Chilled water Increased flow in 1-1/2" pipe

[jtimmer]

I am working on a system that has a 1-1/2" common header from 3 individual pumps. We didn't do the original install but the 9 air handlers that were originally selected didn't perform so we re-did the heat load calculations for the new units and the chilled water to nearly all of the units has increased. The total GPM we are now looking at with a full load is close to 110 GPM. We have plenty of head from the pumps, I think the inlet was around 30 and the outlet sits at about 70ish on the original system. This is a constant flow system with 3 way valves at each air handler. I am thinking with the increased flow the velocity and pressure loss will obviously increase but the 1-1/2" header should still "work" its not ideal I know, but the customer doesn't want to have to rip out a bunch of piping to upsize to a bigger header as space is at a premium as well. What would the negative effects of leaving the 1-1/2" piping in vice up-sizing be, any thoughts would be appreciated.

Side note there are two air handlers near the pumps (within 30ft of pipe or so) that have about 30 GPM running through them, so the main section of header will run closer to 80 GPM after these branches.

Thanks

 

[Artisi]

The friction loss thru' your system will increase as the square of the flowrate:
Q2/Q1^2. Not knowing the original flow you will need to calculate the new losses.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[LittleInch]

The negative effects are basically that the if the pump is still the same, then the flow rate may not increase to your units further away.

all depends on how much "spare" the systems had and where the 3 way valves went - are they modulating or on/off?

YOu say the total flow is now close to 110 gpm. What was it before??

A network diagram or schematic with before and after would help a lot to understand what has changed and whether the flow in the system is actually higher or not.

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

 

[jtimmer]

Thanks for the responses. The 3 way valves are on/off so the flow is either going through the coil and then to the return or straight back to the return. Both routes run through a constant flow fitting in the return line.

I will have to look up the numbers to be sure, but the original flow was closer to 90 GPM so its about a 20 to 30 GPM increase.

 

[EdStainless]

You should estimate the flow velocity in the piping.
If it is too high then you are wasting power and maybe eroding the piping.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[Artisi]

if original flow was 90 and you are planning on 110gpm then the increased friction component of the installation will increase by:-
110/90^2 = 1.49 which is a 50% increase - lots of luck.


It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[bimr]

One of the reasons that the original system probably did not perform as expected was that velocity was too high. Now, you are increasing it. Bad idea.

In addition to what the others said, you can also expect problems with noise and water hammer at the proposed velocities.

 

[jtimmer]

Thank you for all the responses, I did some more digging and the original flow was about 85 while the new flow will be about 113. The header pipe size is actually 2 inch which I am more happy with, still not ideal if we were doing this brand new. The problem we are trying to correct wasn't really with a lack of flow to the AHU's, but with the coils themselves. The AHU's are being replaced and we wanted to look at the pipe sizes and pumps as well since the flow will be increasing.

 

[Artisi]

Does not change the fact that the head loss from friction is still the square of the change, and in the new scenario it is even worse.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[LittleInch]

Apologies if you've done this, but to see whether it's going to work you need to calculate all your losses to and from the furthest AHU. Even a 2" NPS pipe at 113 gpm is about 3.5m/sec so at the edge of normal acceptance. You need to calculate all the sections with their branches and gradually reducing flow rates, then the increased (or maybe decreased) pressure drop across the new AHU, increased PD across the flow valves ( assume fully open), then the increased PD in the return line ( also 2"?) - but should be equal to the losses in the inlet system - then the extra PD across the chillers.

Is it a 2" NPS pipe or 2" OD?

Only then can you work out if the pressure / head differential of your pump, which could drop by 10-15% depending on the pump curves, is sufficient for your new load. The length of the pipe to the furthest AHU could be critical.

It's the system that counts, not just this bit of pipe.

Let us know how it gets on.

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

 

[Artisi]

As all the data is not available to us, we can only make a few assumptions -- the OP said that the pump discharge pressure was 70ish, so if the flow is now increased from 85 to 113gpm it is fair to assume without any other info the new total head on the pump is likely to increase to 123ish psi at 113gm.

jtimmer, can you show the pump curve?

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[LittleInch]

Artisi - you missed the differential - inlet 30, outlet 70 so differential 40 (psi? ft? m?)

Not all the increase comes from the pipe so increase in required pressure isn't as much as a pipeline. my point about pump curves is what ids the differential head/pressure at the new higher flowrate? It won't go up....

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

 

[bimr]

As LittleInch posted, the updated flow information shows the fluid velocities are at the edge of normal acceptance and will probably work.

However, you need to determine if the pump has adequate head to pump against the higher system head caused by the increased velocity.

Note that the pumped flow will decrease as the system head increases. The change in pumped flow can be determined by plotting the system curve onto the pump performance curve.

 

[LittleInch]

It all depends on how much head is being lost across the constant flow valve at the furthest part of the system. If this is a small number/percent, then you may not have enough spare in the pump to cope with the extra head required. If its a big percent of the pump differential head then you might just get there.

For this increase in flow, if the original designers did their job properly and optimized the pumps and flow control then you may not have enough and may need to swap out your pumps for some bigger ones.

Let us know how it goes.

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