Phast

Release Notes for V6.7

DNV SOFTWARE Palace House, 3 Cathedral Street, London SE19DE, UK http://www.dnv.com/software Copyright Det Norske Veritas. All Rights Reserved. No reproduction or broadcast of this material is permitted without the express written consent of DNV. Contact software.support@dnv.com for more information.

Phast v6.7 Release Notes

Phast v6.7 Release Notes

Contents
New Features and Major Enhancements...................................................... 3
Added output for toxics mixtures.................................................................................................. 3 Warehouse Fire Model.................................................................................................................. 3 Not enough material model enhancement................................................................................ 4 Mixture lethality contours enhancement.................................................................................. 6 ERPG equivalent mixture concentration contours enhancement............................................. 8 Rename K1/K2 liquids to ADR Class 3................................................................................... 9 Removal of ADR Class 3 from some fire fighting systems................................................... 10 User defined reaction rate for non PGS15 extinction systems............................................... 12 In-Building model improvements ............................................................................................... 14 Improved output report for discharge results .............................................................................. 15 New numerical solver for UDM outdoor dispersion model........................................................ 15 Improved correlation for initial droplet size ............................................................................... 16 Modelling of solid effects for CO2 .............................................................................................. 17 Materials Property update ........................................................................................................... 18 Update to toxic Probit values................................................................................................. 18 Pools on Water....................................................................................................................... 18 Modified Pool Fire Location in case of bunds ............................................................................ 19 External CAD connection wizard ............................................................................................... 20 Help menu additions ................................................................................................................... 23

Minor enhancements..................................................................................... 24 Workarounds and Bug Fixes........................................................................ 25

Known Issues, Alerts and Workarounds for v6.7 ....................................................................... 25 Bug Fixes for v6.7....................................................................................................................... 27 Discharge ............................................................................................................................... 27 Dispersion .............................................................................................................................. 27 GUI ........................................................................................................................................ 27

Making the Most of the Documentation...................................................... 29

Phast v6.7 Compiled Help........................................................................................................... 29 Phast v6.7 Release Notes ............................................................................................................ 29 Technical Reference.................................................................................................................... 29

For System Administrators: Technical Notes ............................................ 30

System Requirements.................................................................................................................. 30 Internet Protocol v6 (IPv6).......................................................................................................... 30 Temporary folders....................................................................................................................... 30

Managing Study Folders............................................................................... 31 Memory Management................................................................................... 32

Phast v6.7 Release Notes

New Features and Major Enhancements

This section describes the main new features in Phast v6.7

Added output for toxics mixtures

Phast v6.7 now generates additional output for mixtures containing toxic components: Graphs: o o ERPG concentration contours Mixture lethality contours in addition to component lethality contours

Report: component probit/lethality versus distance and mixture lethality versus distance

These new additional results are available for the Vessel/Pipe Source, User-defined Source and the Warehouse Fire Model scenarios. These enhancements are further detailed below for the specific case of the warehouse scenario.

Warehouse Fire Model

Phast v6.7 includes a number of modifications and enhancements to the Warehouse Fire Model. The enhancements include Extension to the model for instances when there is not enough material present for the specified fire duration, Enabling mixture lethality contours for warehouse fire models, Enabling ERPG equivalent mixture concentration contours.

The modifications include Renaming K1/K2 liquids to ADR Class 3, Removing ADR Class 3 materials from appropriate fire extinction systems, Enabling user specified reaction rate for non-PGS15 fire extinction systems.

The sections below describe these modifications and enhancements in further detail.

Phast v6.7 Release Notes

Not enough material model enhancement

In previous versions of Phast if there was not enough material present the warehouse scenario would fail and a message was displayed in the Message Log window as follows:
15:43:14 ----- Calculating Model Doors Open- 900 m2/1800 s for Weather Deelen, nacht:Deelen - F 1.5m/s... 15:43:14 ***Error: WHFIRE 4: Specified fire duration is larger than time 1031.11 s at which all warehouse materials are burned 15:43:14 *******Run Failed*******

In Phast v6.7 the model has been enhanced to reset the fire duration time to the actual duration it takes for all materials to be burnt and the model will now run successfully.
15:47:39 ----- Calculating Model Deelen, dag:Deelen - D 9.0m/s... Doors Open900 m2/1800 s for Weather

15:47:39 ---Info: WHFIRE 2001: Specified fire duration is larger than time 1031.11 s at which all warehouse materials are burned. So fire duration is reset to the time 1031.11 s at which all warehouse materials are burned

Phast v6.7 Release Notes

The message displayed will only be seen if display information messages and display warning messages are selected in the Miscellaneous tab from the Options>Preferences>General Preferences menu.

Phast v6.7 Release Notes

Mixture lethality contours enhancement

For each warehouse fire scenario, the warehouse fire model predicts the dispersion of the combustion mixture released from the warehouse. This mixture is modelled as a mixture of the toxic components NO2,, SO2 and HCl. It is labelled in the materials tab as the Warehouse Reference material. In previous versions, the following toxic graphs were produced: NO2,SO2,HCl separate probits versus distance NO2/SO2/HCl mixture dose/lethality versus distance NO2,SO2,HCl separate footprint contours for probit/dose/lethality In v6.7 additional toxic graphs are produced, so all the following combinations are now available: NO2,SO2,HCl separate probits/lethality versus distance NO2/SO2/HCl mixture dose/lethality versus distance NO2/SO2/HCl mixture footprint contours for probit/dose/lethality NO2,SO2,HCl separate footprint contours for probit/dose/lethality An example can be seen in the Outdoor Toxic Lethality Footprint below where version v6.6 displays the NO2 & HCl lethality footprint.

Phast v6.7 Release Notes

The v6.7 plot displays the Warehouse Reference mixture footprint in addition to the NO2 & HCl lethality

Please note that in the above screen shot the formatting of the plots has been adjusted from the default (right-click on legend and choose Properties) to help distinguish between overlapping lines.

Phast v6.7 Release Notes

ERPG equivalent mixture concentration contours enhancement

In Phast v6.7, users now have the option of specifying ERPG averaging time for the warehouse fire model. A new tab has been introduced to the Warehouse Fire model, the Location tab. This tab is analogous to the Location tab in the Vessel/Pipe source model.

Phast v6.7 Release Notes

Rename K1/K2 liquids to ADR Class 3

In previous versions of Phast the labels read K1/K2 liquids in the fire extinction systems. These have been renamed to ADR Class 3 all throughout the product. This has been carried out to be consistent with the PGS15 terminology as described in the Reference Manual BEVI Risk Assessments (formerly the Purple Book) The screen below shows the input dialog for the warehouse fire model. The help file has also been updated.

Phast v6.7 Release Notes

10

Removal of ADR Class 3 from some fire fighting systems

In previous versions of Phast, there was no logic preventing users from selecting synthetic packaging for K1/K2 liquids even though the Fire-fighting system would not involve these, according to the PGS15 methodology. This can be seen in the following screen where Fire-fighting system 2.1b Fire brigade mobilization<6 mins ADR cl 3 not stored in synth pack or no ADR cl 3 has been selected yet in the K1/K2 liquids input section the user can still select synthetic packaging.

Phast v6.7 Release Notes

11

PGS15 logic for allowing ADR Class 3 liquids has now been fully implemented. This logic is as follows: None allowed is only to be selected for system 3.1 Both synthetic and other packaging allowed are available for all systems except 2.1a,2.1b,2.2a,2.2b and 3.1 Synthetic Packaging allowed only is available for systems 2.1a, and 2.2a Other packaging allowed only is available for systems 2.1b, and 2.2b

The screen below shows that for Fire-fighting system 2.1b, the not stored & Stored in other packaging are now the only available options, as expected from the details above.

Phast v6.7 Release Notes

12

User defined reaction rate for non PGS15 extinction systems

In previous versions of Phast user defined reactions rates are only available to PGS-15 Fire-fighting systems. Once the PGS-15 option is unticked the option of user defined reaction rates is no longer available as can be seen in the Phast v6.54 image below.

In Phast v6.7, users now have availability of the user-defined reaction rate when PGS-15 Fire-fighting systems are not selected.

Phast v6.7 Release Notes

13

Users need to select either Stored in Other Packaging or Stored in Synthetic Packaging in order for the Reaction rate calculation field to be enabled. Then, the user defined calculation can be selected and the desired value may be entered, as can be seen below.

Phast v6.7 Release Notes

14

In-Building model improvements

The following enhancements have been implemented in Phast v6.7 to the in-building release model: In v6.7, the release duration is set so all liquid will eventually evaporate from the building for both instantaneous and continuous releases. This corrects an inconsistency previously present where liquid would not always evaporate for instantaneous releases. For continuous releases the release duration from a building can never be less than one air change. The release duration can never be more than the maximum release duration as set in the parameters (currently 1800 seconds).

For indoor releases, the vapor fraction multiplication factor remains in use (current default value of 3) and the material remaining is presumed to be trapped as liquid in the building. This multiplication factor determines the fraction of vapor that can be released from the building and can be modified in the Discharge parameters group. In addition to dispersion calculations, the in-building model also carries out explosion calculations, (which are not included in the risk calculations). Previous to v6.7, the venting parameter of the indoor confined explosion model was a constant. The venting parameter is now a variable based on the laminar burning velocity. Please note this correlation is only applicable for low-strength enclosures which cannot withhold an overpressure higher than 0.1 bar (National Fire Protection Association NFPA 68) for laminar burning velocity not higher than 0.6m/s. Currently, there is no implementation for high-strength enclosures; this is listed as a potential enhancement for a future version. For further information on the in-building release model theory, please refer to the technical documentation which is available with the Phast installation

Phast v6.7 Release Notes

15

Improved output report for discharge results

The layout of the discharge report produced a by Vessel or pipe source has been improved in order to produce a more clear overview of the output results. A clear distinction is made between stagnation data (data at upstream end for long pipe), orifice or pipe exit data before atmospheric expansion, and final data (after atmospheric expansion). See the screen dumps below for the case of a chlorine leak.

Improved 6.7 discharge report

New numerical solver for UDM outdoor dispersion model

The UDM Version 1 model is the model available in v6.54 and earlier versions, and the UDM Version 2 model is an enhanced model, introduced as an option in v6.6. The default in Phast v6.7 has changed to the UDM Version 2 model. The UDM Version 1 model was based on synchronised solution of two separate sets of differential equations for droplet and plume-dispersion variables. The UDM Version 2 model adopts a more rigorous approach by solving simultaneously a single set of droplet and plume-dispersion variables using a differential-algebraic solver. This eliminates previous erroneous droplet oscillations and allows accurate location of cloud transitions. For further information on the UDM Version 2 model theory, validation and verification, please refer to the technical documentation which is available with the installation

Phast v6.7 Release Notes

16

Improved correlation for initial droplet size

The Phast outdoor dispersion model (Unified Dispersion Model, UDM) requires as input an initial value of the averaged droplet size (Sauter Mean Diameter, SMD) to perform dispersion calculations. This initial droplet size is the droplet size following expansion of the release to the atmospheric pressure. Phast v6.54 made use of the 1994 CCPS droplet correlation to estimate this initial droplet size for continuous and instantaneous releases. The 1994 CCPS correlation involves both a mechanical and flashing break-up prediction of which the minimum is chosen as the droplet diameter. When compared with experimental data, obtained through a series of joint industry research projects where DNV was involved, the results showed that the 1994 CCPS model tended to under predict liquid droplet size for a range of materials, as shown in the figure below. 1994 CCPS Correlation

10000

Predicted SMD [microns]

1000

100

Cardiff_Cyclo-Hexane Cardiff_Water Cardiff_Gasoline Cardiff_Butane Cardiff_Propane INERIS_Butane VKI - R134-A STEP - Propane HSL - Propane INERIS - Butane Ecole de Mines - Water +30%-deviation -30%-deviation

10 10 100 Measured SMD [microns] 1000 10000

A modified CCPS droplet correlation has been developed and implemented in Phast v6.7 as the default choice for continuous releases. This modified CCPS correlation chooses the mechanical break-up droplet size for sub-cooled releases and the flashing droplet size for superheated releases, instead of choosing the minimum size as in the original CCPS correlation. Better predictions compared to experiments are obtained with the modified CCPS droplet correlation, as shown in the figure below.

Phast v6.7 Release Notes

17

Modified CCPS Correlation

10000

Predicted SMD [microns]

1000

100

10 10 100 Measured SMD [microns] 1000

Cardiff_Cyclo-Hexane Cardiff_Water Cardiff_Gasoline Cardiff_Butane Cardiff_Propane INERIS_Butane VKI - R134-A STEP - Propane HSL - Propane INERIS - Butane Ecole de Mines - Water +30%-deviation -30%-deviation

10000

Please note the modified CCPS droplet correlation is intended to be used in conjunction with the Unified Dispersion Model (UDM) Version 2, which is the new default UDM model in Phast v6.7. Additionally, it was also shown that the modified CCPS droplet correlation resulted in the most accurate UDM predictions of rainout. For further details, see the document Dropsize_theory_validation.pdf included as part of the Phast technical documentation.

Modelling of solid effects for CO2

For the most accurate CO2 atmospheric-expansion and dispersion calculations the UDM Version 2 contains a flag whereby it automatically switches to the equilibrium, no rainout dispersion parameter option for droplet modelling. There is a warning to this effect in the message log window and in the commentary report when this takes place, as shown below.
13:45:47 +++Warning: UDMA 1106: Solid CO2 modelling enabled, but droplet modelling not handled. Dispersion modelling will use equilibrium no rainout

The inclusion of CO2 as a Hazardous Substance in the Seveso Directive has been described in a HSE paper available from the HSE's web site (http://www.hse.gov.uk/seveso/co2-hazardous-substance.pdf). This paper states that the Phast CO2 model currently has the greatest potential to accurately model the release of CO2. For further information on the modelling, see also the files Thrm_theory.pdf and Xprp_Theory.pdf included as part of the technical documentation on the release DVD.

Phast v6.7 Release Notes

18

Materials Property update

In v6.7, there have been a number of updates to the property system. For further details than provided in this section, please refer to the Property_Database_Property_List.pdf document provided with the technical documentation The updates can be grouped as follows:

Update to toxic Probit values

In Phast v6.7 the A, B & N values for nitrogen (N2) and carbon dioxide (CO2) have been updated as shown on the table below.

Component Nitrogen (N2)

A -65.7

B 1

N 5.2

Carbon Dioxide (CO2) -90.778 1.01 8 The above values for nitrogen correspond to the probit function recommended in the Reference
Manual BEVI Risk Assessments (formerly Purple Book) for inert gases such as nitrogen.

The above values for carbon dioxide are derived from values recommended by the HSE for SLOT (1% probability of death), SLOD (50% probability of death) and N. The HSE SLOT DTL and SLOD DTL values for carbon dioxide are 1.5 x 1040 and 1.5 x 1041 respectively (units ppm for concentration c and minutes for time t), where the dangerous toxic load DTL is defined by

DTL =

N c dt

Please note that the above values are not used by default, since both nitrogen and carbon dioxide are flagged as inert materials. The materials Flammable/Toxic physical property setting needs to be set to Toxic to enable toxicity calculations (and associated risk). Previously, if using dangerous toxic load values greater than 3.4 x 1038 you receive an error during toxic effect calculations. The SLOT DTL and SLOD DTL values for Carbon Dioxide are 1.5 x 1040 and 1.5 x 1041 respectively. The dangerous toxic load model currently uses single precision floating point values which have a maximum definable positive value of 3.4 x 1038. Over time all of the DNV Software models are being updated to use double precision floating point numerical values but at this time the dangerous toxic load calculation is awaiting update. It is worth noting that this problem tends to exist for materials which are not strictly classified as toxic but are considered harmful at relatively large concentrations and exposure times based on their asphyxiation potential. Please note that when calculating toxic load using these toxic coefficients changes should be made to the model inputs as follow: 1. Change the Toxic parameters > Toxic Tab > Probit Calculation Method to 'Use Probit' 2. For SLOT use 1% lethality, for SLOD 50%

Pools on Water
The table below lists the material properties used in Phast for pools on water. These are all nonDIPPR physical properties. The last two properties in the table concern ammonia only, and for this case the properties are obtained from Raj and Reid as described in the PVAP theory manual. For the remaining first four properties a thorough literature search has been done and new values have been

Phast v6.7 Release Notes

19

implemented in Phast v6.7. For further details on the actual values please refer to the Property_Database_Property_List.pdf document provided with the software.

Property Material-water interfacial tension, Lw (N/m) Heat of solution, Hsol (J/kg) Solubility in water, fs (kg/kg) Heat transfer coefficient, hs (W/m K) Reaction with water (0, none, 1=Raj and Reid) Liquid-water enthalpy coefficients AL,w , BL,w, CL,w, DL,w
2

Affected Pool Scenarios Spreading of instantaneous releases on water Dissolution in water Dissolution in water Boiling on water Reaction with water (=1, for ammonia only) Reaction with water (ammonia only)

Modified Pool Fire Location in case of bunds

Phast v6.6 applies circular pools centered at the rain-out point, as seen in the image below. In reality the release location may be anywhere within the bund; if the release location is outside the bund, no bund should be specified.
Modifications to Pool Fire and UDM bund modelling
vapour-plume centre-line (always unaffected by bund)

bund

Bund wall = point of rainout = pool centre

Bund radius

If the release location is indeed at the centre of the bund, the formulation in previous versions of Phast may have given too conservative results; particularly for the cases of a large bund, rainout locations at or near the bund wall and formation of a large pool filling the entire bund. Phast v6.7 provides a modified less conservative method, which applies the assumption that the release point is at the centre of the bund both for the rainout criterion and the subsequent pool fire calculations.
vapour-plume centre-line (always unaffected by bund)

bund

Bund wall = point of rainout (pool spreads towards upwind edge of bund)
Bund radius

If rainout occurs outside the bund, this enhancement will have no effect. If the droplets hit the bund wall, rainout is assumed to occur inside the bund and the resulting pool is assumed to have spread back, in the upwind direction If rainout occurs within the bund, without hitting the wall, it is assumed that the pool will spread without bund effects until the downwind radius of the pool reaches the downwind radius of the bund and then spread back upwind, ultimately with the pool constrained by the bund centered at the release point.

Phast v6.7 Release Notes

20

The previous versions of Phast assumed the pool centre to be equal to the rainout location. This simplistic latter assumption has been maintained in Phast v6.7 for the purpose of the pool evaporation calculations and the pool vapour added back to the UDM cloud. However the early and late pool fires will apply the new, more correct location of the pool fire, with the downwind edge of the pool fires moved to the downwind bund wall (but reduction of radiation due to shielding by the bund wall is not considered).

External CAD connection wizard

In earlier versions of Phast users who wished to import CAD files had to go through a lengthy process to create a CAD Server Schema file (CSD file). Phast v6.7 has a new wizard that will create the CSD file in fewer steps. The CAD wizard can be accessed from the root node of the study file while on the map tab, then selecting the Input option and the CAD option will then be available, as shown in the following image

Phast v6.7 Release Notes

21

This will then bring up the wizard where you can select the type of CAD connection to be used and the file to be used.

Once the file is selected the wizard will create the CSD file and will include all layers available. As you can see below the location of the CSD file is displayed

Phast v6.7 Release Notes

22

Once the Finish button is pressed, the tree will show the new connection created.

Phast v6.7 Release Notes

23

Help menu additions

The Help menu now includes links to the Release Notes as well as the Technical Documentation for Phast as shown on the image below

Phast v6.7 Release Notes

24

Minor enhancements
Specification of breach location larger than the total pipe length will produce a fatal error, resulting in a red bordered Vessel/Pipe Source and no consequence results in v6.7 whereas the previous behavior was to automatically relocate the break location to the end of the pipe. This change applies to both long pipeline models: the vapor phase long pipeline scenarios and 2-phase long pipeline scenarios. In calculation of ambient temperature, now dry air specific heat instead of wet air is used in the code in line with UDM theory (Appendix A.3 in UDM theory manual). Particularly in cases which experience large elevation changes this may have some effect. This has less effect if UDM version 2 (default) is in use for dispersion calculations The default value for the parameter for maximum step size for instantaneous releases has been reduced from 1000s to 100s. This will affect results for instantaneous releases for step sizes larger than 100 seconds. This will also result in more smooth results in the far-field both for dispersion and subsequent hazardous effects particularly toxic results. Modified pool fire maximum burn rate mixing rule - incorrect mixing rule is applied in case a mixture contains both components with defined maximum burn rate and undefined maximum burn rate. In Phast v6.7 a mixtures maximum burn rate is now set to undefined when any of the components are undefined. An undefined mixing rate as input to the pool fire model POLF, means that POLF will calculate the burn rate itself. Corrected density calculations for jet fires. The saturated vapour density at atmospheric pressure is required for two-phase jet fire calculation. For some materials, particularly mixtures, liquid density could previously be erroneously returned as vapour density by the property calculation if the saturation temperature at atmospheric pressure is very low and this would subsequently lead to under-predictions of jet fire radiation. This is prevented now by using ideal gas density if no valid vapour density is found by the property calculation for a 2-phase discharge. If this occurs, the JFSH warning 1050 is given. Thus if this warning is given in Phast v6.7, increased hazards may occur for the jet fire compared to v6.6. Warehouse fires, with warehouse material including Br and F. The modelled combustion product released to the atmosphere consist of HCl, HBr and HF. Prior to v6.7, HBr and HF were classed together with HCl to conserve mass in line with previous CPR-15 regulations. To obtain more accurate toxicity calculations it is however more appropriate to conserve number of moles instead of mass. Thus in Phast v6.7, HBr and HF are classed together with HCl such as to conserve moles in line in more recent PGS-15 regulations. In case Br and/or F are present in the stored warehouse material, this will affect the mass fraction of HCl per unit mass of active warehouse material, the mass fraction of HCl in the modelled released combustion product (HCl, NO2, SO2 mixture), and the mass release rate of HCl/NO2/SO2 mixture.

Phast v6.7 Release Notes

25

Workarounds and Bug Fixes

Known Issues, Alerts and Workarounds for v6.7
Please find below a list of known issues, alerts and workarounds for v6.7. These comprise items that are behaving as designed though are not as intuitive as intended as well as issues related to external components where we have workarounds to suggest.

Feature Description

Alert/Workaround

SI10384: when a file created in versions previous to This can be remedied by saving the PSU file in v6.7 is opened and the map is viewed there may be a v6.7, closing the file and then reopening the file. GIS error on the message log and the map may be invisible SI10311: when exporting a report, the save as dialog The workaround is to click the save button and does not show a field to enter a name in Vista then rename the saved file to the desired name SI10150: When using external connections, the data If you get this message in the log window, please complete check does not spot if an External check the locations for the external connections Connection has moved from its original location. This and modify accordingly means that you can have an External Connection with Feature Classes showing as complete in the node tree, but not displaying in the GIS views. On opening a map or risk contour plot you get the following message in the log window:
"GIS Error: Unable to connect to the database. Please verify that your warehouse-connection parameters are correct and try again."

SI10536: Import of certain AutoCAD files results in The example.dwg import was created using the undesirable display of vectors zig zagging the object. new CAD import wizard (see page 20). This method imports all layers. To workaround the An example of such a drawing is: problem, use the manual CAD import method and exclude the problematic layer when defining feature classes. Alternatively, remove the layer in AutoCAD.

Phast v6.7 Release Notes

26

Feature Description

Alert/Workaround

This problem does not occur for Vessel/Pipe and SI10525: It has been discovered at a late stage of the User defined sources which do not use case lists. Phast 6.7 development project that a bug is present in the area of producing graphs of radiation lethality from a jet fire resulting from a vessel/pipe and user defined There are two possible workarounds for this issue: source release when using Case Lists. The first workaround is to convert sources that use caselists into a set of independent sources. This can be done by right clicking a case list and choosing "Convert all caselists to models". From the resulting models, it is possible to produce lethality graphs. If you also switch on the calculation of Dose and Probit, this overcomes the issue and lethality plots can be viewed from caselist results.

The problem arises if the Vessel/Pipe or User Defined Source has a case list in use and you choose to calculate lethality results.

Unfortunately, for the above situation lethality ellipses are not plotted when graphs are generated. An issue has been discovered relating to the use of WMS GIS data sources in the map in Phast and Phast Risk. Import of some map imagery can result in illegible images. For example, a map which should appear in this way: There is no workaround to this issue at this time but will be addressed in an update of the 3rd part GIS mapping software Geomedia by Integraph in a future release of Phast and Phast Risk.

Can only be displayed in the following unusable style:

Phast v6.7 Release Notes

27

Bug Fixes for v6.7

There are several bug fixes to the modelling and the user-interface for Phast v6.7. For the most up-to-date list of known bugs and other issues, see the Technical Support area of the DNV Software website: http://www.dnvsoftware.com. In order to access the support information, you must enter a valid user name and password, which you can obtain from your local DNV Software office.

Discharge
10044 Premature termination for time-varying vapour line ruptures In earlier versions there were cases of premature termination of time-varying line rupture cases. In v6.7 this has now been largely addressed for cases where the storage state is pure vapour. For cases that did not depressurize fully, the release duration, expelled mass and hence risk may increase in v6.7. However, results are not expected to change for cases which previously depressurized successfully. Discharge from vessels and short pipes (leak, line rupture, relief valve scenarios) DISC warning 1005 converted to DISC error 32. The error is now provided in case entropy is not conserved for property calculations associated with choked orifice flow. This appears to be mostly caused by picking up incorrect vapour density instead of liquid density and vice versa. This is now an error because previous results were shown to be physically incorrect by a large margin. To model cases which show this error it will be necessary to adjust inputs and/or use the user-defined source model.

J415

Dispersion
2110 UDM instantaneous dispersion taking excessively large downwind steps Very large steps taken by instantaneous dispersion model results in discontinuities in the toxic and concentration results. This has been resolved in reducing the default maximum step size for instantaneous dispersion calculations from 1000 seconds to 100 seconds. Droplet trajectories for pressurised instantaneous releases These were incorrectly calculated, resulting in rainout too far downwind. This would result in clouds leaving the pool behind too soon.

10365

GUI
8812 Problems connecting to some GIS database formats Previously there were problems connecting to GML, WCS and WMS sources.

9128

Warehouse can be deselected when changing stop criterion

In v6.54 if you switch the dispersion stop criterion from 'mixed basis' to 'risk based' then it can change the model selection. This can result in changes in the risk results and may not be noticed by the user so they finish with incorrect results.

9343

Warehouse model - Coordinates do not change after "move" button Previously, if a user tries to move the warehouse model in the graph by using the "move" button, it visually changes location on the map, but when opening the model > geometry tab, the coordinates are still the same original values. Results are also given with the old location so the risk results may be wrong without the use noticing.

Phast v6.7 Release Notes

28

9838

Not possible to import multiple materials into a PSU Previously multiple weathers and models can be exported and re-imported into a different PSU, but this is not possible with materials (each has to be imported as a separate PSU file). Thus it is not possible to transfer multiple materials defined in a study (not in the matls.xpu file) to another study: they need to be transferred one at a time and this can be a slow process if there are many materials. Flammable Mass in the v6.6 fireball report is not the mass actually used in the model For v6.7, in the Fireball report there are now two fields: Potential Fireball Mass (total mass in vessel), and Actual Flammable Mass (mass used for fire-ball calculations, three times vapour fraction (if < 100%). Negative areas for dose, probit, lethality contours The area calculation could result in a ve or +ve value in the legend depending on the order of the points. Now only +ve values are calculated and reported. The Arcview DLL doesn't work in v6.54 (or v6.6) Corrected in v6.7 The flammable lethality zone on the Map tab in graphs is incorrectly calculated. Changing the lethality on the jetfire reporting levels grid had no effect on the jetfire lethality Effect zone on the map. The Pool fire report included the radiative fraction for general fires even if not used Now the correct radiative fraction is reported (as output by the pool fire calculations) for all types of fires (general, luminous, smoky fires). This only affects the pool fire report and does not otherwise affect the results. Save as with results not working properly with psc format with memory saving on The supplementary results files we not being written the results folder as required by the minimise memory option. Warehouse fire scenario remains blue while copying to other study Because the parameters may vary between studies then the results must be regenerated in the new study and not copied from the existing run of the model. Save as overwrites existing file on Vista and Windows 7 with file extensions on It was easy to think you were saving to a new file and instead the existing file was being overwritten so your original data was lost. The problem only occurred on Windows Vista and version 7 when the user had activated the display of file extensions (non-default).

10225

10227

10235

10290

10379

10440

10469

10513

Phast v6.7 Release Notes

29

Making the Most of the Documentation

The Documentation folder on the Phast v6.7 CD contains a wide range of information on the program, and should enable you to find answers to most of the questions you may have about Phast. Most of the information is supplied in the form of password-protected *.PDF files. You can obtain the password for these files by contacting DNV Software Technical Support at Software.Support@dnv.com The contents of the folder are described below, in the order in which they appear inside the folder.

Phast v6.7 Compiled Help

The Online Help for Phast v6.7 is in the form of a *.chm filea compiled HTML Help file. The file will be installed with the rest of the Phast files when you install the program, and you will normally access it using the Help menu, but you can also open and read the file directly. There is a separate version of the Help for each product. In order to read the files, you must have Microsoft Internet Explorer Version 5 or later installed on your PC, although it does not have to be your default browser. You must also have some additional files that are specific to Microsoft HTML Help; these are installed automatically during the Phast setup process, so you should be able to read the files directly if you have already installed Phast.

Phast v6.7 Release Notes

These Release Notes are supplied in *.PDF form, for future reference.

Technical Reference
The Technical Reference subfolder contains *.PDF documents which give details of several aspects of the modelling. These details include verification and validation of the models. In Phast v6.7 this documentation in installed and is accessible from the Help menu in the program.

Phast v6.7 Release Notes

30

For System Administrators: Technical Notes

These notes give technical details to help you improve the performance of Phast v6.7 on your system. They are aimed mainly at secondary support personnel (e.g. IT professionals), but may be useful for expert users.

System Requirements
Operating system Internet Browser Display Screen depth CPU Media Pointing device Disk Space Spreadsheet . CPU Speed Recommended Memory (minimum) Phast 2.8+GHz 3GB (2GB) Windows XP SP3, Windows Server 2003, Windows Server 2008, Microsoft Vista, Windows7 IE 5.0x. This does not have to be set as your default browser, but you must have at least this version installed in order to run the program. 800x600 resolution color 65536 (16-bit) and higher 32 or 64 bit multi-core CPU from Intel or ADM like Intel Core2 Duo DVD-ROM drive or internet connection MS Mouse or compatible The installation requires 285 MB of disk space The program is capable of using Excel spreadsheets for input/output of data

These values are based on the user experiencing acceptable performance. Depending on your needs you may consider to select a more powerful machine. The indicated required disk space is very dependent on the existing system. Our values are based on a clean operating system with no additional software.

Internet Protocol v6 (IPv6)

Testing of v6.7 has shown that network licenses set up with client machines using IPv6 are not supported when remotely connecting to the network. A new version built with IPv6 libraries can be made available on request.

Temporary folders
The program uses hard disk space to write temporary, or scratch files which are subsequently deleted once the run has completed or in some cases when the program has been closed down. By default the program uses the subfolder C:\DNVUser\Temp to write temporary files during the calculation of results, and also to write temporary files that are used during the reporting of results. For large Study Folders, these temporary files can require a lot of hard disk space. Phast will crash if there is insufficient hard disk space for it to create the temporary files, so you should save your Study Folder before running calculations or viewing results, and avoid running Phast if the free space on your hard disk is low

Phast v6.7 Release Notes

31

Phast uses a third party product called Seagate Crystal Reports to produce text reports. Crystal Reports always uses the temporary folder set up by your operating system for its own scratch file space, and by default this is generally the C:\Temp folder.

Managing Study Folders

Some general points: You should always save your Study Folder before you run any calculations. This will ensure that you do not lose any changes that you have made to the input data in the unlikely event that a condition emerges in the calculations that means that the program must be terminated. It is unwise to give the same name to more than one Model in a given Study Folder, Phast will allow you to do this, but you will find it difficult to distinguish between the results for the Models in Reports and Graphs, and it is not good practice. You should use only letters, numbers and underscores (_) in the names of Study Folder files, and should not use special characters such as commas or quotation marks.

Phast v6.7 Release Notes

32

Memory Management
Phast can use a lot of memory (RAM) when running very large studies/study folders, and also when viewing large combinations of text reports and graphs. You should follow these simple rules to avoid problems with memory: Close all out-of-date Reports and Graphs if you are no longer using them. You can always reopen the Report or Graph for the Model if you need to view it again; if the name of the Model is displayed in blue in the Study Tree, then that Model has results that are available for viewing. Use the option in the Results tab of the General Preferences to save results to files instead of storing them in memory. Reports and Graphs will take longer to generate if you choose this option, but it will reduce the memory requirements considerably. For very large studies (for example those containing more than 100 models and 3 associated weathers) ensure before you run that you have sufficient hard disk space for temporary file storage and also for runtime memory requirements.

Runtime memory is memory which will be used by Phast for storing results as it performs its calculations; it is a combination of fixed RAM plus virtual (hard disk substitute) memory. You can adjust the total available memory size using the System option in the Control Panel. Consult your IT professional before altering any of these settings. During calculations for a very large study, you may see an "out of memory" message (or similar) appear on the screen. If this occurs it is important you stop the calculations immediately and save your work. If you do not stop the calculations, Phast may crash, and you will lose any changes you made to your Study Folder since the last time you saved your work. The process of generating Graphs can sometimes require a large amount of memory (especially graphics memory). Models which involve the rainout and evaporation of a flammable material can produce a particularly large number of data points that need to be plotted, with large demands on the Graphing function. If you use the data culling option (on the Graphs tab section of the Preferences dialog, which is under the Options menu), then the Graphing function will ignore unnecessary data points (i.e. data points which do not describe significant changes in the behavior), and will therefore save memory and time. See the online Help for more details of data culling. It has not been explicitly mentioned that the application is single-thread software and increasing cores or CPUs will not improve calculation. However, multi-cores would benefit the overall user experience as the OS, Outlook, or other applications can use the other system capacities. In this case, the bottleneck may be the RAM or disk read/write speed (local or remote).