Relief Valve Exhaust Piping for different operating pressure lines

[RobsVette]

Hi Guys,

I am having an arguement with someone here in our office and have a question regarding best practice for steam relief valve discharge piping connections.

The question is "can discharge piping from relief valves of different operating pressure steam lines, for instance 150 psig medium pressure and 15 psig low pressure, be tied together into the same vent header? My thoughts on this subject has always been that they should not be because of the back pressure produced in the header by the discharging steam would typically be higher in a discharge from a 150psig valve as opposed to a 15 psig valve.

Best practice would be to take all relief valves from the medium pressure system and run them out through a seperate header and take all teh relief valves from a low pressure system and run them out through a seperate header. I assume through some careful engineering and coordination you could run them all together if you are careful of the back pressure but was wondering if anyone had any code references or insight on this?

thanks,
Rob

 

[davefitz]

You would need to first determine the applicable or governing code for the piping system, then read that particular code. And you also need to read the valve mfr's requiements for the specific type of relief valve .

For the example you gave , the answer likely is that you cannot connect the 15 psig vent to the 150 psig vent . In general , the backpressure of the vent must not cause a reduction in rated exhaust flow and must not prevent the lifting of the relief valve at its set pressure.

Most relief valves have "acoustically choked flow" across the valve orifice plate, which implies the exhaust pressure must be less than 50% of the set pressure in order to achieve rated flow , using absolute pressure. The 15 psig ( 29.7 psia) relief valve must then have a back pressure not greater than 14.9 psia ( 0.2 psig) to avoid restricting its relief flow. That implies the common vent pipe would have to be really big, ,sometimes called an aquaduct.

A high backpressure can also prevent some types of relief valves from lifting altogether.

"Whom the gods would destroy, they first make mad "

 

[moltenmetal]

If the header is sized to give a total accumulated backpressure not more than the maximum permitted back-pressure on the lowest set relief device, then you can combine the discharges into said common header. It is necessary to determine which devices would be relieving at the same time by examining the causes of relief events, and it is this coincident relief flow that determines the sizing of the header.

Will that requirement generate a header of practical size, relative to two separate headers? Maybe, maybe not. For steam, with the destination of the relief being the atmosphere, two separate headers may be the most economical. For process reliefs, your destination is probably a common knock-out drum, flare or incinerator or scrubber or other such device, in which case either way you're ending up in a common header, so you have no choice but to size that header based on the back pressure requirements of the lowest set pressure relief device.

For relief valves, the highest permissible header back-pressure which will not affect the device's rated relieving capacity depends on the design and construction of the device. For ordinary devices, the highest back pressure permitted is generally 10% of the set pressure. For balanced bellows devices it can be quite a bit higher, but those devices are more expensive and can be less reliable.

 

[Drexl]

Generally you can combine outlet, see for example EN 764-7 ch 8.2.3.

It shall be shown that the back-pressure doesn't prevent the operation of the safety valves as described in previous posts.
The conservative and often correct way to size the outlet is then to assume that all valves are blowing simultaneously.
However if you have a case with many, say valves going to the same header and can demonstrate that the relieving scenarios are independent of each other I don't see why you shouldn't be able to size the combined header for less than sum of all flows.

As always the applicable codes and standards is what should be followed.