The authors applied a Markov chain model to calculate the transition probabilities between degradation levels, evaluated as three grades from Levels 1 to 3, in the periodic inspection records of air-cooled heat pump chillers installed in many office buildings. The changes in the gradation levels included degradation (e.g., Level 1 to Level 2), no change in condition (e.g., Level 2 to Level 2), and repair (e.g., Level 3 to Level 1) based on actual maintenance activities. Our analysis revealed that the calculated transition probabilities varied according to the inspection item and the age of the facilities. By applying Markov chains to the repair process as well as degradation, the transition probabilities of repair could be modified to simulate different repair plans, such as reducing the number of repairs or increasing the number of observations with no repairs. Next, a Monte Carlo simulation was performed to compare the repair plans with the changed probability of repair and the reference plan with the calculated transition probabilities. The repair plans were formulated as follows. (1) Repair Plan 1-1: repair all of Level 2 and Level 3 (2) Repair Plan 1-2: repair Level 2 and Level 3 with half the probability of the reference plan (3) Repair Plan 2-1: observe all of Level 2 and repair all of Level 3 (4) Repair Plan 2-2: observe all of Level 2 and repair Level 3 with half the probability of the reference plan (5) Repair Plan 3: observe all (no repairs) The authors counted the total numbers of repairs and observations in the simulation results and found a negative correlation between the two, with one increasing and the other decreasing, confirming that the calculated results can be used to compare repair plans. Repair Plan 1-1 can be divided into inspection items for which the number of repairs increases and for which the change is small. Repair Plan 1-2 increased observations 1.7 - 2.9-fold and ultimately reduced repairs by 5 - 22%. Repair Plan 2-1 increased observations 2.1 - 7.9-fold and decreased repairs by 15% - 74% except for pressure gauges and water systems. Repair Plan 2-2 increased observations 1.7 - 8.8-fold and decreased repairs by 16 - 81%. Repair Plan 3, which observed all items, increased observations 2.1 - 17.8-fold. As described above, a Monte Carlo simulation applying Markov chains to periodic inspection records allows quantitative comparisons of conflicting repairs and observations, which is useful for selecting repair plans.

We developed an open-cycle type adsorption thermal storage system. Owing to the adsorbent in HAS-Clay, the thermal storage system can utilize low-temperature waste heat. To demonstrate offline heat transportation in motor factories, the thermal storage tank charged in a co-generation system was transported in a tractor and used to dehumidify the supply air in an air handling unit (AHU) painting. Generally, the AHU cools and dehumidifies the air with chilled water and then heats it with hot water to obtain required during painting air with the specified temperature and humidity. The energy consumption of cold and hot water in the AHU decreased with dehumidification, which was realized using the adsorbent thermal storage tank. Consequently, the CO₂ reduction effect of 57% was confirmed from the demonstration test. Further, painting was confirmed to be an economical application of the adsorbent thermal storage system.

The calculation procedure of “pseudo-atmospheric transmittance” based on the month and weather in a particular location is described in this paper. The pseudo-atmospheric transmittance with regard to the month and weather in Shimonoseki was calculated based on the hourly data of total solar radiation on the horizontal plane of the Shimonoseki Local Meteorological Observatory in 2021. As an example of the application of the pseudo-atmospheric transmittance, the hourly total solar radiation was predicted using the pseudo-atmospheric transmittance, the calculated position of the sun, and the weather forecast. The hourly total solar radiation in the western part of Yamaguchi Prefecture (including Shimonoseki City) from March 13 (Sun) to March 19 (Sat), 2022 was predicted by using the weather forecast texts announced by the Shimonoseki Local Meteorological Observatory at 17:00 on the days before the forecast target days. The accuracy of the predicted solar radiation largely depends on the weather forecast information. On March 15, 16, and 18, when the real weather was almost as forecasted, the errors in the predicted total daily solar radiation were -8.0, -5.0, and -0.9%, respectively. On the contrary, on March 13, when the difference in the real and forecasted weather occurred, the error in the predicted total solar radiation was 74.0%. The weather forecast information with a high spatiotemporal resolution is required to improve the prediction accuracy.