Helidyne LLC has developed an innovative, positive displacement rotary TFC expander for use in geothermal power generation. The expander is intended to replace the turbine currently employed in traditional plants as the prime mover. One major advantage offered by Helidyne technology is ability to process brine directly as it emerges from the geothermal well. The "total-flow" strategy makes possible recovery of all available thermal energy by flashing the brine entirely within the expander cavity instead of inside an external flash tank. Direct brine processing potentially doubles plant overall output for a given resource and significantly decreases plant capital cost.
Helidyne's geothermal expander design involves an unique arrangement of four interleaving helical rotors that create a large internal working cavity -see conceptual animation of Fig.1. The rotors are coupled directly to an AC generator, foregoing the extreme gear reduction commonly used in <1MW turbo scenarios, to produce electricity. The Helidyne unit is robust, economical, and capable of removing scale deposits created during the flashing process.
Figure 1: TFC Expander Animation
The simple design offers many advantages over the comparable total-flow twin screw expander (Fig. 2) including: high flow rate with twice the cycles per revolution; high expansion ratio by means of a proprietary liquid injection valve; absence of a high tolerance housing-to-rotor clearance; scalable to large sizes approaching 5MW; oil free design with no physical rotor contact required for synchronization; self-cleaning non-binding rotors, to name a few.
Figure 2: Twin Screw
For decades the industry has understood the potential of a total-flow prime mover, but every attempt has encountered design deficiencies preventing commercialization. Experts have stated:
"Basically, the Total-Flow concept offers the potential for maximum recovery of available energy, but achieving this potential depends on the choice of expander." (Austin & Lundberg, A Status Report on the Development of the Total-Flow Concept, 1978)
Dr. DiPippo, in his text Geothermal Power Plants states:
"The notion of a total-flow system arises from a desire to avoid the irreversibilities associated with the flashing process needed for either a single or double-flash plant." (DiPippo, 2008)
"If a way could be devised to use the geofluid directly as it emerges from the well in the prime mover, be it a turbine or some other specially designed device, significant [cost] savings would be achieved." (DiPippo, 2008)
For this reason, many have considered the total-flow method as the Holy Grail of geothermal power production. Coming to the same conclusion, Helidyne LLC was founded and, after almost four years of development, will demonstrate the performance of its experimental expander on a geothermal well by the end of the second quarter, 2012.
Beginning in the early 1970's, Roger S. Sprankle won a US patent for the concept of using the twin screw expander as a geothermal total-flow prime mover. His strategy relied heavily on scale build-up as a means of improving efficiency by reducing leakage between the rotor-housing running clearance. This approach proved ill conceived with discovery that scale accumulated predominately on the rotors themselves instead of on the desired inner wall of the housing. In some cases, rotor scale was sufficient to cause catastrophic seizure due to zero relative velocity at their point of convergence, as shown in the animation of Fig.3. By the late 1980's, the concept of using a twin screw total-flow device was abandoned altogether for these and other reasons discussed later.
Figure 3: Twin Screw Animation
Recently however, twin screws have found success, not as total-flow devices, but rather in closed loop binary ORC configurations. Companies such as Electratherm now offer commercially available units up to 50kw for low temperature waste heat recovery, solar thermal, and geothermal applications.
Figure 4: 50kw ORC Unit
Helidyne's tehnology is different than the twin screw apparatus. The Helidyne expander operates with an extremely close running clearance (<.002"/50 microns). Excess scale formation is automatically removed by the unique sweeping action of the rotors (Fig 5). Rotors do not bind because scale is immediately removed as each chamber passes through its cycle. In addition, the rotor surface material is non-corrosive and virtually inert to the acids or caustics characterizing the chemistry of typical geothermal brine. Superb running clearances have been demonstrated on existing prototypes, and the manufacturing process is readily scalable to MW sizes.
Figure 5: Self Cleaning Design
A quad-rotor arrangement is preferred for geothermal applications over the tri-rotor geopressure design -see cross section comparison in Fig. 6. Geothermal applications require a high volume flow rate at low pressure, just the opposite of geopressure. Reduced rotor strength of the low pressure 4-rotor geothermal machine is offset by its 4-fold increase in volume flow rate relative to a high pressure 3-rotor geopressure machine of the same overall size.
Figure 6: Cavity Volume Comparison
Another attempt in the 1980's at total-flow design was the bi-phase turbine. Developed and tested by the Douglas Energy Company, the machine involved several nozzles directed tangentially towards a rotating drum containing numerous impulse blades (Fig.7). Pressurized steam is converted to high velocity inside the nozzles which in turn accelerates the liquid portion. Resulting kinetic energy is converted to mechanical energy by directing the liquid/vapor jet stream at an impulse wheel coupled to a generator. The now separated steam is piped to a traditional turbine coupled to the same generator driven by the impulse wheel (Fig.8). The goal here is the same with all total-flow devices: to harness the expansive energy of the flash process instead of losing it in separation tank.
Unfortunately, this approach proved problematic for geothermal applications, but has found success in replacing refrigeration pressure reduction valves and in waste heat ORC applications. Energent is one company at the forefront of this technology.
One of the drawbacks of the concept is the high degree of slippage between the expanding vapor and the liquid causing loss of useable energy despite isentropic flashing inside the nozzles. In addition, erosion of the impulse blades by the jet stream requires costly titanium alloys to minimize wear.
Figure 8: The Bi-Phase Turbine (Unknown, 1997)
By contrast, Helidyne's energy conversion process is more direct. Rather than relying on the dynamic principles of turbo machines, the Helidyne device converts thermal energy directly to mechanical energy via isentropic flashing of brine inside the expander cavity. Vapor pressure acting on the rotors produces torque. A complete thermodynamic analysis of this two phase energy conversion process can be found in the PDF link below. In short, Helidyne's expander is a true total-flow device. The entire expansive energy of the brine is harnessed by the rotors themselves without involving nozzles, impulse wheels or turbines. As stated previously, direct processing enables drastic improvement of overall system efficiency and reduction in plant capital cost.
Plant Type Comparison
To fully appreciate the difference a total-flow expander can make to geothermal plant output, a comparison will be made between conventional plants and the total-flow principle. There are three geothermal plant types: Dry Steam, Flash, and Binary. An elementary description of each can be found at the DOE's website: Department of Energy: Geothermal Technologies Program.
An in-depth commentary about each type and its accompanying loss mechanisms follows.
Dry Steam Plants
Dry steam plants are by far the most desirable, yet require the rarest of geo resources, one where brine flashes deep inside the well to produce high quality (dry) steam upon reaching the surface. In other plant types, brine must be pumped as either saturated liquid or the "mixed-flow" state in order to prevent dissolved solids from precipitating on well walls and plant piping. Consequently, pumping losses in dry steam plants are drastically reduced because steam naturally exits the well at working pressure, leaving only pumping of injection well and cooling water as parasitic loads. Also, because steam contains 15 times more energy per unit weight as water undergoing a 60° F temperature change, pumping loads are proportionately reduced as compared to an equivalent flash or binary plant. Because most known dry steam geothermal resources have already been developed, discussion of this type plant will be omitted hereafter.
Flash Steam Plants
Flash geothermal plants operate by throttling liquid brine from the well into a steam separator or flash tank located at the surface (Fig. 9). The separation process is necessary because traditional turbines are unable to process excessively "wet" steam in the region below the saturation curve on the T-S diagram (Fig. 10). This operating requirement is attributable to the dynamic nature of turbines. Specifically, as vapor expands across the turbine it accelerates to extreme velocities. Kinetic energy of the supersonic vapor is captured by the turbine blades and converted to shaft work. Saturated vapor forms water droplets during isentropic expansion that can severely erodes the blades. Every 1% increase in average moisture content in the vapor causes roughly 1% drop in turbine efficiency. (DiPippo, 2008) Therefore, plant designers strive to obtain high quality steam containing less than 10% liquid.
Figure 9: Flash Plant
Conventional flash plants utilize throttled flashing, a “constant enthalpy” (isenthalpic) process in which half the resource delta-T (high-low temperature difference) is lost. Optimum flashing temperature lies at the midpoint between the source and sink, as depicted in Fig. 10, leaving only half of the original geothermal energy available for producing shaft energy. For a detailed and complete analysis of the flashing process see the PDF link below.
In short, a single-stage flash plant re-injects 50% of the original geothermal energy back underground without performing useful work simply because of the inefficient isenthalpic flashing process. Serial flashing in 2 or 3 serial stages somewhat reduces this loss.
By contrast, flashing under controlled “constant entropy” (isentropic) conditions inside the expander cavity converts the theoretical maximum available thermal energy to useful shaft work.
Figure 10: T-S Diagram for Water
By replacing conventional steam-driven devices with a total-flow expander, the otherwise wasted energy can be harnessed as shaft energy rather than lost in the flash tank (Fig. 19).
A typical flash plant requires well temperatures of at least 360°F to provide a turbine inlet-outlet pressure drop sufficient for acceptable efficiency. However, at the more common lower geothermal temperatures, the saturation pressure of water quickly drops off resulting in low working pressures as shown in Fig. 11 which compares the saturation temperature of water with the corresponding pressure taken from the steam tables:
Figure 11: Saturation Graph
For example, at 240°F the saturation pressure is 10 psig, while at 382°F it climbs quickly to 185 psig. With a 358°F geothermal resource, the turbine inlet-outlet pressure difference is about 38 psi for a condenser temperature of 212°F. By comparison, a 300°F resource, after flashing and under the same condenser temperature, yields only 17 psi pressure difference, less than half that of the 58°F hotter resource. Clearly, small temperature variations affect operating pressures significantly.
Turbine conversion efficiency falls off with decreasing pressure as a consequence of their dynamic nature. By contrast, the energy conversion efficiency of positive displacement expanders is largely independent of the inlet-outlet pressure difference.
Like flash plants, a binary plant follows a thermodynamic cycle involving a phase change. The Organic Rankine Cycle (ORC) used in a binary plant accommodates low temperature geothermal reservoirs by virtue of a low boiling point organic working medium. Brine itself does not flash to steam, but rather exchanges heat with an organic binary fluid such as isobutane or refrigerant R245. The two fluids never mix, being fully contained within their respective closed loops as show in the diagram of Fig.12.
Figure 12: Binary Plant
The primary objective of using an organic working medium is reduction of plant size and cost while enabling efficient turbine operation.
From a theoretical standpoint, physical properties of the working fluid are immaterial to thermal efficiency which is fixed by the laws of thermodynamics as strictly a function of source/sink temperatures and the type of thermodynamic cycle employed, e.g. Brayton vs. Carnot.
However, physical properties of the fluid, particularly the boiling point, do affect power output relative to plant size. Power versus size is a function of fluid properties as demonstrated by the following equation:
P = expander shaft power output
m· = mass flow rate of the organic working medium
Δh = change of enthalpy between expander inlet and outlet
Cp = specific heat of working medium, similar value for most gases
ΔT = temperature drop across the expander, proportional to resource delta-T
Mass flow rate is the chief variable determining shaft power for a given geothermal delta-T. Operating a closed system at higher pressure raises gas density and hence the mass flow rate achievable from a given expander cavity volume and shaft speed. Doubling system pressure while halving the specific volume, say, would approximately halve the expander size while maintaining constant power.
Low boiling point working mediums, typically refrigerants, raise the pressure and density of the working medium by a factor of four or more above a water-based system. High gas density and resulting high mass flow rates reduce machine size by a similar factor.
Applied to a positive displacement expander, such as the rotary screw machine, the concept is understood using the equation for hydraulic force:
Where F is force, P is pressure, and A is the area exposed to the applied pressure. As shown in Fig. 13, the two pistons produce the same force with either high pressure concentrated on a small piston or low pressure applied to a large piston. Thus machine size is inversely proportional to operating pressure.
Figure 13: Pressure vs. Force
Summarizing, the high vapor density of a binary plant permits a smaller turbine to produce the same power at higher efficiency than possible in a much larger water-based system under low temperature geothermal conditions.
As an added benefit, the binary organic vapor expands across the turbine into the superheat region instead of the "wet region" (compare Fig. 10 with Fig. 14) which promotes turbine efficiency and longevity.
Figure 14: T-S Diagram for Organic Fluid
Utilizing an organic working fluid in an ORC plant provides several key advantages in addition to reducing plant size and offering a friendly turbine environment: maintaining brine in the liquid state throughout the entire loop prevents scale formation in the water circuit, especially important in the boiler where scale on exchanger walls impedes heat transfer and compromises efficiency.
By contrast, the Helidyne expander operates efficiently directly on well water without a boiler or organic fluid at very low pressure differences (15 psi). It can achieve high volume ratios, large mass flow rates, and can be manufactured in large MW sizes with lightweight rotors approaching 5 ft in diameter. The device is truly a low pressure, large volume, positive displacement energy converter. This unique capability, along with the self cleaning feature, proves extremely valuable in low temperature waste heat recovery, solar thermal, and solar ponds applications in addition to geothermal.
Despite the advantages of using an organic working fluid, binary plants suffer an efficiency penalty similar to flash plants. Binary boiler losses are theoretically analogous to flash tank losses, each representing a 50% loss of available thermal energy.
Figure 15: ORC Boiler
The diagram of Fig. 15 represents a binary plant boiler. Organic fluid boils at constant temperature by absorbing heat across a large temperature gradient from the brine. Miss-matched sensible-to-latent heat transfer accounts for a 50% loss of the thermal energy, equivalent in magnitude to single-stage flashing as revealed in the thermodynamic analysis (PDF).
To highlight the inefficiency associated with the ORC boiler heat transfer process, consider the sensible-to-sensible counter-flow heat exchanger shown in Fig 16.
Figure 16: Liquid-to-Liquid Heat Exchanger
A very small temperature gradient under this scenario indicates excellent “temperature matching” between the two fluids, minimizing losses associated with irreversible heat transfer. Unfortunately, this efficient sensible-to-sensible process cannot be implemented by the Organic Rankine Cycle (ORC) binary plant which intrinsically relies on sensible-to-latent heat transfer.
In conclusion, a binary system provides practical solutions to problems engineers face when generating power from a low temperature geothermal resource, notably by increasing power output for a given plant size. Nevertheless it suffers the same tremendous loss of available thermal energy as a flash plant.
For decades it has been understood that a total-flow TFC plant is the ideal approach to geothermal power production both from a thermodynamic standpoint and from reduced capital cost. (Austin, Higgins, & Howard, The Total Flow Concept for Recovery of Energy From Geothermal Hot Brine Deposits, 1973) The TFC cycle is the ideal thermodynamic cycle for any two phase system to the same degree that the Carnot cycle is the ideal cycle for any single phase system
As depicted in Fig. 17, brine remains in the saturated liquid state right up to the intake valve of the expander. Brine is isentropically flashed inside the expander where temperature and pressure drop to condenser conditions while simultaneously extracting shaft work in the process. This serves a number of purposes including hindering scale formation in plant piping and implementing the necessarily high expansion ratio required in a water-based system while eliminating feed pumps, receiver tanks, super heaters, mechanical seals, and boilers.
Figure 17: Total-Flow Plant
The “Trilateral Flash Cycle” (TFC) total-flow method takes full advantage of the resource delta-T and its corresponding expansive energy (Fig. 18). Notice in Fig.18 that expansion takes place on the "wet" side of the T-S diagram, just the opposite of a turbine that requires steam to remain dry during expansion. The quality of TFC exhaust vapor is very low, usually <10%.
Figure 18: Total-Flow T-S Diagram
The concept of initiating expansion from the liquid state inside positive displacement devices was first conceived in the 1920's and more recently labeled the “Trilateral Flash Cycle” (TFC) by researchers from City University in London at the Centre for Positive Displacement Technology. The organic rendering of this cycle (OTFC) has been utilized in a twin screw expander achieving performance enhancements reported by the London researchers:
"Provided that two-phase adiabatic expansion efficiencies of at least 75% could be attained, a (TFC) system using light hydrocarbons as working fluids, could recover almost double the power from a hot liquid stream than was possible from either steam or indirectly heated simple ORC systems." (Smith, Stosic, & Kovacevic, 2005)
Helidyne's independent thermodynamic analysis has confirmed this same finding, even when using water as the working medium in an efficient expander. (See PDF link)
It must be noted however, that a machine of practical size has a minimum inlet temperature for water that is higher than organic-based systems. The water-based temperature limits will be discussed later
Efficient twin-screw performance requires the addition of lubrication to the working fluid which precludes open-loop operation, essentially relegating conventional TFC to a closed-loop binary system. The Helidyne machine has no lubrication stipulation, operating oil-free directly on brine in an open system.
As mentioned earlier, the geothermal industry has long recognized the benefits of a total-flow device, and every effort has been made over the last 35 years to adapt the conventional twin screw expander to the task. However, after lengthy assessment researchers conclude that:
"The main reason for the failure of the screw expander in total-flow applications was that the volume ratio needed for efficient expansion of water is too large to be practically attainable." (Smith, Stosic, & Kovacevic, 2005)
In a separate report by the same authors:
"It was soon found that such expanders were unsuitable for the huge volume flow rates and volume ratios required for the expansion of water to normal condensing temperatures."(Smith, Stosic, & Kovacevic, 2005)
"The main obstacle that positive displacement expanders (PDEs), of any design, must overcome is that of a limited volume ratio, i.e., the ratio of the exit volume to the inlet volume." (DiPippo, 2008)
The Helidyne expander design is unique in that it can overcome both obstacles of limited expansion ratio and limited flow rate as cited above.
Concerning volume flow rates, the Helidyne expander has a larger cavity size and delivers twice the cycles per revolution as the twin screw machine of comparable size. Large rotors (<5ft dia.) can be readily manufactured at low cost with clearances largely independent of size.
High Expansion Ratio:
Attaining high expansion ratios in the Helidyne device is understood in reference to the familiar piston/cylinder arrangement shown in Fig.19. The reciprocating model explains Helidyne's rotary arrangement because both machines rely on the same positive displacement principle. As the animation (Fig.19) demonstrates, a metered quantity of brine (red) enters the cylinder under constant pressure.
Figure 19: Piston Cylinder Animation
Isentropic flashing commences when the intake valve closes. A proprietary injection valve (not shown) facilitates variable expansion ratios ranging from 10:1 to 500:1 making it tunable to changing brine and condenser conditions. Variations of seasonal ambient temperatures, and perhaps even the effects of relative humidity in the case of evaporative cooling, are thereby compensated.
Cyclic metering of brine into the individual expander chamber is the key to obtaining expansion ratios of almost any magnitude and Helidyne's 4-rotor design is the only known rotary expander capable of fully implementing the TFC strategy.
Figure 20: TFC Expander Animation
As shown in the animation of Fig.20, metered brine in the liquid state (red volumes) is injected at the expander intake to expand within the cavity before exhausting as a 2-phase fluid (blue volumes). The process is identical to the piston cylinder process shown in Fig.19, only rotary.
The traditional twin screw expander cannot utilize this technique because multiple chambers are exposed to the intake simultaneously, making discrete metering into individual chambers difficult if not impossible. As a consequence, TFC twin screw machines are restricted to organic fluids having inherently low expansion ratios to better match their “built-in” 5:1 ratio.
Helidyne will eventually offer both total-flow water-based and organic-based versions of its expander. Each offers advantages and disadvantages depending on operating temperatures. Generating electricity from low temperature resources involves cost-benefit issues that affect plant performance and economics.
Figure 21: Temperature Comparison
Water-Based TFC Systems
As portrayed in Fig. 21, there is a lower practical limit to well temperatures that can be effectively used in a water-based system. Expanding low temperature water down to sub atmospheric condenser temperatures involves enormous expansion ratios that must be accommodated by the expander to achieve high efficiency.
The scenario shown in Fig. 22 specifies an expander inlet temperature of 250°F. Full expansion down to a condenser temperature of 80°F requires a 5,000:1 expansion ratio resulting in an extremely low machine energy density, i.e. enormous machine size. In other words, expander size would be inordinately large relative to the power produced. While theoretically efficient, the fabrication cost of an oversized machine would render the overall plant economics problematic.
Figure 22: Water Volume Ratio, Pressure, vs. condenser Temperature
Practical machine size under low temperature conditions can be only achieved by limiting the expansion ratio to around 200:1 with a commensurate reduction of machine efficiency due to incomplete expansion. Referring to Fig. 22, a 200:1 expansion ratio needs a condenser temperature of only 180°F. However, significant power gains are still possible at lower condenser temperatures. For example, a condenser temperature of 160°F adds an additional 25% output power. But further power gains at lower condenser temperatures quickly diminish relative to rapidly increasing machine size.
Cost-benefit analysis considers the method of cooling available at the geothermal site in determining the choice of water-based or organic-based. A water-based system may be best for arid regions where air cooling condenses steam at 160°F, a reasonable temperature that minimizes condenser size and parasitic losses. Sites with ground water may consider an organic binary system to take advantage of lower evaporative cooling temperatures (80°F) to produce additional power.
Organic-Based (OTFC) Systems:
The only way to generate power below 235°F in a practical size machine is to replace water with a low-boiling point working medium such as isobutane. Under this scenario the TFC cycle enables efficient liquid-to-liquid heat exchange versus wasteful sensible-to-latent heat exchange of conventional ORC turbine systems. The Helidyne expander implements the TFC principle equally well in organic closed-loop environments. However, unlike a water-based system, the organic cycle requires two heat exchangers to facilitate both the hot and cold sides as shown in Fig.23.
Figure 23: Organic OTFC Diagram
Helidyne achieves a 2-fold gain in plant power compared to standard technologies operating on the same delta T. The 2X improvement arises from the TFC principle, the 2-phase equivalent of the 1-phase Carnot cycle representing the theoretical ideal energy conversion process. This statement applies to both practical-sized water based systems with condenser temperatures above 160°F to enable complete expansion, and to binary systems at lower condenser temperatures.
Plant Efficiency Comparison:
Both ideal and actual conversion efficiencies are shown in Fig. 24 for the various plant types previously discussed. The chart highlights the advantages of a total-flow Helidyne TFC system relative to conventional flash and binary plants.
“Ideal” figures assume zero losses anywhere in the system except for inescapable flashing losses (flash plant) or vaporizing losses (binary). This includes zero heat or piping losses, an ideal isentropic prime mover, and a lossless generator. “Actual” figures take into account the numerous losses associated with a real-world plant.
Figure 24: Plant Comparison Chart
Actual realized conversion efficiencies for 3-stage flash and binary plants are 23% and 13% respectively. (DiPippo, 2008) Such lackluster figures are the consequence of enormous thermal losses sustained by throttled flashing or boiler heat transfer. Implementing a total-flow TFC system without flashing and boiler losses yields conversion efficiencies reaching 40% or higher –an improvement of 2 to 3 times over conventional plant types.
Today's geothermal plants are typically designed using a centralized approach to take advantage of the high conversion efficiency innate to large ( >20 MW) turbo-generator sets. Efficiency gains from a few large centrally-located turbines justify pumping brine through miles of above-ground insulated piping as shown in Fig. 25.
Figure 25: Centralized Plant
In recent years however, a new approach has been introduced of installing smaller, scalable modular units at the wellhead (Fig. 26). Small ORC turbine power plants (<1MW), such as theUTC PureCycle are not sufficiently efficient to make the modular approach economically feasible despite reduction in the pumping load and heat losses of extensive piping networks.
Figure 26: Decentralized Plant
High efficiency of the Helidyne expander makes the modular approach economically feasible, both in terms of reduced capital cost and increased energy extraction from the geothermal resource. Multiple small units are ganged to match the variability of individual wells to reduce construction time and cost and eliminate custom engineering to suit the specific resource.
This so-called "rapid deployment strategy", attempted by Raser Technologies, Inc. in 2008, used 50 small 250 kW UTC unit's in a centralized plant. Lack of sufficient thermal energy from the production wells ultimately doomed the Raser project.
In conclusion, Helidyne's total-flow expander is an attractive alternative to traditional flash and binary methods. The ideal total-flow TFC concept is realized in a simple low cost machine that obviates the enormous thermal losses entailed by throttled flashing or boiler heat transfer losses of conventional plants. TFC efficiency improvement doubles power production without increasing capital costs.
Austin, A. L., & Lundberg, A. W. (1978). A Status Report on the Development of the Total-Flow Concept. University of California, The LLL Geothermal Energy Program. Livermore: Lawrence Livermore Laboratory.
Austin, A. L., Higgins, G. H., & Howard, J. H. (1973). The Total Flow Concept for Recovery of Energy From Geothermal Hot Brine Deposits. UCRL Report No. 51366, Lawrence Livermore Laboratory.
DiPippo, R. (2008). Geothermal Power Plants (2nd ed.). Burlington, MA: Elsevier.
Smith, I. K., Stosic, N., & Kovacevic, A. (2005, 9 22). Power Recovery From Low Cost Two-Phase Expanders. Retrieved 12 30, 2010, from Welcome to the City University Web Server for Staff Personal Pages: http://www.staff.city.ac.uk/~ra601/exp.pdf
Smith, I. K., Stosic, N., & Kovacevic, A. (2005, 9 22). Screw Expanders Increase Output and Decrease cost of Geothermal Binary Power Plant Systems. Retrieved 12 28, 2010, from Welcome to the City University Web Server for Staff Personal Pages: http://www.staff.city.ac.uk/~ra601/grc2005.pdf
Sprankle, R. S. (1973). Patent No. 3,751,673. United State of America.
Unknown. (1997). A Biphase Turbine at a Geothermal Well: Economic Benefits. Caddet Renewable Energy (Technical Brochure No. 52.).