6 replies to this topic

[ashetty]

Hello everybody,

I’m in the process of verifying the line sizes of the fuel gas header/sub header and all the branch lines. These were sized previously by other engineers & I’m doing a final cross check.

The line of concern is one of the branch lines from the fuel gas header. It's a 2" line routed to the 44” Hydrocarbon flare header near the inlet of the flare K.O.D as shown in the attached sketch.
The purpose of this line is to protect against vacuum forming in the flare header.

If the relieving gases are at a high temperature , there is a possibility that this hot gas may be cooled by ambient conditions which subsequently may result in condensation and create partial vacuum (Due to internal contraction) in the flare header. On sensing high temperature the ON/OFF valve will open.

I need the flowrate required to prevent the formation of vacuum, so I can verify the line size (shown as 2 inches). I checked the datasheet for the ON/OFF valve, but since it is a linesize valve the flowrate has not been specified.

Suggestions would be helpful.

Thank You.

Attached Files

•  Sketch.xls   36.5KB   140 downloads

[fallah]

Sizing based on worst case:

-Calculation of maximum percent of volumetric flowrate of hot gas that may be condensed,considering simultaneously maximum temperature of hot gas and minimum ambient temperature.
-Sizing fuel gas line based on calculated maximum volumetric flowrate.

One question to be raised:
Why didn't you use VRV (vacuum relief valve) instead of On/Off valve in fuel gas line?

[ashetty]

Hello Fallah,

Thanks for the response.Yes, i think your basis is the right way to go.I`ll try to work on that.
I was not involved in the design of the system, but we wouldnt want air ingress into the flare header. Hydrocarbon or nitrogen pressure is a better choice than a PVRV, I would think.

Thanks.

Shetty

[Art Montemayor]

Shetty:

Thank you for furnishing the schematic and the basis of the scope of work. This helps tremendously in getting to the core of the application and being able to analyze it - as well as the proposed solution.

However, I see several possible flaws and "over-engineering" in this application. Fallah has come close to getting to the center of the concern and the correct (in my opinion) answer. Allow me to point out what I detect from your schematic and explanation:

1) with the fuel gas at 35 kg/cm2, I have to presume that the flare drum is at approximately 5 psig - at most. Correct me if I'm wrong, because you failed to note that value. If that is the case, then you have sonic (choked) flow through the On/Off valve. This is easy solution, because you can quickly size the correct line based on constant mass flow rate (choked).

2) I don't agree with the instrument logic. I believe it is potentially hazardous because, as you noted, there is a possibility of drawing a vacuum and air could be sucked into the flare drum. (I doubt you could ever draw a partial vacuum in the flare drum - but that's another point I'll note below). The reason I identify this as a potential hazard (assuming you could draw a vacuum) is because you are trying to monitor the temperature and use that signal as the means of activating the On/Off valve. This system does not work. You will find that the time lag for measuring the temperature and generating the signal to allow fuel gas to enter will be such that you could create a vacuum instantaneously and the temperature sensor would fail in detecting the correct temperature in time. That is why the method used to break a vacuum - as fallah has inferred - is to USE PRESSURE - as the key signal to activate a vacuum breaker valve. This is done on blanketed storage tanks all the time. Visit our Tank Blanketing Forum (Paul Ostand) and you will gather the necessary instrumentation logic.

3) There is a strong possibility that it may be impossible to draw a vacuum in the flare drum (& header) if you have the presence of a mixture of gases that contain components that have sufficient vapor pressure at ambient temperature to generate a positive pressure. In other words, I'm assuming that you always have a mixture of "cats and dogs" going to flare. Some of these components may not condense at all at the ambient (or colder temperatures. I think you should analyze the worse mixture possible and see it that is a possibility. I believe your explanation assumes that you have a pure gas that is condensable under the lower temperature conditions - and Yes, under this basis, the gas could totally condense and cause a partial vacuum. My point is that in flare systems it is rare that you are flaring a pure gas whose properties you can easily identify.

4) Another instrument flaw is that you show that the 35 kg/cm2 fuel gas will instantaneously be connected with the lower pressure flare system. How do you know that the flare system can take the 35 kg/cm2??? I don't think it can. Then, how do you shut off the 35 kg/cm2 fuel gas once you have enough fuel gas in the flare system? Without an emergency shutoff valve on the fuel gas (based on the flare system pressure) I would not install such an On/Off valve device. Again, that is why we don't use that method in blanketing storage tanks - which, by the way, are much more weaker against pressure than your flare system. That is why I think you should visit the Tank Blanketing Forum.

I hope this quick analysis is of some help.

[fallah]

Hello Fallah,

Thanks for the response.Yes, i think your basis is the right way to go.I`ll try to work on that.
I was not involved in the design of the system, but we wouldnt want air ingress into the flare header. Hydrocarbon or nitrogen pressure is a better choice than a PVRV, I would think.

Thanks.

Shetty

Dear Shetty

I meant only VRV instead of on/off valve in your sckech and rest remains unchanged.Therefore in the case of vacuum VRV would suck fuel gas through connecting line.

Regards

[fallah]

Dear Art,

Thanks a lot for your complete explanations over my comment.As i read in the shetty sketch the fuel gas header pressure is 3.5 kg/cm2g (not 35).Shetty is kindly requested to confirm.

Emergency shut-off valve wouldn't solve the problem ,in my opinion,because of probability of instrument malfunction (HAZOP viewpoint) and design pressure of downstream should be same as upstream.Of course, in this case,if design pressure of flare header would be 3.5 kg/cm2g,no problem would be happenend.

Regards

[ashetty]

Thank you Art and Fallah for your detailed insights.

Yes a good diagram can greatly simplify things. I have learnt over time that effective communication is very important for engineers.

Firstly, I confirm that the fuel gas header pressure is 3.5 kg/cm2g. Design pressure of flare system is also 3.5 kg/cm2g. So no problems on that front.

I agree with you that in the event of multiple PSV relief (Cooling water failure,General power failure etc) there may be zero possibility of the mixed gases condensing at the minimum ambient temperatures. I`ll have to go through the relief load summary and alalyze individual PSV relief cases.

Also, I agree that using pressure control would make much more sense. As it stands, there are some flaws in the logic. This vave will open and remain open. I have made suggestions based on responses to this thread, and it does appear that we will go for changes in the design.

Im actually wondering if this provision is required in the first place, because we do have a purge line to the flare to maintain the +ve pressue. But I think they went in for double measures, assuming that the purge line valve was accidently closed, or that the purge gas flowrate was not sufficient. If that were the case, i think pressure control would be more reliable, considering the faster response time.

I have never visited the Tank Blanketing Forum, but I did so for the first time today. Yes, plenty of information to be found there.

Both of you have been very helpful. Art, i think your "quick analysis" covers all possible scenarios.


Regards,

A.Shetty.