Minimum pitch temperature

[Yobbo]

Dear readers,

I looked at the FAQ's and the historical discussions, as well as in some text books about heat exchangers, but I couldn't find a substantiated answer to my question. My question is the following:

I want to rate an existing heat exchanger for a somewhat different application. I have no dedicated software for this at my disposal. I use the e-NTU method for this.

My question is the following:

"What is the minimum pitch temperature that should be applied for a heat exchanger and on what grounds is the limitation of this minimum based?"

One textbook recommends 5 deg C. I would like to know though what the grounds are for this limitation? I mean a guide line in itself is no reason for a limitation.

I hope someone in this realm of heat transfer experts is able to lift a little corner of the veil.

Thanking you in advance for your attention and your effort I remain.

With best regards,

Karel Postulart

Karel Postulart, The Netherlands
Nuon Power Generation

 

[EmmanuelTop]

I believe you are referring to the "pinch" temperature, which is by definition the minimum temperature difference between the hot fluid and the cold fluid, anywhere along the heat exchanger. It can be at the cold-end or at warm-end of the exchanger, but also anywhere between these two physical locations.

Pinch temperature depends on the exchanger design (type), heat transfer area, and geometry / flow pattern (approach to the ideal counter-current flow). For the same flows and inlet conditions of hot and cold streams, a 1-pass shell & tube exchanger will have different pinch temperature than a 2-pass or a 4-pass S&T exchanger. Spiral heat exchanger will have different performance than shell & tube exchanger, as well as any other exchanger type/design. For example, TEMA E-shell exchangers are typically designed with 15 degC temperature difference.

In design of heat exchanger networks, the "pinch" temperature has different meaning. There is a lot of information online.


Dejan IVANOVIC
Process Engineer, MSChE

 

[IRstuff]

The exact value of the limiting temperature difference is basically arbitrary, but the closer you get to zero difference, the less efficient and effective the heat exchanger is and less worthwhile it is to use that heat exchanger.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529

 

[georgeverghese]

Many shell and tube units operate with 10-15degC temperature approach; but 5degC sounds good for minimum permissible. Plate frame HXs' often operate at 2-3degC temp approach; while cryogenic heat exchangers operate at less than 2degC (1degC or less is often quoted). The controlling case temp approach will be the smaller of either that seen at the hot end of the HX and the cold end.

The selection of temp approach must be weighed against other factors such as no of tube side passes, no of shells used, HX-TEMA type complexity, total surface area to be provided, cost of utilities, cost of the heat exchanger, and plot space - there may be others also.

In some cases, there may be a preference to use a co current flow type HX, while in most other cases, mix cocurrent and countercurrent may be acceptable.

In some 2 phase units with strange Q-sat temp profiles, there may be an internal temp approach which may be less than that at either terminus.

 

[Yobbo]

Thanks EmmanuelTop, IRstuff and georgevergese for your replies so far.

I realise I made a mistake with the title and in my text. Indeed it is "Pinch temperature" instead of "Pitch temperature". I do understand the general concept of the pinch temperature. Choosing a minimum pinch temperature in the sizing process results in an economical use of transfer area. Choosing it too small will lead to an unevenly large heat exchanger.

The heat exchanger in question is an existing shell tube heat exchanger and was originally sized with a pinch temperature of 5 deg C. I realise I didn't formulate my question properly. The question is the following:
When an existing heat exchanger is rated for a different situation it may turn out that it becomes so efficient that the resulting pinch temperature is much lower than the said 5 deg C. For example 1 deg C. The flows have become higher, so there's no case of possible switching from turbulent to laminary flow or vice versa. I am aware of the fact that a higher flow will result in a higher pressure drop over the heat exchanger. Are there any other reasons of physical nature (thermodynamically or hydraulically) that will limit the use of this existing heat exchanger under that circumstance ( i.e. with a much lower pinch temperature)?

I would appreciate your additional comment on this.

With best regards,

Karel Postulart

Karel Postulart, The Netherlands
Nuon Power Generation

 

[EmmanuelTop]

Karel,

This is all now very theoretical. Depending on what changes you make to the fluids (change type of fluid and/or their flows), the exchanger may slightly under-perform or over-perform compared to the base case. You may get higher performance (duty) of the exchanger but with higher pinch temperature, simply if you alter inlet temperature of one of the fluids. For example, if you increase hot fluid inlet temperature, LMTD will increase and you'll get more duty out of the exchanger but depending on the flowrates and exchanger design/geometry the pinch temperature difference might actually increase. The other way round is also possible under certain circumstances, by e.g. swapping from liquid-liquid to gas-gas exchanger where temperature approach is always higher due to lower heat transfer coefficients in gas service.

For a properly designed exchanger, I don't think it is possible to obtain substantially lower pinch temperature - unless the original design was way below the optimum. Much will always depend on the exchanger type/geometry (i.e. for TEMA E-shell with 2-pases, the cold fluid outlet temperature can never be higher than hot fluid outlet temperature, due to the element of cross-flow in these exchangers, making it impossible to cross this gap). On the other hand, with Plate & Frame exchanger or brazed aluminum exchanger you can get very tight temperature approach (less than 1 degree for BAHX).

You have two options - either to consult the exchanger supplier for re-rating of the equipment in new/altered service, or to have a heat transfer expert running simulations in specialized software (e.g. HTRI) to see what the performance prediction will be under new conditions.




Dejan IVANOVIC
Process Engineer, MSChE

 

[IRstuff]

I'm not sure that you are yet asking the right question. The pinch temperature, as described, is the result of the two mass flows and inlet and outlet temperatures. If the calculated pinch temperature is small, then so long as the overall performance requirements are met, it is what it is. The question is whether the heat transferred is being efficiently transferred. Q is proportional to htc*ΔT, where the heat transfer coefficient is a function of mass flow, etc. Generally, htc is less than proportional to mass flow, i.e., doubling htc may require much more than double the mass flow, which requires more power to move and re-cool the mass flow.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529

 

[Compositepro]

The only reason for the rule of thumb that is being discussed is economics. The closer the pinch temperature is to zero the more energy is transferred from one stream to the other (i.e., more efficient). But the size and the cost of the heat exchanger grows exponentially as pinch temperature approaches zero. If you have a free heat exchanger pinch temperature really does not matter very much, unless it is too high, which means it is not large enough for your application.

There are other considerations like flow velocity being very low, resulting in fouling of the heat exchanger.

 

[marty007]

If you have an existing heat exchanger, then as Compositepro says, pinch temperatures really don't matter mechanically speaking.

The only mechanical detail that I'd bring up would be for if this is a fixed tubesheet heat exchanger and you have significantly changed the flow conditions from the original design. You may need to take a closer look at the mean metal temperatures of your shell vs. tubes, and examine if you have any new issues related to differential thermal expansion.

It's impossible to tell from the information provided if this is a real concern, but I always bring it up when people talk about changing heat exchanger operating conditions.

 

[georgeverghese]

If your existing HX is oversized for a particular service, then controlling case approach temp will decrease to suit the duty and the exchange surface provided. You will find the LMTD will drop considerably - the limiting case will be when the LMTD correction factor, otherwise called the F factor will decrease and approach 0.75 for the HX configuration you have now (in most new designs the LMTD F factor ranges 1.0 to a min allowable 0.8 - see heat transfer texts for F factor chart for the HX configuration you have).

In some cases where the process fluid temp must be limited to a certain range throughout the entire turndown range, a different design approach must be adopted. If you can disclose details of this application and why you have a concern with decreasing approach temperatures ( which can be translated to an LMTD F factor approaching 0.75 or so ), then we may be able to help.