Wiki

Options

New Case Case Status
Log In

Wiki

Options

 
Knowledge Base Articles»KB100031 - APR Fault Recogniti…
  • RSS Feed

Last modified on 10/17/2019 1:39 PM by User.

Tags:

KB100031 - APR Fault Recognition Documentation

Concept of Fault Detection using APR and Fault Signatures

  • APR detects abnormal behavior or a SYMPTOM at the sensor level.
  • A SYMPTOM is an observable change to a variable caused by the FAULT.
  • A FAULT is a malfunction or incorrect operation of a piece of equipment or sensor(s).
  • SYMPTOMS from two or more sensors can be used to identify a single FAULT
    • Example:  Consider the sealing water from a pump.
    • Increased seal leak out flow and increased seal leak out temperature can be used together to confirm that a pump seal failure fault is occurring.
    • APR is used to detect the abnormal increases in flow and temperature.
    • The APR indicators form a pattern or a signature that points to a FAULT.
  • There are three possible APR states
    • HIGH – Higher than normal, above the gray band
    • NORMAL – Within the normal gray band
    • LOW – Lower than normal, below the gray band
  • Fault Example: Detection of a boiler feed pump seal failure
    • Instrumentation
      • In Flow
        • Hot water to the inner seal is measured as the pump suction temperature
        • Cool injected water temperature and flow are measured
      • Out Flow from the outer seal (leak off)
        • Flow and temperature measured
    • FAULT
      • When the inner seal begins to fail it allows more hot water to mix with the injected cool water. This causes the symptoms of increased leak off flow and increased temperature.
    • FAULT RECOGNITION = Failing seal  which produces more flow at higher temperature
      • Seal out flow - HIGH
      • Seal out flow temperature - HIGH
      • Seal inlet temperature and flow - NORMAL
  • One sensor detecting abnormal behavior often points to several possible FAULTS.
    • Abnormally high Boiler feed pump seal out flow can indicate:
      • The beginning of seal failure
      • Increased injection water flow
      • Failure of the flow measurement
    • Use of more symptoms provides better confirmation of the fault.
    • Quite often, there are fewer sensors available than a user might desire.

Three states (high, normal, low) of two sensors (seal leakoff temperature and flow) of the above example form seven possible faults. The following table demonstrates the concept of a FAULT GROUP. This group of points (sensors or calculated points) which can be used to distinguish between multiple faults, is a truth table. A signature based on two sensors is shown in the following table:

FAULT GROUP

Assuming Inlet Temperature and Flow are normal

 

FAULT SIGNATURE

FAULT

Order of likelihood

Out Flow

Out Flow Temp

NONE

1

Normal

Normal

Failure of flow sensor

2

Low

Normal

Failure of  temperature sensor

3

Normal

Low

Initial failure of seal (1 of 2 indicators)

4

High

Normal

Initial failure of seal (1 of 2 indicators)

4

Normal

High

Confirmed failure of seal (2 of 2 indicators)

4

High

High

Failure of seal and bad temperature sensor (low probability)

5

High

Low

Failure of both flow or temperature sensors  (low probability)

6

Low

Low

Creating a Fault Signatures in Asset Editor

Overview of Creating a Faults

Creating faults in a group which uses similar sensors allows comparison of the distinctions between the different fault signatures. After opening an asset in Asset Editor the faults will be created on the Faults tab. There are two approaches to creating a Fault Group, local and global. A Fault Group can be created for the existing asset only.  This Fault Group is local, based on the points in the asset and is for the specific asset only. A Fault Group Template can be created that can be applied to similar assets in a fleet. The Fault Template will be referred to as a Fault Template. Such a Fault Template is global, based on generic points and can be imported into a fleet of similar assets which are attached to generic points. An example of a generic point would be “Compressor Discharge Pressure”. This point would be associated with the specific points, GT1 CDP in a first asset, GT2 CDP in a second asset, and GT3 CDP in a third.  A Fault Template using point “Compressor Discharge Pressure” would then find the associated asset specific point in each asset.

Fault Group Templates can be created in two ways. Generic points can be selected into a truth table and saved as a Fault Group Template. Alternatively an asset specific Fault Group truth table can be created using asset specific points. Generic points can then later be connected to the asset specific points. With connected Generic points the asset specific Fault Group can then be converted to a Fault Group Template.

Creating Faults Local to the Asset

In Asset Editor open an asset and go to the Faults tab.

 

Create a Fault Group

The first step in creating a fault is to create a fault group which will contain one of more faults. Clicking the New button allows the user to enter a name for the fault group and add the points to the group that will be indicators of the faults. The Edit button will allow the user to edit contents of an existing fault group.

Selecting Sensors (Points)

Click Add Points. Points are selected from the points in the asset.

Select points and press OK to view the Truth Table in Fault Builder. Asset specific points are selected to go into the truth table.

Setting the Fault Signature in the Truth Table

Each row in the truth table should correspond to a different fault.

Each fault needs the following:

·        Fault name

·        State for each point: No Concern(0), High Concern(1), Low Concern(-1), Not Used( )

·        Any notes associated with the fault. This could be actions to take when the fault occurs.

·        Any Web Link to a learning web page. Typically learning web pages will be put under the EtaPRO web directory.

·        Whether the Fault is in service. If in service, the Fault will evaluate whenever the asset runs.

Fault Templates for Global use

A user can create templates in a Fault Library so that they can be used in a fleet of similar assets. Fault Group Templates can be built based on asset specific fault groups or they can be built using generic tags. To build using generic points, click on Fault Library button to open the library. Right click to add a Category then right click on the selected category to add a Fault Group Template.

Right clicking allows the user:

·        Add a Category. There is only one level of Category.

·        Add a Fault Template. A Fault Template can contain a number of faults.

·        Edit a Category or Fault Template

·        Delete a Category or Fault Template

Creating a Fault Template for Global use

A Fault Template looks the similar to a local Fault Group truth table. The difference is the points. The local Fault Group uses asset specific points. The Fault Template selects points from a list of generic tags. Click Add Tags to get the list of available generic points. Initially this list is empty. Generic entry items can be added by clicking the Edit Tags button which allows, deletion, addition or editing of the generic tags. These generic points must be created prior to the creation of the Fault Template.

They will then be connected to asset specific points using the GenericID column in the Points Tab.

Once generic points have been created, click Fault Library. Select any Category and right click Add Fault Template. Then Click Add Tags. Select the generic points to be used in a Fault Template and click Ok.

The Generic Points are added as columns of the truth Table in a fashion similar to the asset specific fault signature.

Importing Templates into Specific Assets

A Fault Template can be imported into a specific asset and used to create the local asset specific faults that will be saved and executed within that specific asset.  To do this, populate the GenericID column in the Points tab with points that match the generic points in the Fault Template.

Next, click the New button to create a new Fault Group.

Then click, Open Fault Template to get a list of available Templates and click on the desired template.

The Fault Template uses the Generic ID to connect to asset specific tags and create an asset specific Fault Group which will get saved with the asset. The header has the Point ID, point description and Generic ID. The fault signatures which were imported, are shown below the header.

Testing the Fault Pattern in Asset Editor

Use the Offline Analysis tab to create expected values. The Concerns tab will have a Faults column similar to the Concerns viewer. The number tells the user how many faults the point is connected to.

Right clicking on the Show Fault Tags, displays only the points used in the calculation of the fault and the pattern of the expected signature in the … column. Comparing the bar charts to the expected signature shows when the fault occurs, when those patterns match. The filter which selected these points is at the bottom of the window and be deleted with the x to show all the points in the asset.

 

Creating a Template from an Asset Specific Fault Group

Once an asset specific fault group has been created, it can be saved as a Fault Template as follows. First all points used in the Fault Group must have a GenericID. When the Fault Group is opened, any missing GenericID’s will be easily seen. Then click the button, Save As Fault Template.

Viewing Faults in Concerns Viewer

When faults are detected, they can be viewed in Concerns Viewer in either the Concerns tab or the Faults tab.

Viewing Detected Faults

The Faults tab displays faults that have been triggered. Once triggered a fault becomes unacknowledged and remains so until the user clears the fault. In this regard the management of faults is the same as management of concerns. Right clicking on the fault allows the user to open the asset or show the fault points in the Concerns tab.

The Concerns tab has a Faults column with numbers like 3/7. This means that there are 3 triggered faults that use this point. There are a total of 7 faults which use this point. Click on the pulldown in the Faults column to display the Faults that use that concern. Right clicking on a Fault will display Show Fault Tags.

Clicking on Show Fault Tags will display only the concerns used by this Fault. The filter at the bottom of the page was automatically created to select the points in the Fault. It can be removed by clicking the x.

The status of the Fault is changed by selecting the pulldown in the Faults column then clicking on the Status pulldown to get the dialog box for writing a note and changing status.

The comment with go into the notes of all the points associated with the fault. The fault Notes can be configured to contain actions to be takes to resolve the fault. There is also a link to user configured web pages with further information to help the user understand the problem and take action to resolve it.

Lifecycle of a Fault

Once a fault is triggered, its status goes from clear to unaddressed. During the resolution of the fault, it can be changed to any another status except clear. Comments are written at the time status is changed. When the fault status is set to clear, the fault is no longer displayed in the Faults tab. If an asset had triggered a fault and the asset was then turned off, that fault would remain visible and unaddressed until it was cleared by user.

Concerns Tab Functionality

When the concerns associated with a fault are displayed, the expected pattern at the right can be compared with the actual concern states.

Actions that Can Be Triggered by Faults

Triggered faults can used to send an email or write an entry to EPLog. The columns Email and EPLog in the Points tab of the Asset Editor set the conditions for sending an Email or a EPLog entry. None means no action will occur. High means the action will occur whenever the high concern of the point goes from clear to unaddressed. Similar action occurs for low. High or Low means the action will occur when either a High or a Low Concern goes from clear to unaddressed. Note that the action (Email or EPLog entry) occurs only once, when the concern is first identified and the status is changed from clear to unaddressed.

The Email addresses and the messages to the Email and EPLog are set on the Asset Properties tab of the Asset Editor using the buttons, Email Settings and EPLog Settings.

Automated Concern Messaging

EtaPRO APR is able to provide automated messaging of Concerns in two ways, based on an asset or based on an asset point. Triggering of automatic messaging occurs when an asset point enters an unaddressed concern status or the total concern count for an asset increases or exceeds a fixed integer value. Once triggered an automated message is sent to an email and/or an EPLog entry.  The messages are configurable based on whether the triggering occurred at the asset or point level.  At the point level, point triggering is turned on and off in the points tab of the Asset Editor by selecting None, the High concern, the Low concern or either HighOrLow.

Email Message Configuration

Triggering is set by the selection of any combination of the check boxes. The email subject can use specific text along with a variety of variables. In the Asset Concerns tab, a Point ID must be inserted within the parenthesis.

The Point Concerns tab provides the ability to configure and email subject and message. In this tab when the variable is used without the point id argument, the Asset Point triggering the message is assumed to be the point in the argument. Example: If the triggering point ID is BFA268,  %Name(BFA268)% will return ID FAN A AMPS. %Name% on the Point Concerns tab will return the same thing.

Finally the email recipients can be added. Click Add, then Manage. Add email addresses to a local list. From that list, add email addresses to Contacts or put them in Groups to form the email list for the asset.

EPLog Message Configuration

Automated messages to EPLog are triggered in the same way as email. Select the Log to receive the messages. Configure the messages in the same fashion with the same variable as was done for email.

In this example the APR log was selected. This Log has several specialized columns that were added from the Administrative Preferences item of the EPLog menu of EPLog. These are called Attributes. The values of these attributes can come from fixed text or numbers or variables and can be different for Asset Concerns or Point Concerns.

Point Details 1 to 10 are defined in the Asset Editor Points tab as text. These can then be used to feed detail information about the point to the EPLog attributes.