Help with Storage Tank Bolt Failure 5

[bottle77]

My company installed two 10,000 gal 304 SS tanks Just outside our plant back in February November 2018 (time flies). They are roughly 8.25' in diameter. One is holding a chemical that is 7.9 lbs/gal the other holds a chemical that is 8.69lbs/gal. The material comes in at 140F, and we try to keep it that way, though it may dip down to ambient temperatures (60-80F).

The tanks were installed about a foot apart on a level high strength concrete pad. 6 holes were drilled in the 2 inch angle iron flange around the tank bases and drilled down roughly 18" into the concrete. Then 24"x0.5" anchor bolts were epoxied into the holes and nuts were used to tighten the tank down to the concrete. After installation the tanks were equipped with band heaters at the bottom to help keep the material warm.

Last week the operators came in to find that the tank holding the 7.9 lb/gal liquid had snapped 3 of its bolts and bent a fourth considerably. Half of the base is sitting up 1.75". They emptied the 4,500 gallons of stored material out of it, and now we're trying to determine the cause of the failure, and how to remediate it. The other tank was just under half full at the time, probably also around 4,500gal.
Both tanks have vents to keep them at atmospheric pressure, so I don't think a pressure buildup could have caused this. The welds to the outer flange have been broken in a couple spots. The other tank doesn't really show any signs of stress, though there are some hairline cracks in the concrete around the flange. Could this be a failure from thermal expansion? The broken bolts appear to be sheared off. It was windy the evening it failed, but there was at least 35,000 Lbs of material sitting in the tank. I would think that would be enough to "hold it down" as it were. I would also think other parts of the tank would be damaged from wind before the base.

Any advice would be appreciated. Not sure where to start. If pictures of a different vantage point could help let me know. I am a recent grad with a ChemE degree. I don't know what analysis, if any, was done before they decided on how they wanted to install them outside of what was practical given space constraints. Tank schematic should be attached to the post

 

[bottle77]

The chemical identities might help. The tank that broke bolts is holding nonylphenol and the tank that didn't is holding epoxy-resin.

 

[Geoff13]

You should be talking the company that sold you these tanks. Presumably Pacific Tank Engineering or Douglas Tank Sales.

The liquid in the tanks won't help hold down an anchored tank until after the anchor bolts have broken. To use the weight of the product the tank must lift slightly so the bottom can try to raise the liquid. However the tank can't lift due to the bolts, so all the load goes to the bolts until they fail.

 

[bottle77]

We are. They said it looks like an overpressured tank, but out vents and check valves are pvc. Wouldn't they fail before the steel would deform?

 

[MJCronin]

The choice of a "angle iron" with long spans between anchor bolts seems very suspect to me...

The tank that you purchased seems to be built to no particular code or standard. There seems to be no pressure or temperature limitations. How did you set the pressure relief valve ? The angle iron at the tank bottom seems to be more of a housekeeping device suitable for a dairy rather than for a process chemical facility.

This tank configuration would not meet the requirements of the more robust API-650 or API-620 design codes. The angle iron anchorage would not be permitted under the API codes.

I believe that your anchorage problems are due to a thermally induced rotation of the angle and a prying action on the bolts ... The angle is simply a bad idea

How are the chances of using individual 4 or 6 welded "anchor chairs" instead of the angle iron ?

Oh, ...wait .... your boss just discovered that there is no money in the budget for repairs.... sorry

Does the tank have a truly flat bottom ? (I can't tell) and what is the design pressure ?

MJCronin
Sr. Process Engineer

 

[Geoff13]

According to your attachment the bottom is only 0.11" thick. That will deform very easily once the bolts break.

I think they're on the right track suspecting an overpressure event. That was my first reaction.

If the 3" nozzle at the roof apex is the vent, it won't take much to overwhelm it's venting capacity.

 

[MJCronin]

You bought an extremely inexpensive, thin walled tank (possibly for dairy or winery service ? ) ....Your design sketch shows some shell courses are .078"

This is the YUGO of flat bottomed tanks ...

Very thin walled tanks will exhibit extreme shell rotation at the tank bottom. It is my opinion that this rotation is a contributing factor in your anchorage problem

A thicker walled tank would be required for a unit built to API codes .... 0.1875 " is the specified minimum.

The minimum shell thickness recommended by API has been specified to prevent excessive rotation at the bottom of the shell.

https://www.eng-tips.com/viewthread.cfm?qid=217201

MJCronin
Sr. Process Engineer

 

[3DDave]

Is it possible the concrete is shrinking? I see a large number of cracks. If nothing else I see nothing to make up for the differential expansion between the metal of the tank and the concrete.

I see in one reference that 1 part in 1000 can occur; at 8 feet, that's about 100 inches, so 0.10 inch concrete shrink is a possibility.

The other consideration is the thermal expansion coefficient for steel is (more or less) 12ppm/deg F; if the tank hold-downs were drilled at 30F then this is about 100F delta. Over 100 inches that's (12inch/10e6inches-degF)(100degF)(100inches) = 0.12 inches. Substitute values for what the actual conditions were.

So between concrete shrinking and the thermal expansion from the installed condition that's the possibility of nearly 1/4 inch of displacement across the diameter of the tank.

One other thing to check is to see that the loads on the vertical leg of the angle are in line with the flange. If they are not restricted to being there then the horizontal leg/flange will act to constrict the thermal expansion, due to being in contact with cold concrete, and tend to rotate the vertical leg outward. This can be solved by thermally insulating it so the horizontal leg is as warm as the rest of the tank. Alternatively choose a different flange so the expansion loads will be symmetrical, such as an easy-bend T with the vertical of the T set horizontally outward.

Edit to add:

I agree on the over pressure - even a slight over pressure on this will dome the bottom of that tank and pry it up. That's over 30,000 square inches; so 1 psi would net 30,000 pounds of tension at the angle; with the leverage, those bolts are seeing twice that force, so 15,000 pounds per bolt. That will do some bending.

However, the thermal part still sticks with me if you are sure the other bolts were sheared and did not fail in tension.

The tank maker should be able to tell you how far the bottom of the tank will dome before taking a permanent set and what the pressure to do that is. I bet the pressure is a small number.

The only surprise is if the bolts are arranged as a square. If a bolt fails then the tank should tend to tilt to the opposite one, relieving the load and preventing the opposite bolt from failing. I'd expect two bolts to remain OK, but maybe the prying action is still too much for that configuration across the chord between them.

 

[bottle77]

Thanks for all the info and advice, @MJcronin, @Geoff13, and @3DDave.

 

[bottle77]

@MJCronin The only thing they asked me about the tanks before buying them was “will the chemicals eat through them” and “what is the cheapest way to heat them.” Beyond that I wasn’t consulted about buying or installing them. I assumed that they had the experience and knowledge to do it, and was told they were installing it how the manufacturer recommended when I asked. The boss and operators were the ones who installed all the piping and venting. To be fair, if they had asked I probably wouldn’t have known how to size/design the pressure relief since my education focused more on running/monitoring processes than installing equipment, but I would have researched it. Anyway, the tanks were bought used, and my understanding was that they either held milk or water previously. There is money in the budget for repairs since before I posted on here the boss was looking at replacing the tanks entirely. If we installed the anchor chairs from that schematic, would we need to tighten/loosen the bolts as the temperature changes from season to season? Should we make those chairs out of SS like the tank? The tank had a truly flat bottom, now it is domed out , which seems to support the consensus of over-pressurization. Boss said in an email that they have a 1’ ball valve that is opened during filling and emptying, and 0.5psi check valves for when the product is just sitting. After we emailed him, the manufacturer said the venting should be 1.5x the inlet, so we need to increase our venting. Manufacturer recommends laying the tank down, sledgehammering the bottom flat, increase the venting, and then it could be put back into service. Don’t know what the design pressure is, but I bet it is atmospheric. I’m a greenhorn so could you tell me what YUGO stands for? (google couldn’t tell me. It was suggesting it meant Yugoslavia)

 

[bottle77]

@Geoff13 copied from above: . The tank had a truly flat bottom, now it is domed out , which seems to support the consensus of over-pressurization. Boss said in an email that they have a 1’ ball valve that is opened during filling and emptying, and 0.5psi check valves for when the product is just sitting. After we emailed him, the manufacturer said the venting should be 1.5x the inlet, so we need to increase our venting. The manufacturer is saying to increase the venting to 3”, but it seemed like you were saying that wouldn’t be enough. Is there a better method for figuring out venting capacity?

 

[bottle77]

@3DDave The hole drill temp was probably ~40F, so Δ100F sounds right. No one was around when they failed so I can’t say definitively how they failed. I can say that one of them was found a considerable distance away, so maybe it was tension? Is there a good way to tell from the fragments? There were 6 bolt holes drilled roughly evenly spaced around the angle iron. Three outright failed, one bent, and the other two look normal. The tank bottom looks domed now, and it is tilted up on the failing side.

 

[bottle77]

I think we need to increase the venting on the other tank, and maybe switch to the anchor chairs to keep it from failing in a similar manner. Once again thanks for the advice, and let me know if I can clarify anything else

 

[LittleInch]

Well a few things here which should be considered.

The drawing doesn't show any nozzle other than the 3" presumably vent nozzle at the top.

Was this a free vent to atmosphere or any other restriction? What's a 0.5psi check valve? Was that on the vent??

I assume you mean the inlet is a 1" nozzle or a 12" nozzle? 1 inch seems very small for a tank this big.

This tank is 26 ft tall on an 8 foot wide base - pretty tall and slim. Wind loads will be an issue, especially if you have two together that close. Heaven only know what sort of interaction is going on there depending on wind direction and buffeting. 6 puny little bolts on a spot welded 2" angle seems very low. Were these tanks ever designed to be outside?

you say the base was flat - but the tank base has a slope on it of 1:50. Now when the erected it did they need to pack up one side to make the sides vertical? 1: 50 would look very odd for such a tall tank and be both visibly out of plumb and on a spirit level.

there's a typo on the drawing "TPYE" instead of TYPE.....
If the tank moved and tilted a bit in the wind then the weight of contents could have bent the floor, but yes over pressure is a key reason for this.

BTW I don't know what YUGO is either. Or what a "Van Stone style" flange is.... (had to go and look that up)



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[MJCronin]

https://www.caranddriver.com/featur...history-of-the-yugo-the-worst-car-in-history/

---WORST CAR IN HISTORY !!

MJCronin
Sr. Process Engineer

 

[bottle77]

@littleinch and geoff13 It doesn’t look like I said it earlier, the inlet is 2”.
I’ll see if I can get some answers from the boss and operators on how they leveled the tank and how they installed the vent. I can ask the manufacturer, but I’m fairly certain they were designed to be outside just set on some packed gravel. That’s what I remember from a conversation, but it’s been a minute since we talked about it. The windspeed where we’re at averages between 6 and 8 mph depending on the month with maxes around 20mph. The strongest we’ve had this year was almost 30mph. The wind max the day/night before the bolts broke was 23mph.
@littleinch do you have an anchorage suggestion, or resource that can help specify an adequate anchorage?

 

[bottle77]

Thanks for the link MJCronin. So Yugoslavia was involved after all.... I take it then you don't think much of these tanks... I'll ask what standard they were built to.

 

[bottle77]

So check valves have a spring inside that keeps the valve closed, but pops open when the designed pressure is reached to keep the tanks at roughly atmospheric. There's one to let air out and another to let air in. in the middle is the vent valve that is opened during filling. it is one inch and connected to a one inch pipe at the top of the tank check/vent apparatus:

The three inch flange in the top is connected to a two inch inlet pipe. the flange in the bottom goes to our operations.

I asked the boss, and it sounds like he's made up his mind to take these tanks out for now and use them for lighter materials later, and buy new "heavier duty" tanks and install them. Asked if he wanted me to look into anything and was told he has it handled.... sounds eerily familiar to what I was told when the company bought these tanks.

I guess they'll see what happens....

Here's some images of what the other tank base is looking like right now:

 

[MJCronin]

Ask your boss if he has an MBA .... If the answer is yes, ask if his concentration was in finance or marketing ....

MJCronin
Sr. Process Engineer

 

[bottle77]

He doesn't have a college education if I remember right. Definitely finished high school though. I don't think he would care for me asking so I don't know if I can confirm....

He started this company ~8 years ago with some seed money from his father who was/is the CEO of a construction company. He's slowly gathered folks as he's run into them in the industry. I actually got hired because I called in and asked if they had any work for a chemE as there weren't many entry-level positions open in the area. Said he had wanted someone to take over the lab tests for awhile, but until that point had been doing the formulating himself.