High velocity in heat exchanger tubes 2

[tim02]

In normal the operation of our cross flow heat exchanger the rate of heat transfer goes up proportionally to the cold side flow and we control exit temperature that way with a PID loop program operating the cold side throttle valve. But we seem to be able to reach a condition when the flow rate reaches a certain point the rate of heat transfer starts to drop, so the throttles open more and the rate goes up more so the rate of transfer drops even more and continues into an out of control situation.
I seem to remember something from an Advanced Fluid Flow course (about a zillion years ago) where high velocity flow drastically reduces efficiency but I don't remember any details or even what the phenomenon is called.
Can someone explain to me the mechanism by which cooling water flow above a certain velocity in the cold side tubes will drastically reduce the heat transfer ability? And what solutions either in design or controls do you suggest?

 

[JOM]

tim02,

increasing velocity should increase heat transfer coefficient, not decrease it; but you know that...

a dumb question - how do you know the heat transfer rate is going down? by the outlet temp on the hot side?

it's just that the exit temp on the cold side may decrease with increased flow, but the heat transfer is increased...

is there any possibilty of fouling occuring on the hot side?

just some thoughts, no solution for you Cheers,
John.

 

[hacksaw]

At the risk of misunderstanding your question, the minimum approach temperature for your exchanger design, limits what the control valve can accomplish.

 

[ardilesd]

Think that JOM's approach is correct. If you guess that heat transfer is going down (because exit temperature tends to decrease), your controller should reduce the flowrate, not increase it !!.
There is a first order effect in the relation Flow-temperature and a second order effect in the Heat transfer coefficient with flow (because the other side and delta T also count), so , only the first should be used.
Regards

 

[tim02]

Let me clarify some more.
The hot side exit temperature starts to climb as the flow rate goes up.
The cold side exit temperature (and delta T) goes down.
The more the cold side control valve opens to increase flow in attempt to bring hot side (process control) temperature back down into spec., the higher the hot side exit climbs and the smaller the delta T across the cold side until it is almost zero. The normal control response only amplifies the problem and rapidly drives itself out of control.
We take manual electric control of the throttle valve and forced it more closed which drives the delta T on both sides back up, and process exit temps come back down with less flow.
I have been told by a corporate facilities engineer that there is a phenomena that occurs at very high cold side water velocities in the tubes that can cause an inability to transfer heat from the tube wall to the cooling water. But he didn't have any more than annecdotal information and I cannot seem to find any documented corroboration. But I also can't find any other explanation of this flow to heat transfer proportionality reversal effect

 

[hacksaw]

tim,

if you can control it in manual then your controller settings need attention. inverse acting processes do exist. they are characterized by an initial dynamic action opposite to the steady-state behavior. without a very complex control scheme about all you can do is to detune the controller and accept the setpoint error.

restated: if you have a tight control objective on the hot side exit, then the control gain, integral, and derivative are set for rapid response. this can destabilize the valve operation.

suggest that you try, turning off the derivative and only using a little bit of integral. the controller gain should be 1 to 2 (to start with).

you can also add valve limits to precent the valve from going full closed: if the valve closes then you lose controllablity until the hot side exit temperature is returned below setpoint manually - has to do with so-called reset wind-up.

 

[vpl]

Tim

This is going to sound a bit dumb, but you've made sure your controller wasn't wired up backwards, right? Sometimes that happens and the sensor is valiantly trying to tell the valve to close, but the wires are literally crossed....

it pays to check out the simple things before looking into the weird and unexplainable. Patricia Lougheed

Please see FAQ731-376 for tips on how to make the best use of the Eng-Tips Forums.

 

[tim02]

Thanks Guys,

Hacksaw, In manual we are closing the control valve to get more cooling which is intuitively backward but works. until we get back into a lower flow range and then starts working proportionally again.

Patricia, it's never a dumb question unless you don't ask it and check it out, because it will end up being the problem, right? But we did ask and we have checked. They are all wired in the correct direction.

We have several of these heat exchangers with 2 parallel, independently throttled, stages in each HX and they all drive into this situation on occasion. And it's always the hottest stage that goes first.

But it's beginning to seem like no one else has heard of this supposed "high water velocity - inverse thermal efficiency" effect.

Tim

 

[JOM]

Tim.

Looks like a job for Sherlock Holmes (or Mulder n Scully).

Just to be crystal clear, can you tell us the typical temperatures on both sides at both inlets and outlets when the HX is performing as desired? Also, what are the two streams? What is the internal arrangemt - no of passes on tube side?

>But it's beginning to seem like no one else has heard of >this supposed "high water velocity - inverse thermal >efficiency" effect.

My thoughts are that whatever phenomenon was occurring was been masked and the observations, taken at face value, led someone to jump to a wrong conclusion. "There's a logical, scientific explanation, Mulder."

We musn't abandon our confidence in the basic heat transfer principles. Increasing velocity will increase heat transfer co-efficient - believe it. If somehow the oppposite occurs, there will be an axplanation, but it won't be a reversal of the basic principles.

In milk pasteurisers (a heating job), milk proteins deposit on the HX surfaces, heat transfer reduces so the controller calls for higher temp of the heating stream, the protein deposition gets worse cos of the higher temps and the controller calls for more heat....unsustainable and the HX eventually has to be cleaned.

I know that's not your situation, but I just use it as an example to show that something other than simple heat transfer can be at work.

Smart of you to ask for fresh ideas. Have you ever found the harder you stare at a problem the more the solution avoids you?

I know of an example of a HX behaving opposite to plant personnel's intuition, and the answer was surprising and only discovered through expert analysis. I'll see if I can write it up in detail. The answer was in leaking tubes which no-one knew about. When was the last time your HX was inspected internally?

C'mon, Scully, let's go.
Cheers,
John.

 

[hacksaw]

Tim02,

The inverse action you have described is normally found in reboilers and other such equipment where a phase change is possible. That problem is well known.

What is unique to your problem is an apparent lack of phase change.

What is the fluid state of the hot stream and where is the temperature sensor located realtive to the discharge of the exhanger?

 

[berry]

hi guys,
in my opinion,there can be a turbulance effect due to high velocity which can make the heat transfer goes down.
may be you know in many cases, this effect should provide the heat transfer to increase up to intended values.
but,according to high velocity,there can be no useful attitude for the water or generally let's say refrigerant to release its thermal energy to go out.
we have to check the pressure drops and piping diameter if it is enough for the heat transfer within the sufficient turbulance effection.

 

[25362]

Hacksaw may have indeed hinted at the possible reason of the strange cooling water behaviour when he speaks of phase changes. On this line of thought I'd venture:

If the H/E is located sufficiently high so as to create a vacuum at the water return nozzle, air may be released from warm water.

The geometry of the exchanger may be so that at higher water superficial velocities, we get some kind of segregation, so some tubes may be "seeing" a water-air mixture, and others only water, with a consequent drop in HTC.

Don't take this idea too seriously, it's just an hypothesis easy to be checked.

 

[reactorshell]

25362,

I do have the similar sentiments as u, that the possibility of a reduced HTC as a result of some vapour-liquid mixture within the HX. In fact, I am in the midst of doing some testing on this possibility for one of the HX in the plant i'm working in. Any comments on this subject, anyone, because i'm definitely not an expert on this.

 

[quark]

I too had a problem of that sort once in our vapor absorption system condenser. What I thought at that time was that the contact period between the two fluids was less(it was a guess and was not backed by any calculation). I controlled the valve towards closing position and the thing worked.

Yet I agree with JOM that the real occurence might have been masked for I still think my idea was a weird one.

 

[JOM]

I want to upload an unusual heat exchanger behaviour case, but it really needs a diagram.

Does anyone know how I can insert a .gif file in the message. I've noticed .jpg pix in other threads. How do you do it? Cheers,
John.

 

[quark]

Perhaps this could be of help if the picture is on a webpage.

[img]http://www.sitename.com/picture.gif[/img]

Just above Submit Post button there are 3 options. Process TGML is one option where you will find some tags to put into your post for subscript, superscript, bold etc,

Regards,

 

[25362]

While trying to find a reasonable solution to this puzzling question I imagined the H/E as a square box. tim02 speaks of a (one-pass?) crossflow exchanger. In these exchangers both fluids usually enter and flow at right angles to each other, and tubes sometimes have perpendicular radial fins to enhance heat transfer on the shell side.

As a result they behave in a strange manner: the shell side unmixed fluid crosses the unit leaving it with a profile of temperatures, showing a coldest "corner" where the outlet of the warm fluid crosses the incoming cold water, and a warmest corner at the other shell outlet end.

Water also leaves with a profile of temperatures, the highest at a corner diagonally opposed to the "cold" corner of the shell side fluid.

This translates in a peculiar delta T map: two diagonally opposed corners with minimum delta T (the "cold corner" and its diagonally opposite) and the others with a maximum delta T.

Because of the particular geometry, the large number of tubes, and typical header arrangements, water distributes unevenly between the middle tubes and those at the periphery.

The overall coefficient of heat transfer influenced by the shell side fluid is low, and the increased water velocities don't improve it much. If we admit that the increased water flow rates asked for because of higher oulet shell fluid temperatures, tends to re-distribute among the tubes, we may find that the overall heat transfer drops even when the outgoing mixed fluids show a larger delta T.

This is only an hypothesis trying to show that there is nothing wrong with heat transfer theory and that hydraulics may be the reason for this fata morgana trying to lead us astray.

 

[hartof]

Hi ,

I just also want to write down my propably stupid thought,
can temperature increases due to high friction between water and tube wall caused by extremely high velocities ?

 

[JOM]

hartof,

that's a brilliant thought, not stupid at all

basic physics says there ought to be friction-generated heat, but I've never, ever heard it mentioned in fluid flow

my guess it would need extreme velocities to have a noticeable effect Cheers,
John.

 

[25362]

Especially when timO2 says that water comes off cooler at high rather than at low flow rates.

To me it seems that somehow cooling water doesn't seem ""to see" all the warmer tubes at high (beyond some limit) flow rates as it does at lower flow rates.