With a 9¾ mindset going into an outage, we are leveraging the unique moment of the interchange between production versus planned downtime. We have written about the importance of post-outage reviews in a blog called Cheers…to the last outage that ensures we evaluate the effectiveness of the outage's execution.
Missing from these and other downtime reflections, nothing has been written on what good start-up from a bad start-up. Whether coming out of a downturn, a repair turn, or a major planned/unplanned outage, one generally feels whether it transitioned back into production successfully.
What is considered a good start-up from an outage?
I recently heard an executive refer to coming out of an outage as the best yet. This instinctual or gut opinion was generated based on the effectiveness to achieve a targeted production rate in a desired amount of time. But what were the quantitative attributes that substantiated this opinion so that we can do it again?
Under a micro-lens, the effectiveness of the outage can be measured by start-up time or the time to make the first produced product. But this is short-sighted because it is simply starting up and not getting to the desired output. Instead, using a 80/20 rule as a macro-lens can measure the start-up's success.
Using the 80/20 rule for start-up effectiveness
Consider a time-series plot where the y-axis is OEE (Overall Equipment Effectiveness) and a production unit has a targeted OEE. We will call this percentage “Zeit” or the target in German. Additionally, on this y-axis is 80% of the Zeit. We will call this percentage the “Decke,” or the ceiling in German. For example, if the Zeit is an OEE of 85%, the Decke would be an OEE of 68%.
The x-axis is a simplistic timescale, and we will consider the unit of measure an hour. For simplistic purposes, we assume an operating unit is scheduled to be at the Zeit of 85% within 24 hours. We will call this duration of 24 hours a “Zeilzeit” or targeted time in German. Using the 80/20 rule, 20% of a Zeilzeit of 24 hours is 4.8 hours. We will call time this the “Wand” or the wall in German. You can remember the rallying cry of this time-series plot as we want to be through the ceiling by the time we get to the wall. Or, wir wollen durch die Decke sein, wenn wir an der Wand ankommen.
In this example, we want to achieve an OEE of 68% before 4.8 hours and achieve 85% OEE within 24 hours.
Good Start-up = AND[(Wand <20% when Decke >80%), (Zeit achieved before Zeilzeit)]
Start-up Effectiveness
If the Zeit of 85% OEE is not achieved within a Zeilzeit of 24 hours, the start-up is not unsuccessful. Additionally, if the targeted Decke of 68% was not achieved within a Wand of 4.8 hours, the start-up was not good. This is an AND statement, whereas we need both. We need to have a good ramp-up to production AND achieve the targeted run rate by a certain period. If we fail to meet both of these measurements, it is a "Bad Startup." If we achieve both, we can classify it as a "Good Startup." From a performance review perspective, one could imagine striving for 90% start-up effectiveness over a year. We will call this value Start-up Effectiveness (Image 1).
Example 1 - Start-up Effectiveness
What we have shown here is one way to determine if the start-up after downtime was effective or ineffective. We have demonstrated another way to use the 80/20 rule to establish a targeted run rate in a period while ramping up to a targeted production rate. There are plenty of other ways to do this, but this is one that you can objectively indicate if the start-up was good or not and trend over time.
Comments