Super heated steam to blame?

[bmemmott]

We have a HX at our plant that has both 600# and 160# steam attached. We have been running on 600# steam and I am currently looking at the economic implications of switching to 160# steam. My problem is that when looking at historical data, the required flowrate for 160# steam is 33% higher than when 600# steam is used. I am not sure what is going on. I would think that as 160# steam has a higher heat of vaporization, less steam would be required. I just wondered if anyone had some ideas on why this is the case.
Thanks

 

[CJKruger]

bmemmott, how do you know the flow rate ? My guess is that the orifice plate coefficients are based on 600# steam, and gives you the wrong reading for 160# steam.

 

[bmemmott]

We have two different flowmeters, one for the 600# and one for the 160#, and both are calibrated for there respective pressures.

 

[rmw]

Why does your thread title mention superheated steam and then your post doesn't mention if one or both of these steam sources is superheated. It could be that you are actually using more steam because you are getting better heat transfer. What about the duty on the heater, is the heated product inlet/outlet temperatures or flow rate different when on one steam or the other? According to what limited information you are giving, I would suspect that you have SH in the 600# steam and are getting better heat transfer and better temperature rise with the 160# steam but I'd have to have more information to confirm that.

rmw

 

[Joesteam]

I was always taught that superheated steam has lower heat transfer than saturated steam. It is the latent heat that gives the bang for the buck, and the thin film of condensate on the tubes give a better heat transfer than that for a gas that does not condense. Maybe he has superheat on the low pressure steam??

Joe Lambert

http://www.control-associates.com/

 

[danberry]

Joesteam is correct; superheated steam doesn't transfer well and it takes a lot of surface to just cool it to saturated temperature.

The 600# steam provides slightly more energy than the 160# steam when condensed to a fixed condensate temperature. What is the shell pressure for each case?

If you are using more 600# steam, perhaps you are providing more heat to the heated side than with the 160# steam.

Orifice plates are unreliable until proven otherwise. Measure the condensate instead.

 

[CJKruger]

danberry and Joesteam. I was also taught that superheated steam doesn't transfer heat well, but I no longer believe this is totally true.

If the delta temp is sufficient, even superheated steam quickly condenses on the cold tube. The delta T should however be based on saturated steam temperature, and not the superheated inlet temperature.

 

[Latexman]

The backpressure on your condensate system is probably the same whether you are using 160# or 600# steam. Assuming the condensate is sat'd at that backpressure, sat'd 600# condensate has 38% more heat content than sat'd 160# condensate. Thus, you need 33% more 160# steam to make up the difference to not affect your process.

Good luck,
Latexman

 

[rmw]

While what CJKruger says is true in basic heat transfer theory, you have to look at what is going on inside the heater. Heat transfer is more than just delta T. It is a lot of Reynolds number and other fancy name guys number considerations, Prantl et al.

So while the delta t difference of superheated steam gives a higher driving force or heat tranfer rate than the delta t difference of its equivalent saturated steam at condensing temperature, and yes there is some heat transfer rate at the sensible heat transfer from superheat to saturation, it is the latent heat transfer as the steam condenses that give the high heat transfer rates-shall we say the bang for your buck.

But it is what is going on around the tubes while this sensible heat transfer is taking place that is what is important. When saturated steam gives up its latent heat and the good heat transfer happens, the volume change from a pound of steam to less than a teacup of water (I am not going to get steam tables out here to be more specific) that draws more steam into the heater creating steam flow and velocity around the tubes resulting in higher Reynolds numbers.

So sure, superheated steam has a higher driving force for a given amount of sensible heat transfer, but the time required that the steam has to sit there in contact with the tube surface to transfer that heat is enormous compared to steam entering at saturation and collapsing to condensate immediately upon touching a cool tube, bringing some more steam right in to do the same microseconds later.

Superheated steam blankets the tubes and slows down the flow velocities in the tube bank to a point that the reynolds numbers become so low that to repeat a line that I do remember from my heat transfer text, 'for Reynolds numbers below 2000, it is debatable whether or not heat transfer even occurs at all' or words to that effect.

rmw

 

[25362]

To bmemmott, historical data may refer to different heat load conditions. You should check all the factors, including the process side, not just the recorded steam consumption.

 

[quark]

Latexman,

The heat capacity of the condensate can be 38% more but that need not get tanslated into 33% more steam flow.

The extra heat content (considering a back pressure of say 160psi or below) is about 135btu/lb. The latent heat of steam at 160#a is about 850 btu/lb. So, the steam flow should only increase by 135/850 = 16%

Further, the difference in enthalpies of steam at 600# and 160# is 8.33btu/lb, the steam flowrate should not vary much if the outlet condition is same in both cases.

Apart from 25362's suggestion, I would also check for the flow compensation with both temperature and pressure if 600# steam is in super heated condition.

 

[Latexman]

Right. If the OP would fill in the info. gap on superheat, I'm sure more meaningful advice would be forthcoming. What is T and P of both steams?

Good luck,
Latexman

 

[CJKruger]

rmw, I think maybe you misunderstood my post (or maybe I wasn't clear).

My point was that the temp driving force for superheated steam is NOT more than for saturated steam.

There is not really a "desuperheating zone" for superheated steam. The superheated steam in contact with the cold tube immediately condenses on it. Superheated steam then condenses into the saturated condensate around the tube. Thus, the tube always sees saturated condensate around it, even for superheated steam, and the temp difference should reflect it.

 

[rmw]

No, it wasn't unclear, and I solidly but respectfully disagree with what you stated. Super heated steam can't condense until its temperature is reduced to its saturation point. That is basic thermodynamics or the laws of physics. The heat transfer that occurs during that process is only sensible heat transfer. Once saturation is reached, then the steam will condense and only then. That is when latent heat transfer occurs. Superheated steam by definition can't condense until it reaches saturation and it can't reach saturation until it transfers enough heat away to get there.

Heat exchangers that are designed to handle superheated steam such as power plant feed-water heaters have desuperheating sections where there is no moisture present and, actually where moisture isn't wanted because it is very detrimental. In order to get any meaningful sensible heat transfer at all, (or said differently to get the type of Reynolds numbers that produce good sensible heat transfer) the velocities in a DSH zone are very high and if condensation does begin within the DSH, it will cut the tubes to ribbons. So normally DSH zones are designed to have a 5-10 deg F outlet temperature above saturation to insure a dry exit as the steam exits the DSH and enters the condensing zone of the heater. The heat transfer rates even at those Reynolds numbers (velocities) are still very low overall, in the 50-150 B/hr/ft^2/F compared to the condensing zone which starts at 500 and can be higher.

Many believe that superheated steam condenses, but sorry, I've tested heaters both ways; with and without SH and I have seen the difference.

rmw

 

[Latexman]

Didn't we go round and round on this about a year or so ago?

Why does a cold glass of ice tea sweat? Same thing. Local temperatures condense vapors too.

Good luck,
Latexman

 

[rmw]

Huh?

rmw

 

[Latexman]

RMW,

I believe CJKruger is speaking about condensers that are fed superheated steam. If the tube is less than the saturation temperature, condensate will form on the tube and superheated steam will transfer heat into the condensate and condense into it. The key is the local tube temperature. Condensers that have tube temperatures that are everywhere lower than saturation temperature do not have desuperheater zones. I suspect these are handling steam with much less superheat than what you are talking about.

I believe you are speaking of desuperheaters whose tubes everywhere are above the saturation temperature. They are not meant to condense steam, or they will quickly be destroyed by erosion corrosion.

It's apples and oranges, right?

Good luck,
Latexman

 

[rmw]

Latexman,

Heat exchangers that transfer heat by condensing steam work (that is to say make their duty) by condensing steam, which collapses the steam vapor volume which draws more steam to be condensed into the condenser to be condensed which condenses and collapses the steam.......etc, etc and now we have a performing heat exchanger.

While a glass of tea will condense moisture out of the atmosphere and a cold tube will in some finite amount of time condense steam from superheated steam around it. However, the rate at which that happens is drastically different from the action of condensating heat transfer.

Remember for a point of discussion, if superheated steam encounters saturated condensate on the surface of a tube, in order to transfer heat to that condensate, and in so doing desuperheat itself, it has to boil away some of the condensate, which then has to be recondensed by the tube bundle adding load to the heater.

The heaters I mentioned are condensers that operate at high pressure and that have special zones to desuperheat the steam so that it can do what is described in the first paragraph above. It is not that they won't condense that steam without desuperheating it first rather that the rate at which they do the condensing would be so slow that the heater would have to be tens of dozens of sizes larger.

So, a heat exchanger of any type that is designed to transfer heat by condensing steam and reaping the benefit of latent heat transfer is greatly hampered by the presence of superheat-any superheat, obviously the more there is the worse it is. The slow heat transfer rates of the sensible heat transfer (as compared to the heat transfer rates of the condensation) has the effect of blanketing the tube bundles, slowing down the velocities through the tube bundle and overall choking the heater.

But, yes, don't misunderstand that I am saying that a cold tube wont eventually condense the steam around it. I just can't afford to build a HX big enough to operate that way.

It is apples and oranges and that is why I asked the OP about his steam conditions. If he is trying to get his HX to do a certain duty with saturated steam at one of the two pressures given and superheated steam at the other, no matter which one it is, then he is trying to compare apples and oranges. That is what I am trying to draw out of him/her.

rmw

 

[Joesteam]

Everyone is right, going back 30 years to my engineering class: Q=UA(dT).

With superheated steam the dT is greater, however the U is MUCH smaller, hence the lower heat exchange rate with superheat vs. saturated steam.

A gross estimate of the 'U' for superheat I have seen is 16 while that for saturated is 205.

Then there is the BTU's available from the phase change.

Joe Lambert

http://www.control-associates.com/

 

[CJKruger]

rmw, I used to think the same as you, but in most cases now think that approach is incorrect. I agree with Latexman's explanation.

The best description I found on the internet related to this issue, is Section 3.3.4 from:

http://www.wlv.com/products/databook/ch3_3.pdf

They conclude with: "It is both simpler and more conservative to assume that condensation will occur directly from the superheated vapor, using the saturated temperature and a condensing heat transfer coefficient in the rate equation..."