4 replies to this topic

[SawsanAli311]

Dear All, 

 

Setting the number of compression stages is a strong function of the discharge temperature limitation needed to protect the compressor system as well as power consumption across each stage. For a multi-stage reciprocating HC  injection compressor (currently being designed), vendor confirmed that four stages would be required for the inlet suction pressure of 1.5barg to be increased to a normal rated discharge pressure of 200 barg.

 

However, when it comes to setting the discharge design conditions of each stage (temperature and pressure), specific excursion margins are applied on top of the maximum operating discharge pressure and temperature in order to allow for potential increase in the operating conditions in case of increase in the backpressure seen by the last compression stage and such that sufficient margins are maintained below the alarm and trip set points. 

 

Now, the question is:

 

I set the discharge discharge design pressure at the last stage to be 250 barg with  discharge design temperature of 180 C. I was questioned by the rotating engineer that now the number of stages will have to be increased to five stages instead of 4! I explained to him that design conditions are set for the system's rating while we use the rated operating parameters in determining the number of compression stages. The 250 barg design pressure was specified to allow some margin for potential increase in backpressure due to wellhead injection pressure requirements knowing that with an increase in the CR from 200 to 220 for example, due to the large temperature term as X+273.15 K , the change in the temperature term would not be highly increased. 

 

However, when it comes to power consumption, would really the discharge PSV set pressure at each stage have to be used when estimating the power and rod load?

 

Therefore, when determining the number of compression stages, we will be basing our selection based on rated conditions or we shall confirm the power requirements even when discharge PSV set point is actually reached? 

[Art Montemayor]

Sawsan:

 

A reciprocating discharge temperature for an oil-lubricated cylinder should be limited to 300 F.  That is what I have always designed for and my shutdown temperature has been 350 F (149 ºC).  While some engineers may argue with this as conservative, I point to others - such as Exxon specifications and many more as those agreeing with this specification.  I would never design for 180 ºC on an oil-lubricated machine - and much less on a non-lube one.

A compressor cylinder is basically a pressure vessel.  Like all pressure vessels, the cylinder has a maximum allowable working pressure (MAWP).  Therefore, since the cylinder is a pressure vessel, it must have a pressure relief valve to protect it and all surroundings - especially human beings in the vicinity.  Normally the cylinder’s MAWP determines the setting of the relief valve that is downstream of the cylinder; however, it is possible that the downstream piping or aftercooler(s) will have a lower MAWP, which will determine the relief valve’s set pressure.  The MAWP of the cylinder should be 10 percent, or at least 172.4 kPa (25 psi), greater than its operating pressure.  I personally recommend that the MAWP be 20% greater than its operating pressure - especially in remote, unattended compressors.

 

I don’t understand what you mean by “CR”.  I keep repeating, after many years, that our members should cease using acronyms that are undefined.  If you mean “compression ratio”, then you are wrong.  Your total compression ratio for the 200 barg discharge is 80 and that for the 250 case, 100.

 

As I’ve explained to you in your thread regarding compressor seals, you should study, read, and fill in all of the API 618 Data Sheet.  Note line #32 on page 7.  By doing that, you should immediately become aware of the important points that have to be considered, detailed, and specified in the design and purchase of a reciprocating compressor.  Refer to the attached workbook to get my comments on how the number of correct stages for a reciprocating compressor are arrived at.  As Bobby Strain correctly pointed out in your previous thread, you need to give your compressor vendors all the information you have.

 

From what little information and limited scope of work you have supplied, I am inclined to believe that you may indeed have an application for a 5-stage machine rather than a 4-stage one.  I have not done any detailed calculations on this, but I would be glad to receive your submittal of your calculations and check them out if they are presented in a readable and logical manner complete with all basic data.

 

The discharge pressures I estimate for the 4- and 5-stage machines (without interstage pressure drop) are:

 

4 stages: 8, 23, 68, & 202 bara
5 stages: 7, 16, 40, 100, & 252 bara

 

 Reciprocating Compressor Stages.xlsx   24.7KB   103 downloads

 

[SawsanAli311]

Thank you very much Mr.Art the highly useful and valuable explanation,,

 

The key issue is that in case we set the relief valve set pressure at the discharge of the last stage oat 20% above discharge pressure of 200 barg then we should definitely guarantee that compressor's discharge temperature would still not exceed 150 C when it has to experience an increase in backpressure up to and below its PSV set point  for example (240 barg)?

 

Additionally, that also means that our power consumption estimation shall also capture the relief valve set point obviously, kindly please correct me if I am wrong. 

 

For the 200 barg, we contacted vendors and they confirmed the stage pressures as : 6.5 , 23.64, 122.39 and 201.57 barg subsequently.  The discharge temperature was maintained below 150 C as per API 618. 

 

I was thinking that in case compressor experiences high backpressure leading to being close to 240 barg (PSV set point at last stage), the increase in compression ratio across each stage will have to be re-distributed to accommodate such increase, leading to higher power consumed as well as higher discharge temperature which can be theoretically predicted via the isentropic or the polytropic calculations (which can use the efficiencies back calculated based on vendor's provided real data). I am aware that the calculated temperature would be in reality higher due to the volumetric efficiency changes as a result of change in compression ratio. 

 

Kindly please advise your opinion 

[Art Montemayor]

Sawsan:

 

My basic opinion on this topic to date is that there is an unrelated importance being given to the setting of the final discharge PSV on your proposed compressor.  It is my opinion that this is leading to a confused idea of what role the PSV plays vis-a-vis the main process that the compressor takes part in and whether it should be a factor in the design of the compressor.

 

As I have mentioned previously: the main and primary purpose of installing a PSV on the discharge of each of your compressor’s cylinders is to PROTECT the cylinders.  The PSV is not intended to relate to the process taking place around it.  It is not intended to participate in your process needs and much less to serve as an instrument within that process.  It is there solely for the purpose of safely protecting your equipment and you.  The set pressures of the PSVs are related to the MAWP of each of the corresponding compressor cylinders - and should not be based on or related to the required process pressures taking place.  The logic I employ here is that you should specify the maximum, normal, operating pressure of the compressor’s discharge at its last stage in accordance with what you normally need to operate your process downstream.  If your process usually operates at 200 barg and at times can elevate itself to 240 barg and still function in an acceptable, safe mode, then you should design your compressor to operate in accordance with meeting the gas requirements at 240 barg.  Your final stage cylinder’s MAWP should, therefore, be rated for somewhere between 10 or 20% above 240 barg in order to conform to good and safe operating practice.  That means your PSV would be set at the MAWP of the cylinder it is protecting and would not function unless there was an emergency or process upset situation increasing the compressor’s final discharge pressure.  If I were to specify the operations of the compressor, I would specify my compressor driver to safely operate during the time that the PSV is functioning.  I would also have relief and pressure alarms with a timed shutdown of the compressor and process.  Depending on how your process functions and what is allowed would determine whether that logic and procedure would be the best practice in your operations.

 

The basic design philosophy I employ is that the process is meant to continue on a steady, normal basis until there is an upset or an emergency.  At this point you must furnish a safe pressure relief and the ability to immediately gain control of the situation by a shutdown.  Depending on your operation, the actuation of the PSV might (or might not) be sufficient alarm for an immediate shutdown.  You normally size the compressor driver to take you safely through the PSV’s actuation without tripping out due to an overload.

 

I am attaching two documents for your perusal and study.  They take you through the steps normally used in specifying and designing a reciprocating compressor.  One of the articles is not to my complete satisfaction in logically explaining the algorithmic manner of designing a compressor but it does touch on some important points.  I hope these help you understand the various steps and procedure that your future compressor supplier has to go through.  It should help you understand what are the important points that you should specify and communicate to your supplier.

 

 Reciprocating Compressor Calculations.docx   2.62MB   84 downloads

 Improving The Process For Sizing Reciprocating Compressors.docx   3.22MB   78 downloads

 

[SawsanAli311]

Thank you so much Mr.Art for your great help!