Superheated steam in heat exchangers 1

[Keef]

Can anyone point me in the direction of some information regarding the use of superheated steam in heat exchangers. I know that it reduces the heat transfer rate as the bulk of the heat transfer is obtained by the change of phase, which only occurs after the steam has lost its superheat.

I do not have a particular problem, more I am looking for general info about applications and how to calculate heat transfer areas.

 

[speco]

Keef,

Well, it depends. If the cold surface is below the saturation temperature of the steam, chances are that a nominal amount of superheat will not make any difference. The exchanger will operate basically at the saturation temperature.

However, if there is a LOT (say a few hundred degrees) of superheat, then it's a different ballgame. The steam side heat transfer coefficient would then be based on a dry gas in the desuperheating portion of the exchanger, until it reaches a point where the wall temperature is at saturation. The dry gas coefficient is MUCH lower than the condensing coefficient. Chances are that the amount of surface required would then be greater than if just condensing the steam, even though the temperature difference in that portion is greater.

The amount of steam required would be slightly less than if just condensing. It would be a function of the total enthalpy difference between the superheated enthalpy and the saturated liquid enthalpy (assuming that you would be controlling the steam with a simple trap at the outlet).

Regards,

speco

http://www.stoneprocess.com

 

[phex]

as the heat capacity of steam aside from the heat of condensation is very low, desuperheating will take just a very small part of the exchanger surface. thereafter you will have normal condensing heat transfer at saturation temperature. ballpark estimate would be 10% of exchanger surface are lost to sensible heat transfer (which sports lower heat transfer coefficient by a factor of 100). hx software like HTRI can calculate the lost surface due to superheat. my advise would be to use some such software to verify those assumptions.

as a rule of thumb, as heat transfer is much higher with condensing steam, it is not advisable to use highly superheated steam in heat exchangers. you'll lose more than you gain due tothe higher temperature.

hth,
chris

 

[25362]

The presence of non-condensables in condensing steam is sometimes of more importance in that it significantly reduces the condensing film-side heat transfer coefficient. From Ludwig's Applied Process Design for Chemical and Petrochemical Plants, Volume 3, Gulf Publishing Company, 1% mol non-condensable in steam reduces the pure steam estimated coefficient to 60-70%, 5% gas to 20-30%, and 8% of non-condensables, to 12-25%.

 

[rmw]

Regarding what Chris and Speco said above, while it is all true, it is not necessarily always true all at the same time.

Yes, superheated steam has sensible heat that has to be removed prior to condensation, assuming condensing is the goal, where, as you state, the latent heat of the steam is recovered at rapid heat transfer rates. The question of heat transfer rate is the problem.

Heat transfer takes surface area, and the higher the heat transfer rate, the lower the surface area required. That's the basics.

Some heat exchangers, such as power plant feed water heaters (FWH) are compelled to have to handle superheated steam before condensation can occur, due to the conditions at the turbine extraction point. These higher pressure FWH's have a dedicated zone in them called the desuperheating zone (DSH) that are secifically and carefully designed to do this with the least amount of surface area possible, as the condensing zone is really where it is at, and surface area costs money.

In order to maximize heat transfer, in this area of the heater the velocities are very very high. Reynolds number is a dimensionless heat transfer parameter that is a function of velocity, among others. As a basic saying, for sensible heat transfer, (convection) the higher the reynolds number, the better and faster the heat transfer.

Hence FWH's are designed to have very high reynolds numbers, so high in fact that vibration caused by hydrodynamic swirl, or whip of the tubing excited by the velocity becomes the limiting factor for DSH design. The highest heat transfer coefficient that I am accoustomed to seeing in DSH's is in the 160 btu/hr-ft^2/F range at best, which is not very high, compared to condensing rates in excess of 500.

Now, consider a Hx that is not specifically designed for convection heat transfer, meaning that the velocities are not very high, favoring low pressure drop in order to get the saturated steam in and exposed to all the heat transfer surface, and the htc steam side plummets rapidly. I recently figured a Roberts type sugar evaporator, both by hand calculations, and using

http://www.processassociates.com,

and found that the OHTC (low reynolds numbers tube side also due to subcooled juice) at conditions was 52, compared to a decent average of 450 when the steam finally got cool enough to condense in the back side of the calandria.

Bottom line was that a lot of the available surface of the evaporator was devoted to desuperheating, instead of sugar juice boiling.

Superheat in this case robbed the evaporator of a lot of available capacity, and put a large percentage of its surface area to doing low grade heat transfer.

25362, except for the fact that there is at least a little delta F in the sensible heat transfer, superheat has a "blanketing" effect, acting much like NC gases, especially in surface condensers, where SH steam was never anticipated to be.

Bottom line; if you can get the steam to saturation before it enters the Hx, the better the heat transfer.

 

[Montemayor]

I don't know how engineers today study heat transfer, but 40 years ago we learned the facts: you can't expect to get any use out of superheated steam in heat transfer. And it still holds today. Don Kern says it in "Process Heat Transfer" and every other recognized author on the subject. I don't understand why engineers today can't see the facts.

The only reason(s) superheated steam is made is for:
1. To operate superheated steam turbines and other drives;
2. To make sure that the steam doesn't condense in the steam mains that transport it to its use as a saturated heating medium.

There may be other uses, but those are the main ones in a process plant.



Art Montemayor
Spring, TX

 

[rmw]

Art,

The reason I was knowledgable enough to launch into the tirade I did in the post before yours is that a client of mine was told by a Phd/ME that superheated steam didn't matter (in the evaporator mentioned). He is the industry expert, and I am just a BS/ME, who barely graduated, and a newcomer to that particular industrial segment. I had to dig out the old college text books, and other reference material to reteach myself what I learned 30 years ago. Thank goodness for the Internet, because it made the review much easier, and Eng-Tips, and the processassociates site that I found here helped a lot, but not until after I had retaught myself basic hand calculated heat transfer.

At least I did not have to use a slide rule this time.

If the PHD's are giving such advice, how are us BS's going to be able to go around cleaning up after them, and go about containing the damage.

You stated the facts in a forceful way. Thank you.

rmw

 

[25362]

Some additional comments about steam superheat in heat exchangers:

The condition of steam superheat by a few degrees exists in many HE. For example, by control valve throttling prior to entering the heating unit, and, in particular, in those HE condensing the overhead vapours of vacuum columns in crude units where steam may enter at temperatures about 50[sup]o[/sup]C, or more, above its dew point.

Vacuum unit vapours also contain, besides steam, hydrocarbons and non-condensable gases. Down to its dew point, steam would be considered a non-condensable and treated as such. In these cases the dirt (fouling) coefficient may become an important factor in the design of the condensers.

 

[Montemayor]

rmw:

What you describe has also been my experience. Unfortunately, perhaps because of age or senility, I assume that all my engineering colleagues have undergone similar experiences. Often, I find they haven't; and it it isn't their fault - not having the opportunity to confront hands-on problems is a handicap, not a crime. But it is essentially in the hands-on field, however that most of these problems that we describe come up. Superheated steam gives one the appearance of having tremendous heating capabilities when one merely looks at it's temperature level. However, as you've pointed out, not all is roses. And you haven't mentioned the problem that if you can't handle the resulting condensate, you'd better be informed how terribly inconsistent it is to expect that superheated steam will release only superheat without itself condensing. This is an even more difficult problem to resolve in the field. The design of heat exchangers is so conservative by nature that factors such as inerts (as 25362 mentions), fouling, and other items will quickly make any attempt to use superheated steam a bad dream.

I wish it were a simple and straight-forward application but I'm afraid, as we say here in Texas, "that dog won't hunt".
Regards

Art Montemayor
Spring, TX

 

[abeltio]

please see thread 391-84863

there is a lot of discussion on the issue...

there are basically 2 different applications in heat exchangers for superheated steam... industrial and power generation...
the industrial side is what most of the postings have been referring to... i.e. the sensible heat in SH Steam is basically useless...
in power generation, when the pre-heaters get really big... we are talking behemoth's that weight more than 10 ton each... designers get really picky and as each kJoule counts against heat rate (efficiency) and output those pre-heaters have specially designed "De-superheating" sections where the superheated steam leaves it's sensible heat and the pre-heated feedwater get's the very last drop of heat before entering the boiler...

see the following site for a pic of the above...

http://www.struthersind.com/hp.htm

there are 300+ sites when searching google for:
desuperheating section

HTH


saludos.
a.

 

[kyong]

I like rmw's first posting.
When superheated steam(at leat a few hundred degree higher than saturation temperature) is given to you any way,
you can utilize it by specially designed desuperheating zone. Desuperheating zone of feed water heater has many baffles to increase velocity of the steam so that heat transfer coeffecient becomes feasible. On the other hand, if a heat exchanger is designed to receive saturated steam and substantially superheated steam is pushed into it, surface area available to condensation will be reduced drastically, which nobody would like to have.

kyong

 

[rmw]

Albetio, that was a good reference for a picture, and description of what I had earlier described. I lament their recent bankruptcy. I have good friends that work, (or worked) there.