If a model is calibrated according to these limits, it is sufficiently close to the real response of the building. Obviously, this threshold can not guarantee that all input data entered are correct.
There are many other ways to calibrate a model, some of these follow manual procedures, other graphical, others are done by calculating some indexes, for more details refer to the sources [8][9].
Uncertainty and sensitivity analyses study how the uncertainties in the model inputs affect the model response. The uncertainty analysis aims to quantify the output variability and have a great impact on reliability of the model during the calibration process.
Generally, uncertainties create by different sources, the main are presented in the following table according to [11].
Category | Factors |
Scenario uncertainty | Outdoor weather condition Building usage/ occupancy schedule |
Building physical/operational uncertainties | Building envelope properties Internal gains HVAC system Operation and control setting |
Model inadequacy | Modeling assumptions Simplification in the model algorithm Ignored phenomena in the algorithm |
Observation error | Metered data accuracy |
There are different methods for uncertainty and sensitivity analysis which can be divided into two main approaches, external and internal method.
External method is divided in two different categories, local and global approach. The first category considered one at a time (OAT) method which is one input data changed while other inputs are keep constant. Then uncertainty in the parameter which kept fix can evaluate and study how this variation influence in the output.
Global sensitivity instead is based on changing more than one parameter simultaneously and study how variations of parameters impact on the uncertainty of whole space.
When energy models are correctly calibrated and validated, these could be used to evaluate and test strategies and solutions to improve the current condition.
Many lacks are derived from the State of the art related to the model calibration of historical buildings.
This research, through the investigation of the environmental indoor performance in historical buildings, is aimed to improve the conditions for the conservation of the cultural heritage and provide a better comfort for the visitors.
For this purpose, it is necessary to evaluate different intervention solutions, whose effects can be foreseen through the creation of simulation models.
The data simulated by these models, however, at an initial stage are far from the real response of the building, due to the numerous causes of uncertainty that are inherent in the creation and insertion of the inputs of the model itself. In order to implement a model with a fair degree of accuracy, a procedure called "calibration" is required.
The intent is to provide guideline in order to reduce the gap between simulated and measured data in historical buildings. When is not present HVAC systems, compare the energy consumption is not possible. In these cases, the model must be calibrated using other available data. Among these some may be the hygrothermal data acquired through environmental monitoring campaigns, using temperature and relative humidity sensors.
From the SoA many lacks are founded about the calibration process in historical buildings. In the following paragraph is described a methodology to use for this specific scope and during the next stages of the research many other unresolved aspects will be addressed.
In order to perform an accurate calibration of historical buildings without HVAC system, measured temperature and absolute air humidity can be used to make comparison with the simulated data. The temperature is the main parameter to be considered, while the absolute humidity could be calculated and represent a counter-test if the calibration has been carried out correctly, furthermore, can give information about the correct estimate of the air changes input. The surface temperature data can be also taken into consideration, if sensors have been positioned on the walls; this data may be an additional parameter to compare in the calibration process, especially when analyzing in detail specific parts of the building.
In such respect, methods to define which parameters affects greatly the results of the simulations as well as metrics to carry out the validation model should be defined. In the following paragraph a brief description of part related to Calibration process has been outlined