Process Alarms: Dead weight or the back bone of productivity?

Around the globe, production is fundamental for economics and growth, and a key factor for employment. The process industry has several years invested largely in automation and digitalization to obtain global competitiveness and increased productivity. We are stepping into the industrialization 4.0, enabling the opportunities included in Internet of things, service and people. But how about reliability and maintenance? Is an alarm system dead weight or the back bone of productivity?

But to give this post a fare start, what is an alarm?
Well, looking into the etymology of the word, one can read the following description: “The word came from the old French à l′arme meaning “To the arms”, “To the weapons” telling armed men to pick up their weapons and get ready for action, because an enemy may have suddenly appeared”. One may conclude that an alarm needs to be followed by a reaction, to defend themselves to a potential enemy, to evacuate a building when the fire alarm sounds or - to be put in our circumstances - to perform a task to keep the process running.

To go further, you can look into the different standards for alarm management available, which includes ISA 18.9, IEC 62682 and EEMUA 191, where they managed to define an alarm in a simple yet clear approach: “an audible and/or visible means of indicating to the operator an equipment malfunction, process deviation, or abnormal condition requiring a response”.

So far so good right? The most interesting part in that definition is the very last three words, “…requiring a response”. Do all your alarms comes with a response? And some food for thought regarding that statement. How many alarms should one object be able to create? If a temp sensor on a pump passes its alarm limit, should the sensor, pump, the outgoing pressure sensor or the process generate the alarm? Or as more common: all of them?
The last one causes alarm flooding, which is a common concept in the industry. But it really needs drainage to keep operators from drowning.

An alarm list should normally be pitch black: Empty. An alarm should come with an identified response. It′s not an information feed, it′s an action list - To the arms!

Texaco Milford Heaven in 1994 suffered from a sever explosion, causing 26 injured personnel and material damage for over 100 million USD. The accident was thoroughly investigated and the key factors was summarized as following:

• The last 11 minutes before the explosion, the operators had 275 alarms to identify, acknowledge and response.
• Too many alarms, furthermore lack of prioritization 

Okay, you may say that this will not happen to you. Sure, I honestly hope so. But there are many other reasons why you should improve you alarm system:

• Improve operator work environment
• Ease identification, prioritizing and responses to alarms

Which will give you:

• Lower risks for personnel and equipment damage
• Lower risks for unplanned production stops / losses
• Increased product quality and delivery reliability

Do you still having troubles to see the potential? Please reflect if you ever seen any of these behaviors:

• Operator makes acknowledge or silence alarm without looking and/or any other response?
• A missed alarm causes an incident or almost incident?
• Alarm sound is muted?
• Operators do not know the meaning of a specific alarm or what they are supposed to do?
• An alarm is disabled or in some other way deactivated?

These are all characteristics of a poor alarm status. So what can you do about it?
Well, it can be consider as a two phase rocket, starting with the basics:

1. Alarm philosophy. Good alarm strategy needs to be built-in. To form an alarm philosophy will not only tell you which alarms that should be considered as alarms and how they should be prioritized. It also will serve as a guideline for all your consultants and in-house programmers when building a new application.
2. System clean up. Based on your alarm philosophy, your current system needs to be cleaned and functional to increase the usability. To ensure reliability, you′ll need continuity and to go from reactive to proactive.
   
   

It all comes down to three steps:

Step 1: Diagnostic and analytics
With efficient tools and alarm fingerprints, we can fast and reliable get an reality check of the alarm status. This is then used as input for analyzing your situation and what suitable actions may be.
 
Step 2: Improvements
Identified actions such as System 800xA alarm features i.e. Alarm Shelving and hiding, and besides that your system is cleaned according to your alarm philosophy.

Step 3: Monitoring
We need to monitor your systems, to ensure output and to sustain an good alarm status. This should be followed by more iterations for continuity and constant improvements.

My advice is to really put efforts into this manner. Working continuously with alarm management is eliminating the balance act between positive and negative cash flow. The balance between success and failure. 

Reach out! We′re here - close to you. With expertise, tools and benefits to be proactive and unleash your true potential!
 
Process alarms are the back bone of productivity. So embrace it.