Departures from this treatment o f a generic dataset are identified in the following discussions f o r each dataset.
1. Burro
The data report (Reference 14) contains summary sheets f o r
the trials, which serve as the basis for most of the data placed in the MDA. -
The following comments should be noted:
The mixture of methane, ethane, and propane that makes up the LNG is reported in the summary sheets.
calculate the molecular weight o f the mixture, but all other properties in the MüA are For pure methane.
We use t h i s imiormation to
The source diameter is calculated by assuming that the s p i l l rate is equal to the total evaporation rate.
a function of the rate of release.
Using an evaporation rate per unit area of 0.085 !qg/m 2 /s for LNG on water, the diameter varies as
The source containment diameter is set equal to the diameter of the water test basin created for the series of experiments.
The relative humidity is that termed “downwind numiditi>” in the sunmary sheets.
The temperatures and wind speed used in calculatizg the
Monin-Obukhov length are obtained from the “upwinà .tertical profile”
data.
Copyright American Petroleum Institute
--`,,-`-`,,`,,`,`,,`---
A P I P U B L * 4 5 4 b 9 2 W 0 7 3 2 2 9 0 0 5 0 5 4 9 5 437 W The domain-averaged wind speed and turbulence data use the average values listed on the summav sheets.
The Pasquill-Gifford stability class values are assigned on the basis of characterizations such as "neutral", slightly stable", etc.
The following correspondence uas assumed:
unstable, slightly unstable neutral
slightly stable
C D E
Peak concentrations (both "instantaneous" and average), and the Q
Y' for the average concentration distribution along each monitoring arc are found as described for the generic dataset- Cross-sections o i concentrations along each arc, which are plotted in the data report, provided the means for defining the windows for averaging the
concentrations.
2. Coyote
The structure and documentation for the Coyote dataset is nearly the
same as that for Burro, so that the process of preparing the MDA is virtually the same. The only departure is in specifying the Pasquill-Gifford stability class. The data report for Coyote (Reference 151 does n o t provide the
stability classification. However, a later summary of a dataset for
dispersion modeling (Reference 2 6 ) does report the stability class for each of the three trials in the MDA.
3. Desert Tortoise
The information placed in the MDA for this experiment is taken from the data report (Reference 16). The following comments should be
noted:
The exit pressure is assumed to equal the ?ressure measured prior to the.point of discharge. I t is n o t the tank Fressure that is listed.
41
Copyright American Petroleum Institute
--`,,-`-`,,`,,`,`,,`---
4.
A P I PUBL*454b 9 2 0732290 0505496 373 The wind sgeed and the temperatures used to calculace the Monin-ûbukhov length are those measured a t site Gal.
The domain-averaged wind data are those reported as :he average values at 2 rn.
The spill rate is the rate actually listed, with no adjustments for the results of the mass-ẻlux estimates made from the data obtained along the arc at 1OOm downward of the point of release.
Peak concentrations and values of Q ẻ o r the arcs at aistances of 100 m and 800 m downwind o f the point of release are obtained in the mamer described f o r the generic dataset.
Y
Goldfish
The MDA f o r the Goldfish experiment was prepared front information contained in Blewitt (Reference 171, and from information obtained directly from Mr. D. Blewitt of MCO, one of the sponsors or' the experiment. The following porints should be noted:
Some chemical properties listed f o r HF vary among several references.
Handbook (Reference 27) is 7460 calJmol, which is equivalent to 1.558~10~ J/kg. But Large's Handbook o ẻ Chemistry (Reference 2 8 ) and a basic chemistry textbook (Reference 291 !ist the latent heat of vaporization as 1.8 KcaUmol, which is equivalent to
3.76~10 J k g . Several of the models that a r e evaluated also list the physical properties of HE. PI-IASù uses 1.266~10~ JAg at 293 K, and the user's guide f o r S í contains an example in which the latent heat of vaporization for HF i s 3.732~10 JAg. These differences may be due to different assumptions by the references regarding the degree ol polymerization or' the HF. We have chosen the number used by SLAB for this progerty o17 HF, because personnel at LLNL have developed S i l and have conàucted the Goldfish tests.
The latent heat of vaporization listed in Perry's
5
5
The diameter of the discharge orifice is not Listed in any of the rer'erences f o r this experiment. The values useci in tne MDA were obtained from Mr. D. Blewitt of AMOCO.
42
Copyright American Petroleum Institute
--`,,-`-`,,`,,`,`,,`---
A P I P U B L * 4 5 4 6 9 2 m O732290 0 5 0 5 4 9 7 2 0 T m
Concentration measurements for short sampling tines (of order Is) are not available. Instead, averaging times are assumed to be tither 68.6s. or 58.3s. depending on the sampler ?osition. As a result, we only characterize the peak concentration for the
averaging time associated with all samplers in a particular arc. A
corresponding value of r is calculated from Lhe concentrations reported at monitors along the arc during the same period that contains the peak concentration. The methodology follows that for the generic case except no averaging is performed. Hence, the averaging time associated with Q is either 88.3 or 69.6s.
Y
Y
5 . Hanford Krg5
KrgS is a radioactive gas, and was released in very small quantities both as a continuous release and as an instantaneous release. The
instantaneous releases were accomplished by sealing a small volume of the gas in a quartz vial, and then dropping a weight onto the container to crush it.
The continuous releases were accomplished by adding a very sinall amount of Kr8’ to a cylinder of compressed argon gas, and releasing the mixture at a
controlled rate. In both cases, the Krg5 was quantified in terms of its disintegration rate: Cils for the continuous releases, and Ci for the
instantaneous releases. Concentrations downwind of the release were measured as radiation counts, and converted to the equivalent Ci/m 3 (actually expressed as p~i/m~).
Using a half-life of 10.4 years, we calculate that there are
kg-moles of Kr8’ associated with 1 Ci. Because the instantaneous release made use of only 10 C i , the mass and volume of the gas in the vial was very small.
The continuous release rate did not exceed 0.0388 C i / s of which amounts to approximately 7.8~10 kg-nole/s. However, because the Xr8’ was
introduced into an argon carrier-gas at an unspecified mixing ratio, we do not know the initial dilution of Kr .
-10
85
3ased on these considerations, we have modeled the Hanford Kr 85 trials with a neutral-density gas, released at a small rate.
to be “dry air”, which ensures that aense-gas eifects will ,?oc be significant in the simulations. E-ission rates (kg/s) or total aass released (kg) of the The gas is taken
4 3
Copyright American Petroleum Institute
--`,,-`-`,,`,,`,`,,`---
A P I PUBL*454b 92 0732290 0505498 L4b
"ary air" are estaclished by arbitrarily assigning 1 k g O ẻ niaSS :3 1 Ci, that the instantaneous releases are modeled as i ẻ 10 kg ( - 113 kg-aolel ox'
"dry air" were released, and the continuous releases are moaeied ôs if 0.03SS kg/s (at most) of "dry air" were released.
so
Smaller amounts of " a r y air" could have been modeled. For example.
we could have assigned 1 g of mass to 1 Ci.
To gauge the effect of choosizg 1 kg/Ci rather than 1 g/Ci in obtaining modeled normalized concentrations (that is, C/Q in Crc/rn3 o r Nui3], w e ran the
This may alter model predictions.
The results are:
S U B model both ways.