Dear all,

Do any one know how to determine the flow rate of the N2 used for blanketing? Can i just base on the standing storage loss formula (from EPA) and determine the N2 flow for blanketing?

Your help is greatly appreciated.

Rgds,

Siak:

I have given - and repeated - a detailed method that I've often used to estimate the Nitrogen quantity required to maintain a nitrogen blanket on a given storage tank or process vessel right here in our Forums. If you perform a SEARCH function, I'm sure you will find what you are looking for.

I don't call the nitrogen consumed for blanketing a "flow rate"; it is used in batch quantities - as required to offset leaks, pump-outs, thermal effects, etc.. A rate is estimated for sizing of the nitrogen make-up feed valve.

Siak,

Yes Art is correct. The quantity of gas is dependent upon liquid transfer rates, as well as thermal contraction or expansion. API 2000 details the requirements.

Also, if this is not enough for you, look on my website for additional information.

dear siak,
We do nitrogen blanketing of various pressure vessel like secondary reformer in our ammonia plant,the thing is u have to know the design pressure of the equipment whom u want to provide blanketting n then apply apt flow rate comensurating with that pressure n u''ll see how ur equipment got blanketed

A lot of companies will have their own specifications for what flowrates to use.

For example, a client I am working for at the moment recommends a flowrate of 60 l/min nitrogen.

60l/min might do wonderfully for a 100m^3 tank, yet be totally inadequate for a 1000m^3 tank. It may also be somewhat dependent on the level of safety you are striving to achieve, though perhaps that wouldn't be a variable for API-650 tanks.

I think the "normal" flowrate is ZERO. There is generally no reason to maintain a flow of "sweep gas" through a vessel. You maintain the vessel's pressure at some small positive value and air/oxygen is excluded. Purge gas flow is induced only when the tank's vapor space pressure falls. That typically occurs when the tank is emptying or during thermal cycling as it cools. As previously indicated, there are established methods for calculating these values.

A huge safety alarm has gone off in my head. I consider anyone who sets a general design flow rate for make-up nitrogen into a blanketed tank as a recommendation for any tank as foolhardy and a dangerous person to have around. Let me give you an example:

You have a nominal 50,000 gallon (189,271 liters) storage tank blanketed with nitrogen and you start to pump out liquid product into a tanker truck (lorry) at a rate of 150 gpm (568 lt/min). And lets suppose that the process engineer who specified, bought and installed the nitrogen make-up feed valve was “too busy” drinking coffee or pinching the bosses secretary and simply specified the company’s “standard” 60 lt/min valve instead of sitting down, doing his job and logically studying the possible design constraints and hurdles. If we are standing alongside the storage tank at that moment, we all know that we probably won’t be around much longer since the storage tank is going to “suck-in” and collapse with a huge and thunderous roar. This would not be the first time this kind of tragic and fatal “accident” would happen – and unfortunately it probably won’t be the last, either.

Anyone who has witnessed a collapsed tank or seen the results of one will testify to the inherent danger that exists when a conventional storage tank is subjected to a slight vacuum. And a goodly partial vacuum can easily be obtained in a storage tank by simply running a pump-out, centrifugal pump on startup without throttling the discharge. What many young engineers fail to take into consideration is that the pump-out pump is usually capable of pumping double (or triple) the rated capacity IF A DEVELOPED HEAD IS NOT APPLIED TO IT. In other words, if the discharge is wide open, the pump will probably try to pump to “the end of the curve” – a phrase used to describe when the load on the pump is removed and it tries to match the conditions imposed on its discharge: no pressure! And how many process engineers took THAT into consideration when they specified their make-up nitrogen feed valves?

The total, net effect is felt inside the tank as its liquid level falls at a rate that was not “designed” – while the nitrogen make-up feed valve tries to allow nitrogen into the tank at a rate for which it was designed: a rate that is far below that required to maintain a positive pressure in the tank. The result is a sudden and excessive vacuum. This is where a vacuum breaker will literally save you – if one was installed.

It is for these reasons that I don’t describe the nitrogen makeup as a flow rate. Although I consider it to be a batch type of operation, the actual, real rate at which it is required can be very high. Doug, I believe, is well aware of these process potential traps and rightly brings out the notion that the "normal" flowrate is ZERO. I totally agree with the concept. Ideally, during the majority of its operating life, an inerted storage tank has NO need for nitrogen. And there is no need for purge gas during normal operation. But a process engineer is certainly needed to logically design the system and prevent any hazardous events from occurring.

Every inerted tank is a specific and unique application that should be studied and engineered based on its own merits and underpinnings.

Of course Art and lets not forget that API 2000 will provide the proper flowrate of blanket gas. This standard is very good in many respects, albeit conservative.

Paul:

I've looked in my API Standard 2000, 5th Edition and I can't find any reference to blanketing procedures or venting due to blanketing. Section 5.3.2.2 states "A gas-repressuring line with a suitable control and source of gas may be provided to avoid drawing air into the tank. If a gas-repressuring system is used, it shall be used in addition to the vacuum relief devices, and no capacity credit shall ve allowed."

Then, in Section 5.3.2.3 it states that the "requirement for venting capacity for maximum liquid movement out of a tank should be equivalent to 5.6 Scfh of air for each 42 US gallon barrel per hour of maximum emptying rate for liquids of any flash point." This is nothing more than stating that a standard cubic foot of blanket gas should be used to replace a cubic boot of out-going liquid.

I think the 5th Edition is the latest. Is there a recent edition that has finally incorporated the subject of tank blanketing and its safety aspects? Perhaps I should be upgrading my API Standard.

Art, that is the latest, and I believe current one. You are correct in that it does not address blanketing as such, but does address the addition of air in response to pump off, and contraction.

What I was refering to is that the equations give the flow rate of air and we use that flow rate, corrected to the replacement gas sg for the blanketing flowrate.

Paul