A group of researchers at UC Santa Barbara is looking to play a major part in that reduction. Professors Bud Homsy, Igor Mezic, Jeff Moehlis, João Hespanha, and Rich Wolski are leading research efforts into integrated building design, in which active control of indoor airflows could greatly improve the ventilation and efficiency of heating and cooling in buildings.
Current technologies—existing hardware combined with energy-efficiency modeling tools and algorithms, such as the Department of Energy’s “EnergyPlus”—offer energy savings of 10% to 30% when applied to the retrofit of extant buildings and the design of new ones. Far greater savings can be achieved, however, with modern analysis and control tools based on dynamical systems and control theory, when these tools are used to optimize the performance of the building as a fully integrated system.
Large buildings equipped with heating, ventilation and air conditioning (HVAC), data centers, and a myriad of sensors and wireless communication devices are complex systems whose operation includes multi-physics and multi-scale effects. Building systems are dynamically uncertain with respect to both the energy load and the environment, with dramatic changes in the number of occupants in the building, their energy demand, and ambient weather conditions.
A smart building containing an array of sensors and an integrated, optimized control system could dynamically adjust lighting and HVAC flows based on actual, real-time presence rather than scheduled occupancy. Energy and money are not wasted on cooling and lighting empty rooms. Mixing of airflows is a particularly important topic for energy efficiency in buildings, since good mixing avoids the wasteful temperature stratification which occurs, for example, when hot air rises to the ceiling and cold air falls to the floor.
Recent research in Mezic’s laboratory has established modeling techniques that enable design of efficient mixing in a variety of ambient conditions—the airflow can be directed to cool or heat the occupants much more efficiently. By combining sensor networks, active ventilation, adjustable lighting, and adjustable windows and doors into integrated and optimized compound systems, much higher quality living and working environments are possible, with 50% greater energy efficiency than current systems offer.
Data centers are an important, special case, whether as occupants of dedicated areas in general-use buildings or as single-purpose data center buildings. In data centers, there are few people but large heat loads from high power-density server arrays. Data center growth is causing the energy density of buildings to grow at a rate that is expected to be proportional to internet and server demand—and that demand is expected to double in the next few years.
Efficient energy use in data centers, far more than in typical office spaces and commercial buildings, requires especially efficient ways of redistributing cooling energy. Data centers present special and unique challenges, due to the coupling between the computer use and cooling systems. Mezic and Rich Wolski are spearheading an effort by faculty in UCSB’s Mechanical Engineering and Computer Science departments, developing new approaches and solutions for this special class of buildings.