Q&A: Avoiding downtime with a systems-level approach

This approach will continue to mature, evolve and become democratized, bringing operational efficiency and reliability from the plant into the boardroom. 

metamorworks/iStock
metamorworks/iStock

Robert Golightly

 

Robert Golightly is product marketing lead for the AspenTech Asset Performance Management suite. Previously, he managed the product marketing function for the company’s Advanced Process Control and Manufacturing Execution Systems product lines. Golightly’s professional background includes work for SaaS provider FineTooth and Pavilion Technologies.

Q: What causes downtime in a manufacturing plant? Is it often the result of a single asset failure, or something bigger?  

A: When it comes to unplanned downtime, the problem could be the result of a single asset or multiple asset failures. However, the real underlying issue is what exactly causes downtime. When comparing a plant to the human body, for example; we now know dental health has a huge impact on your heart. A tooth may seem like a small thing compared to how your entire body operates, but if you neglect this asset, a larger asset (your heart) can shut down the entire system. There’s a similar trend in complex production systems — the way you operate the asset has the biggest impact on likelihood of failure, not the asset itself. 

Most facilities were designed decades ago and operation processes have since changed dramatically. So, when you think about the high likelihood of mismatches in a manufacturing plant, it’s not just about maintenance, it’s about how you operate the asset. As a result, downtime can often be traced back to plant configurations wherein the failure of a single asset is a derivative of a much bigger business issue.

Q: Explain a systems-level approach and how it works? 

A: By addressing how things work together and interrelate, a systems-level approach can help avoid downtime to achieve operational excellence. The real need for a systems-level approach is simple: there’s usually more than one thing going wrong in a plant at a given moment. 

Failure of systems and assets usually come in floods, and we seldom understand how it will impact the business or what the larger effect of the downtime will be right away. When a failure occurs, how do you prioritize what asset to get back up and running first? How do you know the first thing you should be working on? How do you know what the impact will be on the people down river from you? The systems-level approach can be compared to the adult in the room who tells us, "this issue is more important than that one and here’s why," so we can make the best business decisions and avoid longer periods of downtime. 

Q: Describe some real-world situations in which a systems-level approach worked.

A: We worked with a Saudi Arabian chemicals company that had numerous levels of redundancy — a backup power feed and eight production units — to ensure they did not experience failure or unplanned downtime. However, one lightning storm took out the main power feed and the backup feed and the plant went down for a day and a half. When we looked at the system to run simulations, we learned that the initial level of redundancy wasn’t enough to protect the business. By looking beyond one "bad" asset and examining how the current state of all assets will impact the likelihood that some event could take production offline again, the company was able to achieve a more contextual view of its top 10 contributors to downtime. 

When we talk about a systems-level approach, we think about "alarm flooding" and the domino effect that occurs after one failure. This approach lets companies muffle out the noise and focus on the real alarms and cause for concern, not the secondary ones that can be dealt with later. 

Q: When did operators begin acknowledging the need for a systems-level approach and how do you see the trend evolving over time?

A: In the process industries, the trend of using a systems-level approach started back in the 50s and 60s. Before this time, the operators sent "runners" who would have to take a bicycle from the control room to make changes to valve settings in the plant! But being able to integrate distributed control systems (DCS) was a major step toward adopting a systems-level approach and now there’s a DCS in every refinery on the planet. 

Much like the industry itself, the systems-level approach has evolved over time. How so? First, we broadened the scope and scale of the information we were looking at. Then, companies began using more sophisticated tools — with help from machine learning — to analyze data and detect anomalies. Finally, through the ability to pull data from different assets, we’re now able to achieve a balanced and fair view of the manufacturing plant, bringing us to today’s era of smart manufacturing. 

Now we’re bringing artificial intelligence (AI) into the mix, so it’s a logical continuation of initiatives that started decades ago with technology at the helm. As we move through the new year, it will become more apparent technology has an impact on tactical business decisions, now and in the future. A systems-level approach will continue to mature, evolve and become democratized, bringing operational efficiency and reliability from the plant into the boardroom. 

Q : Overall, do you think there is a widespread adoption of this approach or do you find most plants still operate in the traditional way? 

A: Whether a systems-level approach is adopted or not boils down to the size of the company; if you’re producing at a grand scale you can often afford to make the investment. Overall, there is growing implementation of a systems-level approach, but pockets of seemingly lackluster adoption remain across industries. 

We’ve recognized that the sluggish adoption is a result of the slower implementation of newer technologies (machine learning, AI, advanced analytics, etc.). For instance, when you look at the discrete industry, it has the furthest to go to catch up to digitalization and the weakest infrastructure compared to others. Right now, the biggest companies have the deepest pockets to conduct research and make investments in advanced technologies to propel them forward. However, as machine learning and advanced analytics evolve, and costs around advanced technologies continue to decrease, it will open the door for investment opportunities from other businesses across industries — leading to even broader adoption. 

More in Asset Management