Real-world experiments to reduce electricity costs by automatically reducing peak demand provide information that can be used to convince end-users these measures lead to significant monetary savings.
The state of California and the U.S. Department of Energy jointly sponsored a 2003–2006 study by the Lawrence Berkeley National Laboratory. The “Introduction to Commercial Building Control Strategies and Techniques for Demand Response” provides information on a range of control strategies that can be implemented to reduce both peak and average electrical usage. It looked at 56 different sites, including office buildings, a library, schools, a detention facility and retail establishments. Definitions from the report clarify key concepts.
Demand response (DR) strategies: “Demand response is dynamic and event-driven and can be defined as short-term modifications in customer end-use electric loads in response to dynamic price and reliability information.”
Two strategies for responding to DR signals initiated by an electric utility are “demand limiting,” in which a portion of the load is cut back, and “demand shifting,” which is implemented by changing the time at which the electricity is used.
Shared burden: “DR strategies that share the burden evenly throughout the facility are least likely to have negative effects on building occupants.”
Granularity of control “refers to the amount of floor area covered by each controlled parameter (e.g., temperature).” A more granular system allows the temperature setpoints throughout the facility to be separately adjusted according to the needs of each particular area.
Adjusting the central cooling or heating equipment, for example, can cause uncomfortable temperature extremes in some areas and troublesome temperature swings in sensitive areas like computer data rooms.
A November 2007 article in the ASHRAE Journal—“BACnet at Georgia Tech” by Donald P. Alexander, Cornelius Ejimofor and David G. Holmberg—details a 2006 project at the Georgia Institute of Technology, in which automated load shedding would cut back the air conditioning in response to a given trigger. Georgia Tech is a large energy user because in addition to heating and cooling classrooms, dorms and offices, it supports many research projects.
According to the article, “Large facility owners generally negotiate yearly or multiyear electric service agreements and often are given lower base rates in exchange for sharing the risk of price ?uctuations. This agreement may take the form of a capped rate or may go as far as a real-time market rate where the facility owner sees every price spike in the electric market. Price spikes due to normal summer heat, as well as unforeseen events such as power plant emergency shutdowns, occasionally can reach ?ve times the base utility rate, or higher. The challenge for the facility owner is how to reduce power consumption during periods of peak pricing, while maintaining mission critical building loads.”
For this project, Georgia Tech agreed to be charged at the fluctuating price rates, known as the real-time price (RTP). The university partnered with its electric utility company and a controls manufacturer to tie into the existing building control network and automatically trigger demand limiting when the real-time price rose above a certain level. It chose the electrical engineering instruction building, which doesn’t house research labs that must be held at very stable temperatures. Because the HVAC system was already under digital control, the university was able to interface it with the utility’s pricing database through the Internet. The control system was programmed, so when the price level for the following two hours was calculated to be above the agreed-on trigger level, the HVAC system would go into energy-savings mode. The technique was to adjust the temperature setback points in five different zones throughout the building. Since various areas in the building have different thermal load demands, the temperature set points are at different levels. Researchers chose to respond to load shedding commands by widening the heating and cooling offset ranges from ±2°F to ±5°F, rather than changing the setpoints. The nighttime range when the building was unoccupied was set to ±9°F.
The data in the study was extrapolated to predict that if the same strategies were extended campus-wide, the annual savings in energy costs would be $150,000. In addition, a “bene?t, beyond cost avoidance, is that Georgia Tech now has a method based on the dynamics of electrical cost to alter control parameters to allow more saving strategies in the future.”
This subject is very important, and there is so much information packed into these reports that this column will continue to cover them next month. Meanwhile, readers may wish to review the full reports.
The Georgia Tech report can be found at www.fire.nist.gov/bfrlpubs/build07/PDF/b07028.pdf.
The California report can be found at drrc.lbl.gov/drrc-pubs-auto-dr.html.
BROWN is an electrical engineer, technical writer and editor. He serves as managing editor for SECURITY + LIFE SAFETY SYSTEMS magazine. For many years, he designed high-power electronics systems for industry, research laboratories and government. Reach him at firstname.lastname@example.org.