6 replies to this topic

[j_r_m_c]

Hello

in my company we are thinking to connect some tanks to flare.

the inbreath will be N2
and the out breath to the flare with a check valve to avoid the return.
simple!!
but unfortanely the N2 blankting can fail and O2 can enter the tank, and then during the out breathing we will putt O2 in the flare system.

what you reccomend? i read some opinions about install O2 analiser in the flare line.
can i read some documentation about this?

thus anyone have exprience with this?
Regards

[proinwv]

Let me answer your question with a question. Would the analyzer be able to to respond and take the necessary action in advance of any O2 entering the flare?

You are correct in that a N2 blanket can fail and if it does, the tank must be protected from vacuum as well as the process from the ingress of air. A second blanketing valve, in parallel with the first, and set and a slightly different pressure will offer redundancy but will not help if there is a failure in the supply of N2. If the backup valve were connected to a different source of N2 that also would offer more protection.

I could say that controls could be added to monitor N2 supply to the blanketing valve, as well as tank (positive) pressure indicative of a functioning blanket valve and system. O2 could be monitored at the vacuum vent inlet as the precursor.

But will these be responsive enough for your system or personnel to take action in time? I think that you realize that you must deal with the possibility that O2 will enter the flare system.

I am hopeful that someone will add some experience based recommendation to my simplified response.

Also, you might consider posting this question in another, more appropriate, forum, but please reveal that it is also posted here. The administrator may be helpful in this and I will contact him about it.

[DB Shah]

Dear JRMC

"in my company we are thinking to connect some tanks to flare."

As per your description of exsiting system, I presume that the tanks are atmospheric storage tanks. Kindly confirm. If this is the case kindly note that the atmospheric storage tanks are designed for very low pressure, in the range of ~ 200 to 400 mmWCg. Normal backpressure at the flare KOD is >2000 mmWCg.

Confirm the tank design pressure v/s Flare header operating/maximum pressure.

DB Shah

Edited by DB Shah, 03 January 2011 - 06:15 AM.

[proinwv]

I agree, you must know the MAWV.

[demank]

"but unfortanely the N2 blankting can fail and O2 can enter the tank, and then during the out breathing we will putt O2 in the flare system."

O2 is the same dangerous as air (containing oxygen) with may cause to tank corrosion, isn't it?
The use of nitrogen is to prevent the ingress of air into tank.
Why don't you spare nitrogen supply?

In case of nitrogen (or other blanketing gas) failure, then let the air fill the tank as blanket (IMO).
So I think no need to install Oxygen analyzer.

[proinwv]

demank, I don't understand why you are recommending,

"In case of nitrogen (or other blanketing gas) failure, then let the air fill the tank as blanket (IMO).
So I think no need to install Oxygen analyzer."

I thought that they did not want O2 into the flare.

[Art Montemayor]

Paul:

I believe that I recognize a potentially important and hazardous situation with which I have had past field experience in resolving. Therefore, following your request, I am offering my experience and recommendations on how to handle such a scenario and hope that it can be of service to all.

Unfortunately, the Original Poster (OP) has not given us a good, accurate, and complete description of what his problem is. The OP has failed to:

• State the type and physical constraints on the tank(s) in question. Are the tanks Storage Tanks or Process Tanks? What are the Maximum Allowable Working Pressure (MAWP) and Maximum Allowable Working Vacuum (MAWV)?
• What are the liquid fluids stored in the tanks? What are their flammability points? What are their ignition points and conditions?
• What are the sizes and the locations of the tanks within the process area?
• How is the gas blanket routed to the flare? Is a blower required to do this? Or is the tank pressure sufficient to overcome the flare pressure requirements?
• What type and make of pressure relief and vacuum relief are used on the tanks?

All of the above information plays an important role in arriving at a practical, reliable, and safe operating resolution regarding the possibility of blanketing gas availability failure.

Since we don’t have all the necessary Basic Data yet, I can’t recommend specific and direct solutions to the described problem. However, I can describe what I did to resolve a similar situation.

I have always been very concerned about the possible loss of nitrogen blanketing in a storage tank and the subsequent creation of a partial vacuum that would draw atmospheric air into the tank. This is something that has also bothered a lot of other engineers. The fact that it is very costly to justify a continuous and available back-up Nitrogen supply makes it important to prepare oneself for a possible blanketing gas failure.

My application was a Tetrahydrofuran (THF) storage tank with Nitrogen blanketing. The THF is a very volatile and flammable liquid. It was very important to keep Oxygen (air) away from it. Our storage tank was an API 350 design and we had no flaring requirement. Please refer to the attached workbook to see what we did to protect ourselves from this situation.

Basically, I can add the following comments to this thread:

• Normally, as you have mentioned, we accept the introduction of air into an API tank in order to avoid a vacuum and save the tank. We later purge and re-establish the required nitrogen blanket. HOWEVER, this is done when we are made aware of a partial vacuum condition taking place within the tank. Normally, this is never the case – unless we instrument the tank and relief valves. Owners – unfortunately – still do not like to spend money on storage tanks and there is always a minimum of instruments and thought that go into a storage tank installation. Nobody seems to care – until an accident or disaster occurs. This is why I did what I describe in the workbook.
• Normally, the fluid in the storage tank will not be “damaged” or subject to combustion if air is introduced into the same tank – at least not right away. Oxygen in contact with a combustible does not constitute a flame or explosion – at least not until the “third” ingredient, an ignition, is introduced. In other words, a blanket of methane gas in a storage tank mixed with air will not explode or catch fire until a spark is introduced. If the tank is grounded and ignition points are eradicated from it, the condition may be a nervous one, but it can be operated safely with precaution until original conditions are re-established. This is done every day out in the “oil patch”, where crude oil is blanketed with natural gas in an API tank with a vacuum breaker that admits atmospheric air. I do not recommend this for a chemical processing facility, but it is a tolerable and operable situation that can be handled with precaution.
• The OP states that a check valve will be used in the out-breathing line to the flare. This check valve, although recommended, lends more pressure design requirement to the Tank’s MAWP due to the pressure drop it adds to the system. A “standard” atmospheric design tank cannot be used. DB Shah has correctly raised this alarm.
• The OP seems to be concerned with putting Oxygen in the flare system. I do not believe this is a dangerous possibility. Any Oxygen will simply not burn in the flare – but it will contribute to the efficient burning of any combustible that is fed to the flare (as is the case, 100% of the time). I would not be concerned with this point – unless there is something else the OP is not telling us.
• Installing an Oxygen analyzer as the OP describes does nothing to mitigate any hazard. It is analogous to worrying about the hen house after the fox has left. If there is a legitimate concern for having Oxygen in the flare line, then it should be PREVENTED and not analyzed.
• The plain fact that you need a positive driving force (a pressure in excess of approx. 2,000 mm WC = 2.85 psig) as DB Shah has noted, means that there is probably sufficient time and opportunity to do something manually in the event of a blanketing gas failure in keeping the tank(s) at such a relatively high pressure. By installing, recording, alarming, and monitoring the pressure inside the tanks the operators should be able to safely control such a scenario. However, they have to be given the engineered and designed equipment to be able to react to such a situation.

I hope this experience and comments help.

Attached Files

•  Protecting Storage Tanks from Blanketing Failure.xls   54.5KB   240 downloads