New Toolkit for Outdoor Lighting Baseline Updates

The primary goal of this study is to develop a new toolkit for conducting an outdoor lighting baseline study for California. This project and subsequent baseline study is crucial for updating the assumptions used in energy efficiency program design, codes and standards development, and for validating baseline energy load profiles in eTRM. Given the significant advancements in outdoor lighting technology since the last baseline study in 2003, an update to the information used to establish nonresidential outdoor lighting energy consumption is both necessary and warranted.The project team developed a framework to categorize different types of data pertaining to outdoor lighting into static and dynamic attributes and different data collection methods into high fidelity and low fidelity methods. The team implemented a hybrid approach using both high and low fidelity methods to balance accuracy and resource constraints. This involved developing a sample of sites, applying both high and low fidelity methods to a sub-sample, and using inference models to enhance the data for sites collected only using the low fidelity methodsThe project team conducted a comprehensive review of promising high fidelity and low fidelity data collection methods. The integration of both methods supports the development of an inference model to interpolate across gaps in low fidelity data. Using the inference model will enable the project team to combine the precision of high fidelity data and the scalability of low fidelity data by helping train predictive relationships between data collected using low fidelity methods and true values confirmed via high fidelity methods. This allowed the project team to estimate key static and dynamic attributes for a much larger number of sites by applying calibration adjustments to data collected using low fidelity methods. Based on the analysis, stakeholder engagement, and evaluation of various data collection approaches, the project team offers the following actionable conclusions and recommendations:1. Future data collection efforts should incorporate greater use of low fidelity data collection methods because they demonstrate potential for scalability and ease of implementation, and can enable rapid, cost-effective data acquisition, particularly for static attributes.2. The inference model proposed in this study should be used as a core component of future baseline efforts. The model can be trained to calibrate or predict key lighting attributes across the full dataset by integrating data from high fidelity and low fidelity collection methods. This approach will help reduce overall costs and labor requirements while preserving accuracy.3. Conduct a focused follow-up project to pilot the shortlisted high fidelity and low fidelity data collection methods in real-world conditions to assess each method’s viability, cost, and accuracy.