Project Drawdown defines smart thermostats as Internet-connected devices in households that reduce the heating and cooling demand of homes by using sensors and intelligent settings to maintain building comfort. This solution replaces conventional home thermostats.
The built environment accounts for a substantial portion of global energy consumption and expense, and there is significant room for improvement in how the energy used to heat and cool buildings is managed, especially in the residential sector. Less than half of homeowners actively manage their energy use, and the vast majority of thermostats currently on the market are not data-driven.
In recent years, a new type of thermostat has been introduced—a “smart thermostat,” “learning thermostat,” or “wi-fi thermostat.” This type connects to the Internet to allow settings control from anywhere, and learns from user behavior to optimize energy settings, saving 10–15 percent of energy needs while improving comfort and convenience.
The analysis below examines the financial and emissions impacts of high adoption of smart thermostats instead of non-programmable and programmable thermostats over the period 2020–2050.
Total Addressable Market
Much of the energy savings benefits of smart thermostats requires Internet at home, so for the total addressable market, the number of Internet-enabled households was used. This was calculated by estimating the relationship between the number of Internet-enabled households and the global average gross domestic product (GDP) per capita using United Nations and International Telecommunications Union (ITU) data. Projections for GDP per capita were then used to project how the number of Internet-enabled households would change over 2020-2050. This indicated, for instance, 1,163 million Internet-enabled households in 2018.
Current adoption of smart thermostats is roughly 3 percent of the market globally, primarily in the US and Europe (37 million households).
Impacts of increased adoption of smart thermostats from 2020-2050 were generated based on two growth scenarios, which were developed by extrapolating projections from published historical sources. Each scenario was assessed in comparison to a Reference Scenario, where the solution’s market share was fixed at the current levels.
- Scenario 1: This scenario uses the projection from the Berg Insight data. A second-degree polynomial curve ft was used which had an r-squared value of 100 percent.
- Scenario 2: This scenario uses the projection from the Bloomberg BusinessWeek data. A second-degree polynomial curve ft was used which had an r-squared value of 99.9 percent.
Emissions numbers came from electricity and fuel consumption average values with grid emissions factors and natural gas emissions factors using data from the Intergovernmental Panel on Climate Change (IPCC).
First costs for conventional and smart thermostats were derived from the averages of 17 and 18 data points, respectively, from retailer websites covering the US, UK, and EU. No installation costs were assumed. This produced an average conventional price of US$38 and an average smart thermostat price of US$182. A 13 percent learning rate was applied to the solution, which came from data on air conditioners. Operating costs were taken as the electricity cost for cooling and fuel costs of heating homes using data for the US, EU, and China.
The smart thermostats solution was integrated with others in the Buildings Sector by first prioritizing all solutions according to the point of impact on building energy usage. This meant that building envelope solutions like insulation were first, building systems like building automation systems were second, and building applications like heat pumps were last. The impact on building energy demand was calculated for highest-priority solutions, and energy-related smart thermostats input values were reduced to represent the impact of higher building envelope solutions.
Scenario 1 forecasts that 1,453 million households could have installed a smart thermostat by 2050. The climate and financial impacts for this accelerated adoption of smart thermostats are significant: 7.0 gigatons of carbon dioxide-equivalent greenhouse gas emissions avoided. The marginal capital cost of this would be US$155 billion, but it would save US$1.8 trillion in lifetime operating costs due to reduced energy consumption for space heating and cooling. Based on the financial impacts alone, it is clear that global adoption of smart thermostats is economically viable and could provide a significant return on investment. The impacts of Scenario 2, where 1,589 million homes adopt are higher, at 7.4 gigatons carbon dioxide and US$2.1 trillion in lifetime savings. The cost is only marginally higher at US$172 billion.
The benefits that smart thermostats provide are substantial enough that it is not unlikely that they will become a replacement technology for mechanical or programmable thermostats, although widespread adoption will take time.
The high up-front cost of smart thermostats has inhibited growth; other barriers to adoption include access to Internet and appropriate heating and cooling systems. In many developing countries with warm climates, the lack of centralized digital air-conditioning systems inhibits high adoption, but the data collected indicate that the financial benefits are similar between heating and cooling, so smart thermostats could grow with the increased use of newer air condition systems.
As new competitors enter the market, the price of smart thermostats is expected to drop, and as rates of household Internet access and centralized heating and cooling systems grow, adoption will accelerate. Growth could be further driven with government and utility support and through the development of programs that demonstrate to consumers the benefits of smart thermostats.
 From the AMPERE MESSAGE3 Model.
 Current adoption is defined as the amount of functional demand supplied by the solution in 2018. This study uses 2014 as the base year.
 All monetary values are presented in 2014 US$.
 Although we use the term “priority,” we do not mean to say that any solution is of greater importance than any other, but rather that for estimating total impact of all building solutions, we simply applied the impacts of some solutions before others, and used the output energy demand after application of a higher-priority solution as the energy demand input to a lower-priority solution.