Available online at www.sciencedirect.

com

ScienceDirect
Energy Procedia 105 (2017) 244 – 249

The 8th International Conference on Applied Energy – ICAE2016

Geothermal energy extraction from abandoned wells

Marius Røksland, Tommy A. Basmoen, Dan Sui*
Petroleum Engineering Department, University of Stavanger, Stavanger 4036, Norway

Abstract

Below the surface of the earth, there exist geothermal resources with the potential to make a significant contribution to
the global demand for energy in an environmental friendly way. Seemingly an endless resource, it could be capable of
replacing other sources of energy if treated wisely. A main challenge with the industry is related to the capital expensive
costs of drilling the geothermal wells, hence the introduction of abandoned petroleum wells is of interest. In this paper,
we seek to investigate the potential amounts of heat extracted from beneath the surface, by retrofitting a double pipe
heat exchanger in an abandoned petroleum well. The working fluid of choice is proposed to be circulated down through
the annulus, then up through an inner, insulated geostring. By making use of a numerical simulator, we asses parameters
of interest and obtain knowledge on how they affect the outlet temperature of the circulating working fluids and the
accumulated heat extracted from the geothermal wells.
© 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
© 2016 The Authors. Published by Elsevier Ltd.
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Selection and/or peer-review under responsibility of ICAE
Peer-review under responsibility of the scientific committee of the 8th International Conference on Applied Energy.

Keywords: Geothermal energy; Heat transfer; Sensitivity analysis

1. Introduction

The world’s demand for energy is bound to increase. In order to supply the growing population with the
required comfort, environmental friendly and renewable initiatives should be considered above non-
renewable energy sources. The CO2 content in the atmosphere is set to rise, hence the power generation
capacity installed from renewable resources should continue to represent the majority of additions [1]. An
energy source well worth considering in a process towards a more sustainable and renewable energy
production is geothermal, which depends on heat extraction from the subsurface. Geothermal heat has
enormous energy potential, hence reliable and steady output is usually expected. The earth’s capability to
replace the lost heat makes the resource renewable.

* Corresponding author. Tel.: 0047-46696188.

E-mail address: dan.sui@uis.no.

1876-6102 © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the scientific committee of the 8th International Conference on Applied Energy.
doi:10.1016/j.egypro.2017.03.309
 Marius Røksland et al. / Energy Procedia 105 (2017) 244 – 249 245

Three types of power plants are recognized in commercial geothermal electricity production, making use
of steam turbines and generators. Dry Steam-power plants operate high enthalpy resources, where the
temperature is sufficient to vaporize the extracted fluid. A flash-steam power plant vaporizes extracted fluid
in a flash-tank, before directing it through the turbine. For the low-enthalpy resources, a Binary cycle power
plant circulates a secondary fluid with low boiling point in a closed loop, which vaporizes after heat-
exchanging with the extracted fluid. The potential for utilizing a binary-cycle design today is high, as the
low- to intermediate enthalpy resources are most abundant. The possibility of utilizing geothermal for
heating purposes is an alternative exploitation. Hot fluid recovered from the ground can be piped several
kilometers to supply buildings with required heating. Other alternative application areas for the geothermal
heat is de-icing of roads, agricultural drying, heating for greenhouses, dry out fish and crops, etc.
The novel idea we investigate in this paper, makes use of abandoned petroleum wells for geothermal
purposes. With the drilling process already accomplished, the geothermal project can potentially shrink its
costs with 50% [2]. When retrofitting an already designed wellbore, the casings and depth are set. By using
a modified model and a developed simulator, an investigation of a double pipe heat exchanger retrofitted
to an abandoned petroleum well, is carried out. By heat-exchanging, the circulating working fluid will
gather heat-energy going downwards then, bring it up through an insulated geostring. The presented heat
transfer model takes many parameters into consideration, like formation properties, working fluid
properties, wellbore architecture, etc. The sensitivity analysis carried out by the simulator provides valuable
knowledge on construction design, fluid properties, geological conditions, etc. which affects the extracted
heat and temperature.

2. Properties of working fluids

2.1. Specific heat capacity Cp, thermal conductivity k and Viscosity μ

The specific heat capacity is defined as measurable physical quantity, related to the ratio of heat removed
from or added to a system, to the resulting temperature change. It is difficult to estimate or predict for
different temperatures. The thermal conductivity is a materials ability to conduct heat. Higher values of k
implies better heat conduction across a material, hence these materials are often used for heatsink- and heat
exchanger purposes. The thermal conductivity of a material may vary with temperature and convection. In
the case of high temperatures in fluids, convection is often present and this results in an increase in thermal
conductivity. The thermal conductivity in a geothermal well influences the conduction of heat in radial
direction. Dynamic viscosity is a fluid property of interest, as we examine heat transfer mechanisms in the
geothermal well. It is a fluid property of great importance to several industries, but here it will have an
impact on the heat transfer mechanisms taking place in the circulation process. [3].

2.2. Nusselt number,Prandtl number and Reynolds number

This equation introduces a correlation between the convective heat transfer coefficient and the Nusselt
number. It is commonly defined as the ratio between occurring convection and conduction across a
boundary [4, p. 787]:

݄‫ܦ‬
ܰ௨ ൌ ሺͳሻ
݇

where h is the convective heat transfer coefficient, D is the equivalent diameter of the pipe considered.
Prandtl number approximates a number for the ratio of momentum diffusivity to thermal diffusivity [5]:
246 Marius Røksland et al. / Energy Procedia 105 (2017) 244 – 249

ߤ‫ܥ‬௣
ܲ௥ ൌ ሺʹሻ
݇

Reynolds number Re describe different flow regimes occurring in a flowing medium, given as [4]:

ߩ‫ܦݒ‬
ܴ௘ ൌ ሺ͵ሻ
ߤ

The different phases are distinguished in terms of the Re-value, as follows: [Re < 2300] for Laminar flow,
[2300 < Re < 4000] for Transient phase and [Re > 4000] is for Turbulent flow.

3. Model description
3.1. Regions

To be able to properly estimate outlet temperatures of

the extracted fluids circulating the wellbore, it is highly
convenient to consider separated working regions. As
the fluid would experience different influences during
a full circulation, it is important to account for which
influences are of importance in different scenarios, like
when flowing downwards contra upwards. In turn, each
part will influence the wellbore temperature, fluid
temperature, casing temperature, etc. In our case, we
divide the well- bore into five regions: the drill pipe Figure 1. Schematic of a geothermal wellbore
(geostring), the drill pipe wall (geostring wall), the annular region, the interface between the annulus and
the formation and the formation itself.

3.2. Mathematical model

To be able to estimate temperatures at any given time and depth of the wellbore, an energy equation which
accounts for conservation of energy is formulated in cylindrical coordinates, presented in more detail [6]:

߲ܶ ߲ܶ ߲ܶ ͳ ߲ሺ‫ݍݎ‬௥ ሻ ߲‫ݍ‬௭
ߩ‫ܥ‬௣ ൬ ൅ ‫ݒ‬௥ ൅ ‫ݒ‬௭ ൰ ൌ െ ቆ ൅ ቇሺͶሻ
߲‫ݐ‬ ߲‫ݎ‬ ߲‫ݖ‬ ‫ݎ߲ ݎ‬ ߲‫ݖ‬

where T is temperature, r is the radius, z is the depth, vr is the linear velocity in radial direction, vz is the
linear velocity in axial direction, qr is the heat flux in radial direction and qz is the heat flux in the axial
direction. By making use of the heat flux definitions for heat transfer occurring through several surfaces
per unit time area, we can in turn obtain the following equation that addresses the problem:

߲ܶ ߲ܶ ߲ܶ ݇ ߲ܶ ߲ଶܶ ߲ଶܶ
ߩ‫ܥ‬௣ ൬ ൅ ‫ݒ‬௥ ൅ ‫ݒ‬௭ ൰ ൌ ൅ ݇ ଶ ൅ ݇ ଶ ሺͷሻ
߲‫ݐ‬ ߲‫ݎ‬ ߲‫ݖ‬ ‫ݎ߲ ݎ‬ ߲‫ݎ‬ ߲‫ݖ‬

A continuity equation for incompressible fluid is presented below for the actual scenario:
 Marius Røksland et al. / Energy Procedia 105 (2017) 244 – 249 247

ͳ ߲ሺ‫ݒݎ‬௥ ሻ ߲‫ݒ‬௭
൅ ൌ Ͳሺ͸ሻ
‫ݎ߲ ݎ‬ ߲‫ݖ‬

The Equations (5) and (6) are together with the boundary-and initial conditions presented below, the main
equations and correlations on which the model operates.

3.3. Boundary and initial conditions

To limit the possible solutions to the relevant differential equations, one needs to consider boundary
conditions and initial conditions. Being a set of conditions specified for the behavior of the solution at the
boundary of its domain, they are important for our model in determination of the solution. The boundary
and initial conditions are given in [5].

3.4. Overall heat transfer coefficient

By taking into consideration conditions appearing on both sides of a heat-transferring material, U calculates
the overall heat transfer through the material. We consider the overall heat transfer with respect to the heat
transfer direction of interest, which is calculated as follows [7].

‫ݎ‬௢ ‫ݎ‬௢ ‫ݎ‬௢ ͳ ିଵ

ܷ௢ ൌ ൬ ݈݊ ൬ ൰ ൅ ൅ ൰  ሺ͹ሻ
݇ ‫ݎ‬௜ ݄௜ ‫ݎ‬௜ ݄௢

‫ݎ‬௜ ‫ݎ‬௢ ‫ݎ‬௜ ͳ ିଵ

ܷ௜ ൌ ൬ ݈݊ ൬ ൰ ൅ ൅ ൰  ሺͺሻ
݇ ‫ݎ‬௜ ݄௢ ‫ݎ‬௢ ݄௜

where Uo, ro, ho are based on the outer region and Ui, ri, hi are based on the neighboring inner region.

3.5. Determination of heat transfer coefficient h

The Reynolds number Re is of great importance to the model. The different phases of the circulation process
are associated with the Reynolds numbers in terms of the present diameters and accompanying velocities.

Conditions in the geostring Theoretical study regarding this laminar phenomenon by [8], shows that the
Nusselt number Nu might be treated as a constant (Nu = 4.36). For turbulent phenomena, different
equations have been suggested to achieve optimal accuracy, as the turbulent conditions may vary, see [8]
and [9] for more detailed discussions.

Conditions in the annulus Convection heat transfer appears both on the inner and outer tube surfaces.
Regarding laminar conditions, the heat transfer coefficients related to the different surfaces are separately
calculated utilizing the Nusselt number-correlation, given by (1). Regarding turbulent conditions, this case
would be dependent on both Re and Pr , and calculated through the Dittus and Boelter equation [9]:

ସ ଵ ߤ ଴ǡଵସ
ܰ௨ ൌ ͲǡͲʹ͹ܴ௘ ହ ܲ௥ ଷ ൬ ൰ ሺͻሻ
ߤ௦

where μs is surface viscosity.

248 Marius Røksland et al. / Energy Procedia 105 (2017) 244 – 249

3.6. Heat energy equation

In order to estimate the power continuously extracted as heat from the well, plotted as a blue graph in
Figures 2-4, the following equation is used:

ܳ ൌ ݉‫ܥ‬௣ ȟܶ ൌ ݉‫ܥ‬௣ ሺܶ௢௨௧ െ ܶ௜௡ ሻሺͳͲሻ

where m is the mass and ∆T is the temperature difference between extracted fluid and injected fluid.

4. Simulations and sensitivity analysis

The numerical simulator calculates the temperatures extracted from the geothermal wellbore by solving
finite difference equations according to Equations (5)-(6). Due to the limited space, the detailed input data
is given in Table 5.1 in [11]. The simulation time for each simulation of the fluid were 300000s, enough to
complete a full circulation with a good margin.

Density of fluid For the density of the fluid, we

can see that it has a very little effect on the outlet
temperature, according to the model. The trend is
consequent towards lighter fluid exhibiting a
slightly higher outlet temperature, as presented in
Figure 2. For example, considering a density
variation between 800 kg/m3 and 1600 kg/m3, it is
observed from the plot that it only represents a
difference of 1.96°C. The difference in density is
100% between 800 kg/m3 and 1600 kg/m3, this Figure 2. Temperature and heat with respect to different densities
variation causes a decrease in outlet temperature of
2.48%, after completion of more than a full circulation in the double pipe heat exchanger. The heat plotted
as a blue graph indicates that the heat energy gained also decreases with a close to similar trend as the
temperature, when increasing the density.

Thermal conductivity of fluid The simulator illustrates that an increase of the thermal conductivity
increases the heat exchange between the working fluid and the formation down through the outer pipe. The
fluid obviously gains more heat, which is indicated by the blue graph. The outlet temperature after one
circulation also increases as expected. Seemingly, the rate at which heat is obtained decreases for higher
thermal conductivities, compared to the rate of increased temperature. By considering the increase between
0.3 W/m°C to 5.0 W/m°C, an increase in the outlet temperature is from 63.02°C to 87.38°C. This is an
important parameter to account for when enhancing the performance.

Specific heat capacity of fluid Specific heat capacity is another critical parameter which influences the
heat transfer efficiency. The red graph in Figure 4 shows decreasing outlet temperatures for increased
specific heat capacities. It is obvious that for lower heat capacities a more desirable outcome, in terms of
higher outlet temperature, is achieved. However, there exists an opposite trend for the heat energy gained,
since Cp is proportional to heat. Therefore, considerations of Cp is important when making decisions on
the working fluid used. Considering the red graph which represents the outlet temperature, an increase of
267%, from 1500 J/kg°C to 5500 J/kg°C, leads to a decrease of 33% in the outlet temperature.
 Marius Røksland et al. / Energy Procedia 105 (2017) 244 – 249 249

Figure 3. Temperature and heat w.r.t. thermal conductivities Figure 4. Temperature and heat w.r.t. specific heat capacities

5. Conclusion
The economical aspect of a geothermal energy project would benefit from utilization of an abandoned
petroleum well. The enclosed sensitivity-plots reflects the importance of several parameters relevant to the
project, and should be closely evaluated with respect to the geothermal area of application. For the
circulation of water as working fluid in a 4500m deep well, temperatures in the range of 70 – 80 °C can be
expected, which in turn could operate a binary cycle power plant. An enhanced working fluid would gather
more heat, and generate more desired amounts of electricity if needed. Economical, environmental and
realistic aspects from the proposed introduction of enhanced fluids should be evaluated against the potential
return from the project.

Acknowledgement
Authors would like to thank for financial support through the Plogen program.
References

[1] Global trends in renewable energy investment 2016, Tech. rep., Frankfurt School of Finance & Management (2016).
[2] H. L. Xianbiao Bu, Weibin Ma, Geothermal energy production utilizing abandoned oil and gas wells (Renewable Energy 41,
80-85 2012).
[3] S. Sudarsanam, Why the viscosity of liquids decrease with increasing temperature while that of gases increases with increasing
temperature?
[4] R. H. T. Yunus A. Cengel, John M. Cimbala, Fundamentals of Thermal-Fluid Sciences, fourth edition in si units Edition, McGraw-
Hill Education, 2 Penn Plaza, New York, NY 10121, 2012, iSBN: 978-007-132511-0.
[5] E. R. Santoyo-Gutie ́rrez, Transient numerical simulation of heat transfer processes during drilling of geothermal wells, Ph.D.
thesis, University of Salford, Salford, UK
[6] E. R. Santoyo, Transient numerical simulation of heat transfer processes during drilling of geothermal wells, Ph.D. thesis.
[7] C. L. . N. Sayma, Heat Transfer, 1st Edition, Vol. volume, Ventus Publishing ApS, 2009, iSBN: 978-87-7681-432-8.
[8] F. Incropera, D. Dewitt, Introduction to heat transfer, Tech. rep., New York, USA (1990).
[9] B. Petukhov, Advances in heat transfer, T.F. Irvine and J.P. Hartnett Vol. 6.
[10] W. K. Sons, H. Perkins, in Handbook of heat transfer, Chapter 7, W.M. Roshenow and J.P. Harnett (1972). ͒
[11] M. Røksland, T. A. Basmoen, Geothermal energy extraction from abandoned petroleum wells, University of Stavange

Biography
Dan Sui is Associated Professor at Department of Petroleum Engineering, University of
Stavanger, Norway. She has received her doctorate in Mechanical Engineering from
National University of Singapore (2006). She has then worked as a researcher at Norweigian
University of Science and Technology and SINTEF. Her main research interests are drilling,
new energy production and modelling.