See below ...

Mark Goodson Wrote:
-------------------------------------------------------
> In trying to determine Blood ethanol from breath
> ethanol levels, by legislative fiat the government
> has declared that everyone has a partition
> coefficient of 2100. Yet, we all know that the
> number varies between about 1500 and 3000. '

The constant used in the CFK equation is on the order of 200 with some variation (e.g. 217 as opposed to 200). My previous response covers this issue. As an engineer, you understand that while any variable by definition will be different for different scenarios, it is the sensitivity in relationship to the other variables that matter. You also have to understand how accurate you have to be to answer a specific question. Even a model with an uncertainty of 50% has value, but it may have limitations under certain applications.

>
> This variability is the problem with Coburn
> Forster Kane. You have to assume a Haldane
> constant, You have to assume a blood mass based on
> body mass. You assume a certain partial pressure
> of CO - never mind that the CO is changing, and
> that as ventilation changes, so will atmospheric
> CO concentration. How much body weight was lost in
> the fire, and was it muscle or adipose tissue or
> dermis? Was the deecent impaired by conditions
> that would alter CO uptake?

The CFK equation is coupled with CO concentrations where the sensitivity of the variables of the CFK equation tend to "wash out" with higher CO concentrations. Our analysis technique does include the blood mass as a senaitive variable. This is not assumed, but calculated. We also have constrcted the model to take into account transient concentrations of CO with time. The CO concentrations are modeled with respect to the ventilation conditions, which is the primary mechanism for CO production in compartment fires as shown by Beyler (Havard), Gottuk (Virginia Tech), Lattimer (Virginia Tech), and Gann (NIST).

With respect to your last concern, the time constant associated with CO uptake and death is very short when compared to the time constant associated with degrading of the body. A person would expire under the initial thermal conditions associated with the comsumption of the body. The CO concentration in the blood has been shown to be very stable over long periods of time, so this assumption is reasonable in this context.

>
> I tried this stuff in the 1980's. It was a
> difficult task then. Even with matlab or similar
> programs, I see huge problems, given the
> assumptions that have to be made. My examp0les
> were persons who died from fugitive CO with no
> thermal destruction of the body.

Yes, you need to solve the equations numerically as opposed to analytically, but this is not new nor difficult, especially for an engineer. It would seem that your difficulties and the sensitivity of the variables was the result of a lack of fundamental knowledge of CO production in fires. This fundamental knowledge was gained in the 1990 as a result of a number of fires that resulted in very high CO concentrations in the blood that could not be explained based on the current level of knowledge of CO production at that time. Thus, your efforts in the 1980's would not have included this more recent knowledge.

Douglas J. Carpenter, MScFPE, CFEI, PE, FSFPE
Vice President & Principal Engineer
Combustion Science & Engineering, Inc.
8940 Old Annapolis Road, Suite L
Columbia, MD 21045
(410) 884-3266
(410) 884-3267 (fax)
www.csefire.com