciate your time and sharing of knowledge.

-Eric

[LazyPup]

When it is necessary to do a theoretical demand in leu of actual demand figures I find it is always best to inflate the figures slightly to insure the system is adequate.

When designing pipe layouts for single family residential systems it is pretty basic and there are some rules of thumb that apply, however, when designing pipe layouts for multifamily structures or commercial layouts the codes have a series of tables that assign a specific volume demand which is rated in "fixture units". We must then analize each point of demand individually to determine the total fixture units for that area, then determine the size of the branch line required to serve that portion of the structure.

The method of determining supply line size by fixture units is very similar to the method used to compute drain line diameters. In fact, determining drain lines is more critical as an oversized drain line will cause more problems than an undersized one.

I also agree that there appears to be an error with the meter.

It is obvious that the meter was changed, but you still retained the same main line from the street, the same internal distribution piping, and basically the same demand load, so it should be easy for the water company to check billing records for a period of a year or two prior to the installation of the new meter and compare that average to the readings they are getting with the new meter.

[Dean]

Very interesting stuff but it seems like a lot of complicated answers for a simple "shut in" test.
I just make sure everything is turned off and watch the meter for movement then if it proves water flowing through, I find out where it's going.




Sounds like the toilet might be the problem.

[LazyPup]

dean,,

In my region they are now installing digital meters that do not have a flow indicator.

By taking ten minutes to work out the math I can then conclude the reported consumption is nearly 1000 gal/day which is simply too great to warrant searching for a leak.

While we could guess a toilet, they already indicated in the post that the toilet valves had been checked. Given that a toilet typically fills 1.5gal/min that would work out to 90gal/hr. In order for a toilet to cause a loss in the amount in question it would have to remain in full fill mode for nearly 11.1 hrs/day. Again, the math indicates it is simply to great to even bother double checking.

In fact, the math indicates the loss to be neary 5 times the normal daily consumption for a structure this size. Given that most water meters are inside the structure, we can instantly rule out a leak on the main line from the street, as any water leaking there would not be indicated on the meter.

A leak of 1000 gallons would fill an area of 1600sq.ft and inch deep, which would mean it would be 1/2" to 1" per day in a typical basement. Any leak of that proportion would be blatantly obvious so there should be little need to worry about zone pressure drops or fittings dripping. Either there is another major unreported consumption, such as lawn irrigation or the meter is faulty.

[Dean]

I don't remember ever seeing a digital water meter (but then there are many other things I don't remember either) so please excuse my ignorance but can you not run a simple "shut in" test with one?
Seems like a digital meter should somehow still show water flow passing through.

So Eric, can't you just run a "shut in" test with your meter?

[ewmyers]

Hi Dean, I did a "shut in" test right off the bat. At first I just made sure everything was turned off, then looked at the low-flow indicator on the meter, and it wasn't turning at all. That's when I called the water district to come out and take a look, and they found the pressure drop in the house. They didn't want to check the meter until that got resolved.

Now that I can visualize the magnitude of the discrepancy, I can pretty much rule out a leak anywhere on the property. Even if there was a constant flow of water, but just below the ability of the low-flow indicator to register, it still wouldn't account for the amount of water use that shows on the meter. There is just no way that a 1,000 gal/day leak would not become very obvious very fast.

So my meter isn't a digital one, but they read it by driving by and picking up a radio broadcast from it. The numbers on the bill match the numbers on the meter, so that's not part of the problem. I will call the water district tomorrow and ask them to come test the meter.
-Eric

[ewmyers]

So here is the latest update. First, I tested my water meter's accuracy, and was not impressed. I read the meter, drew 15 gallons through the hose, then reread the meter. It should have shown an increase of 2 cubic feet, but instead showed 3. Then I drew 7.5 gallons and checked the meter. No change, the meter hadn't incremented at all. I repeated the first experiment, and the meter showed another 3 cubic foot change. OK, the meter is wonky (thats a technical term).

Next I called the water district, and was able to verify that the meter is supposed to show cubic feet, rather than having been recalibrated or some such thing. The bill is actually in CCF (hundred cubic feet), and they get that by simply reading the first 4 digits and not the last 2. Well, that appears to be my main problem, because when my meter said 026301, the bill showed the reading as 2630, not 0263. They are investigating, and will get back to me on Monday.

LazyPup, I used your math and numbers on them, and they were quickly convinced of the magnitude of the problem. THANK YOU!

-Eric

[ewmyers]

The Water district called me back just now and confirmed that their readings were off. They corrected my account in their system and are refunding the overcharged amount. They seem less open to the idea that the meter is wonky, and I think I will retest before pushing that issue with them.

Thanks again everyone!