Hydrostatic pressure rise 8

[op9]

I am interested in knowing how to estimate or calculate the expected pressure rise in a closed pipe full of liquid propane or butane (initially pressurised well above its vapour pressure of course) when it is heated. My particular case at the moment is a length of 32mm nom. bore steel pipe, about 15 mtrs long, filled with liquid butane at 900kPag. If it is initially at say 25deg C, what would be the expected pressure rise when heated to 45deg C. Obviously being an incompressible fluid the usual gas laws do not apply, but the pressure DOES rise very sharply. Any help would be appreciated.

Rod

 

[butelja]

I can tell you from experience that most hydrocarbon liquids in carbon steel schedule 40 pipe will exhibit a pressure rise in the range of ~40-100 PSI/°F, or ~500-1240 kPa/°C. This is independent of pipe length. It is mainly of function of the pipe diameter/thickness ratio, the liquid compressibility, and liquid coefficient of thermal expansion. The limiting case for the pressure rise, using an infinitely rigid pipe (D/t=0), usually is about 1.5 times what it is in schedule 40 pipe.

The pressure rise is usually so severe that only a very few °C temperature rise will be sufficient to blow a gasket somewhere in the system. Therefore, a thermal relief valve or expansion chamber (accumulator) is usually installed to protect any section of pipe that can be blocked in full of liquid.

 

[pkhenry]

Why is it that the propane or butane are not reaching their equilibrium, i.e. reaching vapor pressure at the temperature. If you initially fill the line to a pressure above the vapor pressure the line pressure will drop to the vapor presure for that gas temperature. Conversely once the gas is at it's vapor pressure for the specified temperature any further increase in temperature will result in a rise of pressure to meet the vapor pressure. If you need some properties for propane or butane try <A HREF="

http://www.standby.com/"

TARGET="_new">

http://www.standby.com/</A>

under the technical data.

 

[butelja]

Phenry,<br>
If the pipe is full of a 2 phase (liquid/vapor) mixture, then your answer is correct. However, if the pipe is 100% full of liquid, then there is no additional space for any of the liquid to expand into vapor. This is when the severe pressure rise can occur, likely leading to a leak somewhere in the system. This is true not only for hydrocarbons, but any liquid.

 

[benriemens]

Hello all,
Butelja speaks of accumulators. Are those accumulators specialy designed for this service or are they &quot;general&quot; purpose accumulators?

Does anybody know links to manufacturing companies? Or has anyone more info on this subject??

Grtz
Ben

 

[butelja]

For process applications, I typically have used accumulators that have a metal bellows rather than an elastomer bellows. Some of the manufacturers are:

http://www.flexicraft.com/index.asp

http://www.metalbellows.com/

For noncritical and/or noncorrosive applications, an elastomer bellows may be more cost effective.

 

[TD2K]

I've just done some PSV sizing for an ammonia system for thermal relief and was unable to find the coefficient of explansion listed. However, it's easy to calculate:

beta = rho'/rho*dT)

beta is the cubical coefficient of expansion
rho' is the change in density between two temperatures
rho is the initial density
dT is the change in temperature.

For the density, I used the PVT data for ammonia in Perry's.

I got a beta of 0.0014 per deg F. I also checked the approach for water (which is commonly available) and got virtually the same number as what I've seen published, 0.0001 per deg F.

 

[TD2K]

Opps, make that:

beta = rho'/(rho*dT)

 

[mharper]

mdry,
I have not yet found anything on the expansion of ammonia. I have been searching for days. I really need to get something so I can get my spreadsheet finished. I have found several different values for the density of liquid ammonia, can't really find an answer on the speed of sound of ammonia, and the k-value for the worksheet isn't quite working out. anyone got any help!?!?!?

 

[rusman]

Thermal expansion as a result of the blocked-in liquid exposed to the higher fluid operating temperature within the system. For example, blocked in liquid and expose to hot sunny condition. If the operating condition of the fluid is above hot sun condition, then it is not a problem, as the temp won't rise above its initial operating condition. However, if there is a steam traced around the pipe (this is particularly true for the black product liquid, eg LSWR, in order to maintain the temp above its pour point or cloud point).

Normally, if a blocked in liquid having total volume of in excess of 500 l, then a thermal relief valve, TRV (size is normally 3/4&quot; by 1&quot; or sometime 1&quot; x 1&quot; though it is quite rate using this size) is required. The set pressure of TRV is maximum flange design pressure (MAWP/Design pressure) plus 20 %. Refer to B31.3 about this. Though sometime, the set point is the same with flange design pressure based on ANSI B16.5.

Based on previous experience and calculation, every 1 deg C increase in fluid inside temp, a pressure rise of 100 psi is expected. The calculation of pressure rise is independent of fluid volume as hilited by Butelja. It does depend on the behaviour of the fluid (cubical expansion), properties of pipe.

Another option of not having a TRV is by installing a permanent drilled check valve of small bypass of the block valve so that the rise in pressure within the block-in liquid can be expelled to the other reservoir/system/piping.

TRV is normally discharge back to the tankage of to the open/closed drain system. It can also be cascaded to the next line provided the TRV is installed and finally discharge back to the tankage.

I've a MS Excel spreadsheet for this calculation.

Hope this help.

Thanks

 

[22091965]

I am working on PDO projects in Oman and we have been asked to give a solution to a similar problem over here, which I am describing below:

There was a rupture on one of PDO's flowlines. The flowline was &quot;blocked in&quot; at appropriate places, along the length of the pipeline to repair the damaged portion of the flowline.

The blocked in section of the pipeline(6&quot; API 5LX42)is about 4 km. in length carrying crude (High GOR about 1000, 25% BS&W) of API 38 density. The initial temperature of the crude (at the time when the pipeline was blocked) was 18 Deg.C

I am required to estimate the final pressure rise and the eventual pressure of the pipeline if the temperature of the blocked in section reaches a temperature of 40 Deg. C.

It is estimated that this pipeline will stay blocked for about 9 hrs during the daytime before the rupture is taken care of and flowline is &quot;up and running&quot;.

Please suggest solutions on how to work this problem out.

Thanks

 

[TD2K]

Your problem 22091965 is very different from a pipeline line blocked in full of liquid. The gas will greatly decrease the pressure rise compared to a liquid full pipeline.

I would take a first pass look at it by estimating the amount of gas/liquid you have when it is initially blocked in at 18C. You'll need to have a process simulator or some tool to estimate the amount of gas at this temperature and the initial pressure.

Then, restimate the amount of gas at 45C. The change in liquid volume in this case should be minor. You can put the additional gas into the vapor phase and see what the pressure is. As the increase in pressure will affect the amount of vapor released, it will be a trial and error solution.

You should be able to also do a quick check of the 40C final figure. You know the surface area of the pipeline that can 'see' the sun and the radiation from the sun (likely close to 350 BTU/hr/ft2 there). Work out how much heat that is, look at the mass of the pipeline and its contents and see what the dT is for 9 hours. You can also consider heat convection/conduction from the surroundings if that could be significant.

 

[TD2K]

Mharper,

I know this is really late, if you haven't gotten the data you are looking for, densities of ammonia can be gotten from Perry's handbook for Chemical engineering or the ASHRAE series.

 

[DaveHouston]

This is to butelja.
I am curious to know if the the pipe diameter/thickness ratio is used to calculate the change in pipe diameter due to internal pressure.
If you ignore changes in pipe geometry, this problem can be solved by following a line of constant liquid density from the inital T&P to the final T. Graphs of liquid density as a function of T&P are available for propane and NH3 from ASHRAE.
If you account for thermal expansion of the pipe, the radius and length of the pipe drop out when you divide the initial volume by the final volume.

 

[FEBA]

rusman,

I agree with your explanation. However, I prefer use 90% of the ANSI B16.5 as a security factor.
Can you send me your spreadsheet for TRV calculation ??

Thanks a lot !