Checking Existing Bursting Disc For Two Phase Release Scenario 1

[chemks2012]

Dear all,
I have a dilemma and need your help
I have an existing bursting disc [independent] on the reactor in the existing plant which I want to check if it has got enough relieving capacity for the two phase release scenario I am looking into.

Reactor design pressure: 6barg
Bursting disc: 10” bursting disc set at 2.75barg and discharge line 10”
Bursting disc discharge pipe: 1.5m vertical (1off) rise straight from reactor nozzle, 3off long radius elbows and total discharge length of 6m discharges to a dump tank. [Note: I am sure, you don’t need sketchJ]
Two phase release scenario: complete failure of a reactor steam coil below liquid level causing two phase flow release
Two phase release rate: 480kg/sec [of which liquid release rate is 478 kg/sec and vapour is 2kg/sec]

I believe, there are two options and it seems only one option could be right?

Option#1: With above two phase release rate [480kg/sec], the two phase flow pressure drop [using Lokhart Martinelli model for two phase flow pressure drop], in the discharge pipe line is about 90% of set pressure i.e. 2.5barg. This proves that the given 10” bursting disc OK.


Option#2: If I use Omega Method or HEM Method to estimate the required bursting disc diameter, it seems that the 10” bursting disc is not enough and I need about 23” bursting disc!!!


Which option is correct? Am I missing anything?

Waiting for expert comments
Thanks.

 

[Latexman]

Option # 2 is RAGAGEP for relief device sizing.

I don't recall ever seeing Lockhart Martinelli recommended for relief device sizing. Is there a recent credible reference that does?

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[Latexman]

Maybe you should reconsider the scenario. Is the steam coil made of pipe or tubing? What is your testing and inspection frequency? Are there any reasons (vibration, fatigue, corrosion, etc.) that justify considering a complete failure? If not, a slow developing leak may be much more credible. In this case, we use a 1/8" orifice with 0.7 flow coefficient to simulate the leak.

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[MikeClay]

Due to the length of the relief pipe, the rupture disk cannot be sized using a discharge coefficient. The hydraulic capacity of the relief system needs to show that it is capable of relieving. In a two phase situation, a two phase flow correlation is needed. Lockhart Martinelli is one choice. Just make sure that the dump tank vent does not add too much pressure drop.

--Mike--

 

[Latexman]

Personally, I think Lockhart Martinelli (LM) is a poor choice in this case.

LM's data correlated +/- 50% with their correlation. Not so good.

LM is recommended for cocurrent horizontal flow. At least 1.5 ft of the 6 ft total is vertical; there may be more.

LM used experimental data from 1" and smaller pipe. LM is recommended for 4" and smaller pipe. This is 10" and bigger pipe.

LM pressure drop predictions are high for low flows and are low for high flows. Most relief systems are characterised as high flow.

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[chemks2012]

Thanks Mike

Thanks very much Latexman,

Few basic queries before I go deep down.

The vessel design pressure is 6barg and as per vessel design code we can allow 10% accumulation i.e. vessel can withstand total 6.6barg.

Now the 10" BD set pressure is 2.75barg and supplier has given the BD tolerance data as +0% and -10% and suggested that bursting disc maximum operating temperature is 2.75 X 0.9 (tolerance) X 0.95 (safety factor) = 2.35barg.

However, for the release rate calculation, I have ignored all above data thinking irrelevant to my calculation and considered relieving pressure of 2.75barg + 10% overpressure = 3.025barg = 4.025bara. However, as you can see 10% has not been mentioned by supplier. Can I assume 10% overpressure here? I assume in HEM method where is says upstream stagnation pressure is the relieving pressure. Am I correct?

Also, let's assume, if I consider a scenario wherein I assume an orifice leak of 10mm, this disc is OK using HEM method at relieving pressure of 4.025bara.

Once, HEM model confirm disc is OK, do you think I need to perform other calculations to show bursting disc discharge line pressure drop?

Thanks and regards
KS

 

[Latexman]

It is not a fire case, so up to 10% accumulation applies.

The relief line plus rupture disk must be shown to be adequate.

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[Latexman]

Can I assume 10% overpressure here?

You can. With Code allowable accumulation of 10%, your overpressure can be a lot more, yp to 140% if I calculate correctly.

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[chemks2012]

Hi Latexman
You hit the nail on the head. Thanks.
Yes,it's 140%. But my point is how is it possible that bursting disc will open way before it sees 6.6barg (7.6bara) anyway and hence I doubt if could consider upstream stagnation pressure of 7.6bara. If you say yes I can, then why don't we do the same for relief valve?

I doubt if i can even allow 10% over pressure (i.e. 3.025bara) as disc tolerance is +0% to -10%.

Thanks

 

[Latexman]

The current disk will burst at 2.75 barg. Way before 6.6 barg. Code allows you to use a sizing pressure of up to 6.6 barg. So, in your calculations use 6.6 barg. Code allows it. Then, as long as the installation (pipe, fittings, rupture disk, everything) flows more than required by the scenario, all is good.

By the way, is your steam pressure > 6 barg?

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[chemks2012]

Hi Latexman
Thanks for your reply, much appreciated.
Sorry which code? API 520?

So I would say over pressure of 140% and need to use 6.6barg as the relieving pressure? Please comment.

Yes, steam is at 15barg flowing through 1.5"pipe however, the steam flow is restricted by an orifice of 20mm which feeds to a 3" coil in the reactor.

So,when I said full rupture, I have taken the full flow through that 20mm orifice with (allowed 0.9 discharge coefficient).

Thanks a lot in advance

 

[Latexman]

The pressure vessel Code that applies to the reactor. API 520 is a standard, not Code (law). If you don't know the Code the reactor is, that needs to be discovered before calculations are of any value.

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[chemks2012]

Hi Latexman,
Vessel (pressure vessel) code is PD5500.

So let's say if I allow 140% over pressure to the bursting disc, this should be applied to an independent relief valve which has set pressure of 2barg and in this case over pressure of 230% should be applied?

Thanks
KS

 

[Latexman]

Correct.

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[chemks2012]

Thanks a lot Latexman.
Much appreciated your help.

Regards,
KS

 

[chemks2012]

Hello all,

Below is the refined latest version or an extension of this thread discussion.

Just to refresh the discussion with all latest data...


A reactor has an independent 10” bursting disc [BD] set at 2.75barg [with tolerance +0% to -10%]. Vessel design pressure is 6barg with allowable accumulation of 10%.

If I take advantage of overpressure allowable, I can allow 140% so that vessel pressure should be limited to 6.6barg during two phase relief i.e. 2.75 barg + 140%(2.75 barg) = 6.6barg

BD is designed for two phase flow and the required two phase flow release rate is 150kg/sec. If I wish to allow safety factor of 2 i.e. if I wish to design BD for the capacity of 300kg/sec.

I have checked pressured drop at the BD discharge line at 300kg/sec and it is about 1.3bar.

My queries:

Now pipe stress analyst need some info on maximum pressure that could be seen by the BD discharge line and I am stuck and need your help on following please.

1) What would be the vessel pressure during relief? I believe it is 5.5barg [6.6 – 1.3 = 5.3barg]. Please comment. However, I believe the pressure of 5.3barg is all THEORETICAL and vessel will never see pressure of 5.3barg as the set pressure of bursting disc is 2.75barg and it will burst at 2.75barg

2) What maximum pressure would be seen by BD discharge line? I believe it is 6.6barg [5.3 + 1.3 = 6.6barg]. Please comment. However, I believe bursting disc discharge line pressure of 6.6barg is all THEORETICAL and the discharge line will never see pressure of 6.6barg as the set pressure of bursting disc is 2.75barg and it will burst at 2.75barg and the line will see the pressure of 2.75barg + 0.5bar [assumed pressure drop] = 3.25barg


Your help would be greatly appreciated.
KS