There were no silver bullets that magically transformed PPL Electric Utilities' deployment of an automated meter-reading (AMR) system from an anemic effort into a sure-fire winner. It took almost two years of intensive work to develop the details and gain business acceptance for the project, which involved installing a power line communication system serving the utility's 1.2-million customers.
The business case included both hard and soft benefits. By following the revenue cycle process from the meter reader to the call center, PPL (Allentown, Pennsylvania, U.S.) identified the hard benefits, which involved such areas as full-time equivalent (FTE) reductions, vehicle costs and call center improvements. However, even with a thorough identification of the hard benefits, the business case was marginal at best.
The key to developing a successful strategy was gaining acceptance of the soft benefits. Particularly important were the benefits associated with theft deterrence and detection. No utility really knows or wants to admit how much electricity is being illegally diverted. Sources have suggested the industry standard is 1% to 3%. Even conservative estimates of savings through AMR will yield significant benefits.
PPL's experience during implementation was that benefits associated with theft deterrence were difficult to document. Even with instructions to look for evidence of tampering, installers found few instances, although this could have been due to guilty parties removing evidence of tampering once PPL announced it was coming to their area.
PPL needed a more economic solution to reading meters that could support the future retail market. A mobile system clearly would have been the most economical choice but would not have met future customer data needs.
A fixed network radio frequency, although capable of providing interval data, became infeasible when coverage concerns required some customers to be served by a mobile system. Consequently, a fixed network PLC system developed by DCSI (St. Louis, Missouri, U.S.) was chosen so that full communication was possible to all customers — an important consideration for future retail choice. The DCSI system was supplemented with a 6200-unit fixed network radio frequency solution developed by Comverge (East Hanover, New Jersey, U.S.).
Utilities historically have often been reluctant to deploy AMR projects, largely because of the perceived risks. Stranded investment, regulatory and business case concerns add to the risk equation. In fact, large-scale AMR projects have not demonstrated the same risk profile as many other company-critical information systems. For example, customer systems are complex installations that require a number of interfacing systems and significant change management efforts. The history of customer system deployment is littered with failed projects, significant cost overruns and huge business disruptions. As a result, it's imperative to do your homework before embarking on a large-scale AMR project.
|Risk Category||Description||# Identified|
|Product||System fails to perform||10|
|Supplier||Ability to perform||6|
|Change management||Ability to capture AMR process benefits||11|
|Benefits/business case||Benefits less than projected||5|
PPL's business case development included an extensive risk management strategy and plan, which identified 60 distinct risks (see table) to the project, ranging from technology failure to vendor insolvency. Each risk had a course of action identified should the risk materialize, and each team leader had the responsibility to track a portion of the risk plan.
Once a decision was made to deploy the AMR system, PPL found three major areas that posed a risk to the project. First, a contract needs to be in place to ensure the requirements of the business case will be met. Ideally, the utility and the vendor — DCSI in this case — are partners in the project, and each has incentives to achieve a successful outcome. DCSI must accept all the risk if the system doesn't perform to business requirements. Second, small-scale tests should be implemented as part of the overall deployment strategy to address specific questions or concerns.
At PPL, small-scale tests (10,000 meters) were built into the schedule to address proof of concept and system operation under adverse conditions.
As the project gets underway, logistics management becomes a major challenge in maintaining the schedule. At times, it seems an AMR project has a million moving parts, including meter control and management, meter exchanges and personnel management — all of which require tight management control and processes. Strong project management skills by both the utility and the vendor are key to avoiding schedule snarls.
The third risk area involves customer impact. A carefully crafted communication plan is necessary to inform customers of the deployment schedule to avoid excess calls from curious or concerned customers. Customers also need to be informed about the process for allowing installers to access meters with inside locations. Any questions or problems regarding the accuracy of the new system can be addressed by maintaining manual reads in parallel with system reads for several billing periods.
IT Due Diligence
Prior to selecting a vendor, PPL conducted a series of visits to companies that had deployed full-scale systems. Those visits were tremendously useful in helping to fashion an IT strategy. Key in this approach was the desire to minimize the impact on the company's billing system.
Billing code tends to be a lot more complex than the supporting metering systems. PPL targeted few changes in the billing system, so the new system provided the same data for billing in the same manner as readings uploaded from the manual process. To facilitate the interface with the billing system, as well as other interfacing systems, PPL developed a “middleware” solution. It managed the timing of data exchanges between the various interfacing systems including AMR, the manual metering application, billing, continuous load data and the deployment applications used by both DCSI and Comverge. This application simplified the interface design, insulated PPL's customer system from change and provided an alternative data source for meter data.
Startup, Implementation and Schedule
Securing the resources to install the new system as well as properly managing material and inventory is critical to a successful project startup. However, the key to staying on schedule is having the material when it's needed. Meters should be tested before they are released to the field.
PPL upgraded its bar code system to sample meters so that it would not need to test every meter. Without this upgrade, the meter lab would have been a big bottleneck in the supply chain. It's also imperative to have a system in place to ensure the physical integration of the meter and module. Otherwise, the meter will be reading for one account, and the module will be sending back usage data on another.
Once a supply of meters is in the hands of an installer, getting the meters installed efficiently requires the ability to group clusters of meters in a certain geographic area. Like many other companies, PPL's configured meter routes reflect the location and density of new development, available time and resources to read the meters. Previously, the ability to complete a route was more a matter of local knowledge rather than optimized programming. Meter route lines often arbitrarily bisected developed areas, frequently resulting in an installer completing one portion of a developed area only to leave to complete a route in a more rural location. To remedy this, the old meter routes were remapped in geographic grids allowing clustering by density, which significantly improved installation rates.
During implementation, PPL incurred some extra cost to avoid disruptions and ensure the project would be completed in the time frame desired. In particular, as installers replaced certain meters, they discovered damage to the meter base. In Pennsylvania, it's the responsibility of the customer to repair the meter base before a new meter can be installed. Anticipating that resolving these issues with customers would cause lengthy delays — or hinder the efficiency of the deployment — PPL decided to absorb the cost of repairing or replacing any faulty meter bases. In PPL's AMR deployment, more than 3000 faulty or deteriorated meter bases were found and replaced at company cost. At the peak of deployment, PPL installed more than 4000 meters a week. In fact, PPL's original 36-month deployment schedule was accomplished in only 30 months.
The size of PPL's project implementation team was surprisingly small. There was a project director and five team leaders responsible for the following areas:
This leader was responsible for the identification and resolution of technical issues associated with new metering hardware and communications equipment. This individual was supported by vendor and nonproject PPL metering support staff.
This position was most critical, as it managed both meter supply and implementing the overall schedule. PPL relied on a separate contractor to supply the meters and personnel to deploy the system. At the deployment peak — approximately 4000 meters/week — the implementation team consisted of 150 installers — both vendor and PPL personnel.
A certain number of installed meters must be read manually before being turned on in a fully automated mode. Once this occurs, the operations leader assures the system provides any billing reads necessary, troubleshoots the system and resolves any data and communication issues.
- Change management
This effort focused on making sure the new technology caused minimal impact to business processes and the people supporting them. The PPL change management team consisted of one experienced individual working closely with call center and field maintenance personnel on accessing AMR data and service and maintenance procedures.
- Information Technology
During the peak of system integration efforts, the IT team included 13 FTEs to develop the necessary interfaces to the company's billing system and several other systems related to managing meter inventory. Current IT support is about four or five FTEs.
The collective effort involved a project director, four or five program leaders, 10 to 12 analysts, IT supports and business analysts who managed the entire effort, not including vendor personnel. The principal solution provider also had an on-site team as well as a home office team providing software support. The vendor's on-site team was close in number to PPL's team.
In regards to system operation, PPL chose to operate the system in-house, which required transferring the operating knowledge from the vendor to in-house staff. PPL operators were engaged in the traditional knowledge-transfer approaches, such as training and hands-on experience. Especially effective was the combination of system acceptance testing with knowledge transfer.
DCSI and PPL devised a systematic approach involving 164 discrete tests. PPL operators, with DCSI support, completed the tests and documented the results. The appropriate PPL team leader then presented the results at a peer review between PPL and DCSI.
Although this process was somewhat time intensive, it established a strong sense of accountability and ownership, and proved to be an effective knowledge-transfer tool.
In developing the original business case, executive management wanted the project justified on the basis of the current business (i.e. don't include any evaluation of future benefits). However, now that the project is meeting the requirements of the business case, there's a renewed interest in leveraging the investment beyond the benefits of its original scope.
An unanticipated benefit was how the two-way communication could be used during storm-restoration efforts, which allowed local storm managers to determine if all meters were communicating before dispatching crews to another area.
The ability to record momentary outages is another benefit. This can be used to predict reliability and target more focused maintenance efforts, especially in the area of tree trimming. By monitoring electric use patterns, PPL can better protect revenue and block meter processes. Improved load data provides the opportunity to more optimally size feeders and line devices. Interval reads support time-of-use rates, and customers could have access to their own usage data and choose their own billing dates.
Bernie Bujnowski joined PPL Electric Utilities in 1978 and has been director of the AMR project at PPL since January 2004. When PPL Electric Utilities reorganized around an Asset Management-Resource Management model in 2001, Bujnowski was appointed resource manager. He also has experience in delivering technical solutions to the business, serving as the project director in the company's customer system implementation. He also served as project sponsor of a recently deployed work management system. His operating experience comes largely from the customer side of the business as call center manager and director of customer services. firstname.lastname@example.org