Substation Asset Performance Management

by Mark Biagi, Bentley Systems

An effective system for managing asset integrity and reliability is the most important barrier against failure in complex industrial facilities such as electric utility infrastructure. Throughout electric utilities, assets are being pushed to their limits with increased loads on aging infrastructure and enhanced expectations of reliability. Owner-operators are dealing with the conflicting pressures of minimizing operating costs to increase profitability while demonstrating high standards of asset governance and meeting regulatory obligations. To succeed, owner-operators need a better strategy for managing and maintaining their complex assets-a strategy that supports more economical and impactful decisions.

Asset Performance Management

Asset performance management (APM) is gaining wider acceptance as a way of addressing these conflicting requirements. APM describes the subset of an asset management strategy that relates approaches to operations as opposed to conventional reactive or time-based approaches. APM also defines a category of services and software products that can be applied tactically to plan and execute a program of improving asset integrity and reliability.

The principles of APM also are aligned directly with the new ISO 55000 standard, which sets a new benchmark for asset management best practice. This standard, released in January and building on the Institute for Asset Management’s PAS 55 approach, has gained the attention of owners and operators of infrastructure assets and has piqued the interest of shareholders, stakeholders and the insurance industry, who are equally interested to know how well their assets are being managed. Many infrastructure owner-operators are engaging consultants to help them understand where on the scale of asset management maturity they reside and what they need to do to get measurably closer to ISO 55000. Often the answer comes down to APM.

APM unites and adds value to many existing systems and processes that have become widely accepted across industrial facilities. Owner-operators understand the need and the value of having such tools as enterprise resource planning (ERP), enterprise asset management (EAM), maintenance management system (MMS), condition-based monitoring (CBM), document management systems (DMS) or some combination thereof.

Why does industry need another three-letter abbreviation? To shift from a reactive or time-based maintenance regime to a risk-based and reliability-centred approach requires the ability to straddle the conventional boundaries of transactional (e.g., ERP) and time-series (e.g., condition monitoring) systems to drive better decision-making.

Asset performance management unites the five fundamental elements of an asset management strategy shown in the figure, namely:

  • Management strategy. What are the business and performance requirements for the asset?
  • Asset context. What and where is the asset?
  • Asset performance. What is the asset’s condition and how is it performing?
  • Risk assessment. How can failure occur? What is the likelihood? What are the consequences?
  • Risk mitigation. How are scheduled and unscheduled maintenance (and incidents) managed?

There are countless reasons organizations might not be managing their assets effectively, especially not to ISO 55000 standards. The business objectives might have changed such that the asset is becoming a liability (for example, new emissions legislation changes business objectives). Likewise, the asset context might not be well-understood as an organization might have acquired assets that are poorly documented or that have been modified without updating the engineering information. The asset performance might not be well-understood as sources of field information such as inspections or condition monitoring might be ineffective. Risk assessment might not have been carried out adequately to understand the ways in which failure can occur, which applies to failure of physical assets and to meet the business objectives. Last, the risk mitigation methods employed might not address the condition and potential failure modes adequately.

Notice that the figure does not mention maintenance. For many organizations, maintenance is just something that must be done, like housework. Maintenance tasks are often an aggregation and accumulation of all the individual tasks that are recommended by all the individual vendors of equipment that make up the system. For an organization to be effective to the standards of ISO 55000, all inspection and maintenance become risk mitigation, namely that each inspection and maintenance activity should be aligned with the indicators of specific failure modes and driven by the likelihood and the consequence of those failure modes given the asset’s condition, operating context and business objectives. This inter-relationship between conflicting requirements and disparate sources of information is fundamental to asset performance management.

APM for Substation at Exelon

Exelon is one of the largest electric companies in the U.S. with close to $19 billion in annual revenue and more than $55 billion in assets. It has the largest market capitalization in the electric utility industry and is ranked No. 1 in gas and electric utilities by Fortune magazine. The ComEd division delivers electricity to several million customers in northern Illinois.

Although one of the strongest points in a power system is the electric substation, it still contains weak points or points of failure that could lead to loss of load. By knowing how to calculate the reliability of substation configurations, an engineer can design a system with the best overall reliability. But determining the reliability of a substation also can be important for existing installations because it can help locate weak points that might contribute to overall system unreliability and therefore help prioritize maintenance and capital replacement strategies.

Several years ago, ComEd commissioned Kinectrics Inc. to conduct a study of 31 classes of substation and transmission assets. The study developed a score for each of the 57,000 assets, where zero represented an asset that had failed beyond repair and 100 represented a brand new asset. This asset health index score initially was used to drive capital replacement plans and help prioritize maintenance work for the following year.

The Challenge

The study was valuable, and ComEd envisioned many more potential uses for the asset health indices, but the study results were becoming out-of-date. ComEd tried to manually recalculate the indices but found the data volume overwhelming for its small internal staff. ComEd needed a way to update the indices fairly regularly with minimal effort so that it could make decisions based on the current health of its assets-not the health as it was assessed several years ago.

ComEd was impressed with Bentley’s AssetWise Performance Management’s ability to gather data from multiple disparate process control, supervisory control and data acquisition (SCADA), computerized maintenance management systems (CMMS) and homegrown systems and use the data to generate alarms for condition-based indicators. The next step was to leverage the software’s data connectivity and flexible customization tools to develop complex health index calculations.

The Bentley team took a sample health index spreadsheet for one of the asset types and turned around a prototype in less than two weeks for what the final health index system would look like. Based on this, they developed a production system in less than three months for generating health indices for all 31 asset types and 57,000 assets.

The developed solution defined and created a Health Index Worksheet for each of the 31 identified asset classes. Individual components then were created for each Health Index Worksheet to define the desired calculation for that component. For example, one component might be to count the number of infrared measurement problems on a circuit breaker over the past year. Another might be to determine the most recent trip time measurement on a circuit breaker. A third might be to find the maximum load peak readings over the past 60 months on a transformer.

Each component data value is then normalized by referencing a factor table and deriving a factor ranging from 4 (excellent health) to zero (very poor health). Each factor is then multiplied by a weighing factor to obtain a score for that component. Some asset classes could have as few as 14 uniquely calculated components while others could have as many as 90 or more components, depending on the complexity of the equipment. Next, each component score is calculated in a health index calculation. The component score is divided by the maximum possible score to derive a health index expressed in percentage (zero to 100) for each component. Finally, all the component scores are combined and divided by the maximum possible scores to determine the overall asset health index in percentage (zero to 100).

Also, a calculation is made on the data that was found during the component calculations to evaluate how complete the data was during the calculations. For example, if a health index for an asset was found to be 95 percent-quite good-while the data availability was only 65 percent-moderate-then the confidence of the health index might be somewhat suspect. Assets in a given class then can be ranked by their health index score and the poorest ones (lowest percentage) identified for further analysis.

The data is provided by automated daily interfaces to several legacy systems, providing:

  • Updates to the assets;
  • Indicator readings (measurements) for all the measured parameters;
  • Work orders (created against the assets); and
  • Problems identified in the field for each asset.

All factors, weightings, parameters and components are maintainable and customizable by Exelon personnel, allowing what-if scenarios to be run at will. Health indices can be run on an asset or the entire asset class. The developed framework and logic allows other asset classes and health indices to be added easily. Finally, the calculation programs can be modified by Exelon information technology personnel to provide a robust, flexible solution into the future.

The Result

The inputs to the health index calculations are collected automatically every night from the source systems. Monthly or quarterly this will be augmented with manual data from Excel spreadsheets, and updated indices will be calculated.

ComEd plans to use the much more timely health index data for various purposes, which may include:

  • Putting in place a short-term capital replacement plan for old assets with low health;
  • Developing a five- or 10-year capital plan for replacing as-sets projected to deteriorate over the coming years;
  • Considering the deferral of maintenance for assets soon to be replaced;
  • Increasing the priority of maintenance work for lower health assets that are not scheduled for replacement;
  • Performing root cause analysis for young assets with low health to determine why they are problematic and if they are recoverable;
  • Displaying the geographic dispersion of low health assets to determine if there are economies of scale in grouping maintenance or replacement of assets in close proximity with one another;
  • Using histogram and regression analysis to determine if there are key factors that are causing poor asset health;
  • Using trend analysis to determine the rate of change of asset health for individual assets and overall asset classes;
  • Calculating the probability of asset failure in the near term based on current health; and
  • Combining probability of failure with an analysis of the criticality of failure to identify high-risk assets that required urgent risk mitigation strategies.

Conclusion

By generating asset health indices in a much more timely and automated fashion, Bentley’s AssetWise Performance Management is providing ComEd with much better visibility into the overall health of its extensive physical plant. Health indexing is bridging the gap between short-term corrective work driven by condition-based maintenance and longer-term capital planning, which used to be driven by periodic one-time studies or last year’s budget and available capital. The foundation has been set for fact-based decisions on how to find the right balance among ongoing maintenance, capital replacement and overall risk mitigation.

Asset performance management unites the five elements of an asset management strategy.

Mark Biagi is solutions executive of process and power at Bentley Systems.

More PowerGrid International Issue Articles
PowerGrid International Articles Archives
View Power Generation Articles on PennEnergy.com
Previous articleCase Study: Thermal Energy Storage Saves California Winery Nearly $30K a Year
Next articleSmart Grid Asset Management-an Integrated Approach to Maximize Asset Utilization

No posts to display