Reservoir Leak Test

[SWarner]

We will be performing a leak test on a 55 million gallon concrete reservoir and are interested how others have measured the water loss. We are specified to follow ACI 350.1-01/350.1R-01 Tightness Testing of Environmental Structures with the HST-050 criteria. This criteria allows for a loss of 27,500 gallons per 24 hours. This computes to 0.017 inches per 24 hours due to our very large footprint. At this time we are considering pressure transmitters, ultrasonic transmitters, a simple ruler within a still on the reservoir wall or some combination of all three. Our biggest concern is the wave action of the water affecting results.

 

[JedClampett]

Good luck with that. Before ACI 350, we were allowed twice that much leakage, and it was still hard to measure. But we were talking about an 1/8 to 1/4 of an inch. Are you sure about that number (.017 inches)? It seems very low.

 

[cvg]

typical method for large reservoirs if leakage needs to be measured is to install a leak collection and detection system with a sump and a pump. short of doing that, I don't know how you will determine if it is leaking or ever be able to locate a leak of this small volume. I agree your 0.017 inches sounds a bit low unless this is a very shallow reservoir.

 

[SWarner]

Sorry for the typo. The allowable loss of height is 0.174 inches. The footprint of the reservoir is 259,750 square feet. The height at maximum capacity is 29 feet. So I guess you could say it is a shallow reservoir. With an allowable loss of 0.050% per 24 hours, the allowable head loss is 29 ft * 12 in/ft * 0.0005 = 0.174 in. Does this help with ideas?

 

[LittleInch]

Leave it for a week and have a large tank next to it so you can adjust for rain or evaporation.

I did a fire water tank liner once and that took a day to notice we were loosing water, fixed it then took a week to confirm that the fix had worked. Wasn't 55 million gallons though!

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[JedClampett]

We adhere a scale on the wall or some other visible location and use that for our measurements. We float a shallow pan of water in the vessel to get the evaporation. Don't forget to tether the pan to the wall or you're going to lose it.
As I said, it's tricky. The water is moving up and down, so you have to use some judgment. Maybe double the test time to get a more noticeable difference. But no matter what you do, it's never going to be exact. If it's less than a 1/4 inch per day, you're probably going to have to pass it and if it's an inch, it's going to fail.

 

[cvg]

assuming this is an open reservoir, you might expect to lose a 1/4 inch of water per day just to evaporation. So you could potentially lose 40,000 gallons per day in addition to any leakage.

 

[btrueblood]

You could measure piezo head to the required precision (.17 inches or better) with a low-pressure, precision transducer mounted at (and measuring) a partial depth. In other words, get a 0.25% or 0.5% precision transducer with a 0-36 inch WC range, and mount it some 30-36 inches down the tank wall (or pressurize it and run a bubbler rig, cheaper that way). By only measuring the change in the upper level, you have reduced the range of the transducer and improved its resolution relative to the full depth.

 

[dicksewerrat]

Why not place a series of floating booms in the reservoir to still the wave action. Maybe 4-5 sets of booms, 10 feet apart in a half circle, all terminating at a wall.

Richard A. Cornelius, P.E.

http://WWW.amlinereast.com

 

[MEM1]

Do a search for overnight water balance test by Jay Ham. He was at Kansas State University when he did the study, but he is currently at Colorado State. I am currently attempting to do a seepage study on a synthetically lined pond at an ethanol plant. Maximum seepage is 1/64 inches/day. It is proving to be extremely difficult. Trying to measure seepage at numbers so close to zero is near impossible when the is no easy way to mount stilling wells so there is no movement. Shoot me an email if you want to discuss further s m u r r i s o n @cox.net. Take out the spaces.

Scott

 

[racookpe1978]

At that very low levels of "possible leak" detection, watch for thermal expansion as well.

If just the upper surface (6 inches or so) heats/cools from the original temeprature, you probably can't detect it in a level gage if the beginning and ending thermometers read the same.

Try placing four thermometers at your evaporation pan. One right under the surface (less than 1 inch from the surface). One 12 inches down. The next halfway between surface and bottom. The last right near the bottom. Hopefully, in a still open pond, your heat will be absorbed right near the surface, your evaporation will happen in the top few centimmeters, the daytime surface heating (and night time cooling) won't mix too much lower down.

 

[JedClampett]

I stole this from somewhere, but it's still good:
"They pretend to build leak tight reservoirs and we pretend to measure it."

 

[RHTPE]

This is not my area of expertise, but I have a suggestion anyway.

Fill a clear vinyl tube with water. Drop one end into the reservoir well below the water surface. Run the other end to a shaded vertical surface at the same elevation as the water. Fix the tube to this surface with a scale behind it. Set your benchmark elevation.

0.174 inches should be discernible in a small diameter tube. I would run the entire length of tubing so that it is below the water level of the reservoir. By locating the reservoir end of the tubing well below (12" or so) the reservoir water level it should mitigate the wave action. Shading or burying the tubing should eliminate thermal effects. Perhaps a few drops of ordinary motor oil added to the observation end of the tube would help define the water level and make it easier to see.

Sometimes the best test or measurement technique is simple and far from sophisticated.


Ralph
Structures Consulting
Northeast USA

 

[chicopee]

I am somewhat in line with RHTPE, however I would have several clear tubes let's say 5' long secured and positioned around the reservoir. Mark the original water level and then come back 24 hours later for additional readings.