Heat transfer regime of Superheated steam with surface metal temperature under steam dew _ Latexman

[Rspain]

Dear Mr. Latexman,
Dear Sirs,

It is clear you have a good understanding of the underlying principles of heat transfer with steam.

Reding your posts (thread391-155676), I have a doubt and wonder if you could please help me.

It seems now clear to me that superheated steam (transfering heat through a metal wall to water) will behave differently in two different scenarios:

1) gas-metal interface (metal surface temperature on steam side) above saturation temperature of the superheated steam will lead to single-phase superheated convective heat transfer as if hot air, thus with moderate delta temperatures (100ºC), low heat transfer expected in comparisson with saturated steam in similar conditions.

2) gas-metal interface (metal surface temperature on superheated steam side) under saturation temperature of superheated steam, steam will collapse into saturated condensate in contact with the metal surface, thus causing depression and speeding up heat transfer rate.

From different readings (drying food with superheated steam, heating undercooled water tank with superheated steam, etc.), it seems clear to me that behaviour of superheated steam will be similar (as for condensation rate) to that of saturated until Tsat is reached. According to your understanding/experience, regardless degree of superheating (kinetics involved in superheated steam condensing are extremly fast)?

Reading the Perry's 11-11 you refer to, I'm a bit puzzled.

It states: To determine whether or not condensation will occur directly from the superheated vapor, calculate the surface temperature by assuming single-phase heat transfer. By equating the heat flux calculated with global heat transfer coefficient with heat transfer flux (in steady state) on superheated steam side assuming single-phase heat transfer, it is possible to deduct metal surface temperature on superheated steam side, and check if the assumption (above dew point) is correct or not.

The thing that puzzles me is:

1) "If Tsurface > Tsaturation, no condensation occurs at that point and the heat flux is actually higher than if Tsurface ≤ Tsaturation and condensation did occur"

This I don't understand. I would have expected a lower heat transfer rate, in analogy to hot air vs saturated steam. Where is my mistake? is it a misprint, or am I doing something wrong? is it perhaps considering a scenario of very big delta temperature? Would, in such scenario (very big delta T), condensation and consequent water film formation slow heat transfer in comparisson with single-phase high delta T superheated convective heat transfer?

2) I understand, from a transitory point of view, that the condition for superheated steam condensing when coming into contact with the metal surface can be calculated as above indicated. However, once steady state is achieved, wouldn't the film/condensate layer exhibit a temperature gradient, such that the relevant temperature (for determining if superheated steam condenses or not) would no longer be the metal surface temperature but instead the outer surface temperature of the condensate film in contact with metal matter?

Thank you (and any other person who can add light) for your kind help,

With best regards,
Roberto

 

[Rspain]

PS: I would just like to feel more comfortable in estimating if superheated steam will immediatelly condense or not in different scenarios (flow through pipe with heat loss to surrounding air, slightly superheated steam being fed to a paper preheater, etc.)

 

[Latexman]

I don't understand the second half of that sentence either. Unless others can explain, perhaps you can e-mail the author, Kenneth J. Bell and let us know what he says.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[Rspain]

Thank you Latexman, for sure I will share my findings.

If any other person could share his knowledge on the subject, would very much appreciate.

Thanks and kind regards,
Roberto

 

[Rspain]

NOTE: for the purpose of further information, from HANDBOOK OF HEATRANSFER (M.Rohsenow;P.Hartnett;I.Cho - Third Edition), attached

 

[georgeverghese]

Agree, the statement (1), which is an abstract of the notes below eqn 11-26 in Perry 7th edn, does sound odd.

But it is consistent with the text which follows on (in the same para), which seems to imply that the author believes that condensing htc's are lower than desuperheating htc's. And we know this is not true.

 

[Rspain]

Dear Latexman,
Dear Georgeverghese,
Dear All,

First of all, thanks for Latexman for his suggestion, I am in contact with Dr. Bell and I am about to further study the info he has just sent me in an attempt to fully understand. I will share my findings asap. By the way, he is extremely kind person.

The questions seem not trivial at all, but rather extremely complex.

I attach an interesting paper dated 1925.

Argument A) Worth noting in page 59, comments (1), (4), (5), (8) & (9) - suggesting that indeed Superheated steam renders higher heat flux than saturated steam at the same pressure.

Argument B) In opposition to above, statements (3), (6) & (7).

In favour of Argument A, again in page 60 (left column), when talking about the advantages of superheating steam after the boiler for transit through pipes until arrival to process: "Even if the heat actually lost in transit is greater than in the case of saturated steam...". Thus clearly suggesting that supeheated steam will experience bigger heat losses through pipes than saturated steam at the same pressure.

Coming back to the core question; If we assume:

1) Superheated steam will condense if coming into contact with a metal surface temperature colder than its Tsat***
2) From heat transfer flux point of view, such an event (above) will happen in a fashion essentially similar to saturated steam at Tsat

It follows that there are only two possible scenarios in steady state when using a liquid coolant (at a temperature below Tsat of superheated steam all along the heat exchanger) to cool superheated steam:

i) Metal surface temperature on steam side is below Tsat
ii) Metal surface tempreature on steam side is above Tsat

Regardless any other consideration, it seems trivial that, being the heat flux the same (in steady state) in all media (coolant, metal wall, condensate film if existing and steam side), the overall heat flux can be evaluated (it is the same) by looking at the heat flux through the metal wall, which "ceteris paribus" on coolant side, will necessarily be bigger in scenario (ii).

Now, it seems clear that scenario (ii) can only be achieved with superheated steam (since if using saturated steam the heat source is already at Tsat, thus scenario [ii]) results simply impossible from a thermodynamical point of view).

Hence, if we assume a certain heat flux (i.e. 100W/squared meter) using saturated 1bar steam, the only possible way to increase heat transfer flux with the same pressure and flow regime in the steam side (assuming every media other than steam remains the same) is by using superheated steam.

It follows quite trivially from above assumptions (i) & (ii); that heat flux with superheated steam should always render equal or bigger heat flux than saturated steam at the same pressure.

Hence, it all pivots around assumption (i).

There seems to be plenty of evidence that assumption (i) is sensible.

By googling Superheated steam heat flux it is easy to find several studies showing that:
a) when direct heating sub-cooled water with superheated steam, the behaviour of superheated steam is essentially the same to what would be expected from saturated steam at the same pressure until the sub-cooled water reaches Tsat of the superheated steam
b) when drying a wet material with superheated steam, the heat rate is identical to that expected from saturated steam at the same pressure until the water film surrounding the material reaches Tsat of the superheated steam

There is much additional evidence supporting assumption (i).

The only piece of information I have been able to find suggesting that assumption (i) might sometimes not be true, is the statement from HANDBOOK OF HEAT TRANSFER - Warren M. Rohsenow: “Miropolskiy et al. [30] found that superheated steam flowing in a tube did not always condense when the wall temperature was less than the saturation temperature. Condensation did not occur unless both the vapor temperature and quality were below certain threshold values”.
Thus, it seems clear that the necessary conditions for having supeherheated steam rendering a heat flux lower than the one that would be provided by saturated steam at the same pressure, is to have a single-phase convective heat transfer regime on the steam side with metal surface temperature under Tsat BUT WITH NO condensation.
If I understand correctly, this would be a meta-stable kinetically impeded state of superheated steam.
So far I have been able to read and understand, predicting when such impeded regime will occur is far away from being trivial, thus to predict is superheated steam will:
1) Render lower heat flux
2) Render same heat flux
3) Render bigger heat flux
Is simply not an easy matter.

I might have made some mistake above, I have not had enough time to study all the info that I have recently received and researched. If so, I apologize.
I will read with more calm the notes from Dr. Bell and will revert when I feel more confident of having understood correctly the notes and its implications.

With kindest regards,
Roberto

 

[Rspain]

PS: interesting attached investigation

Looking at page 44, figure 16 "Experimental average wall temperatures", if I understand correctly, in the dry section of the superheated steam side of the heat exchanger, heat transfer coefficient is clearly lower than in the CSH, where condensation occurs. However, it appears as if the wall temperature is higher, thus clearly suggesting bigger heat flux (bigger delta T overweights lower HTC).

Does that make sense for you...?

 

[Rspain]

As for: "But it is consistent with the text which follows on (in the same para), which seems to imply that the author believes that condensing htc's are lower than desuperheating htc's. And we know this is not true."

I believe that what is being stated by Dr. Bell is rather that the heat flux of superheated steam with condensation is lower than when single-phase convective heating against metal surface temperature above Tsat (in steady state), in spite of superheated steam's HTC being much lower. This is consistent with the fact that a higher metal surface temperature must necessarily imply bigger heat flux. As for HTC being smaller in dry superheated steam heat transfer, there is no discussion in this point and every single source I have found agrees on this.

It is only in the scenario of dry superheated steam single-phase convective heat transfer with metal surface below Tsat (meta-stable) with no condensation that it is possible, in steady state, to have a lower overall heat flux with superheated steam (and therefore a lower metal surface temperature than would expect with saturated).

BR,
Roberto

 

[georgeverghese]

Okay, will wait to hear from you what Dr. Bell has to say about his notes following on from eqn 11-26 in Perry.

 

[georgeverghese]

After I fleshed out eqn 11-26 with some notional numbers, I'm now convinced of the veracity of the statements made in the text following this equation. Here is one of the 2 cases I ran, with each case consisting of a calc for the desuperheating zone and another calc for the condensing zone.

Rearranging equation 11-26 and adding another obvious but critical equality. I've assumed the process gas is on the shellside and coolant on the tubeside. Have used conventional US units for all numbers.

ho (Tvap - Ts) = U (Tvap-Tcool) = hi ( Ts-Tcool) = Qzone

Case 1
Desuperheating Zone
Assume ho = 100 btu/hr/ft2/degF, Tvap = 300degF , hi=50btu/hr/ft2/degF, Tcool=80degF

Here, we derive U = 33.3btu/hr/ft2/degF, and from this

we have U(300-80) = 100(300-Ts), from which we get Ts = 226.7degF.

Also desuperheating zone heat flux = 33.3(300-80)= 7326btu/hr/ft2 and MTD = 300-80=220degF

Condensing Zone

Assume Tsat = 200degF. Since this is condensing zone, assume Ts = 180degF

For a given HX, the value for hi and Tcool is essentially unchanged in this zone

So Qzone = 50(180-80) = 5000 btu/hr/ft2 < 7326btu/hr/ft2

Also we assume that ho is now say 300btu/hr/ft2/degF for condensing conditions

Hence ho ( Tvap - Ts ) = Qzone or 300(Tvap - 180) = 5000, from which we get Tvap = 196.7degF

and MTD = 196.7-80= 116.7degF

Using another set of numbers for case 2, I get the same observations as stated in the text following eqn 11-26, even if I may base these on these 2 cases only

a) That heat flux is indeed lower in the condensing zone compared to that in the desuperheating zone.
b) It is therefore conservative to dump the entire desuperheating zone heat load into the condensing zone heat load and use the U and MTD for the condensing zone to derive the required surface area for the HX.

Obviously, working out the desuperheating zone surface area separately with the local value for U and MTD will result in some savings in total HX area.

The notes indicate these observations are universally applicable to all pure component desuperheating - condensing HX applications, regardless of whether the condensing phase is on shellside or tubeside.

 

[Rspain]

Hi All,

I think it is then clear and we all agree that, unless violation of principle (superheated steam will condense if metal surface temperature is below Tsat - ref. Miropolskiy et al. [30]), the following follows:

1) Superheated steam at a given pressure will always result in a heat flux equal or bigger than the heat flux that would result from saturated steam at the same pressure
2) Convective single-phase heat transfer of Superheated steam (dry wall desuperheating) always renders bigger heat flux than wet wall de-superheating

BR,
Roberto

 

[georgeverghese]

Agreed, these conclusions are valid when the coolant side htc remains more or less constant and there is only a small difference in the coolant side mean temperature (ie between desuperheating and condensing zones coolant temps).