Enhanced Detection and Diagnostic Capabilities of Data Mining Methods to Identify Energy Efficiency Enhanced Detection and Diagnostic Capabilities of Data Mining Methods to Identify Energy Efficiency The Pacific Northwest National Laboratory (PNNL), under contract to the Washington State Attorney Generals Office (AGO), requires technical support to the project titled, Data Mining Methods to Identify Energy Efficiency Opportunities in Small-/Medium- Sized Commercial Buildings. The proposed scope of work described below is to support PNNL in developing data mining methods. The intent of this research is to demonstrate the enhanced detection and diagnostic capabilities of data mining techniques when applied to whole building interval data in terms of being able to identify actionable measures that will reduce the energy use and cost in the building. In essence, intelligent filters are to be developed, which are anchored on physics/engineering-based understanding of the behavior of building operation and their heating, ventilation and air conditioning (HVAC) systems. This is a proof-of-concept study to be demonstrated on a relatively small number of small- and medium-sized commercial buildings that have no building automation system and have a limited range/type of HVAC systems. Background Several electric utilities, as part of the Smart Grid Advanced Metering Infrastructure (AMI) have launched a multi-billion dollar program to install smart meters in several million residential and commercial customers. AMI allows customers to have almost real-time access to customers electric use data. The data would typically be in the form of whole building interval data (15 min aggregates for commercial customers) similar to the type of data already implemented for commercial customers with installed load greater than 200 kW. The question now is What can the customer do with this data? Small- to medium-sized buildings account for over 40% of the total floor space. Most commercial buildings are operated inefficiently; among the reasons are improper operational practices, lights and HVAC equipment left on in unused spaces, and spaces are needlessly cooled or heated during unoccupied periods. This is due, in part, to lack of quantitative information by the building manager or operator as to the buildings expected performance and the benefits and costs of cost-saving strategies specific to his or her facility. Data mining techniques, on the other hand, have proven to provide extremely valuable insights into general hidden trends and patterns in customer energy use behavior and to identify ways to incentivize customers to reduce their energy use. While the former techniques have been primarily applied to individual buildings, the latter have been almost exclusively used population of customer data, mostly residential customers.
|Effective start/end date||3/26/13 → 3/31/14|
- Pacific Northwest National Laboratory (PNNL): $50,000.00
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.