Milorad Bojić

Artificial neural networks for the prediction of the energy consumption of a passive solar building

Artificial neural networks (ANNs) have been used for the prediction of the energy consumption of a passive solar building. The building structure consists of one room with an inclined roof. Two cases were investigated, an all insulated building and a building with one wall made completely of masonry and the other walls made partially of masonry and thermal insulation. The investigation was performed for two seasons: winter, for which the building with the masonry-only wall is facing south, and summer, for which the building with the masonry-only wall is facing north. The buildings thermal behaviour was evaluated by using a dynamic thermal building model constructed on the basis of finite volumes and time marching. The energy consumption of the building depends on whether all walls have insulation, on the thickness of the masonry and insulation and on the season. Simulated data for a number of cases were used to train an artificial neural network (ANN) in order to generate a mapping between the above easily measurable inputs and the desired output, i.e., the building energy consumption in kWh. The simulated buildings had walls varying from 15 cm to 60 cm in thickness. The objective of this work is to produce another simulation program, using ANNs, to model the thermal behaviour of the building. A multilayer recurrent architecture using the standard back-propagation learning algorithm has been applied. The results obtained for the training set are such that they yield a coefficient of multiple determination (R2 value) equal to 0.9985. The network was used subsequently for predictions of the energy consumption for cases other than the ones used for training. The coefficient of multiple determination obtained in this case was equal to 0.9991, which is very satisfactory. The ANN model proved to be much faster than the dynamic simulation programs.

Influence of thermal insulation position in building envelope on the space cooling of high-rise residential buildings in Hong Kong

At the present time, thermal insulation is almost not used in fabric of tall residential buildings in Hong Kong, as their fabric slabs usually comprise concrete layer covered on each side by plaster layers. This study investigates into the influence of an existence of the thermal insulation layer in the outside walls on the yearly cooling load and yearly maximum cooling demand in two typical residential flats in a high-rise residential building by employing HTB2, detailed building heat transfer simulation software. During the investigations, the thermal insulation layer up to 15 cm thick was placed either at the inside, or at the outside, or at the middle part of the outside wall structure. Then, the concrete layer was up to 40 cm thick. The simulation predictions indicate that the highest decrease in the yearly cooling load of up to 6.8% is obtained when a 5 cm thick thermal insulation layer faces the inside of the residential flat. The highest decrease in the yearly maximum cooling demand of 7.3% is recorded when a 5 cm thermal insulation layer faces either the outside or the inside of the flat; this depends on the flat orientation. However, much weaker reductions in the yearly cooling load and yearly maximum cooling demand are found when the thickness of thermal insulation is increased above 5 cm and the thickness of concrete is increased above 10 cm.

Thermal insulation of cooled spaces in high rise residential buildings in Hong Kong

Thermal insulation is seldom applied to the fabric of high rise residential buildings in Hong Kong. Where it is used, it is placed in the fabric component at the side that faces the outdoors or non-air conditioned spaces, such as kitchens, bathrooms and entrance lobbies. The paper reports on the results of an investigation into the effects of including a thermal insulation layer in the fabric components that separate cooled spaces from the outdoors and from non-air conditioned spaces. In the investigation, the effects of placing the insulation layer at the indoor side, in the middle and at the outdoor side, or the side of the adjacent non-air conditioned space, of the fabric components were evaluated and compared. For the doors, the insulation layer was always put between the two face panels. The detailed building heat transfer simulation program HTB2 was employed to calculate the yearly cooling loads and the maximum cooling demand in the year of two typical flats in a high rise residential building. The simulation predictions show that the highest reduction in the yearly cooling load, by 9.1%, and in the maximum cooling demand, by 10.5%, would be achieved when a 50 mm thick thermal insulation layer was placed at the indoor side of the walls that enclose the cooled spaces. However, increasing the thickness of the insulation layer beyond 50 mm and of the concrete layer to above 100 mm would only lead to insignificant further reductions in the yearly cooling load and the yearly maximum cooling demand.

Numerical simulation, technical and economic evaluation of air-to-earth heat exchanger coupled to a building

An air-to-earth heat exchanger (ATEHE) consists of pipes buried in soil. We have evaluated the technical and economic performance of an ATEHE coupled to the system for heating or cooling of a building that uses 100% fresh air as heating or cooling medium during winter and summer. The soil is divided into elementary layers. The problem solved, is non stationary; however, steady state-energy equations are used for soil layers in each time step. It is found that the use of the ATEHE covers a portion of the daily building needs for space heating or cooling. The cost of the ATEHE energy is lower for summer than for winter.

Simulation of a solar domestic water heating system using a time marching model

This paper presents the modelling and simulation of a solar water heating system using a time marching model. The results of simulations performed on an annual basis for a solar system, constructed and operated in Yugoslavia, which provides domestic hot water for a four-person family are presented. The solar water heater consists of a flat-plate solar collector, a water-storage tank, an electric heater, and a water-mixing device. The mathematical model is used to evaluate the annual variation of the solar fraction with respect to the volume of the storage tank, demand hot water temperature required, difference of this temperature and preset storage tank water temperature, and consumption profile of the domestic hot water demand. The results of this investigation may be used to design a solar collector system, and to operate already designed systems, effectively. The results for a number of designs with different storage tank volumes indicate that the systems with greater volume yield higher solar fraction values. The results additionally indicate that the solar fraction of the system increases with lower hot water demand temperature and higher differences between the mean storage water and the demand temperatures. However, when a larger storage tank volume is used, the solar fraction is less sensitive to a variation of these operation parameters.

Energy from a two-pipe, earth-to-air heat exchanger

Solar energy accumulated in the soil may be utilized with an air-to-earth heat exchanger (ATEHE) which has two pipes buried in the soil, one made of PVC and one of steel. During the winter, air is heated; during the summer, it is cooled and then used in an air-conditioning device. To obtain the mathematical model of the ATEHE, we divided the soil and pipes into elementary volumes, used steady-state energy equations, and applied a time-marching method. We determined how the season, soil thermal conductivity and pipe spacing influence energy transfer from the soil to the ATEHE and also the steel-pipe contribution to this energy transfer.

Energy performance of windows in high-rise residential buildings in Hong Kong

At the present time in Hong Kong in its hot and humid climate, a single pane, clear glazing is most often used in windows of tall residential buildings. During this study, we employed HTB2, detailed building heat transfer simulation software, to evaluate a decrease in the yearly cooling load (Q) and in the peak cooling-load (D) in two residential flats due to different glazing single pane types and different flat orientations. The investigated glazing types were clear glazing, tinted glazing, reflective glazing, and tinted and reflective glazing.

Influence of envelope and partition characteristics on the space cooling of high-rise residential buildings in Hong Kong

Abstract Nowadays, many apartments in residential buildings in Hong Kong are equipped with air-conditioners to provide comfort cooling for the living, dining and bedrooms. Air-conditioners serving the living and dining rooms would be operated mostly in the evenings and sometimes during daytime as well, but those serving the bedrooms could remain operating overnight. Other rooms, such as bathrooms and kitchens, will not normally be air-conditioned. The paper describes a study into the influence of the building envelope construction on the space cooling loads in residential buildings. The yearly cooling loads of apartments were predicted by using HTB2, a detailed building heat transfer simulation program. The study includes modifications to the base-line characteristics of walls and doors of two typical apartments. The modified walls have an additional 50 mm thick thermal insulation and/or a thickness of up to 400 mm of concrete. Modified doors have an additional 38 mm thick thermal insulation. The simulation predictions indicate that the highest reduction in the yearly cooling load could be obtained when the modified walls and doors were used at suitable locations. In addition, the study results show that the yearly cooling load would increase when the thermal insulation was applied to external walls without increasing the thickness of the concrete layer, and when the concrete layer was thickened but without addition of insulation.

Locating air-conditioners and furniture inside residential flats to obtain good thermal comfort

Majority of residential apartments in Hong Kong may be equipped with up to three window-type air-conditioners that operate during 7 months of hot and humid weather. The units are placed in the living rooms and bedrooms. The positions of these units and furniture inside these small rooms are directed in some extent by the architectural design, however, residents have possibility to decide about their actual position. The occupants should locate their air-conditioner with respect to the room furniture to avoid the cold air re-circulation and draft, and to obtain good thermal comfort for residents. To accomplish this, flat manufacturers should give some recommendations to the occupants, and architects should include such concerns in their design. For three positions of an air-conditioner unit and one position of a bed and a cupboard in a typical Hong Kong residential bedroom, the air velocities, temperatures, and air diffusion performance index are determined by using the computational fluid dynamics (CFD) software FLOVENT. The cases with the maximum benefit and maximum consequences in terms of thermal comfort in the bedroom are identified.

MILP optimization of a CHP energy system

To support its technology, a factory may consume heat and power. We can produce them inside the factory or buy them from an outside plant. For production of these energies, we use a CHP energy system consisting of a steam boiler and a backpressure turbine. Using a steady-state, bottom-up approach, a network of energy modules and mixed integer linear programming (MILP), we have developed a software package KOM9. Depending on the unit costs of different kinds of energies, we found the cost region diagram. This diagram gives an energy plant operator a recommendation for use of an optimum energy consumption of particular composition. This yields the minimum energy expense and, at most, 65.6% in cost saving. Regardless of an inexpensive production of the power, we show that sometimes it may not be optimal to operate the turbine. This is the case when refuse steam for heating purposes is available.