Storage Thermal Latent Tanks Failure

[forensiclab]

Hi guys,

There is a total of 5 STL tanks in operation which are connected to chillers for district cooling. During the night when cooling usage is low, cool brine will be passed through the tanks to freeze up the ice balls inside to store the latent heat. During the day, the ice balls will be melted to support the chiller.

The tanks were put into commission since 1993. Recently we found leaks on two of the tanks. We then proceeded to remove the insulation and realised that there were cracks together with warping/buckling on the wall of the tanks.

As there is a substantial amount of condensation occurring, we first thought that it might be due to corrosion that leads to lost in material. After we did a thickness testing, the reduction in thickness of the tank wall was not significant.

I could not figure out what could have caused this failure. Someone please enlighten me.

Many Thanks!!

 

[MikeHalloran]

Uh, 21 years in brine service?

Some photographs might help.


Mike Halloran
Pembroke Pines, FL, USA

 

[dhengr]

Forensiclab:
Well..., I think that crack over there is one of these kinds and that crack over here is one of those kinds of crack. They are obviously not the same, one goes up and the other goes down. What else do you need to know? Oh..., maybe some photos and a few sketches of the tanks, with realistic proportions, details of tank construction and welding, piping, stl. matr’l. specs., etc., etc., with dimensions and the like. How do the brine pipes go through the tank wall and can freezing start right inside the tank wall? Do the tanks (fluid/water) freeze solid, or what size and shape do the ice balls/cylinders grow to? Freezing ice can exert very large forces on the tanks or pipes. What about the relative expansion and contraction of the various parts and the ice, how do they push on the tanks? How does the expanding ice push on the tank, how much movement, what type of buckling does this cause? Are the tanks open to the atmosphere? What do various details look like, in terms of hard spots, inflexible details and corners, where some tank wall movement causes high stresses and cracking. Is insulation closed cell insulation, so condensation is less likely to penetrate to the steel tank skin? Is there any consistency in crack locations and patterns, w.r.t. tank and pipe details? Rather than freezing solid ice, should the fluid be stirred so the whole volume gets to 32.1̊F and then the brine flow slows down, and holds the temp? Have you spoken with the engineers who designed the original system? My goodness, you ask a question like your OP, and then provide so little needed info. for a meaningful discussion, what kinda answers do you expect, except a bunch of questions?

 

[LittleInch]

As dhengr says "Someone please enlighten me". Reading this carefully is sounds as if you have a brine line going through a tank which allows ice to grow on the inside of the tank. Ice growth can very easily damage things if left unchecked or you may simply have flexed the tank too many times at temperatures which the material is not best suited for and as a result got fatigue cracks at some stress concentration point. How do you determine when there is enough ice and not too much? was this operation a retrofit or just some mode of operation which has happened over time or was it always part of the planned operation??

There are plenty of people here who can offer opinions, but need much more data than you have given up so far.



My motto: Learn something new every day

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

 

[chicopee]

I'll be interested to see pictures. Per chance, do you have bronze or brass valves at the bottom of the tanks?

 

[Duwe6]

If you can clean and dry a cracked portion of your tank, and warm it above freezing: go the welding supply store and get some dye penetrant "Visible Red". You'll need a can of dye, one or two cans developer, and 2-6 cans cleaner. Be ready to photograph the cracks - red lines on a white background, when they first pop up. There are a bunch [eq. >>20] of us that can diagnose fatique cracking from the patterns. And that is probably what you have.