Design Pressure of Piping Systems

[Achman]

Hi Guys,
I am designing a pumping and piping system to B3.1.3 pumping water at 900 kPa. The dead head pressure of the pump is 1300kPa.
I am trying to find ways to set the design pressure lower than the pump dead head pressure. If we put instrumentation and controls in place to shut the pump at 1000 kPa, does the code allow us to use 1000kPa as design pressure ?
If this is not allowed, is it acceptable to install a PSV to bring down the design pressure to 1000kPa ?
Thanks for the input.

 

[XL83NL]

Yes

 

[LittleInch]

The problem is whether the instrumentation and controls can actually do what you want. Pump shut off can occur very fast if the action to initiate this event is also very fast (e.g. valve closure). Even if valve closure is slow, the final pressure rise can still be very fast because it is only the last bit of the valve closure that actually stops the flow.

Also it might make starting the pump more difficult. Many pumps like to start against a shut in head to reduce motor load, but if the instrumentation trips the pump every time you can't start the pump.

Having a normal operating pressure of 9 bar and a design pressure of 10 bar gives you virtually no head space to set your alarms or trips. You could easily find the pump tripping a ot due to small changes in pump inlet or outlet head.

so in theory xl83nl is correct, but in practice and taking into account other issues.... not so sure.

Why can't you just design your system to 13bar? Class 150 is usually 18 to 19 bar depending on temperature.

PSVs are notorious for either "chattering" or simply opening at up to 10% below set point. Even pilot operated one only give you 3-4% below set point as a margin. Also you don't want these going off every time you start or stop the pump - they tend not to re-seat properly and then just leak all the time.



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[junkpile]

You could, but I wouldn't. Best route would be to either install a pump with lower output pressure, or just design your piping to be able to handle the pump max.

 

[KevinNZ]

B31.3 allows pressure over design pressure for short periods. See 302.2.4

 

[Achman]

Thanks For the Responses,
The system is a DN 600 HDPE pipeline. In most of the sections of the pipeline the surge pressure drives the rating but in some sections the pump dead head pressure drives the rating. Using higher PN ratings for the pipeline to be able to handle the pump dead head pressure is extremely expensive. ASME B31.3 doesn't allow short time over pressure for the non-metallic piping. We have pressure transmitters and flow meters on the pump discharge and I am thinking on using some controls to shut of the pump if flow is zero or when pressure goes up with a PRV as backup and just wanted to make sure this is allowed by the code?
The design code is AS 4041 which refers to B31.3 for non-metallic piping.

 

[LittleInch]

That makes more sense.

Even so the margin between operating pressure and design/troop pressure is too low in practice. You will need a pilot operated relief valve for sure. Code compliance doesn't mean much if your system keeps tripping or leaking through the relief valve

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[davefitz]

It really sounds as it the better choices are either redesign the pump for a lower dead head pressure or increase the design rating of the piping. Some codes permit short term 15% overpressure during events similar to a pump dead head event, but some valve codes do not permit more than 10% overpressure .

If those choices are not selected, then consider other precautions that the codes define for mis matched design pressures . For the common case of a natural gas pipeline desing pressure being much higher than the downstream consumer pressure, once set of rules is to have all of the following protections ( if full capacity releif valves not provided):
- fast acting stop valve that will close within 1 second based on a hard wired pressure switch signal
- 2 independent pressure reducing valves in series
-small leakage relief valve
-scheduled calibration of the presure switch

"...when logic, and proportion, have fallen, sloppy dead..." Grace Slick