- Auto-DR Successful
- Real-time Pricing Reports Outline Policy Recommendations
- Research Project: Establish the Value of Demand Response
- Research Project: Incentives and Rate Design for Efficiency and Demand Response
- Status Update on DRRC Research Projects
- DRRC Collaboration Opportunities
- DRRC Creates Research Roadmap for California
- Calendar
- Publications
- Subscribe
Auto-DR Successful
Large Commercial Buildings Reduce Load by up to 30%
Large commercial buildings have often been viewed as strong candidates for demand response (DR). A few large commercial facilities often account for a disproportionate share of the system load. Experience has shown that lighting, heating, ventilation and air conditioning (HVAC), and other facility loads often can be temporarily shifted, limited, or shed in order to reduce demand at critical times without minimally impacting building operations or tenant comfort.
DR program success has been difficult to achieve or sustain because of the complexity of managing multiple, complex building systems. While problems start with determining how to notify a building manager of a pending event, how the manager responds to that event poses even more problems. Manually turning loads on or off and changing set points is labor-intensive and costly. Manual control can also produce inconsistent results from one event to the next, and create adverse customer impacts in the process. A better, more automated way to manage facility load and demand response is needed.
Automated Demand Response or Auto-DR, a research program managed by the Demand Response Research Center (DRRC), is designed to link facility energy management control systems (EMCS) with external utility-generated price or emergency signals. The signals initiate pre-programmed, customer-defined strategies to shift, reduce or shed loads for brief periods of time. Pre-defining and automating the customer response through the facility’s EMCS can substantially reduce cost and complexity and provide a more consistent and reliable demand response. Auto-DR also provides facility managers with a valuable feature not included in many DR options–the ability to “opt out” or “override” a DR event if it comes at a time when the reduction in end-use services is not desirable.
A report titled “Findings from the 2004 Fully Automated Demand Response Tests in Large Facilities” (LBL-58178) is now available. Released in September 2005, it ) describes the results of the second season of research to develop and evaluate the performance of new Automated Demand Response (Auto-DR) hardware and software technology in large commercial facilities. The overall goal is to support increased penetration of DR through the use of automation, better understanding of DR technologies and customer-driven response strategies.
How Does Auto-DR Work?
Figure 1. Aggregated demand savings for five buildings, September 8, 2004.
Auto-DR provides a price or activation signal through an interface directly to the commercial facility’s EMCS. Two options for interfacing with an EMCS were identified for the first and second year’s tests: (1) a software-based Internet gateway and (2) a low cost hardware-based Internet relay.
Many retails and other multi-site organizations use Internet gateways to communicate with their buildings. Internet gateways use the Internet Communication Protocol (TCP/IP) to customize their connection to the EMCS. Unfortunately, not all EMCS provide Internet gateways and those that do don’t provide standardized interfaces. Consequently, the TCP/IP link for gateway-equipped EMCS has to be programmed by the customer for each system, which can often be a complex, costly process.
To overcome this problem, the DRRC provided test participants with a low cost Internet relay. The Internet relay provides the EMCS with a contact closure that can be activated remotely over a LAN, WAN or the Internet using standardized Internet protocols (IP). Internet relays can be used to communicate with any commercial building’s EMCS if the building has a connection to the Internet.
Figure 2 illustrates both the gateway and relay configurations used during the Auto-DR field trials.
Figure 2. Auto-DR communication configurations.
Results From Field Trials
In the tests, the DRRC price server depicted in Figure 2 acted as a proxy for a utility or ISO message unit. For the first two test-years, the DRRC price server used fictitious electricity price signals that mimicked a critical peak price or rate (CPP), to automatically activate customer programmed DR strategies. For the 2005 third-year test, the DRRC provided real prices for PG&E customers enrolled on an actual CPP rate.
During Auto-DR events, facilities are encouraged to shift, limit or shed loads to avoid high-cost peak period charges. Based on the rate incentives, facilities may also be encouraged to shift load to lower-cost off-peak time periods. Pre-cooling is one example of shifting loads.
Over three years of testing, the DRRC research team was successful in recruiting, configuring and testing Auto-DR strategies on more than 10 million square feet of facility floor space. During 2003, tests were conducted on five large commercial facilities. The 2003 tests were conducted in November, during mild weather. The test achieved a shed of nearly 10 % from the 5 MW demand under control among the five building. A late summer test achieved a maximum savings of nearly 1.5 MW, or about 24% of the total load for all five sites. No complaints were registered as a result of these large reductions.
In 2004 the test group expanded to 18 sites, including a total of 36 buildings. Maximum savings per site reached 1.8 watts per square foot with an average peak load reduction of 0.5 watts per square foot, which was equivalent to 7% of the average building load. Overall, the participating facilities’ Auto-DR strategies yielded peak demand reductions that ranged from a few percent to over 30% of total building load.
According to DRRC researchers, the largest savings were observed from strategies that used cooling zone set point increases, although lighting, anti-sweat heaters and other HVAC strategies also contributed.
After three years of successful tests, DRRC researchers observed that automating DR is likely to foster greater participation in various markets. Automation decreases the time needed to prepare for an event, increases the number of times a facility may be willing to shed loads, and perhaps improves the size of the DR response.
For More Information:
Mary Ann Piette, mapiette@lbl.gov
Dave Watson, dswatson@lbl.gov
Findings from the 2004 Fully Automated Demand Response Tests in Large Facilities Piette, M.A., D.S. Watson, N. Motegi, and N. Bourassa, Lawrence Berkeley National Laboratory. LBNL-58178. September 2005.
Development and Evaluation of Fully Automated Demand Response in Large Facilities Piette, M. A., O. Sezgen, D. Watson, N. Motegi, (Lawrence Berkeley National Laboratory), C. Shockman (Shockman Consulting), L. ten Hope (Program Manager, Energy Systems Integration CEC). CEC-500-2005-013. January 2005.
In the next issue: Auto-DR Case Studies: Strategies for Responding to Pricing and Emergency Events.
top