Desuperheating necessary/senseful 2

[virk]

With a shell&tube heat exchanger we need to heat an ammonia/water-solution. On the tube side will be the heating steam. The condensing pressure of the steam will be about 2.8 bar absolute. The steam is provided with a pressure slightly higher and/but with a temperature of about 371°C. Customer is talking about desuperheating the steam, before it enters into the heat exchanger. I do not think it to be neccessary to desuperheat the steam. It costs money, causes pressure drop somehow and additionally I think that heat transfer without this desuperheating is even better. Heat exchanger will be designed for to withstand that temperature in any case.
What do you think about that?

 

[25362]

May be a visit to the following threads will be helpful:

thread391-122033
thread931-223195
thread391-326958
thread301-86408

 

[katmar]

A small amount of superheat in the steam can usually be accepted, but in this case I believe you have too much and you do need to do something about it.

The way I look at it - and I accept that it is not totally rigorous - is to look at the ratio of heats to be given up and the relative heat transfer coefficients.

Some base numbers:
Heat in 2.8 bara 371°C steam = 3215 kJ/kg
Heat in 2.8 bara saturated (131°C) steam = 2720 kJ/kg
Latent Heat of condensation at 2.8 bara = 2170 kJ/kg

To cool the superheated steam to the saturation temperature inside the reboiler requires the removal of 495 kJ/kg (3215-2720). After cooling, the steam gives up its latent heat. The sensible heat is 19% of the heat given up by the steam ( 495/(495+2170) ). The heat transfer coefficient in the zone where the sensible heat is being taken up is probably 25 to 30% of the HTC in the condensing zone. You are therefore using about 45% of your heat transfer area to transfer 19% of the heat.

If your exchanger is sized accordingly then there is no problem, but if not it can lead to under performance of the exchanger. Like I said, this is not a rigorous analysis and you could do full designs on both scenarios if you have the software available, but it does give something of a qualitative feel for the situation.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[hacksaw]

agreed most HEX are sized for some superheat but not that much. A majority of the useful heat transfer only occurs after the steam condenses.

some desuperheating occurs with when the incoming steam line is un-insulated, but that introduces operating problems that should be avoided.


be sure to include adequate pipe run after desuperheater

 

[FredRosse]

Superheated steam convection coefficients (pure gas convection) are far lower than condensing heat transfer coefficients, however there is other actions often with your posted condition.

This problem often depends on the HX tube metal temperature (or the fouling film layer temperature) at the steam interface. If the metal is below saturation temperature, then there will be condensation on the metal surface, and the turbulent superheated steam flow will often strip the condensate from the surface, and entrain the condensed water into the steam flow stream. These small entrained droplets then are in good thermal contact with the superheated steam, heat transfer then evaporates the water droplets, and desuperheats the main steam flow. Inside tubes at fairly high velocity this can desuperheat the steam flow very quickly, as long as the tube metal temperature in contact with the superheated steam is at least 5C (10F) below local saturation temperature. For these cases all the desuperheating is done without any external desuperheating apparatus, and with essentially the same heat transfer surface area as a pure condensing HX design.

The relevant parameter here is the surface to steam interface temperature, which must be significantly below the local saturation temperature, if that is so, and the steam flow has good turbulence, then hardly any extra surface area beyond the pure condensing design is necessary.

 

[abeltio]

With steam inside the tubes, I would use a vertical heat exchanger so the level of condensate inside the tubes will "limit" the useful heat transfer surface. Horizontal tubes will have an impact on the distribution of temperatures leading to bowing (hogging) of the tubes, loss of efficiency, very hot top tubes and potential damage to the tube bundle.

The steam pressure is quite low, for this service I would certainly consider steam on the shell side with a desuperheating section with impingement inlet plate as required.
Having water + ammonia on the shell side will create other problems, what happens in case of a leak? Containment is usually better within the tube side.
Also, from the heat transfer coefficient point of view, tube side water and shell side steam would provide the most compact HX (it is much easier and cheaper to increase velocity, thus heat transfer coeff in the tube side than on the shell side)

saludos.
a.

 

[virk]

We have built and still build this type of heat exchanger (Steam in u-bundle tubes and boiling out of water-ammonia-solution) since nearly 20 years without any problems. Heat transfer we know and have our experience and concerning leakages we until now never had a serious problem.
My "problem" was related to desuperheating or not. Actually I think (in opposite to the opinion here ) that the heat exchanger will transfer more heat, if the entering steam will not be desuperheated.
But most probably the desuperheater will be installed anyway; so this matter is no longer of interest.

Kind regards

virk

 

[abeltio]

Ah, ok! A reboiler boiling out the water ammonia mixture is a different application (not a regular heat exchanger) where there is a free surface of liquid. The original post mentioned a "heat exchanger", I was thinking BEU, instead of a "reboiler" like an AKU

saludos.
a.