Elizabeth Bray, P.E.
Times have changed and how we manage substation equipment has to be a lean mean operating machine. Lean in the way of operating efficiently for necessary maintenance along with limited resources, mean in use of creativity and technology to keep the equipment operating safely. This is not a novel approach; however, the challenges of reducing operational maintenance cost each year with no additional resources to grow the effort alters the dynamics of how Southern Company Transmission is surviving in the area of substation asset management and making progress in the area of condition-based maintenance and asset health as it relates to substation asset management with a one-woman team who is creatively using technology to grow a condition-based maintenance and asset health program for the substations in the Southern Company territory.
In this era of technology there is a constant demand to do more with less resources under the expectation that the resources are to be a jack of all trades without being just a master of one. The challenges are growing in the utility industry. A few that should resonate with any reader are how to attract young talent, retain employees, and how to handle the aging asset population with shrinking operating and maintenance dollars.
Some of these issues can be addressed with the move from a time-based maintenance approach to a condition-based maintenance approach. An example of time-based maintenance is when a substation oil-filled circuit breaker gets a preventive maintenance check every two years. The condition-based maintenance approach would schedule maintenance on this oil-filled circuit breaker based on triggers such as breaker operation speed, number of breaker operations, or after a cumulative fault energy is reached. The difference is focusing maintenance dollars on corrective actions instead of preventive or reactive actions. Even though it seems like a good approach to change maintenance to condition-based it is not simple to change a process quickly.
Introduction to the Challenges
Condition based type of maintenance program has proven a winner across many industries with the utility business seemingly the last one to the table. At least the writer’s utility is facing these challenges. Readers who have not faced the following challenges should consider themselves lucky. Why is the electric utility behind? One reason is the shifting of gears to change how the utility has operated its maintenance programs for the length of time they have been in business. Another driver is money. Depending on how the accounting rules apply this shift is considered operating and maintenance (O&M) dollars and capitalization of such an effort is very difficult. Another challenge is data. Having access to data to determine an asset’s condition as well as the analytical tools required to make this an automated process, require the need for dedicated resources. Dedicated resources are a challenge and it does come down to a need for dedicated resources with dedicated dollars.
The reason for this paper is to share the challenges of resources around moving condition-based maintenance forward. Condition-based maintenance is a part of substation asset health which leads to managing the substation assets from a maintenance and a capital replacement standpoint. The goal here is to help others realize the challenges and learn from the team of one to advance their own programs as a team of one or with additional resources to advance their program.
The Team of One
A sports team cannot win games alone; it is a team effort. However, in the eyes of this author the condition based maintenance program is a team of one. The challenge is to succeed and show value. The ability to show value is to help executive see the need to increase the team membership to more than one player.
The team of one has multiple roles. At the time this paper was written there were around 1000 assets (transformers, circuit breakers, and battery banks) that are a part of the program. The team of one has a few jobs such as historian administrator, backup historian administrator, main historian programmer, head of dashboard configurations, only predictive analytic model builder, contact for the field installations, sole person to vet the active alarms, contact for monitoring vendors, holds list of all online monitors, tracker of monitor related issues, contact for vetting potential future monitors, in charge of geomagnetic disturbances (GMD) data reporting, author of monthly reports, and have several industry roles as well. This is all the same person.
How does this “team of one” operate? One key is work prioritization. This is a “first in first out” process with a few exceptions. Priority is given to GMD during a solar flare event. Another priority is compliance work which involves battery monitoring data and soon to involve GMD data reporting. Several online monitor projects also receive priority due to reporting and alarms only issued by the team of one. Typically, five to ten monitors are brought online each week. The team of one does understand that the integration of new data does take away from other responsibilities. This prioritization does step on toes; however, no additional funding has been allocated in the stepping on toes and everyone has to understand that a team of one can only work so fast.
Another key is a sense of humor. This is a stressful job. There is a lot of balancing and depending on other groups to get data set up correctly. Those that provide this support are awesome and without them, the team of one would be data-less, and proving a case would be very difficult, if not impossible. The sense of humor is needed to lighten the mood and the ability to laugh at yourself is very important. A wise engineer once told the team of one that you can talk to yourself but just do not answer your own questions. The team of one cannot follow that wise advice and holds team meetings usually in the car or during a run.
A third key to keeping the team of one afloat is the use of technology. Technology comes in many forms. One is online monitoring equipment such as dissolved gas analysis (DGA) monitors or SF gas monitoring. Technology around the data historian is an important key in keeping the team of one able to automate alarm notifications. This gives the team of one the ability to rely on the data historian like a member of the team. There is an increase in monitoring data. No complaints because this is welcome growth to show the value in this process. The data from these monitors is the team of one’s responsibility. No alarms are sent to the control centers and all alarms notifications are programmed by the team of one using historian software tools. These notifications are sent to subject matter experts and maintenance engineers. This enables the team of one to not monitor the monitor data and reference to the data when vetting notifications is required. Today it is about visualizing your data and letting the data tell the story. Software tools allow that story to develop whether it’s simply viewing the historical and current data in a user-friendly format via the internet or the ability to slice and dice data from multiple locations to show asset health, maintenance history, asset information, and available data for that device. The ability to use software technology helps share the story of condition-based maintenance to the subject matter experts and allows them to learn more about the assets they manage.
Another key is the use of data analytics. Currently, the team of one is limited on data analytics mostly due to resource restraints. Advancing this field will be key to success around condition-based maintenance and understanding the health of our substation assets. Analytics involve algorithms.
Managing Substation Assets
One way to show value is to produce asset health indices for substation assets. The team of one worked with in-house subject matter experts to develop asset health algorithms for transformers and circuit breakers. These algorithms are currently reviewed annually. These algorithms generate health scores and are currently based on available data. Available data varies but usually consists of offline test records, asset nameplate data, and limited relay data. For transformer assets, health algorithms are currently tied to offline test data and asset nameplate data. For breakers, the asset health algorithms are based on limited offline test data, nameplate data, subject matter expert ranking, and limited relay data. The added challenge is the relay data and getting the event file parsed to extract key pieces of data such fault magnitude values. The team of one will admit that there are many holes in the offline test records and there are challenges that are being addressed from a maintenance standpoint for the need to accurate data recording into the asset management system. The algorithms will grow and become more complex as more types of data are integrated into the condition-based maintenance program. The breaker asset health scores are 90% automated to pull data from sources and display the scores on a website while storing the scores in the data historian. The transformer asset health scores are 100% manual and the team of one has some work to do to advance the automation process as time allows.
This approach to applying condition based maintenance was not intended to be a team of one. A dedicated resource was added to the organization to explore the thought of condition based maintenance and the proof was in the pilot five years ago. To the credit of the team of one’s organization the complications due to politics and funding even with a successful pilot that showed value and the potential for future value added to the organization were not enough to push forward with growing the team of one.
The future is not charted at this time. The goal is to move this effort forward to all substation assets quickly and give subject matter experts, maintenance engineers, and operations another tool in their toolbox to better understand their substation assets for maintenance and future capital replacement projects. The added knowledge that this program will provide will be important from a safety, maintenance, and operational standpoint and drive by a team of many.