7. Alarms

Alarms are how the plant talks to the operator. In the OTS they come from the alarm limits of gauges and transmitters, are announced on the panel and the screens, and live in an alarm summary with the acknowledgement discipline of a control room.

What you will learn

• Setting LL, L, H and HH limits on gauges and transmitters
• The life of an alarm: active, acknowledged, back to normal
• The alarm summary window
• Using alarms in interlocks and objectives

Prerequisites

• Page 5

Where alarms come from

Every analog, digital and level gauge, and every transmitter, has four limits: LL (very low), L (low), H (high) and HH (very high), each with its own enable box, in the object's editor. The sample plant's PIT-001 has LL 100, L 150, H 400 and HH 450 kPa; LIT-001 has LL 0.15, L 0.25, H 0.4 and HH 0.5 m.

The gauge computes its flags every step; the session watches the flags and turns each change into an alarm record and a journal line: PIT-001 H alarm ACTIVE (value 752.3 kPa) and later PIT-001 H alarm cleared.

Priority follows the level: HH and LL are High, H and L are Low.

The life of an alarm

stateDiagram-v2
    [*] --> ActiveUnacked: limit crossed
    ActiveUnacked --> ActiveAcked: operator acknowledges
    ActiveUnacked --> NormalUnacked: value returns
    ActiveAcked --> [*]: value returns (leaves the summary)
    NormalUnacked --> [*]: operator acknowledges (leaves the summary)
    NormalUnacked --> ActiveUnacked: limit crossed again (same record)

An alarm leaves the summary only when it is both back to normal and acknowledged. An alarm that returns to normal before anyone acknowledged it stays listed as normal, unacked, so a fleeting excursion during the trainee's coffee break is still visible when they come back.

Where the operator sees them

• The banner on the operator panel and the alarm line on the screens show the newest unacknowledged alarm and blink red until everything is acknowledged. The count of active alarms sits beside it.
• Gauge tiles and bound ISA bubbles show the level in red.
• Acknowledge on the panel acknowledges everything at once (the shortcut of a busy operator).

The alarm summary

Summary... on the panel, or the alarm line on a screen.

Columns: raised at (simulated time), tag, level, priority, value and units when raised, state, cleared at, acknowledged at. The list is sorted with the newest first; Active only hides alarms that are back to normal.

• Acknowledge selected, or double-click a row, acknowledges one alarm.
• Acknowledge all acknowledges everything.

Each acknowledgement is an operator action in the journal: Acknowledged PIT-001 H.

Alarms elsewhere in the OTS

• Interlocks (page 8) take Alarm active as a condition: "trip when PIT-001 HH is active for 10 s".
• Objectives (page 4) take the same condition: an Avoid on HH is the usual "do not let it get that far".
• The OPC UA server (page 9) publishes the counts of active and unacknowledged alarms and accepts acknowledge all.
• The exercise report counts alarms raised.

Try it

1. Run the sample at 10x. Take PID-012 to manual with 0 % output: PV-001 closes and the pressure climbs.
2. At 400 kPa the banner shows PIT-001 H ACTIVE. Do nothing.
3. At 450 kPa PIT-001 HH joins it. Open the summary: two rows, both unacknowledged, HH with priority High.
4. Acknowledge H only. Put the controller back in AUTO; the pressure falls.
5. When the pressure drops below 400 kPa the H alarm leaves the summary (acknowledged and back to normal) and HH stays as normal, unacked until you acknowledge it.

The next page turns the HH alarm into a trip.