13 replies to this topic

[go-fish]

Hi,

I am designing a heating system which uses LP steam to heat a process fluid in a shell & tube heat exchanger. On exchanger hot side outlet, I am using a condensate drum and a condensate return pump with level control valve on the discharge. For conservative approach, I am assuming that there is no sub-cooling and the condensate at pump inlet is saturated liquid, and hence, has same temperature and pressure (3.8 bara, 142 deg. C)as at exchanger outlet. Since, the condensate liquid-vapor phases are assumed to be in equilibrium inside the drum, I am taking vapor pressure at pump inlet same as source pressure i.e. 3.8 bara. All these assumptions are resulting in setting the drum elevation high (about 4 m)in order to get adequate NPSHa.

I searched for pictures of packaged condenser receiver modules on internet. In the packages, the elevation difference between the drum and the pumps is generally less than a meter.

Am I wrong somewhere in my assumptions or am I comparing apple and oranges?

Thanks

[Padmakar Katre]

Dear,
As you mentioned that you are designing a system it means its a new system. When you quote to increase the drum elevation based on an inadequate NPSHa I understand this as per the vendors reply i.e. minimum NPSHr is more than what you have with the present elevation. Now some of the points I could speak are as below,
As fas my experience is concernde I haven't seen this kind of arrangement to have a dedicated pump for the condensate pot/vessel. The condensate pots/vessels are mounted with relative elevations with reboilers/heat exchangers. The condensate is from different users is collected in a flash vessel which is relatively at a low pressure and at a higher elevation from where the condensate is pumped to the De-areator in Boiler area. If you have the only condensate pot/vessel with need to have the bottom pump to transfer the coondensate then the only option remains is to elevate the Heat exchanger relative to the required elevation of condensate pot to meet the NPSH requirement. Hope this gives you a little idea.

[ankur2061]

go-fish,

This is one of the most obvious problems faced while pumping condensate in condensate recovery systems due to layout constraints & the high vapor pressures encountered. As suggested by Padmakar if you recover flash steam then your problem gets mitigated to a great extent since it will significantly reduce the pressure. But I believe in your case the condensate pressure is low in the first place so no economics in recovering flash steam.

The most commonly observed solution for this kind of a problem I have found is using a 'pressure-powered pump' (PPP) instead of a centrifugal pump. In such a pump a higher pressure motive fluid (steam in this case) would pump out the fluid collected in the pump pot. The pump does not require any power source & the higher pressure steam pushes the condensate to its final collection vessel. For more details for PPP visit the web site of "Spirax-Sarco" & "Armstrong". I have seen this system work without any hitch in many plants where I have handled condensate (steam / vapor heat transfer fluid).

Hope this helps.

Regards,
Ankur.

[go-fish]

I was thinking of using PPP initially. However, a very senior engineer in our company suggested to use conventional condensate pot and centrifugal pump arrangement since the amount of condensate to be handled is about 10 tph. PPP should be used for small condensate rates.

I did some more researching and found some literature by spirax sarco on condensate pot and electric pump arrangement. It mentions that the condensate should not be more than 100 deg. C. In my case, the LP steam supply is 3.5 barg and I am assuming the steam control valve reduces it to 2.8 barg (10psi pressure drop across CV) which has a saturation temp. of 142 deg. C. I am not sure if I am making wrong assumptions that the condensate pot is also at 2.8 barg, 142 deg. C or there is a reduction in pressure and temperature otherwise how will the condensate be limited to 100 deg. C. The condensate pot has a vapor balance line joining back on the vent on the shell side of HE. Just wondering if this has something to do with bringing down the thermal properties of condensate. Also, are steam HEs designed to use only latent heat for heating or sensible heat in the saturated condensate is also used because in that case the condensate could be subcooled and not at 2.8 barg, 142 deg C.

[ankur2061]

go-fish,

There is nothing like a perfectly isothermal process in the chemical process industry. There will be certain amount of heat loss from the condensate system irrespective of the insulation because nobody is trying to achieve a perfectly insulated system in such an application. The heat loss would contribute to a very small degree of sub-cooling (2-3°C) from the saturation temperature. All in all you will still not be able to achieve a sub-cooling from 142°C to 100°C unless you provide some cooling medium.

The design intent of condensing steam heat exchangers is always to utilize the latent heat & sensible heat is not considered when performing thermal rating calculations of condensing steam heat exchangers. The rating software I have used (b-jac) always used to provide the saturation temperature of the condensate corresponding to the saturation pressure indicating that the heat transfer mechanism considers only the latent heat.

Any application of utilizing both the latent heat & sensible heat in condensing steam heat exchanger applications would constitute a special case & I personally haven't come across such a case.

Hope this helps.

Regards,
Ankur.

[go-fish]

Does that mean I should reduce the pressure of the supply steam further (near to 0 barg)? In my application cold side fluid needs to be heated near to 50 deg. C. This way, I can take advantage of more latent heat in steam at lower pressures, however, the required pump differential for the condensate pump will increase.

[joesteam]

As I understand it, you are balancing the condensate pump receiver with the steam pressure and using the liquid level to control the pump. Therefor your condensate temperature will be the saturation temperature of the steam/condensate at whatever pressure the system happens to run at (at low flow requirements, you will get much more than 10% drop through the control valve).

They do make some multi-stage turbine pumps used for this high temperature condensate that will reduce the effects of cavitation. They are costly.

Another method may be to use steam traps into an atmospheric receiver. The flashing WILL reduce the temperature of the condensate to 100^oC, although you will loose the BTU value of the flash to the atmosphere.

[Art Montemayor]

Fish:

Check out the attached sketch in Excel and verify that you have the same type of installation.

From what you originally stated, you should have no problem pumping out the 3.8 to 3.5 bara steam condensate. I don't know what section of the Sarco Steam manual you refer to (you should always be specific in your references), but you are limited in condensate temperature only by the type of pump/driver you are using. With a conventional centrifugal pump and electric motor, you should easily be able to pump the hot saturated condensate back to the steam generation facilities.

Attached Files

•  Steam Heater.xls   44KB   159 downloads

[go-fish]

Fish:

Check out the attached sketch in Excel and verify that you have the same type of installation.

From what you originally stated, you should have no problem pumping out the 3.8 to 3.5 bara steam condensate. I don't know what section of the Sarco Steam manual you refer to (you should always be specific in your references), but you are limited in condensate temperature only by the type of pump/driver you are using. With a conventional centrifugal pump and electric motor, you should easily be able to pump the hot saturated condensate back to the steam generation facilities.

Please see the attached file. My systems looks like that.

Attached Files

•  Steam Heater(1).xls   46.5KB   105 downloads

[kkala]

I am designing a heating system which uses LP steam to heat a process fluid in a shell & tube heat exchanger. On exchanger hot side outlet, I am using a condensate drum and a condensate return pump with level control valve on the discharge. For conservative approach, I am assuming that there is no sub-cooling and the condensate at pump inlet is saturated liquid, and hence, has same temperature and pressure (3.8 bara, 142 deg. C)as at exchanger outlet. Since, the condensate liquid-vapor phases are assumed to be in equilibrium inside the drum, I am taking vapor pressure at pump inlet same as source pressure i.e. 3.8 bara. All these assumptions are resulting in setting the drum elevation high (about 4 m)in order to get adequate NPSHa.
I searched for pictures of packaged condenser receiver modules on internet. In the packages, the elevation difference between the drum and the pumps is generally less than a meter.
Am I wrong somewhere in my assumptions or am I comparing apple and oranges?

Below are relevant notes (of course subject to criticism).
1. You have rightly taken 2.8 Barg as normal operating pressure in the exchanger. This is also normal operating pressure in the drum (Δp through equalizing line is negligible).
2. Equalizing line counterweights feeding steam pressure variations, so that condensate flow through the pump depends only on static level differences.
3. An horizontal usual centrigugal pump of 10.8 m3/h (condensate) could have NPSHr ~ 5 ft, or 1.5 m (from a chart, not so precise), so NPSHa ~ 2.1 m (Art Montemayor seems to agree to it).
Estimates may not be quite precise, but indicate you can search and find a pump of NPSHa pretty lower than about 4 m. You may have applied additional margins for NPSHa, e.g. another 0.6 m (necessary?), but even so NPSHa ~ 2.7 m.
4. As joesteam pointed out, there are special pumps of lower NPSHr (also: vertical pumps, double suction pumps). Of course they are expensive and probably hard to find at such low flowrate.
5. Elevation difference of less than 1 m between drum and pump may mean that drum has an open vent to atmosphere. In this case condensate temperature is 100 oC and the atmospheric pressure (~10.8 m of condensate) is added to the static pressure, thus highly increasing available NPSH.
A "lifting condensate" scheme of Sarco uses "Ogden" pump after collecting condensates into a vented receiver; condensates flow there by gravity, after handling any flash steam (Sarco, Practical steam trapping).
6. In case that condensate ends up in an atmospheric tank for treatment (e.g. deoiling, iron removal) before reintroducing it into the deaerator, atmospheric flashing upstream the pump can be a choice. Yet using it downstream the pump for useful heating is a more efficient alternative.

Edited by kkala, 01 March 2010 - 01:37 PM.

[Art Montemayor]

Fish:

This thread is starting to take on the features of a lot of other prior threads that simply go on, and on, and on, without a clear and concise description or definition of the issues and scope of work. Without a clear and accurate schematic, we on the Forum have not had a clear and definitive idea of what you are doing or what you are proposing.

Take a serious and dedicated look and study of my attached Rev2 schematic drawing and you will become aware of all the important issues you have totally left out of the original query. By not taking YOUR time and effort to accurately describe in detail what you have, you have provoked what could be an endless thread going around and around without a clear definition of your destination because you haven't defined your starting point until now.

I have taken this time and made this effort to make a strong point of what seems to bother or confuse a lot of young chemical engineers in industry. You are not the first - and unfortunately, probably not the last - to misunderstand what is happening in a heating process where there is a change of phase. A simple steam system is such a process - albeit, it doesn't turn out to be so simple. Your steam system (as you have now described it with the aid of my original schematic) is, in reality, a little complex. It is complex because you are controling a related variable and not a direct variable - while going through a phase change. Please make careful study of the fact that in order to control the temperature of the outlet process liquid you must control the resulting steam pressure in the heater shell in order to set the heat transfer rate that corresponds to the desired set point. This resulting steam pressure is not obvious and must be calculated through a heat balance around the heat exchanger. This is the same effect that normally takes place in a reboiler - especially in a thermosyphon reboiler.

All of the above means that you really don't have a condensate pumping problem. The pump application is a rather simple one - whether done by a steam trap, a steam-driven pump, or an electric motor-driven pump. The problem is resolving the vapor pressure existing in the condensate trap or drum - whichever is the case. Without knowing what is the lowest or average steam vapor presure, you are unable to determine if you can apply a steam trap or not. Clearly you do not have 3.8 bara condensate in the steam trap (and by the way, please be consistant with your basic data: you first stated the steam pressure was 3.8 and then you changed to 2.8 bara). You cannot assume that the condensate at the pump inlet is saturated liquid with the same temperature and pressure (3.8 bara, 142 deg. C). You must take into consideration that you MUST throttle the steam prior to its entry into the heater shell not only because you require to control the flow, but you also have to control the heat transfer (delta temperature mean).

Do not feel inferior or downcast about not considering these facts in your application. A LOT of engineers have continuously failed to take the various factors involved in this presumed simple heat transfer operation. It turns out that it isn't that simple. I hope this explanation helps to clear up a lot of the unseen factors in this application. That is why I keep insisting for the Original Posters (OPs) to be detailed and clear in their queries and furnish schematics and/or drawings as much as possible from the very beginning.

Attached Files

•  Steam_Heater Rev2.xls   52KB   126 downloads

[go-fish]

@ Art, Kkala

Thank you for taking deep interest in my query. Apologies for prolonging this thread. I have added more information in the attached revised ( Rev 3) schematic. Actually, some of the data is based on assumptions.

To produce HE loadsheet, I have to provide HE shell side inlet pressure (bara) and the heat exchanged (MW) to the exchangers group. Therefore, I assumed 3.8 bara at HE inlet after assuming del P across the steam throttle valve as 0.7 bar (about 15% of CV inlet pressure). I have not done any rigorous Heat Exchanger calculation using a simulation software at this stage. Please correct me if I am wrong in my assumptions.

However, even if the vapor pressure of 3.8 bara at pump inlet is not confirmed, my understanding is that the conditions should still be saturated liquid. For liquid-vapor in equilibrium, the source drum operating pressure and the vapor pressure cancel out while calculating NPSHa and only static head difference should play any significant role in NPSHa. Please correct if I am wrong in my understanding.

NPSHa = Source Pressure Head - Vapor Pressure Head - Frictional Losses Head + Static Head Difference - 0.6 m

Attached Files

•  Steam_Heater Rev3.xls   49KB   107 downloads

[kkala]

[quote name='go-fish' date='02 March 2010 - 01:23 PM' timestamp='1267554790' post='36777'
...To produce HE loadsheet, I have to provide HE shell side inlet pressure (bara) and the heat exchanged (MW) to the exchangers group. Therefore, I assumed 3.8 bara at HE inlet after assuming del P across the steam throttle valve as 0.7 bar (about 15% of CV inlet pressure). I have not done any rigorous Heat Exchanger calculation using a simulation software at this stage. Please correct me if I am wrong in my assumptions.
I believe your assumptions are right .
However, even if the vapor pressure of 3.8 bara at pump inlet is not confirmed, my understanding is that the conditions should still be saturated liquid. For liquid-vapor in equilibrium, the source drum operating pressure and the vapor pressure cancel out while calculating NPSHa and only static head difference should play any significant role in NPSHa. Please correct if I am wrong in my understanding.
Your understanding is same as mine. There is equalizing line too .
NPSHa = Source Pressure Head - Vapor Pressure Head - Frictional Losses Head + Static Head Difference - 0.6 m
Right. You have taken a 0.6 m margin, which is usual. .
Principles of previous posts do not change if steam operating pressure changes a bit.
Comments on the above welcomed.

[kkala]

A few more remarks on the subject:
1. Check whether seen web cases of condensate drum & pumps (of less than 1 m elevation between) have apparent open vent on the drum to atmosphere (point 5 of kkala post, 1 Mar 10). It would be interesting to know.
2. Place antivortex ring on the drum to pump suction, to avoid steam ingression.
Drum LLL height had to be "generously" set for same reason.
3. For the steam pressures reported, globe control valves should be specified for ΔP=10% of upstream pressure and at any case not less than 0.7 bar, according to reliable practices. So your assumption of ΔP=0.7 bar is OK. This is for max process flow (cosidered in your design), being 70% - 80% of max valve flow Q (usually considered by Instrument Dept). We call this flow of 70%Q - 80%Q max controllable flow.

Edited by kkala, 03 March 2010 - 05:43 AM.