Civil Engineering Reference
In-Depth Information
the building can have a considerable deviation from the selected and used material
characteristics in original energy modeling during the design stage.
In 2010, Dominguez-Munoz, Cejudo-Lopez and Carrillo in their paper
“Uncertainty in peak cooling load calculations” (Dominguez-Munoz et al. 2010 )
represented the effects of multiple (more than 25) input components on the overall
cooling system output. They showed that the peak cooling load uncertainty due to
effects of the chosen input elements could be as high as ±15 %. It should be noted
here that the possibility of occurrence of the peak cooling load below 4 % and above
96 % of the mean value was shown to be extremely small and almost negligible.
Prior to that in 2002, McDonald in his PhD thesis “Quantifying the effects of
uncertainty in building simulation” (McDonald 2002 ) showed for a specifi c offi ce
building and based on the defi ned uncertainties (20 %) in input parameters such as
conductivity of external wall insulation, and conductivity of external wall brick face
the calculated energy consumption standard deviation was ±7.77 kW h.
In 2012, in my PhD thesis “The effects of sub-optimal components in cooling
systems energy consumption and effi ciency” (Khazaii 2012 ) I showed the envelope
elements of the building are not the sole sources of uncertainty in energy consump-
tion of the building and other parameters such as testing agencies test tolerances
allowance for testing the HVAC equipment have considerable impact on the output
probability distribution of the energy consumption of the whole building. Of course
there are other uncertain elements that could be added to this process, such as light-
ing power (watt) tolerance allowance, which was missing from my research, and
could make the results even more considerable.
These researches show the importance and need of new approach in energy mod-
eling different from the current industry-wide adopted method, and introduction of
new revisions to the existing software for building energy modeling simulation to
make it capable of performing a probabilistic energy modeling instead of current
deterministic method. The above mentioned researches also show the need for
emphasizing on the fact that the commercial software writers should get themselves
familiar with the relevant concepts such as uncertainty and sensitivity and imple-
ment the required provisions in their software.
13.2
Implementing Uncertainty and Sensitivity Analysis
in Energy Modeling
Uncertainty analysis and sensitivity analysis are two main topics in branch of sci-
ence known as statistics. Unfortunately up to a few years ago these powerful tools
did not fi nd their ways into the building architecture and HVAC engineering world.
As it was stated earlier in the past few decades there were some attempts on using
these tools to disclose the effects of uncertain envelope elements on building energy
consumption, but I believe very soon these concepts will be developed further and
will be used more universally in building energy modeling and design.
Search WWH ::




Custom Search