تجزیه و تحلیل حساسیت در مورد سطح نور در روز و عملکرد انرژی ادارات پیرامون با استفاده از سایه های خودکار

This paper presents a comprehensive global uncertainty and sensitivity analysis of daylighting and energy performance for private offices with automated interior roller shades using an advanced integrated thermal and lighting simulation model. The purpose was to identify the more important factors with respect to building thermal and lighting energy performance so as to facilitate decision making in building design stage and simplify further investigation such as optimization analysis. Seven studied parameters were selected: window-to-floor ratio, shading transmittance, shading front and back reflectance, space aspect ratio, insulation thermal resistance and glazing type. The performance metrics include useful daylight illuminance (500–2000 lux), annual lighting, heating and cooling demand per unit floor area and annual source energy consumption per unit floor area. The uncertainty analysis is based on the Monte Carlo method with Latin Hypercube Sampling, showing the possible ranges in these performance indices. The sensitivity analysis uses a variance-based method in the extended FAST implementation. Application of the analysis to perimeter private office spaces for the climate of Philadelphia showed the first order and total order effects of each studied parameter to determine the building parameters that have the most significant impact. Results are presented for different facade orientations.

Commercial buildings, primarily office buildings, consume a large amount of energy and present a rapid increase in total primary energy consumption. The situation of high energy requirements on one hand and limited energy resources on the other hand has sparked a lot of research activities concentrating on early stage building design as well as on retrofit efforts for energy saving purposes. Conventionally, previous studies used approaches based on evaluation of several alternative design options to identify the best solution [22] or analysis of influence coefficients in terms of a base case to determine the important design parameters [17] and [38]. In the design procedure, it is beneficial to identify the importance of design parameters correctly in order to efficiently develop design options or reach optimal design solutions. Recently, more advanced sensitivity analysis approaches have been employed in determination of the most important parameters in relation to building performance. Heiselberg et al. [12] used the elementary effects method to investigate which design parameters are the most important among the 21 selected factors to change in order to reduce the primary energy consumption. Their results showed that lighting control is one of the two most important parameters that will have the most significant effect. Mechri et al. [21] employed the Monte Carlo method with Latin Hypercube Sampling (MC-LHS) and the Analysis Of Variance-Fourier Amplitude Sensitivity Test method (ANOVA-FAST) for uncertainty and sensitivity analysis of heating and cooling energy needs. The first order effect of each studied factor was calculated and the envelope transparent surface ratio was distinguished as the most significant factor for both heating and cooling energy needs. In some studies [6], [43], [15] and [44], the MC-LHS method was also used to calculate the sensitivity indices such as the Standardized Regression Coefficient (SRC) and the Standardized Rank Regression Coefficient (SRRC) given that the model coefficient of determination is higher than 0.7 [32]. In the above mentioned studies, windows have gained enough importance as an influential envelope element. Usually one or more factors related to windows (including transparent surface ratio, U-value and solar heat gain coefficient) were considered in the analysis. However, those thermal parameters were considered only for the concern of heating/cooling load or energy consumption. A series of studies have focused on the glazing optical properties and the resulting effects on lighting, heating and cooling needs. It is well known that utilization of daylight in perimeter office spaces introduces opportunities for energy savings [18], [7] and [29]. The daylighting performance is affected by many interfacing factors such as glazing size and properties, shading properties and control [35], [25], [39] and [3], room aspect ratio and orientation [8]. These factors also affect the thermal loads and hence the energy performance of space. In order to improve the overall performance via most effective approach in design, a sensitivity analysis should be applied to an integrated building model, especially when dynamic control of lighting systems or shading devices is in use. Furthermore, the different window properties are often correlated and therefore require great carefulness in defining the probability density functions and respective ranges. For example, a window with a high solar heat gain coefficient might also have a high U-value. Such properties are not completely independent and the combination between them is not really random. This paper presents a comprehensive global uncertainty and sensitivity analysis of daylighting and energy performance for private offices with automated interior roller shades using an advanced transient simulation model. Studied performance indices include useful daylight illuminance, annual lighting, heating and cooling demand and annual source energy consumption. The uncertainty analysis is based on MC-LHS method showing the possible ranges in these performance indices. The sensitivity analysis uses a variance-based method in the extended FAST implementation. First order and total order effects of each studied parameter were calculated to determine the building parameters that have the most significant impact on the performance indices.

This paper presented a sensitivity analysis of five building performance metrics to seven selected design parameters. The purpose was to identify the more important factors with respect to building thermal and lighting energy performance so as to facilitate decision making in building design stage and simplify further study such as optimization analysis. An integrated transient daylighting and energy model (with annual runs) developed in a previous study with two improvements was employed to simulate the building performance. A perimeter private office in Philadelphia was used as a case study. The studied building performances included UDI (500–2000 lux), lighting, heating and cooling demand and source energy consumption. Considering the possible influential factors, seven design parameters are selected: window-to-floor ratio, shading transmittance, shading front and back reflectance, space aspect ratio, insulation thermal resistance and glazing type. Glazing properties, such as visible transmittance, solar transmittance and U-value, are not continuously available from market and more importantly, they are not independent with each other, therefore they were grouped as one factor – glazing type- in the analysis. An uncertainty analysis was performed first to evaluate the necessity for further sensitivity analysis. The analysis is based on MC-LHS method with 140 runs. Significant dispersion of the building performance is evaluated, indicating that the studied factors should be carefully designed to achieve certain design targets. A sensitivity analysis was then performed: based on the characteristics of the building model, a variance-based method was selected. Using the extended FAST module in SimLab 2.2, a sample set with length of 460 data was generated and the first and total order sensitivity indices were computed for each studied design factor. The following conclusions are drawn from the results: (i) The uncertainty analysis shows high dispersions of all investigated performance metrics (13–68%), indicating that decisions should be made very carefully in the design stage to ensure that the various building performance indices remain in the preferred range. (ii) For all evaluated performance metrics, window-to-floor ratio and glazing type have a significant impact. (iii) For useful daylight illuminance, window-to-floor ratio has a great influence with first order sensitivity index varies from 0.38 to 0.54 and total order sensitivity index varies from 0.75 to 0.91. The impact of glazing type is usually involved with other design parameters, showing low first order sensitivity index around 0.02 but high total sensitivity index varies from 0.17 to 0.53. For annual lighting demand, the impact is similar to that of useful daylight illuminance. (iv) For annual heating demand, window-to-floor ratio makes a contribution varying from 0.2 to 0.39 (total indices vary from 0.35 to 0.56), and window type similarly varying from 0.27 to 0.41 (total indices varies from 0.42 to 0.55). (v) For annual cooling demand, window-to-floor ratio shows high first order impact ranging from 0.39 to 0.53 with total impact around 0.7. The glazing type also shows significant impact with first order sensitivity index (0.31 to 0.44), and total sensitivity index around 0.55. (vi) When it comes to annual source energy consumption, different orientations have their own slightly different import factors. For south facades, the most important factor is the shade transmittance (first order index: 0.34, total order index: 0.47), followed by glazing type (first order index: 0.08, total order index: 0.29) and window-to-floor ratio (first order index: 0.06, total index: 0.25). West facades have the same import factors in different order: glazing type (first order index: 0.24, total order index: 0.6), window-to-floor ratio (first order index: 0.09, total order index: 0.32) and shade transmittance (first order index: 0.18, total order index: 0.16). North and east facades have different important factors in the order of glazing type, window-to-floor ratio and front side shade absorptance. (vii) For different orientations, the rank of factor importance changes slightly because of different outside conditions such as incident solar radiation and the resulting different shading schedules and lighting operation. The value of sensitivity indices strongly depends on the range of inputs. When the range changes, both the indices’ value and importance rank may change. This paper was focused on a typical perimeter private office with one external facade. The interior roller shade and lighting system were controlled with certain algorithms, so the results may not applicable to other spaces with different conditions. Finally, some of the studied parameters will also have a significant impact on thermal and visual comfort, which should also be investigated in a future study.