Probabilistic Analysis of Failure Risk in the Primary Geothermal Cycle

The implementation of renewable energy sources, and geothermal energy in particular, is becoming increasingly important in Germany. However, geothermal power generation is a high risk, capital intensive technology and its future use will depend on how successfully it can be integrated within the German power grid infrastructure. For this to happen, its inherent operational risks must be reduced to a level that will guarantee a safe, available and affordable geothermal energy production over a plant’s lifetime. To operate successfully in the deep, hot, saline conditions that are associated with the Northern German Basin, a geothermal power plant will need to incorporate an Enhanced Geothermal System (EGS). The objective of this study is to identify and statistically model the main causes of failure in the primary cycle of an EGS and how likely they were to occur. In so doing, it is hoped to reduce the probability of downtime in such geothermal power systems in order to achieve higher plant online availability. INTRODUCTION A research project was set up at the Clausthal University of Technology to investigate the damage to equipment and subsequent system failures observed during the operation of deep geothermal power plants in Germany. There are three potential regions for geothermal energy production in Germany, namely the Upper Rhine Graben area, the Bavarian Molasse Basin and the Northern German Basin (Figure 1). The latter area is the main focus of this study, where the relatively deep geothermal reservoirs are characterized by high temperatures and saline formation waters. Figure 1: Regions of potential for hydrogeothermal exploitation within Germany (Pester et al. 2010). A theoretical system analysis of a survey of geothermal power plant operators in Europe was carried out in order to define the main causes of failure in their geothermal systems. Due to the hot, high salinity operating environment, the most common causes of equipment failure are corrosion, scaling, erosion and fatigue. The resulting damage to plant means high maintenance and repair costs are incurred with a corresponding increase in plant downtime and a subsequent loss of power generation revenue. In order to cope with the harsh operating conditions, the materials used throughout the EGS must be designed to meet the 2 challenges outlined above (Fichter et al. 2009, 2010-1; Schröder & Hesshaus 2009). Experimental datasets from investigations into fluid-rock wash out were used (Bai et al. 2010) to define the fluid composition in the Northern German Basin and to characterize the various fluidmaterial surface interactions in an EGS. These interactions were described probabilistically as rates of corrosion, scaling, erosion and fatigue of the different material surfaces, such as carbon steel, stainless steel, etc. The downtime of a geothermal power plant will depend on the speed and magnitude of these fluid-material interactions. To estimate the downtime probability of each single aggregate, as a function of the surface materials and the circumfluent milieu, the rates were varied in a Monte Carlo simulation. The results of the probabilistic modeling were implemented in a decision tree to visualize which fluid-material interactions had the most influence. Finally, a sensitivity analysis was performed to characterize plant downtime as a function of the fluid-material surface interactions (Fichter et al. 2009, 2010-1, 2010-2; Schuyler 2001; Stapelberg 2009). TYPICAL DEEP GEOTHERMAL SYSTEM The typical deep geothermal system consists of three loops: primary, secondary and tertiary (Figure 1). The role of the primary loop is that of carrying the hot geothermal fluid from subsurface to surface and transferring the heat to the secondary loop (power plant) and the tertiary loop (district heating and cooling system). The present paper focuses on the primary loop. A simplified primary loop consists of the production well, the electrical submersible pump, the separator, the heat exchanger and the injection well. The circulating fluid recovers the heat from the hot reservoir. High production rates, high temperatures, significant mineralization and possibly high solids loading are to be expected from deep geothermal systems. The chemical equilibrium of the highly mineralized geothermal fluid can be altered by pressure and temperature changes, and also by changes in material surface properties encountered by the fluid along its path to surface. Changes in mineral content of the reservoir over time, depending on the initial condition of the circulated fluid composition and the washed out zone will lead to continuous mineralization of the re-circulated fluid. Figure 2: Schematic of a deep geothermal system, showing its primary, secondary and tertiary loops. Changes in the chemical composition of the geothermal fluid over time can lead to scaling and corrosion of various components of the geothermal plant. These phenomena are enhanced by frequent plant start-ups and shut-downs, which also cause temperature cycling within the system that has a negative impact on the integrity of the well completions. To counteract these mechanisms and to ensure minimum system downtime, suitable hardware components and materials should be selected. The economic feasibility of geothermal power production strongly depends on system reliability (Gedzius et al. 2010; GeoHyBe 2009; Fichter et al. 2009, 2010-1, 2010-2; Schröder & Hesshaus 2009; Paschen et al. 2003). DATABASE USED FOR THIS STUDY The data used in this study were sourced from worldwide literature, from European plant operators, experimental data sets, and from the E&P service industry. Analysis of the data helped to define the type and origin of plant failure and damage, the fluid composition at the plant and the interaction of this fluid with the plant’s material surfaces (Table 1) (Bäßler et al. 2009;