Energy performance research

In 2009 Newsham et al. analyzed a database of 100 LEED certified (v3 or earlier version) buildings. In
this study, each building was paired with a conventional "twin" building within the Commercial Building
Energy Consumption Survey (CBECS) database according to building type and occupancy. On average,
LEED buildings consumed 18 to 39% less energy per floor area than their conventional "twin" buildings,
although 28 to 35% of LEED-certified buildings used more energy than their "twin.”[37] The paper found
no correlation between the number of energy points achieved or LEED certification level and measured
building performance.

In 2009 Scofield published an article in response to Newsham et al., analyzing the same database of
LEED buildings and arriving at different conclusions. In his analysis, Scofield criticized that Newsham et
al.'s study only considered the energy per floor area instead of a total energy consumption. Scofield
considered source energy (accounting for energy losses during generation and transmission) as well as
site energy, and used area-weighted energy use intensities, or EUIs (energy per unit area per year), when
comparing LEED and non-LEED buildings to account for the fact that larger buildings tend to have larger
EUIs. Scofield concluded that, collectively, the LEED-certified buildings showed no significant source
energy consumption savings or greenhouse gas emission reductions when compared to non-LEED
buildings, although they did consume 10-17% less site energy.

Scofield in 2013 analyzed 21 LEED-certified buildings in New York City. He found that buildings that
had achieved LEED Gold used, on average, 20% less source energy than did conventional buildings.
Buildings with LEED Silver or LEED Certified ratings actually used 11 to 15% more source energy, on
average, than did their conventional counterparts.

In 2014, Fuertes and Schiavon developed the first study that analyzes plug loads using LEED documented
data from certified projects. The study compared plug load assumptions made by 92 energy modeling
practitioners against ASHRAE and Title 24 requirements, as well as the evaluation of the plug loads
calculation methodology used by 660 LEED-CI and 429 LEED-NC certified projects. In general, energy
modelers considered the energy consumption of plug loads of equipment that are constantly running (such
as refrigerators) as well as monitors and computers predictable. Overall the results suggested a
disconnection between energy modelers assumptions and the actual performance of buildings. In
conclusion, the study suggests LEED or ASHRAE to develop guidelines for plug loads calculations.

Energy model might be a source of error during LEED design phase. Stoppel and Leite evaluated the
predicted and actual energy consumption of two twin buildings using the energy model during the LEED
design phase and the utility meter data after one year of occupancy. The study’s results suggest that
mechanical systems turnover and occupancy assumptions significantly differing from predicted to actual
values.

Most of the current available energy consumption analysis of LEED buildings focuses on LEED v3
(2009) or even earlier versions certified buildings instead of the newest LEED v4 (2014) certified
buildings. According to Newsham et al., these analyses should be considered as preliminary and should
be repeated with longer data history and larger sample buildings, including new LEED v4 certified
buildings. Newsham et al. also pointed out that further work needs to be done to define green building
rating schemes to ensure more consistent and substantial energy consumption reduction success at the
individual building level in long term.