Transient analysis S+T heat exchanger to evaluate rupture disks vs PSVs for tube rupture 2

[SNORGY]

A colleague and I have an opportunity to undertake a study to evaluate the transient pressure rise in S+T heat exchangers due to guillotine tube rupture. These units are liquid-liquid, not liquid-gas. We are not in possession of CFD software like AFT Impulse, and I am of the opinion that we need a tool like that. We do have Aspentech and we can run it in dynamic mode, I suppose, but I'm not convinced it's the right tool.

A proposal has been made to do these calculations manually. My colleague has an advanced degree in mathematics, but I do not. That said, we recently did a transient analysis study for high pressure liquid piping flowing through various control valves - manually - and (1) I was the one that did the math and (2) it withstood peer review scrutiny towards establishing the requisite system safeguards.

I am of the opinion that, while we could either purchase software or indeed perform calculations manually, it's probably a moot point due to the very short timescales involved that would only prove that non-reclosing devices are forced in the design. We might be better off putting our attention into safeguarding the relief effluent system. However, I would nonetheless like to ask if someone might know of a specific technical reference that I could use for these calculations, or else, if there is specific software available that will not cost upwards of $10 K?

 

[EmmanuelTop]

Recently we asked several providers of transient flow software to provide us with quotations for a few licensing options. AFT Impulse was ~1,800 USD for 6 month lease option. That is quite affordable, I think - not only for this particular case but it is a great value for money in my opinion. Downside - applicable for liquid systems only.

Apparently the dynamic case can be accurately modeled in Aspen Hysys (see attached article). They key thing is to define fluid properties and system volumes, so that the model can accurately predict pressure rise in time.

Dejan IVANOVIC
Process Engineer, MSChE

 

[SNORGY]

Thanks ET.

 

[georgeverghese]

If the relief stream is either substantially liquid or all liquid downstream of the relief device, would avoid hooking this stream in to the overhead gas HP flare / relief network. A dedicated liquid / 2phase relief collection header going into the flare drum through another feed nozzle would be preferable. This avoids high velocity liquid slugs that could rip the HP gas flare lines off their supports in the event of a coincidental plant ESD or other similar high throughput gas flaring event.

You'll need plenty of room between maximum operating press on the HX LP side and RD set press if you want to minimise the risk of accidental rupture. From an operations safety perspective, setting HX LP side design press = HP side would be better.

 

[SNORGY]

Thanks for that, georgeverghese.

Absolutely nothing to be found fault with concerning your recommendations.

These exchangers are, apparently, existing units already equipped with RD's, with the effluent system (details for which are as yet undisclosed to us) already in place. All we are being asked to investigate is the comparative response times between RD's and PSV's in a tube rupture case to see if the RD's can be replaced.

 

[georgeverghese]

So these people arent too happy with these RDs' - these are notorious for rupturing during liquid hammer transients, which could happen if some one closed the LP exit valve too fast.

To throw into the basket of options to study for faster reponse relief devices, a buckling pin activated relief valve may be considered also? These can be setup to enable quicker replacement and have better integrity in indication when the pin has buckled in comparison to rupture discs.

 

[SNORGY]

Thanks, ET & gv.