An energy benchmarking model based on artificial neural network method with a case example for tropical climates

Energy benchmarking is an important step in evaluating a buildings energy use and comparing it with similar buildings in similar climates. Depending on the benchmarking results, extra measures can be taken to reduce energy consumption when the subject building has been assessed to consume more than other similar buildings. This study presents the current state of energy benchmarking‐related research and available tools. An artificial neural networks (ANN)‐based benchmarking technique is presented as a highly effective method. The model specifically focuses on predicting a weighted energy use index (EUI) by taking into consideration various building variables, such as plug load density, lighting type and hours of operation, air conditioning equipment type and efficiency, etc. Data collected from laboratory, office and classroom‐type buildings and mixed use buildings in Hawaii are used to present the ANN‐based benchmarking technique. The developed model successfully predicted the benchmarking EUI for the buildings considered in the study. The model coefficient of correlation was 0.86 for the whole building benchmarking analysis, indicating a good correlation between the measured EUI and the ANN predictions. Additionally, the use of ANN benchmark model for predicting potential energy savings from retrofit projects was evaluated. Some of the benchmarking input variables were modified to reflect a potential energy savings from a retrofit project and the new input set was simulated with the ANN model. The preliminary results show that the developed ANN model can be used to predict energy savings from retrofit projects. Copyright © 2006 John Wiley & Sons, Ltd.     

An energy benchmarking model based on artificial neural network method utilizing US Commercial Buildings Energy Consumption Survey (CBECS) database

This study focuses on development of an energy benchmarking model utilizing U.S. Commercial Buildings Energy Consumption Survey (CBECS) Database. An artificial neural networks (ANN) method based approach was used in the study. Office type buildings in the CBECS database were used in the benchmarking model development and weighted energy use intensity (EUI) was selected as the benchmarking index. The benchmarking model included input variables describing building's physical properties, occupancy and climate. Yearly electricity consumption per square meter, or EUI, was estimated by the ANN model. The correlation coefficient for each census division benchmarking model varied between 0.45 and 0.73, and mean squared error (MSE) varied between 9.60 and 15.25. It was observed that when the data set for a census division was grouped by different climate zones, ANN benchmarking model provided more accurate predictions. It was also observed that ANN model provides more accurate estimations when compared with predictions obtained with multi-linear regression models. For comparison, the MSE values varied between 10.24 and 40.43. Overall, the ANN model proved itself a better prediction model for energy benchmarking. Copyright © 2006 John Wiley & Sons, Ltd.     

Assessing performance and controlling operating conditions of a solar driven absorption chiller using simplified numerical models

In cooling buildings, the use of solar energy can save around 50% of primary energy. Many studies have demonstrated the interest of such systems. However, developing and improving reliability of new components, design, control, and implementation remain a major concern. The performances of solar cooling systems are greatly influenced by climatic conditions. Indeed they affect both the driving energy of the chiller and the heat rejection. It is important to mention that internal loads and control strategy also have an impact on energy performances. Therefore, assessing the energy performance during the design phase is a key point in evaluating the economic interest of an installation. Moreover, once the commissioning of the installation is accomplished, there is a need to follow through and ensure its performance, since a large number of malfunctions can affect the quality of the system. Actual performances can be very different from those calculated in the design phase.With this aim, the present article deals with the development of an absorption chiller model used in an existing solar cooling system. This installation includes a single effect absorption chiller with a nominal chilling capacity of 30 kW (EAW LB30 chiller functioning with water and lithium bromide), and it cools four classrooms of a University building in Reunion Island which is situated under a tropical climate. This pilot plant is very good monitored and can thus be used to develop and validate the absorption chiller model. The present paper first recalls the absorption principle and presents the pilot plant, the metrology, and the control strategy. Secondly, the experimental results are analysed and the steady state chiller model and also the identification method are developed. Thereafter, the simplex method is used to determine the design parameters of the machine. Finally, the simulation results are presented. The good agreement between the prediction and the experimental results allows the use of the model not only to design an installation but also to follow and control its performances.     

Modelling a solar-assisted air-conditioning system installed in CIESOL building using an artificial neural network

This paper proposes Artificial Neural Networks (ANN) to model a solar-assisted air-conditioning system installed in the Solar Energy Research Center (CIESOL). This system consists mainly of the single-effect LiBr-H₂0 absorption chiller fed by water provided from either solar collectors or hot water storage tanks. The present work describes the total solar cooling systems based on absorption chiller and provided only with solar collectors. The experimental data were collected during the cooling period of 2008. ANN was used with the main goal of predicting the efficiency of the chiller and global system using the lowest number of input variables. The configuration 7-8-4 (7 inputs, 8 hidden and 4 output neurons) was found to be the optimal topology. The results demonstrate the accuracy ANN’s predictions with a Root Mean Square Error (RMSE) of less than 1.9% and practically null deviation, which can be considered very satisfactory.     

Assessment of climate change impact on building energy use and mitigation measures in subtropical climates

Likely increase in energy use in air-conditioned office buildings due to climate change in subtropical Hong Kong was estimated for two emissions scenarios. Towards the end of the 21st century (i.e. 2091-2100), the average annual building energy use would be 6.6% and 8.1% more than that in 1979-2008 for low and medium forcing, respectively. Potential mitigation measures concerning the building envelope, internal condition, lighting load density (LLD) and chiller plant were considered. Thermal insulation to the external wall would not be effective to mitigate the expected increase in building energy use due to climate change. Controlling the amount of solar heat gain through the window would be a better option. Lowering the current LLD of 15 W/m² to about 13 W/m² would result in substantial energy savings because of the reduction in electricity consumption for both electric lighting and air-conditioning. As for the chiller plant, the coefficient of performance (COP) should be improved from the current minimum requirement of 4.7 to at least 5.5 to alleviate the impact of climate change. Raising the summer set point temperature (SST) to 25.5 °C or higher would have high energy saving and hence mitigation potential, which could be readily applied to both new and existing buildings.     

Estimating operating cell temperature of BIPV modules in Thailand

Several models have been developed to estimate the operating cell temperatures of photovoltaic (PV) modules because they directly affect the performance of each PV module. In this study, two prediction models used most commonly, the nominal operating cell temperature (NOCT) model and the Sandia National Laboratory temperature prediction model (SNL), were investigated for their suitability in the prediction of PV module's temperatures for building integrated photovoltaic (BIPV) installation in the tropical climate conditions of Thailand. It was found that, in general, the SNL model tends to give better results of temperature prediction than those of the NOCT model. Nevertheless, both models are strongly over-biased in temperature predictions. The discrepancies of the predictions are basically caused by the dissimilarity of the BIPV installation and the standard installation as specified by the models, rather than the effect of differences in climatic conditions between the temperate and tropical zones. In the worst case, it was found that the highest value of the mean bias error (MBE) is +8 °C, or equivalent to +21% of the mean observed temperature, and the root mean square error (RMSE) is ±10 °C, or equivalent to ±24% of the mean observed temperature. However, although these errors were large, their effects on the accuracy of the final prediction of the electrical power output generated by the PV module over a long term would not be great. The error of the expected generated energy output would not be more than 6% of the averaged actual energy output, which is acceptable for most applications.     

An analysis of climatic influences on chiller plant electricity consumption

Principal component analysis of dry-bulb temperature, wet-bulb temperature, global solar radiation, clearness index and wind speed was conducted, and a two-component solution obtained which could explain 80% of the variance in the original weather data. Clustering analysis of these two principal components resulted in a total of 18 typical day types being identified. A year long monitoring of the daily chiller plant electricity consumption in a fully air-conditioned office building was conducted. It was found that the typical day types exhibited daily and seasonal variations similar to the daily and monthly electricity consumption recorded. Three regression models were developed to correlate the daily chiller plant electricity consumption and the corresponding day types. The coefficient of determination (R²) was 0.86–0.99 showing strong correlation. It is proposed that the day type approach can be used as a tool for weather normalisation and inter-year comparisons in the analysis of energy savings due to building retrofits. It was also found that the typical day types identified appeared to show a slight increasing trend during the 28-year period (1979–2006) indicating a subtle, but gradual change of climatic conditions that might affect chiller plant electricity consumption in future years.     

Development of surrogate models using artificial neural network for building shell energy labelling

Surrogate models are an important part of building energy labelling programs, but these models still present low accuracy, particularly in cooling-dominated climates. The objective of this study was to evaluate the feasibility of using an artificial neural network (ANN) to improve the accuracy of surrogate models for labelling purposes. An ANN was applied to model the building stock of a city in Brazil, based on the results of extensive simulations using the high-resolution building energy simulation program EnergyPlus. Sensitivity and uncertainty analyses were carried out to evaluate the behaviour of the ANN model, and the variations in the best and worst performance for several typologies were analysed in relation to variations in the input parameters and building characteristics. The results obtained indicate that an ANN can represent the interaction between input and output data for a vast and diverse building stock. Sensitivity analysis showed that no single input parameter can be identified as the main factor responsible for the building energy performance. The uncertainty associated with several parameters plays a major role in assessing building energy performance, together with the facade area and the shell-to-floor ratio. The results of this study may have a profound impact as ANNs could be applied in the future to define regulations in many countries, with positive effects on optimizing the energy consumption.     

Solar-assisted single-double-effect absorption chiller for use in Asian tropical climates

Solar energy is accessible throughout the year in tropical regions. The latest development of absorption chillers has demonstrated that these systems are suitable for effective use of solar energy. The utilisation of solar energy for heat-driven cooling systems has significant advantages. Without a doubt, solar energy represents a clean energy source that is available without any additional fuel cost, and that can be proportionally accessible when the cooling load increases during the middle hours of the day. This study focuses on a single-double-effect absorption chiller machine that was installed in Indonesia. The system is driven by a dual-heat source that combines gas and solar energy. This system is characterised by simulating its performance in various conditions in terms of the cooling water (28–34 °C) and the hot water (75–90 °C) inlet temperatures. The reference operating condition of this system is 239 kW of cooling capacity. The mathematical model is validated and shows a good agreement with experimental data. In the operative range considered, simulation results yield a coefficient of performance between 1.4 and 3.3, and a gas reduction ratio from 7 to 58% when compared to a double-effect absorption chiller driven by gas. Based on the simulation results, this system is expected to have a good potential for widespread use in tropical Asia regions.     

Artificial neural networks for energy analysis of office buildings with daylighting

An artificial neural network (ANN) model was developed for office buildings with daylighting for subtropical climates. A total of nine variables were used as the input parameters – four variables were related to the external weather conditions (daily average dry-bulb temperature, daily average wet-bulb temperature, daily global solar radiation and daily average clearness index), four for the building envelope designs (solar aperture, daylight aperture, overhang and side-fins projections), and the last variable was day type (i.e. weekdays, Saturdays and Sundays). There were four nodes at the output layer with the estimated daily electricity use for cooling, heating, electric lighting and total building as the output. Building energy simulation using EnergyPlus was conducted to generate daily building energy use database for the training and testing of ANNs. The Nash–Sutcliffe efficiency coefficient for the ANN modelled cooling, heating, electric lighting and total building electricity use was 0.994, 0.940, 0.993, and 0.996, respectively, indicating excellent predictive power. Error analysis showed that lighting electricity use had the smallest errors, from 0.2% under-estimation to 3.6% over-estimation, with the coefficient of variation of the root mean square error ranging from 3% to 5.6%.     

An analysis of future building energy use in subtropical Hong Kong

Principal component analysis of prevailing weather conditions in subtropical Hong Kong was conducted, and a new climatic index Z (as a function of the dry-bulb temperature, wet-bulb temperature and global solar radiation) determined for past (1979–2008, measurements made at local meteorological station) and future (2009–2100, predictions from general circulation models) years. Multi-year (1979–2008) building energy simulations were carried out for a generic office building. It was found that Z exhibited monthly and seasonal variations similar to the simulated cooling/heating loads and building energy use. Regression models were developed to correlate the simulated monthly building cooling loads and total energy use with the corresponding Z. Error analysis indicated that annual building energy use from the regression models were very close to the simulated values; the difference was about 1%. Difference in individual monthly cooling load and energy use, however, could be up to 4%. It was also found that both the DOE-simulated results during 1979–2008 and the regression-predicted data during 2009–2100 indicated an increasing trend in annual cooling load and energy use and a gradual reduction in the already insignificant heating requirement in cooling-dominated office buildings in subtropical climates.     

Improved energy performance of air cooled centrifugal chillers with variable chilled water flow

This paper considers how to apply optimum condensing temperature control and variable chilled water flow to increase the coefficient of performance (COP) of air cooled centrifugal chillers. A thermodynamic model for the chillers was developed and validated using a wide range of operating data and specifications. The model considers real process phenomena, including capacity control by the inlet guide vanes of the compressor and an algorithm to determine the number and speed of condenser fans staged based on a set point of condensing temperature. Based on the validated model, it was found that optimizing the control of condensing temperature and varying the evaporator’s chilled water flow rate enable the COP to increase by 0.8–191.7%, depending on the load and ambient conditions. A cooling load profile of an office building in a subtropical climate was considered to assess the potential electricity savings resulting from the increased chiller COP and optimum staging of chillers and pumps. There is 16.3–21.0% reduction in the annual electricity consumption of the building’s chiller plant. The results of this paper provide useful information on how to implement a low energy chiller plant.     

Performance modelling of air-cooled twin-circuit screw chiller

A large variety of chiller models are available in the public domain but none can model chillers that comprise multiple and separate refrigerant circuits, despite that chillers of this type are already widely used for their good part-load performance. Presented in the paper is a mathematical model for an air-cooled twin-circuit chiller, with two screw compressors per circuit. The chiller model comprises a series of linked mathematical modules, each made up of a set of thermodynamic and empirical equations for modelling the major chiller components. The coefficients in the component models were evaluated using rated operating conditions obtained from the manufacturer and measured performance data of an existing chiller. The chiller model had been applied to simulate the performance of another set of chiller of the same make and model. Comparison of the predicted and measured performance of the chiller showed that the model could yield accurate energy use predictions over a wide range of operating conditions. The model could also provide good predictions of the variation in chiller performance due to staged operation of the separate refrigerant circuits in the chiller and of compressors in each circuit, which matched with observations made with measured chiller operation data.     

Determination and modelling of energy consumption in wheat production using neural networks: “A case study in Canterbury province, New Zealand”

This study was conducted on irrigated and dryland wheat fields in Canterbury in the 2007-2008 harvest year based on an extensive process of data collection involving a questionnaire and interviews. Total energy consumption in wheat production was estimated at 22,566 MJ/ha. On average, fertilizer and electricity were used more than other energy sources, at around 10,654 (47%) and 4870 (22%) MJ/ha, respectively. The energy consumption for wheat production in irrigated and dryland farming systems was estimated at 25,600 and 17,458 MJ/ha, respectively.In this study, several direct and indirect factors have been identified to create an artificial neural networks (ANN) model to predict energy use in wheat production. The final model can predict energy consumption based on farm conditions (size of crop area), farmers’ social considerations (level of education), and energy inputs (N and P use and irrigation frequency), and it predicts energy use in Canterbury arable farms with an error margin of ±12% (±2900 MJ/ha). Furthermore, comparison between the ANN model and a Multiple Linear Regression (MLR) model showed that the ANN model can predict energy consumption relatively better than the MLR multiple model on the selected training set and validation set.     

Theoretical and testing performance of an innovative indirect evaporative chiller

An indirect evaporative chiller is a device used to produce chilled water at a temperature between the wet bulb temperature and dew point of the outdoor air, which can be used in building HVAC systems. This article presents a theoretical analysis and practical performance of an innovative indirect evaporative chiller. First, the process of the indirect evaporative chiller is introduced; then, the matching characteristics of the process are presented and analyzed. It can be shown that the process that produces cold water by using dry air is a nearly-reversible process, so the ideal produced chilled water temperature of the indirect evaporative chiller can be set close to the dew point temperature of the chiller’s inlet air. After the indirect evaporative chiller was designed, simulations were done to analyze the output water temperature, the cooling efficiency relative to the inlet dew point temperature, and the COP that the chiller can performance. The first installation of the indirect evaporative chiller of this kind has been run for 5 years in a building in the city of Shihezi. The tested output water temperature of the chiller is around 14–20 °C, which is just in between of the outdoor wet bulb temperature and dew point. The tested COP_r,s of the developed indirect evaporative chiller reaches 9.1. Compared with ordinary air conditioning systems, the indirect evaporative chiller can save more than 40% in energy consumption due to the fact that the only energy consumed is from pumps and fans. An added bonus is that the indirect evaporative chiller uses no CFCs that pollute to the aerosphere. The tested internal parameters, such as the water–air flow rate ratio and heat transfer area for each heat transfer process inside the chiller, were analyzed and compared with designed values. The tested indoor air conditions, with a room temperature of 23–27 °C and relative humidity of 50–70%, proved that the developed practical indirect evaporative chiller successfully satisfy the indoor air conditioning load for the demo building. The indirect evaporative chiller has a potentially wide application in dry regions, especially for large scale commercial buildings. Finally, this paper presented the geographic regions suitable for the technology worldwide.     

Hybrid methodology for hourly global radiation forecasting in Mediterranean area

The renewable energies prediction and particularly global radiation forecasting is a challenge studied by a growing number of research teams. This paper proposes an original technique to model the insolation time series based on combining Artificial Neural Network (ANN) and Auto-Regressive and Moving Average (ARMA) model. While ANN by its non-linear nature is effective to predict cloudy days, ARMA techniques are more dedicated to sunny days without cloud occurrences. Thus, three hybrids models are suggested: the first proposes simply to use ARMA for 6 months in spring and summer and to use an optimized ANN for the other part of the year; the second model is equivalent to the first but with a seasonal learning; the last model depends on the error occurred the previous hour. These models were used to forecast the hourly global radiation for five places in Mediterranean area. The forecasting performance was compared among several models: the 3 above mentioned models, the best ANN and ARMA for each location. In the best configuration, the coupling of ANN and ARMA allows an improvement of more than 1%, with a maximum in autumn (3.4%) and a minimum in winter (0.9%) where ANN alone is the best.     

Applying condensing-temperature control in air-cooled reciprocating water chillers for energy efficiency

This paper reports on the modelling and findings of the energy performance of an air-cooled reciprocating multiple-chiller plant under the conventional head pressure control and the new condensing-temperature control in a subtropical climate. The simulation model was validated using the operating data of an existing chiller plant. As noted from this existing air-cooled reciprocating chiller plant, there was a substantial efficiency drop at part-load resulting from the head pressure control. If operating at variable lower condensing-temperatures based on the established operating mode of the condenser fans and compressors, it is shown that the chiller consumption can be maintained below 2 kW/refrigeration ton throughout the entire range of outdoor temperature and part-load conditions, giving an average efficiency of 1.08 kW/refrigeration ton. The energy imposition due to cycling on more condenser fans can be compensated by the reduced compressor consumption. Potential energy savings of 18.2 and 29% in the annual chiller consumption are achievable by applying the condensing-temperature control to two existing chiller plants studied. This supports the need to develop the condensing-temperature control as an improvement to the conventional head pressure control.     

Optimization of water-cooled chiller system with load-based speed control

This study investigates the energy performance of chiller and cooling tower systems integrated with variable condenser water flow and optimal speed control for tower fans and condenser water pumps. Thermodynamic-behaviour chiller and cooling tower models were developed to assess how different control methods of cooling towers and condenser water pumps influence the trade-off between the chiller power, pump power, fan power and water consumption under various operating conditions. Load-based speed control is introduced for the tower fans and condenser water pumps to achieve optimum system performance. With regard to an example chiller system serving an office building, the optimal control coupled with variable condenser water flow could reduce the annual system electricity use by 5.3% and operating cost by 4.9% relative to the equivalent system using constant speed fans and pumps with a fixed set point for cooling water temperature control.     

Vulnerability of wind power resources to climate change in the continental United States

Renewable energy resources will play a key role in meeting the world's energy demand over the coming decades. Unfortunately, these resources are all susceptible to variations in climate, and hence vulnerable to climate change. Recent findings in the atmospheric science literature suggest that the impacts of greenhouse gas induced warming are likely to significantly alter climate patterns in the future. In this paper we investigate the potential impacts of climate change on wind speeds and hence on wind power, across the continental US. General Circulation Model output from the Canadian Climate Center and the Hadley Center were used to provide a range of possible variations in seasonal mean wind magnitude. These projections were used to investigate the vulnerability of current and potential wind power generation regions. The models were generally consistent in predicting that the US will see reduced wind speeds of 1.0 to 3.2% in the next 50 years, and 1.4 to 4.5% over the next 100 years. In both cases the Canadian model predicted larger decreases in wind speeds. At regional scales the two models showed some similarities in early years of simulations (e.g. 2050), but diverged significantly in their predictions for 2100. Hence, there is still a great deal of uncertainty regarding how wind fields will change in the future. Nevertheless, the two models investigated here are used as possible scenarios for use in investigating regional wind power vulnerabilities, and point to the need to consider climate variability and long term climate change in citing wind power facilities.     

An investigation of the potential for natural ventilation and building orientation to achieve thermal comfort in warm and humid climates

The aim of this study was to analyze the most important factor, the climatic conditions with respect to thermal comfort in buildings. The impact of building location and climate and orientation on thermal comfort were investigated.With the help of dynamic computer simulations the different hourly weather data were analyzed. First of all the climate determines the amount of solar radiation and mean outside temperature that a building is exposed to. The climate also influences the amount of energy that is used for heating and cooling but also the amount of energy that is used for lighting. There is solar excess which determines the amount of solar energy that is unwanted in the building. With growing amounts of glass and a glazing system that allows large solar heat gains,the impact of orientation is substantial. A detailed analysis was conducted to evaluate the potentials for improving thermal comfort. Detailed results are given in sample graphics and tables in the study. In a tropical climate the improvement in comfort by NV range between 9% and 41% (Kuala Lumpur in April). For a subtropical climate the improvements vary between 3% and 14%. In a temperate climate the improvements vary between 8% and 56%. The results showed that NV has a good potential in tropical and temperate climates but not in subtropical climates. Especially in Hong Kong it seems to be very difficult to apply NV. The results showed that in particular in the hottest period (summer) the potential for comfort improvements is rather small. The design of climate responsive building envelopes should take this into consideration.