air valves for the suction & delivery of a pumpstation

[waterpipe]

I am reviewing a big water pump station with the main suction and delivery steel pipes of 2000 mm (yes 2 meter)and 1400 mm respectively. There are five pumps in parallel and take their suction lines as branches from the mail suction line.

I asked for an air valve at the end of the main suction line (right before the pipe cap)and another at the beginning of the main delivery side, again before the cap. Now the designer has returned with the answer that there is no need for these valves as there are air vents on the pumps.

I have never seen a water pump station without these A.Vs before. I am concern for the filling and the proper air release especially at the pipe dead ends. I'm also concern for the negative surge at the discharge side and the A.V need.

looking to AWWA M11 manual, there is a recommendation to install air valves on the discharge of pumps. So here's the question:
Any other reference or standard to fortify my recommendation for having these valves?
Thanks in advance.

 

[BigInch]

Rather than looking for a source to backup your arguments, why not look for a reason to have them, or not. Its more interesting than looking for backups to arguments that are only baased on a "suggestion" and may not have any other legs to stand on.

I would assume that the pump vents would be sufficient for priming and starting and the air valves were only suggested for ease of remote autostart when pumps were without automatic vent valves, or no operators were present to vent manually.

Once the start question is decided, I would tend to only place air valves where air valves would be necessary. Offhand, I would think that air valves might have more typical need on the pump suction lines, as you may not have control of the take source and air may or may not be present, and air through the pump would also reduce efficiency. Once the air is removed at the suction, why would you need air valves on the pump discharge, provided that vents were there to handle priming.



"I am sure it can be done. I've seen it on the internet." BigInch's favorite client.

"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.live.com

 

[cvg]

water generally contains significant amounts of entrained air which can collect anywhere downstream of the pumps. However, I don't see any compelling reason to have an (automatic) air release valve on the end of the discharge header. Possibly more useful on the suction header due to lower velocities which might allow air to accumulate. Perhaps you can answer the question - is there significant amounts of air in the reservoir or the suction line and is there some reason air will tend to collect and cause problems in the discharge header due to your design?

 

[BigInch]

Also unlikely to collect air in the discharge, since the nature of the beast is simply that pumps go very near to the lowest points of most systems anyway. Doesn't do much good to place them where the discharge is at a high point that would tend to collect air naturally. I suppose it might do some good if you had a vortex sucking in air, but then that certainly wouldn't be a very good suction design.

"I am sure it can be done. I've seen it on the internet." BigInch's favorite client.

"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.live.com

 

[waterpipe]

Thank you. First let me put a rough sketch of the system (suction reservoir has higher level than the PS):

Taking BigInch advice to look for reasons, first the suction side:
I agree with BigInch that if it's a good intake design, there shouldn't be an air entrapment. But, considering that pump branches are from mid-span of the main suction, what would happen after the half pipe is filled? then we have trapped air on top of the main suction line which can not escape (700mm branch from 2000 mm pipe, so there is plenty space in main suction pipe on top of the branch level).
If air accumulates in the suction line during the operation (poor intake design, low pressure of water at suction side, etc.), again air stays on top (under pipe crown) unless very unlikely being sucked to the pump (from top to mid span of the 2000 mm pipe and then through the pump suction line).
I am also concern about the end leg, right before the thrust block, where air could potentially be trapped (there is a provision for future pump so the dead leg is long).
Then I am also thinking about the construction issue e.g. pumps are not delivered but the the line is ready for hydraulic test, etc.

Getting to delivery side, having an air valve (better to say air/vacuum) means a preventive measure in case of a failure of the surge tanks and occurrence of negative surge. delivery line is a "big inch" and long one (25km) which means high negative surge. I am a bit concern again since the surge vessels are outside around 150 meter away from the PS.
Pump trip, Check valves closure, flow still going, surge vessels 150m away and damn, it's a negative pressure. Better to be on the safe side with an A.V. (I'm going to ask for a detail surge analysis to see the 150m effect).

So, that's my argumentation. Comments are appreciate here.
Thank you all.

 

[BigInch]

Now we're getting somewhere.

Of course, if there is a way to trap air in the suction line, it would need to be vented somehow, either manually, or by some other means.

Surge vessels 150 meters away? You probably have more surge volume in your piping than in the vessels themselves. That's a fair piece away with a lot of fluid mass in there that needs to be accelerated too, which will tend to slow down the release of any surge pressure in your discharge line.

Are they surge relief vessels (one way flow), or are they surge vessels (two-way flow)?

The discharge line: 2m diameter. Will it be designed for full vacuum? If not, DEFINITELY use a vacuum breaker valve in the discharge line. You will most likely get a complete vacuum, ie. vapor pressure remaining only, surge when the pumps trip. If the surge vessels are full, you might could take water back from them to eliminate that negative pressure, or perhaps only from the lines going to them would be sufficient, in place of letting in air through a vacuum breaker valve. Your surge analysis should tell you if you can or not. Other than that, please have someone that knows what the surge analysis means, not just a "data entry clerk", doing the analysis. Ask for some hard evidence of experience, and/or some good questions.

"I am sure it can be done. I've seen it on the internet." BigInch's favorite client.

"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.live.com

 

[waterpipe]

Dear BigInch, Thanks and I fully agree with your sayings. I like your last sentence about the "data entry clerk"! well said.
The vessels are two-way, bladder type. You have seen that I mentioned some concern about them as well in a separate post.

Thank you again.
Nice weekend.

 

[cvg]

tangential outlets on either the suction or discharge headers can allow air to move through the system without being trapped.

A small valve for sampling, releasing air, initial filling can be installed on the dead legs. It might not need to be a combo / automatic valve. Although, I'm not sure what harm there would be if air was trapped in the dead leg of the headers, as long as it does not impede flow through the header.

 

[BigInch]

Do you have a link to those surge vessels? I would guess that they must be RATHER HUGE, given a 2m [Ø] pipeline.
What's the fluid velocity at the pump discharge?


Thanks. Today was absolutely beautiful, but they're saying rain for the next 4 days, which also happen to be a 4-day holiday weekend.

"I am sure it can be done. I've seen it on the internet." BigInch's favorite client.

"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.live.com

 

[waterpipe]

well, there's nothing in the design documents that I have regarding the real vessels but they should look like these:

http://www.youngeng.com/bladdersurgetankwater.html

http://www.charlatte.fr/2006/csae/index.htm

for the exact velocity at the pipe discharge, I should check it when I'm back at the office but the main discharge pipe is 1400 mm. 2m Ø is on the suction side.

Rain on a 4-day holiday weekend of BigInch could mean a lot for this forum!

 

[BigInch]

You got me!

I think you're going to need a whole lot more of those vessles. I'm sure there's no 1.4m Ø pipes anywhere near those pictures.

Note, you may need surge vessels on both suction and discharge sides, if the velocities are high.



"I am sure it can be done. I've seen it on the internet." BigInch's favorite client.

"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.live.com

 

[stanier]

Have you considered flywheels on your pumps to eliminate or ereduce the size of your vessels?

Have you considered the check valves on the pumps . prof ARD thorley recommends non slam check valves where bladder vessels are used. One pump trips and th reverse velocity and local deceleration will dictate their use. Noreva or Mokveld are generally used for such large stations.

website

http://www.ventomat.com

has some good technical papers on air valves and their location. Also you can get their application software for location and sizing of the air valve.