Share on email
Share on twitter
Share on linkedin

Individual Capacitor Monitoring for Intelligent Capacitor Bank Diagnostics

Presented By:

William A Munn and Joshua Nikkel

General Electric and Southern Company Services

TechCon 2022

Capacitor bank monitoring is typically limited to the whole bank, per phase, or per string monitoring. The lack of granular information about individual capacitors hampers attempts to predict capacitor bank failure and contributes to capacitor bank maintenance that is a lengthy, labor-intensive process of manually testing each capacitor. The results of which are increased bank restoration times, O&M spending, and forced outages. The innovative technology presented in this paper is designed to solve these issues today. The technology is comprised of individual sensors that monitor each capacitor and can detect and report granular changes in the local capacitance, such as those resulting from the failure of a single pack. The health status of Sensors in a bank is collected and stored by a dedicated Data Aggregator and is available locally to field teams and remotely via industry-standard communication protocols.

These sensors are designed with regard to longevity, ease of installation, ease of use, and integration with existing data systems. No wiring needs to be added to the capacitor bank at installation; the sensors harvest power locally and transmit data wirelessly.


High voltage transmission substations require regular maintenance and assessment of assets in the grid. Effective management of the electrical grid assets allows utilities to reach some of the goals such as:

  • Optimize costs
  • Ensure safety for operations
  • Extend the life of transmission line assets
  • Ensure reliability and maintain customer trust
  • Compliance to continually updating regulatory requirements

Condition-based maintenance and reliability-centered maintenance program is considered as a proactive approach to asset management in the energy sector. The condition-based maintenance strategy optimizes the economic benefit of transmission line physical components. It leverages available technology, like digital monitoring and data collection, to extract accurate information about the assets. The data are then extrapolated to predict future failures. This approach may seem basic, yet it is not implemented often enough in actual practice. The conventional approach for even large players was reactive maintenance, which addressed defects and issues after the fact.

Implementation of Condition Based Maintenance Strategy provides below benefits

  • Provides the insight required to shift from costly time-based maintenance to cost-effective and focused condition-based maintenance
  • Reduces ambiguity and the need to speculate regarding budget requirements
  • Reduces OPEX (Operational Excellence) usage on assets to a minimum

When it comes to capacitors: high voltage capacitors are the key elements in improving the efficiency and reliability of the grid. They provide simple and reliable reactive power to improve system performance and quality. In this paper, we will focus on high voltage capacitors, challenges of Capacitor banks, and how Real-time monitoring of Capacitor banks help in Proactive condition-based maintenance strategy. The solution proposed in this paper will help address some of the challenges of Capacitor Banks and has been validated in a field application scenario.

Challenges of Capacitor Banks

capacitor bank image

The statement that “Capacitor banks are generally expected to not require regular maintenance” is not often true. Certain external or internal conditions might result in causing catastrophic failures of Capacitor banks such as sharp transient voltage and temperature variations, internal stress, and pressure caused by overheating. Therefore, regular maintenance and inspection of Capacitors banks is inevitable and varies from Utility to Utility based on their applications.

Why Monitor Capacitor Banks?

Below are some of the reasons why Capacitor Bank monitoring is essential

  • The need to reduce bank restoration time and O&M spending’s.
  • Avoid emergency outages forcing rescheduling of planned jobs.
  • Prevent capacitor failure due to transient voltage & sharp temperature variations.
  • Monitor ageing asset infrastructure.
  • Difficulty accessing remote locations and or during harsh weather conditions.
  • Crew safety and liability is a priority.
  • Loss of asset historical knowledge as company technical experts retire.
  • Avoid financial penalties for non-delivery of reactive power.

Current Methodologies for Failure Detection:

A key concern with respect to capacitor banks is identifying an individual failed capacitor in a bank and reducing capacitor bank restoration times. Currently, a few methodologies are employed to detect individual capacitor failures.

  1. String Monitoring: Current transformers and relays are installed to detect changes in the current of a string of capacitors. This method is able to identify which section of a bank which has a failed capacitor but does not help determine which exact capacitor failed. Identifying which individual capacitors in a given string have failed must be done manually in the field.
  2. Phase Monitoring: Additional capacitors and voltage transformers are installed between the bank and the neutral phase. These protection modules detect leakage current from a given phase to the neutral. When the leakage current exists, it can be determined that one or more capacitors have failed, and in which phase the failures have occurred. Like String Monitoring, this method is able to identify which section of a bank which has a failed capacitor but does not help determine which exact capacitor failed. Identifying which individual capacitors in a given phase have failed must be done manually in the field.
  3. Externally Fused: Each capacitor in a bank is equipped with an external fuse with visible indication. Unlike String Monitoring and Phase Monitoring, individual capacitor failures can be easily identified. However, a single blown fuse usually indicates multiple failures in a group of capacitors. Manual testing of all capacitors in the series group with the blown fuse is necessary. It is important to note that externally fused capacitor banks require a larger space for installation and are not suitable for all applications.

Regardless of which existing methodology is used, a capacitor failure necessitates several person-hours spent identifying the source of the failure detected. This is especially true for large banks where the process may take a field team two or more days and involve two or more days of capacitor bank outage. Additionally, the current methods of failure detection are course, often not triggering until more than one failure is present. These drawbacks were the impetus in pursuing a new method of capacitor bank monitoring.

Capacitor Health Monitoring System

Sensor diagram

Capacitor Health Monitoring System allows to detect the exact failed capacitor Can in a Capacitor Bank. Unlike the traditional methodologies of failure detection, The Capacitor Health Monitoring System monitors Capacitors in real-time allowing utilities ample time to plan outages and detect failures ahead of time. The Capacitor Health Monitoring System is comprised of two main components, the Cap-Can Sensor, and the Receiver. The Cap-Can Sensor performs the majority of the “heavy lifting” by continuously sampling the voltage across and the current flowing through the capacitor. It then computes the impedance in order to provide the indication for the health of the Cap-Can. The Status is indicated on the sensor via LED indicator and a mechanical flip dot. The health status data can be wirelessly transmitted on request to the Receiver. The Receiver then in turn can transmit the data to any other systems via TCP/IP for remote monitoring purposes. The sensor is truly wireless and does not need any additional wiring installed for either power or communications.

Solution Highlights:

  • Visual indication of failure mode
  • RMS Current & Voltage of connected cap cans
  • Historical trending
  • Enables remote notifications and alerts of capacitor failures and warnings
  • Enables remote updates of sensors
  • DNP3 enabled
  • On-premise or server data logs possible

Failure Detection Scheme

The current single Can Sensor will be attached to Capacitor Can bushings and contains a radio chip that can provide Capacitor Can status to the receiver that allows electronic monitoring of the entire bank at one location. This receiver communicates wirelessly with each sensor to gather data and displays it on any attached laptop or tablet. The Sensor will measure the difference of the impedance of shorted stacks within each Can. This difference will cause an electronic signal to be sent to the ground receiver to notify the operators of an impending can failure. Once data is received, the receiver will then exhibit the Can status (either yellow or red) depending on the severity of the stack(s) that failed. This status can be viewed by maintenance crews at the local substation station or can also be transmitted to the central control room via either a separate system or SCADA as a visual or audible alarm. Once the Capacitor Bank is de-energized and the Can replaced the alarm is extinguished and the sensor can be re-used on the new Capacitor Can. The sensors and receiver module can have their firmware updated and future software applications added from the receiver box located outside of the extended reach area without taking an additional outage. The sensor is easy to install using its standoff adaptors placed on the Capacitor Can Bushing. Just screw in the standoff adaptors to the top of the bushings (through animal protection if applicable).

Field Application Study

capacitor unit image

Southern Company made a shift from fused to fuseless capacitor units in the early 2000s. This shift helped improve reliability of the system and to help maintenance by reducing the number of capacitor units in the capacitor bank. The result of this change allowed them to implement an unbalance protection scheme to give a warning to the control center of an issue with a capacitor unit prior to a failure that would cause an outage. The control center could prioritize an outage and send a maintenance crew to the location to replace the failing unit. Changing from fused to fuseless units reduced the number of units in the bank but the visual detection of a failed fuse was no longer available. The relaying scheme could indicate the phase that had a failed unit, but each unit had to be troubleshot to determine what capacitor unit failed.

Southern Company then installed the Capacitor Health Monitoring System as pilot program, placed 48 sensors in a single capacitor bank at Carrolton, Georgia. The developed Capacitor Unit Failure Sensor (Cap Sensor) was attached to each unit. The Cap Sensor measured the voltage and current to detect the change in the impedance of shorted packs within each unit. This difference caused an LED light to be illuminated and a non-volatile/mechanical indicator to change states to notify the maintenance crew which unit has failed.

The Carrolton capacitor bank has been monitored in real-time by Capacitor Health Monitoring System since November of 2018. Over the course of a year, a total of 3 capacitors have been identified and confirmed in various stages of failure prior to the entire bank being tripped. Alerts of impending Cap Failure allowed Southern Company to plan and replace the Capacitor Units prior to catastrophic failure and unplanned outages. In effect, the Capacitor Failure Monitoring System prevented unplanned outages, decreased failed capacitor unit identification times, and overall helped to reduce O&M cost to Southern Company.


As more utilities are moving away from reactive-based maintenance strategies to adapting to condition-based maintenance strategy for asset management, predictive maintenance has been a key aspect for consideration. The ability to not only detect failure but be able to predict or prevent failure has started to gain importance. This Capacitor Health Monitoring System aids in not only providing utilities the ability to detect the exact capacitor failure and reduce the bank restoration times, but also allows for predictive maintenance. By gaining access to the health status of the individual capacitors remotely, utilities can schedule a maintenance outage ahead of time. Moreover, utilities that install Capacitor Health Monitoring Systems such as the one proposed in this paper will begin aggregating granular data from their capacitor banks. This large amount of data can be used to drive future insights in preventing capacitor failures. The result of this technology, therefore, is a more reliable and intelligent capacitor bank that experiences less failures, requires less maintenance when failures do occur, and supports future development in capacitors.

Join our email list

We use cookies to give you the best online experience. By using this website you agree with our cookie policy.