Tip: Set categories to Research Papers to search all published papers. Add tags to search for keywords.
SPOT: A Smart Personalized Office Thermal Control System
Problem
Heating, Ventilation, and Air Conditioning (HVAC) accounts for about 40% of the energy consumption in buildings. By changing the indoor air temperature of a building to be closer to the outdoor air temperature—for example, maintaining the building at a warmer temperature during summer months—HVAC energy consumption can be reduced by 10-40%. However, this comes at the cost of a reduction in individual comfort.
Solution
We have designed and implemented SPOT: a Smart Personalized Office Thermal control system. A SPOT device is placed in individual office spaces to heat or cool the immediate area to a comfortable temperature when an occupant is present. This allows the temperature of a building to be set to a value lower than normal in winter and to a value higher than normal in summer.
The first version of SPOT used 6 parameters to predict personal comfort: air temperature, radiant temperature, humidity, air speed, clothing level, and activity level. We made three iterations on SPOT’s design to improve its balance between energy conservation and personal thermal comfort, as described next.
SPOT
The first iteration of SPOT uses a Microsoft Kinect and a variety of sensors to measure the six parameters mentioned above. SPOT is able to calculate the amount of clothing a person is wearing using data from the infrared sensor and the Kinect. Once a user enters a room, SPOT measures these parameters, then controls a radiant heater to heat the workspace to a comfortable temperature. More details can be found here:
P. X. Gao. SPOT: A Smart Personalized Office Thermal Control System, MMath thesis, University of Waterloo, May 2013.
P. X. Gao and S. Keshav. SPOT: A Smart Personalized Office Thermal Control System, Proc. ACM e-Energy, May 2013.
SPOT+ (SPOT Plus)
SPOT+ improves upon SPOT by performing predictive control rather than reactive control. That is, SPOT+ will begin heating a workspace 10 minutes before a user walks in, so when they arrive the workspace is already at a comfortable temperature. It will also predict when a user will leave, so that it can begin cooling earlier to save energy.
After deploying SPOT+, we found that it reduced energy usage by 60% compared to a fixed temperature setting, and it reduces personal thermal discomfort from 0.36 to 0.02 (in the ASHRAE comfort scale) compared to SPOT.
P.X. Gao and S. Keshav. Optimal Personal Comfort Management Using SPOT+, Proc. BuildSys Workshop, November 2013. (Winner of the Best Student Paper Award.)
SPOT* (SPOT Star)
Our third version improves on SPOT and SPOT+ in 5 distinct ways:
- It provides both heating and cooling, unlike the previous iterations which only provided heating.
- It uses a speed-controlled desktop fan instead of a radiant-heater, making it possible to rapidly change the room temperature in response to discomfort.
- It is far less intrusive than the prior systems, because it does not use a camera.
- SPOT* is about an order of magnitude less expensive – while SPOT/SPOT+ were $1000 a unit, this prototype model costs only $185 a unit, with costs dropping to below $100 in mass production.
- Flexible placement of its software components allow for a balance between cost, privacy, and data durability.
In our deployment, we found that SPOT* improved user comfort by 78% compared to traditional HVAC systems.
Rabbani and S. Keshav, The SPOT* System for Flexible Personal Heating and Cooling, Poster, Proceedings of the 2015 ACM Sixth International Conference on Future Energy Systems (e-Energy), July 2015, 209-210. Best Poster Award
Rabbani and S. Keshav, “The SPOT* Personal Thermal Comfort System,” Proc. ACM BuildSys’16, November 2016.
Controlling an HVAC system with SPOTs deployed
To integrate SPOT systems into everyday use, we explore how to make HVAC systems SPOT-aware. We propose a control strategy to update the temperature set points for an HVAC system using the following factors: occupancy status in each room, preferred comfort requirements of occupied rooms, zones which have SPOT systems in it, outside temperature, and thermal properties of each room.
In a simulation setting, when users have homogeneous comfort requirements, we find that our system provides 45% savings in energy during the summer, and 15% during the winter compared to current predictive HVAC systems. When users have heterogeneous comfort requirements, our system provides 50% improvement in comfort in the summer and about 30% in winter, on top of significant energy savings.
Kalaimani, M. Jain, S. Keshav, and C. Rosenberg, ”On the Interaction between Personal Comfort Systems and Centralized HVAC Systems in Office Buildings, ” J. Advances in Building Energy Research, August 2018, V7:p.1-29.
Mitigating the impact of occupancy prediction errors in HVAC performance
Many commercial buildings use model predictive control (MPC) to control their HVAC systems – the model predicts outside air temperature and the number of people that will be in each zone of a building on a given day and adjusts the HVAC system accordingly. A prediction model cannot be perfect however – when the prediction errors of a model increase from 5% to 20%, the performance of the HVAC controller, as measured by occupant comfort and building energy use, becomes worse than that of a simple static scheduler that changes the temperature setpoint at the beginning and the end of the day.
We found that by employing the SPOT Aware strategy for HVAC systems, we stay in the acceptable region of occupancy comfort 95% of the time as opposed to only 83% when there are prediction errors in the MPC system. Thus, installing SPOT systems can not only save energy, but also make building occupants more comfortable, even in the presence of forecasting errors.
M. Jain, R. Kalaimani, S. Keshav, and C. Rosenberg, ”Using Personal Environmental Comfort Systems to Mitigate the Impact of Occupancy Prediction Errors on HVAC Performance,” Energy Informatics, 2018 1:60, https://doi.org/10.1186/s42162-018-0064-9, December 2018.
Range prediction for electric bicycles
Problem
A significant barrier to the adoption of e-bikes is range anxiety, or the fear of running out of battery with no place to recharge. Currently, e-bikes do not display the estimated range available. Instead, a digital display shows battery voltage, but it is difficult to estimate the remaining range of a bike based on this value. Though e-bike manufacturers do publish the maximum range of their models, we found that this number is not an accurate predictor for all riders, depending on how aggressively they ride. Note that any additional hardware required for range prediction must be inexpensive, in order to keep overall prices low.
Solution
Using data from a fleet of 31 sensor-equipped e-bikes used in the University of Waterloo WeBike project, combined with OpenStreetMap data, we evaluate two range prediction models for e-bikes. The first model is a simple one, based on the average battery consumption from past trips. The second model is a more complex linear regression model that considers the characteristics of the anticipated route (such as off-road percentage, the number of stop signs, and the number of traffic lights), as well as battery temperature.
Evaluation
We found that the more complex linear regression model didn’t perform much better than the simpler one. Using real trip data, our predictions using the simple model were usually within a 10% of the actual remaining range at the end of the trip.
These results should be of interest to e-bike manufacturers because a simple on-board prediction technique can be implemented by measuring battery voltage, battery current, and mileage. Since most e-bikes have an odometer built in, by making this odometer data accessible and deploying additional sensors at the battery, our technique can be implemented inexpensively.
L. Gebhard, L. Golab. S. Keshav, and H. de Meer, “Range prediction for electric bicycles,” Proc. ACM e-Energy 2016.
Control of Electric Vehicle Charging
Problem
Electric vehicles (EVs) pose a challenge to the electrical grid in two ways.
- First, large-scale introductions of EVs pose a significant load to the grid. An EV can be charged with a load of up to 19.2kW (with Level 2 chargers), whereas a typical North American home has an average load of 1kW – this means a single EV could impose a load as large as that imposed by nearly twenty average homes.
- Secondly, the load posed by an EV is variable by time and location: its load on a grid will unpredictably disappear when it is being driven. It might then charge at a different location, re-appearing at a different part of the electricity distribution network.
Solution
Since a typical EV charger is located within 3km of the nearest substation, the transmission delay between any charger and its connected substation is less than 1ms. As such, we can design a distributed control algorithm that adjusts the charging rate of an EV every few milliseconds, in response to the load being placed on the overall distribution system. For example, if an EV is charging at a rate that affects the reliability of the grid, its charging rate can be decreased.
Three papers were written on this subject. The first paper introduces the problem and describes how the congestion control problem for a grid distribution system is similar to the congestion control problem in the Internet.
- O. Ardakanian, C. Rosenberg, and S. Keshav. Real Time Distributed Congestion Control for Electrical Vehicle Charging (invited paper), ACM SIGMETRICS Performance Evaluation Review 40.3 (2012): 38-42.
By using a mathematical framework originally developed for rate control in the Internet (TCP), each EV charger in the grid can independently update its charging rate, while ensuring that the overall load on the grid stays at an ideal level, the allocated rates for each charger are proportionally fair, and that these allocations are optimal. The second paper in this series focuses on a static network scenario, in which the non-EV load is constant, and a fixed number of EVs are connected to chargers.
- O. Ardakanian, C. Rosenberg, and S. Keshav. Distributed Control of Electric Vehicle Charging, Proc. ACM e-Energy, May 2013. Winner of Best Paper Award.
The third paper goes into detail about the dynamic network scenario, which involves variable home loads and number of plugged-in EVs. Since the dynamic network scenario can be decomposed into a series of static intervals, the static control algorithm described above can be extended to be used in a dynamic network.
- O. Ardakanian, S. Keshav, C. Rosenberg. Real-Time Distributed Control for Smart Electric Vehicle Chargers: From a Static to a Dynamic Study, IEEE Transactions on Smart Grid, vol.5, no.5, pp. 2295-2305, Sept. 2014.
Evaluation
We show that in a test setting, only 70 EVs could be fully charged without control, whereas up to around 700 EVs can be fully charged using our control algorithm. This work was further extended to integrate EV charging control with control of distributed storage, while accounting for distributed solar generation. Details can be found here: O. Ardakanian, S. Keshav, C. Rosenberg, “Integration of Renewable Generation and Elastic Loads into Distribution Grids,” Springer, 2016.
WeBike: A three year study on e-bikes as a mode of sustainable transport in a Canadian City
Context
E‐bikes are revolutionizing transportation in China and parts of Europe, yet little is known about user patterns in North America, particularly Canadian cities where uptake tends to lag. To fill this gap in knowledge, the University of Waterloo studied a sample of e-bike riders over three years in Kitchener-Waterloo, Ontario, Canada. The field trial, called WeBike, amassed over 150GB of data on e-bike usage by faculty, staff, and students from the summer of 2014 until spring of 2017.
Learnings
Three papers were published by the University of Waterloo analyzing this data – the first two are quantitative in nature, and the last one is qualitative.
Quantitative data collected from the e-bikes included GPS, acceleration, and battery charge and discharge data. Based on our analysis, we draw several conclusions:
- Over 6000 trips, students made more e-bike trips than faculty and staff members, were more likely to ride in the evening, and had lower average speed trips.
- The primary purpose of e-bikes in the trial were used for commuting.
- Most trips lasted less than 20 minutes.
- Most trips had an average speed of 15–23 km/h (while in motion) with a mean of 18.9 km/h.
- The most prominent charging times for all riders was between 4pm and 7pm (likely coinciding with when riders returned home from school or work).
- Participation in the WeBike field trial did not significantly change participants’ sentiments towards various modes of transportation.
- There was little correlation between anticipated use of the e-bikes and actual use.
I. Rios, L. Golab. and S. Keshav, “Analyzing the Usage Patterns of Electric Bicycles,” EV-SYS Workshop at ACM e-Energy, 2016.
C. Gorenflo, I. Rios, L. Golab, S. Keshav, “Usage Patterns of Electric Bicycles: An Analysis of the WeBike Project,” Journal of Advanced Transportation, October 2017.
Some notable findings in the qualitative study include:
- Over time, nearly all participants self‐reported an increase in the total number of trips taken on their e‐bike, due to
- 1) Increased level of comfort with the technology and battery range
- 2) The motor allowing them to travel further than a regular bike or by foot
- 3) The e‐bike being generally viewed as faster than walking, cycling, and public transit
- Security concerns, specifically theft, were consistently cited. Participants felt the e‐bike was more susceptible to theft or vandalism because it looked expensive.
- Participants felt that e-bike usage facilitated more physical activity than driving a car.
- All participants stated that once the study was over, they would continue to ride the e‐bike they had been given as compensation for their participation in the study.
S. Edge, J. Dean, M. Cuomo, and S. Keshav, “Exploring e-bikes as a mode of sustainable transport: a temporal qualitative study of the perspectives of a sample of novice riders in a Canadian city“, Canadian Geographer/Le Geographe Canadien, April 2018.
Recommendations
- The general population in Canada is still unaware of e-bikes and their potential, so e-bike manufacturers should consider educating potential customers on e-bike usage.
- E-bike use does not cease in the winter months. Hence, e-bike manufacturers in countries with winter weather should consider offering built-in fenders and lights for safer winter cycling.
- E-bike manufacturers should target sales to non-bike users, such as seniors, rather than trying to displace sales of regular bicycles and for increasing physical activity for individuals with health conditions or limited mobility (e.g., those related to aging).
Robust and Practical Approaches for Solar PV and Storage Sizing
Problem
As grid-provided electricity continues to rise in cost, individuals, small companies, and building operators are increasingly looking to offset grid usage using solar photovoltaic (PV) panels and energy storage. When constrained by a budget however, it is difficult to predict the optimal ratio of solar PV panels and storage that one should buy to completely or partially offset grid usage.
Solution
We tested three ways to determine the possible sizing(*) required to meet the anticipated load in an off-grid system. Out of the three methods–simulation, optimization, and stochastic network calculus–we found that the easiest method, simulation, also provided the best results.
(*) sizing: the power or energy size of the storage in kW/kWh, and the size of solar generation in kWp.
Evaluation
We are currently building a web application to help predict the best sizing for a solar energy system. The tool will be linked here when completed.
Citation
F. Kazhamiaka, S. Keshav, and C. Rosenberg, Robust and Practical Approaches for Solar PV and Storage Sizing, Proc. ACM eEnergy 2018, June 2018.
Simple Spec-Based Modelling of Lithium-Ion Batteries
Problem
The best-known lithium-ion battery model whose parameters can be calibrated entirely from a battery’s manufacturer-provided specifications (ie. internal resistance, nominal capacity, voltage at full charge, etc) was proposed by Tremblay et al and is widely use. However, it has some shortcomings. For instance, it has low fidelity at high charge and discharge rates. Moreover, it only models a cell, and does not model the battery management system. This makes it unsuitable for evaluating practical storage systems.
Solution
We propose an alternative, the Power-based Integrated (PI) model, whose parameters can also be calibrated entirely from manufacturer specifications, but has much higher fidelity across a wide range of charge/discharge rates. This model is freely available in the public domain as a Matlab system block compatible with Simulink simulation software. Access the block here.
Evaluation
We suggest using our model when:
- Simulating an energy system with storage by modeling its power flows. In this case, it is desirable to have a battery model that uses power as input because power is conserved (note that the Tremblay model uses current as input). The PI model uses power as input.
- Modeling the battery management system (BMS) – the BMS protects the cells in a battery from being damaged due to improper use such as under/over-charging. Modelling the BMS in the PI model prevents the simulated battery from being used in unrealistic ways.
- Low error – based on validation performed in the experiment, we see that the PI model has a mean absolute voltage error of less than 0.1V across a wide range of C-rates.
Citation
F. Kazhamiaka, S. Keshav, C. Rosenberg, and K.-H. Pettinger, ”Simple Spec-Based Modelling of Lithium-Ion Batteries, ” IEEE Transactions on Energy Conversion, Vol 33, No. 4, December 2018.
