Thermal waterhammer: steam sparger 2

[fvincent]

Have you ever experienced strong vibrations in water tanks provided with steam spargers?

The sparger is made of several holed pipes forming an hexagon on the bottom of the tank. The holes are 6 mm and were drilled on the top of the pipes

Low pressure waste steam from batch processes is used to heat water from 75oC to 95oC.

The tank itself is 8m diameter and 9 m height. Cold water comes from top and hot water is pumped from the tank bottom.

Level can change a lot, because heated water is transferred to a second hot water tank (reservoir)only when there is steam to heating it in the main tank. Cold water, on the other hand, is continually supplied to the main tank

A pair of chicanes inside the tank were installed to avoid pumping cold water. These chicanes are so located:
- lower one: a horizontal circular ring at 2.5 m from the bottom, inner diameter 6 m and outer diameter 9 m
- higher one: a horizontal disk at 3.0 m from the bottom,

A set of column is placed to support the chicanes. Some radial nervures (flat bars) were welded to the lower chicane.

Well, bangs can be heard and the tank is shaking a lot.

Internal inspection of the tank showed that some nervures detached from the chicane. One has even gone to rupture!

Well, I had heard of collapsing effects of steam bubbles and thermal waterhammer, but I was quite surprised with the intensity.

Do you think a small injection of air to the steam could cause a benefical cushioning effect?

Do you think the chicanes should be drilled to let eventual confined steam to flow to the upper part of tank, so avoiding hammer effects against such chicane?

Other ideas??

Thanks

fvincent

 

[davefitz]

Just an idea:

Use multiple spargers in parrellel as follows: each sparger to be oriented vertically, and the sparger would be co-avial with a larger diameter flow nozzle. The cold water would rise thru the annular space at the bottom of the flow nozzle, mix with the steam bubbles from the sparger, and exit the top of the flow nozzle. This procedure breaks the warming process into 2 parts; first within the annular space , second as the hot water ejected from the flow nozzles mixes with the tank water.

 

[fvincent]

thanks davefitz

It is an idea for future however

As for immediate modifications:

Is it probable that large steam volumes might be forming just above the flat chicanes and either abruptly condensing or suddenly escaping so that in both cases water impacts against the surfaces? And if yes:

- Do you think that air injection (in the steam flow) could reduce the magnitude of such thermal waterhammer effects?

- Do you think that drilling some holes in the chicane or just removing them could minimize the problem, by allowing steam to leak upwards thru such holes ?

regards

fvincent

 

[abcmex]

The problem is the steam, the collapsing bubbles create this problem, I doubt that any of the methods will help you.
You may:
reduce the size of the holes (sparger) and limit the steam pressure; this may reduce the size of the bubbles and consecuently noise and your heating capacity.
Install eductors (ventury type) inside of the vessel.

Kind regards
Check the link:

http://www.process-heating.com/CDA/...ures/BNP__Features__Item/0,3156,20425,00.html

 

[fvincent]

Thanks for your suggestions, abcmex

However I have to take in account that steam pressure is only 1 bar g and the hydrostatic pressure at the bottom of tank is 0.8 bar g. So only 0.2 bar is left as "driving force".
Moreover the steam generation is not constant, depending on heat release in a batch process of vegetable oil hydrogenation. Steam mass rate changes abruptly from zero to up to 12 t/h

Venturi injector might not operate very well under these conditions, don't you think so?

For future modifications I think however that I could explore solutions like the one you have kindly suggested.

My present concern is how to diminish the strength of the thermal waterhammer.

I am not convinced of the efficacy of air injection to promote some cushioning of implosions, either... But I would like to hear more about other experiences on this solution..

As to eliminating the chicanes, I think they are submitted to strong impacts because non condensed steam bubbles will accumulate under them. When these bubbles (coalesced) finally condense or escape water would impact violently against the metal. That could explain why many of the nervures are so damaged.

I would appreciate a lot your comments.

thanks

fvincent

 

[TBP]

Spirax Sarco used to publish a series of booklets on a variety of steam & compressed air topics, and I only have 3 of them. One is "Direct Steam Injection - Heating Liquids With Live Steam". It indicates that a larger number of smaller holes are better than fewer larger holes. It has an chart that shows 1.6mm, 2, 3 4 & 5mm holes. It doesn't show 6mm. "The holes should be as small as possible." "Note that the total area of the holes is almost always much less than the cross sectional area of the pipe in a well designed system." It goes on to say "Holes should generally be equally spaced, and drilled just below the centre line on one or both sides of the sparge pipe."

There are a couple of sizing rules of thumb for the holes:

1/ A 1/8" hole will pass 8 lbs/hr, with an 8 PSI differential."

2/ A 1.6mm hole will pass 1.6 kg/hr, with a 1.6 bar differential."

Regarding the mixing of air to reduce noise, I once saw this recommended in a very old engineer's book (1880's). I've never had the occasion to try it, but as I recall, there was a lot of excellent information in this particular book, so I'd be surprised if it didn't help.

 

[25362]

I concurr with abcmex's opinion on the mechanism of steam bubbling. As an analogy we could envisage a kitchen boiling pot, in which, at the beginning of the heating process we actually hear the implosion of the rising steam bubbles when they coalesce upon meeting with colder water.

Some thoughts: to improve heat transfer -with less bubble implosion effects- one should aim at larger specific surfaces, namely, smaller bubbles.

From gas-liquid dispersion studies we learn that to keep within the "bubble flow" regime range for a tank of more than 2 m diameter, the bubble velocity should be lower than 4 cm/s. Velocities higher than 5 cm/s might bring the sparging steam into the "churn-turbulent" range generating bubbles much larger than their usual 3-8 mm in diameter.

For the sake of good operation, beside doing what abcmex has recommended, one thing to do would be to keep the tank level and the steam pressures at constant values, and, as you suggest perforate the chicane as a last resort. About the chicanes themselves, apart from removing them altogether as you are considering, is it possible to lift them up so as to make their distance from the sparger larger than 2.5 m ?

As for the possible air cushioning effects I believe that your suspicion is right; steam being lighter (less sp. gravity) than air would tend to collect nearer the lower side of the chicane anyway.

Good luck.

 

[fvincent]

Thank you both TBP and 25362,

The mention to a book dated 1880 was quite amazing and shows us that we are not inventing the wheel...

The proposition about the localization of the hole (only below center line) will certainly reduce dangerous behaviour inside the steampipe (sparger).

I think that adding this to a possible reduction of holes diameter as indicated by you and by 25362 could result fine.

As to level and pressure stabilization, 25362, I can say that the tank is atmospheric and a certain variation of level takes place because water is pumped from (heated water from the bottom) and to (cold water to the top). The variation of level associated with these flowrates however is quite low, because steam injection takes place during 10 to 20 min each four hours.

The air segregation from steam however physically reasonable would probably never happen due to high turbulence and high diffusity, don't you think?

Thanks all





fvincent

 

[25362]

Just below the lower chicane the water turbulence may have subsided, and given the needed residence time, steam bubbles may preferentially attach to it.

Besides, on heating (atmospheric) water, dissolved air would be released. Neither this effect nor any steel (weakening) corrosion effects have been noticed or reported.

Whether the internally diffused-and-heated air may help in breaking the steam bubbles into smaller units is unknown to me. The easiest thing would be to give it a try. Do you agree ?

 

[fvincent]

Thanks, 25362

Yes, I do agree with you!

The best I can do now is just trying some air-to-steam mixture (near 1% of air) and check the results.

regards

fvincent

 

[GenB]

I do not think that the air injection will help but will add rust/oxigenand pitt the steel.
Your problem may be in the operating cycle, large amount of steam is used at once creating cold condensate, then replenish with cold water(make-up water). you may want to add some heat to this water. as it is done with deaerators. heat the incoming makeup water. also isolate the return condensate, return water should drop into the steam side instead of the water side.
'hope this helps.
ER

 

[TBP]

I initially thought about corrosion problems from adding air/oxygen as well. But if a carbon steel tank is being used to heat raw water, corrosion will be pretty spectacular, whether air is added or not.

 

[fvincent]

Corrosion IS a problem indeed... but not for the tank which was made out of SS. The sparger and the chicanes, however, are made out of CS and so, as you've pointed out, do undergo severe corrosion. That will be corrected in the near future, I hope.

Regards



fvincent

 

[GenB]

I can not believe it, corrossion was just a comment!
ER

 

[fvincent]

Never mind, generalblr: it was just a comment of yours not a witch spell... Corrosion was happening before it!! ;-)

Would you suggest to replace CS pieces by SS, or do you think epoxy painting could prevent this kind of corrosion?

regards




fvincent

 

[TBP]

I'll bet that by the time you fool around with a coating on carbon steel, it will be cheaper just to go with stainless steel from the beginning.

 

[fvincent]

TBP, I guess you are correct, but the 'beginning' is over a long time ago... So, what are my options now? What is feasible?
regards

fvincent

 

[TBP]

I guess I didn't word that very well. You have an existing stainless tank, but some misc bits are carbon steel, and corroded. I meant that if it were my installation, and it's time for replacing any carbon steel components that are corroded, I'd simply go with SS for anything in contact with the water, rather than fool with coatings on CS. I've never been a big fan of coatings.

 

[kirsner]

fvincent--The "overpressure" due to the collapsing bubbles is due to the size of the bubbles and the ambient pressure surrounding the bubbles. Making the holes in the sparger smaller would help, so would interducing the steam at a higher elevation in the 9 meter tall tank to reduce the ambient pressure. For a rapid condensation collapse to occur, the water must be subcooled with respect to the steam at least 20oC--probably more. This might supply another motivation for introducing the steam toward the top of the tank where the water is warmer.

Introducing air into the steam is also an idea as it cushions the bubble collapse. (Don't know what a "chicane" is so no comment there).

Hope you let us know what works. kirsner@kirsner.org

 

[TBP]

Mr. Kirsner! I was thinking about "condensation induced water hammer" when this string started