Proceedings of the ASME 2018 International Mechanical Engineering Congress and

Exposition
IMECE2018
November 9 - 15, 2018, Pittsburgh, PA, USA

IMECE2018-87738

THERMAL BEHAVIOR OF SOILS UNDER TIDAL EFFECT: A CASE STUDY IN

GUAYAQUIL, ECUADOR.

Daniel Moreira José Macı́as

Facultad de Ingenierı́a en Mecánica y Instituto Nacional de Eficiencia
Ciencias de la Produccı́on Energética y Energı́as Renovables, INER
Escuela Superior Politécnica del Litoral Quito, Pichincha, 170507, Ecuador
Guayaquil, Guayas, 090903, Ecuador [email protected]
[email protected]

Ruben Hidalgo-León Guillermo Soriano∗

Centro de Energı́as Renovables y Alternativas Centro de Energı́as Renovables y Alternativas
Facultad de Ingenierı́a en Mecánica y Facultad de Ingenierı́a en Mecánica y
Ciencias de la Produccı́on Ciencias de la Produccı́on
Escuela Superior Politécnica del Litoral Escuela Superior Politécnica del Litoral
Guayaquil, Guayas, 090903, Ecuador Guayaquil, Guayas, 090903, Ecuador
[email protected] [email protected]

ABSTRACT NOMENCLATURE
This research analyzes the tidal effect in the thermal prop- TRT Thermal response test.
erties of the ground for a case study in Guayaquil, Ecuador. GHE Ground heat exchanger.
A thermal response test (TRT) performed near the shore of the BTESS Borehole thermal energy storage system.
Guayas river presented periodic fluctuations in the thermal be- INOCAR Oceanographic Institute of the Ecuadorian Navy.
havior concurrent with the tide cycle. First, an analytical solu- MLWS Mean low water springs
tion for tide-induced water table fluctuations was used for the HVAC Heating, ventilation, and air conditioning
determination of the phreatic level for the days of the test. The RMSE Root mean square
analytical model accounted for the horizontal distance from the HDPE High-Density Polyethylene
shore, the ground porosity, and permeability. Afterward, a geo-
metric mean model was used to predict the thermal conductivity
of soil considering the groundwater level fluctuations. Finally, a INTRODUCTION
correlation between the effective thermal capacity of the ground The use of borehole heat exchanger (BHE) is an option to
and the phreatic level in the soil was found. typical HVAC systems in buildings that is sustainable and with
low environmental impact. This type of technology is suitable
for commercial and residential buildings [1] and feasible in cities
with tropical climate where the use of air conditioning always is
∗ Escuela Superior Politécnica del Litoral ESPOL, Ecuador quite large. In the design of a BHE system is key to determine

1 Copyright
c 2018 by ASME
certain ground thermal properties such as ground temperature,
thermal diffusivity, soil thermal conductivity, and BHE thermal
resistance [2]. These measurements must be made precisely be-
cause errors can lead sizing errors of the BHE with possible con-
sequences in its capital cost and system performance. To deter-
mine these properties, the most commonly used method is the
thermal response test (TRT).
In Guayaquil (Ecuador), ten boreholes measurements at ten
representative locations were performed to study the feasibility
of replacing cooling towers with BHE systems [3]. The depths
of the boreholes were between 40 m and 60 m. The results ob-
tained from the TRTs showed significant differences in the ther- FIGURE 1. BOREHOLE ANALYZED TEMPERATURE RE-
mal properties of the soil between the two areas under study SPONSE AT q0 =2.61 kW
(Deltaic-Estuarine zone and Cayo formation). A variation of the
TRT was proposed in [4] where the authors presented a method
for extending these tests to other BHEs at the same site.
In boreholes, ignoring the presence of groundwater flows
can affect the measurements in TRTs. Nguyen et al. [5] demon-
strated that the groundwater flows considerably reduces the per-
formance of BTESS where the effect of this fluid is much greater
than seasonal variation of ground surface temperature . In Boad-
ing, China, a BHE thermal experiment was carried out under
the effect of groundwater flow where BHE performance was im-
proved with groundwater flow [6]. Hu et al. [7] presented an
improved analytical model which considered the effect of inho-
mogeneous groundwater flow in multiple-layer geologies. This
model was experimentally verified and the error on the outlet FIGURE 2. SCHEMATIC DIAGRAM OF TIDAL INDUCED WA-
fluid temperature of GHE was less than 0.3◦ C. In a study by TER TABLE FLUCTUATIONS
Zhang et al. [8], the thermal conductivity in a 100 m borehole
(in Tangshan locality) was measured by taking into considera-
tion variations in groundwater flows. Here, the authors deter- equal at the shore, we described this by:
mined that the acceleration of groundwater flow is beneficial in
raising the heat exchange rate. In the measurements presented h(x0 ,t) = D + Acosωt (1)
by Soriano et al. [3], the authors found periodic fluctuations in
the thermal properties of the ground concurrent with the tide as
shown in Figure 1. The borehole is located at 5 m of the Guayas where A is maximum amplitude and ω is wave frequency.
river shore with the following site coordinates (UTM-WGS84) This model assumed a homogeneous, isothermal and incom-
and depth: X=624152.13, Y=9755373.32 and 50 m. pressible flow in a rigid porous medium. The fluctuations of
water table decay exponentially inland. Nielsen [10] presented
The present paper analyzes these fluctuations in the thermal
an analytical model for tidal motions of the water table, in this
properties of the ground and an analytical model is proposed and
research the author defined the length scale for significant varia-
compared with experimental results.
tions in x direction as:

METHODOLOGY s
2KD
Tidal-induced water table fluctuations model L= (2)
ηω
The configuration of tidal induced water table fluctuation
is shown in Figure 2 [9]. In the figure h(x,t) is the water table
height, D is the inland water table height. The tidal height varies where K is the hydraulic conductivity and η is the soil porosity.
periodically between high and low tide (HT - LT). This tidal con- These constants define soil properties.
figuration, forced a groundwater flow resulting in a varying water In this analysis, we used an analytical solution for tide-
table height. Since the elevation of water table and tide level are induced-water table fluctuations in a sloping beach developed by

2 Copyright
c 2018 by ASME
Teo et al. [11]. Their research was based in one-dimensional The Equation 5 can be written as a linear function of the
models presented previously by Nielsen [10] and Li et al [9]., in- logarithm of time.
cluding higher order components and redefining two perturbation
parameters, the shallow water parameter (ε) and the amplitude
parameter (a). ∆T = mln(t) + b (6)
The shallow water parameter ε=D/L, represents the ratio of
the water table height to the linear decay length and it depends
where ∆T = T̄ f (t) − To is the temperature difference between
interely on soil properties:
fluid and undisturbed soil temperature; m=0.25q0 π −1 λe−1
f f is the
slope of linear function; and b represents the intercept, which
r contains the constant parameters of the model. The slope allows
ηωD to determine the effective thermal conductivity of soils, and it
ε= (3)
2K can be found through a linear regression between the tempera-
ture difference and logarithmic scale time. Thermal diffusivity
The amplitude parameter a=A/D represents the ratio of the can be calculated by using the intercept and the slope of the lin-
tidal amplitude to the average height of the water table. Teo et ear regression.
al. [11] (for the special case of a vertical beach β = π/2) shows
a solution which can be written as:  
R f +R p
 
b−q0 2
−γ 
 
m
r2

   1h √
−X 2 1
H(X, T ) = ae cosθ1 + a 1 − e−2X + e− 2X αe f f = e (7)
4 2 4
i √2
cosθ2 − e−2X cos2θ1 − ε 2 aX1 e−X The heat capacity can be determined by the definition of
3

1

1
 
X
 thermal diffusivity:
cos θ1 − π + ε 2 a2 −1 + 1 + (4)
4 3 2
√ √
 π 
e−2X − 2Xe− 2X cos θ2 − + e− 2X sinθ2 + λe f f
4 Ce f f = (8)
√

 π  −2X αe f f
2Xe−2X cos 2θ1 − −e sin2θ1
4
In this research we proposed a simple method without de-
where X=x/L,
p Z=z/D, H=h/D, T=ωt are non-dimensional √
vari-
parting from the original idea of the slope method and without
ables; L = 2KD/(ηω), ε = D/L, θ1 = T − X, θ2 = 2T − 2X
delve into details of the groundwater flow.
and θ3 = 2T − X are transformation variables.
In this approach, the soil can be considered as a compound
medium. Its structure is porous composed by minerals grains,
Thermal properties models and a space filled with water and air in different volumetric pro-
The previous analysis of borehole response test used the portions. The effective thermal properties can be expressed as
slope technique [3]. The average fluid temperature in the re- function of each components properties.
sponse test can be expressed as: In the case of thermal conductivity, a large number of models
can be found in literature, and there is not an universally accepted
procedure for its determination, but the upper and lower limits of
its possible values can be determinate through the series (9) and
q0 q0
  
4α
T̄ f (t) = ln(t) + ln −γ parallel (10) models [13–17].
4πλe f f 4πλe f f r2
  (5)
R f + Rp
+q0 Rb + + To
2 1 N 
φi

=∑ (9)
λe f f i λi
where q0 is the linear heat transfer rate; λe f f is the effective ther-
mal conductivity of soil; α is the thermal diffusivity of soil; Rb ,
R f , R p are thermal resistances of borehole, fluid and pipe respec-
N
tively; To is the undisturbed soil temperature. This model is valid λe f f = ∑ (φi λi ) (10)
for values of αtr−2 ≥5 [12]. i

3 Copyright
c 2018 by ASME
where φi is the volumetric proportion; and λi is the thermal con-
ductivity for each component. Since the large difference between
the limits, we used the geometric mean law for predicting the ef-
fective thermal conductivity [18].

N  
φ
λe f f = ∏ λi i (11)
i

For a saturated soil, it can be expressed as:

η 1−η
λe f f = λwater λsoil (12)

where η is soil porosity; λwater =0.56 Wm−1 K−1 ; and λsoil =2.5
Wm−1 K−1 (for silt soil) which is determined by an analysis of
soil mineralogy after a borehole perforation [15]. Since, the
borehole analyzed is very near at the shore, the consideration of
effective thermal conductivity was also carried out on the grout
of the well.

Estimation of tidal effect on thermal capacity

We determined thermal diffusivity values through Equation
(12) using the data of temperatures collected form borehole re-
sponse test. The effective heat capacity was calculated using
Equation (8). The procedure followed is shown in Figure 3.
Water table height was estimated through Equation (4) using
the data of tidal elevation of Guayas river for the same days when
was conducted the borehole response test [19]. In order to know
the influence of tidal variations in effective heat capacity, a linear
regression was carried with water table height and the estimated
effective heat capacity. FIGURE 3. PROCEDURE FOR ESTIMATING SOIL THERMAL
For the estimation of error, we used RMSE approximation PROPERTIES
between the difference of temperature calculated by the model
and the data obtained by the TRT.
zone. We identify a periodic fluctuation of this temperature after
test, as shown in Figure 1.
" #0.5 The test lasted 48h, starting on March 24, 2014; in Table 1
1 n we listed tidal level for those days. In this way, the tide level
RMSE = ∑ (∆Tmodelo − ∆TT RT )2
n i=1
(13) is superimposed with the TRT data. The temperature difference
fluctuations have a very similar period as tidal wave of Guayas
river, being that tidal wave crest coincides with the temperature
difference fluctuations valley. This can be appreciated in Figure
RESULTS AND DISCUSSION
4.
The tidal-induced Guayas river level variation influence on
the measured temperature in a borehole response test near at the
river shore are presented where the distance to the shore was 5 Slope technique without tide effect consideration
m. The Temperature difference between average water tempera- The slope method was applied at the fluid-soil temperature
ture and undisturbed soil temperature was register by a TRT in a difference data of TRT, as shown in Figure 5. Using Table 2, the
borehole near at the shore of Guayas river in a Deltaic-Estuarine borehole parameters and the method explained in previous sec-

4 Copyright
c 2018 by ASME
 TABLE 1. TIDES IN GUAYAS RIVER, [19] TABLE 2. BOREHOLE PHYSICAL PARAMETERS, [3]

24/03/2014 25/03/2014 26/03/2014

Monday Tuesday Wednesday Parameters Value
Hour Hour Hour Well depth 50 m
MLWS MLWS MLWS
hh:mm hh:mm hh:mm Undisturbed soil temperature 29o C
0:09 4.49 P 1:20 4.31 P 2:35 4.23 P Ground thermal conductivity 1.73 Wm−1 K−1
7:09 0.23 B 8:12 0.28 B 9:32 0.31 B Pipe (HDPE)
12:58 4.42 P 14:07 4.37 P 15:19 4.40 P Pipe thermal conductivity 0.5 Wm−1 K−1
19:31 0.35 B 20:48 0.40 B 22:18 0.39 B Well diameter 113 mm
Internal diameter 19.5 mm
External diameter 25.4 mm

FIGURE 4. INFLUENCE OF THE TIDE ON TRT

FIGURE 6. VALIDATION OF LINEAR MODEL RESULTS VER-

SUS EXPERIMENTAL RESULTS

Model proposed including tidal effect

Considering the values listed in Table 3 and following the
procedure set in Figure 3, we estimated the effective thermal
properties of soil. In this way, a strong correlation was found
between effective heat capacity and water table height values as
shown in Figure 7.
For the effective thermal conductivity and effective ther-
mal diffusivity, we obtained values of 1.13 Wm−1 K−1 and
7.2x10−7 m2 s−1 respectively. These values are in the range of
FIGURE 5. APPLICATION OF THE SLOPE METHOD IN THE
previous reported values for this soil type in water saturated con-
LOGARITHMIC SCALE OF THE FLUID-SOIL TEMPERATURE
ditions [13].
DIFFERENCE RECORDED IN THE TRT
This method had a RMSE approximation of 0.33o C between
model predicted fluid-soil difference temperature and TRT data.
tion, we obtained a thermal conductivity and thermal diffusivity The correlation model allows better fitting with the experimental
of λe f f = 1.23 Wm−1 K−1 and α = 0.03 m2 s−1 respectively. An values as shown in Figure 8.
approximation was performed between the model and TRT fluid-
soil temperature difference data, as shown in Figure 6, obtained
an RMSE of 3.35o C. It is easy to appreciate that both data has SUMMARY AND CONCLUSIONS
the same trend, but they are displaced vertically, which directly The effect of tidal-induced water table variations on TRT of
affects the determination of thermal diffusivity. a BHE near at Guayas river shore was evaluated. The TRT re-

5 Copyright
c 2018 by ASME
TABLE 3. PHYSICAL PARAMETERS USED IN ANALYTICAL
WATER TABLE HEIGHT MODEL

Parameter Value
Average tide level(MLWS) 2.33 m
Permeability 5 m day−1
Tide amplitude 3.86 m−1
Wave frequency 4 π day−1
FIGURE 8. VALIDATION OF THE MODEL, CONSIDERING THE
Soil porosity 0.55 TIDAL EFFECT
Soil type Silty
fer potential in the soil analyzed, therefore, a BHE selected with-
out considering the effect of tide in this zone, could be under-
sized. Future research will be focused on developing this topic.

ACKNOWLEDGMENT
The authors of the present work to thank to Instituto Na-
cional de Eficiencia Energética y Energı́as Renovables (INER)
of Ecuador for providing the necessary data for validate the pro-
posal model.

REFERENCES
[1] Chen, J., Xia, L., Li, B., and Mmereki, D., 2015. “Sim-
FIGURE 7. LINEAR REGRESSION BETWEEN SOIL HEAT CA- ulation and experimental analysis of optimal buried depth
PACITY AND WATER TABLE HEIGHT of the vertical u-tube ground heat exchanger for a ground-
coupled heat pump system”. Renewable Energy, 73,
pp. 46–54.
sults were analyzed using two strategies, the slope technique and [2] Jensen-Page, L., Narsilio, G. A., Bidarmaghz, A., and John-
a new proposed method including Guayas river tidal-effect. The ston, I. W., 2017. “Investigation of the effect of seasonal
proposed method adjusts better to experimental data and main- variation in ground temperature on thermal response tests”.
tains the simplicity of the slope technique without characterizing Renewable Energy.
the underground flow. [3] Soriano, G., Espinoza, T., Villanueva, R., Gonzalez, I.,
In the proposed method, the heat capacity is calculated using Montero, A., Cornejo, M., and Lopez, K., 2017. “Ther-
the water table instantaneous height as function of porosity, soil mal geological model of the city of guayaquil, ecuador”.
permeability, soil mineralogy. In the parallel form, thermal con- Geothermics, 66, pp. 101–109.
ductivity is predicted using a geometric mean model for porous [4] Raymond, J., Lamarche, L., and Malo, M., 2016. “Extend-
medium. After this, heat capacity and thermal diffusivity are es- ing thermal response test assessments with inverse numer-
timated using the proposed model. ical modeling of temperature profiles measured in ground
The temperature difference between average fluid temper- heat exchangers”. Renewable Energy, 99, pp. 614–621.
ature and the undisturbed soil temperature was calculated and [5] Nguyen, A., Pasquier, P., and Marcotte, D., 2017. “Bore-
compared with experimental data. This methodology showed hole thermal energy storage systems under the influence of
a RMSE of 0.13o C which is superior to the RMSE of 3.35o C groundwater flow and time-varying surface temperature”.
achieved by the slope technique. Geothermics, 66, pp. 110–118.
Finally, the values of thermal conductivity and thermal diffu- [6] Wang, H., Qi, C., Du, H., and Gu, J., 2009. “Thermal per-
sivity obtained from proposed method suggest a lower heat trans- formance of borehole heat exchanger under groundwater

6 Copyright
c 2018 by ASME
 flow: a case study from baoding”. Energy and Buildings,
41(12), pp. 1368–1373.
[7] Hu, J., 2017. “An improved analytical model for ver-
tical borehole ground heat exchanger with multiple-layer
substrates and groundwater flow”. Applied Energy, 202,
pp. 537–549.
[8] Zhang, Y., Zhang, J., Yu, Z., Guo, L., and Hao, S., 2016.
“Analysis of the influence of different groundwater flow
conditions on the thermal response test in tangshan”. Envi-
ronmental Earth Sciences, 75(22), p. 1444.
[9] Li, L., Barry, D., Stagnitti, F., and Parlange, J.-Y., 2000.
“Groundwater waves in a coastal aquifer: A new governing
equation including vertical effects and capillarity”. Water
Resources Research, 36(2), pp. 411–420.
[10] Nielsen, P., 1990. “Tidal dynamics of the water table in
beaches”. Water Resources Research, 26(9), pp. 2127–
2134.
[11] Teo, H., Jeng, D., Seymour, B., Barry, D., and Li, L., 2003.
“A new analytical solution for water table fluctuations in
coastal aquifers with sloping beaches”. Advances in Water
Resources, 26(12), pp. 1239–1247.
[12] Ingersol, L., Zobel, O., and Ingersol, A., 1948. “Heat
conduction, with engineering and geological applications”.
MacGraw-Hill, New York.
[13] Banjac, M. J., Todorović, M. N., Ristanović, M. R., and
Galić, R. D., 2012. “Experimental determination of thermal
conductivity of soil with a thermal response test”. Thermal
Science, 16(4), pp. 1117–1126.
[14] Chen, S. X., 2008. “Thermal conductivity of sands”. Heat
and mass transfer, 44(10), p. 1241.
[15] Dong, Y., McCartney, J. S., and Lu, N., 2015. “Critical re-
view of thermal conductivity models for unsaturated soils”.
Geotechnical and Geological Engineering, 33(2), pp. 207–
221.
[16] Lovell, M., 1985. “Thermal conductivities of marine sed-
iments”. Quarterly Journal of Engineering Geology and
Hydrogeology, 18(4), pp. 437–441.
[17] Johansen, O., 1977. “Thermal conductivity of soils, cold
regions research and engineering laboratory report 637”.
Hanover, NH: US Army Corps of Engineers.
[18] Woodside, W., and Messmer, J., 1961. “Thermal conduc-
tivity of porous media. i. unconsolidated sands”. Journal of
applied physics, 32(9), pp. 1688–1699.
[19] INOCAR, 2018. Instituto oceanográfico de la
armada - tabla de mareas puertos del ecuador.
http://www.inocar.mil.ec/web/index.php/productos/tabla-
mareas. [Accedido: 08-dic-2017].

7 Copyright
c 2018 by ASME