Distributed and localized cooling with thermoelectrics

Abstract

G. Jeffrey Snyder is a Professor of Materials Science and Engineering at Northwestern University in Evanston Illinois. His interests are focused on engineering of electronic and thermal properties, and he is well known for his work on thermoelectric materials. He has developed new methods of electron band structure engineering and microstructure engineering of thermal and electrical properties of complex materials. His interdisciplinary approach stems from his background in solid state chemistry at Cornell University and the Max Planck Institute for solid state research, applied physics at Stanford University, and thermoelectric materials and device engineering at NASA/Jet Propulsion Laboratory and California Institute of Technology (Caltech). G. Jeffrey Snyder is a Professor of Materials Science and Engineering at Northwestern University in Evanston Illinois. His interests are focused on engineering of electronic and thermal properties, and he is well known for his work on thermoelectric materials. He has developed new methods of electron band structure engineering and microstructure engineering of thermal and electrical properties of complex materials. His interdisciplinary approach stems from his background in solid state chemistry at Cornell University and the Max Planck Institute for solid state research, applied physics at Stanford University, and thermoelectric materials and device engineering at NASA/Jet Propulsion Laboratory and California Institute of Technology (Caltech). The energy use and dramatic environmental impact of refrigerants collectively imposes the need to transform cooling technology to address climate change in a significant manner. Heating, ventilation, and air conditioning (HVAC) accounts for over 25% of electricity use in the USA, and by 2060 energy use from cooling could overtake heating. Refrigerant management, i.e., replacing the high global-warming-potential refrigerants themselves, is the largest (10%) component of the climate change solution proposed by Project Drawdown (see Note S1). Enhancing energy efficiency through smart, efficient HVAC systems could transform energy usage and facilitate rapid integration of renewable energy. In the current centralized architecture for heating and cooling, a large fraction of the energy consumed is simply wasted. HVAC systems are typically designed to heat or cool an entire building or vehicle in the worst-case scenario,1Sekhar C. Anand P. Schiavon S. Tham K.W. Cheong D. Saber E.M. Adaptable cooling coil performance during part loads in the tropics—A computational evaluation.Energy Build. 2018; 159: 148-163Crossref Scopus (9) Google Scholar and this results in continuous operational inefficiency.2Paliaga G. Zhang H. Hoyt T. Arens E. Eliminating Overcooling Discomfort While Saving Energy.ASHRAE J. 2019; https://www.ashrae.org/technical-resources/ashrae-journal/featured-articles/eliminating-overcooling-discomfort-while-saving-energyGoogle Scholar Such “overcooling” and then reheating is ubiquitous because platform designs emphasize reliability, simplicity, and initial cost even if energy consumption, user comfort, and health are more valuable to the operators and occupants. Such energy intensive, centralized heating and cooling has been the design theme for many decades even though component technologies have advanced significantly. Heating or cooling only where and when it is needed (referred to as distributed HVAC) (Figure 1) builds upon the theme of localized as opposed to centralized heating or cooling. Distributed HVAC has the potential to dramatically improve the energy efficiency and reduce the overall energy consumption of a variety of platforms including residential and office buildings, vehicles, storage containers, and warehouses. Instead of heating/cooling an entire building, everywhere, all the time, it can be much more efficient to heat or cool only the occupants or objects that need it, and only when and where discomfort is sensed.4Andersen M. Fierro G. Kumar S. Andersen M.P. Kim J. Arens E.A. Zhang H. Raftery P. Culler D.E. Well-connected microzones for increased building efficiency and occupant comfort.in: CEEE Summer Study on Energy Efficiency in Buildings. 2016Google Scholar Electric vehicles will likely be an early adopter of distributed HVAC (Figure 1) because every watt spent on climate control results in reduction of the vehicle range or adds to battery cost and weight. Distributed heating and cooling systems are anticipated to sense occupants’ discomfort and then automatically control heating or cooling of seats, steering wheel, windows, etc., as well as direct airflow where it is needed. Where today’s car can require up to 5 kW for the HVAC system, tomorrow’s might only use 100–200 W to cool the car seat and other accessories.5Bell L.E. Cooling, heating, generating power, and recovering waste heat with thermoelectric systems.Science. 2008; 321: 1457-1461Crossref PubMed Scopus (3909) Google Scholar The whole building environment does not need to be at the same temperature; rather, the occupants’ local environment and the building’s global environment can have separate temperatures. Advances in controls and sensing, supported by the internet-of-things, have reached a level that can address the requirements for implementing distributed HVAC in the built as well as the transportation environment.6Wen J.T. Mishra S. Intelligent Building Control Systems - A Survey of Modern Building Control and Sensing Strategies.First Edition. Springer, 2018Crossref Google Scholar Examples for these advanced controls would be net-zero commercial buildings such as “The Edge” in Amsterdam, and One Angel Square in Manchester, UK. Smart building control systems can also achieve grid integration, wherein a building can participate in demand response to lower energy costs and/or shift loads to times when electric power is available from renewable resources. Separate HVAC zones for individual rooms is known to provide better comfort and improved energy efficiency. Distributed HVAC for large buildings typically involves hot and cold water to store heat and move it from room to room. Smaller buildings and individual homes can simply distribute the humidity and temperature-controlled air to select zones when it is needed. Such systems have relatively low cost but only take limited advantage of the distributed HVAC concept. Micro-distributed HVAC on-demand concepts might include individual bed, seat, desk, floor, and surface heating and cooling.4Andersen M. Fierro G. Kumar S. Andersen M.P. Kim J. Arens E.A. Zhang H. Raftery P. Culler D.E. Well-connected microzones for increased building efficiency and occupant comfort.in: CEEE Summer Study on Energy Efficiency in Buildings. 2016Google Scholar Enabling this distributed architecture will be a new generation of associated sensors and controls. Individual micro-climates would then unobtrusively follow each individual throughout the day. The micro-distributed heat pumps for both heating and cooling could reject heat to the room HVAC controlled by the macro-distributed system. Although a micro-distributed cooling device will add heat to a room, the target room temperature can be raised leading to an overall decrease in room HVAC needed. Solid-state heat pumps, such as thermoelectric (TE) Peltier coolers will enable micro-distributed HVAC. These economical small semiconductor devices can instantaneously provide heating or cooling with no moving parts, no noise nor harmful liquids or gases. Thus, solid-state TE heating/cooling has potential to provide transformative effects on the environment, platform design, user experience, and energy demand. The coefficient of performance (COP), j, defined as the heat removed, Qc, divided by the electrical power consumed, W, is limited by thermodynamics to be less than that of a Carnot engine, ΔT/Tc, where Tc is the temperature (in Kelvin) being cooled and ΔT is the difference between Tc and the heat rejection temperature Th.ϕ=QcW=TcΔTϕr For a typical thermoelectric cooler (TEC) (shown in Figure 2), the fraction of Carnot COP, jr, is 10%–20% and increases with device ZT. Thermoelectric (Peltier) coolers are commonly used in systems requiring only 100 W of cooling or less because of cost and low COP at large ΔT. However, thermoelectrics can be competitive at low ΔT (Figure 3). Although vapor compression cycles (VCCs) can exceed 50% of Carnot under ideal conditions, their COP does not improve substantially when operating at smaller ΔT like a TE system. The USA requirements for EnergyStar certification, COP > 3.7 for ΔT of 8K (EER = 12.5 under EER test conditions) (see Note S2), is only 9.8% of Carnot. This is close to a thermoelectric figure of merit of 0.5, which is easily obtainable with commercial TE coolers even considering heat exchanger losses.7Wang D. Crane D. LaGrandeur J. Design and Analysis of a Thermoelectric HVAC System for Passenger Vehicles.in: SAE 2010 World Congress & Exhibition. SAE International, 2010Crossref Scopus (11) Google Scholar Larger ΔT might be needed for humidity control (or required for refrigeration) where VCC systems can maintain good COP, whereas TE systems tend to have a constant jr. Complex heat exchanger designs can be devised that reduce the ΔT requirement, making the COP of TE systems competitive to VCC. In a micro-distributed HVAC system, a change of only a few degrees could be sufficient to provide comfort. Advanced heat exchanger technologies5Bell L.E. Cooling, heating, generating power, and recovering waste heat with thermoelectric systems.Science. 2008; 321: 1457-1461Crossref PubMed Scopus (3909) Google Scholar that reduce ΔT will be an enabling technology for the use of small, distributed cooling systems. The cost of small heat pumping systems is typically dominated by the heat exchanger. Thus, the cost in US dollars per W of a micro-distributed HVAC system (such as the seat in Figure 2), is typically controlled by heat flow system cost and ZT (which determines the relative COP).10Leblanc S. Yee S.K. Scullin M.L. Dames C. Goodson K.E. Material and manufacturing cost considerations for thermoelectrics.Renew. Sustain. Energy Rev. 2014; 32: 313-327Crossref Scopus (302) Google Scholar For systems meeting these minimum efficiency requirements, cost becomes the main driver. Large VCC systems have cost, COP, and ΔT advantage. The cost of the VCC reduces slowly as its size is reduced, making them easy to oversize.1Sekhar C. Anand P. Schiavon S. Tham K.W. Cheong D. Saber E.M. Adaptable cooling coil performance during part loads in the tropics—A computational evaluation.Energy Build. 2018; 159: 148-163Crossref Scopus (9) Google Scholar The cost of TE, in comparison, typically scales linearly with size,10Leblanc S. Yee S.K. Scullin M.L. Dames C. Goodson K.E. Material and manufacturing cost considerations for thermoelectrics.Renew. Sustain. Energy Rev. 2014; 32: 313-327Crossref Scopus (302) Google Scholar making them ideal for small systems. Thus for cooling power of 1 kW or more VCC is always used, but for applications requiring 100 W or less, thermoelectric Peltier coolers are preferable.11Vining C.B. An inconvenient truth about thermoelectrics.Nat. Mater. 2009; 8: 83-85Crossref PubMed Scopus (624) Google Scholar Improvements in TE materials, heat exchangers, and controls as well as cost reductions will gradually push the market for Peltier cooling to larger systems starting with small refrigerators and micro-distributed HVAC units. Heat pumps work like a refrigerator in reverse, providing thermal heating power greater than the input electrical power (Qh/W > 100%) by pumping heat from the cool ambient. Heat pumps can be highly efficient heating —even more efficient than using waste heat that is available at seemingly no cost (e.g., from power plants).12MacKay D.J. Sustainable Energy-without the hot air. UIT Cambridge Ltd, 2009Google Scholar Even more localized are wearable thermoelectric devices13Kishore R.A. Nozariasbmarz A. Poudel B. Sanghadasa M. Priya S. Ultra-high performance wearable thermoelectric coolers with less materials.Nat. Commun. 2019; 10: 1765Crossref PubMed Scopus (75) Google Scholar to provide immediate heating and cooling at an exact location on the body. Beyond improving comfort by managing body heat within healthy levels,14Zhang H. Arens E. Zhai Y. A review of the corrective power of personal comfort systems in non-neutral ambient environments.Build. Environ. 2015; 91: 15-41Crossref Scopus (194) Google Scholar such systems can affect the perception of thermal comfort simply by adding user control.15Zhou X. Ouyang Q. Zhu Y. Feng C. Zhang X. Experimental study of the influence of anticipated control on human thermal sensation and thermal comfort.Indoor Air. 2014; 24: 171-177Crossref PubMed Scopus (30) Google Scholar By making a slightly warmer room more comfortable, overall energy demand can be reduced. The flexible technology rapidly being developed for wearables can also be incorporated in furniture, beds, and other accessories in direct contact with the occupant. For portable cooling systems, their power supply will be an additional issue. Again, for effective use an integrated system design including the heat exchanger is imperative.13Kishore R.A. Nozariasbmarz A. Poudel B. Sanghadasa M. Priya S. Ultra-high performance wearable thermoelectric coolers with less materials.Nat. Commun. 2019; 10: 1765Crossref PubMed Scopus (75) Google Scholar Much of the research in TE is focused on new thermoelectric materials. For distributed HVAC ZT > 0.5 might be sufficient providing opportunity for non-toxic and recyclable materials. Devices will need to be durable as well as inexpensive. However, for such ideas to be economically transformational and impactful to society requires a convergence of technology, design, and consumer behavior. Systems are likely to first appear in the automotive industry because this convergence is ingrained, and energy efficiency needs of EVs are so pressing. For the built environment, aspects of a distributed HVAC concept have been proposed and even demonstrated on a small scale, but the concept has not proliferated because of a lack of convergence. Beyond a techno-economic analysis to estimate market impact, a coherent collaboration is needed to gauge consumer and regulatory acceptance as well as life cycle analysis to predict the impact on sustainability. Ultimately, the convergence of all stakeholders—engineers, designers, builders, owners, regulators, and users—is needed for the adoption of energy-efficient distributed HVAC technology in residences to have an impact on personal comfort, society, and climate change. We acknowledge support from an NSF ERC Planning Grant 1936896 . G.J.S. acknowledges NSF DMREF award# 1729487. We thank Tom Radcliff and Christian Werner for helpful discussions. For L.B.’s contributions, this material is based upon work supported by the National Science Foundation Graduate Research Fellowship Program under grant number DGE-1842165 . Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. G.J.S., S.L.B., D.C., H.P., C.E.F., A.R., and S.P. devised the idea with other participants at a workshop organized by S.P. G.J.S., S.L.B., D.C., and LB analyzed the COP reports. All authors wrote and revised the manuscript. Download .pdf (.21 MB) Help with pdf files Document S1. Figure S1 and Notes S1 and S2