8 replies to this topic

[veerblue]

Hello Folks,
 
I am working on the hydraulics of a propane pump, pumping propane to LPG mixer. For the same I have to decide the inlet pressure of the mixer in order to prevent flashing. There is one design case at 42°C, and I have to do the same at 10°C.
 
Case I : at 42°C
 
The feed to the LPG is propane and butane, with mixing ratio 50:50. The properties of the components are as following:

 

            

                        Vap P kPa(a)       Vap P kPa(g)        Density kg/m3          Viscosity cP

 

Propane              1519                       1418                       563.4                     0.083

Butane                 485                         384                         540.5                     0.141

LPG                     1040                       939                         501.2                     0.1077

 

 
 
In the datasheet of the mixer, the operating pressure is mentioned as min 1690kPa(g), in order to prevent flashing. The mixer pressure drop is mentioned as 100kPa. It follows that the minimum outlet pressure can be 1590kPa(g), which is much higher than any of the individual components vapor pressure.
 
Moreover, the vapor pressure of the product (LPG: 939kPa(g)) is also much lower than the outlet pressure of the mixer ( 1590kPa(g)).
 
Since the outlet pressure of the mixer is much higher than the operating pressure of the LPG tank, there is a control valve installed in the downstream of the mixer.
 
The schematic is attached with the post.
 
I have to decide the inlet pressue of the same system at 10°C. The individual properties are as following:
 
 

at 10°C

                                Vap P kPa(a)      Vap P kPa(g)      Density kg/m3   Viscosity  cP

Propane              662                         561                         515.7                     0.114

Butane                 155                         54                           590.6                     0.180

LPG                     484                         383                         562                        0.145

 

 

 

How should I decide the inlet pressure of the mixer?
In perticular, for the first case (at 42°C), the margin above the higher individual vapor pressure component (propane) is 272kPa (1690-1418 for propane). What can be the basis of such margin?
 
Any insights on the topic would be really helpful.
 
Regards,
Mahavir

[Bobby Strain]

So, why would it be any different?

 

Bobby

[veerblue]

_{Dear Bobby,}

 

_{The figure 1690kPa for the minimum inlet pressure of the mixer is based on the vapor pressure of the lighter component propane (1418 kPag), in order to prevent the flashing of propane.}

_{Since the temperature has changed to 10 degC now, the vapour pressure of Propane has also changed to 561kPag. Now, since the vapor pressure has changed, the required minimum inlet pressure shoud be changed as well. I hope it makes sense.}

 

_Mahavir

Edited by Mahavir, 06 December 2013 - 01:17 AM.

[PaoloPemi]

(as general consideration) for a pure fluid or a mixture in liquid state,

near bubble point  if you reduce the temperature you reduce

also the vapor pressure and (as conseguence) the possibility

of flashing at some specified pressure,

from your description it would seem that if the system works without

problems at 42 C from the point of view of flashing it should work

also at 10 C (all other parameters being the same),

however consider that going from 42 C  to 10 C also other 

propertiers as density and viscosity will change and this may

affect operating.

Finally, to calculate true vapor pressure of HC mixtures (bubble point)

as well as other properties you may use a EOS as PR or SRK

which provides more accurate estimates,

you may use a library or a simulator for that (I use PRODE).

[S.AHMAD]

1. If the hydraulic system works at 42C then it works also for system at any lower temperature.

2. You should be worry if the new working temperature is higher then 42C.

3. Understanding the relationship between vapor pressure and temperature explains the reason.

Edited by S.AHMAD, 09 December 2013 - 03:07 AM.

[curious_cat]

 

 

Since the outlet pressure of the mixer is much higher than the operating pressure of the LPG tank, there is a control valve installed in the downstream of the mixer. 
 
The schematic is attached with the post.

 

I don't see a schematic. In any case, are you sure it does not flash over the control valve? What's the max deltaP over the valve? You seem to have barely 150 kPa or so available?

Edited by curious_cat, 09 December 2013 - 05:50 AM.

[S.AHMAD]

Mahavir

1. You do not need any control valve downstream of the mixture valve. However you may need flow control valve for the propane and butane streams so that you can mix the right mixture of LPG composition (50:50?).

2. The pressure of the LPG storage at 10C will be the same as the the vapor pressure at that temperature (i.e. 383 kPag - as per you data)

3. The pressure at downstream of the mixer shall be the LPG storage pressure PLUS line & fitting pressure drop from mixer to storage PLUS also any elevation change. Let say the pressure is 450 kPag. If the mixer is set at 100 kPa then the pressure upstream the mixer shall be 550 kPag.

4. The flow control valve for the propane & butane will auto-close or open to give the pressure 550 kPag at the mixer inlet regardless of the pump discharge pressure (of course the pump discharge pressure must be high enough to overcome all the pressure drop downstream of the pump). Let say the pump discharge pressure is 1800 kPag. Then the pressure drop across the flow control valve shall be 1800 - 550 = 1250 kPa. (assuming negligible piping & fitting frictional loss)

5. Hope the above short explanation is good enough for your understanding. Propose you draw and attach the schematic diagram so that I can verify the accuracy.

6. GOOD LUCK

Edited by S.AHMAD, 10 December 2013 - 03:47 AM.

[veerblue]

Dear All,

 

Thanks a lot for taking your time to delebrate upon the problem. As I did the calculation for 10C, the differential pressure came out to be significantly less than 42C case. For the 10C case the frictional losses increased (by 30kPa) due to increase in density and viscosity, but the static losses were significantly less (decreased by 280kPa, due to decrease in vapour pressure of both the upstream and downstream components) than that in the 42C case. Hence the governing case is still 42C case.

 

Dear Mr. Ahmed,

Yes we do need the control valve in the upstream of the static mixer in all the component lines to control the flow rate. But, we do need a control valve in the downstream of the mixer as well in order to maintain back pressure to prevent flashing of the lighter component. The inlet pressure of the mixer needs to be a minimum of 1690kpa in order to prevent flashing (as per the datasheet). The mixer pressure drop is around 100kPa, resulting in downstream pressure of 1590kPag. This pressure is too high for the product tank (LPG: operating pressure 939kPag). Hence we do need a control valve in the downstream of the mixer. In my case the pressure drop is coming around 520kPa for the control valve. I hope I am clear.

 

Mahavir

 

P.S. : Can someone help me figure out how to upload the schematic image. Thanks in advance.

[S.AHMAD]

Dear Mahavir

 

1. If yo think that you need the downstream control valve that means you do not fully understand the hydraulic system.

2. Suggest you study hard or ask other experienced process design engineers. Do not waste money installing something that is not necessary.

3. Hope you heavily consider my point.

4. If you are worry about the pump discharge pressure is too high then the right thing to do is save electricity by trimming the impeller size.

5. Finally I would like to say that most young engineers believe that pressure is CREATED by pumps. As a matter of fact, pump only deliver the volume flowrate that meet the pressure requirement of the hydraulic system. That is why in designing a pumping system we need to develop the SYSTEM CURVE and find the right centrifugal pump having the right performance curve that meet the requirement of the system curve.

Edited by S.AHMAD, 20 December 2013 - 05:26 AM.