In this study, we analyze the load situation of district heating and cooling (DHC) in Japan for many facilities and investigated how energy can be saved. We surveyed and analyzed the actual condition of the DHC load, which comprises approximately 3.6million square meters in Japan, an average cold heat of 55W/m² at maximum load, an average thermal average of 34W/m², cold heat of 249MJ/m² at yearly load, thermal energy of 121MJ/m², and a load factor of 10%. We reported the actual situation there is a lot of low load, the annual integrated value of the load exceeding the load factor of 10% is only 55% and energy saving in low load parts such as overtime is very important for promoting low carbonization in the future. We show the averaged load duration curve that serves as a reference for load prediction from the wide range of the DHC area; the time variation of the cold heat load ratio in the intermediate term, the summer season, and in winter; and the reference value of the daily cumulative load and the correlation between the cold thermal load of the day and the average daily temperature is high and it is possible to predict a constant load. We reported that the cold heat load in 2016 decreased by approximately 28% per year and the peak cold heat load of the year by approximately 25%, the thermal annual loads and the thermal maximum load in 2016 by approximately 8% annually compared to 2010. We proposed a method for investigating the approximate heat source capacity partition (minimum partial load capacity partition (MPCA)) with a lower load operating range using the load duration curve in DHC . We reported that the efficiency changes greatly depending on the capacity partition method and reported the capacity partition reference value in DHC. We showed that it is important to consider the improvement gained by considering the efficiency of the heat source including partial loads such as overtime that occur many times in year. In order to improve the efficiency of the low load part, we examined the effects of the heat source fragmentation and inverterization, and a case of the latter was reported of DHC maintaining the system COP at 1.15 in primary energy conversion at a load of 10% or less. In the trend toward low load buildings, it is necessary to accumulate more than a few buildings to make the efficiency of the very small load part a certain scale, which shows the necessity of load consolidation and load accumulation by DHC. We believe that the need for low carbonization by DHC will further increase.

View full abstract

Charging and discharging characteristics of the phase change material (PCM) changes depending on the temperature of the phase change and the quantity installed. It is supposed that the supercooling that occurs prior to the solidification has an effect on the charging characteristics of cold heat. It is necessary when searching for the efficient use of PCM to understand the charging and discharging characteristics by simulation. In this study, for a solidified PCM, a model that can express supercooling and its removal and a method to calculate the thermal conductivity for the process of phase change progress were proposed. The specific enthalpy distributions were calculated from the one-dimensional equation of the thermal conduction of the PCM. The temperature distributions were calculated using its T-h characteristics. Fairly good agreements were obtained between the calculated and the measured periphery temperatures of the PCM. Furthermore, in the case of the temperature of the cooling medium near that of the phase change of the PCM, it was shown that the calculated result shows that the charged cold heat does not increase because of the non-removal of supercooling.

View full abstract

A district heat sharing network system is expected to reduce energy consumption and improve machine efficiency in not only a building but also a district. However, an optimal operation of the system is not easy to determine because of the system complexity such as the many available combinations of thermal outputs and highly nonlinear configurations. Some recent studies that aimed to optimize an operating schedule of more two buildings simplified the optimization problem with linearization of the system configuration, and pre-determination of operating priority of the each heat source machine. Although the linearization would be suitable for reducing the computation costs drastically, some nonlinear models (e.g. artificial neural network (ANN)) have been commonly used in recent years. It is also expected that statistical modeling will be widely used for data assimilation of complex systems and fault detection. Thus, the optimization method should be more flexible to integrate the complex models, in comparison with the linear programming and other mathematical programming methods because the methods have some limitations such as linear or convex configurations. To handle complex calculation conditions and optimize many decision variables such as the thermal outputs of each heat source machine at each building in a certain district, ε -constrained differential evolution with random jumping (εDE-RJ) was proposed in this series of studies. εDE-RJ could handle almost all functions such as concave, multimodal, and discrete configurations. In addition, ANN was used to predict the temperature variations of the shared thermal energy storage that was used in a district comprising of six buildings. ANN was also used to reduce the computation costs against a detailed physical model. In this paper, εDE-RJ combined with ANN was compared to a reinforcement learning method such as Q-learning which has been increasingly popular to optimize energy system operation in recent years to minimize the operating cost and improve system efficiency. Some parameters such as the learning rate and the variable ε_Q of the ε-greedy method of Q-learning were evaluated to determine an optimal combination of parameters that will yield a quasi-optimal solution. It was concluded that the learning rate should be large a value such as 0.90 to make the results more stable. In addition, the parameter ε_Q should be decreased from 1.0 to 0.1 instead of being assigned a constant value (0.8). The result obtained by employing ε DE-RJ with 160,000 iterations was 1,394,063 yen/day. In contrast, the result obtained by employing Q-learning was 1,707,947 yen/day. Therefore, εDE-RJ was superior to Q-learning in this optimization problem because the difference in the results was 18.3% and the computation time of ε DE-RJ was only 476 s, whereas the time of Q-learning was approximately 5900 s.

View full abstract

The aim of this study is to develop an optimal operation technology for an HVAC system. To this end, an HVAC system is appropriately operated on the basis of the outside air and loading conditions to eliminate wasteful operation. Moreover, an optimal operating method is examined for a heat source system, including variable air volume control of a cooling tower fan. The optimal set points for the cooling-water temperature and cooling-tower-fan frequency were analyzed using an energy simulation tool. A relational formula for optimal operation of the heat source system was derived by analyzing the result. The approximated relational formula was used to determine the optimal set point of the cooling-tower-fan frequency from the outdoor wet-bulb temperature and the load factor of the chiller. A method in which the cooling-tower-fan frequency was considered the set value was compared with another method in which the cooling-water temperature was considered the set value. The results show that the former method was superior when the input of the outside air wet bulb temperature and the load factor of the heat source machine differ from the original value, and the former method was more accurate approximation. Further, the energy consumption of a heat source system was determined using various control methods. The calculation results show that the energy consumption could be reduced using the optimal operating method with the help of the relational formula. In addition to the simulation, the energy-saving effects of the method were confirmed from a field test.

View full abstract

In CFD analysis, in the HVAC engineering field, it is necessary to reproduce with high accuracy the airflow characteristics of outlets when predicting the airflow distribution in a room. However, most of the outlets installed in the room have complicated shapes. Considering the calculation time and workload for CFD analysis, it is difficult to reproduce the detailed shapes of the outlets at the design stage. Therefore, in this research, we study the airflow characteristics of square ceiling diffusers whose airflow properties are complex, and study how to set the CFD parts for general-purpose use. In this report, we conducted an experiment to visualize the airflow surrounding the diffusers, and derived the airflow characteristics. We then examined the method of setting the boundary conditions when applying the momentum method to a structural grid CFD, and compared the analytical results with the experimental results. We confirmed that it is possible to calculate the spread radius and the temperature distribution inside the room with practically satisfactory accuracy in the design stage.

View full abstract

In recent years, the adoption of air-conditioning systems using ceiling radiation in offices has been increasing. Many previous studies showing the effectiveness of the ceiling-radiation air-conditioning system have been conducted, mainly by evaluating the whole-body thermal sensation and local thermal sensation based on experiments and actual measurements under a cooling condition. However, there are few studies focusing on the relaxation effect of the set value of room temperature due to the decrease of the mean radiation temperature (MRT), and the perceivable heat balance of the human body surface. Therefore, the purpose of this study was to reveal the heat dissipation characteristics of the human body by estimating human physiological quantities while reproducing micro climates around the human body in detail using CFD analysis coupled with a numerical thermoregulation model of the human body, targeting the ceiling-radiation air-conditioning system introduced in office spaces. The results show that, since the MRT in the ceiling-radiation air-conditioning system is less than that of the ceiling supply air-conditioning system, the ratio of radiation heat loss from the human body in the ceiling-radiation air-conditioning system is greater than that in the ceiling-supply air-conditioning system. Furthermore the temperature setting could be raised by 1.6℃.

View full abstract

In this survey, the dispersion of suspended particulate matter (SPM) was examined. In the previous investigations, SPM was also examined multiple times in several buildings. However, these past examinations were limited to large particles. In this study, smaller SPMs (10nm -10 µm) were investigated by using a portable small size scanning mobility particle sizer (SMPS) and a laser particle counter (LPC). In addition, the volume dispersion was calculated. The results were as follows. 1) the distribution of small size SPM (<1.0 µm) showed a log-normal distribution, 2) larger SPM distributions also approximated a linear function with a slope greater than -3.0, 3) these SPM distributions can be converted into quadratic functions in double logarithms, 4) particles with 200-300 nm diameter account for the highest volume concentration in smaller particles, and 5) the point of inflection appeared to be 1 µm, and this number is approximately one order of magnitude larger than what is generally stated.

View full abstract

It is sometimes difficult for us to apply the commissioning process to HVAC systems newly installed throughout the year during the production phase under the restrictions of the existing contract agreement. Therefore, we should continuously monitor the operational state of a HVAC system after completion and optimize operational setting points as required to prevent energy wastage. On the other hand, energy visualization system such as BEMS has been spreading rapidly in recent years and faults that deteriorate energy saving performance of a HVAC system can be detected more easily than before with such a system. Aiming to improve the energy usage efficiency of a target building, we applied the commissioning process to its water thermal energy storage system by utilizing BEMS data to unveil the operational state after completion.　In this paper, a series of setting point optimization results which focused on the heat source during summer is shown. For heat source operation and the temperature profile of the water thermal storage tank, satisfactory results were obtained after adequately activating the heat-source outlet temperature control which was not in operation before commissioning process was applied.

View full abstract

In the architectural plan of a building, which is a head office located in an urban area, we aimed to create a next-generation environment-friendly office that achieves both a pleasant and comfortable working environment and environment-friendliness. To realize this concept, a perimeter aisle in the floor planning and reverse slab in the sectional planning were adopted. By adopting state-of-the-art environmental technology, we constructed a "self-stabilizing green building" to create a stable indoor environment with relatively low energy. By measurement survey on thermal environment and energy data analysis, we confirmed self-stability of the thermal environment, extension of the renewable energy use and reduction of the energy consumption. In this paper, we present the design and basic performance verification of the building.

View full abstract