Design and experiments of a thermoelectric-powered wireless sensor network platform for smart building envelope

Highlights

• First self-powered wireless sensing platform entirely inside the window frame.

• Milli-watt level power management with thermoelectric energy harvesting.

• Energy equilibrium analysis proposed for the self-powered design at any location.

• Platform robust to various sensors and contributes the smart building envelope.

Abstract

A new thermoelectric-powered wireless sensor network platform is presented for the low-cost environmental sensing in building envelopes through thermoelectric energy harvesting and ultra-low power management. It is designed and prototyped entirely inside a window frame without compromising architectural aesthetics. This self-powered sensing platform is achieved by maximizing the harvested energy from building envelopes and optimizing the wireless sensing unit’s energy consumption. It harvests milli-watt level thermoelectric power from the temperature gradient across building envelopes through thermoelectric generators with an optimized thermal connector. The harvested energy is voltage-boosted and regulated through two integrated circuits that are tailored for ultra-low-power input. A low power system-on-chip is used to supervise the environmental sensing and wireless data communication. The energy consumption is tailored by adjusting the system sleep time to match the harvested energy. The proposed platform is prototyped in a window frame and thoroughly tested, where 1.5 mW of power is harvested from thermoelectric generators under 6 °C of temperature difference, and a 33.4% efficiency to the battery. In the meantime, 0.42 mW power is consumed by the wireless sensing unit under a sampling period of 2 h, which reaches the energy equilibrium state. The energy equilibrium algorithm can project the battery energy based on historical weather conditions, so as to achieve the self-powered condition given a geographic location. This smart building envelope systems include the unique innovations in self-powered system architecture, thermally optimized internal structure, and milli-watt level power management.

Introduction

A grand challenge emerging in the energy system is to address the critical national needs of energy security, economic competitiveness, environmental responsibility, smart serviceability, and workability simultaneously. The building sector is the largest consumer of electricity. It was reported that the buildings sector accounts for about 76% of electricity use and 40% of all United States primary energy use and associated greenhouse gas (GHG) emissions [1]. Globally, about 30% of total energy and 60% of electricity are consumed on buildings [2]. The primary energy consumptions in buildings are heating, ventilation, air conditioning, lighting, and other major appliances (water heating, refrigerators, dryers, etc.) [3]. The share of electricity use in building sectors has grown dramatically from 25% of the United States annual electricity consumption in the 1950 s to 40% in the early 1970 s, and to more than 76% by 2012 [1]. In this sense, smart building management becomes significantly important to achieve optimal interior comfort with minimal energy expenditure. The smart building concept emerged at the beginning of the 1980 s and has been continuously evolving during the past decades [4], [5], [6], where the ability to adapt to the dynamic environments is considered as the central aspect in its definition [7]. Research shows that a potential energy saving of 34.78% could be achieved by the smart buildings comparing to conventional buildings, through the optimal control of major building components like the heating, ventilation, air conditioning, and lighting systems [8].

The concept of the smart building has attracted significant attention in the civil engineering industry during the past decade and is estimated to worth 53.45 billion by 2022 [9]. Significant attention has been paid to the controlling algorithms of the smart building management systems (BMSs) thanks to the boom of artificial intelligence, with many discussions about energy consumption forecast [10], [11], [12], smart grid technology [13], and in-room occupancy detection [14]. In contrast, the importance of the sensor network was underestimated with fewer research and discussions. Traditional systems rely heavily on wired sensors to monitor the environment, which comes with substantial workforce and time in both installation and maintenance. Therefore, the concept of the wireless sensor network (WSN) becomes more and more popular in BMSs to ease the installation and maintenance, to enhance the flexibility in dangerous areas, and to reduce the cost of sensor deployments [15]. However, most of these sensors are powered by either replaceable or rechargeable batteries, which means they are exposed to limited battery life, environmental hazards, and complex replacement procedures [16]. For certain applications in the building environment, it could be difficult to replace the battery of WSN and could result in an enormous scale of maintenances cost.

The energy harvesting technologies provide an excellent solution to elongate the battery life and reduce the maintenance cost of a WSN such that the self-powered wireless sensor network (SPWSN) system can be achieved. Energy harvesting technologies use power generating elements such as solar cells [17], piezoelectric elements [18], electromagnetic components [19], [20], and thermoelectric elements [21] to convert light, pressure, vibration, and heat energy into electricity. A comprehensive review of energy harvesting applications in the buildings can be found in [22], from which photovoltaics and thermoelectricity are two primary energy sources for the WSN systems in buildings. A solar energy based SPWSN was proposed by Tzortzakis et al. [23] for the environmental monitoring and smart city design, with the corresponding field test conducted on the campus of the National Technical University of Athens. The solar panels provided abundant energy supply up to 3.5 W but also resulted in a relatively bulky size of the prototype. Moreover, since solar energy enjoys a much higher power output than other energy harvesting sources, the proposed system is not robust to the cases where solar energy is not available. The thermoelectric (TE) energy serves as a potential alternative when solar energy is not available. Compared to solar energy, TE energy harvester suffers from much lower power and voltage output, which cast challenges to the conversion efficiency and energy equilibrium for the WSN. Recently, a new TE energy harvester for the building energy management WSN was proposed by Wang et al. [24] by harvesting the wall-mount heater with 60 °C surface temperature, and a 25% end-to-end efficiency was achieved with 1.7% duty cycle. A higher efficiency TE-based WSN system is later proposed by Guan et al. [25], with the end-to-end efficiency ranging from 44.2% to 75.4% under different open-circuit voltages. The system was built based on a two-stage boosting scheme with tens of microwatts to several milliwatts power input from thermoelectric generators (TEGs). However, both systems mentioned above are tested in the lab and did not discuss the potential applications in the building industry.

The building envelopes, as the skin of buildings and the main barrier against the external environment, are crucial to the environmental sensing and energy harvesting [26]. Some well-known applications include the building integrated photovoltaic-thermal roofing systems [27], sun-powered smart window blinds [28], and wind energy harvester inside double-skin façades [29]. However, for sensing applications inside building envelopes, no physical energy source, such as solar or wind, is available. This limitation creates challenges to energy harvesting and WSN applications for the building envelopes. The traditional building envelope designs tend to break the heat transfer and minimize the thermal infiltration by stacking as many thermal resistant materials as possible. Therefore, a thermal break up to $25°C$ can be easily formulated across the profile, as is schematically shown in Fig. 1. The temperature difference across the building envelope could provide tens of micro-watt to a few milli-watt level of thermal energy harvesting and is a good power supply for WSN nodes. However, due to the challenges of thermal energy harvesting, i.e. the ultra-low power and voltage output, and the limited space inside building envelope profile, no study or design of self-powered WSN system has been observed in the building envelopes.

In this paper, a thermoelectric-powered wireless sensor network (TPWSN) platform is proposed for low-cost environmental sensing in building envelope components and is prototyped into a building window system. The platform is designed entirely inside the window frame with no wired connections to the outside and therefore has no visual compromise to the window’s architectural aesthetics. To the best of the authors’ knowledge, it is the first self-powered WSN system inside the building window systems and robust to other building envelope structures such as façades and roofs. With a slight modification of the window structure and a minor compromise of its thermal performance, the platform takes advantage of the temperature difference across the window frame for the thermoelectric energy harvesting. It achieves the energy equilibrium between the consumption and the harvesting, which significantly elongates the limited battery life of traditional WSN systems. In energy harvesting, a thermally optimized energy harvesting structure is proposed inside the limited space of window frame for an optimal power output, which is supported by heat transfer analysis through the finite element method. The milli-watt level thermoelectric power is voltage-boosted and regulated through two integrated circuits tailored for ultra-low-power input. In the meantime, a low power system-on-chip (SOC) is applied to supervise the environmental sensing and wireless data communication, and is thoroughly tested for efficiency and energy consumption. The total energy consumption is tailored by adjusting the system sleep time to match the harvested energy so that the self-powered design could be realized. Comprehensive discussions of the design methodologies and experimental validations are elaborated for energy harvesting and the wireless sensor subsystems. An energy equilibrium algorithm based on the test performance is proposed to predict the battery energy level and achieve the self-powered design by properly engineering key design parameters. The self-powered system architecture and energy equilibrium algorithm, the thermally optimized internal structure, and the milli-watt level power management of the proposed TPWSN serve as valuable references and guidance to the design and applications of the next generation smart building envelope systems.

The rest of the paper is organized as follows. Section 2 focuses on the architecture and the design philosophy of the TPWSN system. The energy availability across the window profile and the thermally optimized harvester are first analyzed and designed in Section 2.1. The energy harvesting subsystem and wireless sensor network subsystem are designed and discussed in Section 2.2 and Section 2.3. The energy equilibrium algorithm is theoretically formulated in Section 2.4 based on the key design parameters to aid the self-powered design. In Section 3, the subsystems are integrated into the window frame and tested under a controlled temperature gradation, where the amounts of energy consumption and energy harvesting are continuously monitored and calculated. Section 4 presents the energy equilibrium design and the battery energy prediction based on the case study of the historical temperature data in New York City to theoretically achieve eternal battery life. Such a self-powered design could be extended to any location with proper engineering of the key parameters. Section 5 discusses the cost analysis and potential limitations of the TPWSN system under real applications. Section 6 summarizes the paper with conclusive remarks.