Understanding Pressure Loss in Closed-Loop Cooling Systems

[gravitate]

A certain part of a closed loop cooling water systmem has to have some of the pipes made to have a larger diameter to allow more flow. Does this mean though that the overall pressure in the system will drop? I know the headloss in the widened pipe will decrease and flow increase. But will you need to add an additional pump to the system?

Also for another cooling water system we have to add a heat echanger which is also connected to another system another system. I take it this will also need a pump to accommodate for the head loss in the pipe and heat exchanger?

 

[Travis_King]

System pressure drops due to change in elevation and dynamic losses (say from pipe friction or valves). You won't need a bigger pump because you changed pipe diameter.

Heat exchangers can add large head losses to systems, depending on the arrangment. We don't know your pump, so we can't tell whether a heat exchanger would create enough head loss to dead head the pump.

 

[gravitate]

Thank you for your reply. But I am still unclear.

YOu have a pipe with a specific flow rate. You want to increase the diameter of the pipe to increase flow. Where is the additional energy coming from to increase the flow? Surely the pump will have to work harder?
Or is it because you have decrease in headloss that compensates for the increaser of flow?



I remember at uni I could not get my head round this stuff. Still can't so it seams !

 

[Travis_King]

So, brief review of centrifugal pumps (bear with me if you know this, I'm just trying to cover bases, and it might be helpful to someone else looking on):

1. They run on curves based on the head that must be overcome.
2. That head is dependant on three main things: Elevation change between suction resevior and pump suction, elevation change between pump discharge and final discharge, and pipe friction losses.

Because you have a closed loop system, the first two don't really matter at all. What you are worried about is the losses in the system. So for this you have your valves, your bends/elbows, your pipe friction due to velocity, and any equipment which will add to your overall system losses (for instance, that heat exchanger).

So assuming your pump is not on a variable speed drive (is it?), it will churn out a given capacity (gpm, m3/hr, whatever) based on whatever losses are in the system, and where those losses put the system on the pump curve.

Changing the diameter of the pipe does one thing: changes the velocity of the flow. 4" pipe will allow just as much flow as an 8" pipe, it'll just have to go faster. Now, what one means when he misleadingly says "8 inch pipe allows more flow than a 4" pipe", what he means is that an 8" pipe will result in lower friction losses through the pipe run than a 4", since the flow velocity will be lower.

Now, changing pipe diameter does not alway have a significant effect on flow capacity. If you are pumping 200 gpm through a 6" line, you'll see around 2.25 ft/s as a velocity, in Sch. 40 CS pipe, this amounts to around 0.3 ft/100 ft of straight run (give or take).
If you bump that up to an 8" line, you'll see something like 1.3 ft/s, and .075 ft/100ft.
In a system with 500 feet of pipe (let's call it straight pipe for now, it gets hairier with elbows and valves, but the point is still valuable), your difference in overall system head is only 1.125 ft of head. For most pumps, this is an almost insignificant amount. Certainly not enough to warrant the extra expense for the 8" pipe over 6", right?

However, for, say, 800 gpm in those same lines you are looking at:
6": ~9 ft/s and 4 ft/100 ft head loss
8": ~5 ft/s and 1 ft/100 ft head loss.

Now, over your 500 foot pipe run, you've got a difference of 15 ft of head. Now that's pretty noticeable!

So it really depends on the system (pump curve, number of valves, branches, bends, equipment, etc).

What I'm getting at, somewhat long-windedly, is that simply increasing the pipe diameter will not necessarily give you noticeably higher flow.

Adding a heat exchanger, however, will change your flow significantly. Again, we don't know the system conditions, so it's hard to say if the heat exchanger will require a pump. Some process water lines run at very high pressures (high head systems) and depending on what's downstream of the exchanger, may not need an additional booster pump. Others will, especially when you are adding a heat exchanger to an existing line, as the main line pumps were probably not sized to accommodate future equipment loads.