Allowing Municipal Aggregations To Obtain, Under Data Security Agreement, Aggregated Customer Data Which Fails Privacy Screen Under Consideration In New York As Further Comments Sought

The New York PSC has issued a request from stakeholders for further input regarding the appropriate balance between the benefit of making more utility customer aggregated usage data available and the need to maintain consumer privacy.

The PSC previously adopted a 15-by-15 standard, under which aggregated customer usage data is not considered sufficiently anonymous to share publicly unless (1) the aggregated group contains at least 15 individual accounts and (2) no one account represents more than 15 percent of the total load.

As part of a proposed Utility Energy Registry (UER), DPS Staff, as an initial starting point for consideration, proposed to have the UER populated by municipal tax district (in New York City, Con Edison would use zip codes) with utility rate service classes grouped into three buckets (Residential, Small Commercial and Other). Staff proposed that each rate group be populated semi-annually with kWh (monthly), Installed Capacity (ICAP) tag (Capacity Tag), the total number of customer accounts along with the number of accounts served by Energy Service Companies (ESCOs) and/or with blocks (i.e., non-eligible Community Choice Aggregation (CCA) customers), and the number of accounts on a time of use (TOU) rate. In light of the recent Commission order requiring CCAs to guarantee savings for low income assistance program participants (APPs) that are part of the CCA’s program, Staff is also proposing to add APP counts to the data sets.

Staff’s proposed UER data sets were then used by the utilities to apply the Commission’s 15-by-15 privacy screen for two separate months. In general, the results indicated a high failure rate (80-100%) for the Other bucket, a relatively high (35-80%) failure rate for the Small Commercial bucket, and a few failures in the Residential bucket.

A working group agreed to conduct additional analyses using hypothetical screens of 6-by-40 (a screen which considers aggregated usage data sufficiently anonymized if the data set contains at least 6 customers and no one customer represents more than 40 percent of the total load) and 4-by-50 to further assess the UER data and to explore avenues to balance the benefits of data transparency with privacy protections. The alternative screens resulted in lower failure rates

Given the above, the PSC said that further input is requested regarding the appropriate balance between the benefit of making more aggregated data available and the need to maintain consumer privacy.

"Among the methods that should be explored are: (1) for data sets that do not pass the privacy screen at the most granular level (i.e., municipality) combine the data sets with other municipalities (i.e., county); (2) recalibrate the privacy standard (in general or for certain aggregation groups) to optimize the benefits of making more information available (e.g., improved local energy planning, improved targeting of clean energy products and services) while continuing to maintain adequate privacy protection; (3) where aggregation group does not pass the privacy screen, allow the CCA Administrator, and/or municipalities, to obtain the data from utilities pursuant to Data Security Agreement," the PSC said

Case 17-M-0315 et. al.