Liquid Pyrolysis Model

[SD]

 Fire Dynamics Simulator

 Current Date     : September 11, 2014  00:22:51

 Version: FDS 6.0.1; MPI Disabled; OpenMP Disabled

 SVN Revision No. : 17534

I am trying to understand liquid pyrolysis model inorder to eventually test the effect of supression. 

From FDS Userguide, p 80, Sixth Edition, I see that including 'BOILING_TEMPERATURE' tells FDS to use the liquid pyrolysis model. But, I could not find much information on the BOILING_TEMPERATURE. Hence my questions that follow:

1. Is the boiling temperature, as the name suggests,  the same as the temperature at which the liquid fuel vaporises? If that is the case, it is the same as flash point and not the autoignition temperature of the fuel. Am I correct?

2. Although my goal is to work with solid combustion, I am using liquid combustion to simplify my model. Towards this goal, I want to use a liquid fuel that has a flash point that is close to wood (about 300 C). Can I manufacture such a fuel and use it in FDS? I was looking through some previous discussions and I got to a conclusion that it is best to use a liquid fuel which is already part of the FDS library. When I try to look for such a fuel in the user guide or technical reference guide, I cannot find a fuel that has such high flash point. Where can I find the list of fuels available in FDS? In FDS Userguide, p 116, Sixth Edition, I can only find a limited gas and liquid species. Usually, there is a lot of discussion on diesel or dodecane (C12H26) on the forums. Are these fuels available in FDS? What if I wanted to use vegetable oil as my liquid fuel?

Thanks all for your time and kind support. 

[SD]

To add to my question above, the reason I am interested in knowing how FDS  treats 'BOILING_TEMPERATURE' is because:

1. FDS6.0.1 does not let me use a BOILING_TEMPERATURE less than approx 20 deg C. Why? Assuming the default ambient temperature is 20 deg C, I would think that if the BOILING_TEMPERATURE is less than 20 C, then no fuel should evaporate, and hence no combustion should take place.

2. When the liquid fuel evaporates to vapor, does it automatically undergo combustion and release heat in FDS? I am assuming the answer is yes.

3. I ran a liquid pyrolysis case with BOILING_TEMPERATURE=300. The pool fire is burning. I am confused as to how FDS lets the fuel burn to begin with. Based on the definition, assuming the default ambient temperature in FDS is 20 C, the fuel should not vaporise because the BOILING_TEMPERATURE is very high and set to 300. In that case, if no vapor is produced, how does the fire initiate and how does it even begin to produce heat?

4. Later in the simulation when the sprinklers are turned on, the HRR goes down. I am happy with this. In FDS, am I right to say that sprinklers do not extinguish the fire directly, but rather, the sprinklers bring the local temperature down because the water absorbs the heat (latent heat of vaporization) and changes phase. If the local temperature is low, and if the set BOILING_TEMPERATURE is less than the local temperature, then the fuel stops to burn because it cannot generate any vapor. However, if the local temperature is higher than BOILING_TEMPERATURE, then the liquid fuel continues to burn, because liquid can vaporise and FDS automatically allows the vapor to burn.

5. This is how my fuel is defined:

&REAC FUEL               = 'DODECANE'

      HEAT_OF_COMBUSTION = 44400.

 SOOT_YIELD = 0.179  

 FUEL_RADCAL_ID = 'DODECANE' 

 /

There is no heat of vaporization defined here. Shouldn't the heat of vaporization be included to accurately model liquid pyrolysis? If I am able to include the heat of vaporization, then I guess my analysis of the how I interpreted combustion will take place will be different and more complicated.

Please share your thoughts. I am trying to read FDS's mind. I may have missed this information in the user guide, but please feel free to point to relevant sections or topics of the guides where I can find  this information.

[Simo]

Boiling temperature is not the same as flash point, nor autoignition temperature. The boiling temperature is, as the name suggest, boiling temperature! It is defined as the temperature, where the vapor pressure of the liquid is equal to the external pressure on the liquid. Flash point is the temperature, where the liquid vaporizes enough to creage a flammable mixture above the surface.

So, do not try "read FDS's mind". Read the guides, and also the Tech guide.

Liquid pyrolysis model is based on the use of mass-transfer number telling how much vapor leaves the surface, depending on the temperature, gas phase concentration and flow conditions. As you correctly stated, FDS let's that fuel burn immediatelly. Phase change energy (heat of vaporization) is calculated by the pyrolysis model, and specified for MATL, not for the gas phase fuel.

If you observe the sprinklers put down the fire, that's good. Regarding the mechanism, please think a moment what is the actual suppression mechanism of water-based system? What else it is than cooling down the gas and material surface, replacing fuel and oxygen by water vapor, and
blocking radiation? FDS accounts for all of those. (How accurately, is another question).

I would recommend modelling solid pyrolysis using the solid pyrolysis model. It is much more reliable and probably more understandable way to model the things you are interested in, than using the liquid model for the purpose.

[SD]

Simo,

Thanks for your response. Now, I understand that BOILING_TEMPERATURE in FDS refers to the boiling point at 1 atm (for most cases). 

The reason I am using the liquid pyrolysis model as opposed to the solid pyrolysis model is because I have a design HRR curve and I need to reproduce that curve using a certain amount of fuel. Initially, I did start with the solid pyrolysis model and the configuration and geometry of the solid fuel greatly influenced my HRR and it was very difficult to reproduce the design HRR that I am interested in. So, I switched to liquid pyrolysis model since it easier to control the curve using the parameters in liquid pyrolysis model. 

Based on the definition of BOILING_TEMPERATURE in FDS that you helped me with, how do you explain this:

I ran a liquid pyrolysis case with BOILING_TEMPERATURE=300. The pool fire is burning. I am confused as to how FDS lets the fuel burn to begin with. Based on the definition, assuming the default ambient temperature in FDS is 20 C, the fuel should not vaporise because the BOILING_TEMPERATURE is very high and set to 300. In that case, if no vapor is produced, how does the fire initiate and how does it even begin to produce heat?

Thanks much.

[Randy McDermott]

Vapor IS produced.  The boiling temperature of water is 100 C at 1 atm.  Go outside (not in D.C.) on a day that is 20 C with low humidity and sweat will evaporate off your skin.  This is because air is a multi-component system of species.  See, for example, Bird, Stewart, and Lightfoot, "Transport Phenomena".

--
You received this message because you are subscribed to the Google Groups "FDS and Smokeview Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fds-smv+u...@googlegroups.com.
To post to this group, send email to fds...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/fe6b21e2-7a4c-44b7-8abe-dab714cc195b%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[SD]

Thanks Randy, that adds to my understanding a bit. I guess, the answer that I am trying to find in the fds documentation is: How does including or not including the 'BOILING_TEMPERATURE' parameter change the physics of the problem.

From FDS Userguide, p 80, Sixth Edition, I see that when the parameter 'BOILING_TEMPERATURE' is invoked, FDS uses the liquid pyrolysis model.

In my model, I am trying to replicate the phenomenon that a liquid fuel can never burn unless the surrounding temperature is above the flash point of the liquid fuel. So, in reality although the liquid fuel may vaporize at a temperature below its boiling temperature, it cannot become a flammable vapor mixture unless the temperature reaches its flash point. 

I was hoping I could use 'BOILING TEMPERATURE' as a guide to set this criteria. But, I guess I cant. Is there an alternative?

Similarly, when fire suppression occurs due to water sprinklers, I want suppression to happen based on the BOILING_TEMPERATURE. If the water brings the local temperature to below the flash point, I want the liquid in that local cell region to not undergo combustion. I guess I cant do this either based on the BOILING_TEMPERATURE or can I?

Thanks.

[Simo]

You cannot create a model like that using the default combustion model which assumes that fuel gas and oxygen always react when they meet.
You could try to model that using the finite rate combustion option, but I do not recommend it in this case. It would be serious research project.

If your main goal is to reprodice an experimental HRR curve, and then let the sprinklers prevent it from burning, why don't you just prescribe the HRRPUA curve and use the  empirical suppression model (E_COEFFICIENT)?

[Brad Casterline]

I just want to add that if you do use E_COEFFICIENT, use a solid fuel and &MISC POROUS_FLOOR=.FALSE. (per Dr. Floyd a liquid fuel might not act like a 'solid floor').
In the guides it is stated the E_COEFFICIENT must be determined experimentally. I have briefly looked at the referenced FMRC test but I do not have the math experience to translate the results to an FDS E_COEFFICIENT (without trying anyway). For my own curiosity however, I have modeled quite a few scenarios using a range of coefficients and flow rates, taking the long way around to a better understanding. Not very scientific but it gives me something to do.

Brad

[SD]

Simo,

It is a compartment fire with windows.

Yes, part1 of my main goal is to set up a case to reproduce an t^2 HRR curve which attains its maximum at 2 minutes. Part 2, is that given the setup in part 1, if the watermist system is turned on at Time=30 seconds, how will my HRR be different. I am comparing the two HRRs curves by placing them on top of each other to see if my suppression system is effective or not. 

I want the suppression to depict something real (in my case it is a fire caused due to burning of wood, clothes, materials in the compartment). So, I have to choose a realistic fuel. If I choose gasoline as a fuel, I will get a HRR curve after supression that is not realistic as its harder to suppress gasoline compared to say diesel or vegetable oil. So, far I found a liquid pyrolysis model to be most suitable. I am choosing dodecane with a boiling temperature of 360C. I can't think of anything better.

But I like your idea of using E_COEFFICIENT. I tried to look at the FDS user's guide, and I found that it may help me, but I haven't grasped the concept completely. I will take my time and research this method and get back to you.

Brad,

Thanks for your input. The user's guide does say that E_coefficient should be obtained experimentally. I have this issue as well. For now, I have a liquid fuel and I think I will stick to this because I don't have much time to change the fuel and research further. Hopefully, the difference is not too much.

[Brad Casterline]

Really not much difference- it's mostly a matter of time.

--
You received this message because you are subscribed to a topic in the Google Groups "FDS and Smokeview Discussions" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/fds-smv/5MeLiOeu0sg/unsubscribe.
To unsubscribe from this group and all its topics, send an email to fds-smv+u...@googlegroups.com.

To post to this group, send email to fds...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/c4b8b956-9992-4738-9ee1-0ba8bc2823e5%40googlegroups.com.

[Brad Casterline]

Experimenters showed quite a while back that the heat of combustion of a wide range of combustible building contents is 13,100 kJ/kg of O2 consumed. If someone can tell us what the E_COEFFICIENT of that is then I think that would put us both in the infield of the ballpark;)

To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/18E0326D-13CF-4BD6-84A9-1F4E21826DB7%40fsc-inc.com.

[Randy McDermott]

Brad, a friendly correction: ~13000 kJ/kg O2 is from theory.

--

You received this message because you are subscribed to the Google Groups "FDS and Smokeview Discussions" group.

To unsubscribe from this group and stop receiving emails from it, send an email to fds-smv+u...@googlegroups.com.

To post to this group, send email to fds...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/F489AF6B-E7C3-4472-8330-961BF7EFE004%40fsc-inc.com.

[Brad Casterline]

OK thanks Randy. I had read it was theorized several years before it was backed up by calorimetry experiments. 

Soon I will start an E_COEFFICIENT thread (after I have done all I can on my own) so I can stop tracking mud all over SD's thread :)

Thanks again Doc!, Brad 

To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/CAC6wax1pJ%3D-Sf50Z-XU8xT3M8e5u4yi3-iPzUKHkYWyr6ykKJQ%40mail.gmail.com.

[Kevin]

Brad  -- E_COEFFICIENT has nothing to do with 13,100 kJ/kg. The latter refers to the combustion of (typically) hydrocarbon fuels; the former is an empirical model for the reduction of the burning rate of solid objects by water. I added E_COEFFICIENT to FDS long ago after reading various experimental reports in which it was observed that the HRR of a burning object would decrease exponentially in time after the application of a steady spray of water, usually from a sprinkler. But much like the t-squared fire, the exponential decay assumption is very simple way of lumping together a lot of complicated fire phenomena that are related to the exact geometry of the burning item, say, a roomful of "stuff". There has not been much work, that I know of, on modeling fire suppression using this simple exponential decay assumption. But the alternative to using E_COEFFICIENT is to develop a detailed model of the pyrolysis, using solid phase reactions in which the addition of water at the surface will slow the reaction by cooling alone. Not only is it difficult to get the reaction parameters, it is also difficult to model the geometry, or in your case, the introduction of water into a pool of liquid fuel.

I would suggest that anyone looking for a experimental topic might consider conducting a simple experiment with a pan fire and simple water spray. Or if anyone has a reference to such, I'd like to have it.

On Saturday, September 13, 2014 10:03:36 AM UTC-4, Brad Casterline wrote:

OK thanks Randy. I had read it was theorized several years before it was backed up by calorimetry experiments.

Soon I will start an E_COEFFICIENT thread (after I have done all I can on my own) so I can stop tracking mud all over SD's thread :)

Thanks again Doc!, Brad 

On Sep 13, 2014, at 7:37 AM, Randy McDermott <randy.m...@gmail.com> wrote:

Brad, a friendly correction: ~13000 kJ/kg O2 is from theory.

On Fri, Sep 12, 2014 at 11:20 PM, Brad Casterline <bcast...@fsc-inc.com> wrote:

Experimenters showed quite a while back that the heat of combustion of a wide range of combustible building contents is 13,100 kJ/kg of O2 consumed. If someone can tell us what the E_COEFFICIENT of that is then I think that would put us both in the infield of the ballpark;)

On Sep 12, 2014, at 9:51 PM, Brad Casterline <bcast...@fsc-inc.com> wrote:

Really not much difference- it's mostly a matter of time.

On Sep 12, 2014, at 9:42 PM, SD <sai....@gmail.com> wrote:

Simo,

It is a compartment fire with windows.

Yes, part1 of my main goal is to set up a case to reproduce an t^2 HRR curve which attains its maximum at 2 minutes. Part 2, is that given the setup in part 1, if the watermist system is turned on at Time=30 seconds, how will my HRR be different. I am comparing the two HRRs curves by placing them on top of each other to see if my suppression system is effective or not. 

I want the suppression to depict something real (in my case it is a fire caused due to burning of wood, clothes, materials in the compartment). So, I have to choose a realistic fuel. If I choose gasoline as a fuel, I will get a HRR curve after supression that is not realistic as its harder to suppress gasoline compared to say diesel or vegetable oil. So, far I found a liquid pyrolysis model to be most suitable. I am choosing dodecane with a boiling temperature of 360C. I can't think of anything better.

But I like your idea of using E_COEFFICIENT. I tried to look at the FDS user's guide, and I found that it may help me, but I haven't grasped the concept completely. I will take my time and research this method and get back to you.

Brad,

Thanks for your input. The user's guide does say that E_coefficient should be obtained experimentally. I have this issue as well. For now, I have a liquid fuel and I think I will stick to this because I don't have much time to change the fuel and research further. Hopefully, the difference is not too much.

--
You received this message because you are subscribed to a topic in the Google Groups "FDS and Smokeview Discussions" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/fds-smv/5MeLiOeu0sg/unsubscribe.

To unsubscribe from this group and all its topics, send an email to fds-smv+unsubscribe@googlegroups.com.

To post to this group, send email to fds...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/c4b8b956-9992-4738-9ee1-0ba8bc2823e5%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "FDS and Smokeview Discussions" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/fds-smv/5MeLiOeu0sg/unsubscribe.

To unsubscribe from this group and all its topics, send an email to fds-smv+unsubscribe@googlegroups.com.

To post to this group, send email to fds...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/18E0326D-13CF-4BD6-84A9-1F4E21826DB7%40fsc-inc.com.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "FDS and Smokeview Discussions" group.

To unsubscribe from this group and stop receiving emails from it, send an email to fds-smv+unsubscribe@googlegroups.com.

To post to this group, send email to fds...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/F489AF6B-E7C3-4472-8330-961BF7EFE004%40fsc-inc.com.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "FDS and Smokeview Discussions" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/fds-smv/5MeLiOeu0sg/unsubscribe.

To unsubscribe from this group and all its topics, send an email to fds-smv+unsubscribe@googlegroups.com.

[Brad Casterline]

Thank you very much Kevin.

Yes, I was aware that the E-COEFFICIENT is the reduction of the pyrolysis rate, and the fact that it is exponential explains why the idealized t^2 curve does not go smoothly up and then level off when the head activates and stay level forever! That is what I was after; what HRR will be in equilibrium with what water application rate, so that the fire is neither growing nor dying? And I really am interested only in the 'common' solid fuels that comprise building contents. That is why the 13,100 kJ/kg O2 appealed to me as the 'common' HOC. That it has nothing to do with E_COEFFICIENT is the nagging reason I have not started a thread I suppose, now that you state it so clearly!

In short, I am trying to quantify Low, Moderate, and High,- Quantity, Combustibility, and HRR for "Control Mode" of sprinklers based solely on the required water application rates for what NFPA #13 calls Light and Ordinary Hazard. It is mostly a personal endeavor and due to all the variables I would be satisfied with a reasonable range for the quantifications :).

Thank you very much again Dr. McGrattan!

Brad

To unsubscribe from this group and all its topics, send an email to fds-smv+u...@googlegroups.com.

To post to this group, send email to fds...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/fds-smv/501b8026-c322-40b6-adbf-84aedea31d66%40googlegroups.com.

[SD]

Brad,

        You were not tainting my thread, you were adding to the discussion and helping me.

Kevin,

         Thanks for enlightening me about E_COEFFICIENT. Kudos for implementing this in FDS.

Can I ask what E_COEFFICIENT is dependent on? Is it just a property of the fuel or does it depend on other parameters like geometry, size of the fire and boundary conditions? In other words, if E_COEFFICIENT is obtained experimentally for a given test case, then can it be assumed constant for other similar FDS cases when the geometry of domain, boundary conditions and fire suppression flow rate is changed but the fuel is the same?

Simo,

         Based on my limited knowledge I have gathered now, since I have no access an to experimental set up and because of I clearly do not understand the unknowns yet, and given the limited time, I cannot obtain the E_COEFFICIENT value to handle fire supression. Also, I specificially want to use water-mist supression as opposed to sprinklers. And, I have the flowrate, diameter and velocity of the the water droplets, so I might as well use the conventional approach.

Right now I am able to match the design curve to some extent, however the curve looks like a log-curve as opposed to a t^2 curve (if you know what I mean). I only way I can think of obtaining a t^2 curve is to some how control the fuel input to the domain.

Question-1: I would love to use liquid fuel spray with a ramp to obtain the t^2 curve, but I cannot invoke the liquid pyrolysis model. The liquid spray ignores the material line where the boiling temperature is used. So, I cannot use a liquid fuel spray model. Am I right?

Question-2: I currently have a liquid pool fire with a liquid pyrolysis model. Is there anyway I can use a step input (or a ramp function) for my liquid pool fuel? That way I can control to some extent the HRR obtained. I was thinking of gradually uncovering a lid from the top of a liquid pool, but I am not sure how to do this or if there is a better way to model this phenomenon.

I can thank enough everyone for sparing their valuable time to chip in their suggestions. Thanks a lot!

[Kevin]

E_COEFFICIENT depends on fuel type and geometry. It is just a parameter in a simplified model of suppression. There are no published values for it. It is just something we added to FDS as a crude way to model suppression.

[SD]

Thanks for the insight and clarification Kevin.