Flow Velocity?? 1

[MonkeyNuts2479]

Hi all,

I am designing a water circuit where water will be pumped around a cold water circuit. In order to protect the pumps a recirculation line will be installed where a % of the flow will be diverted in case there are any blockages downstream of the pumps.

Flowrate down the 2" main line pipe is about 10 m3/hr. I want to size the recirculation pipe for about 15% of the flow but need the velocity of the water in order to calculate the pipe size required for the recirculation line.

Can anybody advise? My gut is telling me this shouldn't be difficult but the answer seems to evade me!!! I have a feeling I need to two equations and then iterate them....

Many thanks in advance.

 

[BigInch]

10 m3/h * 0.15/(area of the hole in the pipe m2)/3600 = velocity m/sec

 

[MonkeyNuts2479]

Hi BigInch,

Many thanks for your prompt response. I'm trying to calculate the area of the hole in the pipe and therefore have two unknowns (velocity and C.S.A.).

I get the equation you have said but can't seem to get another so I can solve for the 2 unknowns.....

 

[BigInch]

Pick a pipe size, get the cross-sectional area and see if velocity is too high. If it is, pick another pipe size.

 

[micalbrch]

2" main pipe acc. to ASME, 47.5 mm inside diameter. If you use BigInch' equation, it will lead to 1.6 m/s velocity. 15 % of the flow is 1.5 m³/h. If you want the same velocity in the recirculation line you can now easily size it as the diameter is the only unknown parameter. I guess it is around 1".

 

[MortenA]

Monkey

You should propable select a pipesized based on your min pipesize (could e.g. be 50 mm (or 2") since thi is quite small piping. The use an orifice to limit flow (or a shut valve (i think they are called))

Best regrads

Morten

 

[davefitz]

The max permited velocity in the water pipe would be baed on erosion-corrosion effects, so for CS pipe a max V < 20 fps, for 1% cr alloy perhaps V < 27 fps, for SS or inconel V < 70 fps.

To limit the recirc flow to 15% , then the majority of teh pressure drop would be taken acrrss a valve or orifice or capillary tube.

 

[slyfox]

I agree with micalbrch . .You may assume a constant velocity flow since it seems that velocity requirement is not that crucial.

Say v = Q1/A1

where: Q1 (flow) = 10 m3/hr = 0.0027777 m3/sec
A1 (area) = pi (0.0508 m)^2 / 4 = 0.0020268 m^2
v = 1.37 m/sec

From this point, velocity is already known (1.37). . Follow the same principle using the same velocity and 15% of Q1 as your Q2 to solve for A2. . (you should arrive at a diameter of around 20mm). .

Also, the acquired velocity is quite conservative for domestic purposes based on my experience. =)

 

[kacarrol]

Granted I don't know all the details of your system but I'm just going to through this out there as I was asked a while ago to put pressure relief valves on a bunch of centrifugal pumps in a small (<3hp pump) water system because that was the way the company had always engineered their systems.

I questioned the need for the PRV's as they cost around $1500 each and came back with a couple of resources that said that centrifugal pumps under 5-7 horse power don't typically require this type of protection from dead heading (pipe blockage) in the type of system that was being designed. I don't have the particular pump textbook here for reference at the moment. I'll post up the title and page reference tomorrow if you're still interested.

I guess the point I'm trying to make is: have you considered whether this bypass is necessary (maybe consult pump manufacturer)? What are the chances that the system will dead head?

These two threads are related:

http://www.eng-tips.com/viewthread.cfm?qid=196654&page=7

http://www.eng-tips.com/viewthread.cfm?qid=107205

If it turns out you don't need them you could always tell your boss you saved them some money!

Regards,
K

 

[rcchap]

kacarrol is correct. You have to check with the pump manufacturer to determine the required minimum recirculation flow for the pump. You may not need one for a pump this small.

Also, if this is a closed loop chilled water system, I would think that blockage is unlikely. You are primarily concerned with someone shutting a valve, which in itself does not require a recirculation line unless there is an automatic isolation feature or an air operated control valve that could fail closed and isolate the flow.

In any event, if the pump manufacturer does provide a minimum flow requirement, then the pipe size of the recirculation line is much less important than the flow restriction device you must install, either an orifice or a throttle valve (as stated by MortenA above).

To do the sizing, you will need to use Darcy's formula, which relates flow, DP, and resistance coefficient (Cv or K). If the recirc pipe run is fairly short, then most of the pressure drop will be through a valve or orifice, and you can determine the required Cv to generate the desired recric flow. If you haven't done this before, get a copy of Crane Technical Paper 410, and it will give you all the info that you need (and more). Using a throttle valve will give you the ability to adjust the flow, as long as you can measure the flow rate. If you decide to use an orifice, which is more work up front but probably better in the long term since it can't be mispositioned like a valve, you might want to be a little more precice in the calculation by considering the pressure drop in the piping and fittngs as well. Then you can go a step further and iterate the calculation since the recirculation line will affect the pressure in the 2" piping, which will change the DP in your equation.

Finally, you also need to make sure that the loss of process flow through the recirculation line does not adversely affect the operation of the system.

 

[MonkeyNuts2479]

Hi all,

Many thanks for all the info. I have managed to get the problem simulated by a mate of mine and I think I'm going to go with an orifice plate to restrict my flow.

Many thanks again.