Page 270 - Geothermal Energy Systems Exploration, Development, and Utilization

246 5 Geothermal Reservoir Simulation

Stralsund
Rosemanowes
Groß Schönebeck
Soultz KTB
Urach Spa

Figure 5.1 Selected geothermal research sites in Europe.
Geothermal reservoir simulation requires an adequate mathematical representa-
tion of the physical and chemical phenomena during the long-term heat extraction
process. Mainly deterministic methods are used by solving the underlying partial
differential equations (Section 5.1.1). Owing to data availability for deep geo-
reservoirs, the parameterization of numerical models is complicated. Therefore,
quantifying the parameter uncertainty using stochastic methods such as Monte
Carlo simulation is an important part of the reservoir analysis (Section 5.1.2).

5.1.1
Geothermal Modeling

The use of computer modeling in the planning and management of the develop-
ment of geothermal fields has become standard practice during the last 20 years.
During that time models have been developed for more than 100 geothermal fields
worldwide (Willis-Richards and Wallroth, 1995; O’Sullivan et al., 2001). Owing to
geological complexity and the number of processes involved, such as geometry,
hydraulics, thermal effects, geochemical reaction, and stress changes, numerical
methods have been widely used for geothermal reservoir simulation (Zyvoloski
et al., 1988).
The analysis of coupled processes, in particular, feedbacks of mechanical, ther-
mal, and geochemical effects to the flow system, is important for both hydrothermal
(Clauser, 2003) and HDR systems (Tsang, 1991; Bower and Zyvoloski, 1997; Mc-
Dermott and Kolditz, 2006). Numerical THM models have been developed and
applied to several HDR sites such as Soultz-sous-For´ ets in the Rhine Valley (Kohl
et al., 1995)(Hicks et al., 1996) and Urach Spa in the Swabian Alb by McDermott
et al. (2006); Watanabe et al. (2009). More recently, chemical effects have been
included into the coupled analysis (K¨ uhn, 2004; Kiryukhin et al., 2004; Bachler
and Kohl, 2005). One of the key questions hereby is how dissolution and pre-
cipitation processes can change the pore structure and therefore the reservoir
permeability. Discrete fracture network (DFN) models are available for the simula-
tion of fluid, mass, and heat transport, even for realistic geological structures, for
example, (Bruel, 1995b; Kolditz, 2001; Bruel, 2002) for the Soultz HDR reservoir.
Their applicability in the context of fully coupled THM analysis, however, is still
restricted to simplified problems (Walsh et al., 2008). Equivalent porous media
approaches are used for THM analysis of fractured rock instead (Birkholzer et al.,
2008).