7 replies to this topic

[Tran]

Hi every body,

I'm a fresh man in process, I want to size some PSV for a well head platform. As I know this can be done with HYSYS. Someone can advise me how to do ?

Thanks in advance!!

[paulhorth]

Corleone,

If you are a fresh man in process, you should not be doing this job. It takes some experience - not so much in the actual sizing, but in determining what the relief load should be.
However, HYSYS is not needed and should not be used.

The steps required to size any PSV are summarised below - I expect others will add some further comments.

(1) Determine the set pressure of the PSV. This should be higher than the max operating pressure in the system, by a margin of 5% to 10% at least.
(2) Consider each possible cause of overpressure for your system. In the case of a wellhead platform, with no vessels, this is likely to be blocked outlet in the piping with the wells remaining open.
(3) Determine the MAXIMUM nett flow which can enter the system, for each possible cause of overpressure, when the PSV is at its set pressure. For a fire case relief, the flow is generated internally rather than entering from upstream. In the case of flow from a well or wells, this is difficult ,and you need to consider a fully open choke valve or a choke valve with damaged internals. You will need to consult the reservoir engineers about the maximum flow which can be sustained in the well when the wellhead pressure is at the PSV set pressure.They will need to run their model to estimate this.
(4) Establish the fluid physical properties for each relief case. You may have two phase flow - see point (7) below.
(5) at this point, when for each relief case, you have a flow, a mol weight, compressibility, specific heat ratio and a temperature, you can size the valve for each case using equations presented in API RP 520 part 1.
(7) For two phase flow relief, the orifice should be sized using the HEM (Homogeneous Equilibrium) method developed by Imperial College, which has been extensively validated in research.
(8) The relief case which gives the largest orifice size is then the governing case and the PSV is selected for this case.

But most important of all - think carefully about the upset condition which is giving rise to the overpressure. How does it happen? Can it be avoided? Can the system be designed with a high enough pressure so that a PSV is not needed? For example, on a wellhead platform, the piping to the last downstream block valves can be designed for the same pressure as the wellhead itself, then the PSV is eliminated.

Paul

[fallah]

Hi every body,

I'm a fresh man in process, I want to size some PSV for a well head platform. As I know this can be done with HYSYS. Someone can advise me how to do ?

Thanks in advance!!

corleone

paulhorth has clearly submitted the stages of calculating required orifice area of a PSV.

In general,every PSV vendor has its own sizing software.After calculating required orifice area of a PSV by customer using equations presented in API RP 520 part 1,he can use one of mentioned softwares to specify selected area and orifice designation based on API 526 as preliminary PSV sizing.
All these data,required area-selected area-orifice designation,could be reflected in PSV data sheet handed over to vendor.Vendor would check and verify the preliminary sizing done by customer using its own software.

Just the Required Orifice Area can be specified for vendor and in these conditions vendor is expected to perform the PSV sizing up to specifying Orifice Designation and other required data of customer.

Fallah

Edited by fallah, 09 April 2011 - 01:55 AM.

[Tran]

Dear Paul /Fallah,

Thank you very much for your help. I've been given some guidance procedure which is identical to what you specified. I also have some calculation sheet and now I'm following up with work.

Thank you very much for your help.

Best Regards.

[yeek]

I do use HYSYS for determining the relief load but not in the way you might normally expect.

What I will do is set up a normal operation simulation first and we will call this the base case simulation. Perform mass balance here.

Then I will set up the upset condition. For example if it is for a power failure scenario, there will be unremoved heat duty. I will calculate how much unremoved heat duty there is and just add a heater into the simulation which will upset the normal operation mass balance.

That way you can determine how much to relief.

[paulhorth]

Yeek,

Surely, if you want to simulate the loss of cooling due to power failure, you should be deleting a cooler not adding a heater?

If you try to do this with a condenser as part of a column, you will find that the column does not converge, so your simulation will not help.

Furthermore, a simulation will not help to define the blocked outlet case, a tube rupture case, etc. A very important relief case not defined in a normal simulation is blocked outlet of a centrifugal compressor. For this you need the compressor curves, the maximum power of the driver, the maximum suction pressure. I was involved in a recent project where the relief valves on the compressors had been sized for the "normal" mass balance flow, not the maximum flow. When this was discovered, the plant startup was delayed by five months while a series of temporary fixes were implemented, with the ultimate solution being the cutting out and replacement of the relief valves and flare piping. It cost the client and the contractor many millions.
This scenario is not "double jepoardy" because there is nothing to stop the compressor being run at its maximum capacity, to obtain maximum production, then blocked outlet occurring possibly due to a trip elsewhere in the overloaded plant.

Paul

[yeek]

Yeek,

Surely, if you want to simulate the loss of cooling due to power failure, you should be deleting a cooler not adding a heater?

If you try to do this with a condenser as part of a column, you will find that the column does not converge, so your simulation will not help.

Furthermore, a simulation will not help to define the blocked outlet case, a tube rupture case, etc. A very important relief case not defined in a normal simulation is blocked outlet of a centrifugal compressor. For this you need the compressor curves, the maximum power of the driver, the maximum suction pressure. I was involved in a recent project where the relief valves on the compressors had been sized for the "normal" mass balance flow, not the maximum flow. When this was discovered, the plant startup was delayed by five months while a series of temporary fixes were implemented, with the ultimate solution being the cutting out and replacement of the relief valves and flare piping. It cost the client and the contractor many millions.
This scenario is not "double jepoardy" because there is nothing to stop the compressor being run at its maximum capacity, to obtain maximum production, then blocked outlet occurring possibly due to a trip elsewhere in the overloaded plant.

Paul

Well, deleting a cooler and adding the heater doesn't really make a difference does it? I just add the heater the simulate the effect of a relief scenario. Anyway I work with the simulation files provided by the client and i try not to mess it up by deleting stuff. Its also easier to explain to people what i'm trying to do instead of answering inevitable questions ('Why is there a cooler missing in the simulation file you are working with?'). I just add a heater to demonstrate the effect of unremoved duty due to power failure.

You are right, Hysys doesn't work with blocked outlet cases. In fact, for such cases, I don't use HYSYS at all. I don't use it indiscriminately but I'm just saying, with a bit of thinking, you can use it and it can help you.

I've never worked with a compressor before. I'm only four months into the job since graduation so my knowledge is very limited but I'm trying to learn as much as I can. Sometimes I don't even understand the discussions here and I try to read up before I contribute and participate.

[Quench2014]

Very useful information. Thanks all