Bursting Disc Set Pressure and back- pressure determination 2

[s2468]

Hi All,

I have a reactor which has a design pressure of 6.9 barg, where an acid hydrolysis reaction takes place. The batch is reacted to up to a maximum of 125 degrees using a steam heater jacket. The steam comes in at 2.5 barg. The reactor pressure during normal operation is 1.1 barg. The reactor is fitted with a bursting disc which has a set pressure of 2.5 barg upstream of a relief valve which has a set pressure of also 2.5 barg, but I am looking to remove the pressure relief valve and only have a bursting disc where I will replace the current bursting disc. Currently, the relief valve is exposed to vent with an exhaust fan to get rid of the relief gases, so it experiences a back-pressure.

The causes that can lead to overpressure the reactor is the failure of the steam control valve supplying heat to the reactor jacket with the reactor to be blocked in. The steam pressure upstream of the jacket is 2.5 barg and therefore the maximum the reactor can achieve is 2.5 barg which means it cannot overpressurise the reactor.
At present, it is not possible to overpressure the reactor as the maximum pressure of the inlet streams is less than 6.9 barg. But it is good practise to keep a relief device on the reactor. Should the new bursting disc have a set pressure which accounts for the design pressure of the vessel or should It still have a set pressure of 2.5 barg?

Before I remove the relief valve and replace the current bursting disc with an appropriate one, I need to be able to quantify the back pressure the bursting disc will experience due to the exhaust fan to make sure that the bursting disc will be able to handle the back pressure. How can I measure the back-pressure? Would I measure it directly, or use a fan performance curve?

Would greatly appreciate the help

 

[Latexman]

"But it is good practise to keep a relief device on the reactor." It may also be required by law.

"How can I measure the back-pressure?" A well placed pressure gage or transmitter is invaluable. If you are not so lucky, calculations with a fan curve will work.

The burst pressure is up to you. Make it high enough to eliminate small incidents that can be contained.

Good luck,
Latexman

To a ChE, the glass is always full - 1/2 air and 1/2 water.

 

[georgeverghese]

Think there is a flaw in this perception, unless you can clarify :

"The steam pressure upstream of the jacket is 2.5 barg and therefore the maximum the reactor can achieve is 2.5 barg which means it cannot overpressurise the reactor."

From your description, one of the relief scenarios could be abnormal heat input to the reactor as a result of temperature control failure.

This statement would be true only if the contents of the reactor, at the time of relief, has the same saturation vapor pressure characteristics as water - steam.

To complicate things some more, how can be you dead certain (with a probability that exceeds 99.9% say) this exhaust fan will be running at the time of relief?

 

[s2468]

Yes, this would be as a result of a control valve which supplies heat to the reactor jacket with the reactor blocked in.

The contents of the reactor is water.

That is true, I cannot be certain. But won't I need the back-pressure the bursting disc experiences as a result of the exhaust fan to determine what the burst pressure should be in case the fan is on during relief?

 

[georgeverghese]

You say the pressure of the supply steam is 2.5barg - what could the max normal supply pressure be ? - the relief estimates should be based on this pressure.

Yes, since the BD is a dp device, the minimum bursting pressure would be if the fan is on, while the max bursting pressure would be if the fan is off. Also take into account that the set to burst pressure tolerance is much more on a BD than it is for a spring loaded conventional RV.

The current arrangement of BD in series with RV may be due to the corrosive nature of the reactor contents ? There may be some operational availability / flexibility concerns if the BD were to rupture due to corrosion, so materials selection for this BD would be key to maintaining reactor availability for the proposed new arrangement with BD only.

 

[s2468]

The pressure is reduced on entry to the plant to 5 barg by a pressure reducing set. The design pressure of the vessel is 6.9 barg.

Would you suggest I use a pressure gage to find out the maximum burst pressure and the minimum burst pressure.

Yes I agree! The material of the bursting disc is very significant, many are constructed from PTFE which is acid resistant.

 

[LittleInch]

I still can't work out why you want to set the set pressure of a bursting disc at less than the design or MAWP whichever is higher. Your vessel seems to be set up to be strong enough to handle all that can be thrown at it.

Bursting discs have a tolerance on them anyway and are a non re-settable device. The pressure in the vessel would drop to whatever your back pressure is and I guess your water mixture would go into superheated steam and react very violently. Your venting system may not be able to cope with that and could easily break, which is maybe why the relief valve is there to relieve excess pressure only, but maintain asset pressure to prevent mass ejection of superheated steam.

I can't see why this disc is there other than corrosion on the relief valve, but I would just get the right valve and seats myself. Normally bursting discs are installed when a low ish pressure rated vessel can be exposed to a very sudden high pressure event , something alike a HX tube rupture, when a relief valve just can't open fast enough not to burst the vessel. This doesn't sound like one of those so the BD is a puzzle.



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

 

[s2468]

Currently the pressure drop in the piping is greater than 3% of the relief valve' set pressure which therefore means the pressure relief system needs to be reviewed. One solution was to get a smaller relief valve to therefore reduce the pressure drop, but another possible solution was to just remove the relief valve completely and only have the bursting disc as a sufficient pressure relief device.

The set pressure of the bursting disc right now is 2.5 Bar. I was looking to get a bursting disc with a higher set pressure to account for the vessel's design pressure. But I would firstly do this by obtaining a value for the back pressure for when the exhaust fan in the venting system is on to make sure the bursting disc is able to handle this back pressure. I also thought to remove the relief valve, as a bursting disc would be a sufficient pressure relief device and it would not need to be serviced every year like a relief valve.

I understand and agree with what you are saying in terms of a reason for having both a relief valve and a busrting disc, but I do not see the problem with just having a bursting disc. Regardless, the vessel is able to handle high pressures. I don't quite understand why the pressure in the vessel would drop to the back pressure, could you please clarify? And these scenarios would occur during relief conditions?

Thanks in advance.

 

[georgeverghese]

Dont think you'd need to use a PG to estimate what the min and max burst pressures would be for a given set pressure - since the BD is dp device, the actual bursting pressure is just the sum of the built up backpressure and the set pressure. And account for the tolerance in set pressure also to derive these values.
I would work this relief scenario on the basis of a 5barg steam pressure to the reactor heating jacket (given what you've said so far), assuming the final regulating TCV has gone wide open as a result of controller or TCV failure.

 

[s2468]

As the minimum and maximum burst pressure would be determined by the back-pressure as you said it would depend on whether the fan is on or not, I would use the pressure gage to determine the back-pressure for when the fan is on?

 

[LittleInch]

The big difference between a bursting disc and a Pressure relief valve is flow rate and residual pressure in the vessel.

A busting disc is essentially a purposeful full bore rupture evacuating to atmosphere or the back pressure. That pressure will rapidly fall to below your burst pressure whereas a PRV once you fall below your set pressure re-seats and maintains a pressure in the vessel of approx. 2.5 bar

Hence you need to look at what happens once the BD goes bang. You will get a massive instant velocity in your venting system which will be far above what a relief valve will give you. Hence the pressure at the BD exit point will be whatever the back pressure of your vent system is. I can't see this being 2.5 barg, hence the pressure in the vessel will fall rapidly leading to a fairly violent boiling of what is now superheated water.

If your venting system was only designed for a mass flowrate coming out of a relief valve then simply replacing that with a BD could result in a mass flow rate of 10 times or more.

I had a bursting disc go one on a water bath heater which was only at about 10 psig and a very short vent line and you could see the steam cloud / jet from about a mile away. It also went on for some time until the water fell below 100C.

I'm just saying be careful about the difference between relief valves and bursting discs when it comes to the vent system and also how the vessel contents react to a lower pressure than the burst pressure in terms of mass flow being ejected out of the gaping hole in your pressure vessel....

If you have issues with 3% then also look at pilot operated valves taking a tapping from the vessel itself, but if you're setting the pressure much lower than the DP/MAWP I can't see this being an issue? realize it's not perfect, but can't you just allow for it?



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

 

[s2468]

I understand what you are saying, thank you for the insight. In terms of what you say about mass flowrate, relief gases are discharged to the venting system so the fluid in contact with the BD is air.

HCl which has a max pressure of 2.5 barg, and towns water which has a max pressure of 6.3 barg are charged into the tank, so I do not know why the relief devices initially had been set pressure of 2.5 barg. I'm considering having 2 bursting discs in place with a vapour space between due to the corrosive nature in case a leak was to occur and so therefore the second bursting disc can contain the leak. But of course before I decide on this, the back pressure when the fan is on should be measured, especially after considering what you have said.

The first bursting disc should be set to 6.9 barg which is the design pressure of the vessel and the second disc should be set to 5% above the MAWP (6.9 barg) according to ASME. That’s what I had thought - to just allow for this greater than 3% pressure drop, but it would not comply with the standards as I was told.