Language:
    • Available Formats
    • Options
    • Availability
    • Priced From ( in USD )
 

About This Item

 

Full Description

Continuous maintenance of large fleets of building energy models (BEMs) is a new challenge that may soon approach economic parity for large institutions as data and modeling connectivity become streamlined. In the past, BEMs have mostly been used for analyses during design with little or no reuse of the model. The Energy Independence Security Act (EISA) of 2007’s requirement to complete energy and water evaluations for federal facilities is changing this. EISA requirements can be met through performance of ASHRAE energy audits that allow BEMs to be used for identifying energy savings opportunities and energy use breakdowns. Several years ago, Sandia National Laboratories (SNL) developed a fleet of 121 BEMs for site-wide energy assessments. Applications for this fleet has now been expanded to EISA compliance. The authors propose maintaining the BEM fleet on a 4-year cycle. In this process models undergo a quality check (QC) and recalibration whenever a building energy audit is performed for its corresponding building. For recalibration, auto-calibration technology is being used. This paper outlines the first year of efforts to construct a streamlined process and results for the first 5 models. The first BEM underwent both manual and auto-calibration for a direct comparison. Manual calibration incurred more cost and required more time from staff members. Auto-calibration required significantly less effort, slightly less cost, and resulted in slightly better accuracy. ASHRAE Guideline 14-2014 was met by 3 of the 5 buildings after auto-calibration. Even so, significant improvements to Normalized Mean Bias Error (NMBE) and Coefficient of Variation for Root Mean Square Error (CV(RSME)) were achieved for all five buildings. Data and model quality issues are suspected as causes for the non-compliance rather than inadequacy of the auto-calibration procedure. Many improvements to the processes used to prepare data and models have been identified including issues that require major changes to SNL’s energy tracking infrastructure.