Hideki Kikumoto

Bayesian inference for thermal response test parameter estimation and uncertainty assessment

This work was supported by the Japan Society for the Promotion of Science (JSPS) (KAKENHI, grant numbers 26709041 and P16074).

Bayesian source term estimation of atmospheric releases in urban areas using LES approach

The estimation of source information from limited measurements of a sensor network is a challenging inverse problem, which can be viewed as an assimilation process of the observed concentration data and the predicted concentration data. When dealing with releases in built-up areas, the predicted data are generally obtained by the Reynolds-averaged Navier-Stokes (RANS) equations, which yields building-resolving results; however, RANS-based models are outperformed by large-eddy simulation (LES) in the predictions of both airflow and dispersion. Therefore, it is important to explore the possibility of improving the estimation of the source parameters by using the LES approach. In this paper, a novel source term estimation method is proposed based on LES approach using Bayesian inference. The source-receptor relationship is obtained by solving the adjoint equations constructed using the time-averaged flow field simulated by the LES approach based on the gradient diffusion hypothesis. A wind tunnel experiment with a constant point source downwind of a single building model is used to evaluate the performance of the proposed method, which is compared with that of the existing method using a RANS model. The results show that the proposed method reduces the errors of source location and releasing strength by 77% and 28%, respectively.

Impacts of inland water area changes on the local climate of Wuhan, China

A distinct feature of Wuhan is that almost a quarter of the total area of this city is covered with water, leading to its unique hot and humid climate characteristics in summer. However, according to records, water area in built-up zone of Wuhan has been reduced by 130.5 km2 from 1965 to 2008, while the annual average air temperature has been increased by more than 3℃. To investigate the quantitative connection between the water area reduction and air temperature increase, three scenarios were simulated in a summer; to evaluate the impact of water reduction on the local thermal environment in different water areas; and to study the impact of water reduction on the urban heat island (UHI) phenomenon. Meso-scale meteorological models of Weather Research and Forecasting model were applied in this study for quantitative assessment and prediction. With the predictions, this study reveals that the decreased water area could affect air temperature, wind velocity and wind flow direction, energy balance and the UHI intensity. The simulations show that areas with significant wind velocity, wind direction and air temperature differences are distributed among the downwind zones. Moreover, the areas with high UHI intensity are wider and farther from the boundary of urban areas because of the reduction of water areas.

Two thermal performance test (TPT) datasets of a single U-tube borehole heat exchanger with inlet setpoint temperatures of 30 °C and 40 °C

The presented thermal performance test (TPT) datasets were related to the research article “New perspectives in thermal performance test: Cost-effective apparatus and extended data analysis” (Choi et al., 2018), where a new TPT apparatus was developed by adding a solid-state-relay and a proportional–integral–derivative (PID) controller to a thermal response test apparatus. Using the developed apparatus connected to a 50-m-long vertical ground heat exchanger, two TPTs were conducted for 144 h with inlet setpoint temperatures of 30 °C and 40 °C. The raw data were measured at 5 s intervals and consisted of the inlet, and outlet fluid temperatures, and the flow rate. The attached MATLAB script allows users to easily filter the data at user-specified time intervals. Moreover, the execution of code provides two additional quantities: heat injection rate and unit heat exchange rate. The datasets are shared for the following purposes: (1) performance comparison of various ground heat exchangers using the unit heat exchange rate (2) comparison of the control performance of a newly developed TPT apparatus, (3) validation of an analytical or numerical thermal response model, and (4) validation of a parameter estimation algorithm.

Performance verification of typical and design weather year by thermal load calculation targeting office building

We proposed a new type of weather year data for building energy simulations named the typical and design weather year, which can be used for estimating both average and peak energy demand for one year of building energy simulation. The typical and design weather year is generated using a quantile mapping method. In this paper, we made the typical and design weather year for three cities, Tokyo, Sapporo, and Kagoshima, representing a wide range of climatic conditions in Japan, and evaluated its performance by conducting building energy simulations targeting prototypical office buildings. We found that the typical and design weather year was more than twice as accurate in estimating average energy demand as the existing typical weather year data. Typical and design weather year can also estimate peak energy demand with high accuracy. Practical application: The cumulative distribution functions of a target weather data set, on which quantile mapping is performed, are modified to consist entirely of parent multi-year weather data. Therefore, typical and design weather years based on quantile mapping are expected to be useful as versatile weather year data representing the various weather characteristics of multi-year conditions. In this study, we found that the typical and design weather year can estimate both average and peak energy demands in building energy simulations. New type of weather year data named the typical and design weather year can be used as both typical and design weather data.

Large-Eddy Simulation of the Turbulent Boundary Layer Using Generated Inflow Turbulence with Estimated Statistics

We conducted large-eddy simulations of the turbulent boundary layer using the inflow generation technique proposed by Xie and Castro (2008) (XC08) to validate this method in a downscaling simulation. To generate inflow turbulence, we used both exact and estimated statistics. The estimated inflow profiles were relatively accurate and the behaviors of the turbulent intensities and the Reynolds stress were very similar to those generated with exact statistics. Although the inflow turbulence statistics calculated by XC08 were consistent with those of the designed profiles, their instantaneous velocities did not satisfy the continuity equation for fluids. As a result, the turbulence rapidly decayed in the near-inlet region, having more significant effects than the difference in inflow statistics would suggest. Therefore, improvements in XC08 with respect to continuity are needed for practical application.

Numerical Analysis of the Momentum Transport and Temporal and Spatial Scales of Turbulent Coherent Structures in the Urban Boundary Layer Using Large Eddy Simulation

Numerical analysis is performed on an urban model to investigate turbulent coherent structures in urban boundary layer (UBL) using large-eddy simulation (LES). To confirm the accuracy of LES, its results are compared with those of a wind tunnel experiment. Streamwise mean wind velocity and streamwise turbulent intensity calculated using LES are generally identical to results of a wind tunnel experiment. Four-quadrant and correlation coefficient analysis are performed to investigate momentum transport and temporal and spatial scales of turbulent coherent structures in UBL. Four-quadrant analysis results show that strong ejections and sweeps are generated at the top of urban canopy because of inflection point instability of streamwise mean wind velocity. They correspond to drag force above and below the top of urban canopy respectively. Streamwise integral scales are calculated from correlation coefficients. As height increases, streamwise integral time scales decrease, however, streamwise integral length scales remained almost constant.