What Correlation To Be Used In Hysys Pipe Segments For Steam? 1

[AliZirgach]

Just wondering what is the best correlation to be used in a pipe segment of HYSYS for steam? (the closest one to reality for finding the pressure drop of a pipe).

 

[MJCronin]

The problem of pressure drop in steam piping systems is a bit more complex than the simple, straightforward methods of liquid pressure drop (Darcy methodology) as outlined in Crane TP #410

As the steam progresses through the system, it looses pressure and begins to change state, which affects the pressure drop.

Saturated steam systems act differently than superheated steam systems. "Rough" internal

As heat is lost through insulation, the pressure drop is also affected. Make sure that the insulation is accurately evaluated in your model.

Crane #410 has a general approximate method that can be used.

Please complete the thread and let us know of your final comparison.

regards

MJCronin
Sr. Process Engineer

 

[MortenA]

What MJCroning states is true - but HYSYS evaluates properties at each internal elemten - and you could thus just cranc up the number of internal element. The trouble really starts if you reach suturation point and start to experience condensate.

I dont think theres any of the correlations avaialable that has been developed for steam/condensate flow specifically.

Best regards, Morten

 

[katmar]

As stated by Morten, the biggest risk for errors is if you have significant quantities of condensate forming. In general practice where steam lines are properly insulated and trapped the condensate is ignored (or made part of the overall safety factor).

If the steam is dry the answer depends on the pressure drop. If the pressure drop over the line is less than 10% of the absolute upstream pressure then you can safely use the incompressible model - i.e. the Darcy-Weisbach formula for liquids - unless you are looking for extreme accuracy. See the discussion in the example at

http://www.katmarsoftware.com/examples/aioflo-example08.htm

. This example is actually for air, but the 10% rule applies to all ordinary gases and vapors.

If the pressure drop is above 10%, but the velocities are in the normal industrial piping range, then you can generally use the isothermal compressible model which is hardly any more complicated than the incompressible model. The only extra information it requires is the upstream pressure and it takes into account the expansion and acceleration of the gas as it flows down the line.

If you are working with very high velocities - for example with high pressure vents to atmosphere - then you should use the adiabatic model. If you use the isothermal model in a situation like this you would over-estimate the pressure drop (and under-estimate the flow rate). The errors are usually not high and if you are sizing a vent to be safe then this error will work in your favor, but if you are trying to calculate emissions for compliance reasons you would be under-estimating the risk.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[gocougs1]

I typically use the PRO/II default Beggs-Brill-Moody correlation and had good results. The other methods were either way off or practically identical to BBM and are not really worth exploring. The correlation does well at normal velocity but at near choke conditions it can deviate substantially, where I would consider using a different method/approach.

 

[KernOily]

You didn't say the state of the steam(superheated or saturated). If superheated, and you think you won't get cooling down to saturation and thus phase change, use generalized correlations for compressible flow (or use the 10% rule as katmar said). If saturated, use BBM as gocougs1 said. That's what I use and it generally gives agreement with field measurements within 20%, which is Good enough for the girls I go with, those being oilfield girls.

 

[AliZirgach]

Thanks all for answering. My steam is saturated and it becomes two phase along the pipe (coming from OTSGs). What I have seen so far, comparing to plant data, the pressure drop with Beggs and Brill (1979) correlation in HYSYS V7.3 is almost double that of other HYSYS Pipe Segment correlations such as HTFS or Gregory Aziz. HTFS or Gregory Aziz pressure drop seems closer to the plant pipes current pressure drop. If somebody has an experience or know something directly related to what I am talking about I would appreciate sharing it here...