Ammonia Turbomachinery Design Considerations for the Direct Cycle Nuclear Gas Turbine Waste Heat Power Plant ' `

$3.00 PER COPY $1.50 TO ASME MEMBERS 11 1 The Society shall not be responsible for statements or opinions advanced in papers or in discussion at meetings of the Society or of its Divisions or Sections, or printed in its publications. Discussion is printed only if the paper is published in an ASME journal or Proceedings. Released for general publication upon presentation. Full credit should be given to ASME, the Technical Division, and the author(s).


INTRODUCTION
With added emphasis in the United States on fuel conservation and minimum impact on the environment by future power generating systems, there is considerable interest in using waste heat from the GT-HTGR power plant in a secondary power cycle that can raise the overall efficiency of the combined cycle into the range of 50 percent.The attraction of usefully employing the relatively high-grade waste heat from a nuclear closed-cycle gas turbine goes beyond that of achieving higher efficiency.Because the extra power is obtainable without major penalty to the output of the primary helium turbine system, not only is this extra power free of fuel cost, it is also free of the capital cost of most of the rest of the nuclear plant.
The available reject temperature from a nuclear gas turbine of around 250 C (482 F) is too low for utilization of steam in the bottoming cycle, because its very low vapor pressure would result in unacceptably large components.
Previously reported studies (1, 2 ) 1 concluded that supercritical operation of the waste-heat cycle was highly desirable, and while several candidate fluids thermodynamically satisfied the compatibility between the primary system heat rejection characteristics and the goal of maximum bottoming cycle efficiency, it was the size of the power system components, together with chemical stability, that lead to the choice of ammonia as the working fluid.Other principal factors favoring its selection were its good heat transfer properties, high effective specific heat, and high vapor pressure at the sink temperature.With the sizing of the major corn-1 Underlined numbers in parentheses designate References at end of paper.ponent items (turbine and condenser) playing an important role in the selection of the working fluid, this paper presents the design considerations that lead to the conceptual design of the ammonia turbine and pump.Design, performance, and operational aspects of the GT-HTGR waste heat plant, together with component design considerations, have been presented previously (3,4).
The conceptual designs of the ammonia turbomachinery were done for an advanced version of the GT-HTGR plant in an effort to evaluate the economic and performance potential.The advanced GT-HTGR with twin 4000-M6T(t) reactors has a single waste-heat plant, and the reject heat from the plant is dissipated to the environment in natural draft evaporative cooling towers.The projected plant efficiency for this twin-reactor advanced design (L) is 50 percent, the power from the ammonia plant being 920 MWe.The conceptual designs outlined in this paper were done in sufficient detail to: (a) establish their feasibility, (b) satisfy performance and economic requirements, (c) seek industry participation for more detailed design and cost studies, and (d) to assist in plant layout studies.

WASTE HEAT PLANT CYCLE CONDITIONS
The plant performance data are based on a twin 4000-MW(t) reactor arrangement with a core outlet temperature of 950 C (1742 F).The cycle conditions for the GT-HTGR plant with the ammonia bottoming cycle are shown in Table 1.With the heat rejection from the twin reactor arrangement being on the order of 4800 MW(t), the bottoming cycle with an efficiency of just over 19 percent contributes 920 MW(e) to the overall plant output.Cycle conditions and component arrangement are illustrated schematically in Fig. 1.The helium Brayton cycle with recuperation is combined with a supercritical ammonia Rankine cycle for evolution into the binary GT-HTGR plant.Some of the design parameters for the binary plant are set primarily by the design conditions that have been established for the dry-cooled plant (6,7).This is to avoid extensive redesign of the primary system components, the details of which have been discussed in an earlier paper ( 8 ).
Evaporative cooling was assumed as the method of heat rejection from the binary plant, and the wet bulb temperature taken is 9.7 C (49.5 F), which corresponds to an International Standards Organization (ISO) day dry-bulb tem-perature of 15 C (59 F) and 50 percent relative humidity.An ammonia condensation temperature of 32.2 C (90 F) was selected based on an optimization of the entire heat rejection system.
The secondary cycle is basically a supercritical Rankine power cycle as shown in Fig. 2; thus, its components, operation, and controls are similar to Rankine cycle steam power plants.The ammonia turbine inlet pressure of 16.1 MPa (2334 psia) was selected from system studies, thus involving certain criteria to be established (as outlined in the following) to ensure component thermodynamic compatibility for the fixed ammonia turbine inlet temperature of 250 C (482 F).From the temperatureentropy diagram in Fig. 2, it can be seen that for the supercritical cycle, the ammonia precooler heating curve is well matched to the helium heat rejection curve.There is however, a pinch-point in the heat source exchanger which has a significant influence on the surface area requirement of this unit, and, in the cycle studies, a minimum pinch point temperature difference of 5.6 C (10 F) was imposed to ease the thermal conductance requirements of the precooler.With the condensation temperature (hence pressure) established from plant heat rejection characteristics, another important criteria was that of establishing the turbine exhaust condition very close to the saturated vapor line.A few degrees of superheat would have an adverse effect on the already large condenser ( 4 ), and overexpansion in the turbine would result in efficiency loss and erosion problems because of the moisture content.From the pressure-enthalpy diagram in Fig. 2, it can be seen that with a turbine exit quality of 99.8 percent and the aforementioned criteria, the turbine inlet pressure of 16.1 MPa (2334 psia) was established.
The supercritical ammonia fluid is heated in the primary system plant precoolers which are located inside the prestressed concrete reactor vessel.The heated fluid from the precoolers in both reactor units is transported to the single waste heat (secondary cycle) plant.Multiple shell-and-tube condensers reject the heat from the cycle.The secondary cycle ammonia flow is circulated by turbinedriven feedpumps.A surge tank upstream from the feedpumps isolates pressure transients in the system.A simplified flow diagram for the waste heat plant is shown in Fig. 3. Safety studies have shown that the system with direct helium-to-ammonia heating as shown in Fig. 3 is acceptable; however, the turbomachinery de-

TURBINE DESIGN CONSIDERATIONS
In the cycle calculations, the efficiency of the ammonia turbine was assumed to be an optimistic value of 90 percent.The ammonia turbine design outlined in this paper is conceptual in nature and was carried out to: (a) check the efficiency level used in the cycle calculations, (b) develop an understanding of some of the performance and mechanical design problems, (c) solicit vendor inputs in the areas of feasibility and cost, and (d) enable the preparation of a binary cycle plant layout to be made.While there are substantial data in the published literature on the evaluation of different working fluids for electrical power producing Rankine plant (9-l2), there is a paucity of information on ammonia turbomachinery, particularly pursuance of studies leading to a turbine conceptual design.
One of the early design considerations for the binary plant involved the selection of the number of turbines.Realizing the economic merits of a single unit, a conceptual design of a 920-MW(e) turbine was completed.A review of this design concluded that it was indeed technically feasible, but from the overall plant availability standpoint, two turbines would be more attractive.Accordingly two turbines were adopted with a unit rating of 460 MNT(e), and it is details of this ammonia turbine that are presented in this paper.The next major decision involved the choice of a single-or double-flow configuration.A doubleflow arrangement was chosen for the following reasons: (a) it is essentially self-balancing, thus eliminating complex thrust balancing devices; (b) it results in low values of blade stresses; and (c) the double-flow exit system is well suited to the piping configuration for transporting the vapor to the multiplicity of condenser shells.These considerations were viewed as more important than the main merit of the single-flow system which lies in the area of reduced number of disks, blades, and machine structure.A literature search identi- 137.9 (20,000) 137.9 (20,000) r:7 Downloaded From: https://proceedings.asmedigitalcollection.asme.org on 12/02/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use ph fied only one publication on a large electrical power producing ammonia turbine design study (13), although ammonia has been evaluated for smaller ocean thermal energy conversion systems (1a-).The split-flow configuration de- cision is consistent with the 230-MW(e) ammonia turbine conceptual design reported in Reference (13).
The high density working fluid in the ammonia secondary cycle results in a very compact power conversion system, compared with steam turbines and open-cycle industrial gas turbines.The ammonia turbine is characterized by short blades, and high hub-to-tip ratios, that result from a combination of the properties of ammonia and the high degree of pressurizaton, particularly at the turbine exit.The waste-heat plant cycle conditions have been selected so that the ammonia is drysaturated (99.8 percent quality) at the end of expansion, so that in the absence of wetness, the aerodynamic design procedures used are identical to conventional gas and high pressure steam turbine practice.Design conditions for the ammonia turbine are given in Table 2.

TURBINE CONCEPTUAL DESIGN
In previously reported studies (12), rigorous efforts have been made to establish turbine "merit parameters" so that the influence of working fluid properties on the size of the turbine can be evaluated.It is not the purpose of this paper to duplicate this type of effort, but rather to present a practical turbine conceptual design.However, since the size and configuration of the turbine for the GT-HTGR waste heat plant did indeed influence the choice of ammonia as the working fluid, a brief discussion of this is included in the Appendix.The simplistic relationships show that a fluid with a high ratio of vapor pressure to molecular weight, together with a large enthalpy difference between the cycle sink and temperatures, is desirable.
The conceptual design of the turbine was done with the assistance of a comprehensive computer code which has been used extensively for steam turbine design.With the substitution of ammonia properties for steam, this code, which embodies thermodynamic, aerodynamic, and structural subroutines, was used for the turbine blading design.At the onset of the design process, it would initially appear that there is considerable flexibility in the selection of design parameters.However, before the first computer run was made, a brief review of the aerodynamic and structural criteria revealed that, in fact, the choice of basic parameters was indeed very limited to satisfy the requirements of high efficiency with accepted design practice.Some of the important aerodynamic and mechanical variables for turbine design are given in Table 3.

Turbine Aerodynamic Design
Before using the design code, some simplifying assumptions can be made to ensure selection of parameters for maximum efficiency.A summary of the various turbine parameters that will be mentioned in the following is given in Table 4. Unlike gas turbines, the ammonia turbine is not limited by the stresses in the first-stage blading because of the low inlet temperature.However, it is still necessary to have as high a blade speed as possible in order to reduce the number of stages.There are several constraints on the design, some of which are imposed on the turbine by the plant conditions, such as shaft speed.Other constraints are imposed on the turbine by the requirements of a good design.Considering the rear stage of the turbine, two criteria can be used for determining the exit flow area: (a) limiting the centrifugal stress to an acceptable value.Application of this criteria is simplified here since the rotational speed is fixed at the generator synchronous speed of 3600 rpm, and with centrifugal stress being proportional to annulus area times the square of the rotational speed, the maximum allowable exit annulus area is established, and (b) establishing a leaving axial velocity so that the exit kinetic energy is less than 5 percent of the turbine enthalpy drop.
With the rotational speed fixed, the diameter of the turbine (hence blade speed) is initially selected by reviewing the stage loading and flow coefficient.Correlation of efficiency can be based on the stage loading and flow coefficient.Such a simple correlation gives a measure of the vector diagram on stage efficiency, although it must be appreciated that the detailed choice of design parameters, such as blade loading, profile shape, aspect ratio, tip clearance, may all affect the overall efficiency.Performance analysis of many gas turbine rig tests produces an empirical relationship between stage loading factor, flow coefficient, and efficiency.Fig. 4, constructed from test data (1), shows this in a simplified form.With an initial assump-

Tip clearance
Blade height tion of eight stages, it can be seen that the aerodynamic factors are consistent with a high efficiency design.Although the efficiency values shown in Fig. 4 do not include allowances for tip clearance and leakage losses, the position of the efficiency contours is not significantly changed by these factors, and the figure is useful as a preliminary design tool prior to using the turbine design code.Superimposing the same turbine aerodynamic loading factors on similar curve arrays for typical high-pressure steam turbines (16) shows that the 90 percent efficiency goal should be realizable for a conservatively designed multistage axial ammonia turbine.The map of specific speed versus specific diameter for axial turbines (1) shown in Fig. 5 is particularly useful at the conceptual design stage.Initial calculations are based on operating in the middle of the maximum efficiency With the foregoing preliminary design considerations giving the turbine designer an appreciation for influence of the basic parameters, a comprehensive computer survey using the code was initiated to evaluate the blading design for differing number of stages and gas flow path geometries.Details of the selected eight-stage design are shown in Table 2.The first-stage blade height is 58 mm (2.25 in.), the last stage blade height is 126.5 mm (5.0 in.), and with a rear stage tip diameter of 1.46 m (57,5 in.), the maximum tip speed of 275 m/sec (903 ft/sec) is very modest compared with contemporary steam turbines.The degree of reaction for the selected design is 0.55.With relatively low values of blade aspect ratio, the maximum bending stress is only 137.9 MPa (20,000 lb/in. 2 ), and while bending stress per se is not particularly important, it does, at the conceptual design stage, give an indication of the blades ability to withstand flexural vibration.Combined with the centrifugal stress of 88.5 EPa (12,840 lb/in. 2 ), the overall structural loading is consistent with the turbine design life goal of 40 years.

Turbine Performance Estimates
The aforementioned turbine design computer code also embodies a subroutine for performance estimation, including off-design values.The code evaluates blade geometries in each stage, stage efficiencies, and overall turbine (isentropic) efficiency.The various loss correlations used in the code have been well documented (18, 12), From the aerodynamic studies, it was concluded that the minimum number of stages to satisfy the 90 percent efficiency goal was eight.While the blade heights are small (a tip clearance of two percent of the blade height being used in the analysis) and the hub-to-tip ratios are high, two favorable parameters from the standpoint of high efficiency are the high Reynolds number (10 x 10 6 ), and the fact that the Mach num- ber in the turbine does not exceed about 0.70.
The annulus flare is modest, more resembling a gas turbine than a steam turbine, and close control of the blade tip clearances should be possible for this low temperature machine.
Detailed structural analyses are necessary, however, to verify the validity of the foregoing tip clearance assumption, but this was clearly beyond the scope of the described conceptual design activities.If larger tip clearances are necessary, this would have an adverse effect on the turbine efficiency.

Turbine Mechanical Design
A simplified view of the double-flow, eight-stage ammonia turbine is shown in Fig. 6.
The overall length and width are 6.1 m (20 ft) and 3.2 m (10.5 ft), respectively.The 460-MW(e) turbine has four exits as shown, each duct transporting ammonia vapor to a bank of three condenser shells.The turbine layout was prepared to get an appreciation for the overall size and to assist in binary cycle balance of plant studies.In this conceptual phase of the program, the mechanical design aspects of the rotor, bearings, seals, and structures were not addressed, and indeed in these areas, inputs from a turbine company with experience in large power generating machinery are necessary.The choice of bearing type depends on various factors, including dynamic loads.Since the rotor dynamics were not evaluated at this stage of the design, a two-journal bearing an- I rangement with pressure lubricant feed was nominally chosen.Because of the high inlet pressure, a double casing arrangement has been employed, with intermediate pressure ammonia (from the fourth stage) bled into the space between the casings.From Fig. 6, it can be seen that industry practice of splitting the casings horizontally has been adherred to.A thrust bearing has been provided, although the thrust unbalance has been minimized because of the double-flow arrangement.While not engineered at this conceptual design stage, static and dynamic seals are necessary to prevent excessive ammonia leakage during operation and while the turbine is shut down.Because of the low operating temperature, the use of existing low to medium alloys is possible.A typical high-strength low-alloy ferritic steel is suitable for the rotor.For the blades, a 12 percent chrome steel would have the necessary corrosion resistance properties for operation in the ammonia environment.The general response from the turbine industry was, that while an ammonia turbine has not been built and operated, its design would be well within the state-of-the-art, and, with a molecular weight very similar to steam, existing aerodynamic data would be applicable.In Fig. 7, a comparison is shown between a modern reheat steam turbine (20) and the conceptual ammonia turbine outlined in this paper.
From this figure and Table 5, it can be seen that, from the overall configuration and size standpoints, there is considerable similarity.This substantiates the fabricability of the ammonia turbine since existing manufacturing techniques could be directly applied.

AMMONIA FEED PUMP DESIGN CONSIDERATIONS
Unlike boiler feed water pumps in contemporary steam plant, the ammonia pump for the binary plant is not an "off-the-shelf" component, and to assist in balance of plant studies and cost estimates, a conceptual design was carried out.With a secondary plant total   pumping power requirement in excess of 100 NW(e), an ammonia turbine drive for the pump was selected, and the complete pump package is briefly outlined in this paper.For the conceptual design, four parallel pumps were selected, each with a flow in the order of 1.47 m 3 (23,000   gpm), which is similar to existing boiler feed water pumps used in large commercial steam power generating plants.In the cycle calculations, an ammonia pump efficiency of 85 per- cent was assumed, and one of the goals of the conceptual design study was to confirm the feasibility of realizing this efficiency level.
The major problem associated with the subsystem comprising the drive turbine and the ammonia feed pump is the low volumetric flow associated with the drive turbine if high pressure ammonia from the precooler is used.Several alternatives were considered, including the use of radial inflow turbines and partial admission turbines.It was decided to bleed the ammonia from the fifth stage of the main turbine to arrive at an adequate volumetric flow to make a single axial stage drive turbine feasible.Each of the four pumps has a power requirement on the order of 27 MW(e).In order to achieve an adequate NPSH (Net Positive Suction Head), it is necessary to provide for a booster pump, driven off the same shaft as the drive turbine, through a reduction gear.The power requirements of the booster pump are only of the order of 10 percent of the total pumping power.In an earlier study (4), the selection of a low rotational speed, comparable with present-day high performance feed pumps, resulted in six stages being necessary for the centrifugal pump.The trend in the industry ( 21) is toward minimizing the number of stages with higher rotational speed.To satisfy the high efficiency requirements from the basis of specific speed-diameter relations (22), a twostage centrifugal design with an impeller diameter of 384 mm (15.1 in.), in conjunction with a rotational speed of 9000 rpm, was selected.
Ammonia Pump Conceptual Design Design data for the ammonia pump are given in Table 6.Since the pressure at the suction side of the pump is the vapor pressure of ammonia at that temperature, the NPSH available is the static head between the ammonia

1W
surge tank and the pump, which is of the order of 24.4 m (80 ft).NPSH requirements for a specific pump are set by the necessity of avoiding cavitation.NPSH is defined as the absolute total head at the pump suction less the head equivalent of the vapor pressure.The avoidance of cavitation is usually expressed in terms of the Thoma cavitation coefficient and the suction specific speed (23).The NPSH required at the best efficiency point for a given throughput varies approximately as the 4/3 power of the speed.It follows that employment of higher speed entails either increasing the suction pressure to the pump or drastically improving the design of the first-stage impeller and its suction performance.For the conceptual design, the first solution was adopted and a single-stage slow speed booster pump is interposed between the surge tank and the highspeed feed pump.From Table 6, it can be seen that the head for the booster pump is 244 m (800 ft), and this is achieved with a single impeller of 908 mm (36 in.) dia rotating at 1500 rpm.As can be seen in Fig. 8, the booster pump is driven via a reduction gearing from the drive turbo-pump shaft.This drive method ensures that when the feed pump is started so is the booster.The head for the feed pump is 2712 m (8898 ft) and this is achieved with a two-stage impeller, the fluid flow path being clearly seen from Fig. 8.In consultations with the pump industry, it was confirmed that a pump of this type has the potential for satisfying the design goal efficiency of 85 percent.
In-depth studies of the pump mechanical design features, such as the bearings and seals, were not done during the conceptual design, and these areas including axial thrust compensation for the two-stage machine would be resolved by specialist vendors in a followon design effort.

Puma Drive Turbine Conceptual Design
Design data for the ammonia pump drive turbine are given in Table 7.With 204.7 kg/ sec (451.3 lb/sec) of ammonia bled of the fifth stage of the main turbine at 155.2 C (311.4 F) and 6.11 MPa (886 psia), the turbine enthalpy drop of 150 KJ/kg (64.5 Btu/lb) corresponds to a power output of 27 MW(e).Using the ammonia turbine design code mentioned earlier, a blading design was done for a single-stage turbine.With a hub diameter of 782 mm (30.8 in.), the blade height at the exit is 41 mm (1.6 in.), and the blade tip speed of 407 m/sec (1335 ft/sec) is conservative compared with modern steam turbine practice.The projected efficiency for this single-stage ammonia turbine is 85 percent.An important consideration in the design of a single-stage unit is the fraction of the enthalpy drop leaving the turbine in the exhaust.Even with the use of a high efficiency diffuser (75 percent), the losses in the exhaust can affect overall performance, and

(15)
Approx.Overall Height, m (ft) 1.8 (6) in any follow-on effort on this turbine design, a two-stage unit would be evaluated.With an overall length and height of 4.6 m (15 ft) and 1.8 m (6 ft), respectively, it can be seen from Fig. 8 that the ammonia feed pump package is a compact unit.The design, fabrication, and operability of the feed pump unit represent existing technology, and while detailed design is necessary, the development effort to satisfy the high efficiency design goal should be minimal.

AMMONIA TURBINE BUILDING LAYOUT
The sizing of the major equipment items for the waste heat plant were done for the following reasons: (a) to establish feasibility, (b) to assist in balance of plant layouts, and (c) to enable cost data to be generated.A study was initiated to determine the sizes and geometries of the interconnecting ammonia pipework for the waste heat plant.The study consisted of component orientation layouts to establish the best equipment positions to satisfy criteria in the areas of: (a) minimum building size, (b) simple piping arrangement, and (c) component accessibility for maintenance and installation/replacement.Fig. 9 shows a conceptual arrangement of the ammonia turbine building.The two turbines are shown positioned at an elevated location.Each turbine exit discharges the ammonia vapor (dry and saturated) vertically downward to a bank of three condensers.From Fig. 9, it can be seen that the building is dominated by the bulk of the 24 condenser shells.The other major

SUMMARY
The analytical and conceptual design studies of an ammonia turbine presented in this paper lead to the selection of an eight-stage double-flow axial turbine configuration with a rating of 460 MW(e).The stress levels in this machine are commensurate with an operating life of 32 years (40 years with a plant duty factor of 80 percent).With the eight-stage conceptual design, the aerodynamic loading is low, and with the absence of moisture, the 90 percent efficiency goal should be realizable.The num-ber of stages could be reduced, but with a penalty in efficiency, and since the object of the design study was to evaluate the potential of the binary cycle for an advanced GT-HTGR plant, the selection of eight stages is justifiable for a high efficiency machine that is expected to operate for the full life of the nuclear plant.
With the aerodynamic and structural constraints imposed on the design (for high performance and long life, respectively), it is felt that the annulus gas flow path geometry (hence overall machine size) would not change significantly with specialist turbine vendor design.Details of the blading, however, may indeed change since the turbine industry has extensive cascade data permitting optimum blade geometry selection for maximum efficiency.With the high degree of pressurization, particularly at the turbine exhaust, the ammonia turbine conceptual design is a very compact unit, and analogous to modern reheat steam turbines.Since the hub-to-tip ratio is not excessively low in the near stages, the blade roots are operating at a relatively high value of reaction; this permits the avoidance of highly cambered blades.The high density of the ammonia working fluid results in small annulus area requirements, and the small blade heights are of concern, since the turbine efficiency is affected by tip clearances effects and secondary flow losses.
While extensive design and development programs are needed for the ammonia turbomachinery, existing technology from the power generating industry is generally applicable, and with specialized design attention, the conceptual turbine and pump designs p , esented in this paper have the potential for satisfying the high efficiency goals assumed for the cycle analysis.
For the low temperature secondary cycle, no materials technology advancements are necessary, and state-of-the-art design and fabrication methods are applicable.The overall configurations of the ammonia turbine and pump are very similar to power conversion systems being used in modern steam plant.This conventionality of the rotating machinery results in acceptable specific capital costs (^/kwe).
Experience from the chemical and refrigeration industries in the areas of seals, maintenance, and the handling of ammonia components should be generally applicable, thus obviating the learning process associated with other organic and exotic fluids being considered for bottoming cycles.
With a heat source temperature of 250 C (482 F), the system and component studies have shown that a creditable ammonia plant efficiency close to 20 percent is possible.While these studies were directed toward a nuclear closedcycle gas turbine, the binary plant turbomachinery design work presented in this paper could be applied directly to other low-grade temperature systems, such as geothermal, solar devices, and various industrial and marine heat engines and prime-movers.

APPENDIX INFLUENCE OF FLUID PROPERTIES ON TURBINE SIZE
There are several working fluids (including isobutane) which, from the thermodynamic standpoint, yield high efficiency in the GT-HTGR binary plant, however, their character-istics result in very large and impractical components (particularly the turbine and condenser).The very simple relationship established subsequently for the turbine gives some guidance as regards fluid property selection.The symbols are defined in the nomenclature.
The centrifugal blade stress (for untapered steel blades) is given by: f = l.9(-j-N ) 2 A (1) Cŵ here Id = rpm, annulus area A is in ft 2 , and the stress f c is in lb/in. 2 .
The turbine exit area is fixed by the allowable exit Mach number, since this determines the exhaust recovery pressure ratio.With a desire to minimize blade centrifugal stress, the influence of working fluid properties on the turbine area is developed as fol- Fluid sonic velocity = fg R To M (5) and Y JRC pm (6) and substituting sonic velocity becomes g R To iI M 1-R (7)

JCpm
For complex molecules Cpm is high and the or second bracketed term is small compared with unity.This means that y is about constant, and sonic velocity is proportional to T M So, for a constant exit Mach number, the required area, A, is given by: 11 M (8) Pvo Cpm This relationship simply suggests that from the turbine standpoint a fluid with a high ratio of vapor pressure to molecular weight should be selected.
It might seem surprising that the fluid molecular weight appears in the numerator of equation ( 8), implying advantage for low molecular weight.This is because, if Cpm is regarded as constant, the volume flow at a given pressure is fixed.The term /T71 then appears only by way of its influence on sonic speed, lighter gases being allowed a faster velocity for the same Mach number limitation.
Another important fluid selection criteria is the total enthalpy rise per pound between the cycle sink and source temperatures.The required mass flow rate (directly influencing pumping power, component size, and liquid flow pipe size) is inversely proportional to the enthalpy difference.A large enthalpy difference; hence, reduced mass flow is desirable since it minimizes pumping power and results in reduced equipment size.While this statement implies an increase in the number of turbine stages, the conceptual design described in the paper has shown an acceptable eight-stage configuration is possible with ammonia as the working fluid.
The aforementioned simplistic rationale alone suffices to rule out at once isobutane and a number of the better known Freon refrigerants for bottoming cycles, such as the GT-HTGR waste heat plant with power outputs in the hundreds of megawatts.

Fig. 7 4 12
Fig. 7 Comparison of features between ammonia turbine and modern reheat (intermediate pressure) steam turbine

rFig. 8
Fig. 8 Conceptual design of turbine driven ammonia feed pump

Fig. 9
Fig.9Turbine building layout for GT-HTGR waste heat power plant So for a given sink temperature and heat input, the exhaust flow is proportional to 1 Pvo Cpm

Table 3 Turbine
Design Variables

Table 4
Nondimensional Turbine Parameters for Preliminary Sizing

Table 5
Comparison of Ammonia Turbine and Modern Intermediate Pressure (Reheat) Steam Turbine * Steam Turbine data courtesy of BBC, Brown Boveri and Company, Limited

Table 6
Details of Ammonia Pump(s) Conceptual Design AMMONIA FEED PUMP

Table 7
Details of Ammonia Pump(s) Drive Turbine