Lead Performers:
-- National Renewable Energy Lab – Golden, CO
-- Lawrence Berkeley National Lab – Berkeley, CA
DOE Funding: $200,000 in FY15; $200,000 overall
Cost Share: $0
Project Term: 2015 – 2016
Funding Opportunity: Emerging Technologies Lab RFP (Request For Proposals)

PROJECT OBJECTIVE

Building energy modeling (BEM)—physics-based calculation of building energy consumption—is a multiuse tool for building energy efficiency with a variety of use cases from design to code compliance to commissioning and control. DOE’s overarching goal is to increase the effective use of advanced BEM in design while helping to establish its use in building operations. DOE’s 2020 and 2030 goals for BEM use are listed in its draft multiyear project plan (MYPP).

One of the challenges cited in the MYPP is “insufficient characterization of BEM engine accuracy” explaining that “error or degree of approximation associated with various aspects of BEM and various use cases of BEM is not well-characterized or well-attributed. This lack of clarity reduces confidence in BEM as a whole and suppresses the use of BEM in investment and decision-making.” Deviations of BEM calculations from observed data are due as much, if not more, to external errors (i.e., deviations in the inputs to the model) as they are to internal errors (i.e., inappropriate simplifications or implementation bugs in the programs themselves). Acknowledging this, DOE is nonetheless devoting significant near-term attention and resources to the issue of internal errors. This choice is made for practical reasons—the BEM accuracy question is easier to answer when decomposed and internal accuracy is easier to tackle first, especially given the availability of new test facilities at National Laboratories Lawrence Berkeley (LBNL) and Oak Ridge (ORNL). It also serves a psychological purpose—showing that BEM methods and tools are sound will increase confidence in BEM.

ANSI/ASHRAE Standard 140 “Standard Method of Test for the Evaluation of Building Energy Analysis Computer Programs” provides an extensible framework for testing BEM engines and diagnosing internal errors. The Standard 140 framework accommodates empirical tests. However, because of the historical difficulty and cost of obtaining validation-grade empirical data, Standard 140 currently includes only analytical and comparative tests which compare simulated results to analytical solutions of simple configurations and to other simulated results, respectively. A significant amount of testing can be and has been done this way, but the addition of empirical tests will strengthen the entire edifice.

DOE has begun a three-year effort to use the LBNL and ORNL facilities to conduct experiments that can be used as the bases for Standard 140 empirical tests. The tests will be used to characterize the accuracy of the major calculations in modern BEM engines. DOE plans to use these tests to improve the accuracy of its own open-source BEM engine EnergyPlus™ where necessary and expects that other engine vendors will choose to do the same.

To help scope and guide the empirical validation effort, DOE commissioned a roadmap. The purposes of the roadmap are:

  • To identify gaps in the empirical validation of BEM programs, particularly in measured data sets.
  • To explore how the new LBNL and ORNL facilities, and others that may exist both nationally and internationally, could be used to fill these gaps.
  • To define criteria for high-quality empirical validation experiments.

DOE sponsored a workshop on January 28 and 29, 2015. The workshop was attended by 27 experts, including BEM software developers, advanced users, design and engineering practitioners, National Laboratory scientists, and university educators. Participants were asked to write a one-page statement describing their concerns about accuracy for specific uses cases and calculations of BEM as well as for their definition of “success” or “sufficient accuracy." Participants discussed their responses and ideas in the large group as well as in three breakout groups that focused on building envelope, HVAC systems, and integrated whole-building systems, respectively. The roadmap synthesizes the results of the workshop and follow-on discussions.

PROJECT IMPACT

The ability to make definite statements about the accuracy of various aspects of BEM will have three positive impacts:

  • It will improve the reputation of BEM among energy-efficiency decision-makers (e.g., building owners, ESCOs) who will have greater confidence in both contracting with BEM services and acting on BEM-driven recommendations.
  • It will identify aspects of BEM that need additional research and development.
  • By elimination, it will bound the effects of various BEM inputs and identify inputs that require and deserve additional effort to obtain, benchmark, or characterize via analysis.

CONTACTS

DOE Technology Manager: Amir Roth
Lead Performer: Ron Judkoff, NREL; Philip Haves, LBNL; Steve Selkowitz, LBNL

Download the MATERIALS