13 replies to this topic

[KS2010]

Dear all,

I am working on a project of storage tank installation for a flammable liquid. It is an atmospheric storage tank designed for the pressure of +56mbarg and vacuum of –6mbarg. Stored material is flammable and therefore, tank is provided with N2 blanket line at 20mbarg. The tank will be filled and emptied using a single pump.

Please see the attaché P&ID for the tank. Sorry, I could not format it properly which would cause some problem in viewing it. Please not that pump discharge line to tank is 40mm and not 25mm as shown in the P&ID.

There are three pressure protection devices/provisions. (1) 6” Liquid Overflow line – This line is installed with pressure relief valve [PDPV 2355/2] which is set at a pressure of 40mbarg (2) 6” pressure vacuum vent valve [PDPV 2355] which is set at a pressure of +25mbarg and vacuum of –2.5mbarg (3) 24” Emergency pressure relief man way [PDPV 2355/1] is set at a pressure of 45mbarg.

My queries are as below:

Personally I think, the set pressure [40mbarg] of 6” relief valve [PDPV 2355/2] at liquid overflow line is wrong as liquid overflow line would be my primary protection device in the event of tank overfill and therefore, PDPV 2355/2 should be triggered first. Set pressure of PDPV 2355/2 should be greater than the N2 blanket line [20mbarg] and less than set pressure of 6” pressure vacuum vent valve [25mbarg], let’s say 22mbarg. However, if I do so, during tank filling, PDPV 2355/2 will be activated first and NOT pressure vacuum vent valve [PDPV 2355].

How do we decide/calculate the set pressure for PDPV 2355/2?

If the set pressure of PDPV 2355/2 is let’s say 22mbarg and it activates during tank filling, will it create any problem? If it will not create problem, what will be the use of having pressure vacuum vent valve whose set pressure is 25mbarg apart from vacuum protection?

Client wish to offload a tanker at 20m3/hr and empty a tank to reactor at 10m3/hr. I see a problem in tanker offloading due to high velocity. As the pump discharge line to a tank is 40 mm, the line velocity is 4.82m/sec [somehow [client is OK up to a line velocity of 5m/sec] and Raynolds Number is about 220,000 which is very high. Do you think, such a high velocity/turbulent flow will create any problem in terms of static discharge? I can see another problem of higher pumping cost due to higher pressure drop at such a high velocity and it’s very difficult to convince the client to increase the pipe size as they have ordered pipes/fittings etc.

Do you see any other problem in the attached P&ID?

Your help would be highly appreciated.

Regards

Edited by KS2010, 13 October 2010 - 04:36 AM.

[KS2010]

Would appreciate any pointers please!

[Art Montemayor]

KS2010:

I’ve spent some time reading and studying your P&ID and I have some issues and questions as well as comments. Because you submit your P&ID as an object on a Word document, I am unable to pin comments on the pasted drawing. That is why I always recommend Ops submit their drawings on an Excel spreadsheet. It is so simple and direct to make comments and markup a spreadsheet – as well as add sketches and calculations.

You probably haven’t used our SEARCH feature and found a collection of Excel workbooks that I have uploaded for members on many other past occasions, so I’ll attach it to this post once again.

I have done your basic scope of work many times; however, I don’t know your total, specific scope so that I have a lot of questions and comments.

My initial, basic comments are:

• See and study my workbooks. Look at the way I set the DEAD BAND on the tank instrumentation. This is very important – as you are now finding out.
• To get a decent and workable dead band, you MUST HAVE A DEFINITION OF A WORKABLE MAWP AND MAWV for the storage tank in question. If you don’t have the CERTIFIED MAWP and MAWV of the tank (as is often the case in almost every tank application I’ve worked on and assisted), then you must first mechanically rate the tank before proceeding any further. I know this hurts and is embarrassing, but it must be done.
• Supply all the data and information on your transfer pump. You haven’t done this so it is near impossible to comment on the practicality of the controls proposed.
• You are getting a high velocity because your piping is too small. Your pipe is a nominal 1-1/2” pipe and you should be using a 2” nominal size (2.067” = 52.5 mm ID). Always state your pipe as either nominal size or with the ID. You haven’t stated either, so I had to do the calculations from your stated velocity. I have, in the past, standardized on a transfer rate of 150 gpm (34 m/h) and 2” sch 40 pipe for this type of operation.
• What is the capacity and dimensions of the tank?
• It is difficult to read your drawing. How are you controlling your transfer pump flowrate to the reactor?
• How are you protecting your tank against vacuum?
• Why are you vapor balancing the tanker-truck and the storage tank through what appears to be the nitrogen supply instead of a direct, 2”, independent line between the tanker and the tank?
• I am against your using a pressure signal to activate a tank overfill relief. Tank overfilling is a result of excessive volume – not pressure. You should be controlling the volume of fluid in the tank, and not the pressure in order to safeguard against an overfill. I presume you are using a Radar (microware) type of level detection device. That is what I would recommend.

I will furnish more comments when you return with more information as requested above.

Attached Files

•  Art\'s Tank Relief & N2 Blanket.zip   1MB   275 downloads

[fallah]

The storage tank manufactured based on which standard?API 650? or........

Seems the set points are not rightly adjusted.

The overflow configuration on outlet of PDPV isn't a common practice for these tanks.

Hope i can assist more than above based on your further clarification and my more thinking.

Regards

[CMA010]

Scrap the overflow PDPV-2355/2 and go for an instrumented protective system against overfilling, see API 2350 or EN 13616. If PDPV-2355/2 is a breather valve or conservation vent it will not protect the tank against overfilling, those valves are not designed for liquid releases. A non-code pressure relief valve set at 40 mbarg also doesn't sound like a very good idea. Alternatively you could go for an horizontal overflow with a liquid seal. Personally i would not recommend that option. If overfilling occurs you still end up with flammable product in your bund (just like in the current configuration).

Why do you have a PDPV on the nitrogen line to the tanker? And why is it set at 15 / -2.5 mbarg when you want to control the pressure in the storage tank at at 20 mbarg? Are 1" breathers even available?

Keep in mind that the set pressures of the emergency vent en the breather valve should be based on the percentage above set pressure at which they achieve full flow capacity. In the case of 10% above set pressure the emergency vent should be set at 50.4 mbarg and the breather at 44.8 / -5.4 mbarg (or 44.8 and 33.4 / -4.8 for 20%) since you are not allowed to exceed the design pressure. Obviously you can choose "nicer" values, like 50 and 40 / -5 mbarg.

Considering the design pressures of 56 / -6 mbarg, is this tank originally designed acc. BS 2654?

[KS2010]

Dear Art,

Thanks for your reply and explanation.
I am really sorry for not attaching P&ID in excel format but I have to gain skill/speed/accuracy to do so.

Below are my answers and comments/questions against yours.

1) Thanks, I have studied the DEAD BAND criteria.
2) I understand the importance of MAWP [I have also read one of your reply to a member asking about MAWP and design pressure] but unfortunately we don't have that information. However tank is brand new with design pressure of 56mbarg and design vacuum of -6mbarg.

If this is the case what should be the DEAD BAND for this tank based on above design pressure?
3)There will be a single pump for filling and emptying the tank as shown in the P&ID. The tanker offloading [tank filling] will be at max 20m3/hr and process transfer [tank emptying] will be at max 10m3/hr. Here the discharge pipe size is 1-1/2" for both duties. Therefore tanker offloading will be limiting factor for pump. At present, the differential head for tanker offloading duty is about 58m while for the process transfer is about 32m. Client is against using VSD for pump and therefore, we proposed using ROP of about 15mm [calculated] which will give the required pressure drop of about 28m [from the proposed pump/pump curve] in the process line. However, I am worried about pumping cost. Pump supplier has proposed pump motor of about 11kW for this duty [pump motor for 15m3/hr tanker offloading duty is 5.5kw, which is now double for 20m3/hr flowrate]

4) Tank's total capacity is about 45m3 with diameter of 3m and tan to tan height about 6.1m.

5)Controlling the flowrate to the reactor: Massflow meter is installed in the process transfer line + high level signal from reactor will trip the pump

6)Tank vacuum will be controlled by suitable pressure vacuum vent valve [PVDV 2335] which is set at -2.5mbarg vacuum set point.

7) Even I am against using the common line for tanker back venting and nitrogen blanketing but it seems client want to save piping cost.

What are the disadvantages of using common line though?
Is 2" independent line not big for tanker back venting?

8) Yes, even I cant understand why client wants to install the liquid overflow line. They used to have liquid overflow line for many tanks in past but then tanks were not using nitrogen blanket, therefore here I dont see any use of having liquid overflow line. Also there are two level protections, one is the HLA [high level alarm] pot which will trigger the alarm plus stop the pump and second protection is: [as you rightly mentioned], radar type level transmitter to protect the tank against high-high level.

Thanks once again for your time/help.

Regards,

[KS2010]

Thanks Fallah,

Not sure, but tank might be manufactured based on BS EN 14015 [earlier BS 2654].

Yes, even I have never seen pressure relief valve on overflow line and I want something which can replace the existing overflow protection system [6" line + pressure relief valve]. Client has not installed this tank/pipings etc yet.

I think two level protections [as I mentioned in point no.8 of Art's reply] are enough to protect the tank. Or I would say let's have the third hard wired level protection [in case these two fails] but not this overflow line.
What do you think?

[KS2010]

Thanks CMA010,

I agree with you regarding proposal of scrapping the existing pressure relief valve [PDPV 2335/2] at the overflow line but wondering what could be the best option as an replacement. PDPV 2335/2 is the pressure relief valve and I am against it as it will not do a job.

Also, the PDPV set at 15mbarg is on the tanker back venting line [this will allow liquid vapour to vent back to the road tanker during tank filling operation] which is sharing a common line wiht nitrogen blanket line. I am not sure it this valve should be just a pressure relief valve or it should be a breather valve as shown in the P&ID. I have to find out more details about this valve.

Thanks for your idea about pressure settings but I will need to validate them.

Yes, this tank might be built based on BS EN 14015 [earlier BS 2654]

[Art Montemayor]

My main reason for insisting that you obtain a certified rating for your tank is that ultimately YOU are responsible for any safety-related incidents or failures related to the tank’s safety devices. If you are not in a position to 100% confirm that the MAWP and MAWV (or the design pressure and design vacuum) are correct, then you really have no business going any further in setting pressure and vacuum levels for the tank. You should have certified documentation as well as the back-up calculations that demonstrate that the values you are working with are the correct ones.

The DEAD BAND you should be using for this tank should be based on the design pressure and vacuum in accordance with the type of instrumentation that you are using. I cite and show some examples of the dead band and how to apply it on the various examples I drew up and show on the different work sheets. Use this as example and guide. This will quickly tell you if you have a workable instrumentation dead band or not.

As often happens, the embarrassing situation arises when the project engineer becomes aware that he/she has “put on the shoe before the sock”. In other words, a new tank built with certain pressures specified as the design pressure and vacuum – but you didn’t base these values on applying the needed dead band for the instrumentation to work successfully. As I have mentioned previously, this is the unfortunate case many times. And what is worse is that the design pressure and vacuum almost always never are apt for the application. Many times the new tank has to go back to the fabrication shop and be reinforced to take a higher pressure and deeper partial vacuum. You simply HAVE TO HAVE an adequate dead band or your instrumentation will fight each other.

In my opinion, -6 mbarg of vacuum is simply not enough for the design vacuum. And -2.5 mbarg of set pressure is unrealistic. You have to have a wider vacuum span for the vacuum relief to function correctly.

Using a common line for nitrogen fill and for tanker truck balance is not only foolish and ignorant economics, it will cause an inbalance in pressures when unloading. This is the type of foolish “cost savings” that often cause accidents.

The same type of foolish engineering is applied with respect to trying to stop an overflow with a pressure switch. No experienced engineer in his right mind would approve this and I take this opportunity to advise you not to do it and that it is very dangerous to do this. You could easily rupture the tank using this “idea”. One way to protect against a potential overflow is to employ microwave level detectors – on the tank itself and on an overflow surge tank at ground level that is balanced with the tank but kept there to catch any overflow. Once again, this is extra costs, but guess what? We are in the chemical processing business and it is expensive. If your clients don’t like the magnitude of capital monies it takes to have a safe and efficient design, then they should invest in something else.

[fallah]

[size="4"][font="Times New Roman"]
The same type of foolish engineering is applied with respect to trying to stop an overflow with a pressure switch.

Dear Art,

Not being observed the above mentioned method in the originator/other posts.Would you please clarify about the method of stopping an overflow with a pressure switch.

Regards

[Art Montemayor]

Fallah:

I don't know how else to "clarify" my opposition to using a pressure signal to detect a pressure increase within a practically non-compressible fluid. Granted, all liquids are compressible - albeit in very small degrees - but the speed of such a signal would have to be extremely rapid in order to mitigate literally "raising the roof", because that is what would probably fail first: the cone roof would rip off at the seam with the cylindrical shell. That is the weakest part of a fixed, cone roof tank. I have never seen this attempted or suggested - and for good reason.

There are other ways -much more direct, practical, and proven - that are being used in the field. One is to use High Level and High-High Level alarm and shutdown within the tank itself. The other is to install a surge or overflow, external tank and install the level alarms and shutdown there. The shutdown can be with a block valve on the liquid inlet to the tank or turning off the transfer pump. I have used both.

[CMA010]

In my opinion, -6 mbarg of vacuum is simply not enough for the design vacuum. And -2.5 mbarg of set pressure is unrealistic. You have to have a wider vacuum span for the vacuum relief to function correctly.

Why would -6 mbarg not be enough? This is a standard design pressure limit for "low-pressure" and "high-pressure" tanks acc. BS 2654 (Current EN 14015 specifies -8.5 mbarg for "low-pressure" and "high-pressure" tanks).

[KS2010]

Art,

Thanks for your reply. I have further queries as below [in bold fonts].

"In my opinion, -6 mbarg of vacuum is simply not enough for the design vacuum. And -2.5 mbarg of set pressure is unrealistic. You have to have a wider vacuum span for the vacuum relief to function correctly."

As mentioned by CMA010 earlier, why -6mbarg is not sufficint? This tank is designed based on BS 2654 where in 56mbarg and -6mbarg are standard design pressure and vacuum respectively for the atmospheric storage tank.

"Using a common line for nitrogen fill and for tanker truck balance is not only foolish and ignorant economics, it will cause an inbalance in pressures when unloading. This is the type of foolish “cost savings” that often cause accidents."

How will it cause the pressure imbalance when unloading? We intend to isolate the nitrogen blanket line using MV003 [ or installing another manual valve after PI/002 on N2 blanketing line] during tanker unloading and intend to isolate the back venting line by MV006 while using nitrogen blanket. Sorry, am I missing something?

"The same type of foolish engineering is applied with respect to trying to stop an overflow with a pressure switch."

Which pressure switch? We have not usded pressure switch anywhere!! OR are you talking about PDPV 2355/2?

As I mentioned earlier [Point # 8 in my above reply to Art], we already have high level alarm + radar type level transmitter. My question is still as it is: If we have two level protection devices, then what is the use of installing liquid overflow line or PDPV 2335/2?

Thanks and regards

Edited by KS2010, 12 October 2010 - 04:54 AM.

[fallah]

Yes, even I have never seen pressure relief valve on overflow line and I want something which can replace the existing overflow protection system [6" line + pressure relief valve].

Normally,the overflow line in atmospheric tanks would be terminated to a seal leg eqipped with a siphon breaker.