Xcel's smart grid a "learning lab" for dos and don'ts
The company's resulting SmartGridCity project has since indeed set an example in the electricity industry after missing its 2009 target for completion and running three times over the utility's original $15.3 million budget, according to filings with the state Public Utilities Commission.
Now, as other utilities test new ways to improve their own reliability and efficiency, energy experts say they are looking at Xcel's Boulder experiment as both a trailblazer and a cautionary tale.
"It's being used as kind of a poster child of how not to do smart-grid deployment," said Michael Shames, executive director of the San Diego-based Utility Consumers' Action Network, a Southern California watchdog group.
There is still no consensus on what constitutes a "smart grid," but by any definition, Xcel set out to be at the head of the class with SmartGridCity. Beyond two-way meters that periodically relay customers' usage data, Xcel installed equipment that reports outages automatically, substations that communicate with each other at broadband Internet speeds, and plugs that allow hybrid vehicles to serve as batteries for the entire grid.
"It's a very aggressive approach, and they're learning a lot with it," said Don Kintner, a spokesman for the nonprofit Electric Power Research Institute. "I don't think there's anything like SmartGridCity in the U.S."
Other U.S. utilities from San Francisco to Chattanooga, Tenn., have taken more incremental steps, with varying results. Spurred by $3.4 billion in federal stimulus funding distributed in 2009, almost 100 utilities are upgrading some part of their distribution networks, with improvements such as wireless meters and pole-top outage sensors.
Oncor, a Dallas-Fort Worth, utility, has installed 1.2 million smart meters and 1,500 smart switches to detect line breaks remotely. It expects to upgrade more than 3 million customers by 2012, at a cost of about $690 million.
Xcel spokesman Tom Henley said these projects do not compare with Boulder's ambitious approach, which by the end of this year aims to not only make the grid more reliable for 24,000 smart-metered customers but also more capable of handling the future stresses that unpredictable wind- and solar-generation sources will place on it.
"They're not looking at the entirety of the grid," he said. "We're looking at the holistic design of it and not just smart meters."
SmartGridCity received no federal funding because the project was in motion long before the Congress passed the stimulus bill.
Another for-profit, state-regulated utility chose to budget for less technology but cover more customers. San Diego Gas & Electric introduced its smart-meter proposal in 2005 and plans to equip all of its 1.4 million customers with smart meters by the end of next year. The California Public Utilities Commission approved in 2007 a $572 million budget for the program.
San Diego's Smart Meter program promises to lay the foundation for future goals like preparing the city's more than 10,000 rooftop solar panels and 2,000 expected electric cars to plug into the system by the end of next year.
San Diego's program took more than five years to move from idea to implementation because it had to clear a series of cost "gates" before proceeding, said Ted Reguly, director of SDG&E's Smart Meter Program Office.
"To run a big project like this, you need to break it down into littler projects with small goals . . . to be sure you're going in the right direction before you move forward," Reguly said.
State regulators and consumer advocates also monitored major project decisions through a "technical advisory panel," a key concession brokered after Shames' UCAN demanded more oversight. As of this month, the company is still within its regulator-approved budget, according to filings with the California PUC.
SmartGridCity was "the first out of the gate" in testing that kind of technology, Xcel's Henley said. Much of the software and equipment Xcel chose was evaluated for the first time in Boulder, he said.
Instead of soliciting competitive bids — as SDG&E, Oncor and many other utilities did - Xcel chose some of its technology vendors based on how much they "brought to the table," Henley said. Software and equipment vendors donated hardware and expertise in exchange for a real-world lab to test their technologies.
Xcel declined to release the value of these partners' contributions, citing confidentiality agreements. A confidential Xcel estimate valued four vendors' contributions at $21.6 million.
The company still estimates the overall project is worth $100 million, even after its own share of the budget ballooned in two years to $44.8 million — a cost ratepayers will foot.
Own fiber-optic network
Information-technology costs associated with SmartGridCity have reached $19.2 million, according to a PUC filing. For instance, Xcel chose more expensive broadband-over-powerline technology and dug its own fiber-optic network to link its new devices because existing telecommunications companies were unwilling to share their networks at the time, according to Xcel.
The PUC did not treat Boulder's smart-grid project as a separate rate case until this year, after it was clear Xcel's expected costs were three times as high as initially suggested.
"Essentially, this was a decision by Xcel to launch a SmartGridCity pilot project," PUC spokesman Terry Bote said. "We were not asked to give prior approval, . . . but where decision points needed to be made, that's where we became involved."
Xcel points out that its pilot program had successes — power-line sensors prevented 63 unplanned outages in 2009 — long before most utilities had that capability. Henley said the utility will assign dollar figures to more than 60 "value propositions" that SmartGridCity will have demonstrated by the end of this year.
"Now, it's going to be easier for other people to do cost-benefit analyses because of the data they (Xcel) gathered," said Katherine Hamilton, president of GridWise Alliance, a smart-grid advocate. "So I think it was a great learning lab."