Between the Poles

Privacy by Design: Protecting the privacy of consumer electricity usage data

I blogged recently about the rules for protecting consumer electric power usage data collected by utilities in California.  I found this on the Center for Spatial Law and Policy site.  The California Public Utilities Commision, which regulates electric power in California, has defined the rules for providing access to and protecting the privacy of consumer electric power usage data collected by smart meters in California.

At  the Data Analytics for Utilities conference in Toronto last week, Ann Cavoukian, the Information and Privacy Commissioner of Ontario and one of the people who have raised the profile of data privacy, gave an enlightening overview of the important privacy issues as they relate to smart grid, and especially what they mean in Ontario. 

As I have blogged before,  smart grid, expecialy distributed generation and smart meters, is more advanced in Ontario than most other jurisdictions in North America and comparable to Sweden and Denmark in the EU.  Ann Cavoukian provided important information about smart grid in Ontario.   Ontario has invested $1 billion in smart meter infrastructure and is the first jurisdiction in North America to equip every small business and home (except for condos and multi-unit buildings) with a smart meter.   That's 4,770,289 installed smart meters and 4,258,094 customers on time-of-use (TOU) billing. 

Nearly 5 million smart meters means a lot of data and much of it is personally identfiable information (PII), defined as any data that can be tied to an individual, directly or indirectly. Electric usage data tells people when you are using electric power at a granularity of 15 minute intervals and can be used for nefarious purposes.

Privacy = control

Ann Cavoukian's fundamental point is that privacy means that electric usage data belongs to the consumer, and that for the consumer it means being able to control who has access to their data.

When a utility collects electricity usage PPI data for billing purposes, the customer is right to expect that this primary purpose is the only use that the utility will make of the data.  If the utility wants to use the data for another purpose, it needs to ask the consumer for permission.

Privacy in the big data era

The challenge in this era of huge volumes of data that smart meters and other smart devices can generate is reconciling protecting the privacy of consumer energy usage data with  the importance to utilities of analytics which requires access to meter data.

Privacy by design

Ann Cavoukian's recommendation for doing this is what she calls Privacy by Design.  The first principle of which is minimizing the collection of PPI data and using it only for the primary purpose for which it was collected. 

De-identification

But there is a technique which enables the utility to use meter data, simply by aggregating the data so that it is de-identified and cannot be linked to an individual, for example, by aggregating all meter data by postal code and rejectjng any postal codes with less than three meters.  National statistical agencies have been doing this kind of things for years.  For example, Stats Canada does not report demographic data for census divisions  that have less than three respondents.  There are a variety of mathematical techniques for de-identifying data.

Ann Cavoukian said that the principles of Privacy by Design underlie not only Ontario's privacy laws, but also privacy laws in the U.S., E.U., New Zealand and Australia.

Consumer energy usage tracking software - who's the customer ?

Powermeter

Google PowerMeter was a software project of Google's philanthropy division, Google.org.   It was targetted on consumers and was designed to help consumers track their home electricity usage.  PowerMeter users could monitor their energy consumption online via an iGoogle widget, but it required that the utility had to agree to connect their smart meter data with Google.  Later Google started working with h/w manufacturers to develop gadgets to collect data directly from certain types of meters.  Google said that they did not intend to make money with the product, but wanted to contribute to reducing greenhouse gas emissions.   Google PowerMeter was launched on October 5, 2009, and retired on September 16, 2011, after about a two year run.

Hohm

Unlike Google Microsoft intended to make money on its energy management software Hohm through advertising.  LIke Google, Microsoft was interested in the consumer, not the utility, partly because they saw the consumer ultimately controlling demand and secondly because they assumed that consumer awareness of the Microsoft brand would attract users.  Microsoft announced that it was shutting down Microsoft Holm May 31, 2012, citing a “combination of weak customer demand (recession, energy efficiency not mainstream) and the lack of a coherent national energy policy [that] encourages solutions like Hohm.“

Several reasons that have been suggested for the failure of these two products from two of the largest software companies on the planet.   The market for energy management tools was still in a very early stage.   It required the consumer to opt-in, when the only successful utility applications required the consumer to opt-out.  Google.org isn't an energy company and really didn't know how to market to utilities. 

But the one that makes sense to me is the question of whether Google's or Microsoft's real customer should have been the utility or the consumer and whether utilities found the idea of a Microsoft or Google insinuating itself between the utility and its customers with access to electricity usage data attractive.

Opower

Opower is an example of a different approach to "behavioral energy efficiency".  Instead of doing an end-run around the utility,  Opower has focussed on the utility as their direct customer and has aimed at helping the utility make energy efficiency,  including Time-of-Use (TOU) or Peak-Time Rebates (PTR),  palatable or even attractive to the utility's consumer customer base.  As a result Opower has grown since their first customer in 2007 and now claims to be working with 75 utilities, including 8 of the10 largest in the U.S., with their energy monitoring software being used in 15 million homes globally.