7 replies to this topic

[Yasmin]

Could you please guide on how to size condensate return line. Service is high pressure steam of pressure 40 kg/cm2g. Steam temperature is around 255 degC. Quantity is 3500 tons/hr. Pressure of the receiving condensate drum is 4.2 kg/cm2g. The existing condensate return line size is 1.5". This line is leaking frequently especially at the bends. The line thickness at the bends was found eroded.This is probably due to high velocity. I checked out one nomogragh given by Spence. Based on that I have resized the line at 4 ". Should I anticipate any other problem if I resize the line to 4"?

[Diederik Zwart]

Yasmin,

I'm going to assume that you mean 3500 kg/h and not 3500 tons/h, because there's no way you're going to get that amount through the piping you're talking about.

Downstream of the steamtrap the 40 bar condensate is going to flash into approximately 20% steam and 80% condensate (depending on the pressure drop across the condensate line which I estimate to be about 2 bar). More is going to flash along the condensate line until you end up with approximately 25% steam and 75% condensate in the 4.2 bar drum.

Pushing this through a 1.5" pipe will certainly erode the pipewall since we're talking about 50 m/s to 80 m/s velocities. For quick estimate sizing I would use no more than 2 m/s at the start of the line, which would lead me to suggest an 8" pipe for your system.

To do your own calculations:
20% steam + 80% condensate (@4.2 + 2bar) has a density of about 15.5 kg/m3
25% steam + 75 % condensate (@4.2 bar) has a density of about 10 kg/m3

By the way: you're losing a lot of energy if you send all this high pressure condensate to such a low pressure drum. Isn't there a medium pressure condensate drum available you can route it to?

Diederik Zwart

[Yasmin]

Yes, you are right. The flow rate is 3500 kg/hr.
The original line size of 1.5" was provided by the Licensor. What could have been their basis? As far as I can make out, the line is sized for condensate only. It has been assumed that no steam flashes out. Am I correct?

We do not have a high pressure condensate drum. The same drum is used to collect the medium pressure steam's (15 kg/cm2g) condensate return.

[Art Montemayor]

Yasmin:

At the risk of coming into this thread on the tail end of a very good response from Diederik Zwart, I want to address the need for all process and project engineers (especially Chem Engrs) to address the simple and direct process that takes place when you condense steam and separate the two resultant phases created.

If you have a source of 40 kg/cm2g saturated steam heating and condensing in a heat exchanger (or reboiler, etc.), you will always create a saturated, liquid condensate that you must remove - usually through the application of a steam trap. A steam trap is nothing more than a specific, selective valve that is designed to drain only the liquid phase (& also expel troublesome non-condensables in some cases), leaving the higher pressure saturated steam to continue its heating process.

The pressure of the saturated steam (40 kg/cm2g in this case) is the only driving force that can expel the liquid condensate into the condensate header or anything else downstream. As you've stated the pressure driving force in your application is over 35 kg/cm2 (40 - 4.2). When you have this much driving force on a saturated liquid, you will FLASH! For your interest, you should look at a steam chart or Mollier Diagram and see how and why this flash vapor is being generated.

You will have 2-phase flow in your downstream header and it will require a size adequate to handle such a mixture fluid. This is one of the reasons Diederik calls for a bigger header size. He is absolutely correct in allowing for the flashing phenomena. I don't know what led your licensor to design the header (ridiculously small at 1.5" diameter) on the basis of no flashing, but Diederik's reasoning is in the correct direction. I haven't checked out his calculations or sizes, but they follow good logic. The header is definitely going to be much bigger. And you will get resultant flash steam vapors - check this out with some simple calculations or refer to the Mollier Diagram.

Hope this helps out.

[hollerg]

A 2 m/sec design flow is treating the line as if carrying 100% liquid, correct? With the amount of flash steam you can size as if the flow is all flash steam at a higher velocity. Sarco/Armstrong & TLV will have design guides that address the recommended line sizes as a function of flash steam.