Development, Calibration and Validation of an Internal Air Temperature Model for a Naturally Ventilated Nearly Zero Energy Building: Comparison of Model Types and Calibration Methods

Abstract

In this study, a grey box (GB) model for simulating internal air temperatures in a naturally ventilated nearly zero energy building (nZEB) was developed and calibrated, using multiple data configurations for model parameter selection and an automatic calibration algorithm. The GB model was compared to a white box (WB) model for the same application using identical calibration and validation datasets. Calibrating the GB model using only one week of data produced very accurate results for the calibration periods but led to inconsistent and typically inaccurate results for the validation periods (root mean squared error (RMSE) in validation periods was 229% larger than the RMSE in calibration periods). Using three weeks of data from varying seasons for calibration reduced the model accuracy in the calibration period but substantially increased the model accuracy and generalisation abilities for the validation period, reducing the mean RMSE by over 160%. The use of one week of data increased the standard deviation in parameter selections by over 40% when compared with the three-week calibration datasets. Utilising data from multiple seasons for calibration purposes was found to substantially improve generalisation abilities. When compared to the WB model, the GB model produced slightly less accurate results (mean RMSE of the GB model was 1.5% higher). However, the authors found that employing GB modelling with an automatic model calibration technique reduced the human labour input for simulating internal air temperature of a naturally ventilated nZEB by approximately 90%, relative to WB modelling using a manually calibrated approach.

Assessing the Risk to Indoor Thermal Environments on Industrial Sites Offering AHU Capacity for Demand Response

“…A systematic modelling approach outlined in Figure 2, using a methodology based on existing published studies [26][27][28]33] was adopted to capture the complexity of this application. The main aim of this approach is to adequately reproduce, and subsequently predict, the internal air temperature ramps within the indoor occupied space being modelled during shut down of the AHU, based on a number of defined input variables.…”

“…The model was calibrated on one day of actual measured conditions and validated across the remaining 8 days of measured values using the C1V8 process [26], thereby allowing the model to be trained effectively on empirical conditions whilst minimising bias and reducing the risk of over fitting. As the AHU was run on a consistent daily schedule with minimal operational variations, it was found that additional validation days were more valuable than calibration days to gain insights and build confidence in the accuracy of the outputs.…”

“…Effectively capturing AHU and building thermal response dynamics remains an integral challenge to overcome when modelling buildings as an element in DSM flexibility programmes. Considerable research has also been conducted to date into modelling the internal temperature conditions of naturally ventilated office spaces [26,27]. A variety of modelling concepts have been employed in this area, with Grey Box (GB) models [26] and Resistance Capacitance (RC) techniques [28] being the predominant methodologies.…”

“…Considerable research has also been conducted to date into modelling the internal temperature conditions of naturally ventilated office spaces [26,27]. A variety of modelling concepts have been employed in this area, with Grey Box (GB) models [26] and Resistance Capacitance (RC) techniques [28] being the predominant methodologies. In modelling terms, White Box (WB) models are physics-based and mechanistic in operation, typically utilising a large number of building descriptive parameters and system specifications, whereas black box models are generally data-driven and use complex algorithms to map the relationship between system inputs and outputs using large amounts of empirical data.…”

“…GB models are closer to WB models in that they are mechanistic by design, however the iterative physics-based operations are simplified and aggregated. Thus, GB models are less computationally demanding and require fewer building descriptive parameters [26]. RC models are based on the shared relationship between electrical circuits and thermal systems, where the thermal energy flow and storage in a building can be represented by resistors (R) and capacitors (C), similarly to those in an electrical circuit [28].…”