Heating, ventilating, and air conditioning (HVAC) systems account for significant portion of residential buildings’ energy consumption. Previous studies have shown the growing importance of smart thermostats. However, many existing studies are based on reviews, simulations, or large datasets rather than simple real-world studies. This paper studies a forecast informed thermostat control strategy for improving heating efficiency in a single-family residential building in Minnesota with 2,25 ft² area where the building is equipped with a forced air natural gas furnace and a smart Sensi thermostat. With two weeks of study, week 1 was the data collection under a conventional fixed setpoint condition. This served as baseline data. For week 2, thermostat settings were manually adjusted using short-term weather forecasts.  The baseline week produced the model y = 8.946 – 0.1017x, while week 2 produced y = 7.149-0.06102x signifying the furnace runtime improved by 22.34% at 0⁰F compared to baseline week. This finding suggests that integrating smart thermostat data with short-term weather forecasts can provide a practical and low-cost approach for improving HVAC efficiency in residential buildings.

Keywords

 HVAC; smart thermostat; weather forecasting; HVAC efficiency.