Process Safety: Meet Me Under the Clocks

“I’ll meet you under the clocks.” This is such a Melbourne comment. Growing up, the clocks at Flinders Street Station were the perfect meeting point to catch up with friends. At a main train station hub, it is accessible from all train lines. Visually, you can’t miss the line of clocks advising the time of the next train services. I love that it is still a meeting hub and that the clocks still work! Inside the station, there are many digital displays with the platform details and times. We have a multitude of apps that tell us when trains arrive and depart, and which platform they use. But outside, we have the old school, simple version.  

In today’s world, we can get caught up in new digital ways that add layers, which can often be quite complex. In engineering, we strive for the elegant solution, the simple option. The analog clock that does the job. When we apply this thinking to process safety, we are looking to use inherently safer design options. How can we eliminate the hazard? Can we substitute it for something less hazardous? Can we moderate the process conditions to reduce the hazard? Can we minimize the quantity or amount of the hazard present? Last but not least, can we simplify the process to remove room for error? 

I recently worked with a company developing new energy technologies, and their solution is beautifully simple. The perfect example of an elegant solution. Taking established mechanical and electrochemical processes and applying them in novel ways.  

Over my career, I have seen the widespread embedding of distributed control systems (DCS) and programmable logic controllers (PLC). These systems provide a huge leap forward in process control. We have real-time monitoring of multiple parameters and can link actions to specific triggers. Providing for automated optimization that is not reliant on a human sensing some data, thinking about an option, deciding and then acting. These systems have led not only to process improvements but also to greater safety margins and better outcomes.  

So, is there a downside to applying multiple levels of technology? I think so, when I reflect on one of the inherent after-design principles I mentioned, simplification. When we add multiple levels of technology, we increase the complexity of the system, and complex systems can fail in multiple complex ways. We can be misled by the simplicity of applying the multiple layers. One of these multiple layers is the illusion of alarms providing some form of control. How many times in a risk assessment have you heard someone say, “We can put an alarm on that”? The thought is usually that if we alarm it, we have taken some action, but this is not the case. An alarm requires additional human intervention, which can fail for any number of reasons. Such as misunderstanding the requirements, slow response or incorrect action taken. It is so easy in the world of DCS and PLCs to just alarm any parameter we want. In the early days, this led to almost every condition being alarmed, resulting in what we call an alarm flood. During a plant upset situation, the operator received so many meaningless and useless alarms that they could not see the truly important alarms, let alone actually respond to the ones that required immediate action. While much work has been done in plants to rationalize alarms, we still see alarm flooding, and less experienced people just adding more alarms. Apply artificial intelligence on top of that, and we can compound some of the complexity.  

We need to stop and think about the complexity we are building in. To do this, we first need to focus on the outcome we are striving for. Then think about what the elegant solution would look like. We still need to implement the necessary controls, but if we can do something in a simpler way, it will be simpler to operate and simpler to maintain. 

Clocks are simple and work. They provide not only the necessary outcome, but also an iconic Melbourne experience of saying, “I’ll meet you under the clocks.”