Safeti 8.1 Release Notes

RELEASE NOTES
SAFETI TM
Taking hazard and risk analysis one step further
Version: 8.1
Date: April 2018
Reference to part of this report which may lead to misinterpretation is not permissible.
Date: April 2018
Prepared by: DNV GL - Digital Solutions
© DNV GL AS. All rights reserved
This publication or parts thereof may not be reproduced or transmitted in any form or by any means,
including copying or recording, without the prior written consent of DNV GL AS
Table of contents
1 NEW FEATURES APPLICABLE TO ALL SAFETI USERS............................................................ 2
1.1 The tree can show which items are selected for the active Run Row .............................................. 2
1.2 Import of population and plant boundary data from external files ................................................. 3
1.3 Option to include Materials and Parameters data when exporting a workspace ............................ 4
1.4 Option to set the colour to be used for each risk level in the risk contour plot ............................... 4
2 NEW FEATURES APPLICABLE TO USERS OF THE FULL SAFETI PROGRAM ............................... 5

2.1 Modelling of isolation and blowdown for time-varying releases ..................................................... 5
2.2 Consequence Data Results View with summary effect zone distances ............................................ 6
2.3 Parallel run method for running calculations for multiple Scenarios at the same time ................... 7
3 OTHER DIFFERENCES AND BUG FIXES .............................................................................. 9

3.1 Upgrades to licensing software that will allow network and dongle licenses to be used with v8.1 9
3.2 Simplifications in specification of release phase for storage Scenarios ......................................... 10
3.3 Changed approach to generating Fire Scenarios for a Warehouse ................................................ 11
3.4 Area of maximum footprint given in the Cloud Maximum Footprint Graph and GIS Results ........ 11
3.5 More consistent method of defining vessel contents for a Standalone BLEVE blast Scenario ...... 12
3.6 Improved handling of small hole sizes for Long pipeline Scenarios ............................................... 13
3.7 Choice between use of radiation dose or intensity in flammable risk calculations ....................... 14
3.8 Modelling of jet fires for time-varying releases uses settings for the time-averaging method ..... 15
3.9 Clarifications for meaning and application of input data ............................................................... 16
3.10 Bug fixes .......................................................................................................................................... 17
4 PERFORMING A LARGE ANALYSIS ................................................................................... 23

4.1 Why is special attention needed for a large analysis? .................................................................... 23
4.2 At what point should an analysis be considered “large”? .............................................................. 24
4.3 What can I do to make a large analysis easier to perform?............................................................ 24
5 ALERTS AND WORKAROUNDS ........................................................................................ 33

5.1 Database brought in line with changes in password policies ......................................................... 33
5.2 Known Bugs ..................................................................................................................................... 33
| RELEASE NOTE | Safeti version 8.1 Page i

1 NEW FEATURES APPLICABLE TO ALL SAFETI USERS
The following features are applicable to users of Safeti and Safeti Lite, and to users with and without the
extensions for multi-component modelling and 3D explosion modelling.
1.1 The tree can show which items are

selected for the active Run Row
The toolbar in the Run Row Grid Supertab contains a Show Selection
States toggle button that allows you to choose whether or not
the Study Tree should grey out nodes that are not selected for the
active Run Row.
By default the option is turned on, and in this situation the

Scenarios, Route Segments, Weathers, Parameters, Populations,
Obstructions, Boundary and Offset that are selected for the active
Run Row will be shown as selected (i.e. not greyed out), as shown
for the Propane Scenarios in the illustration.
Other types of item will also be shown as selected if they are

referenced by Scenarios or Route Segments that are selected for the
active Run Row, as follows:
Tab in Tree Type of node Nodes shown as selected
Models Model Group folder under a Model Groups selected for the Route Segments
Route Model and all Equipment that are selected in the Model Playlist for the
and Scenario nodes underneath active Run Row
it
Map Types of pool substrate and bund Types selected for Scenarios that are selected in
the Model Playlist for the active Run Row
Types of terrain for dispersion Types selected for Scenarios that are selected in
the Model Playlist for the active Run Row
Building Building selected for Storage Equipment item or

Warehouse that has Scenarios in the Model
Playlist for the active Run Row, and Building
selected for the Population Playlist for the active
Run Row
In addition, Risk Ranking Points are shown as selected if they are selected in the current Risk Ranking
Point Playlist, and a parent node such as a Pressure Vessel or a Study will be shown as selected if it
contains at least one child node that is shown as selected according to the rules given above.
All Material nodes are always shown as selected, whether or not the Material, Mixture or Warehouse
Material is selected for the items in the Models Playlist for the active Run Row, and Population
categories, Vulnerabilities, Risk transects and Risk effect levels are also always shown as selected.
If you turn the Show Selection States button off in the Run Row Grid, no nodes will be shown as greyed
out on the tree and you will not be able to tell from the tree whether or not a particular node is selected
for the active Run Row.
| RELEASE NOTE | Safeti version 8.1 Page 2

1.2 Import of population and plant boundary data from external
files
The Data tab of the Ribbon Bar now includes a Data
source option as shown. This option is enabled if the
node currently selected in the Study Tree is the
Populations set, a folder under the Population Set, or
the Boundaries set. This option allows you to import
data for plant boundary shapes or populations shapes from an external GIS source.
The import process involves a series of dialog screens:
1. Selecting the data source.

You must first select the file or files that contain the data for the shapes that you want to import.
2. Selecting the data tables to import

A GIS database may include data for a number of different features (roads, residential
populations, schools, hotels, etc.), and the data for different features are organised in different
"tables" within the database. You must select at least one of the tables.
3. Applying an optional spatial filter

You can specify that you only want to import the shapes that lie within a given circular region. If
you do not apply this type of spatial filter, the program will import all of the shapes in the
selected tables that have valid, supported geometry.
4. For Population only: Setting values for input fields

You can set values for Density, Population, Category name, or Fraction of population
indoors for societal risk, either by specifying a field in the data source that contains the values
for the different shapes, or by giving a constant value that will be used for all shapes.
When you click on Finish to proceed with the import, the program will display the progress through the
stages in the import, in which it creates nodes in the tree for the different shapes, and set the values for
the input fields.
If you selected more than one table for import, the program will create a separate folder for the shapes
from each table. If a given table contains more than 201 shapes, the program will divide the shape for
that table into subfolders, with up to 201 shapes in each subfolder.
If a shape in the data source is not valid for a population shape or a plant boundary, an error message
will be written to the Output View giving the name of the shape and stating that it is not a supported
type. In the program, population shapes can either be points, rectangles or polygons, and plant
boundaries must be polygons. All other types of shape are regarded as unsupported geometries and will
be omitted in the import process.
For further details of the import option, enter “Data source input” in the Index tab of the online Help.

1.3 Option to include Materials and Parameters data when
exporting a workspace
An exported workspace file is a special format of *.psux file that can have external files that are
associated with the workspace embedded inside it. You can use the Export Workspace option in the File
menu to create a file in this format, acting as a "snapshot" of the state of the analysis at the time of the
export. Exporting a workspace in this way allows you to transfer the exported workspace file with the full
set of results and external supporting data to another machine, or to archive it.
When you select the Export Workspace option from the File menu, a dialog now
appears as shown allowing you to choose which type of data to include in the
exported workspace. The embedded external files can include the database of risk
results, the map image files defined under any Raster Image Sets, and the
System Materials files and System Parameters files that are being used by the
current workspace. The exported workspace file will not include any GIS database
files defined under an External connection, as raster images are the only type of
map image data included in the export.
When you click on OK in the dialog, a File Save dialog will appear, prompting you
for a file name and location for saving the exported workspace file.
You can use the Open option in the File menu to open an exported workspace directly, in the same way
as with any other *.psux file. When you open an exported workspace file, the program will first extract
the embedded data, which involves attaching any risk database to the SQL Server, and placing any map,
materials and parameters files inside a system of subfolders underneath the workspace file. The program
will then load the data from these external files into the workspace, which means that if the subfolders
under the workspace include system materials or parameters data, the workspace will take its system
values from these subfolders, instead of from the central location that is used by other workspaces on
your machine.
Note: The Export Workspace option in the File menu is the most visible way of creating an exported
workspace. However, the Email Software Support option in the Help tab of the Ribbon Bar also gives you
the option to export the current workspace and attach the exported workspace to the email.
For further details of working with exported workspaces, enter “Exporting a workspace” in the Index tab
of the online Help.
1.4 Option to set the colour to be used for each risk level in the
risk contour plot
In the Contours tab of the Risk Preferences dialog, the table
that allows you to set the levels to be calculated for the risk
contours now includes a Colours column as shown that displays
the colour that will be used for each contour in the risk contour
plot and allows you to change the colour. The default colour for
the 10-6 /year contour is set to red, as shown in the illustration.
To open the Risk Preferences dialog, click on Risk Preferences in

the Settings tab of the Ribbon Bar. If you want a particular set
of colours to be used for all workspaces, define the colours in a
new workspace, and then use the Workspace Template > Save
option in the File tab of the Ribbon Bar to update the definition
of a new workspace with these colours.

2 NEW FEATURES APPLICABLE TO USERS OF THE FULL SAFETI
PROGRAM
2.1 Modelling of isolation and blowdown for time-varying

releases
The input dialogs for the Pressure Vessel Equipment item and
the Time varying Leak and Time varying short pipe Scenarios
now include a Safety system modelling (isolation and
blowdown) field as shown in the Time varying releases tab
section.
If the field is set to No, the inventory is assumed to be isolated immediately, at the start of the release,
and blowdown is not considered. This is the approach that was used for the discharge modelling of all
time-varying Pressure Vessel releases in previous versions of the program.
If the field is set to Yes, a Safety systems tab will be present in the dialog as shown below.
Default values for Safety system modelling (isolation and blowdown) and for the two Time fields
are set in the Discharge parameters.
The modelling of isolation

If you set Isolation? to Successful, the discharge calculations will assume that the conditions at the
start of the release are maintained until isolation occurs, i.e. the discharge rate will be constant until
Time to isolation is reached.
If you set Isolation? to Fails, the discharge calculations will model a partial failure of isolation, in which
process shutdown does occur at Time to isolation, but one isolation device fails to close at this time
and the inventory from a neighbouring vessel is added to the inventory available for release. The
discharge rate will be modelled as constant until Time to isolation is reached, and the inventory used in
the discharge calculations will then be increased by Inventory from neighbouring vessel to be
added upon isolation failure.
The modelling of blowdown

If you set Blowdown? to Successful, you must supply a value for the Blowdown diameter. Once the
Time to blowdown actuation is reached, the discharge calculations will model discharge through the
blowdown system. You must set Time to blowdown actuation to be equal to or greater than Time to
isolation.
If you set Blowdown? to Fails, blowdown will not be modelled.

You will see the effect of the operation of
the safety systems in the Flowrate Graph
for the TV Discharge results, as shown.
The illustration shows a constant release
rate until shutdown at 120 s, and then a
rapid decrease in the release rate when
blowdown starts at 180 s.
If blowdown has been modelled for a

Scenario, the Flowrate Graph will also
show the blowdown rate as a function of
time, as shown.
Note: if the Method for calculating

average rate is set to Given time or
Average between 2 times in the Time
varying releases tab, and the maximum
time set for the averaging is less than the Time to isolation, the operation of the safety system will not
affect the value calculated for the average rate, and this means that it will not affect the size of the
effect zones calculated for hazardous effects.
2.2 Consequence Data Results View with summary effect zone

distances
The Results group of the
Home tab in the Ribbon
Bar now includes a
Consequence Data
option, as shown.
The option will be enabled if a node is selected in the Models tab that covers at least one Scenario that
has consequence effect distance results. This option will generate a set of tables with summary data
about consequence effect distances, and with an option in the Ribbon Bar to export the data to an Excel
file. The results are presented in the form of a Grid View, as shown below, which allows easy comparison
of results between Scenarios and Weathers.
The View includes results at a fixed height of interest, and also for fixed levels of toxic lethality, radiation
intensity and explosion overpressure. These fixed values are set in the Parameters, and a particular
Consequence Data View will use values for the Parameter Set that is selected for the Run Row that was
set as active at the time that the Consequence Data View was generated.
For further details of these results, enter “Consequence Data Results View” in the Index tab of the online
Help.

2.3 Parallel run method for running consequence calculations
for multiple Scenarios at the same time
When you use the Run option in the Home tab of the Ribbon Bar to run the calculation for the selected
node in the Models tab or the Run Row Selector, the program will now either perform the calculations as
a series run, running all consequence calculations on a single CPU core, or it will perform the calculations
as a parallel run, running the consequence calculations for multiple Equipment items at the same time
using parallel processing on multiple CPU cores.
The method that is used in a given run depends on the setting for the Run mode option in the Settings
tab for the Ribbon Bar, and also on the number of Equipment items selected for running.
The Run mode options in the Settings tab

The Settings tab of the Ribbon Bar now contains a Run mode option.
If your computer has a processor with more than one core, there will be three
options available under Run mode, as shown. The setting for Run mode determines
how the program will handle the processing of consequence calculations whenever
you use the Run option in the Home tab of the Ribbon Bar or in the right-click
menu.
If your computer has a processor with a single core, the only option available for Run mode will be
Series.
If Run mode is set to Parallel, the program will be able to run the consequence calculations for multiple
Equipment items at the same time using parallel processing on multiple CPU cores, whereas if it is set to
Series, the program will always run all consequence calculations on a single core.
The Parallel run mode has the potential to run the calculations more quickly than the Series run mode,
although in a small analysis the calculations might in fact run more slowly because of the additional time
required for the program to manage the processing between the different cores. With Run mode set to
Smart (recommended), the program will assess the volume of calculations that are currently selected to
be run, and will perform a standard run on a single core if the volume is small and a parallel run on
multiple cores if the volume is larger. When the program is first installed, the Run mode will be set to
Smart (recommended) if your computer has a processor with more than one core.
The Series run method is always used if a single Equipment item is selected for running
The parallel method runs all of the consequence calculations for a given Equipment item on a single core,
which means that consequence calculations will only be run at the same time on multiple cores if more
than one Equipment item is selected for running. For this reason, the program will use the series method
if the selection in the Models tab or the Models Playlist only covers one Equipment item, even if the Run
mode is set to Parallel.
The program behaves differently for each run method

With the series run method, the program will behave as it did for all calculations in previous versions of
the program. Messages about the calculations for the individual Scenarios and Run Rows will be
displayed in the Output View during the course of the run, and the Study Tree and Run Row Selector will
show the progress through the list of Scenarios, displaying an icon by each node after it has been run to
show whether or not the calculations ran successfully. The Stop option will be enabled in the Home tab
of the Ribbon Bar, and you can use it to stop the calculations. There will also be a progress bar displayed
in the status bar along the bottom of the program window, with a Cancel button next to the progress bar
that you can also use to stop the calculations.

With the parallel run method, a Runner window as shown will
be displayed in the centre of the program window. You will not
be able to perform any actions in the program window while the
Runner window is open, and you must either wait for the
calculations to finish, or click on Cancel to stop the processing
and close the Runner window.
The Runner window shows the status of the calculations with a

progress bar and a brief description of the current stage of the
processing, and also displays the messages that are being
generated for the calculations and that would be displayed in
the Output View during a series run.
When all of the calculations are complete, the Runner window will briefly display a message stating that
the calculations are complete and giving the time to run the calculations, and it will then close and you
will be able to use the program window again.
For details of the choice of method and of the behaviour, click on “Series and parallel run methods” in
the Contents tab of the Help window.

3 OTHER DIFFERENCES AND BUG FIXES
3.1 The release includes upgrades to the licensing software that

will allow network and dongle licenses to be used with v8.1
Users of network licenses must upgrade the License Manager
For licence management, Safeti uses Flexera Software’s FlexNet (formerly known as FLEXlm). Safeti v8.1
has been upgraded to use FlexNet 11.14, and this change means that you must also upgrade the DNV
GL License Manager if you want to use a network licence. Network licences are also known as
“concurrent” or “floating” licences and are typically corporate licences shared between several users, and
allocated and managed from a server.
If you want to continue to use your network licences with Safeti, you must take the following steps
before running Safeti:
1. Stop the RMS Licence Service running.
2. Uninstall the version of the DNV GL Licence Manager that is current installed on the server.
3. Install the latest version of the DNV GL Licence Manager that is included with the v8.1 release.
Detailed installation instructions for the DNV GL Licence Manager are included with the installation files.
If you do not upgrade the DNV GL Licence Manager,

then a Safeti licensing dialog will appear as show when
you try to run Safeti v8.1 with a network license.
If this dialog appears even after you have upgraded the

DNV GL Licence Manager, you should contact DNV GL
Software Support for further assistance.
Users of dongle licenses must upgrade the dongle drivers and might need to request a new dongle
If you use a dongle license, you must upgrade the dongle drivers as described in the installation notes.
The dongle type shown on the left hand side of the picture is no
longer supported and will not work with v8.1; if you have this type
of dongle you should contact DNV GL Software Support and request
the latest type of dongle. This older type of dongle is labelled
FLEXID 9-6 and has a green handle that is 4cm/1.5” long; the types
of dongle that are newer and that are supported have shorter
handles as shown in the picture.

3.2 Simplifications in specification of release phase for storage
Scenarios
Several simplifications have been made, affecting different types of Scenarios in different ways.
The “Phase to be released” field is no longer present in the Equipment dialogs

The field has been removed from the Material tab for a Pressure vessel and an Atmospheric storage
tank. This means that is it no longer possible to set a default value for the phase.
The “Phase to be released” field is no longer present in the dialog for some types of Scenario
The field has been removed from the Material tab for the following types of Scenario:
 Catastrophic rupture, for which the release phase is modelled as the storage phase defined for
the Equipment item.
 Spill, for which the release phase is always liquid.
 Vent from vapour space, for which the release phase is always vapour.
The dialog for Short pipe and Time-varying short pipe Scenarios has a single field for the release phase
The Material tab for these two types of Scenario used to have two fields for the release phase, and either
the “Phase to be released” field or the “Phase to be released for 2-phase vessels” field would be enabled,
depending on the storage conditions and the details of the type of scenario. There is now only one
release phase field.
For Time-varying Leak and Time-varying short pipe line rupture Scenarios, the release phase is
calculated depending on the storage conditions and the value for the “Release height from vessel bottom”
For these types of Scenario, the release phase field in the Material tab appears as “Calculated phase to
be released” and is always disabled.
The calculated phase will be shown as Liquid (release below liquid level) if the storage conditions are
liquid, or if the storage conditions are two-phase and the value for “Release height from vessel bottom”
is less than the tank liquid level.
The calculated phase will be shown as Vapour (vapour storage or release above liquid level if the storage
conditions are vapour, or if the storage conditions are two-phase and the value for “Release height from
vessel bottom” is equal to or greater than the tank liquid level.
The Scenario tab for these types of Scenario now shows the “Tank height” and the “Tank liquid level”, to
help you set a suitable value for the release height.
These changes mean that it is no longer necessary with two-phase storage to make sure that the value
that you set for the release phase is consistent with the value for the release height.
For Time-varying disc rupture and relief valve releases from two-phase storage, the “Release height from
vessel bottom” is always disabled
For these types of Time-varying short pipe release, it is no longer necessary to specify the release height
for two-phase storage and the “Release height from vessel bottom” is now always disabled. If you set
the “Phase to be released” to Vapour, the calculations assume that the release is above the release
height, and if you set the “Phase to be released” to Two-phase, the calculations assume that the vessel
contents are no longer separated and a release from any height will be a two-phase release.

3.3 Changed approach to generating Fire Scenarios for a
Warehouse
In previous versions, the Fire Scenarios for a Warehouse were generated when you ran the calculations
for the Warehouse, or when you used the Generate warehouse scenarios option in the right-click menu
for the Warehouse. If a Warehouse had Fire Scenarios present and you changed the input data for the
Warehouse, the Fire Scenario nodes might be deleted when you closed the Warehouse dialog or they
might be retained, depending on whether or not the existing number of nodes was the same as the
number required with the new set of input data.
This behaviour could be confusing, especially when dealing with a Warehouse and its Fire Scenarios in
the Model Playlist, and the behaviour has been changed in the ways described below.
Fire Scenarios nodes are now always regenerated when you make changes in the Warehouse dialog
If you make changes to the input data in the Warehouse dialog, then when you click on OK to close the
dialog, the program will delete any existing Fire Scenario nodes and generate a new set of nodes for the
changed inputs. All of the new nodes will be selected in the Model Playlist for the active Run Row.
All Fire Scenario nodes for a Warehouse will be deleted if you change the dimensions of the building
selected for the Warehouse
You do not define the dimensions of the warehouse directly in the Warehouse dialog, but instead select a
Building for a list of those defined in the Map tab of the Study Tree. If you edit the dimensions for a
Building that is selected for a Warehouse, a warning message will be displayed in the Output View about
the fact that the Warehouse has had its associated building altered, saying that “The warehouse's
scenarios have been deleted and must be re-generated”.
In this situation you must right-click on the Warehouse node and select Generate warehouse scenarios
from the right-click menu. The program will generate a new set of nodes for the changed input data, all
of which will be selected in the Model Playlist for the active Run Row.
Fire Scenario nodes are no longer generated when the calculations are run for the Warehouse
If a given Warehouse does not have any Fire Scenario nodes present when you start a set of calculations
that include the Warehouse, the Fire Scenarios will not be generated during the calculations and nothing
will be calculated for the Warehouse.
If you want to include the Warehouse in the calculations, you must now ensure that the Warehouse has
an up-to-date set of Fire Scenario nodes before you start the calculations.
3.4 The area of the maximum footprint is given in the Cloud

Maximum Footprint Graph and GIS Results
The legend for the Cloud Maximum Footprint dispersion results in the
Graphs and GIS now gives the value of the area for each cloud-
concentration contour and pool footprint displayed. As shown in the
illustration, the value for the area is displayed in brackets at the end of the legend text.

3.5 More consistent method of defining vessel contents for a
Standalone BLEVE blast Scenario
In previous versions, it was possible to set a value for the inventory of the vessel that was not consistent
with the dimensions and liquid content, and the method of defining the liquid content was also unclear.
Some changes have been made to the Storage conditions tab of the Scenario dialog to make it easier to
supply the correct data.
The liquid fraction is now defined on a molar basis instead of a mass basis
A Liquid mole fraction field has been added to the dialog as shown below and will be enabled if the
Specified condition is set to Pressure/bubble point or Temperature/bubble point.
The Liquid fraction mass field is still present, but it is now always disabled and displays the
corresponding mass fraction value calculated from the storage conditions and the liquid mole fraction.
For the model options that require the vessel dimensions, the program now calculates the inventory
The BLEVE blast parameters tab contains a Model option field as shown
that gives a choice of three methods of modelling the BLEVE blast. If this is
set to CCPS second edition (2010) or Brode option (2017), the fields for the
elevation and dimensions for the vessel are enabled and mandatory in the
Storage conditions tab. Given the dimensions of the vessel and the liquid fraction, the program is able to
calculate values for the inventory, so in this situation the inventory fields at the bottom of the tab
section are all disabled as shown above, and they display the values that the program has calculated for
the mass and volume inventory.
If the Model option is set to CCPS first edition (1994), the elevation and dimension fields are disabled,
and you must specify a value for the inventory in the same way as in previous versions.
If you upgrade a file from a previous version that contains Standalone BLEVE blast Scenarios with the
CCPS second edition or Brode options selected, the program will recalculate the inventory based on the
dimensions, overwriting the value that had previously been set for the inventory.

3.6 Improved handling of small hole sizes and improved
presentation of results for Long pipeline Scenarios
Several improvements have been made for the Long pipeline Scenarios. Details of the long pipeline
modelling are available in the technical documentation supplied with the program.
The minimum aperture size for a Location specific breach has been reduced
In previous versions, the minimum aperture size that could be defined for a Location specific breach was
much larger than for a Section breach and was set at 20% of the total outflow area. The Location
specific breach now has the same limit as the Section breach, which is less than 1 mm.
Steady state discharge calculations for small hole sizes can now be performed for gas pipelines as well as
for liquid pipelines
The Long pipe tab of the Long pipeline dialog contains a Small holes method group as shown in the
illustration below. This group allows you to specify how the discharge calculations should be performed
for a Breach Scenario where the hole size is classified as small, and to set the criterion for classifying a
Scenario as a small hole.
The two Steady state options calculate the rate using the initial operating conditions, not taking into
account changes in the pipeline conditions resulting from the release, whereas the Time varying pipeline
model takes into account changes in the pipeline conditions resulting from the release. The time-varying
calculations are time-consuming, and for a small hole, the change in the discharge rate over the time-
period of interest might not be significant. Using the Steady state options can reduce the calculation
times for the Long Pipeline.
In previous versions, the time varying pipeline calculations were always performed for a gas pipeline, no
matter what option was set for the Small hole size discharge method, i.e. the option to use steady
state calculations was only applied to liquid pipelines. The steady state modelling has now been
extended to gas pipelines, and the option set for the Small hole size discharge method will be applied
to all types of pipeline.
Long pipeline discharge graphs are now available for the Rationalised Discharge Scenario for a Section
Breach
In previous versions, the set of graphs for a Rationalised Discharge Scenario generated under a Section
Breach did not include the group of Long pipeline graphs that give the time-dependent discharge results.
These graphs are now included, and the set of graphs for a Rationalised Discharge Scenario now includes
all of the graphs that are included for a Location specific breach.
Reports can be viewed for an individual Rationalised Discharge Scenario

In previous versions, the Report Selector dialog did not allow you to select an individual Rationalised
Discharge Scenario. To view reports for a Rationalised Discharge Scenario, you had to select the Section
Breach in the Report Selector dialog and the Report View that was generated would include results for all
Rationalised Discharge Scenarios. This has changed, and you can now generate a Report View that
covers any set of Rationalised Discharge Scenarios that you choose.

3.7 Choice between use of radiation dose or intensity in
flammable risk calculations
In the Probit Method for the modelling of flammable risk, the program will model the radiation effects
from a fireball, jet fire and pool fire as a set of ellipses, with the innermost ellipse taken as the boundary
of the zone in which there are two types of hazard: a combustion hazard and a radiation hazard. The
options for this method are set in the Flammable risk tab of the Flammable parameters dialog as shown
below.
The innermost ellipse for a given type of fire can be defined either by the Critical radiation intensity
or by the Critical radiation dose.
In previous versions the radiation intensity was used for all

three types of fire, but now the Parameters dialog for each of
the three types of flame includes the field Measure to use for
critical radiation that allows you to choose whether to use
intensity or dose for that type of flame. The illustration shows
the field in the Fireball and BLEVE blast parameters; it has the
same name and the same options in the Jet fire parameters
and Pool fire parameters dialogs.
By default, Measure to use for critical radiation is set to

Dose for fireballs and to Intensity for jet fires and pool fires.
In versions before v8.0 the default value for the critical radiation intensity was 35 kW/m2. It was set to
400 kW/m2 in v8.0 but has now been set back to 35 kW/m2 as shown in the first illustration above.
Note: In previous versions there was a Radiation dose for fireball risk calculations field in the
Fireball and BLEVE blast parameters dialog, but this was not in fact used in the risk calculations for
fireballs. The calculations for all three types of fire used the value for radiation intensity that was set for
the field in the Flammable parameters that was called Radiation level for jet/pool fire risk. The
radiation dose field has now been moved to the Flammable parameters dialog as shown above, and both
fields have been renamed to make their meaning and their relationship to each other clearer. The value
set for critical radiation dose will now be used correctly for any type of fire for which the Measure to
use is set to Dose.

3.8 The modelling of jet fires for time-varying releases uses the
settings for the time-averaging method
In previous versions, the representative discharge conditions used in the modelling of a jet fire for a
time-varying release were those for the average rate during the period between the start of the rate and
the Jet fire averaging time set in the Jet fire parameters (i.e. over the first 20 s with the default value
for the averaging time).
The representative discharge conditions now depend on the setting for the Method for calculating
average rate in the Time varying tab for the Scenario, as follows:
Setting for averaging method Calculation of representative discharge conditions
Average rates or Up to 10 rates The program calculates the average rate between the start of the
release and the Jet fire averaging time. The average rate is
given by the mass released in the time considered, divided by the
time.
Given time The program uses the conditions at the time set for the Time at
which the release rate is calculated field in the Time varying
tab for the Scenario.
Average between 2 times The program calculates the mass discharged over the time period
between the times set in the Time varying tab for the Scenario for
First time value for rate between two times and Second time
value for rate between two times. The average rate is given by
the mass released in the time considered, divided by the time.
For the methods in which the program calculates an average rate, it obtains the discharge conditions
(temperature, velocity, etc.) for the time when the rate was equal to this average rate, and uses these
conditions to represent the release.
This change means that the representative discharge conditions used in the jet fire calculations will now
be consistent with the representative discharge conditions used in the dispersion calculations.

3.9 Clarifications for meaning and application of input data
Some changes have been made in the visibility, enabling and interpretation of aspects of the input data.
Depth of soil cover for a buried Long pipeline is specified from the top of the pipeline instead of from the
centre
In v8.0, the value for Depth of soil cover in the Long pipe tab section for a Long pipeline was taken as
being measured upwards from the centre of the pipeline, and it was possible to define a buried pipeline
with a depth of cover that was less than the radius of the pipeline.
The value for Depth of soil cover is now taken as being measured upwards from the top of the
pipeline, which means that a buried pipeline is always treated as being fully buried.
Temperature for an Atmospheric storage tank no longer defaults to the normal boiling point for the
selected material
In previous versions, whenever you made a selection for the Material in the Material tab section for an
Atmospheric storage tank Equipment item, the Temperature field would automatically be set to the
normal boiling point for that material.
There is now no automatic default value for the Temperature field and you must set a value directly.
The value that you set must not be greater than the normal boiling point for the selected material.
Elevation no longer an input for the User-defined pool source Scenario

In previous versions, the Elevation field was always enabled in the Scenario dialog for a User defined
source Scenario.
The field is now disabled if the Release scenario is set to Pool source (radius). The field is not relevant
to such releases because they are always modelled at ground level.
Vulnerability option not available for a Building type if the Explosion method is set to “3D Cloud
ME/Purple Book Explosions”
In previous versions, the Building vulnerability options tab was always present in the dialog for a Building
type, even if the Explosion method was set to 3D Cloud ME/Purple Book Explosions. The data in this tab
are only relevant for a building when the risk calculations include the risk to Building Populations, and
since the risk to Building Populations is not modelled with the 3D Cloud ME/Purple Book Explosions
method, this tab section is not relevant in this situation and is now no longer displayed.
Automatic event spacing method for Long pipeline no longer has a field for the choice of spacing method
In previous versions, the Event spacing group in the Risk tab of the Long pipeline dialog included an
Automated event spacing method field that was enabled when the Event spacing method field was
set to Automatic.
The list for the Automated event spacing method field only contained one option - Use peak lethality
– and the field has now been removed.
For the Pool vaporisation parameters, the Help states that the choice for “Evaporation of land
correlation” applies to all modelling of pool vaporisation on land
In versions before v8.0, the choice of correlation applied only to the standalone Pool vaporisation
Scenario, but the extensive improvements to the dispersion modelling in v8.0 included extending the use
of this correlation to the pool vaporisation modelling for Source Scenarios. The Help has been updated to
reflect this change.

Descriptions of event tree parameters now make clear that a delayed explosion is always accompanied
by a flash fire
In previous versions, the fields for defining the probabilities of outcomes that involved explosions had
descriptions that referred only to “explosion”, e.g. for delayed ignition there was a Delayed ignition
flash fire field and a Delayed ignition explosion field. This implied that the explosion outcome
involved only an explosion, whereas flash fire effects would also be modelled.
The descriptions have been changed, so that any explosion outcome that also involves a flash fire is
described as “Explosion with flash fire”, as shown.
For the Risk inputs for a Scenario, the Help states that the option to “Ignore risk from fireballs” applies
only to fireball outcomes that do not include pool fires
In versions before v8.0, the option to “Ignore risk from fireballs” was applied to all types of fireball
outcome, i.e. to those with pool fires and to those without pool fires. The risk calculations were changed
in v8.0 so that the option now ignores only the risk from fireball outcomes that do not include pool fires,
and the Help has been updated to reflect this change.
3.10 Bug fixes

The following bugs have been fixed in v8.1:
1 B-16418 Liquid behavior not being modelled for some Atmospheric Storage Tank Scenarios
Description In some situations, the dispersion modelling for an Atmospheric Storage Tank
Scenario with an Elevation greater than zero would not show any liquid present and
would not model rainout and pool vaporization, even though the discharge results
showed the release as liquid. This was an effect of the modelling of the vertical
temperature gradient in the atmosphere which was shifting the temperature of the
dispersing release to above saturation temperature, and the treatment of the
temperature of atmospheric releases has now been adjusted so that this will no
longer happen..
2 B-16667 Results for toxic Scenarios did not correctly state the type of toxic averaging time used
Description The Graphs and Reports that include a statement of the toxic averaging time always
gave the value set in the Toxic Parameters (e.g. “Toxic (600 s)”. This value is only
applicable to a Scenario if the option to Set averaging time equal to exposure
time is set to Use a fixed averaging time, as different values would be used if the
option was set to Averaging time equals exposure time.
Now, if the option for the Scenario is set to Averaging time equals exposure time,
the statement of the toxic averaging time use will appear as “Toxic = exposure
time”. This applies to the Summary Report, the Averaging Times Report, Dispersion
and Toxic results in the Graphs and GIS, and to the Toxic results in the
Consequence Data View.
Note: if the option is set to Use a fixed averaging time and the Graph View is
displaying results for multiple Weathers or Scenarios, the Dose, Probit and Lethality
Toxic Graphs will give the averaging time as “Multiple averaging time”, although all
other Graphs and all Reports will give the correct value, i.e. the value set in the
Toxic Parameters.

3 D-11481 The pool vaporization results for a mixture could depend on the order of the
components in the mixture
Description This applied to the Standalone Pool vaporization scenario, if Multi or pseudo-
component modelling was set to PC modelling. For a mixture, the results such as
evaporation rate, pool temperature, pool radius etc. depended on the order in which
the components had been added to the mixture when the mixture was first defined.
This bug was fixed in v8.0, so results for this type of Scenario will change between
v7.22 and any later version.
4 D-11858 Problem on changing Material when Material to Track was set

D-13280
Description If a mixture was selected for an Equipment item, and one of the components of the
mixture was selected as the material to track for the Equipment or a Scenario, this
component would remain selected if you changed the selection of Material to
something that did not include the selected component. The Equipment and
Scenario would not be shown as having an error in the input data, but the Scenario
would give errors in the dispersion calculations about the chosen component not
being present in the mixture.
This has been fixed so that whenever you change the selection for the Material, the
Material to Track in the Equipment dialog will be reset to be the same as the
Material, and the Material to Track fields in all Scenario dialogs will be reset to
default to the value set in the Equipment dialog.
5 D-12104 Dispersion Graphs always present for time-varying Scenarios even if only the discharge
calculations have been run
Description If you set the Mode to Discharge in the Home tab of the Ribbon Bar, run the
calculations for a time-varying Scenario or a Long pipeline breach Scenario and then
view the Graphs, you would find that the full set of Dispersion Graphs were present,
though all of the graphs were blank except for the Mass Rate Graph.
This has been fixed, and the Dispersion Graphs will not appear in the Graph View.
6 D-12390 Time-varying Scenarios not shown in the Study Tree as run successfully when the
Calculation Mode is set to Discharge
Description If the Calculation Mode was set to Discharge in the Home tab of the Ribbon Bar and
you ran the calculations for a set of source Scenarios, the Scenarios whose
discharge calculations ran successfully were shown with green ticks in the Study
Tree for most types of source Scenario, but not for Time-varying leak Scenarios or
Time-varying short pipe Scenarios. This has been fixed.

7 D-12457 Explosion Flammable Mass not being reported in the Consequence Summary Report
Description In the Explosion Results section of the Consequence Summary Report, the table of
Supplementary data for worst-case explosion results includes an “Explosion
flammable mass” column that was incorrectly being reported as “n/a” for all
Scenarios. This has been corrected and the table now gives the values.
8 D-12545 Risk tab visible in Section breach dialog for a Long pipeline when Mode was not set to
“Risk”
Description The dialog for most types of Equipment item and Scenario includes a Risk tab that
contains inputs that are used in the risk calculations only and are not relevant to the
consequence calculations. This tab should only be displayed when the Mode is set to
Risk in the Home tab of the Ribbon Bar, but it was always present in the dialog for a
Section breach Scenario for a Long pipeline, even when the Mode was set to
Discharge or Consequence. This has been corrected, and the tab is now only visible
when the Mode is set to Risk.
9 D-12591 No graphs available for the Standalone TNT explosion Scenario
Description The Standalone TNT explosion Scenario had results available in the form of Reports,
but the Graphs View for the Scenario would be empty, without any Explosion
Graphs. This has been fixed and Early Explosion Distance and Early Explosion Radii
Graphs are now produced.
10 D-12742 The option to save the workspace could stop working
Description The program uses an SQLite database to reduce memory use by storing results and
other information to the database and only reading them back into memory when
needed. This database is copied into the *.psux file when you save the workspace.
Occasionally the program would try to update the database as the same time as it
was saving it in the *.psux file, and this would cause that save operation to fail,
and also cause the Save and Save As options to become disabled.
This problem has now been fixed.

11 D-12772 Weather dialog was displaying default values from the System Parameters, not from
the Parameters inside the workspace
Description In the Weather dialog, all fields that took a default value were displaying the value
set in the System Parameters, not the value set in the appropriate Parameter Set in
the workspace. This meant that if you edited the Weather Parameters in the
workspace and set values that were different from those in the System Parameters,
you would not see the change in the default values in the Weather dialog. This
problem affected only the display of the default values in the Weather dialog; the
Grid View, the Reports and the calculations all used the correct values, taken from
the appropriate Parameter Set in the workspace.
This problem has now been fixed, and the Weather dialog displays the correct
defaulted values.
12 D-12838 Improved modelling of releases for which the peak pool vaporization rates occur
while the cloud is passing over the pool
Description In conditions of low wind-speed, where the cloud is moving slowly, the peak
evaporation rates for a pool may occur while the cloud is still passing over the pool.
In this situation, the approach used to define observers for the dispersion modelling
meant that the observers being used were not picking up the transient high
evaporation rates, so some of the mass evaporated was not being added to the
cloud, and mass was not being conserved in the modelling. The dispersion
calculations perform checks for mass conservation, and if these checks fail, a
warning message will be generated that states that the modelling is switching to
the equivalent pool approach. With this approach, the total mass vaporised is
modelled as being released at a constant rate over a duration close to the total
duration of the pool, which ensures that mass is conserved but results in lower
concentrations over a greater time-period, missing the periods of higher
concentration from the time-varying pool vaporisation.
Several improvements have been made in this area, in the method used to check
mass conservation and in the approach to the definition of observers to represent
the pool. These changes mean that the warning about failing mass conservation
checks should appear less frequently. In addition, for situations in which the checks
do fail and the equivalent pool is used, some changes have been made to the
modelling of the equivalent pool to reduce the duration and increase the
concentrations.
13 D-12991 Time-varying discharge calculations could stop too soon
Description In some situations the time-varying discharge calculations from a two-phase vessel
would stop prematurely when the liquid level dropped to the height of the hole,
even though the vapour space was still pressurized. This has been fixed, and the
vapour-stage of such releases will now be modelled.

14 D-13004 Results displayed differently in Toxic Footprint Graphs depending on the number of
Weathers selected for Graphing
Description The program uses grids for the calculation and display of toxic contour results, and
there was an error in the way the grids from the calculations were being processed
for graphing. This affected the Footprint Graphs only, and did not affect the other
Toxic Graphs or the toxic results used in the risk calculations.
This error has now been fixed, and the Toxic Footprint results for a given Scenario
and Weather will be displayed correctly in the graphs.
15 D-13073 Error viewing GiS Consequence Results for a combination of Storage and Standalone
Scenarios
Description If you selected the GIS Results option for a Weather, and then selected a
combination of Storage and Standalone Scenarios for Graphing, error and warning
messages would be generated while the GIS View was opening, and no cloud
footprint results would be displayed although other types of results could be viewed
successfully.
This error has now been fixed, and the GIS Results View will display all available
types of results successfully, for any combination of types of Scenarios.
16 D-13144 Maximum Footprint Graph not generated for some Scenarios using Roof/Lee building
wake modelling even though results were available for the Footprint Graph
Description For some Scenarios with Roof/Lee building wake modelling selected, no results
were available for the Maximum Footprint or Max Conc vs Distance Graphs, even
though other graphs such as the Footprint Graph showed that results should have
been available.
This bug has now been fixed.
17 D-13186 Program crashed during the risk calculations if a Scenario had pool results for the first
Weather but not for other Weathers
Description If the set of Scenarios selected for risk calculations included a Scenario that had
pool results present for the first Weather but not for the others, then the program
would shut down without giving any error message.
This bug has been fixed and the crash will not occur.

18 D-13212 Risk calculations not performed for early pool fire effects in some situations
Description This bug applied if all of the following conditions were present:
1. You had a license for the 3D explosion modelling.
2. The setting for Method for radiation vulnerability for a Vulnerability was
set to Use intensity method instead of the default of Use probit method.
3. The value for Exposure time required for damage from pool fire in the
Radiation intensity method tab was set to a value greater than zero.
In this situation, the risk calculations would not model the effects from the early
pool fires. This would affect the risk results for continuous releases with rainout,
with a non-zero probability of immediate ignition.
This bug has been fixed and the risk calculations will now include early pool fire
effects in this situation.
19 D-13218 Vent from vapour space Scenario modelled with incorrect value for liquid fraction
Description In some situation, a Vent from vapour space Scenario could be modelled with an
invalid value for liquid fraction. This would mean that the velocity of the release
used in the dispersion modelling was effectively set to zero, and that Scenarios with
different values for the vent diameter would give identical dispersion results, even
though the discharge results showed different values for the velocity.
This has been fixed and the liquid fraction is now always modelled as zero for a
Vent from vapour space Scenario.
20 D-13281 Risk contributions could be shown as assigned to the wrong Scenarios if the Model
Playlist included Scenarios with the frequency or probability set to zero.
Description If the Model Playlist for a Run Row included Scenarios with the Event frequency or
Event probability set to zero, these Scenarios could cause a mismatch in different
aspects of the program that involve lists of relevant Scenarios, and this meant that
the Risk Results that show the contribution from different Scenarios could show the
wrong Scenario name, i.e. risk that was caused by Scenario A could be shown
incorrectly as being caused by Scenario B.
This has been fixed, and the source of the risk will now be shown correctly in these
results, even if the Model Playlist includes Scenarios with a frequency or probability
of zero.

4 PERFORMING A LARGE ANALYSIS
Performing a large analysis with Safeti 8.1 may require planning and the use of particular techniques and
tools. You should make sure that you understand the issues involved before you start work on a new
analysis that is likely to be large, or before you upgrade an existing large analysis.
4.1 Why is special attention needed for a large analysis?

There are two main factors that can make an analysis with a large workspace file difficult to work with:
Some operations can be very slow with a large workspace

Many operations become slower as the size of the workspace increases, including the following:
 Working with the nodes in the Study Tree, especially with Equipment and Scenario nodes in the
Models tab.
 Using the Grid View to change or view the values for input data.
 Using the GIS Input View, e.g. adding new data, or moving around the view.
 Using the GIS Configuration View in the Run Row Grid to view the set of map-based data
selected for an individual Run Row.
 Exporting input data values to Excel.
 Running the calculations.
 Viewing consequence results for a large number of Scenarios.
 Saving an analysis to a workspace file, if the workspace contains results.
 Opening a workspace that contains results.
 Upgrading a large analysis from a previous version of the program (e.g. from version 6.7.).
For example, saving a large analysis with results to a workspace file may take several hours.
The risk of running out of memory or disk space increases with a large analysis
With most of the operations listed above, there is also a risk that the program may run out of memory
and crash, which will mean that you will lose any work that you did on the workspace after the last time
you saved it to file.
The risk results can occupy a large amount of disk space, and there is the risk of running out of disk
space during the risk calculations, which will cause the program to crash and may also cause problems
for other programs on the computer.

4.2 At what point should an analysis be considered “large”?
As a rule of thumb, the size S of an workspace from the point of view of performance can be assessed as
follows:
where Nscenarios is the number of scenarios that are selected for Run Rows, Nweathers is the number of
Weathers that are selected for Run Rows, and NParameter Sets is the number of Parameter Sets that are
selected for Run Rows. These variables affect the volume of consequence results that have to be held in
memory during calculations, and a workspace can be considered large if S is greater than 1000. If
your analysis is large by this measure, there are steps that you can take to make the analysis more
manageable and reliable.
Other factors that affect the performance are the number of ignition sources selected (including
populations that are being modelled as ignition sources), the number of vulnerabilities defined for the
workspace, and the number of obstructed regions in the Sets that are selected for the Run Rows (if the
Explosion Method is set to “3D Obstructed Regions”). However, the three factors that affect the volume
of consequence results are those that have the largest effect on performance.
4.3 What can I do to make a large analysis easier to perform?

The options for approaching a large analysis are described below, in the order in which they should be
considered or employed in an analysis.
Use a computer with the highest specifications that you can obtain
The recommended specification for a large analysis in version 8.1 is as follows:
Operating system Microsoft Vista SP2 (32 bit version), Windows7 SP1, Windows 8, Windows 8.1
and Windows 10 (32 or 64 bit version).
Type of hard drive Solid State Drive (SSD)

This is the most important recommendation.
Size of hard drive The program itself requires up to 10 GB of free disk space on a standard
Windows 7 SP1 machine, and the input data and results for a single large
workspace can occupy 100 GB of disk space or more.
CPU Recommended: Intel i7, 64-bit CPU from Intel

Minimum: Intel Quad Core 2.7 GHz
The fastest machines we are currently aware of are the DELL Precision 7000
laptops.
Memory Recommended: 16GB

Minimum spare memory: 4GB
Microsoft Excel Safeti requires Excel for the Excel input/output tool
If you are purchasing a new computer you should make sure it has a NVIDIA graphics card that is CUDA
enabled (e.g. the M series), as the calculations for individual risk are able to run on either multiple CUDA
cores or multiple CPU cores, depending on the settings in the IRISK tab of the General risk parameters.
Aspects of the consequence calculations will run on multiple CPU cores if the option to Enable
multithreading for dispersion and toxic calculations is checked in the Workspace dialog, and
multiple CPU cores will be used to run consequence calculations for multiple Scenarios at the same time
if the Run mode is set to Smart (recommended) or Parallel in the Settings tab of the Ribbon Bar.
Build up the analysis gradually, examining intermediate results and performing sensitivity assessments
to see if the number of variables can be reduced
You should run consequence calculations and examine the results before proceeding to the risk
calculations. If the results for particular scenarios or weathers are very similar, you can reduce the size
of the workspace by combining or removing scenarios and weathers.
Having examined the range of consequence results, you might decide to perform some limited runs of
the risk calculations, with a selection of scenarios of different sizes and types, and with different levels of
detail in the modelling of the number of weather directions, populations, ignition sources, vulnerabilities
and obstructed regions. If these sensitivity assessments show that the differences in the levels of detail
do not give significant differences in the calculated levels of risk, you can reduce the size of the
workspace by using a lower level of detail.
Use the Excel input/output tool when setting up input data

The process of exporting and importing input data with Excel using the options in the Tools tab of the
Ribbon Bar is quicker and requires less memory than inserting nodes directly in the Study Tree and using
the input dialogs or the Grid View to set the input values. Having the input data defined in Excel also
means that you can edit and maintain the data easily over the course of an analysis, i.e. to make
changes that affect a large number of Scenarios as assumptions change, and then reimport the data into
the program.
If you need to copy or move a large number of Equipment and Scenario nodes from one part of the
Study Tree to another, it is better to export the data to Excel, edit the Excel file to set the path for the
Equipment items to the path for the new location (i.e. under a different Study or Folder), and then
import from the Excel file. If you try to copy (or cut) and paste a large number of nodes directly in the
Study Tree, the demands on memory may cause the program to freeze, but this will not happen if you
use Excel.
Consider dividing the analysis between separate workspace files and using the Study Manager to
combine the high level risk results
The Study Manager is installed when you install Safeti 8.1. It allows you import risk contour results and
Smoothed FN results from any number of separate workspace files, and then to add these results
together to obtain the full results for the analysis. The Study Manager has its own online Help and
Release Notes, that explain how to use it.
There are two approaches to dividing an analysis between separate workspace files:
1. Dividing different aspects of the input data between different files, but running all types of
relevant risk calculations for each file.
For example, you might define the Equipment for different units in different files, or you might
define the daytime and the nighttime conditions for weather, population and ignition sources in
different files.
2. Having the same input data in each file, but running separate aspects of the risk calculations for
each file.
The Diagnostics option in the Settings tab of the Ribbon Bar allows you to control which aspects
of the risk calculations to run for that workspace file, e.g. whether to calculate only individual
risk, or only societal risk, or both.
For a very large analysis, you might use both approaches.

Use Excel input/output to maintain the input data across multiple files
With both approaches, you will have several workspace files that have a lot of data in common, e.g.
population and ignition data, weather data, etc. Input values are frequently revised during the course of
an analysis, and the Excel input/output tool is the quickest and most reliable way of keeping the values
consistent across all of the files in the analysis.
Save the consequence results to file before starting the risk calculations
If you use the second approach - or if you think that you might use it- you should start with a single
workspace file, run the consequence calculations for that workspace and then save the file with the
results. You can then create a copy of the workspace file for each type of risk results that you want to
run separately, set the Diagnostic options in each file, and then proceed to run the risk calculations for
each file. It is always advisable to run the consequence calculations on their own before proceeding to
the risk calculations so that you can examine the results and make any appropriate adjustments to the
assumptions and input values, but running them separately in this way also makes it easier to adopt the
second approach, as you will not have to rerun the consequence calculations for each separate file.
Use Export Workspace instead of Save / Save As to save the workspace file after running the risk
calculations
The Export Workspace option in the File tab of the Ribbon Bar saves the workspace as a special format of
*.psux file that has the database of risk results embedded inside it, along with any map image files that
are used by the workspace. The Study Manager only allows you to import workspace files that are in this
format, and will give an error message if you try to import a workspace file that was saved with either
the Save or Save As options.
Note: if you use the Console tool to run the calculations for a workspace (as described in a later section),
the Console will always save the calculated workspace in the exported format. This means that you will
be able to import the calculated workspace into the Study Manager immediately after the Console has
completed the calculations, and you do not have to open the workspace in the Safeti program first in
order to use Export Workspace to create a version of the workspace in the exported format.
Divide Equipment between different Run Rows in a workspace to increase the level of detail available
from the Study Manager results
The types of results that are summed and displayed in the Study Manager do not allow you to drill down
to see which individual Scenarios or Weathers or types of effect or types of population contribute most to
the risk levels, and the lowest level of detail that is available in the Study Manager is the level of the
individual Run Rows in a particular workspace. To obtain a full breakdown of the contributions to the risk
levels, you must use the additional forms of results that are available in the Safeti program itself.
However, if you structure the workspace files so that different types of Equipment items or Scenarios are
divided between different Run Rows, the risk results for the individual Run Rows as presented in the
Study Manager will give you good information about the main sources of risk.
Put Route Models and Long Pipeline Models in separate workspace files
The Route Model and the Long Pipeline Model can involve modelling a large number of Scenarios at a
large number of different locations, and the calculations can consume a lot of time and memory.
To make the analysis more manageable, you should place each Route Model and each Long Pipeline
Model in its own workspace file.

Set program and parameter options that will reduce memory usage and the size of the risk results
There are five options that can have a significant effect on the performance in terms of the memory
usage and the size of the risk results.
In the Risk Preferences dialog: Produce detailed risk results
The Risk tab of the Risk Preferences dialog contains an option to

Produce detailed risk results as shown.
If this option is checked, the risk calculations will save the details of
the risk contributions from each individual outcome. This will
increase the volume of the risk results significantly, but will allow
you to see the full details of the risk results, e.g. to analyse the
contributions from different Weathers, wind directions or ignition times.
If the option is not checked, then some forms of risk results will be disabled in the Risk Gallery in the
Home tab of the Ribbon Bar.
In the dialog for the workspace node: Reduce memory usage in preference to calculation speed
The dialog for the workspace node at the top of the

Study Tree includes the option to Reduce memory
usage in preference to calculation speed, as
shown in the illustration of part of the dialog. When
this option is checked, measures will be taken during
calculations to reduce memory usage. This will
reduce the risk that the program will run out of
memory and crash, but it will make the calculations
run more slowly.
The meaning of this option is different for a Phast workspace and for a Safeti workspace; different
memory-saving mechanisms are used by the two programs, because of the different demands on
memory for a consequence analysis and a risk analysis. If a particular Phast 7.* workspace file has the
option checked and you open the file in Safeti, the program will continue to use the Phast-style
mechanisms for the workspace instead of the Safeti-style mechanisms.
In order to change to the Safeti-style mechanisms, you must take the following steps:
1. Open the workspace dialog, uncheck the option, and OK the dialog.
2. Save the upgraded workspace to file.
3. Reopen the workspace file.
4. Open the workspace dialog, check the option, and OK the dialog.
5. Save the workspace to file.
You should make this change immediately after upgrading the Phast file, and before running any
calculations.

In the Flammable Parameters dialog: the number of wind subdivisions per sector
The Flammable Parameters dialog contains the Number of

wind subdivisions per sector, as shown in the illustration of
part of the dialog.
The risk calculations model the distribution of effects from

flammable releases by considering a range of representative
directions. At a minimum, the program will consider each wind
direction or sector specified in the Weather Folder, and if this
parameter is set to be greater than one, the calculations will
also consider more than one representative direction inside
each sector.
By default the parameter is set to 5. If you reduce this number, the calculation times will be shorter but
the results are likely to be less smooth.
In the Flammable Parameters dialog: the population omega factor
The Flammable Parameters dialog contains two population

omega factor fields, as shown in the illustration of part of
the dialog.
The population omega factors are used in calculating the

ignition probability associated with people and their general
activities. The Population Omega Factor is applied to
outdoor Grid Populations, and the Indoor Population
Omega Factor is applied to indoor Grid Populations.
If either of these is set to greater than zero, then each

population will also be modelled as an ignition source, in
addition to the sources defined under the Ignitions folder. Each additional ignition source modelled in the
risk calculations increases the memory usage, as the calculations of delayed ignition are memory-
intensive. By default the factor used for outdoor populations is greater than zero, but you can reduce
calculation times if you set this to zero.
The grid sizing in the Grid parameters dialog
You should normally set the Grid

sizing option to Calculated, as this
ensures that the grid covers the full
area that could be affected by the
Scenarios that are selected for the Run
Row.
The maximum number of cells should

be kept a small as possible, and
usually not increased above the default
of 400x400 cells. Increasing the
number of cells increases calculations
times and memory usage.

Turn off features in the program window that use a lot of memory and reduce performance
There are two types of feature that can be turned off to improve performance:
The Status Checking options in the Settings tab of the Ribbon Bar
The Settings tab of the Ribbon

Bar includes two toggle buttons
for turning on and turning off
the checking of the status of
input data and results. By default they are both turned on, as shown in the illustration.
These checks can be very time-consuming with a large analysis, leading to long pauses after each
change of input values. You should turn off the checks while you are working on the input data for the
analysis, and then turn them back on when you are ready to run the calculations.
The types of View that require constant updating as you move around the Study Tree and change data
Some of the Views in the program window change their content depending on what Study Tree node or
other item is selected elsewhere in the program. As you move around the program between different
items, these Views have to be updated constantly, and this is very time-consuming. You can improve
performance if you close these views whenever you are not actively using them.
The Views that you should close wherever possible are the following:
 The Grid View
 The GIS Input View
 The GIS Configuration View that is displayed when you are working in the Run Row Supertab
To close a View, click on the Close button at the right of the tab for the View, as
shown in the illustration.
To reopen a View after you have closed it, select the option for
that View in the View tab of the Ribbon Bar, as shown.
Avoid generating consequence reports and graphs that cover large number of scenarios
When you select the option to view consequence reports or graphs, the Report View or Graph View will
include the results for all of the Scenarios that are covered by the current node that is selected in the
Study Tree. If there are a lot of Scenarios under the current node, the Report View or Graph View will be
very large and will consume a very large amount of memory.
Whenever you view a consequence Report or Graph, you should try to limit the number of Scenarios
covered to a maximum of about 20. This should ensure that the Report or Graph is small enough to be
responsive as you work with it, and that the program does not run out of memory while displaying it.

Check that you have at least 100 GB of hard disk space free before starting the risk calculations for a
large workspace
The risk results for a single large workspace can occupy 100 GB of disk space or more. Before you start
the risk calculations for any large workspace, you should check that the hard disk has at least this much
space free.
If you wish to remove old database of risk results which might be taking up space on your C: drive, you
can simply delete them from the following locations:
C:\ProgramData\DNVGL\Safeti_Phast_Phoenix\DatabaseStorage
C:\ProgramData\DNVGL\Safeti_Phast_ Phoenix\DatabaseStorage\temp
Consider using the Console tool in order to run the calculations without using the program window
The Console is an applications that runs in a Windows Command Prompt window, and that runs the
Safeti consequence and risk calculations completely separate from the program window. The program
window and the operations of its various features consume memory, so using the Console maximises the
amount of memory available for the calculations.
The Console will run the calculations for all of the workspace files in a specified folder, which means that
you can run the entire analysis with a single command.
Once the calculations have been run in the Console, you can run either the Safeti program or the Study
Manager and view the results. The Console always save the calculated workspaces in the exported
format which means that you will be able to import the calculated workspace into the Study Manager
immediately after the Console has completed the calculations, and you do not have to open the
workspace in the Safeti program first in order to use Export Workspace to create a version of the
workspace in the exported format.
To use the Console, take the following steps:
1. Ensure you have local administration rights for the computer. If you do not, ask your IT
department to arrange this for you.
2. Create or choose a folder for storing the workspace files for the analysis. The path to the
workspaces folder must not contain any spaces.
3. Create or choose a folder for storing the consequence and risk results for the analysis. The
results folder must be different from the workspaces folder, and the path must not
contain any spaces.
4. For each workspace in the analysis, set up the input data in Safeti, and save the workspace to a
*.psux file in the folder for the analysis. The names of the *.psux files must not include any
spaces.
5. Exit from the Safeti program.
6. If you have made any changes to the administration files in setting up the data for the analysis,
you must copy the changed administration files from the folder that is used by the full Safeti
program to the folder that is used by the Console.
The full Safeti program works with administration files located in this folder:
C:\ProgramData\DNVGL\Safeti_Phast_Phoenix\Admin
whereas the Console uses files located in this folder:

C:\Program Files (x86)\DNVGL\Safeti_Phast_Phoenix\configuration

Before you copy the edited files from the ProgramData folder, you should create backup copies
of the original files under the Program Files folder. This will allow you to return to the original
files later, if required.
7. Locate the Windows Command Prompt

application. You can find it in the Start Menu by
searching for “cmd”, as shown in the illustration.
8. Run the Command Prompt.
9. Type in commands to move to the C: drive

and then to the folder for the Safeti program
files, as shown in the illustration.
The folder for the Safeti program files is C:\Program Files (x86)\DNVGL\Safeti_Phast_Phoenix.
As shown in the illustration, you can enter the path entirely in lower case if you wish, which
makes it easier to type.
10. Enter a command of the following form to run the Console:
phastconsole /pdrunrisk [path to workspace files folder] [path to results folder] l s
For example, if the path to the folder of workspace files is:
C:\apps\TestCases\Console
and the path to the results folder is
C:\apps\TestCases\Console\out
Then the command would be as follows:
phastconsole /pdrunrisk C:\app\TestCases\Console C:\apps\TestCases\Console\out l s
The illustration below shows the command for this example in the Command Prompt window.
After you press Enter to start the Console running, messages will appear in the Command
Prompt window about the process of starting the calculations. A separate Console window will
then open for each workspace file as it is run by the Console. These Console windows are similar
to Command Prompt windows, but the title bar contains a special version of the Safeti icon
and the title is the path to the PhastConsole application file. The Console window will close
automatically when the calculations for that workspace are complete.
The Console window displays the messages that would normally be written to the Output View in
the program window during the calculations. These messages are also written to a file with a
name of the form [workspace name]_log.txt in the workspace files folder. For example, if the
workspace file is ChlorinePlant.psux, the log file will be called ChlorinePlant_log.txt.

The calculations in the Console for a given workspace create a new *.psux file in the results
folder that has the same name as the original workspace, and that contains all of the input data
and also all of the results. The workspace files that are created in the results folder are in the
exported format that has the database of risk results inside the file itself. These are the format
of files that you can import into the Study Manager, which means that you will be able to import
the calculated workspace into the Study Manager immediately after the Console has completed
the calculations, and you do not have to open the workspace in the Safeti program first in order
to use Export Workspace to create a version of the workspace in the exported format.
11. Close the Command Prompt window when the calculations for all of the workspaces are complete
At the end of the calculations you will

see messages in the window about
creating a summary document and
processing difference summaries, and
these messages will include some red
error messages, as shown in the
illustration. The Console application
includes features for performing
regression testing (i.e. comparing the
results for the latest run with results
from a previous run), and these
messages are produced by the
regression testing features. These error messages do not indicate that a problem occurred in the
calculations for any of the workspaces, and you can always ignore them.
12. To view the results, you can either import them into the Study Manager from the calculated
workspaces in the results folder, or run the Safeti program and open the calculated workspaces.
In practice it is most convenient to save the commands to run the Console in *.bat files and execute the
*.bat files either from the Command Prompt window or directly using Windows Explorer.
Upgrade your SQL Server

If you run out of space on your SQL server you can upgrade the server. However, it is not likely that you
will run out of space, as the risk results are very efficiently stored and organized. However, if you do
suspect that you need to upgrade your server please contact our Technical Support Team.

5 ALERTS AND WORKAROUNDS
There are some known bugs that have not been fixed in v8.1, and some aspects of the program that
may give behaviour that you do not expect, e.g. because of minor enhancements or minor changes in
the modelling that may produce slight differences in the results.
For the most up-to-date list of known bugs and other issues, you can access the Customer Portal using
the link in the Help tab of the Ribbon Bar. To log into the portal you must enter a valid user name and
password, which you can obtain from your local DNV GL Software office.
5.1 Database brought in line with changes in password policies

As organisations take steps to improve data security, many are introducing password policies that
require long passwords that have particular features to increase the strength of the passwords. The
password for the risk database in Safeti has been changed to bring it into line with these developments
in more-stringent password policies.
One effect of this change is that Safeti 8.1 cannot be run on the same machine as Safeti 7.2 or Safeti
7.21 because the database passwords for the old versions are not compatible with Safeti 8.1. However,
Safeti 8.1 can be run on the same machine as Safeti 6.7 and also on the same machine as Safeti 7.22
which is a new version that has been produced specifically to address the new password requirements.
You can obtain Safeti 7.22 by using the Customer Portal or contacting Software Support, and if you want
to use it, we recommend that you install it before you install 8.1.
This change only affects Safeti. Phast does not use the risk database, and Phast 8.1 can be run on the
same machine as any other version of the program.
5.2 Known Bugs

1 B-16700 Time-varying discharge calculations may fail for a mixture
Description The discharge calculations may fail for a time-varying Scenario if the material is a
mixture, especially if the conditions are near the critical point.
The time-varying calculations are more likely to run successfully if you simplify the
mixture, e.g. to two components that have similar properties. If the calculations
still fail with the simplified mixture, you should perform the time-varying
calculations using a single pure component.
Non-time-varying calculations with the Leak or Short pipe Scenarios may run
successfully for the mixture, especially if you have a licence for the multi-
component modelling and select the multi-component option for the Equipment
item. With the initial release rate from these calculations and with time-varying
results for a simplified mixture or single component, you could make an estimate of
the time-varying behaviour and define this in a User-defined source Scenario.

2 B-16787 Time-varying short pipe Scenarios can take a long time to run or give oscillations in
the discharge results
Description Some Time-varying short pipe Scenarios experience problems with numerical
convergence which can give long run times and oscillations in the discharge
conditions. Pressurized liquid Scenarios are those most likely to experience these
problems.
Editing the Material and setting the Property method template to SRK-
IdealFugacity may improve the performance for such Scenarios.
3 D-11192 Scalloping in Max Footprint graph, depending on grid spacing
Description In some situations, the Max Footprint Dispersion Graph may show “scalloping”, i.e. a
“rippling” effect along the shape of the contours. The appearance of this effect
depends on the settings for the Concentration grid definition fields in the Dispersion
Parameters, and is most likely to happen when grid resolution is high.
4 D-11540 Upgraded standalone pool vaporization Models from v6 not included in any Model
Selection
Description If you upgrade a v6.54 or v6.7 file that includes a standalone Pool Vaporization
Model, the Model will be upgraded as a Spill Scenario under an AST Equipment item,
but it will be removed from the Model Selection which means that it will not be
included when you run the calculations, no matter which Run Row is selected.
The upgrade is handled in this way because the standalone Pool Vaporization Model
cannot be used in the risk calculations, but the Spill Scenario can. The Spill Scenario
will be incomplete on upgrade because there is no event frequency specified, and it
is removed from the Model Selection so that it does not cause problems with the risk
calculations.
You should decide whether or not you want to include the Scenario in the risk
calculations. If you decide not to include it but you do want to be able to run the
consequence calculations for it, you should create a separate Model Selection and
Run Row for it, and only run the consequences calculations for that Run Row.

5 D-11777 Mandation of plant boundary in run row grid
Description In the Run Row Grid, the Plant Boundary field is never shown as mandatory even
if Use free field modelling is set to Free field (plant boundary) in the Flammable
parameters. If a given Run Row has only toxic Scenarios selected or if all flammable
Scenarios are modelled with 100% probability of immediate ignition, then the risk
calculations will run successfully for the Run Row even if no plant boundary is
selected in the Run Row Grid, since the calculations will not involve the modelling of
free field delayed ignition outcomes. However, if the Run Row has flammable
Scenarios that are modelled with an immediate ignition probability less than 100%,
the risk calculations will give an error about a plant boundary not being specified.
6 D-11985 Risk Results not generated when Ribbon Bar is minimized
Description There is a minimize option at the right side of the Ribbon Bar, and if you turn this
option on, the Ribbon Bar will be “collapsed” so that only the names of the tabs are
visible. You have to click on the name of a particular tab to expand the Ribbon Bar
and view and use the options.
When the option is turned on, you will be able to view the list of Risk Results in the
Home tab of the Ribbon Bar, but nothing will happen when you click on one of the
types of results in the Risk Gallery. To view Risk Results, you must turn off the
option to minimize the Ribbon Bar.
7 D-12221 Reports flagged out of date when you change the averaging time for the graph
Description If you have both Reports and Graphs open for a Scenario, and you use the Edit
Settings dialog for the Graphs to change the selection of averaging time, you will
find that the Reports are shown as being out of date, even though the change to the
averaging time is only relevant to the Graphs.

8 D-12256 Limited support for long pipelines under Routes
Description In versions of the program before v7.2, it was possible to insert a Long pipeline
Model under a Route Model. This is no longer possible, but if you upgrade a file that
contains a Route Model with a Long pipeline item defined underneath it, the Long
pipeline item will not be deleted and you will be able to use it in the calculations.
However, some features of the program will not be available for a Long pipeline
under a Route Model, e.g. the Long pipeline will not be included if you export the
workspace data to Excel.
9 D-12286 Some functions are disabled for a Warehouse Model created by copying and pasting
Description If you create a Warehouse Model by copying and pasting another Warehouse
Model, some functions will not work for the second Warehouse Model:
 If the Warehouse Model is in a folder and you right-click on the folder and
select Generate warehouse scenarios, Fire Scenarios will not be generated
for the second Warehouse Model. However, if you right-click on the second
Warehouse Model itself and select Generate warehouse scenarios, the Fire
Scenarios will be generated.
 The Run option will be disabled for the second Warehouse Model, although
it will be enabled for the individual Fire Scenarios under the Model.
10 D-12335 Consequence results are cleared for a Long pipeline item when you change the failure
frequency data
Description If you run the consequence calculations for a Long pipeline and then change any of
the settings for the failure frequency in the Risk tab for the item, the results of the
consequence calculations will be cleared even though the values for frequency are
not relevant to the consequence calculations.
11 D-12362 Jet Fire Contour graph for a standalone Jet Fire Scenario shown as out of date after
checking option to Show Flame Shape
Description If the option to plot side view contours is selected for a standalone Jet Fire
Scenario, the option to Show Flame Shape will be enabled in the Consequence
tab of the Ribbon Bar when the Jet Fire Contour Graph is selected. If you check this
option, the Graphs will be shown incorrectly as being out of date.

12 D-12431 A User-defined Source Scenario generated from a Scenario with multi-rate time-
varying discharge results will give Jet Fire results that are different from those for the
original Scenario
Description The right-click menu for a source Scenario that has results present will include the
option to generate a User-defined Source Scenario. The discharge results for the
selected Scenario will be used as the definition of the release observers for the
User-defined Source Scenario.
If the original Scenario is a Time-varying leak Scenario, a Time-varying short pipe

Scenario or a Location specific breach Scenario, and the Method for calculating
average rate is set to Up to 10 rates in the Time varying releases tab section,
then if you compare the results between the original Scenario and the generated
User-defined Source Scenario, you will find that the results for dispersion, toxic
results and delayed flammable effects are identical, but that the jet fire results are
different. For the original Scenario, the jet fire calculations use discharge conditions
based on the full, detailed discharge results, whereas for the generated User-
defined Source Scenario, the calculations use discharge conditions based on the list
of release observers in the Scenario tab section. These discharge conditions can be
quite different between the two Scenarios, giving different jet fire results, especially
if the release rate changes rapidly at the start of the release. Typically, the
calculations for the User-defined Source will use a higher discharge rate and give a
larger jet fire with greater effect distances.
When you generate a User-defined Source Scenario from a Scenario that has
Method for calculating average rate set to Up to 10 rates, a warning message
about the differences in jet fire results will be written to the Output View.
13 D-12478 Results for a User-defined Source Scenario are not cleared on changing the input data
Description If you run the calculations for a User-defined Source Scenario and then change the
input data for the Scenario, the results will not be cleared and the Scenario will be
shown with a green tick in the Study Tree, showing that it has results that the
program regards as up to date. To be able to rerun the calculations for the Scenario
with the changed input data, you must right-click on the Scenario and select Clear
results.

14 D-12921 The headers in the Grid View might not show the correct units after a change to the
selection of units
Description If you have data displayed in the Grid View, and you open the Units of Measure
dialog and make a change to the selection of units (e.g. from feet to metres), you
will find when you return to the Grid View that the values displayed in the Grid
have been converted to the new units (e.g. they are the correct values in metres) ,
but you may find that the units stated in the header are still the previous units
(e.g. “ft” rather than “m”).
If you change the selection for the Grid Filter and then change back, the units
should be updated in the headers.
15 D-12946 Scenarios for a Long pipeline without valves will give calculation errors about
incomplete input data if you had previously defined valves but left the input data for
the valves incomplete
Description If you set the Number of valves to greater than zero in the Long pipe tab of the
dialog for a Long pipeline, leave the input data for all of the valves incomplete, and
then set the Number of valves to zero, the Scenarios for the Long pipeline will
give calculation errors about the data being incomplete.
In this situation you should set Number of valves to 1 or more, complete the
input data for all of the rows in the table of valves, and then set Number of
valves back to zero. The calculations will then run successfully, and will not model
the effect of any valves.
16 D-13089 The Normal boiling point can be edited for a mixture that is using the Phast64
property method template, and this will give inconsistencies in the modelling
Description If you set the template for a mixture to Phast64 and click on Calculate, the Normal
boiling point will be one of the fields in the list of properties that is enabled and
editable. However, if you change the value, you will find that inconsistencies occur
for Equipment items that have the mixture selected, both in the calculations of
phase and limits in the Materials tab, and in the consequence calculations for
Scenarios.
If you have edited the normal boiling point for any mixture, you should reset it to
the calculated value by editing the mixture and clicking on Calculate.

17 D-13170 The translucency used to display risk contours is the value set in the Risk Contour
Styles, not the values set for the individual risk levels in the Risk Preferences
Description The Contours tab of the Risk Preferences dialog shows the colours that will be
used for the risk contours for each level of risk and allows you to edit the colour.
The inputs for setting the colour includes an A field that is set by default to 100%,
which corresponds with zero translucency. If you reduce the value for A, the Risk
Preferences dialog will display a colour that has the appropriate level of
translucency applied, but the Risk Contour Plots will not use this level, and will
instead use the level that is set for Translucency in the Risk Contour Style dialog.
To open the Risk Contour Style dialog, select Risk (Initial) from the Styles gallery
in the Settings tab of the Ribbon Bar.
The settings in the Risk Contour Style dialog are applied to all contours, and the
default value for Translucency is set to zero. If you change the settings in the
dialog, you must generate a new Risk Contour Plot in order to see the effect of the
changes.
18 D-13218 For very low values of the Concentration of Interest, the Summary Report may give
values for the distance to the concentration even though it was not reaches
Description If the value set for the Concentration of Interest is 1 ppm or lower, then the
Summary Report may give values for the distance to reach this concentration, even
though this concentration was not in fact exceeded during the dispersion. The
program will report a distance to a concentration if the concentrations are within
either 1 ppm or 1% of the target value, whichever is smaller, and this can give
misleading results if the target value itself is of the order of 1 ppm.
19 D-13398 In some situations, changing the height of interest for dispersion graphs can cause
toxic graphs to be regenerated
Description If a Graph View has no results for the Toxic Footprint Graphs at the height of
interest set in the General Parameters, and you use the Edit Settings dialog for the
Graph View, both the dispersion graphs and the toxic graphs will be regenerated,
although the Height of Interest in the Context Panel will still show the value set in
the parameters. This behaviour is incorrect, as changing the height of interest in
the Edit Settings dialog should only affect dispersion graphs, and toxic graphs
should always display results at the height set in the parameters.

ABOUT DNV GL
Driven by our purpose of safeguarding life, property and the environment, DNV GL enables organizations
to advance the safety and sustainability of their business. We provide classification and technical
assurance along with software and independent expert advisory services to the maritime, oil and gas, and
energy industries. We also provide certification services to customers across a wide range of industries.
Operating in more than 100 countries, our 16,000 professionals are dedicated to helping our customers
make the world safer, smarter and greener.
DIGITAL SOLUTIONS
DNV GL is the world-leading provider of software for a safer, smarter and greener future in the energy,
process and maritime industries. Our solutions support a variety of business critical activities including
design and engineering, risk assessment, asset integrity and optimization, QHSE, and ship management.
Our worldwide presence facilitates a strong customer focus and efficient sharing of industry best practice
and standards.

Safeti 8.1 Release Notes

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Safeti 8.1 Release Notes

Uploaded by

Copyright:

Available Formats

RELEASE NOTES

Taking hazard and risk analysis one step further

Date: April 2018

Prepared by: DNV GL - Digital Solutions

© DNV GL AS. All rights reserved

2 NEW FEATURES APPLICABLE TO USERS OF THE FULL SAFETI PROGRAM ............................... 5

3 OTHER DIFFERENCES AND BUG FIXES .............................................................................. 9

4 PERFORMING A LARGE ANALYSIS ................................................................................... 23

5 ALERTS AND WORKAROUNDS ........................................................................................ 33

| RELEASE NOTE | Safeti version 8.1 Page i

1.1 The tree can show which items are

By default the option is turned on, and in this situation the

Other types of item will also be shown as selected if they are

Tab in Tree Type of node Nodes shown as selected

Building Building selected for Storage Equipment item or

| RELEASE NOTE | Safeti version 8.1 Page 2

The import process involves a series of dialog screens:

1. Selecting the data source.

2. Selecting the data tables to import

3. Applying an optional spatial filter

4. For Population only: Setting values for input fields

| RELEASE NOTE | Safeti version 8.1 Page 3

To open the Risk Preferences dialog, click on Risk Preferences in

| RELEASE NOTE | Safeti version 8.1 Page 4

2.1 Modelling of isolation and blowdown for time-varying

The modelling of isolation

The modelling of blowdown

If you set Blowdown? to Fails, blowdown will not be modelled.

| RELEASE NOTE | Safeti version 8.1 Page 5

If blowdown has been modelled for a

Note: if the Method for calculating

2.2 Consequence Data Results View with summary effect zone

| RELEASE NOTE | Safeti version 8.1 Page 6

The Run mode options in the Settings tab

The program behaves differently for each run method

| RELEASE NOTE | Safeti version 8.1 Page 7

The Runner window shows the status of the calculations with a

| RELEASE NOTE | Safeti version 8.1 Page 8

3.1 The release includes upgrades to the licensing software that

1. Stop the RMS Licence Service running.

If you do not upgrade the DNV GL Licence Manager,

If this dialog appears even after you have upgraded the

| RELEASE NOTE | Safeti version 8.1 Page 9

The “Phase to be released” field is no longer present in the Equipment dialogs

 Spill, for which the release phase is always liquid.

| RELEASE NOTE | Safeti version 8.1 Page 10

3.4 The area of the maximum footprint is given in the Cloud

| RELEASE NOTE | Safeti version 8.1 Page 11

| RELEASE NOTE | Safeti version 8.1 Page 12

Reports can be viewed for an individual Rationalised Discharge Scenario

| RELEASE NOTE | Safeti version 8.1 Page 13

In previous versions the radiation intensity was used for all

By default, Measure to use for critical radiation is set to

| RELEASE NOTE | Safeti version 8.1 Page 14

Setting for averaging method Calculation of representative discharge conditions

| RELEASE NOTE | Safeti version 8.1 Page 15

Elevation no longer an input for the User-defined pool source Scenario

| RELEASE NOTE | Safeti version 8.1 Page 16

3.10 Bug fixes

| RELEASE NOTE | Safeti version 8.1 Page 17

4 D-11858 Problem on changing Material when Material to Track was set

| RELEASE NOTE | Safeti version 8.1 Page 18

9 D-12591 No graphs available for the Standalone TNT explosion Scenario

10 D-12742 The option to save the workspace could stop working