Journal of Asian Architecture and Building Engineering

ISSN: 1346-7581 (Print) 1347-2852 (Online) Journal homepage: https://www.tandfonline.com/loi/tabe20

Study on the Heat Load Characteristics of

Underground Structures
Part 1. Field Experiment on an Underground Structure under an Internal
Heat Generation Condition

Kyung-Soon Park, Hisaya Nagai & Takeshi Iwata

To cite this article: Kyung-Soon Park, Hisaya Nagai & Takeshi Iwata (2006) Study on the Heat
Load Characteristics of Underground Structures, Journal of Asian Architecture and Building
Engineering, 5:2, 421-428, DOI: 10.3130/jaabe.5.421

To link to this article: https://doi.org/10.3130/jaabe.5.421

© 2018 Architectural Institute of Japan

Published online: 24 Oct 2018.

Submit your article to this journal

Article views: 470

View related articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=tabe20
 Study on the Heat Load Characteristics of Underground Structures
Part 1. Field Experiment on an Underground Structure under
an Internal Heat Generation Condition

Kyung-Soon Park*1, Hisaya Nagai2 and Takeshi Iwata3

1
Graduate Student, Department of System Engineering, Graduate School of Engineering, Mie University, Japan
2
Associate Professor, Department of Architecture, Faculty of Engineering, Mie University, Japan
3
Research Engineer, Department of Architecture, Faculty of Engineering, Mie University, Japan

Abstract
This study is being conducted to clarify the heat load characteristics of underground structures. The authors
have sought to achieve this by clarifying the heat and moisture behavior of an underground basement and
surrounding ground, and acquiring basic data for computational analysis. This paper presents measurement
results acquired over approximately one year of field experiments on an underground experimental basement
under internal heat generation conditions, commenced in October 2004. The authors also present the results
of analyses on the heat and moisture behavior of the experimental basement and ground, the influence of
internal heat generation on the surrounding ground, the condensation behavior, the interrelationship between
precipitation and the fluctuation of moisture content in the ground, the annual mean heating load per unit
area, and other factors.

Keywords: underground structure; heat and moisture behavior; internal heat generation; moisture content; condensation

1. Introduction moisture content and temperature. Thus, this effect

A number of underground structures have recently needs to be treated as a coupled non-linear problem
been constructed to make effective use of land, to (Matsumoto, 1984). Many studies have investigated
conserve energy, and so forth. From the perspective of simultaneous heat and moisture transfer in porous
the thermal environment, there is a need to precisely mediums (Matsumoto et al., 1988).
evaluate the heat and moisture behavior of underground A number of studies have been conducted on the
basements and the surrounding ground. thermal properties and heat load calculations of
The ground and concrete are generally porous underground spaces. Hasegawa et al. (1992) measured
mediums affected by moisture to some degree. the temperature of a semi-underground basement and
Moisture is one of the main components of the evaluated its thermal performance. Mitalas (1987)
thermal environment in underground structures and proposed a simple heat load calculation method for
the surrounding ground. The moisture in the ground three kinds of basement (deep, shallow, and slab on
and concrete has a great influence on the surrounding grade). Sobotka et al. (1995) analyzed deep basement
area. Further, the moisture content of the ground and heat loss by measurement and calculation.
concrete is strongly influenced by the precipitation In all of these studies, however, the load calculations
and depth of the water table in the ground. For this and evaluations of the heat environments treated the
reason, the moisture content profiles in the ground heat transfer in the ground as simple heat diffusion.
and concrete need to be predicted in order to precisely The influence of moisture was not sufficiently
consider the heat transfer in underground structures. considered. None of the simple methods for heat
Moisture strongly affects heat transfer, and the load calculation have incorporated the influences of
heat and moisture transfers in a porous medium are moisture and precipitation with sufficient generality or
coupled phenomena (Matsumoto, 1978). The effect of accuracy.
coupled phenomena is also non-linear process because There were thus two main purposes of this study:
of the sensitive variation of physical parameters with first, to actually measure and analyze the heat and
moisture behavior of an underground basement
*Contact Author: Kyung-Soon Park, Graduate Student, and surrounding ground through a long-term field
Dept. of System Eng., Graduate School of Eng., Mie University, experiment; second, to develop a simple heat load
1577 Machiya-cho, Kurima, Tsu, 514-8507, Japan calculation method incorporating the influences of
Tel: +81-59-231-9446 Fax: +81-59-231-9452 moisture, precipitation, and so on.
e-mail: [email protected] This paper presents measurement results and
( Received March 16, 2006 ; accepted July 31, 2006 ) details on an experimental underground basement

Journal of Asian Architecture and Building Engineering/November 2006/428 421

and surrounding ground kept under an internal heat the devices used to measure them. Fig.5. shows the
generation condition. The measurement results points where the temperature and water chemical
presented were collected through one-year of field potential were measured in the surrounding ground.
experiments. The noughts and crosses shown in the surrounding
ground in Fig.5. indicate the points for measurement
2. Outline of the Experiment of the temperature and water chemical potential,
2.1 Experimental basement respectively. A thermocouple and hygrometer were
The experimental site was located in Tsu, a city installed in the instrument shelter to measure the
in central Japan. The basement was constructed in outdoor air temperature and relative humidity. A
1997 without any form of air-conditioning. As the pyrheliometer and albedometer were set up on the
temperature was thus left unadjusted, the basement upper part of the ground level to measure the solar
presumably had little influence on the heat and radiation and absorption ratio. The ground surface
moisture environment of the surrounding ground. in the experimental location was kept bare during
The site plan, floor plan, and sectional plan of the measurement. The fluctuation of the underground water
experimental basement are shown in Figs.1. to 3. The table was measured using a water-level gauge placed
basement measured 3.5m × 2.0m × 2.0m (see Figs.2. in a well that had been dug at the experimental site.
and 3), with a floor area and volume of 7m 2 and 2.3 Experimental conditions
14m3, respectively. Once the basement was installed The experiment was commenced on October 22,
underground, the roof was at a depth of 1.0m below the 2004 with the goal of clarifying the heat and moisture
ground level. To clarify the influences of moisture from behavior of the basement and surrounding ground
the ground, the wall of the basement was constructed under an internal heat generation condition.
with reinforced concrete untreated by any form of The indoor air temperature and ventilation rate of the
waterproofing or moisture prevention processing. The basement during the experiment were set to 22°C and 2.2
thicknesses of the ceiling/sidewall and floor were 0.25m times per hour, respectively. The indoor air temperature
and 0.3m, respectively, and blinding concrete and was controlled using an electric hot-air heater regulated
crushed stone were additionally paved below the floor by thermostat, and the ventilation was controlled using
level to a thickness of 0.3m. the variable-air-volume fan mentioned above. The
To suppress the influences of thermal turbulence indoor air humidity was not controlled.
from the outside as much as possible, the entrance of The ventilation rate in this study was calculated
the basement was insulated from heat by a 10cm-thick by the vibration decay method of the carbon dioxide
foam polystyrene board during the experiment. The concentration. Results from a number of similar
ventilation was adjusted with a variable-air-volume fan experiments performed periodically confirmed that the
set up in the upper part of the exhaust cylinder (See ventilation rate was steady during the measurement
Fig.3.). The underground water table at the site was period. Further, a regression analysis confirmed that
located at a depth of about 3m from the ground level. the ventilation rate was uncorrelated with the wind
To keep the floor level higher than the underground velocity, wind direction, and the difference between
water table and to keep the basement completely outdoor and indoor air temperatures. The precipitation,
underground, the basement was banked by about wind velocity, and wind direction were recorded
1m. The soil was banked from each sidewall of the automatically at 5-minute intervals and the other data
basement to a distance 8m in a horizontal direction. were recorded automatically at 1-minute intervals.
2.2 Measurement
Fig.4. indicates the measurement points in the 3. Ground Characteristic
basement. Table 1. lists the conditions measured and According to the exploration of three bores at points
Table 1. Conditions Measured and Measurement Devices

422 JAABE vol.5 no.2 November 2006 Kyung-Soon Park

 Fig.1. Placement of the Basement at the Site (Unit : mm)

Fig.5. Points for Measurement of the Temperature and Water

Chemical Potential in the Surrounding Ground

Fig.2. Basement Floor Plan (Unit : mm)

Fig.6. Moisture Characteristic Curves of Three Types of Soil

Nos. 1, 2, and 3 shown in Fig.1., the composition and

distribution of the soil around experimental location
are thought to be almost the same in both the horizontal
and perpendicular directions. Fig.6. shows the moisture
characteristics of three types of soil measured by Jury
(1973) and the moisture characteristics measured in
the samples. Sample A was taken from the vicinity of
the ground level near the basement and sample B was
Fig.3. Sectional Plan of the Basement (Unit : mm)
taken from the vicinity of the ground level at a distance
from the basement. The measurement results suggest
that the experimental area is plain field sand.

4. Experimental Result
This chapter presents the results of measurements
of the basement and surrounding ground taken from
October 22, 2004 to October 31, 2005. The four-day
omission in the measurement data was the result of a
power failure from February 11 to February 14, 2005.
4.1 Outdoor climate
Figs.7. to 10. show the outdoor air temperature,
relative humidity, precipitation, and solar radiation
for the measurement period. Table 2. lists three sets
Fig.4. Measurement Points in the Basement
of figures for comparison: values measured at the

JAABE vol.5 no.2 November 2006 Kyung-Soon Park 423

site, values measured at a nearby meteorological
observatory, and values for an ordinary year in the
region. Data on solar radiation are not given, however,
as the meteorological observatory performed no
measurements of this condition.
4.2 Precipitation and underground water table
F i g . 11 . s h o w s t h e a n n u a l v a r i a t i o n s o f t h e
underground water table and precipitation. The
maximum precipitation and the lower depth of the
underground water table were 29.0mm/h (70.5mm/
day) and GL-373cm, respectively. The depth of the Fig.10. Annual Variation of Solar Radiation
annual mean underground water table was GL-445cm.
Table 2. Comparison of Values Measured at the Site, Values
The variation caused by rainfall was the only notable
Measured at a Nearby Meteorological Observatory, and Values
variation in the underground water table to result from for an Ordinary Year
factors such as periodic changes. On this basis, we
concluded that the fluctuation of the underground water
table had no direct repercussions on the basement floor.
4.3 Solar absorption ratio
The solar absorption ratio on the ground level was
calculated from the pyrheliometer and albedometer
measurements. Fig.12. shows the annual variation of
the solar absorption ratio on the ground level. The daily

Fig.7. Annual Variation of the Outdoor Air Temperature Fig.11. Annual Variation of the Underground Water Table

Fig.8. Annual Variation of the Relative Humidity

of the Outdoor Air Fig.12. Annual Variation of the Solar Absorption Ratio
solar absorption ratio during the measurement period
changed from 0.69 to 1.0, and the solar absorption ratio
rose markedly after rainfalls. These rises in the solar
absorption ratio were attributable to the decrease in the
solar reflection ratio by the wetting of the ground level
during the rainfalls. The annual mean solar absorption
ratio on the ground level was 0.84 (standard deviation
0.076).
4.4 Indoor air temperature and humidity
Fig.13. and Table 3. show the annual temperature
variation of each measurement point on the inside
Fig.9. Annual Variation of Precipitation

424 JAABE vol.5 no.2 November 2006 Kyung-Soon Park

wall of the basement (See Fig.4.). The influence of the trend was attributable to the differential between the
outside climate pushed the temperature amplitude on water vapor pressure of the outdoor air and saturated
the ceiling and the corner of the upper part of the wall vapor water pressure of the indoor air. The former
to levels higher than that on the other parts of the wall. suppressed the rise of the latter, and the moisture
The annual temperature amplitude on the corner of evaporation on the inner surface of the wall suppressed
the wall, the thermally weak part, was therefore large the fall of the latter when the former rose.
due to the relatively low annual average temperature. Fig.19. shows the variation of the saturated water
Moreover, no differences in azimuthal temperatures vapor pressure of the wall surface and the water vapor
were found in the central parts of the walls (N, S, E, W pressure of the indoor air. The condensation on the
points). inner surface of the basement wall could not be visually
Fig.14. shows the annual variation of the daily mean confirmed during the measurement period. However,
indoor air temperature of R1-R6 and analyzes the noting that the relative humidity of the indoor air was
distribution. The maximum temperature differentials higher than that of the outdoor air from June to August
among the six measurement points in the horizontal (see the dotted line in Fig.17.), we compared the water
and vertical directions were 1.5°C and 1.8°C during vapor pressure of the indoor air and the saturated water
the heating period and 1.0°C and 3.0°C during the non- vapor pressure on the inner surface of the wall in order
heating period, respectively. However, in light of the to analyze the aspects of condensation. We concluded,
short generation time and the low mean temperature based on this comparison, that condensation could
differential between the horizontal and vertical not have developed on the other walls where the
directions (1.0°C or less), the average value of the six saturated water vapor pressure exceeded the water
points was used as the indoor air temperature for the vapor pressure of the indoor air. We also judged,
subsequent analysis. however, that condensation was very likely to have
Fig.15. shows the daily variations of the indoor appeared for a period of about a month on the edge of
and outdoor air temperatures along with the electric the floor at the point where the behavior reversed. In
consumption on a typical day during the heating other words, when the indoor air relative humidity of
period. The variation of the indoor air temperature the basement was maintained at 90% or higher, most
and electric consumption by the electric hot-air heater of the condensation was probably generated on the
in Fig.15. was attributable to the on/off control of the edge of the floor. This compels us to consider moisture
thermostat. The range of this variation remained almost control or latent heat load control in order to maintain
constant throughout the entire heating period. a comfortable indoor environment.
Fig.16. shows the annual variation of the daily 4.5 Amount of electric consumption
mean indoor and outdoor air temperatures. The mean Fig.20. shows the monthly average outdoor
indoor air temperatures for the heating and non-heating air temperature and the monthly integral electric
periods were 21.7°C and 23.6°C, respectively, and consumption of the electric hot-air heater. The annual
the annual mean temperature was 22.2°C. On this electric consumption was 2082kWh, and the maximum
basis, we concluded that the experimental indoor air monthly and daily electric consumptions were
temperature was successfully controlled. 385.94kWh and 13.61kWh, respectively. All of these
Fig.17. shows the annual variation of the relative measurements were taken in January. In addition, the
humidity of the indoor and outdoor air. The annual annual mean heating load per unit area of the basement
mean relativity humidity of the indoor air was 51%, was about 58W/m 2 (including about 19 W/m 2 by
and the mean relative humidities for the heating ventilation). This is 1/4 of the heating load per unit
and non-heating periods were about 35% and 75%, area for the ground building which corresponded to
respectively. In the absence of any measures to control the basement. No cooling load (sensible heat load) was
moisture in this experiment, the relative humidity of generated throughout the year.
the indoor air fluctuated sharply between the heating 4.6 Heat and moisture behavior of the surrounding
and non-heating periods. ground
Fig.18. shows the annual variation of the water Fig.21. shows the annual variation of the daily
vapor pressure of the indoor and outdoor air. The mean ground temperature near the ground level (5cm
daily amplitude of the vapor pressure of the indoor in depth). The average temperature at the ground
air followed that of the outdoor air during the heating level, mean soil-air temperature, and mean outside air
period, but not during the non-heating period. This temperature were 17.3, 20.3, and 16.4°C, respectively,
Table 3. Annual Amplitudes and Annual Mean Temperatures of the Inside Wall of the Basement (Unit:°C)

JAABE vol.5 no.2 November 2006 Kyung-Soon Park 425

 100

80

Relative humidity [%]

60

40

20
Indoor
Outdoor
0
Oct Nov Dec Jan Feb Mar Apr May June July Aug Sep Oct
Fig.13. Annual Temperature Variation of Measurement Points
Fig.17. Annual Variation of the Relative Humidity of
on the Inside Wall of the Basement (Daily Average Values)
Indoor and Outdoor Air (Daily Average Values)

3000

2500

Water Vapor Pressure [Pa]

2000

1500

1000

500 Indoor
Outdoor
0
Fig.14. Annual Variation of Indoor air Temperature at Points Oct Nov Dec Jan Feb Mar Apr May June July Aug Sep Oct
R1 to R6 (Daily Average Values)
Fig.18. Annual Variation of the Water Vapor Pressure of
Indoor and Outdoor Air (Daily Average Values)

Fig.15. Variations of Electric Consumption and Indoor/Outdoor Fig.19. Variation of the Saturated Water Vapor Pressure of the
Air Temperatures (Daily Average Values) Wall Surface and the Water Vapor Pressure of the Indoor Air
(Daily Average Value) (See Fig.4.)
35

30

25
Temperature [ºC]

20

15

10

5 Indoor
Outdoor
0
Oct Nov Dec Jan Feb Mar Apr May June July Aug Sep Oct

Fig.16. Annual Variation of Indoor and Outdoor Air Fig.20. Annual Variation of Monthly Integral
Temperatures (Daily Average Values) Electric Consumption by the Electric Hot-air Heater
for the year. The temperature near the ground level
Figs.22. and 23 show the annual variation of the
fluctuated by about as much as the temperature of the
ground temperature at points A2 and A5 (See Fig.1.).
outdoor air.

426 JAABE vol.5 no.2 November 2006 Kyung-Soon Park

Table 4. lists the annual amplitudes and annual mean Table 4. Annual Amplitudes and Annual Mean Temperatures of
temperatures at every measurement point, measured the Surrounding Ground
at each depth. The daily amplitude of the outdoor air
ceased to have influence from depths of 0.5m onward.
Moreover, the amplitude fell and the delay in the phase
increased at progressively greater depths. As shown in
Figs.22. and 23. and Table 4., the annual amplitude of
the measurement point located farthest away from the
outside wall of the basement exceeded that of the point
nearest the basement, and the annual mean ground
temperature approached the annual mean outdoor
air temperature as the distance from the basement
increased. These findings resulted from the heat
generated internally within the basement, an influence
which increased temperatures outward to about 4.0m
from the outside wall of the basement.
Fig.24. shows the annual variation of the moisture
content. The values were converted from the

Fig.21. Annual Temperature Variation Near the Ground Level

and in Outdoor Air (Daily Average Values)

Fig.22. Annual Variation of the Ground Temperature at Point A2

(Daily Average Values)

Fig.24. Annual Variation of Moisture Content in the Ground

measurement values of the water chemical potential

in the ground using the moisture characteristic curve
of plain field sand, assuming that the equilibrium
relationship between the moisture content and water
chemical potential was not dependent on temperature.
The moisture content of the ground at all depths
rose during rainfalls and decreased after the rains. The
tendency was most conspicuous at the shallow depth
Fig.23. Annual Variation of Ground Temperature at Point A5 of 0.1m. Moreover, the variations of the moisture
(Daily Average Values) content at depths of 2 and 3m remained almost stable

JAABE vol.5 no.2 November 2006 Kyung-Soon Park 427

throughout the entire year. We thus confirmed that highly humid regions, measures for the control of the
the moisture content varied more drastically near the moisture or latent heat load need to be considered in
ground level, where the influences of precipitation and order to maintain a comfortable indoor environment.
solar radiation were substantial. As the depth from 5) The condensation on the inside wall surface of the
the ground level increased, however, the moisture basement could not be confirmed by visual observation.
content remained progressively more stable behavior We concluded, however, that condensation was very
throughout the year. likely to have collected on the edge of the floor of the
In view of these findings, the fluctuation band of basement in summer, when the relative humidity of the
moisture content in the ground is thought to be strongly indoor air remained at 90% or higher.
affected by precipitation but only weakly affected by 6) The influence of the internal heat generation
the depth of the underground water table. extended by about 4.0m from the outside wall of the
In the horizontal direction, however, we find that a basement in horizontal and vertical directions.
simple comparison of moisture content is inadequate, 7) Though the ground at the experimental site had
given that only point C differs in the measurement identical properties, we confirmed that the fluctuation
depth. In this case we observed a changing tendency of the moisture content was dependent on the depth
in the order of C, A2, and A4. On this basis, we from the ground level. In other words, the fluctuation
concluded that the moisture accumulated in the ground band of moisture content in the ground was strongly
of the upper part of the ceiling and flowed onto the affected by precipitation but only minimally affected
outside wall of basement. by the depth of the underground water table.
Experimental measurements are now underway
5. Conclusions for a second year to clarify the effects of the initial
We conducted a field experiments to evaluate conditions of the surrounding ground and internal
the heat and moisture behaviors of an underground heat generation. In ensuing research we will conduct
basement and surrounding ground under an internal nonlinear numerical analyses with simultaneous
heat generation condition, This paper has presented the heat and moisture transfer equations to predict the
details of the experiment and analyzed the results. The temperature and moisture behavior of the basement
main conclusions are as follows. and surrounding ground.
1) The outdoor air temperature and precipitation of
the measured year were higher and lower than an References
ordinary year, respectively. The annual mean solar 1) Matsumoto, M., (1978) Simultaneous Heat and Moisture Transfer,
and Moisture Accumulation in Buildings. Doctoral Dissertation,
absorption ratio on the ground level was 0.84 during
Kyoto National University, Japan.
the experimental period (standard deviation 0.076). 2) Matsumoto, M., (1984) Physics of Environment, Architectural
2) The lower depth of the underground water table Science Series, Vol. 10, Shokokusya, 1st edn., Japan.
was GL-373cm, and the annual mean depth was GL- 3) Matsumoto, M. and Iwamae, A., (1988) An Analysis of
445cm. Thus, the fluctuation of the underground water Temperature and Moisture Variations in the Ground Under Natural
Climatic Conditions. Energy and Buildings, 11, pp.221-237.
table was not found to have direct repercussions on
4) Matsumoto, M. and Kotera, N. (1988) An analysis of heat storage
the basement floor. Moreover, the variation caused in the wet soil solved as a coupled problem. Energy and Buildings,
by rainfall was the only notable variation in the 11, pp.239-247.
underground water table to result from factors such as 5) Ogura, D., Nagai, H., Matsushita, T. and Matsumoto, M., (1997)
periodic changes. An Analysis of Heat and Moisture Behavior in Underground Space
by Quasi-linearized Method. Proc. of CIB W40 meeting in Kyoto,
3) The mean indoor air temperatures for the heating Japan, pp.313-328.
and non-heating periods were 21.7°C and 23.6°C, 6) Hasegawa, F., Nagatomo, M., Yoshino, H., Matsumoto, S.
respectively, and the annual mean temperature was 22.2°C. and Sakanishi, T., (1992) Long-term Thermal Performance
On this basis, we concluded that the experimental indoor air Measurement of Semi-underground Test House with Space
temperature was successfully controlled. The annual Heating. Journal of Architecture, Planning and Environmental
Engineering, AIJ, No. 435, pp.1-10.
electric consumption of the electric hot-air heater 7) Mitalas, G.P. (1982) Basement heat loss studies at DBR/NRC,
was 2082kWh, and the maximum monthly and daily Ottawa. Division of Building Research, National Research Council
electric consumptions were 385.94kWh and 13.61kWh, of Canada, DBR Paper No. 1045.
respectively (all of these measurements were taken in 8) Mitalas, G.P. (1987) Calculation of Below-Grade Residential Heat
Loss. ASHRAE Trans., 93 (1), pp.743-783.
January). The annual mean heating load per unit area
9) Sobotka, P., Yashino, H. and Matsumoto, S., (1994) Thermal
of basement was about 58W/m2 (including about 19 performance of three deep basements: a comparison of
W/m2 by ventilation). No cooling load (sensible heat measurements with ASHRAE Fundamentals and the Mitalas
load) was generated throughout the year. method, the European Standard and the two-dimensional FEM
4) The annual mean relativity humidity of the indoor program. Energy and Buildings, 21 (1), pp.23-34.
10) Sobotka, P., Yashino, H. and Matsumoto, S., (1995) The Analysis
air was 51%. In the absence of measures to control
of Deep Basement Heat Loss by Measurements and Calculations.
moisture in this experiment, however, the relative ASHRAE Trans., 101 (2), pp.186-197
humidity of the indoor air fluctuated rather sharply 11) Jury, W., (1973) Simultaneous transport of heat and mass through
throughout the year. When basements are planned in a medium sand, Ph. thesis, University of Michigan.

428 JAABE vol.5 no.2 November 2006 Kyung-Soon Park