//Home/Support/Questions & Answers

 

Support Services

Questions & Answers
Known Problems
How Toīs
White Papers
Notes RSUM
Fixes

 
Questions & Answers

 
Installation:
I get an error message saying that there is something wrong with my Sentinel Key.
 
Project management:
Where can I find failure data for my project?
I canīt open projects copied from a CD onto my hard disk.
Why do the project files take up so much disk space?
Can I use long file names for RiskSpectrum projects?
Can I use extended characters (e.g.. # & %) or spaces in the name of a RiskSpectrum project?
 
Editing:
Why can't I undo changes?
How can I make use of Drag-and-Drop features when editing fault- and event-trees?
What rules are applied when using function event alternatives?
How do I copy a branch in a fault tree?
 
MCS analysis:
Why does the results differ between a consequence analysis case and a MCS analysis case
Why do I get a warning during MCS analysis saying that I have inconsistent CCF groups defined in the project?
What is a MCS analysis case?
How do I use a MCS analysis case?
How do I create a MCS post-processing specification?
What definition of a house events state overrides another when running an analysis?
Which values does the program use for CCF events with order higher than 3, if I define a CCF group with 5 elements in it?
The probability values generated by the program for CCF events seem to be incorrect. What can be wrong?

-------------------------------------------------------------------------------------------------------------


Questions and Answers - Installation:

Q: I get an error message saying that there is something wrong with my Sentinel Key.
 
A: Before you contact support, please consider the following:

RiskSpectrum is protected using Sentinel keys. The Sentinel keys requires installation of drivers. The drivers are installed when you install RiskSpectrum. You can download drivers here.

Laptops:

The parallel port on Laptops sometimes needs to be activated in order to recognise the Sentinel Key. You activate the port by using the Bios Setup on your computer.

Also, make sure to install RiskSpectrum FT or PSA Professional or update drivers when your laptop is un-docked. Otherwise RiskSpectrum will not find the key when you are undocked.
The parallel port on Laptops sometimes needs to be activated in order to recognise the Sentinel Key. You activate the port by using the Bios Setup on your computer.

Make sure to consult your user manual for your computer before you make any changes to the Bios setup.

You should set it so that the port is activated at all times. Auto detection or similar is not sufficient e.g. it is not recommended to allow the operating system to choose when to activate the port.

Back to top


Questions and Answers - Project management:

Q: Where can I find failure data for my project?
 
A:
  • The T-book (Nordic Nuclear Power Plants). Please contact svenne.skagerman@swedpower.vattenfall.se
  • Spanish database for reliability data collection and Maintenance Rule implementation (BDATA)
  • Offshore Reliability Data (OREDA), http://www.dnv.com/technologyservices/handbooks/index.asp
  • Guidelines for Process Equipment Reliability Data, Centre for Chemical Process Safety of the American Institute of Chemical Engineers
  • Reliability Data, Centralized Data Base (ZEDB), Germany
  • IEEE Guide To The Collection And Presentation Of Electrical, Electronic, Sensing Component, And Mechanical Equipment Reliability Data For Nuclear Power Generating Stations, Published by the Institute of Electrical and Electronics Engineers, Inc.
  • OECD on its behalf is setting up a database relative to the failure of digital I&C namely Computer-Based Systems Database (COMPSIS)
  • Common-Cause Failure Database and Analysis System developed by USNRC and INEEL (NUREG/CR-6268)
  • International Common Cause Data Exchange (ICDE) by SKI, USNRC and OECD
  • NUREG/CR-5497
 
Back to top
   
Q: I canīt open projects copied from a CD onto my hard disk.
 
A: Remove the read-only attribute from the project file. All files that are copied from a CD are marked as read-only. To be able to work with the project, the RiskSpectrum program must be able to write to the project file. Since it can not open the file in write mode, it fails.
 
Back to top
   
   
Q: Why do the project files take up so much disk space?
 
A: By default, the program will calculate and save up to 100 000 MCS for a single MCS analysis. If you are not interested in saving such large MCS lists you can decrease the number of MCS to be saved. This is done in the MCS analysis specification. The size of your result files will decrease accordingly after you re-run your analyses.
 
Back to top
 
   
Q: Can I use long file names for RiskSpectrum projects?
 
A: Since RiskSpectrum is a 16-bit application, it can not recognise long file names. You should also avoid using long directory names for directories where RiskSpectrum projects are located.
 
Back to top
 
   
Q: Can I use extended characters (e.g.. # & %) or spaces in the name of a RiskSpectrum project?
 
A: We recommend only using characters A - Z, digits or underscore character ('_') in the name of a RiskSpectrum project.

Back to top


Questions and Answers - Editing:

Q: Why can I not undo changes?
 
A: Editing Tables:
As you type changes in a record in a table you will see a small "pen" in the far-left column. As long as the "pen" is displayed you can undo the changes you have made by pressing the esc-button. Once you have moved the cursor away from the field you are editing, the “pen” disappears and the changes have been stored into the project file. At this point you are not able to undo the change.

Editing in Dialogs:
Any changes you make to list boxes are stored directly in the project file and can not be cancelled.

Editing of fault- and event tree:
All changes you make in fault trees or event trees graphical structure are saved when you close the window.

Work-around: In the options dialog on the Tools- menu you can select to "Make backup on Open" and "Confirm on Exit". This means that the program will make a backup copy of the project you open and prompt you to confirm to save your changes when you close it. If you press the "No" button all the changes you have made to the project since you opened it are discarded.

 
Back to top
 
   
Q: How can I make use of Drag-and-Drop features when editing fault- and event trees?
 
A: The Record list-Window (activated, for example, by pressing Shift-F2 keys) shows a list of basic events, transfer gates or house events when editing fault trees and initiating events or function events when editing event trees.
From the Record list-Window you can drag-and-drop elements into your fault- or event tree.
For example, you can insert a basic event into a fault tree by opening a fault tree and the Record list-Window, choose a basic event and drag it onto a gate. When you drop it, it will be added as an input to the gate. You also have the possibility to filter the list in the Record list-Window.
   
Back to top
   
   
Q: What rules are applied when using function event alternatives?
 
A: If no function event alternative is specified in a branch point in a sequence, the "success" of the first alternative (input No. 1) will be used for the top-most sequence and the "failure" of the same alternative will be used for calculating sequences of "failing" branches.
If no function event alternative is specified for the top-most sequence in a branch-point, it will be treated as the successful branch for which complemented logic will be used. This means that the program will calculate the top-most sequence as the complementary event of all the function event alternatives used in the "failing" branches).
If you do define a function event alternative for the top-most sequence in a branch-point, it will be treated as "failed", unless there is only one "failing" branch and it "uses" the same function event alternative.
   
Back to top
   
   
Q: How do I copy a branch in a fault tree?
 
A: Select the branch and copy it. Paste it under a gate of your choice.
It is not possible to copy only the inputs to a gate.

Back to top


Questions and Answers- MCS Analysis:

Q: Why does the results differ between a consequence analysis case and a MCS analysis case?
 
A:
There are two reasons for this:

1. The MCS analysis case uses the demodularised cutset-list to calculate the top result. When running a consequence analysis case, all calculations are made using a modularised cutset-list.

2. Events with a failure probability of 1 may also cause differences in the results. An enhanced treatment of Q=1 events has been introduced for the normal analysis case (e.g. consequence analysis case). In the process of the calculation of the top event, basic events with a failure probability of 1,0 are temporarily set to TRUE and the cutset list is minimised before the program calculates the top event. When the top result has been calculated the Q=1 events are restored and saved in the MCS lists. The MCS analysis case operation will be based on the MCS list where Q=1 events are included and that is why the results may differ. We will investigate the possibility of using the enhanced treatment of Q=1 events also for the MCS analysis case.

   
Back to top
   
   
Q: Why do I get a warning during MCS analysis saying that I have inconsistent CCF groups defined in the project?
 
A: If you define more CCF parameters than it is necessary for a CCF group, the program will warn you.
For example, if you define all three alpha factors for a CCF group of only two basic events (definition of only one alpha factor - alpha2 - is sufficient in this case) you will be given a warning saying that the CCF group has more CCF parameters than necessary.
It is not an error and it does not affect the correctness of the MCS results.
   
Back to top
   
   
Q: What is a MCS analysis case?
 
A: A MCS analysis case is used to merge (using logical OR operation) MCS lists produced for analysis cases.
For example, you can merge MCS lists produced for different sequence analysis cases. The MCS lists are minimised during the merge (all duplicate and non-minimal MCS are be removed).
 
Back to top
 
   
Q: How do I use a MCS analysis case?
 
A: a MCS analysis case is used for merging MCS lists. In addition, you can perform actions on the resulting MCS list after the merge (so called post-processing of MCS list).
For example, you can change cutsets by adding or removing events or remove some cutsets from the list. The post-processing actions are defined in the MCS post-processing specification.
 
Back to top
 
   
Q: How do I create a MCS post-processing specification?
 
A: In the MCS post-processing specification you can define what actions should be taken by the program to modify MCSs after the merging of MCS analysis cases.
An action consists of one Type field, where you specify the action type, and five fields, where you can define filter masks. You can insert or delete basic events in MCS or the MCS itself. You can control the scope of the rules by setting a filter for MCS.
If you create the action type: "Include rule" and the filter mask: "*", all defined actions will be applied to the whole MCS list. You can limit the scope by setting the filter mask for the "Include rule" action to some other value.
For example, if you specify the filter mask, "A*" the rules will be applied on MCS containing at least one event with ID starting with the letter A). If you specify more than one ID filter, the actions are applied on MCS matching all filters, only.
You can prevent actions to be applied on MCS by creating an action with an "Exclude rule" action type. Such actions are defined in the same manner as "Include rule" actions.
   
Back to top
   
   
Q: What definition of a house events state overrides another when running an analysis?
 
A: A house events state, specified in a boundary condition set overrides that of the house events definition.
 
Back to top
 
   
Q: Which values does the program use for CCF events with order higher than 3, if I define a CCF group with 5 elements in it?
 
A: Since you only can give up to three CCF parameters for a CCF group regardless of its size, the program assumes zero probability for all CCF events with order higher than 3.
For example, for a CCF group of 5 basic events all CCF events of order 4 (representing simultaneous failure of four basic events in the group) are assigned the probability 0. This means that your CCF parameter for order 4 should include all failure data for 4 or more failing components. This is a reasonable assumption since if 4 out of let's say 10 fail, it is fair to assume that all fail.
 
Back to top
 
   
   
Q: The probability values (generated by the program) for CCF events seems to be incorrect. What can be wrong?
 
A: The first thing to do is to refresh CCF events data by opening the CCF group list window and choose the "Refresh CCFs" command from the Record menu. The program will re-build CCF events according to the current definition of the CCF group and the basic events in it.
When you create a CCF group, the program creates CCF events based on the information from the basic events members of the group.
If you change a basic event belonging to a CCF-group (for example, by changing the event's parameters) the CCF-group needs to be "refreshed" in order to update it's CCF-events.
 
Back to top