A short acronym such as MDMS can hide a list of implementation issues a mile long. With a meter data management system (MDMS), a utility can unlock the full value of its advanced metering infrastructure and obtain operational and customer-side benefits. But implementation, by many accounts, calls for consulting with peers in order to avoid the pitfalls.
So I invite our readers to join us this Thursday at noon EST for a webcast titled "Managing the Data Explosion: The Reality Behind MDM Implementation." (Click on the title to register.)
I spoke briefly with each of our distinguished panelists; a brief synopsis of their MDM issues follows here.
Jonathan Pettit is AMI manager for Oncor, a Texas transmission and distribution service provider and the sixth largest T&D operator in the United States, serving 3.2 million meters across 54,000 square miles. Oncor has deployed more than 2.4 million advanced meters in 2008-present.
By 2008, Oncor's original MDMS needed "a major overhaul," with new functionality, according to Pettit. Among its requirements for the new system, VEE (validation, estimation and editing) loomed large.
"Our concern was getting good estimates," Pettit told me. "Meter-read data is not going to be 100 percent actual data all the time. You can have outages, temporary communication issues."
Settling up with power generators and providing individual customers with usage data via the "Smart Meter Texas" portal sometimes requires estimation. Oncor's new system applies estimation rules that provide fairly accurate estimates, which can be trued up later as actual data becomes available.
"Real money" rides on the ability of the VEE engine to perform accurate estimation, Pettit said. "It's an expectation in the market that that data is accurate."
The MDMS at Oncor "probably has the most horsepower of any IT system in the company," Pettit added. And that drives change management across the company as the AMI/MDMS systems change all business processes, he said.
One lesson learned: change management requires a lot of training and documentation. Another: be prepared to add functionality, which means system-wide updates once or twice per year. Yet another: limit customization, which is expensive and time consuming. Another: you may eliminate meter readers, but now you need to hire data analysts and system administrative support.
"Utilities have a history of modifying everything," Pettit said. "You can't afford to do that anymore. We've set standards in IT that minimize any customization work. If you have to modify something, do it in the integration, not in the application."
Brian Novak is manager of advanced metering systems at JEA, which serves more than 417,000 electric customers in Jacksonville, Florida, and adjacent counties. JEA went live with its upgraded MDMS in November 2011.
"It's all about integration," Novak told me.
The MDMS is the conduit through which multiple data streams flow, such as a CIS (customer information system) and GIS (geographic information service), so if there's a problem, there's a ripple effect across the organization.
"It's incredibly critical that all this is set-up with the right method of transferring data," Novak said. "You need to be sure that if something is changed in one system, that's understood upstream and downstream."
One lesson learned at JEA is that integration can be arduous.
"This is tough work," Novak said. "You don't wave a wand and it all just magically happens. I would have loved, prior to our implementation, to hear some of the ugly things. Implementation matters."
Over-simplification is a hazard; don't underestimate costs, staff time and project implementation time, Novak told me.
"The onus is on the utility—this is your business—to know the MDMS and its implementation issues backwards and forwards," he advised. "Talk to your peers."
That said, getting the job done "in-house" has its pros and cons, he said. If you implement solely with staff, you develop expertise with the system that allows you to make nimble changes. On the other hand, that strategy makes a limited number of people into "big SMEs" (SME = subject matter expert). If you bring in integrators and vendors and they implement your system, one day they'll be gone, Novak pointed out.
Louise Gross is AMI system operator for PPL Electric Utilities, whose 2,200 employees serve 1.4 million customers over 10,000 square miles of central and eastern Pennsylvania.
PPL deployed AMI for the usual operational efficiencies, using power line carrier (PLC) for data backhaul, rather than a wireless mesh network. It encountered no opposition to its smart meter rollout, as it prepared its customers ahead of deployment, Gross said.
MDMS deployed in 2006 to "unlock the full potential of advanced metering." The return on investment (ROI) is attributable to both operational savings and customer-related programs. Operationally, Gross told me, MDMS at PPL is instrumental in four functions: forecasting and settlement, distribution planning and operations, billing and revenue protection. On the customer side, MDMS feeds customer service reps who deal with customer queries, enables personalized energy use strategies and aids program promotion.
Among PPL's lessons learned, according to Gross, was that the utility needs to manage all aspects of an MDMS project. It's advantageous to jointly develop strategies for AMI and MDMS. Internal customers need to be engaged early and often. Achieving business case objectives requires well-established metrics.
All three panelists will share the nitty-gritty of implementing MDMS on the webcast. I invite you to join us for "Managing the Data Explosion: The Reality Behind MDM Implementation." (Click on the title to register.)
Intelligent Utility Daily