California’s three big utilities are being asked to share their customer’s energy data with each other, with their customers, and with the world. The process involves a lot of acronyms (CSV vs. XML, NAESB’s ESPI plus OpenSG’s OpenADE), and a lot of potential for confusion, conflict, and even litigation -- and it’s going to shape how consumers interact with their energy use.
That’s all from the California Public Utilities Commission in San Francisco, where Monday saw smart grid executives and utility honchos hashing out how the bellwether state should enable the sharing of smart grid data in a consumer energy management market.
Data is the lifeblood of that market, but it’s also the private property of each individual customer -- and it’s the utility’s job to keep it that way. That makes the state’s utilities leery of the idea of sharing that data with third parties. Think of everyone from Google and Facebook to anonymous, fly-by-night web-scraping operations, pulling data with or without utility consent and knowledge -- as long as the customer has OK’d the deal, and is using their own data, it’s their decision.
To this complex tangle of concerns, add government mandate. The CPUC had ordered the state’s big three utilities (Pacific Gas & Electric, San Diego Gas & Electric and Southern California Edison) to start sharing customer data, with the customers who own it at the very least, by the end of next year.
At the same time, all three utilities are participating in the “Green Button” challenge from U.S. CTO Aneesh Chopra this September, which asks for a common data format for customers to share their own energy data -- scrubbed of all private data, of course -- with each other, as well as third parties.
It’s a bewildering array of options for connecting people with their home energy use information, and each option has its own unique complications. Let’s lay out how each utility is proposing to do this, while remembering that they have yet to file proposals for actually doing it:
- Brendan Blockowicz, SDG&E's smart meter IT program manager, laid out a two-fold technological approach for the utility’s well-advanced consumer engagement strategy.
First, to meet the CPUC’s demand for machine-to-machine data exchange, SDG&E will allow customers to download a zip file containing a CSV (comma-separated values) file, which can be imported into tools including Excel, as well as transferred to third parties, he said. As one can imagine from the name, this tabular data storage file isn’t something that laypeople are meant to decipher on their own, however.
Second, for the Green Button challenge, SDG&E will let customers download an XML (extensible markup language) document in ASP file format, which once again is “not readable to a human, but is useful to any machine-to-machine interactions,” he said. From there, customers can theoretically run it on apps they’ve downloaded, or send it to third-party apps providers.
- Mark Podorskey, SCE’s senior manager of business design, described a two-stage process much like SDG&E’s, with CSV file format data coming first and XML later. He also laid out some of the steps the utility’s architecture has to take to carry each customer’s request, validate its security, make sure the authorization of the customer is in hand and properly approved, then transfer it to third-party format.
Getting there will require everyone to build to some kind of standard, and "hopefully, everyone’s going to be using ESPI going forward,” Podorskey said. ESPI, stands for Energy Services Provider Interface, the preferred standard of the North American Energy Standards Board (NAESB). That, in turn, is being modeled on Open Automated Data Exchange, or OpenADE, a developing standard for data-sharing that’s expected to dominate the North American market, once it’s finished.
OpenADE is also guiding the XML work utilities are doing to serve the Green Button feature they’re putting on their Web pages. The idea is to create an XML file that can be shared by anyone the customer authorizes to handle it -- at least in theory.
- Saul Zambrano, PG&E’s director of integrated demand-side management products, upped the ante a bit on the Green Button challenge, saying the utility was ready to deploy it on its customer-facing web portals by the end of the month.
“It’s going to be an XML format, predicated on Rev 1 of OpenADE,” he said, throwing out another acronym, for OpenADE’s first revision available for use. Of course, the standard is still developing -- OpenADE is already going into Rev 2 discussions, he said.
At the same time, PG&E is setting up its customer-facing My Energy dashboard to allow download of CSV-formatted data for both electric and gas meters, he said. Interestingly, while PG&E’s version of XML is built to be open to all comers, the CSV data is, “at least for us, a PG&E-specific format,” he said.
That begs the question of how coordinated the PG&E, SDG&E and SCE have been in building their M2M data exchange platforms. Much depends on how each utility has interpreted an OpenADE standard that’s still being created.
“We see it evolving over time,” SDG&E’s Blockowocz said. “We’re already seeing, I don’t want to call them issues, but questions, about the SG (OpenADE) format that might require changes down the road.”
- Speaking of standards, Dave Mollerstuen, standards development architect for home energy management tech startup Tendril, laid out the latest developments. He’s the chairman of the OpenADE group of OpenSG, the industry-government group created to develop national smart grid standards, and Tendril has pilots with 30 utilities, including SDG&E, giving it a lot of room to test new versions of technology.
That’s important, as Mollerstuen pointed out, because of just how in-flux today’s standards are. Take NAESB’s energy usage information effort, which seeks to define a common way of delivering the basic energy data that comes from homes and smart meters.
That involves getting the Smart Grid Interoperability Panel's priority-action plan 10 group (PAP10) to make sure that all its requirements are being met by the NAESB REQ 18/WEQ 19 [Oh, come on --- ed.] standards effort -- in other words, NAESB’s version of an OpenADE-compliant technology. A version of that is expected to be complete by year’s end, he said. But because the PAP10 requirements are more “abstract,” ESBI has to take it upon itself to define a specific model to build on, meaning a lot of back and forth between the two groups, he said.
In the meantime, utilities are struggling with a host of questions about how all this standardized data might spread across the Internet, Zambrano said.
“There’s going to be a lot of innovation around this data,” he said. While PG&E wants to support it, “We only want to do it once … with all the proper privacy concerns addressed to deploy it in a more scaled format.”
That means thinking through some seemingly simple questions, such as who owns the data versus who’s responsible for it. While customers own their own energy data, the utility owns the data out of its meters and from its back-office systems, and can’t turn it over without customer authorization.
But once the customer has signed off on allowing others to access that data, the utility has an obligation to deliver it, albeit in a way that protects data privacy and security. Once the data’s out of its hands, a customer can theoretically do what he or she wants with it. Indeed, customers may allow third parties to access their energy data without telling the utility at all, Zambrano noted.
PG&E is launching a home energy networking pilot next year featuring Silver Spring Networks and Control4. But before it can be expanded to third parties, the utility has to figure out how to certify them to work with such sensitive data -- not to mention how to pay for managing the new system, he said.
Then there are the limits of today’s versions of home energy data. Mollerstuen noted that today’s ESPI standard, for instance, doesn’t “natively capture” power prices or add up how much a customer has used so far this month -- all it does is capture raw energy usage data, usually in 15-minute to hourly increments. Mainly, that was to speed up the process of coming up with a standard, he said.
“It’s been a two-and-a-half-year effort” to get to today’s standard, he noted. “Frankly, we feared it would be a five-year effort if we took on too much.” That means that, at least today, companies are waiting for OpenADE 2.0 to roll out, or looking at extensions to do things like pull price data from websites to calculate running bills.
In the meantime, all that machine-to-machine data is awaiting translation into something customers can actually understand and engage with. Mollerstuen said a 2.0 version of OpenADE is due by the middle of 2012, but it will take another 12 months or so after that to “see new data being available via the OpenADE interface,” he said.
That means that we’ll have to keep watching pilot projects to see how all this technology works in the real world. Mollerstuen said that OpenADE members -- including Tendril -- planned to create “test cases and validation attempts” that could be unveiled in the next month or so.