The Unplanned Outage

Oh Crap! All hands on deck! You’re in another unplanned outage and you are working 16 hour days trying to get the plant up and running again. This time it was a broken agitator shaft, but you’ve dealt with other equipment failures and operator error in the past. Regardless of what caused the outage, you are stuck working to fix it and get back up. Outages are a pain and they are expensive. You have to pay your people overtime to get the problem fixed. You have to spend big dollars to replace/repair the broken equipment. You are also not making any product and therefore not making any money. It’s a trifecta of terrible. What if we could take steps to minimize unplanned downtime? Would you do these steps or at least some of them? I think it is worth it to find the time when things are running well to prevent things from going badly.

First, you must figure out the potential failure points in the plant. On a grand scale it could be three things. Operator error, loss of utilities, or equipment failure (there could be more, but for the sake of this article, let’s go with those three).

In regards to operator error, can you predict what an operator could do to damage or shut down the process? If you can, then implement control system or SOP fixes to prevent them from being able to that damaging activity. Noone wants to shut down a plant, but accidents can happen. Take steps in the software to prevent those accidents.

Regarding the loss of utilities, I don’t know that there is much you can do. You could keep a spare transformer or other large utility items as spares, but these events are relatively rare and it would be really hard to predict which component (yours or the utilities) would fail.

Lastly you have equipment failure. This is preventable or at least predictable. We all do preventative maintenance. Do you do predictive maintenance? Changing oil in gear boxes and replacing consumable parts on a regular basis prevents unplanned down time, but what if there is something that doesn’t get these PMs that fails? What if you could predict this failure? instrumentation has a lot of diagnostic information ready to be used already. You just have to look and have I/O capable to reading and logging that data. You could also use a smart cloud service (like Siemens) to analyze that data for you. There are also SIL (Safety Integrity Level) devices and equipment. These have proven and tested and predictable mean time to failures and can be used to ensure the safe and reliable function of your process. Keeping spare parts and complete assemblies for all your critical components is worth the inventory cost, espessially if you will lose enough paying idle people, not making product, and paying expedite fees to ship an emergency part in from somewhere else.

So, while you are stuck in this unplanned shutdown, use this as an opportunity to make some changes to your inventory, SOPs, and data analysis to prevent the next one. If I could do a little work and spend a little money now to prevent spending a lot of money and doing a lot of work later, I would.

If you like this article (or not) leave a comment and let me know what you think. I will try to cator future articles to the interest of my readers.