Closing in on Zero-Energy Buildings
Producing new commercial buildings that use 80 percent less energy than today's buildings is a new target in the fight against global climate change. If such a building's remaining energy consumption were supplied by clean, carbon-neutral renewable energy, it would be responsible for little or no greenhouse gas emissions. Lawrence Berkeley National Laboratory (Berkeley Lab) researchers are developing technologies to help make this possible.
Buildings in the United States today consume 72 percent of electricity produced, and 55 percent of U.S. natural gas use. They account for about 40 percent of both total U.S. energy consumption (costing $350 billion per year) and greenhouse gas emissions. Therefore, reducing the GHG emissions associated with buildings is essential to reducing overall U.S. emissions.
Before such ultra-efficient buildings become commonplace, researchers will need to develop new, more cost-effective technologies. Demonstration projects dating back to the 1990s suggest that it is possible to build these ultra-efficient buildings currently, but at great expense. Berkeley Lab research is aimed at developing advanced new technologies so that such buildings can be constructed economically and in large numbers.
Scientists at Berkeley Lab's Environmental Energy Technologies Division—in partnership with United Technologies Corporation's Research Center and the University of California at Berkeley, Merced, and Santa Barbara—have embarked on research aimed at developing computer models and building control algorithms and performance monitoring systems for the low energy buildings of the near future. The research is funded by the U.S. Department of Energy and the California Energy Commission's Public Interest Energy Research Program.
The research teams are performing field testing and demonstration work of new technologies on the campus of the University of California, Merced. The UC Merced campus has five buildings which have received Leadership in Energy and Environmental Design (LEED) ratings—four gold and one silver—for sustainable design. The campus has been designed to be a living laboratory, with sensors and instrumentation to support the development and demonstration of energy-efficient technologies and practices.
Some of these campus buildings already achieve a high degree of energy efficiency. "We have a goal that buildings consume half the energy and demand of other university buildings in California," says John Elliott, UC Merced's Manager of Engineering, Energy, and Sustainability Facilities. "We have statistically significant benchmarks based on data from other campuses, and we have been phasing in this goal." The campus's first round of buildings were designed to use 80% of the energy benchmark, and the current round of buildings are being designed to use 65% of that benchmark. The next phase of buildings will be designed to use 50% of the benchmark. However, according to Elliot, through careful design and energy management, many campus buildings have already surpassed those performance goals. The campus is therefore the perfect place to begin testing new hardware and software technologies designed to drive building energy efficiency to even higher performance levels.
The Ultra-Low Energy Strategy
When buildings are designed so that their systems work together to maximize energy efficiency, they can use substantially less energy than they do on average today—even as they provide heating, ventilation, air conditioning, lighting, and electrical power at the outlet. For this enhanced performance level to be achieved, however, the building needs to be constructed according to specification, which is not always the case currently.
"Today, there is usually no transfer of knowledge and design specifications from the design phase to operations," says Philip Haves, Leader of Environmental Energy Technologies Division's (EETD's) Commercial Building Systems Group. "Very few new buildings are commissioned and most building facilities management staff don't have access to the design specifications of building systems they maintain."
Commissioning is originally a naval term applied to the process of ensuring a ship's seaworthiness. Newly constructed buildings need to be commissioned to determine if the energy savings and performance intended for the building in the original design has actually been achieved. When a building is commissioned, a team of engineers start the various building systems, test them to ensure they are operating according to design specifications, and make adjustments if they are not. Commissioning of both new and existing buildings could save billions of dollars in U.S. energy costs, according to Berkeley Lab research estimates.
Finally, buildings need to be operated by a facilities staff trained in the use of sensors and monitoring systems that provide accurate, real-time information about the energy performance and environmental conditions within the building.
This approach to improving building efficiency selectively applies systems-engineering methods that have transformed other industries, including the aircraft and automobile industries.
Wanted: Real-Time Building Energy Use
PC-based monitoring systems could reveal a new world of real-time information to facilities staff about how energy is being used in buildings. However, no standard hardware and software building design and operating platform exists—nothing comparable to the operating system software common to many personal computers, for example.
A problem today is that "we don't measure the energy use of buildings in real-time," says Mary Ann Piette, the Deputy Head of EETD's Building Technologies Department. "Most energy management systems in buildings are geared toward controlling energy use, but without much feedback about how buildings are actually using energy."
Further, there is no performance standard for buildings—not for energy use, occupant comfort, average maintenance cost, or any other building metric one could imagine. Most LEED standards are design oriented—the builder obtains a LEED rating by specifying various energy efficiency, water, materials, and transportation-related measures. These elements are addressed during construction, however, it is rare for the building to be monitored once it is occupied, to determine how well it really performs relative to the initial goals.
Performance benchmarks would help facilities managers do a better job of operating buildings for energy-efficiency, as well as for other qualities such as occupant comfort. A performance benchmark is a data set of performance metrics for a building's operation, energy use, and conditions.
Comparison of the performance metrics in different buildings of roughly the same size and type can help a facilities manager identify specific problems in a building. For example, perhaps the lighting is efficient, but air conditioning energy use during the summer months is higher than normal for a building of that type, leading to the deduction that a chiller plant is underperforming. Benchmarks for each energy-using system in the building provide managers with the target data they need to ensure that each system is operating efficiently.
Another need is for a broad, deep data set about performance of many (preferably thousands) of buildings in real-time—known as a building informatics repository. This repository would not only help building managers do a better job, it would help building scientists develop better control strategies for managing building energy use.
Berkeley Lab researchers are conducting three preliminary research and development projects aimed at developing a building design and operating platform and a building informatics repository.
Model Predictive Control
Berkeley Lab EETD researchers Philip Haves and Michael Wetter, the University of California's Francesco Borelli, John Elliot of UC Merced, and team members from the United Technologies Corporation Research Center are testing a model predictive control (MPC) of a central chilled water system that provides cooled air to campus buildings.
Using the model predictive control method, a computer model of the chilled water plant simulates the effect of various control options on the performance of the system and provides input to the real-time control system for the chilled water plant so that it can properly supply the cooling needed to maintain comfortable conditions while minimizing energy costs.
The computer models of the chillers, the cooling towers, the storage tank and the pumps are based on the manufacturers' performance data and then fine-tuned to the measured system performance. The team estimates that energy savings of 10 to 20 percent of total HVAC energy use is possible using MPC. Data from 2009 summer field tests at UC Merced will be used to determine the actual savings and demonstrate the technology.
Michael Sohn, Group Leader of EETD's Airflow and Pollutant Transport Group, along with colleagues from UC Merced and United Technologies, are studying the use of sensors and occupancy-estimating methods to control a building's lighting and heating, ventilation, and air conditioning (HVAC) systems.
The guiding concept behind this work is that if the occupancy density of a building is well known, its lighting and HVAC energy consumption in a typical office building during the cooling season may possibly be reduced by 10 to 20 percent. What is needed is a hardware and software tool that determines the occupancy in different parts of the building accurately in real time.
Currently, motion sensors only tell you if a space is occupied, not how many people are there. Carbon dioxide (CO2) sensors are better indicators of how many people are in a room, but they are relatively slow to react—people come and go quickly, but CO2 levels change slowly.
Sensors that can estimate quickly how occupancy is changing in a space may be able to provide data that allow a control system to reduce lighting and HVAC services to unoccupied spaces quickly. For example the control system could allow room temperatures to float beyond their occupancy-based set points and rapidly increase services in time to meet anticipated needs.
The research team will build smart, low-resolution sensor networks and develop data processing algorithms, occupancy dynamics models, and energy control methods. They will test whether these technologies can provide sufficient data on building occupancy to make intelligent HVAC and lighting control possible at buildings on the UC Merced campus. The research will also benchmark the energy savings possible in buildings at different occupancy levels.
Visualizing Building Energy Performance
Piette and her colleagues are also developing visual tools for communicating the energy performance of buildings in real time so that improved information can help facilities staff better manage their buildings. Collaborating institutions include United Technologies, UC Merced, and UC Santa Barbara.
The team is developing tools to compare the measured energy performance of buildings to baselines derived from computer simulation, benchmarking, and previous performance measurements. They plan to develop new methods of estimating quantities that are not measured in the building, such as space heating and cooling loads (the power demand to keep the building properly heated and cooled), to help building facilities staff diagnose problems.
The methods and tools will correct for uncertainties in measurements and simulations systematically, something that is typically not done in the building industry today. The researchers are also developing software for visualizing building conditions and performance designed for decision-makers at different facilities management levels.
Piette's team will test the prototype monitoring system in the Science and Engineering I building and on the central plant on the UC Merced campus.
The goal of this technology development is to reduce energy consumption and electricity demand by 20 percent or more in large commercial buildings, by providing actionable energy performance information to facility managers and operators.
For more information, contact:
- Philip Haves
- (510) 486-6512
- Mary Ann Piette
- (510) 486-6286
- Michael Sohn
- (510) 486-7610
For more information about commercial buildings research at Berkeley Lab visit the Commercial Buildings Research and Development web site.
Model Predictive Control at UC Berkeley
LEED-rated buildings at UC Merced