HEAT EXCHANGER SYSTEM CURVE

[mmmumuw]

Perhaps I missed this class while in college ...
I understand that the actual pumping capacity (and head) is the result of intersection between pump curve and system curve.
there is a pump supplying cold water for a typical heat exchanger process (cooling). By the ampere meter reading, we can point the flow at 5-10% in the left of BEP. It is 50% valve closed.
I can get why after we open the valve fully, the pressure get raised (4 bar fro 3.5 bar).
Has it got something to do with Heat exchanger system curve behaviour? Can you engineers explain, please...
Many Ths,
MW

 

[BigInch]

It ALWAYS has something to do with the system curve behavior.

I would guess that, as you open the valve, flow increases.

Increasing flow in the system requires a higher driving pressure. Let's say your pump is producing 4.5 barg at its discharge flange at flowrate₁. The valve, when partially closed, maybe provides provides a large pressure drop of 1 bar between the pump and HX, thus the inlet pressure to the HX is 3.5 barg at flowrate ₁.

When you open the valve a lot, its pressure drop might be something like 0.25 bar, even though it is passing more flow.

Your pump at the higher flow is producing only about say 4.25 barg and subtracting the valve's new pressure drop of 0.25 bar = 4.0 barg at the HX inlet.

"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

 

[rmw]

Actually, the OP doesn't say where the pressure reading is taken. If it is taken at the Hx downstream of the 50% closed valve, then yes, when the valve is opened, the Hx will see full pump discharge pressure (ignoring losses) but if the pressure reading is taken at the pump discharge, I would expect the pressure reading to drop off as flow increases, assuming that the DP of the Hx doesn't exceed the DP of the valve at 50% plus the Dp of the Hx at flow = 50% valve open.

rmw

 

[mmmumuw]

Thanks for the reply.
The pressure reading is at the pump discharge.
The pump flow/pressure always follows the performance curve line, right? The curve is rather steep. So if the pump flow increases, the pressure must be less, i.e opened valve should release some numbers of friction losses.
Unless, the pump can work outside the curve line, or it is possible that opening valve make the flow decrease (?)

 

[BigInch]

What kind of valve is that??? Broken?

If your pressure is at the pump discharge, I would have to guess that your increasing flowrate is causing a reduction in suction pressure at the pump, which in turn is causing the overall discharge pressure to reduce too.

You'll need to post flowrates, suction and discharge pressures and the heat exchanger inlet pressures at both flowrates to confirm. Why not post a diagram of the system and the pump curve too.

"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

 

[ione]

Your valve seems to work the opposite a control valve normally does.

Opening the valve should reduce minor losses producing a clockwise rotation of the system curve around the point of intersection between the system curve and the vertical axis (zero flow). Result = higher flow rate and lower discharge pressure.

 

[mmmumuw]

My favorite quotes from BigInch:
"What kind of valve is that??? Broken?" Hahaha
Made by BROKEN valve Maschinenbau ...
The installation should be simple. It is just cold water pump to filter, to HX and get back to open tank and go to pump suction.
O ya, the opened valve bring the raise of ampere. So, the duty point is now moving right and flow should be increased.
Pump suction, as the tank is atmospheric, no level change are observed, we can assume that no change in suction pressure.

If the ampere reading is wrong, can the scenario be like this:
1. open valve ==> bigger flow ==> filter overcapacity ==> higher pressure loss ==> higher pressure/lower actual flow.
2. half open valve ==> flow matches the filter capacity

Is it possible? Opened valve makes the pump lose its flow. Seems funny.
I will try to collect the diagram, or picture.

 

[25362]

I wonder whether, two-phase flows, for example, water+air, or some dirt moving around and plugging, could result in erratic behaviour of [Δ]P in heat exchangers. BTW, what type of heat exchanger is being considered ?

 

[BigInch]

No. If suction pressure remains the same, higher flows are accompanied by higher head losses and unless you have a flow element causing some discontinuiety, such as a broken valve, a ripped membrane opened more at high flow, etc. It would be "very unusual" to have any typical flow element giving a lesser direct pressure loss with increasing flow without crossing from laminar to turbulent flow or some similar special circumstance affecting the results. All typical flow elements have a quadradically proportional and increasing head loss when related to flow of (a newtonian fluid), something like dH = k * Q^n. Your filter over-capacity scenario does not explain why that typical relationship would not hold true. Normally you would expect some filter compression and even higher head losses with an even higher total differential pressures. Higher flows through a normal "uncontrolled" flow elements always cause higher differential pressure losses.

Reading a lower gage pressure anywhere in the system is caused when relating the total pressure loss (sum of all differential pressure flow losses of each element) to the reference pressure; i.e. probably the same as the gage's atmospheric pressure, which is why it leads me to think it's a problem in the suction side. Something is affecting the boundary conditions. Either upstream suction pressure, or downstream discharge pressure, on the other side of the HX, is not being held constant. Something is forcing a reduction in absolute pressures, even though the sum of all differential head losses is probably still increasing.

If opened valves cause loss of pump flow or head, you should usually be looking for a loss of suction pressure caused by too high a flow in the suction lines with possible cavitation, or perhaps vortexing, reduced pump efficiency from 2phase flow air ingestion, or some other kind of turbulent losses going on in the upstream half of the system.

"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

 

[25362]

Biginch, who said that the flow has increased upon opening the valve ?

 

[BigInch]

Nobody, but that must be what wants to happen. It for some reason does not, which is the problem we're trying to solve.

"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

 

[25362]

You're absolutely right. Since we do not have a clear picture of the installation, we are free to imagine. Aren't we ?
Imagine the partly open gate valve keeps upstream a big air bubble that is released when opening it fully. Could this bubble introduce an additional "permanent" [Δ]P of a 0.5 bar at the heat exchanger ?


"Logic will take you from A to B. Imagination will take you everywhere." Albert Einstein

 

[BigInch]

Ignoring pipe, valve 50%

1.) P_{pump suction} + [Δ]P_pump = P_{pump disch} = 3.5

2.) P_{pump disch} = 3.5 = [Δ]P_valve + [Δ]P_hx+ P_{hx disch}

3.) [Δ]P_pump + [Δ]P_valve + [Δ]P_hx = [Δ]P_sys

Either, valve losses are reduced more than the differential pressure increase you'd expect in the HX plus the decrease in differential head you'd expect at the pump, or something is happening to the boundary conditions; pump suction pressure, or the HX discharge pressure. The first possibility is more or less obvious, so I chose to warn him of the reduction of suction pressure with increasing flow, as that might not be so ovbious and perhaps a prelude to cavitation. But it's still speculation, as there are a lot of question marks in 1,2,3 above, amongst those are both changes in boundary conditions AND caviation, if not others. We don't even know if the system is on the open or closed side of the HX, or if either side is open or closed.



"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

 

[ione]

If I am not mistaken, the cold water flows through the HX and then it is recirculated back to the suction side: : "It is just cold water pump to filter, to HX and get back to open tank and go to pump suction".
Increasing the flow as result of the valve opening should alter the HX performance, being consistent the other HX's parameters. What I mean is that the increased water flow should exit the HX at a lower temperature than that of the throttled flow. A variation in temperature would lead to a different density thus affecting the discharge pressure.

 

[BigInch]

Seems like a big change in density for water.

Actually I'll let you continue to speculate by yourself. I've had all I want of speculation about this problem. More real data is what is needed.

"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