Alarms are meant to grab operators’ attention. Unfortunately, they may sound alerts too often or unnecessarily. For those that do create a nuisance, operations staff should conduct a nuisance analysis to minimize the disruption caused by these alarms.
Nuisance alarms waste time, resources, and money, not to mention they open the door for potential panic among staff. The ability to pinpoint nuisance alarms based on pattern and frequency can help operators streamline processes and fix any issues.
What is nuisance analysis?
Nuisance analysis tracks alarms by type, time, and frequency. By examining the patterns of alarms a control room receives, an operator can determine which require crucial responses and which garner unnecessary attention.
TiPS nuisance analysis overview
Alarm management looks at both personnel and instrumentation to make an alarm system more usable. Nuisance analysis is a great first stop in alarm management benchmarking.
An operator looking at two or three analyses to scope out a problem may miss a more significant trend entirely. Data gathered from every alarm over an entire month lends greater insight into each incident’s what, where, when, and how. A well-designed nuisance analysis summarizes that data and searches the trends that reveal alarm issues needing further review.
The first step in nuisance analysis is frequency analysis. Operators can use the LogMate® ACE module for nuisance analysis, which is unique to TiPS. Using this function in the control room allows operators to resolve more problems and address more meaningful alarms.
In the past, most analyses looked only at the frequency of alarms, perhaps the top 10 that go off the most often. But to get a better picture, operators need to know not just how often they get an alarm but also when.
Does the number spike on a particular day or even a time of day? That information yields many more clues to what part of a system may be going awry than pure frequency measurements. Better still, this type of analysis will also reveal whether a problem occurs occasionally or is chronic, popping up day in and day out.
The more comprehensive analysis aims to eliminate alarm noise and chronic disturbances that distract from genuine upsets. By reducing noise, operators can diagnose serious problems more quickly.
How does nuisance analysis find chronic alarms? Along with telling operators how often an alarm went off, ACE uses a weighting function that is the product of the number of times an alarm occurs by the number of days in a period the alarm occurs.
A high number of alerts on many days puts an alarm on the top of the potential nuisance list. These are the alarms an operator should look at first. Operators might also elect to dig deeper into alarms registering every day.
While nuisance analysis can give operators the count and the number of days, operators can also filter for a specific period. This information contributes to preparing reports. For example, look at the last two days of the month. An upset can happen at the beginning of the month. However, if it’s no longer causing alarms at the end of the month, the problem has probably been fixed. Chronic alarms that still occur at the end of the month may deserve added attention.
Another element operators need to consider is high daily counts. If a process shows upsets constantly, something is wrong with the process, or something is wrong with the alarm settings. Either case merits a look. Operators can easily check for this situation by filtering for a minimum number per day.
For example, the operator might choose a floor of five alarms per day. That configuration added to previous filters can winnow a nuisance list.
Operators can also check the frequency of a particular alarm during a week. This closer look can extend to the number of days per week and the number of times per day. A deeper exploration of a specific alarm during a single day and each alert for that day will give an even more detailed picture of triggering events.
Drill-down capacity also enables operators to look at a small set of chronic alarms and determine what’s going on. For instance, some facilities will see a high number of alarms each day for every day of the month. That gives them quick visibility into processes and settings that call for an investigation.
Some operations will set their analyses to cover a few preceding days. That may give a view of alarms that go off many times per day on consecutive days. The staff may choose to look into those first. Operators can also filter alarms that rarely occur during a single day.
Best practices for nuisance analysis
Operators are encouraged to run analyses at the end of the month or the first of the following month. However, it is not necessary to wait until a month passes. Weekly analysis of daily alarms may be optimal for high-frequency chronic alarms to allow the operator to address process concerns more promptly.
Nuisance analysis helps eliminate unnecessary delays and keeps your production on track. To find out more about these process boosts, contact TiPS.