4 replies to this topic

[Art Montemayor]

palaniappan:

Depending on the application, what I've often resorted to in the past is to design the fire case under the assumption that the stored (or processed) liquid is Hexane. I have found that this yields a conservative design that responds safely to a variety of fluids and situations and, yet, is well within a reasonable size and cost.

I have faith in the above reasoning as long as it is conservative and covers the worst scenario. I think you will find that it is this "spirit" and scope of design that codes and standards try to impose or foment. I've found that safety is paramount and when it comes to design, one cannot be expected to be a miracle-maker nor divine the "optimum" design down to the decimal point. Rather, this is an area where common sense and a respect for the inherent safety of the operation is of importance.

You haven't identified your fluid, so I can only speculate on the relative safety of what you are doing. My experience has dealt with crude petroleum fractions and fuel oils. Perhaps your situation is different.

Art Montemayor

[pleckner]

What Art is suggesting is used by many people and is the standard for many organizations. However, I don't like it because it may just be too conservative...or not conservative enough.

What I prefer is to do a differential vaporization of a small quantity of the mixture. I will get the heat required to vaporize just about 1% of the mixture and divide by the amonut of material vaporized to get the latent heat. From this, I use the properties of the vapor and of course the vapor rate and determine the relief valve size. But you really shouldn't stop there. You continue this procedure again and again until you vaporize the entire mixture. You might find that the relief valve size goes through a maximum somewhere between the initial 1% vaporization and 100% vaporization.

A lot of calculations? Perhaps but it is the more correct way to do the analysis and with computers, it isn't all that bad.

One note. For more common hydrocarbon mixtures, we must recognize that you probably won't find much of a difference in latent heats between a 1% flash and a 100% flash.

Mixtures above the critical point are a whole new story. You can follow API. There also was an article I believe in CEP not too long ago on this very subject. I could kick myself because I thought I had this readily available but I can't find put my hand on it right now. I'll do some digging and post the reference when I find it.

[pleckner]

Guidoo,

Thanks for posting the name of the article that I couldn't find.