Monitoring Many Alarm States Across Numerous PI Tags with PI Point Arrays in AF

Monitoring Many Alarm States Across Numerous PI Tags with PI Point Arrays in AF

Introduction

Monitoring numerous alarm states across different data streams can become daunting, especially when handling a large number of PI tags. A recent customer case provided a challenging scenario involving over 500 unique fault code indications for each wind turbine, dispersed across 500 tags. The objective was to simplify analyses by consolidating these fault states into a single, manageable attribute. This post explores how the use of PI Point Array attributes in Asset Framework (AF) and Analytics can streamline the process, offering a smart alternative to creating gigantic analysis sets.

Scenario Overview

For context, the customer needed to monitor and trigger event frames/notifications based on the overall fault state of a wind turbine, avoiding the cumbersome route of managing 500 individual attributes and analyses per turbine. The solution leveraged PI Point Arrays to efficiently aggregate fault codes and assess the overall fault state.

Configuration Steps

1. PI Data Archive Setup: Begin by creating and assigning unique digital state sets for each fault code tag type. Ensure that the 'good' state is consistent across all tags to facilitate seamless filtering.

2. Asset Framework Configuration: Create an attribute within AF that utilizes the PI Point Array data reference with a value type of String Array. This attribute accumulates all relevant fault code tags.

3. Analysis Development: Implement a periodic expression analysis in AF that discards any fault codes currently in a 'good' state, allowing you to focus on critical alarms only.

4. Attribute Output: Ensure the output attribute incorporates compression to prevent duplicate entries, thus optimizing performance.

Important Considerations

• PI Point Array attributes cannot directly trigger analyses. Consider using another indicative tag that shows any faults which require periodic checks.
• Configuration spans multiple tags into PI Point Arrays for all turbines. Utilize PI Builder for streamlined configuration workflows.
• If applicable, tag naming conventions across turbines allow for substitution parameters, automating template-based configurations.

Assumptions

• Faults sustain their "tripped" status long enough to be detected within a reasonable periodic schedule.
• Fault tags revert to 'good' state correctly post-resolution.

Additional Insights

Leveraging table lookups with array-type attributes is another powerful application, enabling the return of multiple rows from a single column into an analysis-ready format. Such use cases broaden the scope of PI Point Arrays beyond mere fault monitoring.

Comparison with ACE

Some may consider using the Advanced Calculation Engine (ACE) for similar logical arrangements and decision statements. However, using PI Point Arrays offers distinct advantages in specific contexts, notably when the goal is to centralize and simplify attribute analyses.

Conclusion

PI Point Arrays in AF afford a significant operational advantage by consolidating extensive tag arrays into manageable analyses, thus enhancing system efficiency and alert precision. If you’re dealing with complex systems with numerous alarm states, this method offers a streamlined approach worth considering.

Call to Action

Share your own use cases or any other applications where you've successfully employed PI Point Arrays. Your practical insights and experiences enrich the community's understanding and utilization of PI System capabilities.