ZEMCH 2015 - International Conference Proceedings | Page 84

Uncertainty/ Sensitivity analysis
Statistical method of “Monte Carlo” is used to generate probabilistic output. In this method, for
the desired parameters, values from their probability distribution are randomly selected and a
simulation is undertaken. Simulations are repeated in a large number of evaluations with new
values randomly selected (I. Macdonald & Strachan 2001). Sample size of simulation in this study
was taken 100 as recommended by Macdonald (I. A. Macdonald 2009:992–999) for building simulation. In this study, by generating an Excel calculation sheet which operate automatically by
utilization of Visual Basic Application (VBA) codes, all inserted input are applied in calculation of
corresponding global cost and VBA macro enables repetition of calculations for 100 times while
varying input parameters within their given variation range. Distribution of all simulated results
is studied and their uncertainty range is quantified. In the second step sensitivity analysis is performed and through a regression analysis, coefficient of determination (R2) is considered as the
indicator for variance of obtained values. In this way importance of individual input parameters
are compared to determine the most and least influential parameters.

Result
By inserting primary energy needs of 22 cases in the calculation sheet which is created according
to EN 15459, PVs are achieved. Figure.2 indicates primary energy consumption and global cost of
22 studied alternatives. As seen in the scatter plot, wall No.22 with U-value compatible with Lgs.
311/2006, represents the highest energy needs and lowest global cost while wall No.21 with the
minimum transmittance among alternatives (U-value=0.134 w/m2K) does not lead to a reasonable
global cost. Among all simulated cases, wall No.4 and No.5, as highlighted in the graph, are the
most cost-efficient with primary energy consumption of 37.5 and 40 kwh/m2yr, respectively.

Table 4: Rate values and variation interval of input parameters
Base case Value

Range
of Variation

Market interest rate (R)

4.25%

± 15 %

Inflation rate(Ri)

1.90%

±5%

Rate of development of energy cost (Re)

1.95%

± 25 %

Rate of development of price for envelope and technical system (Rp)

-

± 10 %

Rate of development of price for maintenance

-

±5%

Rate of development of price for added cost

-

±5%

Variable Parameter

82

ZEMCH 2015 | International Conference