A few days ago, the standby-to-duty switchover rotation question below arrived in an email.
“What is the current ‘Best Practice’ on which to base the rotation of operating time between installed pumps that are ‘identical’ A vs B?”
Response from AI Chatbot to the Standby-to-Duty Switchover Rotation Question
Just for a laugh, I put the question into the chatGPT AI chatbot, and it returned the following.
“The ‘Best Practice’ for rotating operating time between identical pumps A and B (which is the same as standby-to-duty switchover frequency) is to:
- Use a systematic and equal rotation schedule.
- Monitor and assess the performance of each pump regularly.
- Based on the performance, adjust the operating time as needed.
- Implement a maintenance program to ensure each pump is maintained to the same standard..
This will help ensure optimal performance and extend the lifespan of both pumps.”
It is not a full answer, but the information provided is sound, and none of it wrong. Though the chatbot did not say how to determine standby-to-duty switchover rotation period between equipment. Even so, the correctness and relevance of the content written by chatGPT is quite impressive. It is not yet at subject matter expert level, but it will be one day soon.
Response I Sent to the Standby-to-Duty Switchover Rotation Question
Below is the response I sent to the email query before I put the question into the AI chatbot.
The debate over the best practice for swapping from the duty machine to the standby machine has been around forever.
I’ve heard reasoning to never use the standby unit until the duty unit fails—except to run the standby at a regular schedule for a short time to prove it is ready for operation upon demand—through to running the duty and the standby each 50% of the time.
There are many issues and factors to consider, such as:
- What is the service life of the pumping station—a 10-year service life is a very different situation to a 50-year plus service life.
- What is the historic reliability of the pumping station—how often has it failed or been taken out-of-service due to component breakage or degradation?
- What disruptions arise to the rest of the operation, and their extent of harm caused, when the standby-to-duty swap is done?
- What overloads, excessive forces, and extraordinary stresses are imposed on parts with each standby-to-duty swap?
- What harm and degradations occur to the standby unit and its systems and components when it is not operating?
- Are there viable workarounds if the station standby-to-duty swap fails—if a workaround exists then doing the swap successfully every time doesn’t matter too much.
- How long to complete repairs/maintenance on the out-of-service previous duty equipment once the standby is operating—if it is a long time to maintain the now out-of-service unit then the operational risk of stoppage with only one working pump are massively increased during the entire outage period.
- Does the company reliability and risk culture and practices aim to keep assets in ‘as-good-as-new’ condition and health, or they allow assets to degrade to conditions that require major repairs and maintenance?
- Is the company risk management philosophy to operate between their risk matrix low-risk line and extreme risks where risk management and risk control efforts stop when low risk is considered achieved? Or is it to operate below the low-risk line where a low risk is seen as an unacceptably high risk? When companies operate in the low risk to extreme risk region it means the management accept extreme risks are possible. The consequence of operating in that zone of the risk matrix is there will be occasional catastrophes and many lesser failures.
I wish there was a mathematical proof that showed the best standby-to-duty switchover option for an operation.
What this question you sent me has made me do is to look at the entire situation of standby-to-duty swap differently to the empirical approach done in the past. Following the Plant Wellness Way EAM methodology, you would do the steps below to provide an answer for the rotation time of A and B units that has a math-based foundation rather than ‘guesstimating’ from empirical evidence:
- Using the entire history of the pumping station’s operation (or other similar stations if there is no history for the specific station), do a run chart of the pumping station failures and then plot the frequency distribution of failures.
- Determine the MTBF for the station.
- If the failure distribution is tight; the organization’s reliability philosophy is to keep assets as-good-as-new (or near to AGAN); and risk management is to operate below the low-risk line, then start with a frequency for standby-to-duty switchover at half the MTBF period (if the distribution is very tight then start with three-quarter MTBF for the swap-over period). When reliability of the pumping station is increased, so too is the period between each standby-to-duty swap.
- If the failure distribution is wide; the site’s reliability culture and practices are poor; and risk management operates between extreme and the low-risk line, then start with a frequency for standby-to-duty swap at one quarter the MTBF period. When reliability of the pumping station is increased, so too is the period between each standby-to-duty switchover.
- Once the swap is completed, prove the prior duty unit is AGAN, and if not, return it to AGAN in a timely, rapid manner.
Sorry I could not give you a simple ‘yes or no’ answer. That is why the duty-standby swap debate has always been around—the best choice depends on what an operation can practically do.
Happy to continue a discussion and/or answer questions. The above proposal for setting best standby-to-duty switchover period is how it’s done in the Plant Wellness Way EAM methodology. It is based on analyzing historic facts and operating data and following good operational risk reduction strategy.
All the best to you,
Plant Wellness Way EAM
2 February 2023