9 replies to this topic

[Supriyo Mukherjee]

We are talking of a cryogenic vacuum insulated storage tank for Liquid Nitrogen. Normal Operating pressure of the tank is 8kg/cm2(g). temperature around (-)171 degree Celsius.

 

In this tanks we provide bursting disks rated to burst at around 14 kg/cm2(g). The set pressure for normal PSV attached to the same line is 10.5kg/cm2(g).

 

Now we have observed that the bursting disks rated for pressures around 14kg/cm2(g) pressure ruptures at 7 or sometimes 8 kg/cm2(g). We realised that it is the quality of the rupture disk that is failing us.

 

Now there are two cases once a process engineer gave the temperature of the rated bursting disk is ambient(40 deg C). In that case also the result was same. Generally on the other hand we provide the temperature as saturation temperature at burst pressure.

 

I would like to know if there can be some insight s for from various learned persons that can help out to single out the root cause.

[fallah]

Hi,

 

There are some vague points in your query and lack of a simple sketch is an additional problem to understand your exact issue well. For example you mentioned BD and PSV are in the same line and having set pressures with 3.5 kg/cm2 difference that cannot be logical. Pleae clarify those points and upload a simple sketch....

[Supriyo Mukherjee]

Hello Fallah,

 

The maxm operating pressure of the tank is 8.5kg/cm2(g).

The PSV set pressure kept as 10.5kg/cm2(g).

Bursting Disk is kept for ultimate safety of the tank with an hydrostatic test pressure of ~15kg/cm2(g).

So keeping a safe margin BD is set at 14 kg/cm2(g).

 

I hope it clarifies the system design. I am attaching the sketch.

The basic doubt in my mind is why the BD is bursting at much lower pressure i.e. 7-8kg/cm2(g). Is it the BD quality is poor or is there anything wrong with my specs to vendor in which I gave 40 degree celsius at burst temp in one case and (-) 162.5 deg Celsius in one case. But in both cases the ruspture pressure for bursting disk was around 7-8 kg/cm2g in practice.

 

Please refer page 2 for the sketch.

 

Regards,

 

 

Attached Files

•  Full page photo.pdf   207.07KB   16 downloads

[fallah]

Hi,

 

Bursting disks are temperature sensitive devices, then in my opinion one cannot specify a BD for such different temperatures and same burst pressure and would expect in both cases the BD ruptures at the same preset burst pressure. In fact, a BD cannot cover such both cases with huge difference in temperatures at the same burst pressure.

 

Have you asked vendor to explain about this matter? It might vendor has designed the BD for an average temperature between those of the both cases. Please let's have the BD data sheet if it's possible and if you cantacted vendor let's know the result...

[Supriyo Mukherjee]

Mr. Fallah,

 

Actually we are tried some research on this. The same bursting disk is not specified with diffrent temp's. these are two separate bursting disk.

 

As you correctly said that these are temperature sensitive, so at first we used to give -162 degree celsius as burst temperature(saturation temperature of Liq nitrogen at burst pressure mentioned).

 

Now as it was bursting at pressures much lower so we decided to give ambient temperature as the burst temperature i.e. 40 degree celsius.

But the result was same.

 

We asked the vendor of bursting disk about this problem but never got any satisfactory replies. 

 

But yes recently I gave an idea to the site mechanical team to install a cover (form of canopy) over bursting disk. After this it did not rupture upto 9kg/cm2(g), and as there was no more overpressure in the tank we could not see if it bursts at higher pressures. we thought may be impurities,dust hitting bursting disk may be a problem so we introduced canopy over bursting disk.

 

Regards,

[Neelakantan]

hi

 

AA) there are two aspects in purchase of BD (or for that matter of fact any engineering item); one is correct specification and second is correct product to match the specification; does your purchase spec specify FAT and sample testing?  has the company witness report of test for the lot you bought?

BB) when you say you tested them, what temperature and what rig did you use to test them?

CC) as in PSV specification, the BD burst pressure is again a differential pressure;  since the nitrogen is the stored product and discharged to atmosphere, gauge pressure is okay for the specification.

DD) as Fallah has mentioned, temperature is a critical aspect (as the stress value depends on T); the correct specification is a tricky one here; (this is because, the fluid hold temperature is -171 deg C, but since the PSV/BD are in dead legs there might be only the ambient temperature, ie 40 deg C that you mentioned;) we would be very much interested to know the temperature during normal operation at or around the BD; have you checked it any time?

EE) from your sketch and your write-up I understand that originally the BD assembly is left to atmospheric condition and that now you have put a canopy (without restriction) over the assembly; was there a goose pipe or simply the discharge (and the BD) was open to ambient?

FF) an important query as an operating engineer: I presume you found that the actual tank BD burst even though the operating pressure was much less than the PSV and BD pressure and started intiating the enquiry, test etc? what is the operating mode of the tank? is it while filling this happened? and can you improve your sketch to show the inlet and outlet lines? I am asking this because your sketch shows a pressure relief  control which means while filling nitrogen may be vented to the atmosphere under pressure control; what happens while emptying? how the pressure is maintained? is there any gas balancing or only vaporisation is the mode for pressure? this is important to know because most convex -domed- discs will burst on a much lower negative pressure. is your disk conventional one or reverse buckling?

gg) one thing to note is that a rupture disk is always rated at a range of pressure (~94% to 103%) of the specified pressure unless tighter specs are mentioned (which means a FAT, higher cost...); so your vendor would have given a guaranteed range; what is that range?

 

The above points may not solve the issue, but may point to things...

 

regards

neelakantan

[Supriyo Mukherjee]

Thanks for the insights. Neelkanthan, Most of your understandings are correct.

 

1. Bursting disk were rupturing much below its burst range. So range specification is taken well care off.

 

2. I need to enquire with the quality department whether they have taken FAT or burst test.

 

3. As said earlier inside pipe temperature during operating condition is (-)171 deg C. Outside ambient is generally 35 degree celsius.

 

4. BD was open to atmosphere , no other piping is there. This is what prompted me to install the canopy which is working fine by now.

 

5. As said earlier the pressure rating is 14 kg/cm2 g(+/- 3%). The burst pressure(actual) was 7-8 kg/cm2g.

 

6. At this point of time we  are thinking that absence of canopy was the problem. 

 

 

Regards,

 

Supriyo

[latexman]

This reply is a product of my North America experience.  You may have different circumstances.

 

Rupture discs are typically manufactured to order where each order represents a lot. The ASME Code defines 3 methods of acceptance testing for rupture discs. The most common method requires that at least two discs from the lot be burst tested at the specified disc temperature. The results of these tests must fall within the rupture tolerance.

 

Insist on witnessing the acceptance testing of the next lot of BD's that are purchased.

[Neelakantan]

@ supriyo

 

your point 3 is potently not correct? (-171 deg C); that temp is the storage temp; the line having the RD is a dead-end line with stagnant fluid. its temperature CANNOT be -171 deg C;  if it is -171 on one side and the other side is 35 deg C? not a possible situation.

 

regards

neelakantan

 

PS: why not a snip of the P&ID and the actual photo?

[Supriyo Mukherjee]

Dear Neelkanthan, 

 

I am attaching the P & ID herewith for your understanding.

 

For the question you raised (-) 171 degree Celsius. is right.

 

The tank storage temp is (-) 171 degree celsius.

 

The ambient is 35 degree celsius.

 

Th e pipeline is stagnant fluid so can expect around 30 degree celsius.

 

When the BD bursts, the temperature aroung BD from pipeline were used to be given (-) 162 gegree celsius while providing specifications. 

 

So when I saw that (as  you said the case of stagnant fluid) I changed the spec and gave it ambient temperature and still the bursting disk did rupture at around 7kg/cm2(g). 

 

But the situation is under control after using canopy over BD. Any more insights from members would be good.

 

 

@latexman : you are right someone should witness. which I would try to implement in my organisation.

 

Regards,

 

Supriyo Mukherjee

Attached Files

•  Q-HP TANK.pdf   72.69KB   7 downloads

Edited by Supriyo Mukherjee, 17 May 2014 - 03:46 AM.