Dynamic depressurization in Hysys 2

[mamed]

I am working on depressurization of a natural gas pipeline. The length of the pipeline is 25 kilometers and the diameter is 56 inch. The pressure of the pipeline just before the depressurizing, is about 1000 psia and the temperature is 300 Kelvin.
The natural gas is depressurizing to the atmosphere (14.7 psia). The temperature of the out flow in my simulations in HYSYS is about -110 Celsius degrees. Do you think it is true??
I am using a Relief Valve at the end of the pipeline. Is it true?

 

[apetri]

the discharging temperature depends from the way the natural gas goes from 1000 psia to atm,
we prefer emergency shut down systems (ESD) which includes valves + orifices.
I model the orifice as Constant Energy Unit,
the model is available in PRODE PROPERTIES and predicts reasonable discharging temperatures.
Note that for a 25 km 56 inch pipeline at 1000 Psia there is a LARGE amount of gas to discharge,
perhaps, ou may need to consider alternatives...

 

[LeoChem]

Considering the dimension of the pipe and the size of the relief valve (must be small comparared to the pipe), the time of the depressurization must be long and it means you have enough time to exchange heat with the surroundings.
In short, a slow depressurization is nearly isothermal. The model must be ajusted to predict a temperature higher than -110oC.

 

[mamed]

Do you say that the limiter is the orifice area?? The diameter of orifice area in my simulation is 20 inch. Is it practical?

 

[apetri]

>Do you say that the limiter is the orifice area?? The diameter of orifice area in my simulation is 20 inch. Is it practical?

in a ESD system (block valve + RO) the value of discharging flow (supposed critical, which is true in your case) depends from pin, fluid properties and discharging area (orifice),
if the pipeline contains only gas (no liquid phase) temperatures along the pipeline can be estimated considering JT effect or (better) calculated with flash operations keeping in account the heat exchanged with soil, water, air etc. (this is my preferred way, I create a model in Excel with PRODE)
The maximum time specified to reach the final pressure is critical and determines discharging flow (and orifice area),
a 20 inch. orifice seems very large to me, as said in my previous post we prefer different solutions.

 

[MortenA]

If you are using the depres module in HYSYS SS then those results will not be valid. Your pipeline is too long to disregard the effect of gas from the far end that has to flow towards the end where you depres.

As fas as i know, if you have dynamic hysys AND the upstream license then it should be able to do this - but even "normal" dynamic dont maintain a full mass balance for each timestep (or didnt in rev <7.2). This MAY have been changed in 7.3 - im not 100% sure. I guess you try to blow down the pipeline in the API 521 specified 15 minutes? FORGET IT! Do a rough calculation and see how much gas you need to remove! (Use HYSYS to determine density initially and multiply with the volume.

So i dont think that 20" sound wrong (or right) - i think its you assumption that you can depresusrize within your time (im just guessing 15 minutes). I would also like to state that im certain this is not a requirement for pipelines - but for topside piping.

Best regards

Morten

 

[mamed]

Dear Morten
Thank you for your response

I am using transient simulation in HYSYS and I know about the calculation mechanism.
Exactly my problem is not in the simulation of the process. The problem is that the temperature at the end of pipeline in the case of a 20" orifice diameter (for full depressurization in 90 mins) is less than -110 Celsius degrees, and I do not know that it is practical or not.
Actually, I need your industrial experience and sense.

Best.

 

[apetri]

163 K could be a big problem (for the pipeline),
I did suppose your software did calculate that temperature for the gas at low pressure, after the orifice, which could happen going from 1000 psia to atm.

As said in previous post I prefer to create my own model which allows to control the details,
possibly it will not be extremely accurate
(I input just some rough estimates for the heat exchanged in the different sections of the pipeline)
however the results of the simulations (time steps of 1 minute integrated over 100 minutes or above) are quite reasonable (compared with a specific application developed by a company with large experience in this area and in-field measurements),
I use different tools (PRODE and EXCEL) but perhaps you can do the same assembling a different model with your software, really it should not be too difficult since it's easy for a gas pipeline (for a gas+liquid you would need to consider OLGA or LEDAFLOW),
modern simulators offer many features and allow to complete complex tasks in a limited time,
then the problem becomes to understand what the produced result means )

 

[mamed]

Dear Apetri
Thank you very much for your time and consideration

 

[MortenA]

Most dynamic imulation handles heat transfer based on fairly common film theory and outside resistance eg film if submerged/exposed or if buried as an equvalent heat transfer resitance based on a formula taking into account pipe sized, burial depth, soil conductivity etc.

So this methode has (at least) one weakness: It dosnt work when theres no flow - and at the far end you will se an almost isenthalpic depressurization with no flow (its just "flowing away). Not knowing your excact problem i would ask you to check the temperature just a little bit down the line (seen from the far end). Is it MUCH different (maybe even so much that its ok?

 

[MortenA]

small typo: isentropic not isenthalpic

 

[OldDoc]

indeed it seems a strange result,
perhaps in these cases it would be preferable a step by step solution as suggested by apetri,
at least you will be able to evidence the problem (if any),
unfortunately a (black box) unit allows only a limited access and control.