ALOHA™ and ARCHIE:

A Comparison

Mary Evans

Modeling and Simulation Studies Branch

Hazardous Materials Response and Assessment
Division
Office of Ocean Resources Conservation and
Assessment
National Oceanic and Atmospheric Administration
Seattle, Washington 98115

Report No. HAZMAT 93-2

April 1993
Note

Areal Locations of Hazardous Atmospheres (ALOHA™) and

Automated Resource for Chemical Hazard Evaluation (ARCHIE) are
dynamic models that will continue to be refined over time. The
discussion presented in this document refers to ALOHA version 5.1
and ARCHIE version 1.0.
Contents

Chapter 1 ALOHA™ and ARCHIE: A

Comparison 1

System Requirements
ALOHA ............................................. 3
ARCHIE............................................. 3

Chapter 2 ALOHA and ARCHIE: Calculation

Methods............................................. 9

Source Strength Calculations ............ 10

Inputs............................................... 10
Time Dependence .......................... 10
Non-Pressurized Liquid Tank ..... 11
Two-Phase Tank............................. 13
Pure Gas Tank ................................ 15
Gas Pipe........................................... 16

Dispersion Calculations ..................... 17

ALOHA ........................................... 17
ARCHIE........................................... 19

References ........................................................................... 21
Chapter 1

ALOHA™ and ARCHIE: A Comparison

Introduction
Areal Locations of Hazardous Atmospheres (ALOHA™) and Automated
Resource for Chemical Hazard Incident Evaluation (ARCHIE) are
publicly available air dispersion models primarily intended for use by
emergency response personnel. ALOHA was jointly produced by the
National Oceanic and Atmospheric Administration’s Hazardous
Materials Response and Assessment Division and the U. S.
Environmental Protection Agency’s (EPA) Chemical Emergency
Preparedness and Prevention Office. ARCHIE was produced by Hazmat
America, Inc., and has been approved for distribution by the Federal
Emergency Management Agency (FEMA), the U. S. Department of
Transportation (DOT), and EPA.

ALOHA currently runs either on Macintosh microcomputers or in

Microsoft Windows™ on IBM-compatible microcomputers. ALOHA is
available for $215 to government and non-profit agencies and $610 to
commercial institutions. Registered users may obtain program upgrades
as well as free user support via telephone, quarterly newsletter, or
electronic bulletin board. ARCHIE runs in the MS-DOS operating system
on IBM-compatible microcomputers. It is available for free from FEMA,
DOT, or EPA. Free user support is available via electronic bulletin board.

Both models allow users to enter information about a chemical and the
circumstances of an accidental spill, and will estimate the extent of
downwind dispersion of the chemical. Principal differences between the
models include the following points:

• ARCHIE models fire and explosion hazards (but not shrapnel

hazards) as well as air dispersion; ALOHA models only air dispersion
of volatile chemicals.

1
• ALOHA includes a heavy gas dispersion model; ARCHIE does not.

• ALOHA plots “footprints” of dispersing chemical clouds on a grid;

footprints can also be plotted on a background map in the
MARPLOT™ (either on a Macintosh or in DOS) or BitPlot (in
Windows) mapping applications. ALOHA provides other output in
both text and graphic form. ARCHIE provides text tables as output.

• Along with tables of ground-level and source height centerline

concentrations downwind of a spill source, ARCHIE outputs tables of
arrival and departure times for dispersing chemicals. ALOHA
footprints reflect ground-level concentrations only. ALOHA users
may view graphs of predicted concentration over time for specific
locations.

• ALOHA allows users to choose among a variety of units when

entering information into the model; ARCHIE allows a choice of units
for only a few required input values.

• ALOHA includes a chemical properties library with more than 700

entries. Complete information needed to run the model is included for
about half of these entries; partial information is available for the
remaining entries. Users may add new chemicals and property
information to ALOHA’s chemical library. ARCHIE does not include
a chemical library. However, users may store property information in
ARCHIE in advance of a spill by preparing and archiving accident
scenario files which include necessary chemical property information.

• ALOHA includes a library of U.S. cities, including elevation, latitude

and longitude, and time zone information required by the model.
Users may make additions or modifications to the city library.
ARCHIE does not include a library of locations (and does not account
for location when making computations).

2
• ALOHA accounts for indoor air infiltration when computing
concentration and dose estimates for locations chosen by the user;
ARCHIE does not.

Features included in the two models are shown in Table 1, below. Source
and dispersion algorithms included in the two models are compared in
Table 2. Inputs required to evaluate the types of accidental releases
modeled by both ALOHA and ARCHIE are shown in Table 3.

System Requirements
ALOHA
ALOHA runs on any Apple Macintosh computer with at least one
megabyte of random access memory (RAM) and a hard drive. At least 2
megabytes of hard disk space must be available to load ALOHA for the
Macintosh. It also runs in either Standard or Enhanced mode in Microsoft
Windows, version 3.0 or above. In Windows, ALOHA requires at least 1
megabyte of RAM and about 2.5 megabytes of hard disk space. A 80386
or above microprocessor is recommended for ALOHA. A math
coprocessor chip is recommended but not required for either version of
ALOHA.

ARCHIE
ARCHIE runs on IBM personal computers and compatible computers that
operate under either the PC-DOS or MS-DOS operating systems, versions
2.0 or later. The model requires a hard drive or two disk drives. ARCHIE
requires about 500 kilobytes of disk space, 512 kilobytes of RAM, and
either a monochrome or color 80-column monitor linked to any of several
display adapters.

3
Table 1.
Features included in ALOHA and ARCHIE

Model includes/accounts for: ALOHA ARCHIE

Data Handling
Chemical library √
City library √
Data collected by portable weather √
station
Accident scenario archiving† √ √
Choice of units for all inputs √
Source Strength
User entry of release rate √ √
Pressurized liquid tank √ √
Non-pressurized liquid tank √ √
Pressurized gas tank √ √
Pressurized liquid pipeline √
Pressurized gas pipeline √ √
Stand-alone non-boiling puddle √ √
Stand-alone boiling puddle √
Vapor Dispersion
Choice of roughness lengths √
Above-ground source heights √ √
Effects of inversion √
Neutral gas driver √ √
Heavy gas driver √
Indoor air infiltration √
Concentration at specific location √
Dose at specific location†† √
Fire and Explosion Hazards
Liquid pool fires √
Flame jets √
Fireball thermal radiation √
Vapor cloud fires √
Vapor cloud explosions √
Tank overpressurization explosions √
Condensed-phase explosions √
Type of Output
Text summary √ √
Text tables √ √
Graphs √
Footprint plotted on grid or map √
† ALOHA allows users to save all scenario output as archived files.
Such files can be viewed, but cannot be used as model input. The model
also allows users to save some scenario information, such as location
and chemical information and tank, diked area, and pipe dimensions for
use in additional modeling. ARCHIE users may store all model inputs
and outputs in archive files for future modeling use.

†† ALOHA allows users to choose a dose exponent.

4
Table 2.
Source and dispersion algorithms
in ALOHA and ARCHIE

ALOHA ARCHIE
Unpressurized liquid release from tank
Leak can be from any height on tank Leak is at tank bottom
Tank leaks until liquid level falls below Tank leaks until empty
hole bottom
Uses constant hole discharge coefficient† User can enter value for discharge
coefficient
Pressurized liquid release from tank
Accounts for evaporative cooling in tank Does not account for cooling
Accounts for air ingestion during release Does not account for ingestion
Liquid remains in tank if hole not at All chemical released
bottom
Release rate can change over time Entire release at peak rate
Pressurized gas release from tank
Accounts for tank pressure effect on rate Entire release at peak rate
Non-boiling puddle evaporation
Solar radiation affects evaporation rate Does not account for solar radiation
Ground and air temperature affects rate Accounts for air temperature only
Evaporation rate can vary over time Assumes constant evaporation rate
Boiling puddle evaporation
Accounts for evaporative cooling Does not account for cooling
Heat transfer from ground and air affects Uses correlation of boiling and burning
rate rates
Evaporation rate can vary over time Assumes constant evaporation rate
Gas pipeline release
Models release from pipe connected to Does not account for connection to large
large source source
Models release from pipe of finite length Models release from pipe of finite length
Release rate can change over time Entire release at 75% of initial peak rate
Liquid pipeline release
Does not model liquid pipeline Models liquid pipeline releases
releases††
Gas dispersion
Accounts for ground roughness Does not account for ground roughness
Includes heavy gas model All gases assumed neutrally buoyant
Accounts for indoor air infiltration Does not account for air infiltration
Reports dose for user-specified locations Does not report dose
Fire and explosion hazards
Does not model fire and explosion Models fire and explosion hazards
hazards
† The discharge coefficient is a measure of the characteristics of the edges of the rupture
that affect release rate. Generally, the more jagged the edges, the more impeded the
flow.

†† Releases from liquid pipelines up to 1000 m in length can be modeled via ALOHA’s
tank source option.

5
Table 3.
Information required by ALOHA and ARCHIE to compute source
strength estimates for tank and gas pipeline releases, and to
calculate downwind dispersion of an escaping chemical†

REQUIRED INFORMATION ALOHA ARCHIE

Meteorology and site data
Time of day √
Date √ √
Elevation √
Latitude and longitude √
Building type or infiltration rate √
Stability class √ √
Air temperature √ √
Wind speed √ √
Roughness length or class √
Cloud cover √
Relative humidity √
Liquid tank release
Chemical name ††
Molecular weight †† √
Normal boiling point †† √
Critical pressure ††
Critical temperature ††
Normal freezing point ††
Heat capacity (gas, const. press.) ††
Heat capacity (liquid, const. press.) ††
Liquid specific gravity √
Vapor pressure √
Mass, volume, or height of liquid √ √
Tank dimensions √
Tank temperature √ √
Hole type and dimensions √ √
Discharge coefficient of hole √
Hole location on tank wall √
Ground type (unpressurized liquid) √
Ground temperature (unpress. liq.) √
Diked area (unpress. liq.) √ √
† Chemical property information is required by ALOHA only if the
information is not available in the ALOHA chemical library. Latitude,
longitude, and elevation information is maintained in the ALOHA city
library.

††These values, if not already present in the chemical library, may be

entered by the user.

6
Table 3, cont.

REQUIRED INFORMATION ALOHA ARCHIE

Gas tank release
Chemical name ††
Molecular weight †† √
Normal boiling point †† √
Critical pressure ††
Critical temperature ††
Normal freezing point ††
Heat capacity (gas, const. press.) ††
Heat capacity (liquid, const. press.) ††
Mass of gas or tank pressure √ √
Tank dimensions √
Tank temperature √ √
Hole type and dimensions √ √
Discharge coefficient of hole √
Ratio of specific heats for the gas √
Gas pipeline release
Chemical name ††
Molecular weight †† √
Normal boiling point †† √
Critical pressure ††
Critical temperature ††
Heat capacity (gas, const. press.) ††
Pipe temperature √ √
Pipe length and diameter √
Connected to large source or closed-off √
Hole area √ √
Discharge coefficient of hole √
Pipe pressure √ √
Rough or smooth wall surface √
Vapor pressure in pipeline √
Ratio of specific heats for the gas √
Gas dispersion
Chemical name NG/HG
Molecular weight NG/HG √
Normal boiling point HG
Gas density HG
Gas heat capacity HG
Crit. temp. (or vapor press.) HG
Crit. press. (or vapor press.) HG
Vapor pressure √
Level of concern/Toxic threshold NG/HG √
Release height NG √
NG = Neutral gas dispersion
HG = Heavy gas dispersion
††These values, if not already present in the chemical library, may be
entered by the user.

7
8
Chapter 2

ALOHA and ARCHIE: Calculation Methods

ALOHA and ARCHIE use a variety of calculation methods to make

source strength and dispersion predictions in toxic gas release cases.

The models differ in several fundamental aspects. First, unlike ALOHA,

ARCHIE predicts fire and explosion risks (but does not predict shrapnel
hazard) from liquid pool fires, flame jets, fireball thermal radiation, vapor
cloud fires, unconfined vapor cloud explosions, tank overpressurization
explosions, and condensed-phase explosions. Second, unlike ARCHIE,
ALOHA’s calculations of release rates take into account factors that affect
the rate of the release over time, such as evaporative cooling within a tank
or puddle. Third, ALOHA also incorporates two dispersion models, one
to calculate dispersion of neutrally buoyant gases, and another to model
dispersion of “heavy gases,” denser-than-air vapor clouds. ARCHIE does
not have a heavy gas model.

For more detailed information about ALOHA’s source strength and

dispersion calculations, refer to the ALOHA technical documentation
(Reynolds 1992). Detailed descriptions of ARCHIE’s source and
dispersion equations are appended to the user handbook (Federal
Emergency Management Agency et al. 1988).

9
Source Strength Calculations
Inputs
ALOHA predicts rate and duration of a release from information stored
in the ALOHA chemical library about the chemical of concern and from
user-entered values. This information can include tank, pipe, or puddle
size, amount of chemical, tank or pipe pressure, weather conditions, and
temperature.

ARCHIE predicts rate and duration of release from information entered

by the user. Although ARCHIE does not have a chemical library, users
may pre-store accident scenario files, which may include chemical
property data, tank dimensions, storage amounts, and other pertinent
information.

Time Dependence
ALOHA produces each source strength estimate as a series of up to 100
time steps. An instantaneous release rate is calculated for each time step.
Generally, each step lasts long enough for one percent of the potential
mass of the pollutant to be released to the atmosphere. These 100 steps
are then averaged down into a series of five or fewer time steps, which is
then sent to one of ALOHA’s two dispersion modules. Each of these five
steps must last at least one minute. Because in most cases, meteorological
conditions are expected to change substantially from one hour to the next,
ALOHA expects that the total release duration will never exceed one
hour. If a release continues for more than an hour, users are encouraged
to rerun the model, using adjusted meteorological information.

ARCHIE’s source strength estimates are not time-dependent. That is, the
model predicts release rates that do not change over time. ARCHIE either
calculates a peak or an average rate of release, depending on the type of
release, for a given scenario, then calculates the amount of time necessary
to release all potential mass of the pollutant at that release rate.

10
Non-Pressurized Liquid Tank Release Cases

ALOHA
ALOHA predicts that a chemical will flow as a pure liquid from a tank
whenever its vapor pressure is below atmospheric pressure (whenever
storage temperature is below the boiling point). Any pure liquid release
is expected to be driven by gravitational head and by its own vapor
pressure exerted within the tank. If the liquid level is above the tank hole,
ALOHA allows for the ingestion of a small amount of air. This ingestion
of air is insurance against a complete cessation of flow when the total
tank pressure approaches atmospheric pressure. The model expects flow
to stop once the liquid level has dropped to the level of the bottom of the
tank hole or leaking valve. The liquid will change temperature due to
evaporation within the tank and heat exchange through the tank walls.
The hole may be at any height on the tank wall. The cross-sectional area
of the flowing liquid depends on whether the liquid surface intersects the
hole.

In any non-pressurized liquid release case, ALOHA expects a puddle to

form on the ground below the tank. If the user does not enter a value for
maximum puddle diameter (as would be appropriate if the tank were
diked), ALOHA expects the puddle to spread either until evaporation
rate and spreading rate are balanced or until the puddle has reached a
minimum average depth of 0.5 cm. ALOHA calculates puddle spread
rate as a function of puddle mass, liquid density, and gravity.

In non-cryogenic puddle cases, liquid spills from a tank to form a puddle

that is cooler than its boiling point. ALOHA calculates evaporation rates
for such cases depending primarily on puddle temperature and, hence, on
heat flux between the puddle and its environment. The puddle
temperature is expected to change over time (although in most cases, it
reaches a steady-state value). ALOHA accounts for five energy flux
terms—solar radiation, longwave radiation, ground heat exchange,
sensible heat, and evaporative energy loss—when computing the
evaporation rate from a puddle.

11
If a liquid has been stored below its boiling point, but ambient ground
and air temperatures exceed its boiling point, a cryogenic, or boiling,
puddle will form when the liquid leaks from a tank. In such cases,
ALOHA expects the puddle temperature to remain at the boiling point
and the vapor pressure to remain equal to the atmospheric pressure. The
model expects evaporative flux to be balanced mainly by heat input from
the ground, although the other energy flux terms described above are also
computed. ALOHA also accounts for cooling of the ground beneath a
cryogenic puddle.

In cases of both cryogenic and non-cryogenic puddles formed by liquid

that is continuously leaking from a tank, ALOHA accounts for the effects
of changing leak rate and puddle radius on the evaporation rate.

ARCHIE
ARCHIE assumes that a leak is at the tank bottom in all cases. ARCHIE
calculates an average spill rate, then calculates the amount of time
necessary to empty the tank if flow continues at that rate. Gravitational
head drives flow from the tank, with a puddle expected to form.

When the user does not enter a value for maximum pool diameter,
ARCHIE first calculates a value for evaporation or vaporization rate per
unit area. The model then calculates an equilibrium puddle area by
assuming that spreading will continue until the total evaporation or
vaporization rate equals the tank leak rate.

ARCHIE estimates of evaporation rates of non-boiling liquids and of

liquids that boil at temperatures above 0°C are functions of wind speed,
vapor pressure at ambient temperature, molecular weight, and pool
temperature.

Estimates of evaporation rates of liquids that boil at temperatures below

0°C are made using a correlation between evaporation and burning rates 1
of liquids. This correlation was empirically derived from observations

1 The burning rate of a flammable liquid is the rate at which the depth of a
burning pool decreases over time.

12
made on a set of hazardous chemicals (Burgess et al. 1961). ARCHIE
expects vaporization rates from cold boiling puddles to depend on the
predicted burning rate of the liquid—estimated from its molecular
weight, specific gravity, and boiling point—and on its liquid density.

Stand-alone Puddle Releases

Both ALOHA and ARCHIE can model evaporation from non-boiling,
stand-alone puddles of fixed radius, using the puddle evaporation
algorithms described above, without accounting for puddle spreading.
Only ALOHA models evaporation from stand-alone boiling puddles.
(The model uses the algorithm described above, but does not allow for
puddle spreading). In stand-alone puddle cases, the user is asked to enter
a value for puddle area, and a value for puddle depth, volume, or mass,
in the case of ALOHA, or liquid weight, in the case of ARCHIE.

Two-Phase Tank Release Cases

ALOHA
ALOHA predicts that a chemical will flow from a tank as a two-phase
release of vapor and aerosol whenever vapor pressure of a liquid stored
in a tank is above atmospheric pressure (whenever the storage
temperature exceeds boiling point). Any two-phase release is expected to
be driven by the differential between tank pressure and atmospheric
pressure and by hydrostatic pressure. ALOHA expects the release rate to
slow as tank pressure drops and hydrostatic head is diminished. The
model expects a release to stop once tank pressure has dropped to
atmospheric and hydrostatic head has dropped to zero (i.e., the liquid
level reaches the bottom of the hole, which may be located at any height
on the tank wall). Predicted rate of release also is affected by size of the
tank hole, but not by its discharge coefficient. ALOHA adjusts the rate of
flow through a leaking pipe or valve downward to account for flashing
flow through the restricted space.

In all two-phase release cases, any liquid present above the tank hole is
expected to exit the tank as a mixture of gas and aerosol. The model
expects all aerosol to evaporate before hitting the ground, so that no

13
puddle is formed. For computational simplicity, if the hole is at the
bottom of the tank, ALOHA expects all initial mass to exit the tank, and
ignores the amount of vapor that would actually remain once tank
pressure reached 1 atmosphere (atm). When the hole is above the tank
bottom, the model tracks both liquid and vapor mass remaining in the
tank during and after a release. In these cases, the model expects a
portion of the initial mass, representing vapor at 1 atm pressure, to
remain in the tank at the completion of a release.

ALOHA calculates the rate of evaporation of liquid into the tank vapor
space, and takes into account heat loss from evaporative cooling. It also
takes into account heat exchange through the tank walls. The
temperature within the tank is not allowed to drop below boiling,
however.

ARCHIE
ARCHIE calculates the rate of flow of pressurized liquid from a tank as a
function of rupture area, discharge coefficient of the tank hole, liquid
density, and the differential between atmospheric and tank pressure. The
model ignores the effect of gravitational head, which is expected to be less
important than the tank pressure effect.

The model assumes in all cases that a hole or rupture is located at the
bottom of the tank, and always predicts that all of the chemical (including
vapor) will be released. If a release is through a short pipe at least 4
inches in length instead of through a simple tank hole, the model
accounts for flashing two-phase flow within the pipe. ARCHIE assumes
that whenever storage temperature exceeds boiling point by 6 K, a
mixture of gas and aerosols will be released from the tank, and no liquid
will pool on the ground. At temperatures lower than 6 K above boiling,
the model predicts that a liquid pool will form.
ARCHIE calculates an instantaneous peak rate of release, then calculates
the expected duration of release if all material is assumed to exit a tank at
the peak rate. For this reason, ARCHIE does not take evaporative cooling

14
into account, and generally overpredicts release rate and underpredicts
release duration.

Pure Gas Tank Releases

ALOHA
ALOHA expects the differential between tank pressure and atmospheric
pressure to drive the rate of release of a pure gas from a tank. From the
ratio of atmospheric to tank pressure, the ratio of hole width to tank
length, and the critical pressure ratio for sonic flow (a threshold pressure
ratio value), ALOHA first determines whether a gas flow will be
supersonic (choked) or subsonic (unchoked). If the pressure difference is
great enough, ALOHA models flow as supersonic until the pressure
drops to the point at which flow is subsonic. At this point, ALOHA then
calculates subsonic release rate until tank pressure drops to atmospheric.
ALOHA-calculated rates of gas release drop over time because the tank
pressure is expected to drop as gas exits the tank, and as adiabatic
expansion cools tank contents. However, ALOHA does not account for
the effect of heat flux across the tank wall, or for frictional differences
between tank holes and short pipes/valves. In this case, the model will
produce identical source strength predictions regardless of which rupture
type users choose.

ARCHIE
ARCHIE also differentiates between supersonic and subsonic flow by
comparing the ratio of tank to atmospheric pressure to the estimated
critical pressure ratio. ARCHIE-calculated gas release rates are not time-
dependent. The model first calculates the initial discharge rate of gas
from a tank, then calculates the time required to empty the tank if
material continues to exit at this rate. Thus, ARCHIE tends to overpredict
release rate and underpredict release duration. Expansion cooling and
pressure decline are not taken into account.

15
Gas Pipe Release Cases

ALOHA
ALOHA can model both (1) cases in which a container of very large
capacity is attached to a gas pipe (the steady-state release case) or (2)
cases in which a pipe of finite length is closed off at its unbroken end. In
ALOHA, flow of gas through a pipe is expected to be isothermal, except
for the last 200 pipe diameters. Isothermal flow is assumed to result from
a balance between frictional heating and expansion cooling. Gas moving
through the last section of pipe is expected to expand adiabatically.
ALOHA calculates a friction factor to account for the roughness of the
inside walls of the pipe, defined by the user as either “rough” or
“smooth.” The rupture area may be any value greater than zero and up
to the cross-sectional area of the pipe. Ruptures are expected to be
located at the end of the pipeline.

Both in cases of connected and finite-length pipes, initial flow rate is

modeled as choked flow, and as a function of initial pipe pressure and
temperature, rupture area, and the specific heat ratio of the gas. This flow
rate is used as the steady-state flow rate in cases of release from a pipeline
connected to an infinitely large reservoir. Flow is expected to continue
for one hour, the maximum time for any ALOHA-modeled release. In
cases of releases from finite-length pipelines, release rate is expected to
decline as pipe pressure drops.

ARCHIE
ARCHIE models only cases of gas release from pipelines of finite length
that are unconnected to a reservoir. The user may specify whether a hole
occurs in the line or whether a complete break occurs either at the end or
at some point on the line. ARCHIE incorporates a model that treats a
filled gas pipeline as a volume of compressed, non-flowing gas. It does
not account for the effects of friction along the pipe, and hence does not
differentiate among pipes of different roughnesses. The model produces
an estimate of 75% of the peak release rate in cases of full line breaks, and
the length of time needed to empty the pipeline at that rate of release.
Model authors found that this volume model underpredicts blowdown

16
time and overpredicts release rate, hence producing a conservative
estimate of downwind distance.

Dispersion Calculations
ALOHA
ALOHA incorporates two air dispersion models: a Gaussian model to
predict the downwind dispersion of neutral and positively buoyant gases,
and ALOHA-DEGADIS, a simplified version of the heavy gas model
DEGADIS (Havens and Spicer 1985). Both models compute pollutant
concentrations only at ground level, where a dispersing chemical is most
likely to contact people, and produce a “footprint”, plotted either on a
grid or background map. Each footprint represents the area within
which, at some point during the hour following the beginning of a
release, ground level pollutant concentrations will reach or exceed the
level of concern entered by the user. Both models also produce graphs of
indoor and outdoor concentration and dose expected to be experienced
by people at any specific location identified by the user.

In all cases, ALOHA takes ground roughness into account. In heavy gas
dispersion cases, it accounts for relative humidity, air temperature, and
other influences on gas cloud development. Users may either select a
dispersion module to use in evaluating a particular scenario, or allow
ALOHA to select the most appropriate module, based on the friction
Richardson’s number of the vapor cloud.

ALOHA’s Gaussian dispersion algorithm can account for a release rate

that changes over time, as well as for trapping of dispersing vapor by
low-level atmospheric inversions and by the ground. It can also account
for an above-ground release, when the user directly enters a value for
release rate of a chemical into the atmosphere. The dispersion parameters
σ y and σz , which describe the horizontal and vertical spread of each
vapor cloud, are calculated either from the user’s estimate of the
atmospheric stability class, or, when a portable weather station is being

17
used with the model, directly from measurements of variability in wind
direction.

ALOHA’s Gaussian model treats each release as a series of one to five

short-duration releases of contaminant. Each individual release must last
at least one minute; the total release duration must not exceed 60 minutes.
Each of these release segments, or “clouds”, disperses passively like a
classic Gaussian plume, but with its leading and trailing edges diffusing
upstream and downstream by turbulent mixing.

ALOHA-DEGADIS includes algorithms for prediction of downwind

movement of heavy gases modified from those used by the DEGADIS
dispersion model (Havens and Spicer 1985). A heavy gas cloud is
expected by the model to disperse downwind in a more complex manner
than a neutrally-buoyant cloud. Early in a heavy gas release, the dense
vapor cloud is expected to slump away from the release point, rapidly
becoming increasingly less dense by entraining air, and forming a
secondary blanket centered on the release point. Further from the source,
the cloud maintains its stable stratification, remaining low to the ground
and resisting the effects of air turbulence as it moves downwind. Even
further from the source, the cloud is expected to be so diluted by air that
its further travel is modeled by ALOHA as passive dispersion.

When producing a footprint plot, to speed calculations, ALOHA-

DEGADIS does not account for changing release rate. Instead, the model
treats the largest release rate estimate from among the five time steps as a
steady-state rate. Footprints from short-duration heavy gas releases
hence can be overpredicted. When estimating ground-level concentration
and dose at a location specified by the user, however, ALOHA’s heavy
gas and Gaussian dispersion models account for release rates that change
over time. Time-dependent solutions are feasible in such cases because
the heavy gas concentration and dose equations are solved for a single
point only, rather than for a full grid of points as is necessary when
producing a footprint plot.

18
ARCHIE
ARCHIE assumes that all gases are neutrally buoyant and will disperse
passively downwind. The model does not take ground roughness or
relative humidity into account when making dispersion calculations.

According to its documentation, the ARCHIE dispersion module

incorporates some features of both Gaussian plume and puff dispersion
models. The dispersion parameters σy and σz are calculated from the
user’s estimate of the atmospheric stability class. The model assumes that
for short duration spills, lateral and longitudinal dispersion coefficients
(σx and σy ) are of equal magnitude.

ARCHIE produces a text table including ground level and source height
concentrations at a series of downwind points along the centerline and
evacuation zone widths at each point, a second table displaying
contaminant arrival and departure times at each point in the series, and
an estimate of the downwind distance to the level of concern set by the
user.

19
20
REFERENCES

Burgess, D. S., A. Strasser, and J. Grumer. 1961. Diffusive burning of

liquid fuel in open trays. The Fire Research Abs. and Review 3, p. 177.

Federal Emergency Management Agency, U.S. Department of

Transportation, and U.S. Environmental Protection Agency. 1988.
Handbook of chemical hazard analysis procedures. Washington, D.C.:
Federal Emergency Management Agency Publications Office.

Havens, J. A., and T. O. Spicer. 1985. Development of an atmospheric

dispersion model for heavier-than-air gas mixtures. Report No. CG-D-22-
85. Fayetteville, Arkansas: University of Arkansas Chemical
Engineering Department. Three volumes.

Reynolds, R. M. 1992. ALOHA™ (Areal Locations of Hazardous

Atmospheres) 5.0 theoretical description. Draft NOAA Technical
Memorandum NOS ORCA-65. Seattle: Hazardous Materials
Response and Assessment Division, National Oceanic and
Atmospheric Administration.

21