by the New Jersey Department of Environmental Protection (Reference 33).
It operates within the framework of a L O T E worksheet, and contains
algorithms for treating negatively and neutrally buoyant vapors resulting from either continuous or instantaneous releases. This was submitted to us as a new model late in the program, so there was not sufficient time to incorporate ACTOR in our model evaluation s o f t m e system.
This model was developed
ADAM (Air Force Dispersion Assessment Modell:
developed for the U.S. Air Force by Rag and Morris (Reference 34) as an
improvement to the OBíDG model, which does not account for dense gases or in-plume chemistry and thermodynamic effects.
to include the special chemical properties of üF and liquid fluorine (LF2).
With this change, ADAM can be applied to releases o ẻ nitrogen tetroxide (NsO41, phosgene (COC12), anhydrous ammonia (N"H), cnlorine (C12), sulfur dioxide
(SOZI, hydrogen sulfide (%SI, fluorine (FS), and hydrogen fluoride (HF). For these chemicals, it includes algorithms for chemical reactions, aerosols, and mixing processes for up to 16 types of releases. I t was not included in this demonstration because several chemicals required for the datasets are not among
those treated by ADAM, and no simple facility is provided to incorporate "new"
chemicals in the database.
The ADAM model was originally
The model was recently modified
5 4
Copyright American Petroleum Institute
--`,,-`-`,,`,,`,`,,`---
- - -~
A P I PUBL*:Li546 9 2 W 0 7 3 2 2 9 0 0 5 0 5 5 0 8 915 W
ALOHA (Areal Locations of Hazardous Atmospheres): The NOAA (Xatlonal Oceanic and Xtniospneric Xaministrationl ALOHA air nioael is aesigneci t~ ?roviae
atnospneric dispersion estiniates based on the Gaussian equations.
the ALOHA air sodel is equivalent to a simple Gaussian plume noce1 vith an averaging time o ẽ ten minutes, and does not consider eifects o ẻ cense gases, chemical reactions, liquid-vapor mixtures, and liquid releases. The ALOHA model is the "air" part O ẽ the comprehensive NOAA M O model (Reference 351, and is still under active development. For example, Version 5 has been under development during the past tuo years, and contains algorithms that
approximate the dense gas dispersion model DEGXDIS. It is intended for application to hazardous vapors released at ground-level.
Macintosh machines, AùỷHA is straightfomrd and easy to use.
included ALOHA in our evaluation because (1) version 5 is designed to emulate the DEGADIS model, which is included in the evaluation. ( 2 ) version 4 is functionally equivalent to the simple Gaussian plume calculation coded as GPM in the evaluation, and (3) both versions are designed for the Macintosh computer environment, and are not compatible with the systems used f o r the rest of the models. We note that a windows-based Doc version is under development, but is unavailable for this effort.
Version 4 of
As implemented on
Ue have not
ARCHIE (Automated Resources f o r Chemical Hazard Incident Evaluation): ARCHIE was developed for the Federal Emergency Management Agency ( M I , the U.S.
Department of Transportation (DOT), and the U . S . hvironmental Protection Agency ( D A ) , and is described in a handbook that can be obtained from regional FEbM offices (Reference 36). The primary ?urpoce of ARC3IE is to provide emergency preparedness personnel with several integratea estimation methods that may be used to assess the vapor dispersion, rire, -- and explosion
impacts associated with episodic discharges of hazardous materials into the terrestrial (that is, land) environment. The program is also intended to facilitate a better understanding of the nature and sequence of events that may follow an accident and their resulting conseqyenczs.
The dispersion model within ARCAIE is essentially the Gaussian plume model for point-source releases, with correction terms for relezses of
finite-duration. No heat transfer, chemical effeczs. or dense-sas effeccs are simulated. We do not include ARCXIE in the evaluation becouse t h e Gaussian plume ana 7uẽf calculations o ẽ GThi and INPUT are re?resentative o ẽ t h i s class of dispersion mocieling tecimiques.
55
Copyright American Petroleum Institute
--`,,-`-`,,`,,`,`,,`---
A P I PUBL*KLt51(b 72 W O732290 0505509 851 9
CcU)M (Calm Air Dispersion Modell:
SoftSkills, ùnc. Moser (Reẻerrncr 37) describes it as a generalizad p-ogram for modeling the spreading and dispersion o l dense-gas clouds during calm
periods. We have not included CADM in this evaluation because the number o ẻ
trials in which dense-gas clouds are released during calm or lowwind-speed meteorological conditions i s small.
CADM uas developed by and is available :rom
HGSYCTEM (Heavy Gas System): HGsycr"EM is a recently-developed revision to the HEGADAS model, designed especially f o r modeling releases of hydrogen fluoride
( € F I . The development of the model, first known as HF?RSEMM. focused on three
areas (Reference 38): (1) the modeling of the complex thermodynamics
of EIF/HsO/Air mixtures (Including aerosol eẻfects on cloud density); (21 the treatment of a wide range o f surface roughness conditions (including possible multiple surface roughness conditions); and (3) jet flow and air entrainment for pressurized releases of HF, follửwed by transition to ground-based dense gas dispersion. First, the WADAS model was modified to meet these
oojectives.
from pressurized releases, dispersing initially as an elevated HF plume.
touchdown and slumping of an initially-elevated dense HF glume were also modeled in ủFPLỹME, with a link into HEGADAC to complete the modeling o f the transition from an elevated to grounded dense gas cloud.
estimation model, a pool evaporation model, and a far-field Gaussian model (linked to HFPLUME) were added to provide a more complete source and
dispersion modeling package.
The HFFLüME model was developed and tested to simulate jet flows The
Later, a source
We have not included HGSISTEM in t h i s evaluation because the code was not received until late in the study and there was insufficient time to
thoroughly test the program and correct the few "bugs" that were found during initial test runs. Furthermore, it was found that the foniats o ẻ the outputs vary depending upon which of the major modules is used in a simulation. so that significant work still remains in being able to eẻficiently obtain results from the model wnen simulating the many trials used in the evaluation.
5 6
Copyright American Petroleum Institute
--`,,-`-`,,`,,`,`,,`---
3. DESC3IPTION OF M O D E S EVALUATED
14 models were evaluated-+ of these aro publicly-availaole ( Z O X .
3ritter & McQuaid, DEGADIS, GTf?. 'XEZXDAS, I N P E , OB/CG. ana C L G 3 ) . and 5 are proprietary (AIRTOX, C W , FOCUS, GXCTAR, PIIACT. and 'iTuCE!. Those teraed
"publicly-available" can be obtained from published texts. from their developers, or from the EPA for the cost of reproduction. ùhe source-code f o r these models is distributed along with the user's guides. Those termed
"proprietary" are sold by individual companies, which typically provide
technical support for the product and reference materials, but do not provide the source-code. Primary references f o r these 14 models are listed below: