“What? Oh yeah, that special switch near the electrical panel. OK. I got it; will do.”

Is this a conversation your customers have at the end of the day? No. Are you sure? You can be. This scenario could already be happening with your customers via meter tampering. Do you have a system in place to identify such tampering?

When an automated meter reading (AMR) system is installed, some customers become more savvy and creative in their attempts to save money by tampering with their electric meters. PECO (Philadelphia, Pennsylvania, U.S.) has been able to outpace even the most unscrupulous customers by leveraging data from its AMR system to identify potential theft and tamper conditions.

In 2004, PECO completed the deployment of a fixed network AMR system from Cellnet (Alpharetta, Georgia, U.S.). The PECO AMR network currently reads more than 2.2 million meters daily. This system is designed to deliver both energy consumption and meter-data tamper flags such as reverse rotation, magnetic fields, outage notification/restoration and outage counts.


Prior to deploying an AMR system, revenue protection teams relied on traditional methods to identify meter tampering and potential theft. The meter readers who checked the meters each month were the front-line forces. Beyond the meter reader, techniques such as proactive sweeps and irregular usage investigations were used. Additionally, hot lines and tips are good sources of tampering leads.

PECO no longer uses meter readers as a result of the AMR implementation. While this has eliminated a key identifier of theft, PECO still relies on other traditional methods of theft detection and resolution.

PECO's revenue protection organization conducts proactive area sweeps where theft is suspected or has occurred before. During a proactive area sweep, field personnel physically inspect all services in a specific geographic area. While some theft is identified and resolved, the deterrent value of a visible PECO presence is effective.

Resolving theft at a specific premise is only the first step. PECO also investigates any other premise that might be associated with that thief. This collateral investigation ensures all theft associated with an individual is detected and corrected.


Once the AMR system was installed, PECO quickly realized its need for new, intelligent methods to identify theft and energy diversion. The focus turned to the AMR system.

AMR meters provide various flags and indicators of theft. Reverse rotations, outage-count indicators, unplanned outage notifications and usage on previously cut meters all identify good targets for field investigations. AMR indicators are often corroborated by reports from PECO customers that theft is occurring near their premises.

Throughout the past year, PECO has been working with Cellnet to design several reports that identify conditions indicative of meter tampering. These reports guide the field revenue protection forces in deciding which premises require on-site inspection. This effort has yielded significant improvements over traditional meter investigation methods. Furthermore, there is a measurable increase in metered usage and associated revenue for those meters that have been successfully investigated (Fig. 1).

These reports were piloted during June and July of 2005. Field visits were used to validate the veracity of the reports. The pilot reports successfully identified tampering on 70% of the meters targeted. For many of the remaining 30% where tampering was not validated, usage increased after a field visit by an investigator.

Based on this success, PECO integrated these reports into its existing revenue-protection processes. Cellnet delivers the reports on a monthly or as-needed basis. The reports continue to be successful; so far, theft has been validated at 90% of the premises identified.

  • The Unplanned Outage Report identifies meters where there are numerous or excessive unplanned outages. This report identifies meters with greater than 10 outages occurring during a 30-day period but is not designed to look for specific patterns. In this case, a customer has likely installed an illegal switch that, when activated, reduces or eliminates the flow of electricity through the meter register. This sequence is illustrated in Fig. 2. Approximately 40% of the theft identified by the AMR reports is identified with this report.
  • PECO uses a four-day billing window. Any reading obtained during this window can be used for billing purposes. The Billing Window Report identifies a meter that is turned off either outside of the billing window or during the billing window. Customers who choose to turn off the meter outside the billing window may be trying to convince PECO that the limited usage recorded during the billing window reflects usage for the entire billing period. This sequence is identified in Fig. 3. Customers who choose to turn off their meters during the billing window may be trying to force PECO to estimate their bill. Such customers may hope that the estimate will be low and will reflect less than actual usage. This sequence is identified in Fig. 4. Approximately 35% of the theft identified by AMR reports is identified with this report.
  • The Reversed Meter Report is designed to identify those customers who have removed their electric meter and installed it upside down with the intention of running the register in reverse. When the meter is initially removed, an unplanned last-gasp outage message is generated. This is typically followed up by an unplanned power-up restoration message and a new reverse-rotation tamper flag. This sequence is illustrated in Fig. 5. Approximately 20% of the theft identified by the AMR reports is identified with this report.
  • The Weekend Usage Report is designed to identify customers who regularly tamper with or disable their meters on the weekends. The report highlights meters that exhibit stale usage (no reading after a specific period of time has elapsed) followed by a power-restoration message. Meters are included in the report if this stale usage pattern occurs on consecutive weekends. This sequence is illustrated in Fig. 6 Meters with intermittent communications are eliminated from this report by reviewing the customer's usage on Monday. Meters with communications gaps are expected to register a larger-than-normal volume of usage when communications are restored and the meter “catches up” to reflect weekend usage. Less than 5% of the theft identified by the AMR reports is identified with this report.
  • As you can see, an AMR system and its associated data analysis tools easily identify these scenarios. Additionally, the reports can be quickly modified to reflect changes in customers' tampering patterns.


PECO's revenue-protection organization is responsible for identifying and resolving theft. Once identified, revenue protection personnel visit the suspected premise and verify the existence of theft. The personnel are trained to resolve most theft situations safely and return the premise to an untampered state. This may require the replacement of the meter or corrections by the customer to their service entrance. Customers who fail to cooperate with theft resolution or who are repeat offenders may have their service disconnected.

In cases where theft has been validated, customers are billed for the time and materials necessary to resolve the theft situation. Historical usage data is used to identify the approximate start of theft. The billing department will then rebill customers back to this start time to reflect appropriate nontheft usage and recover lost revenue. During the pilot, rebilled kilowatt-hour usages were greater than the tampered usages by more than 90%.


Theft detection and resolution has produced real, quantifiable benefits. Losses from theft are conservatively estimated to equal 1% of a utility's total revenue. The recovery of part of this loss results in a direct contribution to the bottom line.

Furthermore, more accurately identifying theft can increase the overall productivity of field forces and billing analysts.


While this article has discussed several known methods of meter tampering, it is of utmost importance that no one attempts to tamper with an electric meter. Any alteration of a meter or service entrance as a result of theft automatically places anyone near that meter at risk. This includes the tamperer, other customers and the utility worker required to maintain electric service. Resolution of tampering lessens the chance of a malfunction or unexpected service situation that can lead to damage of property and injury or death.

Furthermore, all personal safety gear must be used when working on energized electrical equipment, including electric meters. Remember, tampering with an electric meter is dangerous, illegal and unsafe, and drives up costs for all utility customers.


The authors would like to thank John Kratzinger and his Revenue Protection team. They are the front line in PECO's efforts to identify and resolve meter tampering and theft at PECO.

Glenn A. Pritchard is a project manager on Exelon's Meter Reading Technology team, where he specializes in finding new uses for PECO's AMR system and its data across multiple business units. Most recently, Pritchard led an initiative to link PECO's AMR and outage management systems. He has been with Exelon and PECO for 14 years, and his experience ranges from distribution automation to reliability engineering. Pritchard holds a BSEE degree from Clemson University, and is a registered professional engineer in Pennsylvania and a member of the IEEE Power Engineering Society.
[email protected]

Jeff Evans is a project manager on Exelon's Meter Reading Technology team, where he is currently evaluating opportunities to improve and expand the utilization of AMR data across multiple business units. He led the initiative to evaluate and recommend the deployment of AMR technology within the ComEd service territory. Evans has held positions with Exelon Energy Delivery's Account Management, IT and marketing organizations over the past 14 years. He holds a BSME degree from Bradley University and an MBA degree from the Kellogg Graduate School of Management.
[email protected]