How Advanced Data Analytics Can Help Optimize Your Asset Health Management

How Advanced Data Analytics Can Help Optimize Your Asset Health Management

Data-driven digital transformation is changing the way utilities do business, with analytics and AI tools driving innovation across the industry.

One key area in which data analytics can be a game-changer is asset health management. Utilities are already transitioning towards Condition Based Maintenance (CBM) to address the ever-increasing load on critical assets and the growing cost of failures. The cornerstone of a successful CBM strategy is the ability to accurately assess and predict the health condition of each and every asset (transformers, cables, etc.). This is where data analytics can provide distinctive added value.

As sensing technology continues to improve and sensors come down in price, they are becoming an integral part of asset health management. But installing the sensors on your transformer fleet and collecting the data is the easy part. The hard part is making sense of these huge amounts of data in real-time in order to provide timely and actionable insight related to asset health. The adoption of big data analytics techniques, like machine learning, can help utilities achieve these goals, while driving cost savings and mitigating risk.

Leveraging Dynamic Data

Data analytics enables utilities to leverage the dynamic (online) data being provided by online DGA (Dissolved Gas Analysis) and other sensors in their transformers. Unlike static DGA sampling performed manually once or twice a year, online DGA sensors take periodic (e.g., hourly) samples and provide real-time results, giving operations teams clear situational awareness at all times. According to a recent Grand View Research study: “The global dissolved gas analyzers (DGA) market is expected to reach USD 1.11 billion by 2024, expanding at a CAGR exceeding 12.0% over the forecast period."

Utilities' adoption of condition-based maintenance (CBM) reflects the inappropriateness of traditional statistical-based tools for predicting transformer failures. This is due to the relatively low rate of failures in power transformers (measured at about 0.1% per 5 years in recent studies). In the absence of a substantial pool of statistical data, utilities require more specific data regarding asset health condition – hence the transition to a CBM-based approach based on the use of online DGA and other sensors. Such an approach takes into consideration the specific conditions of the asset for assessing health and risk, rather than the overall statistics.

The growth of online DGA sensors is being driven by the need to reduce maintenance costs, as well as improvements in data analytics capabilities. Deterioration in critical assets may happen over a period of time (e.g., month) that is much shorter than the sampling rate (annual or semi-annual in most cases). Using online DGA sensors in conjunction with big data analytics, it is possible to identify trends, detect abnormal behavior and get an early warning even before thresholds are reached (“pre-threshold warning”). Early warning makes it easier and cheaper to rectify issues, reducing the cost of failures and improving maintenance budget utilization.

Moreover, by collecting more data from various sources, and then correlating that with various industry standards, tools and models, the accuracy of asset health assessments is significantly higher. This allows utilities to establish solid health predictions (e.g., three months ahead) for decision-making.

Beyond DGA

The probability of transformer failures increases with age and can be caused by a variety of internal and external factors (see chart). Research conducted by industry experts shows that DGA can help detect about 30-50% of the power transformer failures. Accordingly, utilities are in the process of applying other sensors of various types such as partial discharge (PD), load, moisture, IR cameras (for temperature), Tank vibration, GIC, Acoustic emission and RFI, as well as monitoring the bushing and the OLTC. This trend will grow as sensing technology further develops and becomes a commodity.

(Source: http://www.engineeringworldchannel.com/transformer-failure)

Extracting online and offline measurements from various sources and correlating them to establish an accurate health index for each asset requires a robust and comprehensive analytical system that can effectively integrate with all the various systems and sources. By correlating data from various sources and sensors (e.g., temperature, load) as well as results from different industry standards (e.g., IEEE C57.19.01:2000, IEC 60599:2015), utilities can harness the power of data analytics to gain valuable insight into asset health.

Data-Driven Cost Savings

Data analytics helps utilities save costs by establishing significantly more accurate health condition indices, which are the basis for an effective CBM program. This improved accuracy can translate into a significant portion of the annual maintenance budget and of total costs associated with asset management (reactive and proactive).

In addition, by replacing manual sampling with online analysis, utilities can expect to achieve cost reductions in the following areas:

  • Costs associated with unexpected failures such as: outage costs (can reach hundreds of thousands of dollars per day according to some research), loss of revenues, cost of alternative sources of energy, cost of asset replacements and environmental damages, emergency repair and insurance costs, etc. According to industry studies, online monitoring of power transformers may reduce the risk of catastrophic failures by 50% and cut the associated repair costs by 75% due to early alerts.
  • Accurate health indices help assess and in some cases extend the operational life of critical assets based on solid data and models, while preventing unnecessary failure costs and dramatically reducing maintenance costs.
  • Operational savings through improved efficiency in managing and maintaining a large fleet of assets, reducing the investigation time to understand developing problems and their root cause (by applying more data and more analytical tools earlier in the process), and improving the quality of decisions taken by engineers and documenting insights and conclusions.

In light of the damage and financial costs associated with failures of transformers and other critical assets, utilities have a vested economic interest in incorporating advanced data analytics as part of their CBM strategy.

Leave a comment

You are commenting as guest.