Estimating REACTION Rates for Relief Device

[CHD01]

I am interested in the various methods you all use to obtain reaction rates for sizing relief devices; particularly what general methods can be applied that result in safely over-sizing a valve if physical properties are limited and testing is not immediately available (similar to the way API allows the sizing of a tank based on HEXANE in the absence of detailed physical properties).

The methods I have used include:

1. Actual test lab data
2. Using theorectical heat of reaction

3. OTHERS????????????? RULE OF THUMB FOR SAFE OVER-SIZING????

The more you learn, the less you are certain of.

 

[whoislee]

It is not the safety factor that is important here. It is where the relief device is to be used. Consideration of the design limit of the equipment where this relief device is used is more important than the safety factor.
Sizing a relief device is to prevent any dangerous event from occurring. Therefore process design of the equipment and process operating limits have to be considered.

There are 2 sides of a coin
One is to give, one is to take
Give until it hurts with a smile

 

[CHD01]

I believe you may have missed the point of the question, I am asking how others estimate the required relief rate where a reaction is one possible overpressure case but their is limited physical property data. I have given what I have done (lab test work to determine mass flow; and using heat of reaction when some chemistry for the process is known). Or there any shortcut methods that have proven effective and result is safely over-estimating the actual mass flow? The basics of protecting vessels from overpressure and selecting a relief valve to relieve the overpressure isn't the issue here. The concern is recognizing a reaction is possible and then estimating what the mass flow will be. How do you determine mass flow?

The more you learn, the less you are certain of.

 

[StoneCold]

CHD01
I see what you want to do. You want to size the relief so that your vessels are covered for a general runaway reaction. I don't know of anyone who has figured that one out. It would be nice. The DIERS group (Design institute for emergency relief systems) has a book called Emergency Relief System Design Using Diers Technology. They have done a lot of work on relief systems for runaway reactions. They reccomend lab data and have a piece of equipment that they put your sample in to get the parameters needed for thier sizing calcs. (They have software too) You may be able to read their book and incorporate it into what you are doing. I think we got thier book from AICHE online.
Good Luck hope this helps some.

 

[CHD01]

Bradstone: Yes this is what I'm looking for; the test work that I did in the past was done using DIERS tools. Like I stated in my first memo, API allows us to use HEXANE as a worst case for determining relief rates due to fire exposure if the contents cannot be completely defined; it would be nice if we could agree on a worst case basis for accounting for a reaction as well in either an API Tank or a pressure vessel. I was hoping perhaps someone somewhere investigated this possibility.

The more you learn, the less you are certain of.

 

[jay165]

CHD01:

Where in API does it allow you to use hexane? What about chemicals that contain hydrocarbon fractions? Can or should you use the lightest components or the heaviest or an average?

Thanks!

 

[CHD01]

Jay165: Refer to API Standard 2000, "Venting Atmospheric and Low-Pressure Storage Tanks.

The more you learn, the less you are certain of.

 

[CHD01]

Jay165: Sorry, I neglected to address your other question.

Assuming we are talking about API tanks exposed to fire without reaction as a possibility in addition to fire exposure: If you have the physical data (latent heats, etc.) on the components of a liquid mixture you should determine the combined physical properties if they do not separate prior to reaching the relieving pressure . If you have high boilers that will boil off before you reach the set pressure, then you need to treat this separately and determine the required release rate for each fraction coming off as a vapor. I also try to estimate the length of time that a release will occurr for a specific fraction. This is important to know for reacting to the dispersion of vapor that may occur if not separately treated and also for evaluating the release versus the expected response time for corrective action by humans. I hope I interpreted your question correctly.

The more you learn, the less you are certain of.

 

[unclesyd]

You might try the Mathcad or Mathematica forums.

I have seen a lot of work on sizing relief devices for a "dump" system in a HNO3/Organic Process. Most of the work I am familiar with was done with Mathcad, but I also know this person used Mathematica also. Don't know if has published it or not.

http://www.efunda.com

for Mathematica

http://www.mathsoft.com

for Mathcad