Sizing Expansion Steam Header 3

[code1]

Hi,

Can some of the more knowledgable folks please point me in the correct direction.

I would like to size a steam header to reduce the pressure of the recovered process steam (saturated condition) at 600psig to 20psig where it is returned to the deaerator.
The 600psig steam are from 3 separate, but identical lines which will feed into a common header. The 20psig steam is returned to the DA. The total steam flow rate is 3x 6,000lbs/hour.

The steam had been used to do heat up combustion air for a process plant.

What approach should I use to size the steam header?
Thanks.

 

[katmar]

The two basic criteria for line sizing are velocity and pressure drop, which really means a balance between installation cost and running cost. In your case the pressure drop is not too important because you want to get rid of the pressure anyway, so it comes down to a question of what is the maximum permissible velocity.

A velocity that is too high will give you problems with noise and/or erosion. A velocity that is too low will make the pipe expensive. Check the piping handbooks for recommended maximum velocities for your steam conditions - especially the quality (entrained water) of the steam.

Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[davefitz]

I would run the 600 psig steam lines down toward the area of the DA, and use a pressure reducing valve at the DA nozzle. This would likely be an angle valve . Care is needed to ensure the DA has adequate relief valve capacity to match the casualty flow thru the press red valve. This config is used to minimize the diameter of the steam transfer piping.

The max velocity in the outlet section of the transfer pipe ( inlet of valve) should not exceed 250 fps for sat steam, and if the steam is wet, then it may be neccesary to use a lower velocity or upgrade the pipe alloy to P11, for erosion concerns.

 

[MJCronin]

I agree with davefitz, however, I would use P11 or P22 in an "extended nozzle" (just a lengthed section of pipe header with a cap)where the streams are collected.

What size is the inlet to the DA ?

I would prefer a thicked wall for this "extended nozzle"

-MJC

 

[code1]

Hi all,

The 3x6,000lbs/hr lines are provided with 2" or 3" pipes to have a reasonable velocity.

I plan to use either 16" or 20" pipe as the common expansion header, but cannot decide on its volume (and hence its length). The header will be horizontal with two end pipe caps.

A 6" line will carry the LP 20psig steam to the DA which has also a 6" nozzle. The reason of using the common header and line is to minimize the number of openings on the DA and also because the common header will be located closer to the air heaters.

Do I consider the steam specific volumes, flashing of steam and residence time to size the volume? My concern is sizing the volume too small.

 

[insult2injury]

Your referenced 'flashing of steam'. Do you have steam or condensate exiting the air heaters?

I2I

 

[code1]

The air heaters are designed to have condensate exiting at 600psig (ideal case). The heating steam is taken from the steam drum. In reality with heat losses and frictional losses, the fluid maybe a condensate/steam system at a lower pressure when it enters the header. If there is any condensate present, it will then flash. Apology for the term inaccurate term "flashing of steam"

 

[katmar]

code1,

Insult2Injury's question of whether you have steam or condensate, plus your comments on required volumes and residence times, made me realize that your question is not as simple as you first stated it.

It now seems to me that you have 600 psig steam being fed to 3 process air heaters, and what you are trying to do is to recover the flash steam from the condensate exiting those heaters, and to direct the flash steam to your deaerator. If this is the case then you must disregard what I (and davefitz and MJCronin) said before.

Each process air heater must have its own steam trap, or other means of maintaining the pressure in the heater. The condensate from the heaters should be piped to a properly sized condensate collection vessel which can also be your flash vessel. This flash vessel will be sized to disengage the 20 psig flash steam from the condensate and your questions on volumes and residence times become more relevant here. It is actually the superficial velocity of the steam in the flash drum that you should be concerned with.

In this situation your main design problem (apart from sizing the flash vessel) is the sizing of the condensate lines from the heaters to the collection vessel. When you flash 600 psig condensate down to 20 psig you will convert approximately 25% of your condensate to steam. This flash will start to happen in the piping from the traps to the collection vessel. Sizing of these lines is not a trivial exercise and in addition to the flowrates you need to consider the length, changes in elevation and actual routing. It is a problem that should be given to an experienced piping engineer. I would not be surprised if these lines need to be 4" or even 6" diameter. It is not a problem to be solved in a forum such as this. You can get tips here (it is called eng-tips after all!) but detailed designs must be done by competent engineers who have all the information available.

Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[code1]

Yes, I agree that it is relatively non-trivial problem. That's why I am thinking and asking the above questions. Well is just sometimes that the powers-that-are do see it (and other problems) as relatively simple requiring an inexpensive solution (so prob no consultants...). This is a rant suitable for another part of this forum

I know the steam will start flashing along the pipe. I also wonder about the possible errosion, etc of the pipes in the future due to the high two phase flow velocity in the pipes and the subsequent impingement onto the header. But this is the design we are adopting for the plants that we have. So will tackle the process problems (if any) probably a few years down the road.

Thanks for advices above.

 

[insult2injury]

What type of deaerator is used? What connection(s) are available on the deaerator? Location(s)?

I2I

 

[code1]

The deaerator is a vertical tray type integral with feed water storage tank below it.

The process connections to the deaerator are: the boiler feed water, the deaeration steam which is bled from the turbine LP bleed (75psig), demin water makeup and vent.

The process connections to the feed water tank are: the return pipe from the header discussed above, 3 min flow lines from the BFW pumps, and 3 other 2" condensate return lines operating at 75psi.

The DA operates at 20psig, and the equipment are part of a waste heat recovery plant with steam turbine generator.

 

[gerhardl]

katmar: you are of course to the point, giving you a star for saving me many words, and most important for an engineer: knowing when to call in expertice.

Code1: Being no expert myself, and in addition, as katmar says, drawings and full layout is necessary; one suggestion would be for youto contact your local representative from one of the world-wide suppliers of complete steam accessories.

For instance will both Spirax Sarco and Gestra (Tyco) have available layot suggestion and dimensioning tables for different parts mentioned in your question. Both will in addition have available technical papers and manuals for steam equipment.

http://spiraxsarco.com/

http://www.gestra.com/

 

[insult2injury]

code1: If you are going to combine the drains and flash off some steam at the lower pressure header, you may need to consider venting and draining the header separately. If you want to use 1 multiphase inlet to the deaerator, you probably need to keep the header at a higher pressure to prevent flashing until the deaerator inlet. Also, you should review the location of the multiphase inlet to ensure that you will not upset the trays or cause deaerator cracking. Often, these lines are fitted with hydraulic spargers with 1.5:1 or 2:1 ratio on the open area of the holes for dissipation of the stream.

gerhardl: I believe gestra is a flowserve company.



I2I

 

[gerhardl]

insult2injury: you are indeed correct about Gestra, as the link clearly gives. Thank you for the correction!

 

[insult2injury]

gerhardl: with the amount those two conglomerates have purchased, it's hard to keep track.

I2I

 

[katmar]

If I read this correctly, you are intending to use the header as your separation drum? I would not do that. I doubt whether you will find a consultant who would do that for you.

You need to stand up to your bosses because it is NOT simply a question of "will tackle the process problems (if any) probably a few years down the road." Industrial accidents are most unpleasant. People die. Other people are held accountable. Correctly so. Don't let it be you.


Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[insult2injury]

To further katmar's remarks, if the system is not designed correctly, if it unlikely that there will only be a few process hiccups that could be fixed a fews years down the road. More likely, it will cause process failures (ie. inadequate capacity, etc.) that will need to fixed immediately to meet process demands.

I2I