Jump to content



Featured Articles

Check out the latest featured articles.

File Library

Check out the latest downloads available in the File Library.

New Article

Product Viscosity vs. Shear

Featured File

Vertical Tank Selection

New Blog Entry

Low Flow in Pipes- posted in Ankur's blog

Sizing Tank Overflow Line

overflow line gravity flow storage tanks

This topic has been archived. This means that you cannot reply to this topic.
4 replies to this topic
Share this topic:
| More

#1 Oliz

Oliz

    Brand New Member

  • Members
  • 7 posts

Posted 28 November 2014 - 02:30 AM

I've got a task to re-check an existing tank's overflow protection since the old pump supplying the tank has been replaced by two new ones.

 

The tank receives oily water and is used as a settle tank where the water flows to the water treatment facilities and the oil is skimmed off. The tank is an atmospheric tank with floating roof. The overflow line has a syphon break at its top and the size of the line is 14" (I.D=0,3398 m) both upstream and downstream of syphon break. The maximum inflow from the pumps is 700 m3/h.

 

I came across an old post (http://www.cheresour...-overflow-line/) which had a lot of good information but I am still not totally clear if I have understood the problem correctly.

 

Based on equation 3 in P.D Hills article about gravity flow, h is calculated as follows:

 

h > 0.811 * (QL)2 / (g * d4) = 0,811 * (700/3600)2 / (9,81 * 0,33984) = 0,234 m (2,29 kPa)

 

I also calculated the pressure drop from the inlet of the overflow pipe to the top based on darcy weisbach equation (just for comparison) and got a result of 5.21 kPa which equals 0,532 m.

 

Based on the calculated heights I draw the following conclusions.

 

1. If the maximum level of the floating roof is above the height of the U-bend + the height calculated above there will be no problem.

 

2. If the maximum height is not high enough, then the floating roof must be able to handle at least 2,29 kPag (or 5.21 kPag based on darcy weisbach) caused by liquid pushing upwards on the floating roof. At this pressure the full inflow can be accomodated in the overflow line.

 

Next thing, is to check the froude number. As defined by P.D Hills it is calculated like this:

 

JL < (2 * h / d)0.5 = (2 * 0,234 / 0,3398)0.5 = 1,175

 

From this I draw the conclusion that the flow is not self-venting and air will be sucked in through the syphon break which in my application does not matter.

 

Any comments? Have I understood the problem and the theory correctly?

 

 

 

Attached Files



#2 katmar

katmar

    Gold Member

  • ChE Plus Subscriber
  • 663 posts

Posted 28 November 2014 - 06:40 AM

Equation 3 from the Hills article is for the case when the overflow outlet is at the surface level of the liquid, and the liquid overflows directly into the outlet. Your outlet is totally submerged so Equation 3 does not apply.  As far as I can see, you have applied the Darcy-Weisbach analysis correctly.  Your flows are quite high for the pipe diameters you have so I would recommend a substantially sized siphon break.



#3 Oliz

Oliz

    Brand New Member

  • Members
  • 7 posts

Posted 28 November 2014 - 06:50 AM

Equation 3 from the Hills article is for the case when the overflow outlet is at the surface level of the liquid, and the liquid overflows directly into the outlet. Your outlet is totally submerged so Equation 3 does not apply.  As far as I can see, you have applied the Darcy-Weisbach analysis correctly.  Your flows are quite high for the pipe diameters you have so I would recommend a substantially sized siphon break.

 

Thank you for your reply. So, i will use the height and pressure values calculated from the darcy weisbach equation instead of equation 3. The siphon break installed is 3". Is there a way to calculate the size required or is more based on rules of thumbs? I guess what limits you is that if the pressure drop in the siphon break is too large due to large amounts of air beeing sucked in based on the high froude number, the siphon break might not work and there is a risk for the siphon effect at least for a short period of time.



#4 katmar

katmar

    Gold Member

  • ChE Plus Subscriber
  • 663 posts

Posted 01 December 2014 - 10:32 AM

I do not know of a reliable and proven way of sizing siphon break vents, and I suspect that you are correct in that most are sized by rules of thumb or by gut-feel. There is a group at the Korean nuclear research facility that has done a lot of work over the years but their results don't seem to have reached the point of having a robust design method yet. A search on the term "soon ho kang siphon break" will give you some useful hits.

Usually I try to design this type of U-seal with the downleg being self-venting. In this case the liquid level in the downleg is fairly constant and very little air flows in or out of the vent and a relatively small vent is adequate. But this obviously results in large piping when the flows are large and this approach is not always practical or economic.

For large flows and non self-venting downlegs the way I size vents is roughly as follows - You want the siphon to break when the surface level in the tank reaches the top of the U-seal, i.e. when your dimension "h" becomes zero. In fact h will become slightly negative because for air to flow into the U-seal the pressure in it must be less than atmospheric. This negative value of h is called the "undershoot". If you select the undershoot that you want then this defines the pressure in the air space at the top of the U-seal. If the volumetric flow of air through the siphon break vent matches the volumetric flow of liquid in the downleg of the U-seal then no liquid would flow from the tank and the siphon would be broken.

The pressure drop over the siphon break vent is defined by the difference between atmospheric pressure and the undershoot pressure. The maximum volumetric flow of the water can be calculated knowing the height of the U-seal and the length of piping (and its fittings) - and the air flow would be the same in volumetric terms. Knowing the pressure drop and flow rate for the vent allows you to calculate the required diameter for the vent pipe.

Comparing my method with the results published by the Korean group indicates that this is a fairly conservative (i.e. safe) approach because in fact air is already started to be drawn in before the undershoot pressure is reached and you can get away with a smaller vent. As you make the vent smaller and smaller the time taken to break the siphon gets longer and longer until a size is reached where the siphon would never be broken. I have not found a way to predict the delay in breaking the siphon as a function of the vent size and so I use my conservative approach.

If you come across any other design method please share it here as this is a common problem, but one that has no clear answer.
 



#5 Oliz

Oliz

    Brand New Member

  • Members
  • 7 posts

Posted 08 December 2014 - 07:55 AM

I do not know of a reliable and proven way of sizing siphon break vents, and I suspect that you are correct in that most are sized by rules of thumb or by gut-feel. There is a group at the Korean nuclear research facility that has done a lot of work over the years but their results don't seem to have reached the point of having a robust design method yet. A search on the term "soon ho kang siphon break" will give you some useful hits.

Usually I try to design this type of U-seal with the downleg being self-venting. In this case the liquid level in the downleg is fairly constant and very little air flows in or out of the vent and a relatively small vent is adequate. But this obviously results in large piping when the flows are large and this approach is not always practical or economic.

For large flows and non self-venting downlegs the way I size vents is roughly as follows - You want the siphon to break when the surface level in the tank reaches the top of the U-seal, i.e. when your dimension "h" becomes zero. In fact h will become slightly negative because for air to flow into the U-seal the pressure in it must be less than atmospheric. This negative value of h is called the "undershoot". If you select the undershoot that you want then this defines the pressure in the air space at the top of the U-seal. If the volumetric flow of air through the siphon break vent matches the volumetric flow of liquid in the downleg of the U-seal then no liquid would flow from the tank and the siphon would be broken.

The pressure drop over the siphon break vent is defined by the difference between atmospheric pressure and the undershoot pressure. The maximum volumetric flow of the water can be calculated knowing the height of the U-seal and the length of piping (and its fittings) - and the air flow would be the same in volumetric terms. Knowing the pressure drop and flow rate for the vent allows you to calculate the required diameter for the vent pipe.

Comparing my method with the results published by the Korean group indicates that this is a fairly conservative (i.e. safe) approach because in fact air is already started to be drawn in before the undershoot pressure is reached and you can get away with a smaller vent. As you make the vent smaller and smaller the time taken to break the siphon gets longer and longer until a size is reached where the siphon would never be broken. I have not found a way to predict the delay in breaking the siphon as a function of the vent size and so I use my conservative approach.

If you come across any other design method please share it here as this is a common problem, but one that has no clear answer.
 

 

Thank you again for very good information. I read that article from the Korean research group which explained the phenomena very well (http://www.google.co....80642063,d.bGQ)

 

I will try to explain my thought process.

 

At the moment when the pressure drop in the U-tube is overcome, i.e. when the level in the tank is h meters above the U-tube centerline, water will start siphoning out of the tank. Due to the pressure drop, there will be a negative pressure at the top of the U-tube. The flow rate will be at its highest shortly after the start of outflow. The pressure drop due to flow of water will result in a negative absolute pressure at the top of the U-tube, decreasing in the beginning as the level in the tank decreases. In the article the process is divided in 4 stages. At the end of stage A, just when the inflow of air will start to have an effect on the pressure drop in the downward pipe, the absolute pressure will be at its lowest at the top of the U-tube. This pressure will actually be slightly lower than the 5.21 kPag calculated above due to the effect of lower liquid level in the tank.

 

Due to the negative pressure, air will be sucked in and become entrained in the liquid downward pipe and will result in larger and larger pressure drop in the downward pipe hindering the flow of water. At one point, the amount of air sucked in compared to the liquid flow is large enough to break the siphon.  

 

As a result of the research team not having a flowmeter on the siphon breaker to measure the air/water ratio when the siphon broke in the different examples I'll use your rule of thumb. 

 

As long as the pressure drop for the required air flow (in this case in 700 m3/h which is equal to the maximum liquid inflow) through the siphon breaker is lower than 5.21 kPag, enough air will be able to be sucked in to break the siphon with minimal undershoot only dependant on the transient nature of the backpressure in the downward pipe. In my case, the siphon breaker is a 550 mm pipe (ID =  107.1 mm) with two 90 degree LR elbows forming a 180 degree bend. I calculated the air flow @ a pressure drop of 5.21 kPag to 3,900 Nm3/h which is much larger than the required flow so based on this I think that the siphon is large enough for this case. 

 

If the installed siphon breaker is too small, the undershoot will not only be dependant on the transient nature of the backpressure in the downward pipe caused by air inflow but also on the level in the tank that will need to decrease to a certain level before you will reach your desired inflow of air.


Edited by Oliz, 08 December 2014 - 07:56 AM.





Similar Topics