EXCEEDING MAXIMUM FLOW RATE

[imagineers]

ok so I have a plate to plate heat exchanger,

http://www.mcmaster.com/#35115k62/=wzpg0d

it lists the flow capacity at 6gpm. my question is what happens if i try using a 30l/min pump to push fluid through it? I guess for that matter what happens if you try to exceed the flow rate on a pipe in general?

 

[EmmanuelTop]

You'll get more pressure drop through the exchanger (and through piping as well). Theoretically you could increase the flow up to a point when hydraulic resistances (head losses) equal the differential head developed by the pump - if centrifugal type.

It all depends on how the entire circuit looks like. If you can share a sketch, we can comment further.

Dejan IVANOVIC
Process Engineer, MSChE

 

[bimr]

For piping, it depends on the velocity of the fluid and the material of the pipe. This heat exchanger has copper tubing fittings. Copper tubing is a soft material and more sensitive to erosion than other material. If you get above 8 ft/sec velocity in copper pipe, the pipe will erode. With other piping materials, this is generally not an issue, you just have more headloss and cost of pumping.

For the heat exchanger, it seems that it is made of stainless steel, so you probably will not have an erosion problem at your flow rate. The heat exchanger made not have enough heat transfer area to meet the guaranteed heat transfer performance requirements because the water is moving through too quickly.

 

[imagineers]

the pump is here.

http://www.iwakiamerica.com/literature/md_wmd/IALT00034_MDBrochure.pdf

basically fluid coming into the pump and going through the heat exchanger then out again.

Here is a comparison heat exchanger that also can remove about 75000BTU/H and shows a 3psi pressure drop at 7.5gpm (28 l/min) (page 2).

http://www.brazetek.com/docs/pdf/brazed-plate-heat-exchanger-quick-sizing.pdf

So now If I am trying to push fluid through at 40 l/min and the heat exchanger is rated for say 6gpm (23 l/min), what will happen??

 

[EmmanuelTop]

Without any other controls, the flow developed by the pump will balance at the point where differential head developed by the pump equalizes with head losses in the circuit (piping, exchanger, fittings, elevation difference etc.). Assuming this is equivalent to 40 l/min (I'm saying "assuming" because it is impossible to say what the actual flow will be), you'll get approximately 3 times higher pressure drop (~9 psi) across the exchanger and slightly improved heat transfer (more BTU/hr) - as compared to the base case with 23 l/min.

Dejan IVANOVIC
Process Engineer, MSChE

 

[imagineers]

so basially even though the exchanger is rated at 6gpm with roughly a 3psi drop at that rate, WE COULD increase it to 11gpm but would just increase pressure drop? is there a way to calculate what that increase in pressure drop would be? Also, I am assuming there is a point where the flow will just not go through?? I mean I am sure I could not force 500gpm through this exchanger? if that makes sense?

 

[bimr]

The manufacturer should have a published head loss curve.

 

[EmmanuelTop]

You will increase pressure drop and also, as bimr noted, you may get well beyond erosion velocities which will ultimately damage the exchanger and/or piping components. It also depends on the pump how much more flow you can develop. The pump has to follow its Head-Flow curve.

Pressure drop is proportional to the flow squared.

Dejan IVANOVIC
Process Engineer, MSChE

 

[bimr]

If you exceed the design flow, you will not achieve the design ∆T across the heat exchanger. ∆T will be less than design.

∆T (F) = Change in temperatures between inlet and outlet of the fluid

 

[imagineers]

erosion velocities? I would imagine this would have to be quite high for stainless??

 

[imagineers]

Maximum Recommended Water Velocities for stainles steel is 15ft/second

 

[imagineers]

no idea how to calculate that from flow rate of 30l/min

 

[MikeHalloran]

Ignoring erosion and any effects beyond second order, you can use Cv to estimate pressure drop vs. flow, given one such data point.

calculate Cv for heat exchanger		
6	gpm	flow
3	psi	drop
1.7320508076	root(psi)	root(drop)
3.4641016151	gpm/root(psi)	Cv
		
find drop for higher flow		
11	gpm	new flow
3.4641016151	gpm/root(psi)	Cv
3.1754264805	root(psi)	new flow/ Cv
10.0833333333	psi	new dp
		
find drop for super high flow		
500	gpm	new flow
10.0833333333	psi	Cv
49.5867768595	root(psi)	new flow/ Cv
2458.8484393143	psi	new dp

Mike Halloran
Pembroke Pines, FL, USA

 

[georgeverghese]

Please note that this 3psi drop thru the HX for the cooling water side at 6gpm would be for a fully liquid flooded cooling water side - it is not evident how you've oriented this HX or which way you intend to flow this cooling water in order to fully prime this HX

If the arrangement results in a air trap inside the HX on the cooling water side, then the pressure drop at 6gpm would be some higher unknown value. In order to setup the HX properly, you also need to know how the cooling water flows through the passes in this HX.

Which nozzles are used for cooling water, and which is inlet and outlet ?

What about the refrigerant glycol side ? That side will also need to be fully liquid primed. Do you not need to increase flow of glycol refrigerant glycol in tandem with cooling water flow ? In any case, the cooling duty will not be 2x that of the original design cooling duty, since the UA value for this HX is fixed.

If you increase flow from 22 litres/min to 44 litres/min on the CW side for this plate HX cooler, then exit cooling water will not be as cold as for the 22 litres / min case.

 

[imagineers]

thanks guys

 

[MikeHalloran]

There was an error in the third section of the spreadsheet copied above, i.e., used the wrong Cv:

find drop for super high flow		
500	gpm	new flow
3.4641016151	gpm/root(psi)	Cv
144.3375672974	root(psi)	new flow/ Cv
20833.3333333333	psi	new dp

Sorry.


Mike Halloran
Pembroke Pines, FL, USA