Omid Ardakanian

Markovian models for home electricity consumption

Modelling home energy consumption is necessary for studying demand-response, transformer sizing, and distribution network simulation. Using an existing classification, we propose parsimonious Markovian reference models of home load for each class. We derive models for on-peak periods, off-peak periods, and mid-peak periods. These models are derived using traces based on fine-grained measurements of electricity consumption in 20 homes over four months. We validate the representativeness of our models in a specific application.

Distributed control of electric vehicle charging

Electric vehicles (EVs) are expected to soon become widespread in the distribution network. The large magnitude of EV charging load and unpredictable mobility of EVs make them a challenge for the distribution network. Leveraging fast-timescale measurements and low-latency broadband communications enabled by the smart grid, we propose a distributed control algorithm that adapts the charging rate of EVs to the available capacity of the network ensuring that network resources are used efficiently and each EV charger receives a fair share of these resources. We obtain sufficient conditions for stability of this control algorithm in a static network, and demonstrate through simulation in a test distribution network that our algorithm quickly converges to the optimal rate allocation.

Real-Time Distributed Control for Smart Electric Vehicle Chargers: From a Static to a Dynamic Study

At high penetrations, uncontrolled electric vehicle (EV) charging has the potential to cause line and transformer congestion in the distribution network. Instead of upgrading components to higher nameplate ratings, we investigate the use of real-time control to limit EV load to the available capacity in the network. Inspired by rate control algorithms in computer networks such as TCP, we design a measurement-based, real-time, distributed, stable, efficient, and fair charging algorithm using the dual-decomposition approach. We show through extensive numerical simulations and power flow analysis on a test distribution network that this algorithm operates successfully in both static and dynamic settings, despite changes in home loads and the number of connected EVs. We find that our algorithm rapidly converges from large disturbances to a stable operating point. We show that in a test setting, for an acceptable level of overload, only 70 EVs could be fully charged without control, whereas up to around 700 EVs can be fully charged using our control algorithm. This compares well with the maximum supportable population of approximately 900 EVs. Our work also provides engineering guidelines for choosing the control parameters and setpoints in a distribution network.

RealTime distributed congestion control for electrical vehicle charging

The significant load and unpredictable mobility of electric vehicles (EVs) makes them a challenge for grid distribution systems. Unlike most current approaches to control EV charging, which construct optimal charging schedules by predicting EV state of charge and future behaviour, we leverage the anticipated widespread deployment of measurement and control points to propose an alternative vision. In our approach, drawing from a comparative analysis of Internet and distribution grid congestion, control actions taken by a charger every few milliseconds in response to congestion signals allow it to rapidly reduce its charging rate to avoid grid congestion. We sketch three control schemes that embody this vision and compare their relative merits and demerits.

Non-Intrusive Techniques for Establishing Occupancy Related Energy Savings in Commercial Buildings

The design of energy-efficient commercial building Heating Ventilation and Air Conditioning (HVAC) systems has been in the forefront of energy conservation efforts over the past few decades. The HVAC systems traditionally run on a static schedule that does not take occupancy into account, wasting a lot of energy in conditioning empty or partially-occupied spaces. This paper investigates the application of non-intrusive techniques to obtain a rough estimate of occupancy from coarse-grained measurements of the sensors that are commonly available through the building management system. Various per-zone schedules can be developed based on this approximate knowledge of occupancy at the level of individual zones. Our experiments in three large commercial buildings confirm that the proposed techniques can uncover the occupancy pattern of the zones, and schedules that incorporate these occupancy patterns can achieve more than 38% reduction in reheat energy consumption while maintaining indoor thermal comfort.

On the use of teletraffic theory in power distribution systems

Loads on the electrical grid are multiplexed at distribution transformers in the same way that traffic from data sources is multiplexed at a router. This motivates the use of teletraffic theory to size power distribution networks just as it is used to size telecommunication access networks. Specifically, we prove the equivalence between a model of a distribution branch comprised of a transformer and storage that we want to size for a given underflow probability ϵ, and a queuing model that we want to size for a given overflow probability ϵ. Based on this equivalence, we show how existing teletraffic analysis can be applied to size transformers when there is no storage. We compute such sizings using load models obtained from our measurement testbed and load models derived from an electricity demand simulator. We show not only that teletraffic theory agrees well with numerical simulations but also that it closely matches with the heuristics used in current practice by electric utilities, thus validating the use of teletraffic theory.

On the impact of storage in residential power distribution systems

It is anticipated that energy storage will be incorporated into the distribution network component of the future smart grid to allow desirable features such as distributed generation integration and reduction in the peak demand. There is, therefore, an urgent need to understand the impact of storage on distribution system planning. In this paper, we focus on the effect of storage on the loading of neighbourhood pole-top transformers. We apply a probabilistic sizing technique originally developed for sizing buffers and communication links in telecommunications networks to jointly size storage and transformers in the distribution network. This allows us to compute the potential gains from transformer upgrade de- ferral due to the addition of storage. We validate our results through numerical simulation using measurements of home load in a testbed of 20 homes and demonstrate that our guidelines allow local distribution companies to defer trans- former upgrades without reducing reliability.

Synchrophasor data analytics in distribution grids

The deployment of high-fidelity, high-resolution sensors in distribution systems will play a key role in enabling increased resiliency and reliability in the face of a changing generation landscape. In order to leverage the full potential of such a rich dataset, it is necessary to develop an analytics framework capable of both detecting and analyzing patterns within events of interest. This work details the foundation of such an infrastructure. Here, we present an algorithm for detecting events, in the form of edges in voltage magnitude time series data, and an approach for clustering sets of events to reveal unique features that distinguish different events from one another (e.g. capacitor bank switching from transformer tap changes). We test the proposed infrastructure on distribution synchrophasor data obtained from a utility in California over a one week period. Our results indicate that event detection and clustering of archived data reveals features unique to the operation of voltage regulation equipment. The chosen data set particularly highlights the value of the derivative of the localized voltage angle as a distinguishing feature.

Integration of Renewable Generation and Elastic Loads into Distribution Grids

This brief examines the challenges of integrating distributed energy resources and high-power elastic loads into low-voltage distribution grids, as well as the potential for pervasive measurement. It explores the control needed to address these challenges and achieve various system-level and user-level objectives. A mathematical framework is presented for the joint control of active end-nodes at scale, and extensive numerical simulations demonstrate that proper control of active end-nodes can significantly enhance reliable and economical operation of the power grid.

Data-Driven Models for Building Occupancy Estimation

The availability of accurate occupancy information from different spaces in a building allows for significant reduction in the energy consumption of heating, ventilation, air conditioning, and lighting systems. This paper investigates the application of particle filters and time series neural networks to inferring the number of occupants of individual rooms from time series data collected by a set of occupancy-indicative sensors. Our approach is purely data driven and does not require developing customized and complex physics-based models to predict the occupancy level of the many rooms in a building. We evaluate the efficacy of the proposed methods on two data sets, one contains measurements of dedicated sensors while the other one contains measurements of HVAC sensors that are commonly available in commercial buildings. Our results indicate that time series neural networks are superior in this application, estimating the number of occupants with a root-mean-squared error of 0.3 and 0.8 in the two data sets with a maximum of 7 and 67 occupants, respectively.