Paper Conference

Proceedings of eSim 2016: 9th Conference of IBPSA-Canada


A comparison of electricity “smart meter” data with high resolution measured residential electricity loads

Dane George, Lukas G. Swan

Abstract: In the Canadian residential sector, the end-uses of appliances, lighting, and plug loads account for 16% of total end-use energy consumption (NRCan 2012). In an effort to reduce the impacts of this energy consumption, electricity technologies such as solar photovoltaics and smart appliances are being adopted. Their evaluation requires an understanding of residential electricity use patterns. Building simulation tools can estimate the time-step performance of such technologies, but require accurate and representative appliances, lighting and plug-load (ALP) electricity profiles as an input. Submetered datasets lack in quantity and thus overall representativeness of the sector. Meanwhile, large, representative datasets are becoming available through electricity smart-metering programs, but usually consist only of whole-house electricity load and lack summary household characteristics (e.g. occupancy, floor space, appliance descriptions). However, homes which are not electrically heated, cooled and without electric water heating may function as ALP load profiles for simulation. This paper addresses two of these loads with a new method of distinguishing non-electrically heated and cooled homes from a broad dataset of whole-house profiles. The method originates from a comparison of two electricity load datasets: (i) “smart-meter” 15minute time-step whole-house data for 160 homes electricity generation such as combined heat and power and solar photovoltaics (PV) to supply energy end-uses. These buildings presently rely on the electricity grid as an infinite source and sink, and their proliferation necessitates a better understanding of timestep electricity demand. In net-zero communities, distributed generation from solar PV may cause severe peaks and valleys in the community electricity load on the grid. Utilities seek to understand these short-term loads so that they can procure sufficient generating capacity and install adequately sized and placed distribution equipment (e.g. pole transformers). Existing models which employ building simulation software are capable of time-step energy demand estimation of buildings and communities. These typically rely on engineering principles to model spaceheating and space-cooling, but appliances, lighting, and plug loads (ALP loads) are largely driven by occupant behavior and modelling relies on measured or synthetic time-step load profiles. Examples of ALP loads are shown in Figure 1. Since ALP electricity use varies widely across households, community scale modeling requires a sufficient number of ALP load profiles for individual houses to represent a greater community. spanning up to three years, and (ii) “sub-metered” 1minute time-steps for 23 residential homes. This comparison also speaks to the usefulness of whole-house electricity smart-meter information to building performance simulation.
Pages: 26 - 34