Moving to Predictive Maintenance with IIoT in Power Transmission Substations

The presentation “Moving to Predictive Maintenance” was given by Jeff Fleeman, Director, Advanced Transmission Studies & Technologies, American Electric Power, during the ARC Forum session, “Moving to Predictive Maintenance with Industrial IoT (IIoT).”  He focused on predictive maintenance for power transmission substations.  Electric power is one of the most mature industries for maintenance practices, with uptime a key objective enforced by the associated regulatory agencies.  American Electric Power (AEP) has been in this industry for over 100 years, and is now the largest power transmission owner in the USA with over 40,000 miles of transmission lines and 3,500 associated T&D substations in 13 states.

Business Driver

The design of the transmission network has built-in redundancy for the power lines. However, not all substations and associated transformers have this redundancy and therefore become more critical to maintaining high power availability for AEP’s customers.  Assets in the transmission network are widely distributed, and it takes significant time to travel to the assets.  For maintenance, situational awareness has become critical to set priorities and deploy technicians.  Related asset management issues include:

◾ Aging infrastructure and workforce with experienced technicians retiring drives more maintenance activity with less  experienced people

◾ Significant growth in the number of assets with only incremental maintenance budget due to competitive and regulatory constraints

◾ Time based maintenance becoming unsustainable as workload exceeds the capacity of the available technicians

◾ Manually collected and entered inspection data is too limited and slow for the needed real-time monitoring of asset health

◾Lack of commercial off-the-shelf (COTS) solutions for analytics and on-line monitoring of transmission sub-station assets

◾Outage arrangement and permission approval procedures constrain and limit access to take assets out of service for  maintenance

◾Rising expectations for less outages

Previous attempts at on-line inspection and monitoring had been unsuccessful. A common problem involved targeting specific monitoring tasks that were easy, but not the most critical.  These (often vendor-driven) solutions required significant maintenance support, rather than less, and were not scalable across the larger enterprise.  Also, the gathered data was complex, often in proprietary formats, isolated from other data systems, and required on-going engineering analysis and investigation to identify issues.

Transformers have a “bathtub curve” failure rate that grows rapidly after 40 years. Unfortunately, AEP’s population of transformers are entering this area of increased risk.  Allowing outages to increase is unacceptable from both the regulatory and customer satisfaction perspectives.  Replacing them all would be too costly, and the budget is not available to increase maintenance to the degree needed to avoid an increase in outages.  Another solution was needed.


For a solution to this dilemma, AEP partnered with ABB in a four-phase, multi-year program to develop ABB’s Asset Health Centre (AHC) to fit the company’s power transmission and substation assets. The solution automates data collection and analysis, evaluates remaining life, and recommends actions using an asset health index.  Now, maintenance occurs based on condition rather than a time interval.  The program’s goals were:

◾Prevent failures where possible and avoid surprises

◾Optimize the maintenance effectiveness: right place, time, and people

◾Provide data to support a good criteria and rationale for replacing assets

The platform implementation was completed in December 2015. The types of assets covered include transformers, circuit breakers, and batteries.  Now, standardized monitoring packages are included on all new extra-high-voltage (EHV) equipment, and standardized retrofit packages are being rolled out in stages.  AEP started with the bulk transmission assets because they have the biggest cost and could more easily justify the new systems.  Growing economies of scale and new learnings have reduced the cost for each asset added to the system.  The reduced incremental cost enabled wider application in distribution and further economies of scale.

Predictive analytics for assets involves these components:

◾Collect: Unified data communications, analytics and management

◾Predict: Algorithms codify experience to predict and prioritize risk

◾Inform: Alerts, KPIs, and decision support

◾Act: Right process for the associated equipment.

Information from offline systems (age, supplier, and maintenance history), SCADA (real-time operating parameters) and online sensors (specific for asset health) goes into performance models containing algorithms to assess the health of each transformer. Since ABB manufactures substation equipment, and services its own and other vendor’s equipment, its deep knowledge allows it to provide packaged algorithms that can be customized for specific assets.

Though all the IIoT data in aggregate qualify as “Big Data, a “small data” approach is deployed. This focuses the algorithm for a type of asset (like high-voltage transformers) on the data associated with a specific asset (a particular high-voltage transformer) to determine that asset’s health.  It tailors the health assessment process for the asset type and each specific asset.


The key benefit of this asset health center at AEP involves assuring higher uptime by assessing each asset’s risk, and mitigating a problem at the right time prior to failure. This reduces outages, prevents collateral damage and improves safety.  Also, the IIoT data provides the means for a fact-based assessment and justification of when an asset should be replaced.  With AHC, AEP expects to be better able to absorb the maintenance of the aging fleet of substation equipment without a significant increase in maintenance costs.

“Reprinted with permission, original blog was posted here”. You may also visit here for more such insights on the digital transformation of industry.

About ARC Advisory Group ( Founded in 1986, ARC Advisory Group is a Boston based leading technology research and advisory firm for industry and infrastructure.

For further information or to provide feedback on this article, please contact

About the Author:

Ralph Rio

Vice President, Enterprise Software, ARC Advisory Group, Boston

Ralph’s focus areas include Enterprise Asset Management (EAM), Field Service Management (FSM), Global Service Providers (GSP), and 3D Scanning systems & software.

Share This Post

Leave a Reply