What is an asset life model for condition based maintenance
Today, condition-based maintenance (CBM) is entering the mainstream. A typical CBM solution uses real-time sensor data and predictive algorithms to predict potential failures days or even weeks in advance.
Asset Life Model (ALM) is a new type of CBM solution. It is a model that is trained on all the operational data across the life time of the asset. For example, you could collect all the pressure, speed, mud volume, and mud quality data across the life time of a mud pump suction valve, and use those data to train a suction valve life model. Now when you put in a new suction valve, you can predict the life of this valve, starting on the day it is put into service. As that suction valve gets used, the prediction result is adjusted continuously based on the cumulative operating conditions on the valve.
The ALM is effectively a fatigue model. For all mechanical components, their life time depends on the cumulative fatigue. Modern data science enables building these fatigue models by applying all the operational data from the asset.
What are the benefits of ALM? They allow you not only to predict potential failures but also to understand asset performance as a whole. As compared to traditional CBM, which only catches failures as they are about to happen, ALM models the entire life of the asset, so that you know not only when it will reach end of life, but also what operating conditions contribute to the end of life.
In summary, ALM offers the following benefits that are not available from traditional CBM:
Asset Life Tracking: You can track how much life is left in the assets from day-one. This gives the ops team more time to plan maintenance activities.
Failure Forecasting: you can predict how many assets or asset components will fail months in advance. This enables planning and budgeting.
Performance Normalization: You can compare asset quality across different operating conditions because ALM can normalize asset performance across varying operating conditions. This is great for the supply chain team to evaluate vendor quality and enhance negotiation power.
Root Cause Analysis: You can find root causes of performance issues. This helps to improve asset performance for the long term.
What are the limitations of ALM? First, it cannot predict failures from components that are already defective or if the components are installed incorrectly. Other CBM techniques need to supplement the ALM in these cases. Second, if the asset or asset component lasts a long time, then it will take a long time to collect enough data to train the ALM. In this case, historical data could help.
With the advent of modern data science and compute capability, large-data models like the ALM has become not only possible but also cost effective. For the first time in our history, we can gain a complete understanding of our assets.
Click here to learn more about our digital twin solutions using Asset Life Models.