Energy for Sustainable Development 15 (2011) 411–419
Contents lists available at ScienceDirect
Energy for Sustainable Development
Power production from a moderate temperature geothermal resource with regenerative Organic Rankine Cycles Alessandro Franco Department of Energy and System Engineering (DESE), University of Pisa, Largo Lucio Lazzarino, 56122 Pisa, Italy
a r t i c l e
i n f o
Article history: Received 7 September 2010 Revised 4 June 2011 Accepted 4 June 2011 Available online 7 July 2011 Keywords: Moderate-temperature geothermal sources Organic Rankine Cycle Binary cycle Recuperated Rankine cycle Rejection temperature Optimum design
a b s t r a c t Much remains to be done in binary geothermal power plant technology, especially for exploiting lowenthalpy resources. Due to the great variability of available resources (temperature, pressure, chemical composition), it is really difﬁcult to “standardize the technology”.The problem involves many different variables: working ﬂuid selection, heat recovery system deﬁnition, heat transfer surfaces sizing and auxiliary systems consumption. Electricity generation from geothermal resources is convenient if temperature of geothermal resources is higher than 130 °C. Extension of binary power technology to use low-temperature geothermal resources has received much attention in the last years. This paper analyzes and discusses the exploitation of low temperature, water-dominated geothermal ﬁelds with a speciﬁc attention to regenerative Organic Rankine Cycles (ORC). The geothermal ﬂuid inlet temperatures considered are in the 100–130 °C range, while the return temperature of the brine is assumed to be between 70 and 100 °C. The performances of different conﬁgurations, two basic cycle conﬁgurations and two recuperated cycles are analyzed and compared using dry organic ﬂuids as the working ﬂuids. The dry organic ﬂuids for this study are R134a, isobutane, n-pentane and R245fa. Effects of the operating parameters such as turbine inlet temperature and pressure on the thermal efﬁciency, exergy destruction rate and Second Law efﬁciency are evaluated. The possible advantages of recuperated conﬁgurations in comparison with basic conﬁgurations are analyzed, showing that in a lot of cases the advantage in terms of performance increase is minimal but signiﬁcant reductions in cooling systems surface area can be obtained (up to 20%). © 2011 International Energy Initiative. Elsevier Inc. All rights reserved.
Introduction Moderate-temperature water-dominated systems, with temperatures below 130 °C, account for about 70% of the world's geothermal energy potential (Barbier, 2002). The distribution of geothermal energy as function of the resources temperature and the technical resource potential has been evaluated recently by Stefansson (2005), starting from a general correlation between the existing geothermal high temperature resources inferring a total geothermal potential of 200 GWe. Binary technology allows the use of low temperature water dominant reservoirs and makes geothermal power production feasible even for countries lacking high enthalpy resources at shallow depth. For binary plants two different systems currently are state of the art, the Organic Rankine Cycle (ORC) and the Kalina cycle. The binary power plants have the least environmental impact due to the “conﬁnement” of the geoﬂuid. In a binary cycle power plant the heat of the geothermal water is transferred to a secondary working ﬂuid, usually an organic ﬂuid that has a low boiling point and high
E-mail address: [email protected]
vapor pressure when compared to water at a given temperature. The cooled geothermal water is then returned to the ground by the reinjection well to recharge the reservoir (DiPippo, 2008). Such a geothermal plant has no emissions to the atmosphere except for water vapor from the cooling towers (only in case of wet cooling) and any losses of working ﬂuid. Thus, environmental problems that may be associated with the exploitation of higher temperature geothermal resources, like the release of greenhouse gases (e.g. CO2 and CH4) and the discharge of toxic elements (e.g. Hg and As) are avoided. Another advantage of the binary technology is that the geothermal ﬂuids (or brines) do not contact the moving mechanical components of the plant (e.g. the turbine), assuring a longer life for the equipment. Binary plants have allowed the exploitation of a large number of ﬁelds that may have been very difﬁcult (or uneconomic) using other energy conversion technologies (Schochet, 1997; DiPippo, 2004; Bronicki, 2007). Of the about 10,700 MW of geothermal plants installed worldwide, more than 1170 MW use ORC or steam/ORC combined cycles (Bertani, 2010). There exist a great number of studies addressing both the different characteristics of geothermal ﬁelds and the various types of power plants that could be used in their exploitation for electricity production; Barbier (2002), Bertani (2005), Lund (2007) and DiPippo
0973-0826/$ – see front matter © 2011 International Energy Initiative. Elsevier Inc. All rights reserved. doi:10.1016/j.esd.2011.06.002
A. Franco / Energy for Sustainable Development 15 (2011) 411–419
Nomenclature e Ex h H I K(Qcond) m p Q S s T To v x W
speciﬁc exergy (J kg − 1) exergy ﬂow (W) speciﬁc enthalpy (J kg − 1) enthalpy ﬂow (W) exergy loss ﬂow (W) function deﬁning the heat transfer performance of condenser (W K − 1) mass ﬂow rate (kg s − 1) pressure (bar) heat ﬂow rate (W) heat transfer surface (m 2) speciﬁc entropy (J kg − 1 K − 1) temperature (°C) reference temperature (K) speciﬁc volume (m 3 kg − 1) steam quality power (W)
Greek symbols β speciﬁc brine consumption (kg M J − 1) ΔT temperature difference (°C) Δp pressure drop (bar) η efﬁciency ηI First Law efﬁciency ηII Second Law efﬁciency
Subscripts and abbreviations cond condenser, condensation CS cooling system el of the electric system env environment esp of expansion exp of the expander geo of the geothermal brine gross gross power in at the inlet is isentropic liq of the liquid max maximum net net power o reference state out at the outlet PL pressure levels pp pinch point pump of the circulation pump RAN Rankine Cycle RANSH Rankine Cycle with superheater RHE recovery heat exchanger rec of the recuperator rej rejection sat saturation sh superheater ST steam turbine wf working ﬂuid
(2008) provide analyses of the various technological solutions and of the state of the art. A geothermal binary power plant is characterized by high brine speciﬁc consumption and low plant efﬁciencies (5– 10%); First Law efﬁciencies even Second Law efﬁciencies are typically in the 25 to 45% range and by the requirement for large heat transfer
surfaces both for the heat recovery heat exchanger and for the condensation system. The design of binary plants, although widely addressed in the literature (e.g. Gnutek and Bryszewska-Mazurek, 2001; Kanoglu, 2002; DiPippo, 2004; Hettiarachchi et al., 2007; Kaplan, 2007; Kose, 2007; Moya and DiPippo, 2007) is still an area of active research. The author in a recent paper has already analyzed the perspective and the thermodynamic performance of a lot of possible plant conﬁgurations, combining different available organic ﬂuids and recovery cycles for different combination source temperature–reinjection termperature (Franco and Villani, 2009). Different analyses show the possible beneﬁts, in terms of the extent of using the thermal energy of lowtemperature geothermal water, that arise from utilizing hybrid and dual-ﬂuid-hybrid power plants rather than ORC power plants (Borsukiewicz-Gozdur, 2010). Reconsidering the technology of ORC plants, at present this is not at a stage of development capable of providing “standard machinery”, and each installation is designed for the conditions at a given location by the big manufacturers in this ﬁeld, like Ormat, Maﬁ Trench, Siemens and UTC/Turboden. Only recently systematic attempts to standardize machinery have been made: for example UTC Power has proposed The PureCycle® Power System. This is an electric power generating system which runs off any hot water resource at temperatures as low as 90 °C. The hot water can be derived from a geothermal source or other waste heat source. Currently this ORC unit is sized at 280 kW (gross) of electrical power. The ﬂexibility of a modular approach in geothermal power technology is interesting because of employing small, off-the-shelf units, a plant that can be scaled to the local geothermal resource, energy demand and available ﬁnancing. Organic Rankine Cycles seem to be a promising technology in the perspective of a decrease in plant size and investment costs. They can work at lower temperatures, and the total installed power can be reduced down to the kW scale. The market for ORC's is growing at a rapid pace. Since the ﬁrst installed commercial ORC plants in the 80's, an exponential growth has been seen in the past decade. The success of the ORC technology can be partly explained by its modular feature: a similar ORC system can be used, with little modiﬁcations, in conjunction with various heat sources. This success was reinforced by the high technological maturity of most of its components. Moreover, unlike with conventional power cycles, local and small scale power generation is made possible by this technology. Today, Organic Rankine Cycles are commercially available in the MW power range, while very few solutions are actually suitable for the kW scale. Organic Rankine Cycle (ORC) raises considerable interest as it makes it possible to produce electricity from cooler geothermal sources, typically within the 100–130 °C temperature range, exceptionally down to 90–95 °C, often available from below 1000 m deep production well (a case of 75 °C is available too) increasing the number of geothermal reservoirs in the world that can potentially be used for generating electricity. No high plant ratings can be expected for obvious thermodynamic reasons even if improvements should concentrate on cycle and plant efﬁciencies. One of the problems of geothermal binary plants is the rejection of heat at low temperature (thermal pollution). If no adequate water source is available, a dry cooling system must be used. Although such a system solves the problem of water supply, it raises many others. The parasitic power consumption is relatively high because of the need for forced ventilation; a dry cooling system can absorb from 10–12% of gross power (under ideal conditions), to as much as 40–50% if the ambient temperature is very close to the condensation temperature. The capital cost is also quite high; 30–35% of the total capital cost of the geothermal project. A good review on the costs of small geothermal plants is available in the literature (Entingh et al., 1994;
A. Franco / Energy for Sustainable Development 15 (2011) 411–419
Battye et al., 2009). In contrast to other thermal power plants the ratio of auxiliary power to gross electricity can signiﬁcantly vary in geothermal binary power plants depending on site speciﬁc conditions. It typically lies between 30 and 50% but can be also higher depending on the site-speciﬁc (Frick et al., 2010). Even though the basic Rankine cycle is fundamentally the same, there have been several advances in component design and development that have resulted in more efﬁcient turbines and condensers. Several companies are presently working to improve this technology. During the past decade, some new thermodynamic cycles have been also developed that show a better resource utilization compared to the existing cycles. The “regenerative Organic Rankine Cycle” has been proposed and analyzed in some recent papers (Mago et al., 2008; Yari, 2010). The use of a recuperator or a preheater has utility for power cycles employing lower pressure ratios and with a working ﬂuid having a low speciﬁc heat ratio, i.e. a working ﬂuid that emerges from the expander still at a relatively high temperature. This represents the basis for the system presented in this paper. In this cycle the vapor extracted from the turbine is used to preheat working ﬂuid before entering the recovery heat exchanger (in particular in the economizer). The use of recuperative heat exchanger is surely one of the most promising solutions because it can increase the thermodynamic efﬁciency of the cycle, permit the reduction of cooling system dimension and consequently can contribute to the reduction of parasitic losses. This particular cycle conﬁguration is proposed by some manufacturers in the perspective of small power systems (with output power below 200 kW), mainly in connection with use of refrigerant R245fa. This paper presents an overview of current R&D in the ﬁeld of small-scale ORC for the exploitation of geothermal sources with reduced temperature below 130 °C. Therefore, it is the objective of this work to analyze the performance of such those new cycles and to consider the potential improvements that will result in higher cycle performance or lower resource utilization and lower cost of electricity generation. The analysis is mainly oriented to geothermal ﬂuid inlet temperatures in the 100–130 °C range, while the return temperature of the brine is assumed to be between 70 and 100 °C. Low-temperature geothermal technology overview Currently, the total installed power worldwide of geothermal binary power plants is about 800 MWe, representing about 8% of the total geothermal power installed (Franco and Villani, 2009). Currently binary plants are the most widely used types of geothermal power plant with 193 units in operation generating more than 800 MW of power in 17 countries. A complete distribution of the plants installed at the end of 2006 is reported in Table 1 (Valdimarrson, 2006; Bertani, 2010). A lot of them are small plants already in operation while another remarkable amount of binary plants is commissioned or in advanced construction phase (Bertani, 2007, 2010). The geothermal binary plants currently in operation can be divided according to different classiﬁcations: the ﬁrst and more meaningful is surely the dichotomy between “stand-alone” or “bottoming cycles” but another difference is related to the power output. Binary plants can be classiﬁed according to the cooling system used: plants with a wet cooling system, where the working ﬂuid is condensed by cooling water, and plants with a dry cooling system, where the heat is rejected directly to the air (Fig. 1). In the latter case, no water supply is necessary, but a large heat transfer surface is required and the fans of the cooling system consume a signiﬁcant fraction of the gross generated power. Another classiﬁcation is based on the installed power so that it is possible to identify two main groups according to the total power produced.
Table 1 Number of installed binary plants.
USA New Zealand Philippines Iceland Guatemala Portugal Austria Germany Kenya Mexico Japan Costa Rica Australia Ethiopia Austria Turkey China Nicaragua Thailand El Salvador France
Flash + binary
139 10 13 8 1 5 3 3 1 2 2 2 2 2 2 2 1 1 1 1 1 202
10 14 5 7
The ﬁrst group includes medium and large binary power plants (with output power of at least 5 MW). Brine speciﬁc consumption is often higher than 50 kg/s per MW produced. The heat recovery cycle can be a basic or superheated Rankine cycle or a more complex recovery cycle (e.g. a dual-pressure level Rankine cycle or a supercritical cycle). The selection of the working ﬂuid is based on thermodynamic considerations; i.e. on the thermo-physical properties of the geothermal and working ﬂuids, as well as the heat recovery cycle chosen. The working ﬂuids include hydrocarbons (mainly butane and pentane) and synthetic refrigerants (mainly HFCs). Multicomponent working media, where evaporation and condensation occur at variable temperatures, as for example in the Kalina cycle, could increase the thermodynamic efﬁciency are also considered but their use is minimum (Franco and Villani, 2009). The parasitic power loss is quite large by comparison with other types of power plants (combustion or dry or ﬂash steam power plants). Circulation pumps and cooling tower fans consume a considerable fraction of the generated power. The electricity required to run the circulation pumps is relatively constant and is generally between 2 and 10% of the gross plant output, being a function of both the working ﬂuid and the operating pressure. The power usage by the cooling tower fans is strongly affected by operating and environmental conditions, and can vary between 10 and more than 30% of the gross power. The second WF
Fig. 1. Schematic diagram of a binary geothermal binary plant with a dry cooling system. (A: acquifer; CS: cooling system; GE: generator; GEO: geoﬂuid; P: pump; RHE: recovery heat exchanger; T: turbine; WF: working ﬂuid).
A. Franco / Energy for Sustainable Development 15 (2011) 411–419
type of geothermal binary plant is the small “stand alone” power systems (with output power below 5 MW according to the deﬁnition given by Lund and Boyd, 1999), pilot experimental plants or small binary units installed in ﬁelds where previously a ﬂash-steam plant had been used to generate power using less than 150 °C ﬂuids (Angelino et al., 1995; DiPippo, 1999; Schochet, 2000); the installed capacities range from some hundreds of kW to a few MW. Our speciﬁc analysis is related to small size plants which exploit geothermal sources of temperature below 130 °C. These plants often operate through advanced thermodynamic cycles (dual pressure level Rankine cycle or Kalina cycle) and may also use different or unconventional working ﬂuids, such as ammonia–water mixtures (e.g. Husavik, Iceland). The characteristics of this second group of plants are given in Table 2 which is constructed using data from various papers and open-ﬁle sources on the web. The plants reported in the table cover a wide range of geothermal ﬂuid temperatures (74–124 °C) so that brine speciﬁc consumption, which is strongly dependent on the thermodynamic and chemical properties of the geoﬂuid, lies in the range from 44 to 200 kg/s for each MW of electricity produced. The remarkable difference among the various plant performances can be explained in a lot of cases with the differential temperature between source temperature and rejection temperature. Considering the different operating conditions available, it is difﬁcult to identify general criteria for the optimum design of geothermal binary plants. ORC manufacturers have been present on the market since the beginning of the 80's. They provide ORC solutions in a broad range of power and temperature levels. Interesting performance increases can be obtained by means of Kalina cycle even if the apparent efﬁciency increase of the ﬁrst three plants of Table 2 is mainly due to the presence of a water cooling system with a substantial reduction of parasitic loss. Among the various plants the Rankine cycle of Chena Hot Springs use geothermal brine with a 74 °C source and 25 °C rejection temperature with a temperature differential of 49 °C. The same plant is obviously favored by the really low value of condensation temperature that can be approximately 0 °C. Performance of the ORC system is strongly correlated with those of the expander too. The choice of the machine strongly depends on the operating conditions and on the size of the system. Table 3 shows a summary of continuous improvement of the turbine efﬁciency in some of the binary plants installed from 1984 to 2000. In the last years some new turbines (like the Euler Turbine and the Variable Phase Turbine) have been developed that enable the economic and efﬁcient implementation of these cycles that maximize the geothermal resource utilization (Welch and Boyle, 2009).
The future development of moderate temperature geothermal binary plants exploiting quite low temperature sources will be increasingly possible if technology would be capable of providing “standard machinery” even if this will be not simple. Among the various technical solutions, the use of recuperative heat exchanger is surely the one that requires more attention from this viewpoint. Regenerative Organic Rankine Cycle The gross power output of a binary unit increases with decreasing condensation temperatures due to the increasing enthalpy difference in the expansion machine. However, also the requirements for the recooling increase with decreasing condensation temperatures. Another limiting factor in most geothermal sources is the silica scaling risk, which increases as the brine temperature drops. Moreover, regarding the net electricity production, it must be considered that the ambient conditions can signiﬁcantly vary during the year. Regarding the lifetime operation of a binary power unit changes in the reservoir productivity or cooling of the reservoir can have an inﬂuence on the generating capacity of the plant (Porras and Bjornsson, 2010) and in some cases it could lead to unexpected plant failure. A method to partially overcome both the problems of the quite high reinjection temperature and the cooling temperature limit is to add a recuperator which provides some of the preheating heat from the vapor exiting the turbine. The recuperator is applicable when the organic ﬂuid is of the “dry expansion” type, namely a ﬂuid where the expansion in the turbine is done in the dry superheated zone and the expanded vapor contains heat that has to be extracted prior to the condensing stage. Since the temperature after expansion in the turbine remains for most ORC ﬂuids higher than the ﬂuid's condensing temperature, one must extract this additional heat in the condenser. So the recuperator provides some of the preheating heat from the vapor exiting the turbine (Fig. 2). The regenerative binary Rankine Cycle plant exploits the unique characteristics of a binary working ﬂuid to recycle some of the heat that would otherwise be dumped through the condenser. This twophase cycle has been used by some manufacturers like Ormat in geothermal projects all over the world but appears to be particularly interesting in connection with low temperature geothermal resources (b130 °C). The reduction of the heat rejected to the environment permits a decrease of the heat exchange surface area required for the cooling system and consequently of the air-cooled condenser parasitic load and costs. An important percentage of gross power produced aircooled ORC is consumed by parasitic loads mostly due to the
Table 2 Small binary power plants using low-temperature geothermal resources or non-conventional working ﬂuids. Plant and location
Gross capacity (kWe)
Speciﬁc brine consumption [(kg/s)/MW]
Husavik, Iceland Unternhaching, GER Bruchsal, GER Empire, USA Fang, Thailand Nagqu, China Bad Blumau, Austria Wineagle (Susanville), USA Altheim, Austria Wabuska, USA Wendel, USA Birdsville, Australia Neustadt-Glewe, GER Chena Hot Spring, USA
124 122 120 118 116 110 110 110 106 104 103 98–99 98–100 74
Kalina Kalina Kalina RAN RAN RAN RAN RAN RAN RAN RAN RAN RAN RAN
NH3-H2O NH3-H2O NH3-H2O Isopentane Isopentane Isopentane Isopentane Isobutane C5F12 Isopentane R114 R114 (Isopentane) C5F12 R134a
2030 4000 610 1200 300 1300 250 750 1000 750 2000 150 230 250
53 44.2 51.8 90.8 47.4 69 120 105 86 90 128.2 200 120.8 57.9
Wet Wet Wer Dry Wet Dry Dry Dry Dry Wet Wet Wet Wet Wet/dry
(1700) (3400) (550) (1000) (175) (1000) (180) (600) (500) (600) (1600) (120) (180) (210)
A. Franco / Energy for Sustainable Development 15 (2011) 411–419 Table 3 Improvement of turbine efﬁciency over time in binary cycle power plants (from Thomsen, 2006). Year of installation
1984 1985 1989 1993 1996 2000
Steamboat, USA Ormesa, USA Puna, USA Heber, USA Rotokawa, NZ Olkaria, USA
72 75 78 83 84 88
operation of cooling fans and working ﬂuid circulation pump. The loss of 15–20% up to 30% of gross power to parasitic loads is usual for ORC. Moreover the utilization of recuperated solution can compensate better the possible temperature decrease of the geothermal source during the operating life of the plant and the variation of the environmental temperature. The average and the range of variation of the ambient temperature are also important, especially when dry cooling towers are used to condense the working ﬂuid; in this case the condensation temperature greatly affects parasitic power consumption. If the condensation and ambient temperatures are too close, the increase of power consumption in the cooling system severely reduces the net power production. A compromise is necessary between the intrinsic thermal efﬁciency of the recovery cycle and the power loss due to parasitic consumption. The variation of ambient temperature presents a difﬁcult problem because it changes not only annually and seasonally, but also hourly; the condensation temperature cannot follow such a trend. The difference between ambient and condensation temperatures can vary by a factor of three or four between its maximum and minimum values; i.e. during winter nights and summer days, respectively. The recuperated process is still used in many geothermal projects all over the world, such as the 20 MWe Zunil in Guatemala, 1.8 MWe Oserian and 13 MWe Olkaria III in Kenya, but appears to be particularly interesting in the perspective of realization of small size standard machinery for exploiting geothermal sources at moderate temperature in particular below 130 °C. This is because the presence of recuperator or preheater permits, for a given temperature of the geothermal source, reducing the sensitivity with respect to the variation of rejection temperature and of the environmental temperature, that can vary in a sensible way due to seasonal changes.
Fig. 2. Schematic diagram of recuperated binary Rankine cycle. (A: acquifer; CS: cooling system; GE: generator; GEO: geoﬂuid; P: pump; RHE: recovery heat exchanger; T: turbine; WF: working ﬂuid).
Modeling and analysis of small geothermal binary power plants with regenerator The design of a geothermal binary plant needs to take into account the particular type of thermodynamic cycle, the pump and turbine, the recovery heat exchanger and condenser, and the cooling system. For this reason, the process has to consider a large number of design variables and operating parameters. The temperature, pressure and chemical composition of the geothermal ﬂuid, the rejection temperature, the ambient temperature and the maximum rate of energy extraction that can be sustained without a signiﬁcant decrease of the water temperature in the reservoir, can be considered as fundamental variables of the problem. Some of the parameters cannot be modiﬁed (e.g. geothermal ﬂuid inlet temperature) and others lie in welldeﬁned ranges (rejection temperature, geothermal ﬂuid ﬂow rate, ambient temperature). All of these variables are relevant for deﬁning the technical speciﬁcations of the plant (thermodynamic cycle, saturation pressure, maximum temperature), but some are more important than others. The exergy potential of a geothermal resource depends strongly on the geothermal ﬂuid and rejection temperatures. The ﬂuid inlet temperature is a parameter controlled primarily by the characteristics of the geothermal ﬁeld, even if changing well depths can sometimes modify it. On the other hand, the rejection temperature is set so as to avoid scaling problem and is one of the most important factors limiting the complete utilization of geothermal resources (Stefansson, 2005; Mroczek et al., 2000). Identiﬁcation of the characteristics of the geothermal ﬂuids and of the environment is the starting point of the binary plant design, and inﬂuences the speciﬁcation of variables such as the choice of working ﬂuid (cryogenic, synthetic refrigerant, multicomponent medium), the recovery cycle, the condensing temperature, the recovery heat exchanger and its thermal and ﬂuid-dynamic design, and the cooling system. The selection of suitable ﬂuids for use in binary cycle plant is quite a complex problem and cannot be dissociated from the choice of the heat recovery cycle as already discussed by Franco and Villani (2009). The system consists of some relatively simple units, such as the turbine, pump, and heat exchangers, combined with other relatively complex components, such as the regenerator. The basic models for all of the units involve mass/energy balances and heat transfer equations. Four available thermodynamic cycles are analyzed (Fig. 3): 1 Recovery with Rankine cycle: RAN in Fig. 3(A); 2 Recovery with Rankine cycle with superheater: RANSH in Fig. 3(B); 3 Recovery with Rankine recuperative cycle: RAN recuperated in Fig. 3(C); 4 Recovery with Rankine superheatered with recuperator: RANSH recuperated in Fig. 3(D). Potentially the two solutions with recuperators represented in Figs. 3(C) and (D) achieve a better match between the working and geothermal ﬂuids, and the working ﬂuid and the cooling medium; in particular they tend to reduce the exergy loss due to heat transfer. The choice of the working ﬂuid and heat recovery cycle is done on the basis of thermodynamic performance, economic considerations (capital cost of the plant) and adaptability to variations in operating conditions. This last parameter is important because the temperature of a geothermal ﬂuid may decrease after the start of ﬁeld exploitation. Among the various ﬂuids suitable as working media for the present ORC application, four different ﬂuids were selected after ﬁrst screening calculations because they seem to be the most common in ORC applications: R134a, R245fa, n-pentane, and isobutane. Table 4 provides the characteristic properties of the ﬂuids analyzed in Fig. 4 temperature–entropy (T–s) diagram and pressure–enthalpy (p–h) diagram graphs. As can be observed from Fig. 4, they belong to the two different ﬂuids characterized by different types of vapor saturation curves in
A. Franco / Energy for Sustainable Development 15 (2011) 411–419
Fig. 3. T–s diagram of the cycle in the four different cases analyzed: Rankine cycle (a); Rankine cycle with superheater (b); Rankine cycle with recuperator (c); Rankine cycle with superheater and recuperator (d).
the temperature–entropy (T–s) diagram: ﬂuids with positive slopes (dT/ds) like R134a and ﬂuids with negative slopes like isobutane, npentane and R245fa. In the ﬁrst case, since the vapor expands through the turbine along a sub-vertical line on the T–s diagram, saturated vapor at the turbine inlet remains superheated (dry) throughout the turbine, and it is not necessary to resort to a Rankine cycle with superheat. For performance analyses of binary power plants, First and Second Law efﬁciencies can be used. The First Law efﬁciency is conventionally calculated considering the ratio between the net power and the product of the geothermal ﬂow rate and the enthalpy difference between geothermal source and temperature of rejection: ηI =
Wnet W net = Hgeo −Hrej mgeo hgeo −hrej
where Wnet = Wgross −Wpump −WCS
with Wgross = mwf ⋅ Δhesp ⋅ηis;t ⋅ηel
Wpump = mwf ⋅Δhpump ≃mwf ⋅v˜⋅Δp
WCS = K ðQcond Þ⋅ΔTcond−env
in which K(Qcond) is a function representative of the cooling system including heat transfer coefﬁcient and heat exchange surface. The
other important indicator is the Second Law efﬁciency. Eq. (6) is the standard expression for Second Law efﬁciency. ηII =
Wnet Wnet Wnet h i = = Egeo mgeo ⋅egeo mgeo hgeo −ho −To ⋅ sgeo −so
where ho and so are the reference values for enthalpy and entropy (calculated for T = To). The ﬁrst one (Eq. (1)) does not reﬂect the thermodynamic quality of the conversion process, even though it can be used to compute the heat discharge to the environment. The Second Law efﬁciency is more appropriate for assessing the performance of binary plants. Second Law efﬁciency (Eq. (6)) may be deﬁned using a conventional reference temperature To. Since the First and Second Law efﬁciencies are also linked to the brine inlet temperature, it is possible to compare the various available combinations of the source, rejection and condensation temperatures, and obtain an indication about the speciﬁc power of the plant. For this reason another important merit parameter that can be considered in the analysis is the mass ﬂow rate to generate a ﬁxed power output, or speciﬁc brine consumption, which is given by: β¼
mgeo : Wnet
The parameter β is often considered when the minimization of geothermal ﬂuid ﬂow rate (speciﬁc consumption) for a given power is suggested as an objective function for optimal design (Hettiarachchi et al., 2007). Cycle calculations are performed according to the schematization of Fig. 5 with a set of computer worksheets developed by the author for
A. Franco / Energy for Sustainable Development 15 (2011) 411–419 Table 4 Properties of the tested working ﬂuids. Working ﬂuid
CH2FCF3 Molecular weight 102 Class HFC Critical temperature 101.1 [°C] Critical pressure [MPa] 4.060 Shape of saturation Wet curve
134 HFC 154.1
58.12 HC 134.7
72.15 HC 196.6
Recovery Heat Exchanger Q in
W pump W gross
Q out Q cond , out Cooling system
Q cond , in
the thermodynamic analysis and charting of ORCs. The program computes thermodynamic properties of pure ﬂuids and mixtures using a library of subroutines already validated (Franco and Villani, 2009). The regenerative cycle is applicable mainly when the organic ﬂuid is of the “dry expansion” type, namely a ﬂuid where the expansion in the turbine is done in the dry superheated zone and the expanded vapor contains heat that has to be extracted prior to the condensing stage. The condensed vapor preheats the main organic ﬂuid stream as it exits the feed pump according to the balance: mwf ;l Δhliq = mwf ;sh ðΔhsh Þ:
From a theoretical point of view the vapor can be cooled to the temperature of the incoming liquid. In practice an effectiveness of the
Isentropic efﬁciency of turbine: ηis = 0.8 Electric generator efﬁciency: ηel = 0.95 Pinch point: ΔTpp = 5 °C Minimum temperature difference in the recuperator: 5 °C Environmental temperature: Tenv = Tcond − 10 °C
The beneﬁts of these conﬁgurations emerge not particularly in term of cycle efﬁciency and reduction of brine speciﬁc consumption, but only in terms of cooling system dimension reduction. Here we present some performance results obtained using the model described in the previous section; a wide range of operating conditions were covered. In particular, we considered 100–130 °C geothermal 35
Performance evaluations of recuperated binary plant conﬁgurations
B) p-h diagram 3,5
Twf ;in−rec −Twf ;out−rec Twf ;in−rec −Twf ;out−pump
This can be limited at a value below 1 or otherwise a ﬁnite minimum temperature difference between the vapor at the exit of the recuperator and the liquid at the outlet of the feed pump can be considered.
recuperator can be considered and this is referred to the vapor that has a lower speciﬁc heat than the liquid:
The implementation of recuperated conﬁgurations is analyzed and discussed using as hypothesis the production of a gross power of 500 kW. The calculation is carried out using the model described in Fig. 4 with the following additional hypotheses:
A) T-s diagram 463
Fig. 5. Block schematization of the binary plant model.
2,5 2 n-pentane
20 15 10
0,5 0 180
h [kJ/kg] Fig. 4. Thermodynamic phase diagrams of the four tested working ﬂuids.
Fig. 6. Mass ﬂow rate of brine extracted from the well: comparison of basic and recuperated cycles (combination Tgeo = 130 °C; Trej = 70 °C; Tcond = 30 °C).
A. Franco / Energy for Sustainable Development 15 (2011) 411–419
35 30 25 20 15
n-pentane 30 20 10
10 0 100
Fig. 7. Mass ﬂow rate of brine extracted from the well: comparison of basic and recuperated cycles (combination Tgeo = 110 °C; Trej = 70 °C; Tcond = 30 °C).
ﬂuids, 70–100 °C brine rejection temperatures, and 10–30 °C condensing temperatures. In all the cases analyzed, common reference values of ambient temperature (T0 = 298 K) and of geothermal ﬂuid pressure (pgeo = 15 bar) were assumed. The brine consumption (mgeo) is used as general performance indicator, the First and Second Law efﬁciencies are also considered according to the deﬁnitions given by Eqs. (1) and (2). A comparative analysis of Figs. 7 and 10 is interesting to show the possibility of obtaining similar results with reinjection temperatures of 70 and 80 °C. The possibility of maintaining the temperature at a level of 80 °C permits to prevent the silica scaling risk, which is increased as the brine temperature drops. It is clear that there are large differences between the various operating conditions. A 30 °C decrease in geothermal brine temperature increases the speciﬁc consumption by a factor of 2 to 3, and there is a remarkable difference between the best and worst conditions [(Tgeo − Trej) = 130–70 °C and 100–70 °C]. Figs. 6–9 analyze the various cases for the different values of the geothermal brine inlet temperature. In some of the cases the use of recuperator appears not to be particularly convenient from the point of view of the reduction of mass extracted. The same ﬁgures show the importance of the geothermal brine temperature. It is remarkable to show how if for a source temperature of 130 °C a mass ﬂow rate of geothermal brine in the range between 20 and 25 kg/s is sufﬁcient for a net production of 500 kW, a mass ﬂow rate in the range 50–60 kg/s is necessary for generating the same output power for a temperature of 100 °C. Fig. 9 shows how for all the four selected working ﬂuids, in case of the higher performance conﬁguration a given combination of temperatures (Tgeo and Trej), the difference between the obtained results does not depend on the particular ﬂuid used if Tgeo is over 120 °C, while it can be a little bit more meaningful if the brine temperature is in the range between 100 and 110 °C. The efﬁciency
Tgeo[°C] Fig. 9. Trend of the mass ﬂow rate extracted as a function of geothermal brine temperature.
increase that can be obtained with the recuperated solution is not particularly remarkable. A reduction of inlet temperature and increase of rejection temperature by 10 °C (i.e. Tgeo = 110 °C; Trej = 80 °C) entails a relatively modest increase in geothermal ﬂuid ﬂow rate (Fig. 10). Fig. 11 shows the effects of condensation temperature on power plant performance: for larger differences between source and rejection temperatures, good results are obtained with the ﬂuid R245fa while a ﬂuid like R134a appears to be not particularly advantageous. On the other hand, hydrocarbons perform better for smaller temperature differences. Analysis of the computed data also illustrates some further effects of condensation temperature. In particular it may be noted that a reduction in this temperature results in two different and opposite effects, i.e., an improvement in gross thermodynamic efﬁciency of the recovery cycle, but an increase in cooling system power consumption. Considering the whole results obtained in the analysis, we can state that while the reduction of brine speciﬁc consumption and the increase in efﬁciency with the recuperated cycle conﬁgurations are negligible but positive, the reduction of heat exchange surfaces can be signiﬁcant in some cases. The convenience in the utilization of recuperated cycle conﬁgurations appears to be particularly meaningful with a ﬂuid like R245fa while is not remarkable with R134a even if a reduction of the cooling system surface from a minimum of 5% till to a maximum of 15% is possible (Table 5). Conclusions This study shows that the geothermal power plant with a regenerative Organic Rankine Cycle is an interesting and promising option that should be studied in detail. It shows that an interesting method for overcoming partially the cooling temperature limit is to
60 50 40 30
40 30 20
10 0 RAN
Fig. 8. Mass ﬂow rate of brine extracted from the well: comparison of basic and recuperated cycles (combination Tgeo = 100 °C; Trej = 70 °C; Tcond = 30 °C).
Fig. 10. Mass ﬂow rate of brine extracted from the well: comparison of basic and recuperated cycles (combination Tgeo = 110 °C; Trej = 80 °C; Tcond = 30 °C).
A. Franco / Energy for Sustainable Development 15 (2011) 411–419
40 30 20 10 0
Fig. 11. Mass ﬂow rate of brine extracted from the well: comparison of basic cycle and recuperated cycles (combination Tgeo = 100 °C; Trej = 70 °C; Tcond = 20 °C).
Table 5 Percentage reduction of condenser dimension for Trej = 70 °C and Tcond = 30 °C. R134a RAN 130-70-30 120-70-30 110-70-30 100-70-30
8.8% 9.3% 9.8% 4.7%
7.0% 7.2% 6.2% 6.7%
10.6% 10.6% 13.8% 10.5%
4.7% 4.7% 4.7% 3.6%
19.0% 9.4% 14.0% 10.4%
9.2% 6.5% 9.2% 5.0%
13.5% 19.3% 15.6% 11.6%
add a recuperator which provides some of the preheating heat from the vapor exiting the turbine: the regenerative ORC. Regenerative ORC shows an increase on the Second Law efﬁciency compared with the basic ORC in several cases. In some cases the advantages related to the use of regenerative ORC cycle may be important in terms of a decrease in brine speciﬁc consumption (5– 10%) and in terms of decrease of cooling system surface (4–19%). An increase of efﬁciency of the plant is possible but mainly with some speciﬁc operating ﬂuids, as for example R245fa. In other cases it appears surely more important the addition of the superheater section, while the presence of the recuperator is less important even if the possibility of decreasing the dimensions of the cooling system of an amount of 5–15% is possible and this produces a proportional reduction of parasitic power for operation of cooling system, mainly in case of dry cooling system. Moreover the solution appears to be interesting for the opportunities of providing a standard machinery for small size (below 1 MW) geothermal sources due to the major ﬂexibility of the regenerative ORC with respect to possible variation of the combination source temperature – reinjection temperature – environmental temperature, though if in a quite reduced range. The analysis carried out in the present paper clearly shows the potential of regenerative ORC for the exploitation of a large number of geothermal ﬁelds, with source temperature below 130 °C. In any case the convenience of regenerative ORC cycles can be assessed only after the deﬁnition of a size (depending of the reservoir ﬂuid production) and by a detailed analysis of the cost involved in the design of the cycle, which can be quite higher than the conventional ORC cycle due the additional regenerative heat exchanger.
Angelino G, Bini R, Bombarda P, Gaia M, Girardi P, Lucchi P, et al. 1 MW binary cycle turbogenerator model made in Europe. Proceedings of the 1995 World Geothermal Congress, 18–31 May, Florence, Italy; 1995. p. 2125–30. Barbier E. Geothermal energy technology and current status: an overview. Renewable Sustainable Energy Rev 2002;6:3–65. Battye DL, Ashman PJ, Nathan GJ. The development of a basic cost and performance estimator for geothermal power plants. EGRG — Research Grant Report; 2009. Bertani R. World geothermal power generation in the period 2001–2005. Geothermics 2005;34:651–90. Bertani R. World Geothermal Generation in 2007. Geo-Heat Center Q Bull 2007;28(3): 8–19. Bertani R. Geothermal Power Generation in the World 2005–2010 update report. Proceedings World Geothermal Congress 2010 Bali, Indonesia, 25–29 April; 2010. Borsukiewicz-Gozdur A. Dual-ﬂuid-hybrid power plant co-powered by low-temperature geothermal water. Geothermics 2010;39:170–6. Bronicki L. Organic Rankine Cycle in geothermal power plants 25 years of ORMAT experience. UGI News 2007;VI(17):2–8. DiPippo R. Small geothermal power plants: design, performance and economics. GeoHeat Center Q Bull 1999;20(2):1–9. DiPippo R. Second Law assessment of binary plants generating power from lowtemperature geothermal ﬂuids. Geothermics 2004;33:565–86. DiPippo R. Geothermal power plants: principles, applications, case studies and environmental impact. 2nd edition. New York, NY, USA: Butterworth–Heinemann; 2008. 493 pp. Entingh DJ, Easwaran E, McLarty L. Small geothermal electric systems for remote power. Geotherm Resour Counc Bull 1994;23(10). November, Davis, CA. available at http://www.osti.gov/bridge/purl.cover.jsp?purl=/897243-qhLKfx/ (September 2010). Franco A, Villani M. Optimal design of binary cycle power plants for water-dominated, medium-temperature geothermal ﬁelds. Geothermics 2009;38:379–91. Frick S, Kranz S, Saadat A. Holistic design approach for geothermal binary power plants with optimized net electricity provision. Proceedings World Geothermal Congress 2010 Bali, Indonesia, 25–29 April; 2010. Gnutek Z, Bryszewska-Mazurek A. The thermodynamic analysis of multicycle ORC engine. Energy 2001;26:1075–82. Hettiarachchi HDM, Golubovic M, Worek WM, Ikegami Y. Optimum design criteria for an Organic Rankine Cycle using low-temperature geothermal heat sources. Energy 2007;32:1698–706. Kanoglu M. Exergy analysis of a dual-level binary geothermal power plant. Geothermics 2002;31:709–24. Kaplan U. Organic Rankine cycles conﬁguration. Proceedings of the European Geothermal Congress 2007, 30 May-1 June, Unterhaching, Germany; 2007. 5 pp. Kose R. Geothermal energy potential for power generation in Turkey: a case study in Simav, Kutahya. Renewable Sustainable Energy Rev 2007;11:497–511. Lund JW. Characteristics, development and utilization of geothermal resources. GeoHeat Center Q Bull 2007;28(2):1–9. Lund JW, Boyd T. Small geothermal power projects examples. Geo-Heat Center Q Bull 1999;20(2):9–26. Mago PJ, Chamra LM, Srinivasan K, Somayaji C. An examination of regenerative Organic Rankine Cycles using dry ﬂuids. Appl Therm Eng 2008;28:998–1007. Moya P, DiPippo R. Unit 5 bottoming binary plant at Miravalles geothermal ﬁeld, Costa Rica: planning, design, performance and impact. Geothermics 2007;36:63–96. Mroczek EK, White SP, Graham DJ. Deposition of amorphous silica in porous packed beds — predicting the lifetime of the reinjection aquifers. Geothermics 2000;29: 737–57. Porras EA, Bjornsson G. The Momotombo Reservoir performance upon 27 years of Exploitation, Proceedings World Geothermal Congress 2010. Proceedings World Geothermal Congress 2010, Bali, Indonesia, 25–29 April; 2010. Schochet DN. Case histories of small scale geothermal power plants. Proceedings of the 2000 World Geothermal Congress, 28 May-10 June 2000, Kyushu–Tohoku, Japan; 2000. p. 2201–4. Schochet DN. Performance of ORMAT geothermal binary and combined steam/binary cycle power plants with moderate and high temperature resources. Renewable Energy 1997;10:379–87. Stefansson V. World geothermal assessment. Proceedings of the World Geothermal Congress 2005, Antalya, Turkey, April 24–29; 2005. Thomsen PA. New technologies for modular geothermal power plants in oil & gas settings. Dallas, Texas: Southern Methodist University; 2006. Valdimarrson P. The Kalina power plant in Húsavik — why Kalina and what has been learned. Electricity generation from Enhanced Geothermal Systems — Strasbourg, France, Workshop5, 14–16 September; 2006. Welch P, Boyle P. New turbines to enable efﬁcient geothermal power plants. GRC Transactions 2009;33:765–72. Yari M. Exergetic analysis of various types of geothermal power plants. Renewable Energy 2010;35:112–21.