Averting the privacy risks of smart metering by local data preprocessing

More and more renewable sources are integrated into electric power grids worldwide. Their high generation dynamics, however, require power grid operators to monitor electricity generation and demand at a fine temporal resolution. Even small mismatches between supply and demand can impact the power grid's stability, and thus ultimately lead to blackouts. As a result, smart metering equipment has been widely deployed to collect real-time information about the current grid load and forward it to utilities in a timely manner. Numerous research works have shown that power consumption data can, however, reveal the nature of used appliances and their mode of operation at high accuracy. This effectively puts user privacy at risk. In this manuscript, we investigate to which extent the local preprocessing of power data can mitigate this risk. We thus compare the efficacy of different preprocessing steps to eliminate characteristic consumption patterns from the data. Our evaluation shows that a combination of these preprocessing steps can provide a balanced trade-off that is in the interests of both users (privacy protection) and utilities (accurate and timely reporting).