Concrete Water Tank Leak....Now What?

[bigmig]

Does anyone have some suggestions on cast in place concrete repair solutions in regards to water tightness?

I have a 22,000 cubic foot rectangular potable water tank that just went through the specified leak test, and...it leaks in about 5 spots. The owner has described them as "major" leaks. We went through a lot of effort to spec the right mix, Xypex waterproofing admixtures etc., but in the end, we have reason to believe the contractor didn't vibrate the walls like he should have.

So far the research I have come up with suggests:

1)Line exterior walls with bentonite filled boards. Patch interior with polymer concrete and then line with glass reinforced plastic.

2)HLM 5000 spray grade waterproofing

3) Various membranes (Atlas) and other spray on tank linings.

Does anyone have experience or success with a particular method or product? Again, the key words are POTABLE water, so whatever we put in there must be pretty inert.

 

[cntw1953]

Find out the reason for leakage - through cracks or honeycomb. Identify structural importance of such defects. If no structural problem, do chip/saw cut and patch smooth from internal, then provide suitable linners. Otherwise, you have bigger headaches.

 

[kelowna]

Define major! 1 litter a day/hour/minute?

Where is it leaking? Cracks/construccion joints/ through the concrete itself?

For leaks trough cracks/construction joints I have used Xypex, both as a slurry applied on the face of the concrete (kind of painted on) and as a paste forming a 2 inch triangle in a joint betwen wall and base slab. Both applications very sucessfuly and reasonably cheap (some of the options you talk about sound expensive). If I remember right once cured, Xypex is as good or as bad to water as normal concrete.

As with everything good surface prep/application/curing process are very important.

In my case the leaks were very minor, in the order of a couple of cups of liters a day maximum, sometimes just dampness in the concrete wall.

We considered sprayed on membrane (some were epoxy based, other were polymer or plastic) but discarted it. The problem was that they could not guarantee a design life of more than 15/20 years.


In your case, would it not be better to treat the problem areas instead of all the tank?

 

[miecz]

We had a similar situation 3 years ago. Many leaks, furious owner. Had the cracks repaired with an injection system called the "Krystol Crack Repair System." Worked great. Kryton, who makes Krystol, is a competitor to Xypex. Perhaps Xypex has a similar repair system.

 

[nackra]

http://www.sikaconstruction.com/con/con-prod-app-crir.htm

Contractor used Sika product to repair leaking concrete tank on a project. Need to establish limits of defects to for repair method.

 

[dik]

Need to establish the cause... Then repair patch, Xypex, or injection, or whatever... but cause is paramount.

Dik

 

[bigmig]

Thanks for the advice everyone. We are still in the preliminary stages of the solution, so I don't know the volume and rate or leak, or the cause, although obviously these are the "million dollar" questions that will receive attention 1st and therefore direct the answer we come up with.

The reason we are thinking the cause is improper vibration is because the contractor had a form blow out during the wall pour, and proceeded after the blow out with "extra caution" (i.e. limited to no vibration). But that is an assumption that we will need to verify.

 

[cntw1953]

Look for cold joint formed after the blow out, the contractor might have poured fresh concrete over already, or semi-set rough concrete without proper joint treatment. It would be a good location for water to seep through.

 

[rowingengineer]

Concrete (especially concrete with Xypex) can be self healing, if the cracks are “minor”.

I would recommend getting xypex involved as they are quite often willing to provided extra product for free if they think the leaks will give them a bad name.


Arguing with an engineer is like wrestling with a pig in mud. After a while you realize that they like it

 

[JedClampett]

We just got done with a similar issue, except with about a 50 times bigger reservoir. I'd strongly advise sending a diver into the full tank with a dye wand and pinpoint the leaks. Maybe you've already done this, but you weren't specific on how you knew where the leaks were. Otherwise, you're going to repair a lot of good wall and maybe miss some bad spots. Ideally, you send two divers down, one with a dye wand and the other with a camera and a line filled with an expanding polyurethane caulk, such as SikaFix HH. One diver identifies the leak, the other injects it and you can verify the repair on the spot with the dye wand. For rock pockets and other large flaws, an epoxy injection or Xypex might be better than expanding foam.
The contractor should pay for all this.

 

[Zambo]

I assume the tank is buried, which makes me wonder how the leaks have been identified. JedClampett also asks this. Is it a case that groundwater is leaking in? Are there also damp spots or are have just 5 leaking locations been identified?

Another problem I have come across is ground water ingress during consttruction causing washout of fines at the construction joints.

I agree with others that an injection system is the way to go. First you have to select the system either epoxy or foam and then start drilling and pressure grouting. The problem is that once you seal one leak the water can find another path and reappear. Also once the leaks have been repaired people may start focussing on damp spots.

 

[bigmig]

After meeting with the owner, leak rates are in the range of 1000 gallons per 72 hours. A site inspection showed only one leak with visible, trickling water, although non visible leaks may be possible. Other leaks were noticed as dampness, or ponded water on the existing tank lip (see sketch) with no moving water.

The real problem we are about to face is how we poured the new tank. Essentially an existing water tank (concrete, above grade, rectangular) was used as a form, after we removed the roof. We cast a new, independent structure inside the original tank, and extended the height (and volume) of the tank by about vertical feet.

ATTACHED IS A SKETCH of the tank configuration and leak location. There is a good chance the cold joint may be leaking (?) but we cannot see it. Unforunately, the leaky side is the side that is backfilled.

The tank is above grade, with one side backed into a hill.

At this point, due to the fact that the main leaks are hidden, we might have to drain the tank and do a roll-on liner. The other methods are too expensive for a hit and miss application.

I was always taught that concrete is very porous. I probably wouldn't have thought it could be used as a water storage device until I saw the existing tank perform. How much water does new concrete absorb? I'm sure it depends on density, add mixtures and a handful of other factors. Is it possible that a tank this size could soak up 500+ gallons of water in 72 hours?

 

[Gumpmaster]

You normally want to perform the leak test after the tank has been filled for a couple of days. The concrete will absorb some water, and lots of the smaller cracks will fix themselves. A white crystal will form in the smaller cracks, I forget the actual substance (calcium carbonate?), and they'll stop leaking after a couple of days.

If you did the leak test right after the initial filling, see if you can do another one to find out what its leakage rate is now.

Concrete is quite water tight. A well built, well reinforced tank won't have any problems for many many years.

 

[cntw1953]

Water travels, unless there is visible defect at the location of leak, it may come from somewhere else.

If difficult to visually spot the defects, you might try to use "dye" to uncover them as Jed mentioned above, however, I would do differently - drain the tank, and fill the space in between new and old tanks with water and let it seep back in, now you should be able to spot the defective locations easily.

Well, to me, liner is cheaper and less messy. Good luck.

 

[JStephen]

I think the AWWA specifications for prestressed tanks include a leakage allowance. You might compare your leakage to that allowance as a matter of information, although the actual construction may be quite a bit different.

 

[Prepakt1]

Normally for this type of problem We use HLM 5000 brushed on the inside not sprayed, Have also used Xypex in the past but usually just use HLM 5000 these days. Any rough areas of concrete should be covered with a concrete based water plug and then HLM 5000 brushed on( Much thicker than a spray coat}. This should be done when the surface inside is dry. You can also use HLM 5000 on the exterior walls as it does have a good pressure rating and will hold .We have used it this way where access to the pressure side is impossible, sometimes two or three coats but I would not suggest just externally as a total solution.
You could cover the whole inside face or just the areas and seams you think are causing problems but water might still migrate though the concrete from areas not covered.

Intrusion Prepakt /marineconcrete.com