PeHEMS: Privacy enabled HEMS and load balancing prototype

Abstract

Smart grid efficient load balancing and the need for privacy are, in principle, contradictory. While richer information obtained from frequent energy readings help improve both the prediction and the control of the demand, and, effectively, improve the efficiency of the energy equilibrium production problem, it also gives rise to consumer privacy concerns. This is possible by analysing energy signatures to detect appliance usage and home living patterns of behaviour, which in effect cascades to a range of privacy invasion risks. This paper argues that the objective of energy efficiency might not necessarily be contradictory to protecting user privacy. In particular, we introduce a new notion of smart meter privacy which we call reconciled privacy and we connect it with a simple energy management algorithm that caps the energy a home may con sume in 30 minute intervals by using a rechargeable battery system. System benchmarking is underpinned by formulating a methodology to assess a) utility cost savings, b) wholesale energy savings, and c) privacy protection. Our results suggest that the proposed algorithm will protect customer privacy and will improve energy production efficiency as compared with other energy management schemes. This is due to the algorithm's principle of promoting a universal consumption pattern that is close to its average, which in retrospect allows individual usage differences to be absorbed. To support this work, we use data from trials in Bristol city, which forms part of the 3eHouses EU FP7 project, and we present a prototype implementation showcasing the visualisation of privacy and energy control.

Gridchain: an investigation of privacy for the future local distribution grid

“…Let S and D be arrays containing the positions of peaks corresponding to kWh real and kWh reported , respectively. [3,5,8,14,19,21,25,31,36,39,41,44,46] D = [5,7,10,14,19,21,23,25,28,31,33,36,39,41,44,46] In this example, the edit distance is 4, the number of matches is 11 ([5, 14, 19, 21, 25, 31, 36, 39, 41, 44, 46]) whereas the LCS is 5 (the string composed of [36,39,41,44,46] peaks). Finally, we have certified that kWh real = kWh reported holds for all the networks at the end of the specified billing window, being equal to 0.154 kWh.…”

“…Our work is inspired by the work of Kalogridis et al [32][33][34]. They introduce the notion of reconciled privacy where a local battery is used to hide peaks of the consumer.…”

“…In this paper, we designed a system where consumers, together as a group, can hide their consumption pattern by replacing the battery (used in the above approaches) with borrowing / paying back apparent consumption among the Data granularity [66] Course-grained granularity and randomly Differential privacy [30,44,65,67] Added noise may interfere with accurate billing Secure multiparty computation [59] Relatively high computing complexity and overhead Homomorphic encryption [42,70] Local energy storage facilities [7,8,[32][33][34]47,55,64,68] Expensive and not widely available group members. Each customer is billed only for her actual consumption, there is no need for costly batteries, and NILM is made more difficult than without having the system.…”

“…This paper investigates privacy concerns regarding detailed readings for billing purposes. Others have previously suggested that a local supply (battery) can be used to anonymize the usage patterns so that less information about the household can be gained from the energy traces [33,68]. This supply can either be used directly to hide consumption patterns or as an external source to disturb the signal in a differential-private way.…”

Optimal Model for Energy Management Strategy in Smart Building with Energy Storage Systems and Electric Vehicles

“…Various hardware applications for BEMS are proposed in [18,19]. Several works in the technical literature focus on BEMS modeling through linear and non-linear programming, as illustrated in [20][21][22]. Several studies illustrate the programming of energy resources through heuristic methods, mostly with high computational development, as shown in [23][24][25].…”