The Problem
Transitioning to renewable energy is necessary to reduce toxic pollution, prevent the disastrous effects of climate change, and ensure a long-term reliable energy supply. However, clean energy creates its own challenges. For one, common sources of renewable energy such as wind and solar are only available at certain times of day. To use exclusively renewable energy, energy storage during times when these sources are producing well is necessary to keep an uninterrupted supply when they are not available. Energy storage technology, however, can increase the cost by more than double per unit of energy, is inefficient with 5-15% energy losses, and most storage technologies use rare earth metals or other materials that are not environmentally friendly.
The graph below shows the energy usage and availability during a hot summer day. Notice how the temperature tends to peak in mid-afternoon, and remains hot from 12 – 8pm. The net demand trails the temperature by 2-3 hours, peaking from 3pm – 8pm because it takes time for buildings to heat up and need more air conditioning. In the evening, people return home and begin cooking and charging electric vehicles, increasing demand even more. However, the renewable energy supply is dominated by solar, which is most effective between 10am – 4pm. Thus, renewable energy would require significant energy storage for the 4 – 8pm time window to fully power the grid.
The exact availability times for renewables, and the consumption patterns vary over the year. Solar is less available during the winter because of reduced sunlight, but the weather conditions increase the wind power generation. The heating and cooling times can likewise shift because of weather and seasonal factors.
The Solution: Using Energy When It’s Available
The amount of stored energy required can be reduced if the highest energy consumption devices can be operated when renewable energy is plentiful. These include heating, air conditioning, cooking, laundry and dryers, dishwashers, and electric car charging.
Smart meters allow utilities to change the price of electricity throughout the day. The electric industry is forecasting that smart grid technology, in which all buildings are connected via simple smart IoT devices, will soon allow a dynamic pricing model in which the electricity rate changes to match the real-time availability of electricity. When electricity is more plentiful and cheaper to produce, the rate decreases to encourage more energy usage.
Laundry, dryers, dishwashers, and electric cars can be set to run or charge using simple timers or delay settings built into the device, with decisions made using the price of energy at the time. Heating and cooling can also be rescheduled by pre-heating or pre-cooling buildings ahead of time if energy is relatively inexpensive, then using less heating or cooling during the times of highest consumption. In the summer scenario above, the air conditioning could pre-cool buildings between around 10am and 1pm. The building acts as a thermal reservoir, staying somewhat cooler throughout the rest of the afternoon, allowing it to run less air conditioning during the 4pm to 8pm window.
The People Factor
In practice, it is hard to create a mass-shift of energy usage. Activities such as cooking cannot be rescheduled without inconveniencing people. However, since the renewable generation and maximum consumption times vary throughout the year, this requires the user to constantly monitor the electric rates – a huge ask for anyone.
For heating and cooling, the equation gets even more complicated. Pre-heating or pre-cooling too much is wasteful and can be very uncomfortable. Predicting when to cool or heat a building depends on weather and the characteristics of the individual building. Few people would be willing to take the time to figure out when is the best time to heat and cool, and so these potential energy savings cannot be realized if relying on people to run them.
Automation
Fortunately, the availability of low-cost IoT technology and other smart embedded devices means that simple, inexpensive internet connected controllers can be placed in consumer devices. By connecting to the electric rate data, devices such as dishwashers, clothes dryers, and electric cars can be given a deadline by which the user needs the task done, and they can automatically select the best time to do it. They have a consistent duration during which they need to operate, and they can simply pick the time of lowest electric rate before the deadline.
Heating and air conditioning is more complicated. Whereas other devices have a predicted duration, heaters and air conditioners activate whenever the temperature gets too hot or too cold. Weather and building parameters vary, such that it is hard to predict how long it might take to heat or cool a building, and how long the pre-heating or pre-cooling will last. Current smart-thermostats have algorithms manually-programmed to provide some savings over the traditional fixed setpoint type, but cannot attain the same amount of energy and cost savings over ones that can optimize and adapt to the current conditions.
New Research and Development
I am joining the Smart Grid and Cyber Physical Systems team at NIST for this summer to research methods to more efficiently heat and cool buildings in response to standardized pricing and weather data. Our eventual goal is to create a framework through which entire neighborhoods of buildings can communicate with each other and the central grid to optimize overall energy consumption, reducing prices and increasing the sustainability and reliability of the power grid.