Calculating WinCC Pending Alarm Count Filters, Locks, Archive

Calculating WinCC Pending Alarm Count: Filters, Locks, Archive Reload

1. Overview

The "Pending" indicator in the status bar of a WinCC Alarm Control (Classic V7 / V7.5 and WinCC Unified V17 onward) is one of the most frequent sources of confusion in operator HMIs. Operators and engineers often report that the status bar counter does not match the visible row count in the alarm list, that messages appear "stuck" after a WinCC server reboot, or that the counter grows even when filters are applied.

This reference explains the exact counting logic, the role of the Alarmlogging editor, the effects of filters and message locks, the archive-reload behavior after power failure, and how to verify and reset stranded pending states. It covers both WinCC V7 (Classic) and WinCC Unified V20 (TIA Portal V20) so that migration projects retain consistent alarm presentation.

2. Prerequisites

• WinCC V7.0 SP3 or later (Classic) or WinCC Unified V17 / V18 / V19 / V20 (TIA Portal).
• WinCC Explorer / TIA Portal with the project file open in the engineering station.
• User rights to edit the Alarmlogging editor (Classic) or HMI alarm configuration (Unified).
• Read access to the message archive database (*.LDF / *.MDF for Classic; SQLite *.db for Unified Runtime).
• Optional: WinCC / TIA Portal documentation set installed locally or reachable via Siemens Industry Online Support.

3. How the "Pending" Counter Is Calculated

In WinCC Alarm Control (Classic V7 and the WinCC Unified alarm control), the status bar exposes up to four counters, configured through the control's Status Bar properties:

Counting rule (per Siemens WinCC manual, SIMATIC HMI WinCC V7.5 documentation): the pending counter equals the count of all messages currently in the system that carry the state Pending = ON and that belong to at least one alarm class flagged as acknowledgment-capable. Filters, selection lists, and lock states do not remove a message from the underlying pending pool — they merely hide it from the operator view.

4. Filter, Selection, and Lock Behavior

The following WinCC control properties influence which messages are displayed, but they do NOT change which messages are counted as pending:

This explains the question seen on the WinCC engineering floor: "the actually displayed rows are fewer than the pending counter". The "hidden" rows remain in the pending pool until acknowledged, are removed via Hide Acknowledged Messages in the Alarm Control properties, or are archived out by the runtime message wheel.

5. Archive Reload After Power Failure

Classic WinCC V7 introduced a property called "Reload after power failure" in the Alarmlogging Editor. When enabled, on a WinCC Server restart the runtime reconnects to the message archive and re-imports historic rows whose Pending state was 1 at the time of shutdown.

On a Unified runtime (V17+), the equivalent behavior is controlled by the Alarms - archived and Alarms - history settings in the alarm control configuration (see Configuring the alarm control (RT Unified)).

6. Why Messages Stay "Stuck" Pending

The "stuck pending" symptom has three principal causes. Diagnose each before applying a workaround.

6.1 Step-by-Step: Clearing Stranded Pendings

1. Open the WinCC Explorer on the server hosting the Alarm Control.
2. Open the affected picture and double-click the Alarm Control to launch the WinCC Alarm Control configuration dialog.
3. Under Toolbar, click "Selections" and remove every active selection, time range filter, and message-class filter.
4. Click Locks and clear every active lock entry (operator lock list).
5. Click Acknowledge visible to mark all currently visible rows as acknowledged. Acknowledged rows exit the pending pool and the status bar counter drops by their quantity.
6. For rows still pending because they originated from a PLC trigger that has since cleared, use Acknowledge all with administrator rights. In WinCC V7.5 this requires @Alarm_AcknowledgeAll script action or the equivalent DCS user right.
7. If a strikethrough row remains, open the Alarmlogging Editor and either disable Reload after power failure for the affected message wheel, or accept the strikethrough as a documented behavior and provide operators an Acknowledge hotkey.

7. WinCC Unified (TIA V20) Alarm Control

WinCC Unified takes the same pending concept and exposes finer-grained configuration objects. The two settings that map directly to the V7 behavior are:

• Miscellaneous → Alarms - current — selection list of message classes that count toward the pending counter. By default, all classes that support acknowledgment are included. See Configuring the alarm control (RT Unified).
• Miscellaneous → Alarms - pending — which subset of the current alarms is to be acknowledged as pending events. See Acknowledging alarms (RT Unified).

Unified also introduces a per-class visibility flag and a server-side PendingHidden attribute referenced in TIA Openness V20. A pending alarm whose alarm class is set to Hidden in the HMI device configuration is removed from both the visible result set and the pending count.

7.1 Unified Status Bar Configuration

1. In TIA Portal, open the Unified HMI device and switch to Rt Alarm Control under the screen tree.
2. Open Properties → Status bar.
3. Tick Pending, Not Acknowledged, Displayed, and Locked as needed.
4. Under Miscellaneous → Alarms - current, choose All or restrict to a class group.
5. Under Miscellaneous → Alarms - pending, confirm whether only "alarm" classes or additionally "warning" classes contribute to the count. This is project-specific.

8. Key Configuration Parameters (Classic V7 vs Unified V20)

9. Diagnostic Script Examples

WinCC V7 supports VBScript on the Alarm Control to introspect the live state. The following illustrative function returns the total pending count for the active control, useful for cross-checking against the status bar:

' WinCC V7 VBScript - reads pending count via underlying Alarm OCX
Function GetPendingCount(objAlarmControl)
    Dim count, i, msgStates
    count = 0
    ' HMIRuntime.Alarm object exposes the internal message list
    For i = 1 To objAlarmControl.MsgCount
        If objAlarmControl.MsgState(i) = &H00000001 Then ' 0x01 = Pending bit
            count = count + 1
        End If
    Next
    GetPendingCount = count
End Function

' Typical call from a button:
' MsgBox "Pending: " & GetPendingCount(ScreenItems("AlarmControl_1"))

On Unified V20, the same data is exposed through the IHmiAlarm JavaScript API. Iterate Screen.Items('AlarmControl_1').ProcessValues and read the State bitmask for each row; bit 0x01 again represents "Pending".

10. Verification

After any change to filters, locks, or archive settings, verify the pending counter as follows:

1. Open the runtime picture containing the Alarm Control.
2. Read the Pending and Displayed values from the status bar.
3. Remove all selections and all locks.
4. Acknowledge every visible row.
5. Confirm that Pending = 0 AND Displayed = 0. If they differ, run the diagnostic script from Section 9 and capture the script output for trending.
6. Reboot the WinCC server with archive-reload enabled. Confirm that historical pending rows return with strikethrough timestamps and that Pending reflects the recovered count.
7. Cross-check with the archived values via WinCC Archive Viewer or SQL> SELECT COUNT(*) FROM ALGV_VF_BLOB_DATA WHERE STATE & 1 = 1 (Classic, MS SQL backend).

11. Common Pitfalls

• Mixing filters and counts. Operators may believe that switching off a filter "creates" pending alarms. The alarms were already pending; they were simply hidden.
• Non-quittable alarm classes. PLC-driven status messages (e.g., class "Status") set the pending bit but the Alarm Control excludes them from the counter unless the class is configured as Acknowledgment-capable.
• Multiple servers. In a redundant WinCC server pair the pending count is per server. If the project only displays one server's control, the count represents that server's view, not the plant-wide total.
• Strikethrough rows after archive reload. Clear them with Acknowledge all once power is restored or accept them as historical recoveries.
• Misconfigured message wheels. If the alarm archive wheel is shorter than the running time of the process, the archive may evict pending rows before they are acknowledged. Extend the AlarmLogging archive retention period in the Alarmlogging Editor.

12. Field-Commissioning Checklist

• Define, on paper, the alarm classes that count toward Pending for the operator view.
• Decide whether Reload after power failure (V7) / Alarms - archived (Unified) is required. For 24/7 plants without UPS-backed servers, leaving it off avoids revived pendings.
• Configure Alarm Control status bar with Pending, Displayed, Locked, and Not acknowledged.
• Provide operators a documented acknowledgment procedure covering filters, locks, and strikethrough rows.
• Add an event-triggered script that logs the pending count once per minute to an internal tag for historical trending.

Frequently Asked Questions