CONCEPTUAL DESIGN
RANKINE-CYCLE POWER SYSTEM
WITH ORGANIC WORKING FLUID
AND RECIPROCATING ENGINE
FOR PASSENGER VEHICLES
JUNE. 1970
THBRMO ELECTRON
CORPORATION
-------
TMKMKIO BIBCTHON
Report No. TE4121-133-70
CONCEPTUAL DESIGN
RANKINE- CYCLE POWER SYSTEM
WITH ORGANIC WORKING FLUID AND
RECIPROCATING ENGINE
FOR PASSENGER VEHICLES
by
Dean T. Morgan and Robert J. Raymond
Thermo Electron Corporation
Research and Development Center
101 First Avenue
Waltham, Massachusetts 02154
June, 1970
Prepared for
Division of Motor Vehicle Research and Development
National Air Pollution Control Administration
Public Health Service
Department of Health, Education and Welfare
Ann Arbor, Michigan 48104
Contract No. CPA 22-69-132
-------
THBRMO ELECTRON
ACKNOWLEDGEMENTS
In addition to personnel from the Rankine Power Systems
Department of Thermo Electron, the following organizations have
made significant contributions to the program in the primary areas
noted:
1 Ford Motor Company, New Business Development
Office, Scientific Research Laboratory, and
Powertrain Research Department:
Vehicle integration, system performance calculations,
manufacturing considerations, large volume manu-
facturing cost estimates, system and design reviews.
2 Dana Corporation, New Product Design, Corporate
Research and Development:
Transmission design, system performance calculations.
3. Marquardt Corporation:
Burner design parameters.
4. American Oil Company, Research and Development
Department:
Fuel considerations,
5. Control Design, Incorporated, and
F. D. Ezekiel Company:
Controls.
6. Clevite Corporation:
Engine bearing analysis,
7. British Internal Combustion Engine Research Institute,
Ltd. , and American Bosch Company:
Engine intake valving:
8. Monsanto Company, Functional Fluids Division:
Thiophene (Monsanto Cp-34) working fluid characteristics.
ii
-------
THBWKIO BCBCTM ON
TABLE OF CONTENTS
Section Page
1 ABSTRACT 1-1
2 SUMMARY, CONCEPTUAL DESIGN OF A RANKINE-
CYCLE PROPULSION SYSTEM FOR PASSENGER
VEHICLES 2-1
2. 1 INTRODUCTION 2-1
2.2 COMPONENT DESCRIPTIONS 2-2
2.2. 1 System Design Point Characteristics . . . 2-2
2.2.2 Engine (Expander) Design . . . 2-2
2.2.3 Feedpump Design 2-6
2^2.4 Burner-Boiler Design 2-8
2.2.5 Condenser 2-11
2.2. 6 Regenerator 2-11
2.2.7 Automatic Transmission 2-12
2.2.8 Controls 2-12
2,3 SYSTEM PERFORMANCE AND PACKAGING ... 2-13
2.4 EMISSION PROJECTIONS FOR RANKINE-CYCLE
AUTOMOTIVE PROPULSION SYSTEM ; 2-18
2. 5 MAJOR CONCLUSIONS 2-18
REFERENCES 2-19
3 INTRODUCTION 3-1
3. 1 OVERALL GOALS 3-1
3.2 APPROACH FOLLOWED FOR ATTAINMENT
OF GOALS 3-5
REFERENCES 3-16
iii
-------
THERMO ELECTRON
CORPORATION
TABLE OF CONTENTS (continued)
Section Pa.ee
4 COMPONENT CHARACTERISTICS AND
DESCRIPTIONS 4-1
4. 1 INTRODUCTION .' 4-1
4.2 SAFETY CONSIDERATIONS FOR THIOPHENE
WORKING FLUID 4-8
4.2. 1 Introduction 4-8
4.2.2 Flammability and Toxicity of Thiophene . . 4-8
4.2.3 System Design Concepts to Minimize
Hazard from Flammable and Toxic
Working Fluid 4-14
4.3 ENGINE (EXPANDER) DESIGN 4-20
4.3. 1 Performance Estimates 4-20
4.3.2 Engine Configuration 4-27
4.3.3 Expander Intake Valving 4-29
4*3.4 Expander Exhaust Valving 4-37
4.3.5 Engine Bearings 4-39
4.3.6 Final Expander Design 4-42
4.4 FEEDPUMP DESIGN 4-48
4.5 COMBUSTOR DESIGN AND CHARACTERISTICS . 4-54
4.5.1 Combustor Design and Fuel/Air Supply . . 4-54
4.5.2 Emission Levels from Rankine-Cycle
Burners 4-65
4.5.3 Some Considerations in Fuel Selection . . 4-74
4.6 BOILER DESIGN 4-77
4.7 CONDENSER DESIGN 4-90
4.8 REGENERATOR 4798
iv
-------
TMKRMO «L«CTItOH
TABLE OF CONTENTS (continued)
Section Page
4 4. 9 ROTARY SHAFT SEAL AND STATIC SEALS .... 4-104
4. 9. 1 Rotary Shaft Seal . ... 4-104
4.9.2 Static Seals 4-108
4. 10 AUTOMATIC TRANSMISSION 4-111
4. 10. 1 Slipping Clutch . 4rll 1
4.10.2 Torque Converter 4-112
4.11 CONTROL AND STARTUP OF SYSTEM 4-113
4. 11. 1 Controls for System Operation ........ 4-113
4.11.2 System Startup .. .. 4-120
4.11.3 Safety Controls . .... 4-131
REFERENCES „ 4-133
5 SYSTEM DESIGN AND EVALUATION 5-1
5. 1 INTRODUCTION .... ,.... 5-1
5. 2 PERFORMANCE IN REFERENCE AUTO-
MOBILE 5-12
5. 3 PACKAGING OF SYSTEM AND SYSTEM
WEIGHT 5-24
5.4 EMISSION LEVEL FROM THE SYSTEM 5-31
5.5 RELATIVE COST COMPARISON WITH
302-2V FORD ENGINE ^ . . 5-33
5.6 GENERALIZED COMPUTER MODEL 5-34
6 CONCLUSIONS 6-1
-------
THERMO E I. H C T R O M
COIPOIATIOH
TABLE OF CONTENTS (continued)
Appendix Page
A PARAMETERS FOR CHARACTERIZING
FLAMMABILITY CHARACTERISTICS OF
MATERIALS A-l
B API TOXICOLOGICAL REVIEWS OF TfflOPHENE
AND DERIVATIVES AND GASOLINE B-l
VI
-------
THRMMO BLBCTPIOM
1. ABSTRACT
A conceptual design has been prepared of a Rankine-cycle power
system with organic working fluid and reciprocating engine for a low-
emission automotive propulsion powerplant. The goal of the study was
development of a system competitive in cost, performance, and driver
convenience with the internal combustion engine system using current
technology wherever possible.
The component designs and characteristics are presented. The
complete 100 net shaft horsepower system is packaged in the engine
compartment of a 1969 Ford Fairlane; the predicted performance
characteristics are presented. The system is closely competitive,
in 0 - 60 mph acceleration time and in level-grade top speed, with
a 302 cubic inch displacement internal combustion engine with three-
speed transmission. The fuel economy in customer-average mpg is
approximately 20% less than the 302 cubic inch internal combustion
engine.
The use of thiophene as a working fluid, with a moderate maxi-
mum cycle temperature of 550 *F, permits a significant cost reduction
relative to the equivalent steam system. This reduction may permit
the Rankine-cycle system to be competitive costwise with the equiva-
lent internal combustion system, particularly since the stricter
emission level standards will require significant cost increases in
the future in the internal combustion system.
Projection of current burner data indicates a strong potential
for emission levels significantly less than the projected 1980 federal
standards for all three of the major pollutants: unburned hydrocarbons,
carbon monoxide, and nitric oxide.
1-1
-------
THERMO ELECTRON
2. SUMMARY
CONCEPTUAL DESIGN OF A RANKINE-CYCLE
PROPULSION SYSTEM FOR PASSENGER VEHICLES
2. 1 INTRODUCTION
Great emphasis is being placed on air pollution reduction in the
United States. Since the internal combustion engine powered automobile
represents the largest contributor to air pollution, reduction of emissions
from this source represents an extremely important requirement for im-
proving our air quality; strong federal pressureHs being, and will continue
to be, exerted on the automobile industry to reduce pollutant emissions.
The system which, comparatively, offers the greatest potential for
absolute minimum emission of particulates, unburned hydrocarbons, NO,
and CO in a system with range and power equivalent to the internal com-
bustion engine is the Rankine-cycle system. Because of the strong potential
of the Rankine-cycle system as the automotive propulsion system with the
lowest possible emission level, the Motor Vehicle Research Division,
National Air Pollution Control Administration, Public Health Service,
Department of Health, Education, and Welfare is supporting the develop-
ment of such a system. A detailed conceptual design of the system has
been completed and the hardware development of the more critical com-
ponents has been started. The emphasis in the development is on acceptable
performance, packaging, and overall system cost when compared to the
equivalent internal combustion system.
In this summary, a description is presented of the complete design
and characteristics of a Rankine-cycle automotive propulsion system with
thiophene working fluid and reciprocating engine, installed in a 1969 Ford
2-1
-------
THBRIMO BLBCTNON
Fairlane. Projections of emission levels are also made, using results
of a proprietary burner developed at Thermo Electron Corporation.
2. 2 COMPONENT DESCRIPTIONS
2.2.1 System Design Point Characteristics
The sycteni component sizes are based on the peak power re-
quirements for an intermediate-sized automobile (Ford Fairlane) to
give 0-60 mph acceleration of 15 seconds or better and top speed of
— 100 mph. Using these criteria, the engine shaft horsepower less
feedpump power was taken to be 103. 2 hp, obtained from a 184 CID
vapor engine. In Table 1, the design point characteristics of the
propulsion system are presented. Figure 2. 1 shows a simplified
flow schematic and the cycle conditions for the thiophene working
fluid on a T-S diagram. It will be noted that thiophene, due to its
almost vertical saturated vapor line oh the T-S plot and the small
amount of superheat (~40*F) in the boiler outlet vapor, requires a
relatively small regenerator compared to other organic working
fluids; the ratio of boiler heat transfer rate to regenerative heat
transfer rate is 0. 16 for the cycle conditions.
2.2.2 Engine (Expander) Design
Establishing the dimensions of a new engine design depends,
among other things, on the prediction of the indicated and mechanical
efficiencies of the engines at the design condition. Consequently, a
detailed analysis was carried out to determine these efficiencies as
functions of piston speed and load, using measured efficiencies ob-
tained with the 5 hp engine on test at Thermo Electron as a check.
As a result of the analysis, it is possible to plot engine efficiency
versus piston speed at various loads. One such plot is shown in
2-2
-------
THBRIMO ELECTRON
CO*POH*tlOII
Figure 2.2 for an indicated mean effective pressure (IMEP) of 125 psi.
The rapid drop in efficiency with piston speeds above 1000 ft/min occurs
due to inlet valve losses.
From this analysis, a piston speed of 1000 ft/min was selected
for the design condition of 103 bhp at a vehicle speed of 95 mph. The
reduction in engine size which could be realized by selecting a higher
piston speed would probably be more than lost in boiler and condenser
size increases due to lower overall cycle efficiency.
The IMEP and BMEP are determined by the cycle design con-
dition, and the BMEP and the piston speed determine the piston area
required to develop the desired horsepower. A 90°V of four cylinders
was selected as being reasonably compact without either an excessive
number of moving parts or excessive torque variation. The V design
results in a short engine for a given number of cylinders and facilitates
packaging of the system. With four cylinders, the resulting bore is
4.42 inches. The mean piston speed (1000 ft/min at design) and the
engine speed are related by the expression
S = 2 LN
where S = mean piston speed, L = stroke, and N = rpm. The selected
design point rpm of 2000, based on a reasonable bore-to-stroke ratio
(1.47) and on valve train dynamics, results in a stroke of 3. 0 inches.
The basic engine dimensions and specifications are given in Table 2. 1,
and cross-sectional views of the engine are presented in Figures 2. 3 and
2.4 with hydraulic engine valving, slipping-clutch transmission, and
feedpump incorporated.
2-3
-------
THERMO ELECTRON
All materials are identical to those now used in automotive in-
ternal combustion engines. The Rankine-cycle expander differs from
the current automotive internal combustion engine in two very important
aspects of its design: the inlet valving and the bearing design.
2.2.2.1 Variable Cut-Off Inlet Valving. The importance of
having a valving system with variable cut-off is established in a later
part of this paper. Apart from the difficulty in varying the cut-off,
the valving problem is considerably more severe than in internal com-
bustion engines. To avoid excessive losses, the high density of the
Cp-34 vapor at engine inlet conditions necessitates an inlet valve
comparable in diameter and lift (and therefore mass) to the intake
valve of an internal combustion engine. However, the valve event
is much shorter in the Rankine engine than in the internal combustion
engine. The design point intake ratio of 13. 7% corresponds to a
valve event of 60° at most, whereas in internal combustion engines
the inlet valve event is of the order of 120° or more. In a cam operated
system, at a given speed, the acceleration and, therefore, the stress
level are proportional to the lift divided by the square of the valve
event, so that the cam stresses are much higher at a given engine
speed for the vapor engine.
One way of overcoming this problem is shown in Figure 2. 5. In
this system, two concentric inlet valves in series are driven by two
separate camshafts. Cam number 1 driving inlet valve 1 has fixed
timing with respect to the crankshaft. Cam number 2 has variable
timing with respect t.) the crankshaft, and the total valve event is
determined by the overlap of the two valves. In this way, relatively
2-4
-------
THKRMO •LKCTftON
COft'OftATtON
long cam events can be used, giving reasonably sized camshafts.
Figure 2. 2 also shows that the mean valve opening area can be higher
with this approach than with a single valve, which should compensate
for the lower flow coefficient of the two valve system.
Other approaches to variable cut-off inlet valving are shown in
Figures 2.6 and 2. 7. These are both hydraulic devices. A directly-
actuated hydraulic system is shown in Figure 2.6: a cam operated
plunger pump operate0 .-. hydraulic column which acts on a stepped
piston on the inlet valv stem. The pump plunger is constructed
with a helical undercut so that its angular position in its bore deter-
mines its effective strode, thus varying inlet valve duration. This
system is quite similar 10 diesel engine injection systems.
Another hydraulic scheme is shown schematically in Figure 2. 7.
In this system a pump supplies high pressure oil at 2000 - 3000 psi to
a rotary valve (shown as two valves for simplicity in Figure 2. 7). The
rotary valve supplies the high pressure oil to alternate sides of a
piston connected to the inlet valve. Cut-off adjustment is obtained by
moving the rotary valve axially in its housing.
2.2.2.2 Bearing Design. The engine hearing design is strongly
influenced by the type of transmission used. If the engine is coupled
directly to the drive shaft, as in many early steam cars, then journal
bearings relying on hvdrodynamic lubrication cannot be used; roller or
ball bearings rr. jst b-- "'rd, because of the hi^h bearing loads which
could occur at essent.-..y zero rpm. On the .>ther hand, if a con-
venticr.al torque conve -cr were used, bearii,^ sizes, at least on the
2-5
-------
TMKMIMO KLBCTRON
' COt>OI»TIOH
crankshaft, can be comparable to an internal combustion engine of
the same bore, since the peak cylinder pressures are roughly the
same and the inertia loading is lower on the Rankine engine (because
of the 1000 ft/min limit on piston speed). The wrist pin bearing is
more heavily loaded than in the conventional four stroke internal
combustion engine because the load on the pin never reverses; in
this respect, it is much like the wrist pin in a two-stroke engine.
Any single clutch system can load the engine bearings fairly heavily
when the engine is idling at 300 rpm and the clutch is engaged. At
these conditions, with a maximum intake ratio of 80%, the bearing
loading is such that bearing sizes associated with a two stroke diesel
•etf-.the same bore would be barely adequate for the Rankine engine. A
clutch with two forward speeds as well as a reduction in the maximum
intake ratio would alleviate this situation.
2..2. 3 Feedpump Design
The primary factors influencing the selection and design of the
vapor-generator feedpump are that the pump must be positive displace-
ment because of the high discharge pressure; the lubricity of thiophene
is relatively poor and its liquid viscosity low; the pumping rate must
be variable from basically zero to 15 gpm over a 800 - 2000 rpm range;
and the pump must operate with low NPSH without cavitation, since the
NPSH is provided only by subcooling of the liquid coming from the
condenser.
The feedpump selected, illustrated in Figure 2. 8, is a 5-cylinder
piston pump driven by a wobble plate; its characteristics are summarized
in Table 2.3. The selection of a piston pump was based on testing of
2-6
-------
HERMO BLECTRON
several types of positive displacement pumps, including gear and vane-
type pumps, at Thermo Electron Corporation. In general, high leakage
rates (low volumetric efficiency) and high wear rates have been en-
countered for all pump types other than piston, which has given completely
satisfactory performance. With the piston pump all bearing surfaces can
be oil lubricated.
The variable pump rate is obtained by incorporation of variable dis-
placement in the feedpump, permitting the pumping rate to be controlled
at the desired rate regardless of feedpump speed- The method used to
obtain variable displacement is similar to that used in diesel fuel-injection
pumps, in which a ramp undercut is machined in the piston and connected
to the suction side by a port in the cylinder wall. As long as the port is
covered by the piston, pumping occurs on the discharge stroke. As the
port is uncovered by the undercut, the fluid in the cylinder is bypassed to
the suction side as the piston continues its discharge stroke. Rotation of
the piston with the ramp undercut varies the point at which the port is un-
covered, thus varying the effective displacement of the feedpump. Pump-
ing rates from zero to maximum are obtained by 180° rotation of the
piston. In the wobble-plate design illustrated, rotation of all five pistons
is obtained by use of a gear which meshes with gear teeth in the piston
skirts. The gear is rotated by means of a rack and pinion drive passing
external *o the pump through a rolling diaphragm hermetic seal. One-
half inch of rack motion rotates all five pistons simultaneously 180".
oprjng-loaded poppet suction and discharge valves are used. The
suction valve is constructed in the cylinder aiid is made as large as
vbxp tr/ minimize the pressure loss through the valve and the tendency
2-7
-------
THERMO ELECTRON
for cavitation. The smaller discharge valve is located in the cylinder
head. Common suction and discharge plenums for all five cylinders
are incorporated in the housing castings.
The use of five cylinders was based on reducing pressure
transients due to the flow variation from the piston pump. These
transients must be maintained sufficiently small on the suction side
of the pump so that the liquid pressure never falls below the vapor
pressure of the subcooled liquid1. A computer analysis of the pressure
transient behavior indicated that a five-cylinder pump would be required
to prevent cavitation with 20 °F subcooling at the pump suction.
The pump drive could be either crank or wobble-plate. The
wobble-plate drive was selected because of its compactness and
easier packaging with the engine, its lower weight and vibration, its
quieter operation at higher speeds, and its more convenient geometry
for variable displacement incorporation.
2.2.4 Burner-Boiler Design
The burner-boiler design is based on the following requirements:
Reference Cycle Boiler Heat Transfer Rate 1. 58 x 10 Btu/hr
Maximum Boiler Heat Transfer Rate 1.70 x 10 Btu/hr
Burner Design Maximum Heat Release
Rate (HHV) 2. 06 x 10
Burner Design Efficiency (HHV) 82. 5%
Turndown Ratio 15/1
2-8
-------
THBRMO •LBCTRON
In Figures 2.9 and 2. 10, the boiler tube bundle design is presented;
Figure 2. 11 shows a cross section through the burner. The factors
considered in arriving at this design, in addition to heat transfer
performance, were low materials cost, low volume, easy construction,
and low combustion side pressure drop.
The combustion gases, at a temperature of ~3300°F, flow from
the combustion chamber into the center of the tube bundle and radially
outward through the tube bundle. The flow path of the organic through
the tube bundle is illustrated in Figure 2. 12. The organic first flows
through stage 1, from which the combustion gases are exhausted; this
provides the lowest organic temperatures in the boiler at the combustion
gas outlet and a Iv.gh boiler efficiency without air preheat. It is im-
portant that an extremely compact and efficient heat transfer surface
be used in this stage to maximize the boiler efficiency with acceptable
pressure drop on the combustion side. The organic next flows through
the inner stage through which the combustion gases first flow, with a
resultant high heat transfer coefficient. Because of the high gas
temperature and extended surface on the combustion side, coupled
with the high heat transfer coefficient on the organic side, a very
high heat transfer rate can be obtained in the first stage. The organic
next flows through the superheater coil, or stage 3. This stage is a
bare tube coil, since the controlling thermal resistance is on the organic
side and an extended heat transfer surface is not required on the tube.
The characteristics of the three boiler stages are given in Table
2.4; Figure 2. 12 presents the calculated design point temperature and
pressure profiles through the boiler. In the last or third stage, a
matrix made of steel balls brazed together and to the tube is used.
2-9
-------
THBMMO BlBCTttOH
This type of extended surface provides a very high heat transfer £ate
per unit volume and is amenable to mass production techniques.
An intermediate heat transfer fluid (water) is used, with double
"tube construction in the boiler tubes, to positively prohibit hot spots
on the organic side of the boiler' ' Stages 2 and 3 are connected to
the same water reservoir which represents the high pressure side of
the boiler. The water side of stage 1 is separate and represents the
low pressure water side of the boiler.
The combustion chamber design, illustrated in Figure 2. 11, is
based'on design parameters supplied by the Marquardt Corporation,
Derived from experimental testing of a 500, 000 Btu/hr burner of
6 , 3
Similar design. A volumetric burning rate of 2. 8 x 10 Btu/hr-ft -atm
was used in sizing the burner. The burner is constructed integral with
the boiler tube bundle, as illustrated in Figure 2.9. To reduce the
pressure drop, two identical burners are used rather than one longer
burner with the same combustion chamber diameter. The pressure
drop at maximum firing rate for the two burner setup is 1. 5" w. c.
While no pollution measurements are available on the full-scale
burner design illustrated, Thermo Electron Corporation has completed
measurements on a 1/9 scale burner (120,000 Btu/hr) with performance
characteristics similar to the burner illustrated in the design. This
burner is being used on a 5 hp system now on test at Thermo Electron
Corporation. Figures 2. 13 and 2. 14 present the steady state emission
levels from this burner as a function of excess air for burning rates of
105,000 Btu/hr and 50,000 Btu/hr, respectively. It is apparent that the
emission levels are extremely low. To indicate the transient per-
formance of the burner, the burner was oscillated between 50,000 and
2-10
-------
THBRMO ELECTRON
COIPOIATIO*
105,000 Btu/hr burnirg rates with constant fuel-to-air ratio maintained;
CO and unburned } ydrocarbon emission levels were monitored continu-
ously while a bag .ample was collected throughout the run for NO meas-
urement at the e.-.'i. The results are indicated in Table 2. 5. As
indicated in a la; r section, use of these emission concentrations with
the system perfo -mance gives gm/mile emission levels significantly
less than projected 1980 federal limits.
2.2.5 Condenser
The condenser core is similar to a Ford radiator with louvered
fins, except that the flattened tubes have a heavier wall (0. 030" vs
0.005") and are constructed in one integral piece with partitions used
to provide the desired vapor-side flow path. Copper fins with 2. 5 mil
thickness are used, and the tubing is made of carbon steel rather than
brass as in the Ford radiator. The condenser design is illustrated in
Figure 2. 15, with the design point characteristics given in Table 2.. 6.
The frontal area used in the design represents the maximum practical
area for the 1969 Ford Fairlane with some rework of the front-end
frame and grill.
2.2.6 Regenerator
The regenerator design conditions are illustrated in Figure 2. 16
and the regenerator design in Figure 2. 17. The design is based on
obtaining a compact regenerator with geometry suitable for packaging
directly above the engine in the engine compartment and with low
pressure drops on both the liquid and vapor sides. On the vapor side,
a brazed ball matrix extended surface with ] /16 inch ball diameter
is used. The exchanger is divided into four parallel liquid circuits;
2-11
-------
.THRRMO ELECTRON
in each circuit, the vapor passes through four separate stages, per-.
mitting the exchanger to approach that of a pure counterflow exchanger.
2.2. 7 Automatic Transmission
Automatic transmission designs for the system are based on
tailored application of transmission types currently used in automotive
or other vehicular applications, reducing the development effort" and
uncertainty for this component. The Ford Motor Company has supplied
information on one type which uses a standard 12" diameter automotive
torque converter coupled with:modification of a standard manual trans-
mission for forward, reverse, neutral and park control. The Dana
Corporation has provided a two-speed slipping clutch design based
on the type of transmissions they supply for off-the-road vehicles.
Above ~ 5 mph, this transmission locks so the system functions as a
direct-drive system, while still permitting the engine to idle at zero
vehicle speed to drive the accessories. Either of the two transmissions
is feasible, and additional study is required to determine which is
preferable.
2.2.8 Controls
The primary control problem in the system is control of the
burning rate and of the pumping rate to the monotube vapor generator
to maintain boiler outlet pressure and temperature within specified
limits over any type of transient encountered by the system. The
control system is based on operating the boiler as close to quasi-
steady state as possible over all transients; both burning rate and
pumping rate are maintained as closely as possible to the values
corresponding to the instantaneous vapor flow rate from the re-
generator.
2-12
-------
THKRMO BLBCTRON
The feedpvimp control, used to control the pumping rate to main-
tain boiler outlet pressure, is simplified by the fact that the organic
flow rate at any engine rpm is approximately linear with intake ratio.
The feedpump and engine valving are thus operated as a unit directly
by the accelerator pedal; a vernier control working from the boiler
outlet pressure is used to reduce deviations from the design point
and to eliminate any imbalance in the system automatically. A mech-
anical governor is used to limit the maximum intake ratio as a function
of rpm and to govern engine speed at idle.
In order to maintain the burning rate at a value corresponding
to the organic flow rate into the boiler, the burner control uses an
orifice in the organic line to sense almost instantaneously any changes
in the organic flow rate. This signal, along with a similar signal
from an orifice in the fuel line, is used with a diaphragm controller
to provide the proper fuel and air flow rates. If necessary, the fuel-
to-air ratio can easily be varied as a function of turndown to minimize
pollutant emissions at any burning rate. A vernier control operating
from the boiler outlet temperature is used to reduce deviations from
the design point value; at low power levels, when the organic flow is
low, this temperature control becomes the primary control on the
burner.
2. 3 SYSTEM PERFORMANCE AND PACKAGING
The important decisions which have a strong influence on the
overall system performance and cost relative to the internal com-
bust:', -r engine are the method of driving accessories at zero vehicle
speed and the constant intake ratio engine valving. With respect to
2-13
-------
THERMO ELECTRON
the first decision, two alternatives are feasible. An auxiliary, constant
speed engine can be used to drive all accessories, permitting the engine
to be coupled directly to the drive shaft (through a simple gear trans-
mission for forward, reverse, neutral, park control). A more com-
plex transmission can also be used, permitting the main propulsion
engine to idle at zero vehicle speed so that all accessories can be
directly driven by the main propulsion engine; this transmission is
still much simpler than that required for the I. C. engine driven system.
It has been the conclusion of this study that the preferable
approach is use of the main propulsion engine to drive all accessories.
Even though the accessories must be larger when driven by the
variable-speed main propulsion engine, the accessory designs are
identical and the same number of parts must be processed; very
little cost differential exists between the different sized components.
Any cost reductions due to smaller accessory components are more
than counterbalanced by the requirement for an additional engine of
15 -.20 hp to handle short-term accessory peak loads, with governor-
throttle valve control. The system with all accessories driven by the
main propulsion engine, therefore, seems preferable in terms of
simplicity, cost, and packaging; this system has been selected as
the optimum approach.
With respect to engine valving, a detailed analysis with com-
puter modeling of the engine and boiler performance for all operating
conditions has been carried out, comparing a system with constant
intake ratio engine valving and throttle valve control and a system
with variable intake ratio engine valving. The results are summar-
ized in the performance maps presented in Figures 2. 19 and 2. 20
2-14
-------
THKRMO BLBCTMOM
for the constant IR and variable IR (IR = 0.29) systems, respectively,
max '
Alar- j,resented in Figure 2. 20 is the power increment obtained by going
to a system with (IR) = 0. 8, the maximum practical value. It is
max r
apparent that only a relatively small increase in the system maximum
power level is obtained with (IR) = 0. 8, since going to (IR) = 0. 8
max max
requires a much larger feedpump as well as maximum condenser cooling
air at a lower engine and vehicle speed; (IR) = 0. 29 represents the
max
optimum intake ratio. Comparing these two performance maps'for the
same boiler and engine sizes, the following conclusions can be made;
(a) The system with variable IR valving has a peak efficiency of 18. 5%
versus 15.0% for the system with constant IR valving. This 20 - 25%
improvement in efficiency (or in mpg) occurs over a large power-speed
region including the region of 20 -40 hp and 600 - 1200 rpm, where the
system would operate most of the time; and (b) The peak power for the
equivalent sized engine and boiler is much greater with variable IR
valving than for constant IR valving. Calculations of 0 - 60 mph wide-
open-throttle times indicate a 65% increase for constant IR valving
relative to variable IR valving with (IR) =0.8. For these two
max
reasons, there exists an extremely strong incentive for incorporation
of variable intake ratio valving in the engine; this type of valving is
used in the reference engine design presented earlier.
With reference to Figure 2.20, an important consideration
is the part-load performance of the system. Thus, while the design
point efficiency, defined by the peak power requirements for accelera-
tion, is 13. 7%, it increases under part load conditions where an auto-
mobile normally operates. This increase with the variable IR valving
occurs because part-load operation is obtained by reducing the IR below
the design point value of 0. 137, providing a more efficient expansion
2-15
-------
T H g n M O E1.HCTR O_N
CORPORATION
in the engine; reduction in the condenser pressure under part-load
operation also occurs again leading to a more efficient cycle. It
will also be noticed that the region of high efficiency (>17%) is broad;
that is, a high efficiency is obtained over a broad range of engine
power and speed. Thus, while the peak thermal efficiency of the
Rankine-cycle system (18.5%) is much less than that of the I. C.
automotive engine (~ 30% ), the average efficiencies for typical con-
sumer driving cycles are much closer. The Ford Motor Company
has calculated the fuel economies for different driving conditions of
the Rankine-cycle system installed in a 1969 Ford Fairlane 4-door
sedan using the performance map of Figure 2.20 with (IR) =0.8
max
and compared them directly with the fuel economy calculated for the
same body with 302 CID engine and three-speed automatic transmission.
The results are summarized in Table 2. 7 for both steady speed and
dynamic driving cycle operation. For the customer average driving
cycle, the mpg for the 302 CID I. C. engine is 15. 7, versus 12, 7 mpg
for the Rankine-cycle system. With respect to acceleration per-
formance, the Rankine-cycle system tractive effort with single-speed
direct clutch transmission and with torque converter, is compared
for (IR) = 0.8, in Figure 2,21, to that for the 302 CID I.. C. engine
m£Lx
with 3-speed automatic transmission and 2.79 axle ratio (calculations
prepared by Ford Motor Company). At speeds above 43 mph, the
tractive effort of the two systems is practically identical. Below
this speed, the tractive effort from the I. C. engine is somewhat
greater. The 0-60 mph WOT acceleration times are compared
in Table 2.7 for the two systems (11.9 seconds for the I. C. system,
compared to 14.2 seconds for the Rankine-cycle system with 184 CID
engine).
2-16
-------
> ""V, T HBP ?•• *N EL C C T ROM
• • *• ~~ ~' ~~ ~ ~~~*—™^"—~"—
V J*»ORAT|QII
In Figure 2. 21, the tractive effort obtained by use of the Rankine-
system with (IR) = 0.29 and two-speed clutch transmission
max r
as caii ulated by the Dana Corporation is compared with that of the
302 C1D I. C. engine with three-speed transmission. The tractive
effort of the two o /stems is practically identical down to 20 mph. The
0-60 mph acceleration time for the Rankine-cycle system is reduced
to 12. 5 sec. for the combination, which appears optimum from an
overall system viewpoint.
In Table 2. 8; the weight of the Rankine-cycle system reference
design (917 Ibs total) is compared with that of the 302 CID engine
with three-speed transmission (709 Ibs total).
Packaging of the system is an extremely important consideration,
particularly for the first generation prototype or production units when
it is preferable to use as fully as possible the same car body as is used
for I. C. powered autos. Accordingly, the approach followed has been
to package the system completely in the engine compartment of current
automobiles; a 1969 Ford Fairlane has been used for the current study,
and a full-size, complete mockup of the system has been constructed
in the engine compartment of this car with excellent results. In
Figure 2. 22, a photograph of the mockup is presented; in Figure ?. 23,
sketches illustrating the engine-transmission, burner-boiler, and
condenser locations in the system are illustrated. The only change
required in the engine compartment in packaging the system was
mocL^'jaHon of the frame and fender panels at the very front of the
car tc facilitate placement of the condenser.
2-17
-------
THERMO ELECTRON
2.4 EMISSION PROJECTIONS FOR RANKINE-CYCLE AUTOMOTIVE
PROPULSION SYSTEM
Using the burner emission levels obtained from the burner developed
at Thermo Electron Corporation for a 5 hp Ranicine-cycle currently on
test, projections have been made of the emission level of unburned hydro-
carbons, CO, and NO from a Rankine-cycle automotive propulsion system
using 10 mpg fuel economy; the results are presented iii Table 2. 9 and
compared with projected Federal standards for 1975 and 1980; presented
i
also are results for an uncontrolled I. C. engine and the IIEC targets
for emission control for the I. C. engine. The Rankine-cycle system
emission levels are lower than the projected 1980 standards by a factor
of 5 for unburned hydrocarbons,by a factor of 13 for carbon monoxide
emission, and by a factor of 1.6 for NO emission.
2. 5 MAJOR CONCLUSIONS
1. The Rankine-cycle system offers the greatest potential of
any combustion-operated propulsion system for absolute minimum
emissions of particulates, unburned hydrocarbons, nitric oxide,
and carbon monoxide.
2. The use of thiophene as a working fluid, with a moderate
maximum cycle temperature of 550°F, permits a significant cost
reduction relative to the equivalent steam system. This reduction
may permit the Rankine-cycle system to be competitive costwise with
the equivalent I. C. system, particularly since the stricter emission
level standards will require significant cost increases in the I. C. system.
3. An organic working fluid, Rankine-cycle system equivalent in
performance to a 302 CID I. C. engine with 3-speed automatic trans-
mission can be packaged in current automotive engine compartments
with only minor modifications required in the sheet metal and frame.
2-18
-------
THBUMO «i.«CT*OK
c««roi*rioii
Reference
1. D. T, Morgan, E. F. Doyle and S. S. Kitrilakis, "Organic Rankine
Cycle with Reciprocating Engine," Presented at the Fourth Inter-
society Energy Conversion Engineering Conference, Washington,
D.C., Sept. 22-26, 1969. Paper No. 699001.
2-19
-------
TABLE Z. I
DESIGN POINT SPECIFICATIONS
Working Fluid
(toiler Outlet Temperature
hoilar Outlet Preeaure
Boiler Heat Transfer Rate
Boiler Efficiency (HHV)
Engine Displacement
Eni|ine Speed
Engine Piston Speed
Engine llnnepowcr Less Feedpump Power
Engine Thermal Efficiency
Engine Mechanical Efficiency
Engine Overall Efficiency
Engine IMEP
Regenerator Effectiveness
Rrgeneretor Heat Transfer Rate
Condensing Temperature
Condensing Pressure
Subcooled Liquid Temperature
Condenser Heat Transfer Rale
Organic Mats Flow Rale
Organic Volumetric Flow Rate
Feedpump Overall Efficiency
Feedpump Power
Cycle Efficiency
Overall Efficiency
Cp-14
S50T
SOO pela
LSI « 10* Btu/hr
12. 5*
IB4 in1
200U rpm
1000 fl/mln
101. 2 hp
14. 6*
91.5%
77.5*
127.4 pel
90.0%
0.241 . 10* Btu/hr
216.2'F
2V 0 psla
196.2T
1.25 a 10* Btu/hr
7177 pounels/hr
I S. I gallona/min
49.7*
5.25 hp
16.7%
11.7*
TABLE 2. 4
BOIlJCft UFEUNCE DESIGN POINT SPECIFICATION
fltaaei
No.
2
3
1
ToUl
He»t
Rat*
Btu/hr
0. iJ4 M 106
0.313 x I06
0.359 M 10*
I.S76 x 10*
Combuvtloa Ga* Tatnp.
•nt*rinf
1330
1196
1190
-
lea, ring
1896
1190
490
-
Tubing
ft.
17.0
35.0
26.0
78.0
Pr***ur* Drop
Combustion
Sid* In w. c.
0. 136
0.288
2.06
2.48
Orfanic Sid*
pal
21.2
23. 3
2.0
46.5
Tub* Specification*
E»t«rna>l Fin Sp*ciftca,tiona Matria Spaciflcattona
Inner Tub* ID . 0.930* Floa/Inch
OD • 1.000* Fin Thlckn.a.
Ftn Material
Outajr Tub* ID • 1.121* Fin Height
OD • 1.315*
10
0.012*
Copp*r
0.356-
Ball Sit*
Ball Material
Matrix Thlckneee
Matrix Height
(b*tw**n tub**)
3/32*
Carbon St**l
o.s-
0.934*
TABLE 2. 2
ENGINE DIMENSIONS AND SPECIFICATIONS
TABLE 2. 5
TRANSIENT EMISSION DATA
M
O
Configuration
Bore
Stroke
Displacement
BMP (feedpump work deducted)
UrIEP at Design
BMEP at Design
Four Cylinders. 90 • V
4.42 inches
1.0 inches
1*4 in1
10] at 2000 rpm
127 psi
117 psi
TABLE 2.3
CHARACTERISTICS OF FEEDPUMP WITH 15 GPM RATING
(Sited for System with (IRI • 0.29)
Number of cylinders
Volumetric Efficiency
Ov*r*Jl Efficiency
RPM Range for Maximum Pumping Rat*
Total Di*pl*c*rn*nt
Bor*
Strok*
Malarial* of Construction
Housing
PI stone and Valve*
Daaringe
5
90%
80%
800 - 2000 rpm
4.78 In1
1.595 in.
0.478 in.
Ca*tIron
Hardened St*el
Roller and NeecU*
FIRING
RATE
(BTU/HRI
105,000
50, 000
50, 000
50. 000
SO, 000
IOS, 000
105,000
105.000
50. 000
50. 000
SO. 000
50. 000
105.000
105.000
IOS. 000
SO, 000
SO. 000
SO. 000
EXCESS
AIR
(T.I
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
25
• 25
CH
(PPM)
6
-
5
-
4.5
40'
7
5
.5
4
4.5
-
15'
-
6
-
4.5
4
CO
(PPM)
60
-
10
2S
10
mo1
<
-------
1-1117
TABLE 2.4
CONDENSER DESIGN POINT CHARACTERISTICS
Haat RajacttM Rata
(20* fuporhaat
20* fciacoalUf Laavlaf)
Halfht
Dapth
Froatal Area
Caa4ao«r Inlat Praaaura
Daalg* AmalaM Air Tamparatura
Air Praaiura Drap
•oft Fan Power (Two 20 la. O.D. Fana
with 1. SO Ueh Pitch)
Air Mow Rata
1.25 a 10 Btu/hr
SO In.
19.9 U..
3.0 In.
4. 91 ft2
2S pala
9ST
3.45 In. w. e.
7.91 ha
43,000 lh/»r
TABLX 2. 7
PERFORMANCE COMPARISON OF RANKINX-CYCLE AND INTERNAL COMBUSTION
AUTOMOTIVE PROPULSION SYSTEMS
TahUU - 19*9 fcr* Fntrlu* 4-DMr **t»*
W.ight - 3S)9 Ib.
•jratam
302-2V EBfina with Automatic
Truimlifloa
ISO- IV Enflaa with Automatic
Truamlaalon
Baalilaa-Cycla Syatama
1*4 CIO E*flaa (IR) • 0. 1.
• nfla Spaa4
dutch Trajumlaalon
22* CID EnfUa (IR) « 0. 1,
aUfla apaad ""
datch Tru>nUilan
1 *4 CID Eafiaa. (IR) • 0. 29
1.0/1 aad I/I Caar Ratiat
-jr
1M
-
100
1*4
1*0
0-60 mph
AccalarattOB
Tim a, aac
11.9
15.4
14.2
11.4
12. 5
Fual Economy, mph
Staady Spaad, mph
30
27.3
27.0
33.1
-
-
50
20.4
22.9
20.4
-
-
70
16.2
17.6
13.4
-
-
City
13.3
12.9
9.6
-
-
Suburban
It. 0
19.4
15.1
_
-
Gaatomar
15.7
16.3
12.7
_
-
TABLE 2. •
TABULATION OF TOTAL SYSTEM WEIGHT AND COMPARISON WITH 302-2V
INTERNAL COMBUSTION SYSTEM WITH 3-SPEED TRANSMISSION
Eflfina E»p*n4«r Aatamkly
Faadpump
Eafiaa SubayXam
Traaamlaata-a
Bumar-B*lUr
Ra«a««rat*r
Ca*aaa..r
Racftatar with fan. connactora. and watar
CwitraJa. E»»»uat, Daatrical Syatam,
Accaaaory Drivaa, a>4 a**r
Miacallajiaatia Campoaanta
WarkUf nuid am4 Lvbrtcaat
Talal
Raahlaa
Rafaraaca
Daai|n
220
_1»
245 Ika
135 Ika
273 Ika
54 Ika
115 Ika
75 Iba
4« Iba
957 Ika
302 Cu. In V-l
with 3-Spaad
Autamatic
479 Iba
159 Ib.
54 Iba
114 Iba
M4 Iba
2-21
-------
TABLE I. »
iMMPIB CVCfc» WITH INTERNAL COMBUTTtON ENGINE
_^-
ttnliiloa HC
.L»T«1 00
jm./nrt NO
PtojKtW
MM
ttudirdt
0.1
11
0.9
ProJ.ct.J
1910
*ur
-------
2*11-0
rif«r« 2.3 Creic-Sectional Dr»wing of Engln*.
2412-D
y
u
X,.
Tifum 2. 4 Lomfitx**lo»l Scctio* OrtwUf of Enjio..
2-2J
-------
1-315
.. ®
Figure 2. 5 Two Inlet Valves in Series.
1-550
Figure 2. 6 Directly Actuated Hydraulic Valve.
1-551
Figure 2.7 Pilot Operated Hydraulic Valve.
2-24
-------
'6Z3-D
©
•^"'•>:
^05^
Figure 2. 8 800-2000 rpm F.edpump.
2607-D
nan
Z605-D
Figure 2. 10. Top View of Boiler Tube Bundle.
Figure 2. 9 <-'r >•• Section Through Burner-Boiler.
-------
2604-D
1-916
Figure 2. 11 Croa* Section* Through Automotive-Sire
Burner.
1-318
ORGANIC FLOW TO ENGNE
COMBUSTION
GAS
FLOW
t 1
ORGANIC FLOW
TO BOILER
STAGE NO. ' 32
Figur* 2. 12 Organic Flow Path Through Boiler Tube
Bundle.
;
I
5
i
HI
MO-
•0-
'wo-
MO-
120-
100-
•0-
so -
4O -
20 -
•TIAOV ITATI DATA
O'tO.OOO tTU/HN
tun. jp-4
> • MO
• - CO
• -CM.
•
«k^. • -0
0 1O ' 20 10 4<0 M *0
EXCESS AIM nil
Figure 2. 13 Effect of Excel* Air on Emlxlons,
50. 000 Btu/hr.
1-917
•TtADV STATE DATA
Q-Wi.OOO ITU/HN
200-
110-
NO-
MO-
K»-
SO-
40-
20-
0-
—r 1 1—
20 10 40
EXCESS AIM (\>
Figure 2. 14 Effect of Excess Air on Emission*.
105, 000 Btu/hr.
-------
2615-D
-So.o
Figure 2. 15 Condenitr De»ign.
1-1060
Vapar
25 Mi*. 341'F
(550-AP) psii, 2I5'F
Q * 24I.IM Bta/hr
(25-A P) p»it. 230*F
S50 ptia. 1SS'F
it Flaw data » Vapar Flaw Data = 7377 las/ar
Figure 2. 16 Regenerator Design Point Requirement*.
2617-D
0
)0
?s§
fOQQQ OOQQ
'/ OQOO QOOQ
OQOQ W¥¥V
Figure Z. 17 Regenerator Design.
2-27
-------
1-360
2OO 400 600 800 1000 I2OO I4OO I6OO I8OO 20OO 22OO
Flfttr* 2. 18 Performance Map with 184 CID Engine
and Constant Intake; lUtlo of 0. 117.
1-961
ZOO «0 600 800 OOO I2OO I4OO I6OO I8OO 2OOO 2200
figure 2. 19 Performance Map with 184 CID Engine
and Maximum Intake Ratio of 0. 29.
1-1119
1-tPIED AUTO
I.Tt ULI
THtmo tucmoN
•i cm IIMIKINC
1.M OWIKALLJIATIP
rti mr UP TO TOKQUE CONVEKTER
(!.!• P.O. PINION -- 4.10 P.O. OIAN
tOO *PM ENGINE tTALLJ
l.Ot AXLE RATIO
11 B/ia DIAMETER CONVERTER
TMERMO ELECTRON
•1 CIO RANKINE
V*4 AXLE
OIHICT CLUTCH
Ftgur* 2. 20 Tractive Effort r«riu» VeMcIe Sp«»d.
Talrlana at WOT Acceleration.
2-28
-------
1-907
Kigur* Z.ll Photograph of lUnkin* Cycle System Mock up
in 1969 Ford F»irl»n* Engine ( ump»rtm«nt
1-970
f.
t
r
» :
n
i
.1
!
i
»
t
*'- _- -'t -_
i . -t i ' ,- - -
t
/r-v._
/
f '" H .rrr -w
i,1 r\ TT 7" .
r \, Mj-^-4
t i ? ;;:•___ ^
PK^** :^^'
rr\4xf:v : ,
IT*-*
,
* :
1
i '.,i«rv- Kngin* t ompartrr • i
2 29
-------
THERMO ELECTRON
3. INTRODUCTION
3. 1 OVERALL GOALS
Great emphasis is being placed on air pollution reduction in
the United States. Since the internal combustion engine powered
automobile represents the largest contributor to air pollution, a
large reduction in emission from this source represents an extremely
important requirement for improving our air quality; strong federal
arid state pressure is being, and will continue to be, exerted on the
automobile industry to reduce pollutant emissions.
The system which, comparatively, offers the greatest potential
for absolute minimum emission of particulates, unburned hydrocarbons,
NO, and CO, in a system with range and performance equivalent to
the internal combustion engine, is the Rankine-cycle system. Quoting
from a recent article by the Technical Advisory Committee, California
Air Resources Board,
"The promise of the steam engine to produce inherently
low levels of exhaust emissions is well-founded. The
capability to utilize combustion processes which are not
highly productive of the three principal pollutants is con-
firmed by available technical knowledge of flames and
of the formation of those pollutants. "
In Table 3. 1, information prepared by the same Technical
Advisory Committee is presented, summarizing their conclusions
with respect to the emission levels which could be achieved by various
approaches; the emission levels established either by legislation in
California or as goals by various sources are also presented. It is
apparent that the steam car, in the opinion of this Technical .Advisory
Committee, is low in all three of the major pollutants (hydrocarbons,
3-1
-------
THERMO ELECTRON
carbon monoxide, and oxides of nitrogen), and that it is the only system
which is simultaneously low in all three pollutants based on current
technology.
Work at Thermo Electron Corporation in this study and on the
hardware development of a 5 hp Rankine-cycle system, coupled with
recent pollutant measurements performed by other industry groups
from burners suitable for use in a Rankine-cycle system, have con-
firmed this inherent potential of the Rankine-cycle system. In Table
3. 2, the projected pollutant emission levels using the results of
three different types of burners are compared with current and pro-
jected Federal standards. It is apparent that emission levels for
both hydrocarbons and CO are less than the projected 1980 Federal
standards for all three burners. The NO level is less than the pro-
jected i960 standard from the Thermo Electron Corporation measure-
ments and slightly above the 1980 Federal standard from the CM and
Marquardt measurements. These burners have not been optimized
for reduction of NO emission, and it is expected that development
work can reduce the NO emission to a level consistent with the
projected 1980 standards with both HC and CO emissions being signifi-
cantly lower than the 1980 standards.
With this general acceptance of the low pollution level of the
Rankine-cycle system for automobile propulsion, the Motor Vehicle
Research Division, National Air Pollution Control Administration,
Public Health Service, Department of Health, Education, and Welfare,
is supporting the development of such a system. A conceptual design
of the system has been completed; detailed design and experimental
development of the more critical components have been started.
3-2
-------
1-1069
THKRMO BLBCTMON
coiro«*rion
TABLE 3. 1
AUTO EMISSIONS: A SUMMARY OF POSSIBILITIES
(all figures in grams per mile)
Hydro- Carbon Oxides of
Legislation and goals carbons monoxide nitrogen
Prior to control
California Pure Air Act (AB 357) 1966
1971
1972
J9?4
California Low Emission Vehicle Act (AB 356)
Morse report goals for 1975
Interindustry Emission Council goals
Modified conventional engines
Sun Oil Co. test vehicle
Chrysler-Esso engines
Manifold reactor
Catalytic reactor
Synchrothermal reactor
Ethyl Corp. "lean reactor" car
DuPont manifold reactor
Alternative power plants and fuels
Steam car
Gas turbire
Wankel engine
Stirling hybrid
Natural gas fuel
11. 0
3.4
2. 2
1. 5
1. 5
0. 5
0.6
0.82
0.7
<1. 5
1.7
0.25
<0. 7
0. 2
0. 2-0.7
0. 5-1.2
1.8
0. 006
1. 5
80.0
34. 0
23. 0
23. 0
23. 0
11.0
12. 0
7,1
12. 0
<20. 0
12.0
7. 0
<10. 4
12. 0
1.0-4. 0
3. 0-7. 0
23. 0
0. 3
6. 0
4. 0
_
4. 0
3. 0
1. 3
0.75
1. 0
0. 68
0.6
<1. 3
1. 0
0.6
<2. 5
1. 2
0. 15-0. 4
1. 3 -5. 2
2. 2
2. 2
1. 5
So-irce: Technical Advisors Committee, California Air Resources Board
3-3
-------
TABLE 3. 2
PROJECTED EMISSION LEVELS FROM RANKINE CYCLE POWERED CAR
ASSUMING 10 MFC AVERAGE FUEL ECONOMY
i Projected Federal Standards, gms/mlle
1975
HC
0. 5
CO
11
NO
0.9
1980
HC
0.25
CO
4. 7
NO
0. 4
Source of
Emission
Data
Measurements at
Thermo Electron
Corp. on 140. 000
Btu/hr. Burner;
Excess Air=33%
Measurements
by General Motors
Corp. , Excess
Air = 68%
Measurements
by Marquardt Corp,
on 500, 000 Btu/hr
Burner, Excess
Air = 33%
Emission Levels
ppm in Exhaust Gas
HC
15
8
4
CO
60
290
60
NO
40
75.
90
Calculated gms/mile
for 10 MPG Fuel
Economy
HC
0. 05
0. 033
0. 013
CO
0. 35
2. 1
0. 35
NO
0.25
0. 59
0. 57
n
H
a
o
z
o
-j
o
(a) See Section 4. 5 for details of measurements.
-------
THERMO ELECTRON
CO*PO**TIOI
The emphasis in this development is on reducing or eliminating the
problems generally associated with Rankine-cycle automotive pro-
pulsion systems without compromising the emission or performance
features of the system. The problem areas generally associated with
the system which have been considered in the development of the design
presented in this report are summarized in Table 3. 3.
In addition to the goal of reducing the effect of, or eliminating,
the problem areas outlined in Table 3. 3, an additional guideline of the
study has been the application of current state-of-the-art technology where-
ever possible in order to achieve the shortest possible development time
for a prototype research vehicle. Areas where the current state-of-the-
art is not applicable have been identified in the study so that the early
hardware development can be concentrated in these areas. In the
remainder of this section, a description will be given of the approach
followed in accomplishing these goals. The other major sections of the
report present the individual component characteristics and descriptions
and the overall system performance and description for a 100 net shaft
horsepower system competitive in performance with a 302 CIO internal
combustion engine coupled with a three-speed automatic transmission.
The 1969 Ford Fairlane is used as the reference automobile chassis
for the design.
3.2 APPROACH FOLLOWED FOR ATTAINMENT OF GOALS
The approach followed to alleviate the problems discussed above
and !•:> inc rease the competitiveness (in areas other than emission
ieveis) cf the R^nkine-cycle system relative to the internal combustion
system involves use of an organic working fluid with optimum thermo-
3-5
-------
THBRMO Ei-BCTRON
1-1114
TABLE 3. 3
PROBLEM AREAS GENERALLY ASSOCIATED WITH
STEAM ENGINE DRIVE FOR AUTOMOBILES
1. High cost
2. Long startup time
3. Large package size and weight
4. Poor fuel economy
5. Safety problems
. 6. High maintenance and poor reliability
7. Freezing of working fluid at low ambient temperatures
8. Driving of accessories
9. Blowoff of working fluid under peak load conditions
with makeup required
3-6
-------
THBRMO BLECTRON
dynamic characteristics, such as thiophene, in place of water (which
has generally been used in the past). In Figures 3. 1 and 3.2, realistic
cycle conditions suitable for use in a portable power system are pre-
sented on T-S diagrams for steam and thiophene, respectively. The
cycle conditions selected for steam are based on the criterion that
the expansion in the reciprocating engine should not enter the saturation
dome, since this results in a large increase in heat transfer to the
cylinder wall during the expansion, with a resulting decrease in the
engine efficiency.
With respect to the steam cycle, the high pressure steam is
expanded in the engine (process 1-3) to produce shaft power. The
exhaust steam from the engine is then condensed and subcooled in the
condenser (process 3-4). Water from the condenser is pumped back
into the boiler by the feedpump (process 4-5). The burner-boiler
then heats the water to the desired engine inlet conditions (process 5-1)
completing the cycle.
An organic working fluid can be used just as easily as steam in
a Rankine cycle, with the requirement, for an additional component,
a regenerative heat exchanger, and with precautions to prohibit
overheating of the thermally-sensitive working fluid. The regenerator
in an organic Rankine-cycle is desirable to increase the cycle efficiency,
since the exhaust vapor from the engine for organic fluids is superheated
and the energy in the exhaust vapor can be transferred to the feed organic
going to the boiler, thereby reducing significantly the energy to be added
to the working fluid in the vapor generator for a given power output from
the engine. Because of its unique thermodynamic properties, thiophene
requires a regenerator with a heat transfer rate relatively small
3-7
-------
THBRMO ELECTRON
ca*r. otATion
compared to that of the vapor generator. Other organic fluids, particularly
those with high molecular weight suitable for turbine expanders, may
require a regenerator with a heat transfer rate close to or considerably
in excess _of the vapor generator heat input rate.
It is thus apparent that a Rankine-cycle system can be operated
either with an organic working fluid or with steam as a working fluid.
An organic fluid, however, offers a number of advantages over steam
/which_have great significance for some of the problem areas associated
with steam Rankine-cycle systems. A comparison of the principal
characteristics of steam and organic working fluids is presented in
Table 3.4. The advantages accrue from the much lower engine inlet
temperature permissible with the organic working fluid without
affecting the overall system efficiency. Thiophene, for example,
gives a cycle efficiency with an engine inlet temperature of 550°F
almost equal to that of the steam cycle with an engine inlet temperature
of 800°F. This lower maximum temperature has several important
benefits to a practical, low cost Rankine-cycle system with recipro-
cating engine. First, it permits use of less expensive materials of
construction in the engine. Second, it eliminates the need for a
reciprocating seal between the crankcase and power side of the engine
to keep the lubricant separated from the working fluid. Lubricants
are available which are thermally stable and compatible with the
organic working fluid at the low maximum cycle temperature of 550°F.
Lubricant entrained in the working fluid can, therefore, be allowed
to pass through the vapor generator. Both of these features lead to
a significant reduction in the cost of the reciprocating Rankine engine.
In addition, elimination of the reciprocating seal leads to an increased
3-8
-------
TH • II MO • L meTMOM
CORPORATION
2465-D
900
800-
700
;6oo
S 500
I
40O
300
200
i I I I i I
\i
24 6 S 10 12 14 16 18 10
»,EntPOpy(8hj/*F-lb)
1 Working Fluid
' Shaft Power
Fuel
(Y) Temp.Control
@ Pressure Control
f?) Speed Control
Figure 3. 1 Flow Schematic and Cycle Conditions
for Steam Rankine Cycle.
3-9
-------
X7 TM• » M O « L «CTB
~£^X3 CORPORATION
CTMON
2464-D
730
•SO
350
J 4SO
1= 3SO
230
130
-04 -OJ -02 -01 0 010 020
Ttmp Conlrol
Prflturfl Control
ontrol
Figure 3. 2 Flow Schematic and Cycle Conditions for
Organic Rankine Cycle.
3-10
-------
THBRMO BLBCTftON
I-1 071
TABLE 3. 4
COMPARISON OF STEAM AND THIOPHENE AS RANKINE-CYCLE
WORKING FLUIDS IN SYSTEM WITH
RECIPROCATING ENGINE
Thiophene Working Fluid
- Low Operating Temperature
(500°F) and Non-Corrosive
Conventional materials of
construction (carbon steel,
cast iron, aluminum, brass).
Compatible with lubricants
at maximum cycle tempera-
ture.
- Lubricant Sealed in System with
Working Fluid
Bearings and sliding sur-
faces oil lubricated.
Conventional engine construc-
tion (no reciprocating seal
and crosshead piston).
- Low Freezing Point (-40°F)
- Limited Experience
- Flammable/Toxic
- Readily Available/Relatively
High Fluid Cost
- Thermal Decomposition (High
Temperature Limit)
- Regenerator Required
Water Working Fluid
- High Operating Temperature and
Corrosive
Alloy steels required in
boiler and engine.
Lubricant not compatible with
working fluid at maximum
cycle temperature.
- Poor Lubricating Properties
Water lubricated materials
in engine.
Reciprocating seal in every
cylinder.
Crosshead piston required.
- High Freezing Point
.- Extensive Experience
- Non-Toxic/Non-Flammable
- Readily Available/Low Cost
- Thermally Stable
- No Regenerator Required
3-11
-------
THKRMO ELECTRON
life without maintenance. The construction of the engine, in fact, is
identical to that of internal combustion engines and reciprocating
compressors. In Figure 3. 3, a schematic comparison is given of the
construction of an engine with and without a reciprocating seal.
The system is constructed as a completely sealed system with
condenser capacity sufficient for peak load conditions. The only
dynamic seal in the system is the rotary shaft seal on the rear of the
engine,, required for transmission of shaft power from the system.
A seal has been tested at Thermo Electron Corporation which prohibits
leakage of air into, or working fluid from, the system. The working
fluid and lubricant are thus sealed in the system for the life of the
unit. This approach is similar to that of a hermetically-sealed air
conditioning system and should result, with development, in a system
with high reliability and low maintenance requirements.
A very important problem with steam is its freezing point of
32 *F; thiophene, on the other hand, has a freezing point of -40°F and
can thus be used in all parts of the continental United States.
Thiophene also contracts on freezing, so that exposure to temperatures
less than -40°F will not damage the system.
In summary, use of thiophene as working fluid in a completely
sealed Rankine-cycle system for automotive propulsion has the
following advantages:
1. Minimum cost for complete system.
2. High reliability and low maintenance requirements.
3. Capable of cold startup down to -40°F ambient temperature.
4. Capable of storage at temperatures less than -40°F ambient
temperature.
3-12
-------
u>
POWER PISTON
SEAL
SEAL
POWER PISTON
CROSSHEAD PISTON
CONNECTING ROD
CRANKCASE
CRANKSHAFT
Figure 3. 3 Comparison of Conventional Steam Engine with TECO Engine
Using Organic Working Fluid.
-------
THBHMO ELECTRON
The component and overall system designs and characteristics
presented in Sections 4 and 5 have established the capability for:
1. Startup.time of 30 - 45 seconds.
2. Acceptable packaging in engine compartment of current
automobiles.
3. Acceptable weight for current automobiles.
4. Customer average MFC 20% less than internal combustion
- engine.
5. Sufficient condenser capacity for totally condensing, system
under peak load conditions.
6. Conveniently'driving all accessories.
One key difference between an organic and a steam Rajikine
cycle lies in the thermal stability of the working fluid. Water, of course,
is completely stable under the conditions encountered in Rankine-cycle
systems. Organic materials, on the other hand, have a definite and
relatively sharp upper limit to the temperature to which they can be
exposed without catastrophic thermal degradation of the fluid. To
achieve a reasonable cycle efficiency, however, it is necessary to
operate at a maximum cycle temperature close to that temperature
limit. In any direct-fired boiler in which the heat is transferred
directly from the combustion gases (at a maximum temperature of
3300°F) to the organic with only a metal wall between, hot spots will
occur, leading to catastrophic decomposition of the organic working
fluid unless very low peak cycle temperatures are used. At Thermo
Electron Corporation, the approach followed has been to transfer
the energy from the combustion gases to the organic by means of a
thermally-stable, intermediate heat transfer fluid, thereby positively
prohibiting exposure of the organic to excessive temperatures. This
3-14
-------
TNBMMO BLBCTItOM
idea ia not new, but its application has in the past required complicated
and expensive equipment and has therefore been limited to large systems.
The vaporizer concept developed at Thermo Electron, as described in
Section 4, permits use of an intermediate heat transfer fluid to eliminate
hot spots, with only a slight increase in cost over the equivalent direct-
fired boiler; it is geometrically identical to and the same size as the
equivalent direct-fired boiler, and requires no pump for circulation of the
intermediate heat transfer fluid. This;.. vapor generator concept has been
tested with thiophene and with R-22 working fluids in two separate boiler
loops which closely simulate the actual temperature conditions in a
closed Rankine-cycle system for each fluid with no measurable degradation.
With respect to safety, use of a once-through boiler with a small
high-pressure fluid inventory greatly alleviates any potential hazard
from the relatively high boiler pressure. Use of an organic working
fluid can introduce a significant hazard relative to water, however,
because of the flammabUity and toxicity of the working fluid. While
these hazards cannot be completely eliminated, proper mechanical
design of the system coupled with design for minimum working fluid
inventory should reduce the hazard to an acceptable level for the first
experimental prototype. The safety considerations for use of thiophene
working fluid are discussed in Section 4. 2.
In selecting the working fluid for the first experimental prototype,
greater emphasis has been placed on use of a fluid which provides a
minimum cost system with excellent performance and efficiency
characteristics than on minimum hazard. Once competitiveness with
the internal combustion engine has been demonstrated in these critical
areas, with an accompanying extremely low emissions level, a major
3-15
-------
THBRMO ELECTRON
effort on hazards evaluation can be justified, including experimentally
simulated failures in the system as well as working fluid modifications
fo minimize any potential hazards. If competitiveness cannot be
demonstrated, such an effort is not justified. Accordingly, the
decision of working fluid was based primarily on selecting the existing
working fluid which offered the best potential for fulfilling the economic,
performance, and efficiency goals, particularly since the "hazard"
of the system is a" relative factor difficult to evaluate quantitatively
in other than an experimental fashion. An additional factor in the
selection of thiophene as the working fluid was the extensive experience
gained at Thermo Electron Corporation in the development of a 5 hp
system using thiophene as working fluid.
REFERENCES
1. "Post-1974 Auto Emissions: A Report from California, " Environ-
mental Science and Technology, pp. 288-294, Vol. 4, No. 4,
April, 1970.
2. Vickers, P. T. , et al. , "The Design Features of the GM SE-1 01 -
A Vapor-Cycle Powerplant, " SAE Paper 700163, January 12-16,
1970.
3. Personal Communication, April, 1970, Mr. Curtis Burkland,
Marauardt Corporation, Van Nuys, California.
3-16
-------
TMKHIMO BLBCTRON
• OIPOIATIO*
4. COMPONENT CHARACTERISTICS AND DESCRIPTIONS
4. 1 INTRODUCTION
As with any other type of automotive propulsion system, the size
of the system is set by the peak power or torque demands rather than
by the average demands. The component sizes were based on a design
power level of 100 net shaft horsepower at 2000 rpm engine speed,
corresponding to a vehicle speed of 95 mph when directly coupled
to the drive shaft. This maximum engine speed was based on a piston
speed of 1000 ft/min, above which the intake valve loss increases
rapidly. The engine for this power level is a V-4 with 184 cubic inch
displacement.
The engine design is based on use of variable intake valving for
control of the engine power output, thereby obtaining the maximum
possible efficiency from the system under part-load conditions. This
factor is of crucial importance for an automotive propulsion system,
which operates at part-load most of the time with the excess power
required only for acceleration performance. The design point intake
ratio was taken as 0. 137, based on a compromise between engine size,
boiler-condenser size, and system efficiency. For example, for
larger design point intake ratios, the engine size can be smaller for
a given power output; the overall system efficiency is less, however,
requiring a larger boiler-condenser size for a given power output.
The part-load system efficiency and performance map is also dependent
on the design point intake ratio. Increasing the design point intake
ratio lowers the system horsepower at which the maximum efficiency
occurs, while decreasing the design point intake ratio increases the
system horsepower for maximum system efficiency. Thus, the design
4-1
-------
THBHMO ELECTRON
QOKPO«»TIO»
point intake ratio can be used within limits to tailor the performance
map. For example, a lower design point intake ratio with a larger
engine would provide a more efficient system than the reference design
of this study for most driving conditions. In Figures 4. 1. 1 and 4. 1.2,
the performance maps calculated for a design point intake ratio of 0. 137
at 2000 rpm are presented for maximum intake ratios of 0. 8 and 0. 29,
respectively. The maximum intake ratio which can practically be used
is 0.8.
The maximum intake ratio has a large influence on the feedpump
displacement. In Figure 4. 1.3, the maximum organic flow rates are
presented for (IR) = 0. 8 and (IR) = 0. 29. For (IR) = 0. 8,
r x max max v max '
the feedpump must supply the maximum pumping rate down to 300 rpm;
for (IR) = 0. 29, down to 800 rpm; and for a fixed intake ratio with
max r
throttle valve control, the feedpump must supply the maximum'pumping rate
only at the design point speed of 2000 rpm. Feedpump designs have
been prepared for all three approaches for size comparison.
In Table 4. 1. 1, the design point cycle conditions are presented
as calculated with the system performance computer program, and in
Table 4. 1. 2, a summary of the design point conditions is presented.
The condenser fan and line pressure losses have not been included in
the reference cycle design point calculations. With a condenser
pressure of 25 psia and 82. 5% boiler efficiency, based on the fuel
higher heating value, the overall cycle efficiency is 13. 7%. The
design point engine release pressure is 58. 8 psia.
4-2
-------
no
100
80
2. 60
a>
in
w
o
I
r 40
20
in
vO
0 2OO 40O GOO 800
1000 I2OO
Engine RPM
I4OO 1600 I80O 2OOO 22OO
Figure 4. 1. 1 Performance Map with 184 CID Engine and
Maximum Intake Ratio of 0. 8.
-------
vO
0 20O 4OO 6OO 800 1000 I20O I40O I6OO I8OO 2OOO 220O
Engine RPM
Figure 4. 1.2 Performance Map with 184 CID Engine and Maximum
Intake Ratio of 0. ?.
-------
I
Ul
0.4
02
IO.OOO
— 80OO
— 6000 £
— 4OOO
- 2000
2OO 400 600 800
1000 1200 1400 1600 I80O 2000 2200
Engine RPM
i
Ul
Figure 4. 1. 3 Maximum Intake Ratio and Maximum Organic Flow Rate
as Functions of Engine Speed.
-------
TABLE 4. 1. 1
SYSTEM PERFORMANCE COMPUTER PROGRAM PRINTOUT FOR DESIGN
POINT CONDITIONS
• ••«» 1MAKL KAT10 • *13687 PISTON SPEED • 1000*0 •••••
THEfA • <*J»tti7
PI « bUU.UU Tl •.. 530.00 HI • 123.40. S»l.. •. _«31b'53E^Oi VV«i..»l8730
\>d • bH.Kll Tif • 37b.HO H2 • B ' *'d •' *131553E-01 ' V2 4 1*7332
» 2H.y-J6 TCONiJ • 216.20 HVAP • 39*6^3
i IPh.ii'J VSU13 • •16'»'Ktt.-01 HSUb • -'125*84 _-...-
H'J • 77*182 Ht • 43*388 H« • -1?4*03 M9 • "90*236
U^IV • 1.61Jd UhtV • .4bb^2 01 • 5*3661 02 •2*4396"
H.UWHAFL « /J7/*^ WSHAFT • .37.399 WES •48*887
WNtT « 3^«bHH H.P* • 103.15
Jl!
• .796^1 ....-.,.
J \J \J ^
UfUlLt* • elJ.bJfcll HhOILLH • 1576069.0 EFFBOILER • *82500
U
-------
1-1068
TABLE 4. 1.2
DESIGN POINT SPECIFICATIONS
Working Fluid
Boiler Outlet Temperature
Boiler Outlet Pressure
Boiler Heat Transfer Rate
Boiler Efficiency (HHV)
Engine Displacement
Engine Speed
Engine Piston Speed
Engine Horsepower Less Feedpump
Power
Engine Thermal Efficiency
Engine Mechanical Efficiency
Engine Overall Efficiency
Engine IMEP
Regenerator Effectiveness
Regenerator Heat Transfer Rate
Condensing Temperature
Condensing Pressure
Subcooled Liquid Temperature
Condenser Heat Transfer Rate
Organic Mass Flow Rate
Organic Volumetric Flow Rate
Feedpump Overall Efficiency
Feedpump Power
Cycle Efficiency
Overall Efficiency
Thiophene
550°F
500 psia
1.58 x 10 Btu/hr
82. 5%
184 in3
2000 rpm
1000 ft/min
103. 2 hp
84. 6%
91.5%
77.5%
127.4 psi
90. 0%
0. 249 x 10 Btu/hr
216. 2'F
25. 0 psia
196. 2'F
1. 25 x 10 Btu/hr
7377 pounds/hr
15.1 gallons/min
79.7%
5.25 hp
16..7%
13.7%
4-7
-------
THERMO ELECTRON
4. 2 SAFETY CONSIDERATIONS FOR THIOPHENE WOr.--.lNG FLUID
4.2.1 Introduction
As mentioned previously, greater emphasis was placed on the
use of a working fluid which provides a minimum cost system with
excellent performance and efficiency characteristics, than on minimum
hazard. At the same time, it was not desirable to select a fluid which
was obviously unacceptable from a hazard point of view, such as a
highly poisonous material (where slight leakage of the working fluid
would represent a lethal threat to those around the car). The opposite
extreme of requiring complete non-toxicity and non-flammability is
equally undesirable, since this requirement restricts consideration
to one, and only one, working fluid, steam; the impracticality of using
steam for portable power systems has been amply demonstrated.
Thiophene was, therefore, selected as the state-of-the-art working
fluid with characteristics fulfilling the development goals, and with
acceptable flammability and toxicity, at least for the initial demonstra-
tion prototypes. This selection has been justified by two years of con-
tinuous use of thiophene at Thermo Electron Corporation in Rankine-
cycle component and system development without accident or injury
to any personnel and without special safety precautions except to insure
that the development laboratories are adequately ventilated. This work
has involved operation of test loops and disassembly and inspection of
test engines.
4.2.2 Flammability and Toxicity of Thiophene
Relatively little work has been performed on the quantitative
flammability and toxicity characteristics of thiophene. In this section,
a summary of the available information will be presented.
4-8
-------
THERMO «i.«CT*OM
OOHPOiiTIOII
Thiophene has the chemical formula:
H C C H
; ii | '
X/
^c*
In Table 4. 2. 1, a summary of its flammability characteristics and a
comparison with common liquid fuels are presented. Definitions of
the flammability parameters used in evaluating the flammability
hazard of fluids are summarized in Appendix A. It is apparent that
thiophene lies between gasoline and kerosene as a fire hazard, using
flash point as the most appropriate criterion for hazard comparison.
While a higher flash point would be advantageous with respect to low
temperature leaks, it is doubtful if any organic working fluid would
have a flash point above the maximum cycle temperature. Since the
working fluid could be released as a vapor at the maximum cycle
temperature in case of a structural failure in the engine or in the
tubing connecting the boiler to the engine, such as might occur if
the vehicle were involved in an accident, the fire hazard in the engine
compartment with a Rankine-cycle system using organic working fluid
has to be considered as higher than in a conventional internal com-
bustion engine burning gasoline. In Section 4.2. 3, system design
features to alleviate this fire hazard are discussed. It should be
noted that a considerably less flammable fuel can be used for the
Rankine-cycle system than for gasoline-fueled internal combustion
engines.
4-9
-------
THKRMO
1-1072
TABLE 4.2. 1
FLAMMABILITY CHARACTERISTICS OF THIOPHENE
AND COMMON FUELS
Characteristic
Heat of Combustion
(HHV)k Btu/lb
Flash PoinV CC, °F
Fire Point, °F
Autoignition Temp. , *F
Explosive Range
Thiophene
14350
20-F(2)
2OT(2>
7SS«
—
Gasoline
20460
-50 -F("
—
495<"
1.3-6.0%(1)
Kerosene
(No. 1
Fuel Oil)
idsoo
100-165(1)
—
490(1)
1.16-6.-0%(1
Diesel Fuel
(No. 2
Fuel Oil)
20500
100-190(1)
494("
—
4-10
-------
THERMO KUECTROM
Considerable work, though by no means complete, has been
carried out on the toxicity of thiophene. A review of this work is
given in Table 4. 2. 2; in Appendix B, an independent summary of the
thiophene toxicity is given and compared with that of gasoline. It
appears that the only significant problem area with thiophene toxicity
is vapor inhalation; a concentration of 2, 900 ppm caused severe effects
so that an acceptable concentration would be considerably less than this
level. No information is available on the chronic effects of repeated
or prolonged exposure to low concentrations (100 to 1,000 ppm) of
thiophene. It would appear that thiophene is somewhat more toxic
with respect to vapor inhalation than gasoline, but not by a large
factor.
To provide a perspective with respect to the toxicity hazard,
the internal combustion engine in well-tuned 1970 automobiles emits
about 10,000 ppm of CO. In Figure 4. 2. 1, a plot is presented of the
effects of different CO concentrations on humans. Sax describes
an hour's exposure to 1,000 - 1,200 ppm as dangerous, while exposures
to 4,000 ppm are fatal to humans in less than an hour. It is apparent
that the CO concentration in the exhaust from well-tuned internal-
combustion powered automobiles is much greater than that required
for death to humans, even on short-term exposure. It is also apparent
that CO is considerably more toxic than thiophene, based on the available
data. Thus, current automobiles, in operation, continuously emit a
poison in much greater concentrations than are required for human
death. Fatalities due to accidental CO exposure from automobile
exhausts are relatively rare, though they do occur.
4-11
-------
TABLE 4. 2. 2
SUMMARY OF TOXICITY OF THIOPHENE
Source of Information
MapoM »y Yuun|er
Lanoratoriea on *orh
Company, Si. Lou it.
upon anal"|V uilhi *
mat* brhrvtd looml. *
F. flury *™t ». ZrrnliV*
A. r'r.rt.i a ''
Iht Mrnfc In. it.
Rll. f>iiiiun
Oral Tomtcttr
L[l^0 - ». 1 |m«Ai «<
body m«t« (or r»t»; lower
*nd upp*r ILmita * 2.S |m«/h(
Knd 1. • |rn*/hf ; compound
LO^Q • del* required lor %0%
,.. ,. no ,,.'.U..,M '.
,^_
Skin Absorption
••«!f'^TS:.."-»^r
(or tftbbll*
Compound cl»**«d •• »li|hUr tone.
Ala Ir rit«tt*n
CompouMd cltt «d ••
modtrat* akin rrlteal
mAAimufrt «cor w«>
• In 24 hour*.
it tpplicd.
Ey« IrrltatloM
C*nipo«tnd cl*«««d *•
m«d*rat* •r* trriunt
m«aimnm acorc •••
1 10 in 24 houra.
Applied.
Vapor InhaUtion
Concentration o( T/. 000
ppm .'*aull*d in d«ath ol
1'hia >• not un«>p«rt«d tn
by velum*) produced in
the t»*l. Other volarile
•-!•> |*ao1lne *nd
to Ka*e acute lethal e((ecia
tn ronc«nlr«tlona at or n«*r
S% lif volume.
1.900 ppm tauetJ lom ol
deaih of mil*.
Dw( »ai|hi - 11 >g; un* injetdun
ol appromm^tely t gma per day
Firai mfprlion - no e((ect.
0. •< (.mi/hi.
I
I—*
o
-------
1-1074
Figure 4. 2. 1 Effect of CO Concentration and Time
on Carboxy Hemoglobin.
4-13
-------
TNBMMO BLMCTMON
4. 2. 3 System Design Concepts to Minimize Hazard from Flammable
and Toxic Working Fluid
From the preceding discussion, it is apparent that thiophene
does represent a flammability and toxicity hazard. In this section, a
discussion is presented of system design concepts to minimize this
potential hazard.
4. 2. 3. 1 Thiophene Leakage from the System
The system is designed as a completely sealed, leak-tight system,
and must be maintained basically vacuum-tight for satisfactory operation
of the system. Only one rotary shaft seal is used; to minimize the poten-
tial for leakage from this source, a double-face seal is used with pres-
surized buffer fluid of the system lubricant between the seals. This oil
pressure is maintained above the working fluid pressure at all times so
that the small leakage normally encountered with a face seal results in
lubricant entering the system rather than thiophene leaving the system.
All static seals will be the equivalent of an "O" ring vacuum tight seal
using a material (Viton) compatible with thiophene and having a long
"shelf-life" (resistant to oxidation and other environmental conditions).
It should also be noted that the system is at positive gauge pressure only
when operating. When shut down and cooled, the thiophene pressure is
much less than atmospheric pressure so that any leakage is either air-
leakage or oil-leakage into the system rather than thiophene-leakage out
of the system.
The entire system is conservatively designed from stress consid-
erations to prohibit rupture of the various components. Part of the manu-
facturing procedure of the system should be pressure testing of all
components according to ASME code specifications and complete vacuum
leak-testing of the complete assembled system, followed by hot operation
of the system for 15 minutes. Surface coating i an be used on carbon steel
surfaces to prevent corrosion over the 10 year lite of an automobile.
4-14
-------
THBRMO ELECTRON
tOIPOIATIOH
In case of accident, the most vulnerable component is the con-
denser, which is placed in the same position as the radiator on an internal
combustion engine-powered automobile. If this placement proves un-
acceptable, an alternative is to use water as a heat-transfer intermediate,
as illustrated in Figure 4.2.2, so that the thiophene condenser placement
can be in the rear of the engine compartment. This option, of course,
results in a significant increase in the cost of the system and should not
be considered unless experimental evidence indicates that front end place-
ment of the thiophene condenser would result in an unacceptable hazard in
case of front end collisions.
An additional factor of importance is to minimize the thiophene
inventory in the system. The fluid inventory in the design presented in
this report for the 100 shaft horsepower system is 30 Ibs. It is expected
that this inventory can be reduced by 5 to 10 Ibs with refinement in the
system design. It is unlikely that the fluid inventory can be reduced below
about 20 Ibs.
Safety controls must be incorporated to prevent system damage in
event of malfunction of the normal control system. For critical param-
eters, such as boiler pressure and condenser pressure, two levels of
safety control are proposed: one, to detect off-design performance and
shut the system down before a serious operating condition is reached;
the other, final pressure relief to prevent system rupture and to control
the position of thiophene release from the system, minimizing the poten-
tial for ignition of the thiophene. For example, a rupture disc is used on
the condenser to prevent condenser rupture in the event of simultaneous
failure of the normal control system and of the safety control to shut down
the system. The exhaust rupture disc would be ducted beneath the car to
reduce the potential for ignition of the thiophene by electric sparking or
4-15
-------
WATER COOLED THIOPHENE
CONDENSER
THIOPHENE
VAPOR
WATER
PUMP
AAAAAAAAAAA
AA/WWVAA/V
t
THIOPHENE
LIQUID
I
»—•
o
AIR COOLED WATER
COOLER (RADIATOR)
Figure 4. 2. 2 Alternate Thiophene Condenser to Reduce
Front-End Collision Hazard.
-------
THBRMO ELBCTMON
other potential sources of ignition in the engine compartment. The high
density of thiophene vapor and the cooling air movement through the
engine compartment would prevent the thiophene vapor from re-entering
the engine compartment in appreciable concentration.
4.2.3.2 Effects of Leakage
Even with conservative design, leaks can develop in the system,
from- either normal u«-e pr accidents which result in rupture of parts
of the sealed system. It should be noted that the system, when cold,
is at subatmospheric pressure so that the thiophene will not leak from
the system under these conditions.
To aid in detection of small leaks in the system, a small concentra-
tion (ppm) of an odiferous material can be added to the highly purified
thiophene (which has a relatively faint odor). This procedure is identi-
cal to that used for natural gas supplied for residential use, where
highly odiferous mercaptans are added to the natural gas for detection
of leaks before dangerous concentrations are encountered. This
approach, coupled with the subatmospheric pressure of the cold system,
should greatly reduce the hazard when the vehicle is parked in a con-
fined space such as a garage.
The development of small leaks when the vehicle is operating
should also be a minor hazard, since the large volume of condenser
cooling air will sweep the thiophene from the engine compartment with
large dilution of the thiophene vapor in the air.
The principal hazard thus comes from large leaks resulting
from rupture of lines, with the system operating, where flammable
thiophene-air mixtures can be formed and toxic concentrations of
4-17
-------
THBMMO •LBCTMON
thiophene vapor can exist around the vehicle. No easy solution to this
situation exists. To reduce flammability hazards, the engine compart-
ment could be equipped with a CO system which would be released by
a sensor located in the air duct leading to the burner. The sensor could
be a heated, catalytic coil located in the engine compartment which would
burn any flammable thiophene-air mixture which came into contact with it,
thus raising the temperature in the sensing circuit and simultaneously re-
leasing the extinguisher and shutting the burner down. An inertia switch
could also be incorporated for immediate release of the fire extinguisher
system in event of accident. Considerable experimentation will be re-
quired to establish the practicality of this approach.
The toxicity hazard in event of a large leak is difficult to
define. Once the vapor leaves the engine compartment, dilution with
air can be expected, probably reducing concentrations to tolerable
levels around the vehicle. In the passenger compartment, except in a
convertible, an additional dilution would be expected relative to that
outside the car. Experimental work is required to determine the actual
concentrations. The odor of the thiophene, or of odiferous additives, will
provide positive indication of the presence of thiophene vapor, in any case,
so that the vehicle occupants can move away from the vehicle to a point
where the vapor concentration is very low.
The optimum solution to the hazard problem is to find a working
fluid which provides the thermodynamic and economic advantages of
thiophene, and which is also fire-resistant and much less toxic. Such
a fluid does not now exist, however, and a major effort is required
for the development of such a fluid. Discussion with the various chemi-
cal companies involved in thermodynamic fluid development indicates
4-18
-------
THHRMO KUKCTRON
a high probability for synthesis of such a fluid once the practicality of
the Rankine-cycle system from performance and economic considerations
has been established and the incentive for the large investment required
for development of such a fluid can be justified. With development of
such a fluid, the automobile with a Rankine-cycle propulsion system
should be safer than the present internal combustion engine, due to
elimination of the CO toxicity hazard and the use of a fuel considerably
less flammable than gasoline. Even with the thiophene working fluid,
the overall hazard of the Rankine-cycle system may be comparable to
the internal combustion engine system, based on the reduction in deaths
due to CO poisoning, which should not occur at all for the Rankine-cycle
system, and to gasoline fires in accidents, which should be significantly
reduced due to use of a less flammable fuel.
4-19
-------
THBRMiO •t.MCTHOM _
4. 3 ENGINE (EXPANDER) DESIGN
-4»-3."l Performance Estimates
In order to determine proper operating speeds and inlet and
exhaust valve sizes, a simple analytical model of the expander was
constructed, and mechanical and indicated efficiencies were predicted
for a range of speeds and loads.
a. Prediction of Mechanical Efficiency
The mechanical efficiency is defined as follows:
Brake Mean Effective Pressure (3MEP)
c_
Indicated Mean Effective Pressure (IMEP)
Data from Thermo Electron Rankine expanders and internal
combustion engine data were used to derive the following expression
for mechanical efficiency as a function of mean piston speed
[S = (2) (stroke) (rpm) ] and IMEP:
„ f 174 . ,1 ,n-5 5.46 „ _
This expression is plotted for a range of S and IMEP in
Figure 4.3. 1.
b. Prediction of Indicated Efficiency
The indicated efficiency of the expander is defined as follows:
IMEP
i Isentropic Indicated Work
The analysis is based on the following assumptions:
4-20
-------
100
90
80
70
60
50
0
5 * Mean Piston Speed,ft/min
Jf - Mechanical Efficiency, %
'
I . I
20 40
60 80 100 120
IMEP.psia
WO 160 180 200
I
U>
o
Figure 4. 3. 1 Engine Mechanical Efficiency Variation with IMEP for Various Piston Speeds
-------
THBMMO ELECTRON
(1) There are three types of losses in the expander:
throttling through the inlet valve, throttling through
the exhaust valve, and heat losses.
(2) No blowdown losses due to exhaust ports being uncovered
before bottom center occurs.
(3) No recompression occurs.
(4) The expander has zero clearance volume.
(5) The pressure losses are assumed to be such that they
appear as straight lines on the P-V diagram, as shown
in Figure 4. 3. 2.
Assuming incompressible flow through the inlet valve, the
orifice equation can be used to derive the following expression for
the average pressure loss through the inlet valve:
v
A
A.
IV
22 2
V
1 /Si i 360
2 9
V
2
where
6 = cos (1-2 v./v )
A /A. = ratio of picton area to average inlet valve area
* (see Figure 4. 3. 3 for explanation)
g = acceleration due to gravity
S = mean piston speed
C = inlet valve flow coefficient (0.6 assumed)
v. = inlet specific volume
i
v = release specific volume
0 = crank angle for intake opening
4-22
-------
1-1104
iv
theoretical card
— analytica1 model
ev
Figure 4. 3. 2 Sketch Illustrating Analytical Model
Used for Expander.
4-23
-------
Acceleration otL/6.
.4 .5 .6
Intake Ratio,v,/v2S
.7
.8
.9
I
H^
o*
Figure 4. 3. 3 Variation of Piston Area to Valve Area Ratio with Intake
Ratio of Engine
-------
TH
The work lost during the intake process is:
AW.
.
IV
. )(v.) .
IV 1
The exhaust process can be treated the same way, and the
resulting relationship is:
2
ev2
1
S2 1 / EOi/ApAevl
—
ev2
AP
where
ev2
evl
1
EO
area of auxiliary exhaust ports
area of blowdown exhaust port
flow coefficient of blowdown exhaust ports (0. 6)
crank angle through which blowdown exhaust
ports are uncovered (80*)
AP = release pressure - condenser pressure.
The work lost during exhaust is
The heat loss correlation was assumed to be of the usual form of
Nu = Constant x Re .
The bore was taken as the characteristic dimension in the Nusselt and
Reynolds numbers, and the constant was determined by fitting Thermo
Electron data. The exponent, n, was taken as 0. 75, a value used in
internal combustion engines. The above relationship has a heat transfer
4-25
-------
coefficient based on piston area; thus,
Q = hA (T - T ) .
p G w
The mean gas temperature was taken as the arithmetic mean
of the inlet temperature and temperature at beginning of blowdown.
assuming a straight line temperature drop during expansion:
(v - v-)
T_ = T. (v./v ) + —| (T. + T.)
G i i 2 2 v, i 2
The mean wall temperature is based on test data taken from a
single cylinder engine operating at the same inlet conditions with a
number of thermocouples along the cylinder The resulting expression
is a function of intake ratio only, over a fairly small range:
T = 400 + 156 (v./v_)
w i 2
The heat loss in Btu/cycle for a 4 0 inch bore is:
0.75
q.1.48 -
_ - T )
G w
4-26
-------
THKRMO ELECTRON
where
k = vapor conductivity at T
G
(i = vapor viscosity at T .
With the above three expressions, the indicated efficiency can
be calculated for any piston speed and intake ratio. The process is
iterative, in that v and P are functions of the pressure loss through
the inlet valve. The above expressions were incorporated into a com-
puter program for calculating the overall cycle and engine efficiencies.
The overall engine efficiency as a function of piston speed, at a fixed
intake ratio corresponding to an IMEP ** 125 psi, is shown in Figure
4.3.4.
4. 3. 2 Engine Configuration
As a result of the rapid decline in overall engine efficiency with
piston speed above 1000 ft/min (see Figure 4.3.4), this value was
taken as the piston speed at the engine design condition of 103.2 bhp at
95 mph. The IMEP and BMEP are determined by the cycle design con-
dition (piston speed and intake ratio); the BMEP and piston speed de-
termine the piston area required to develop the desired horsepower.
A 90 ' V of four cylinders was chosen as being reasonably compact
without a large number of moving parts. With four cylinders, the
resulting bore is 4.42 inches. The mean piston speed and engine speed
are related through the following expression:
S = 2 LN,
where S = mean piston speed,
L = stroke,
N = rpm.
4-27
-------
1-314
90
80
CE
£70
o
60
50
40
Inlet Valve
Single Inlet Valve
Release Pressure - 75psia
Pi - SOOpsia
Ti * 550 °F
\
\ -
200 600 -1000 1400 1800 2200
Speed, ft/min
Figure 4. 3. 4 Overall Engine Efficiency Variation
with Piston Speed
4-28
-------
THBMMO ELECTRON
A design point speed of 2000 rpm was chosen as being a reason-
able maximum, considering the problem of driving the inlet valves (see
following section). This speed also resulted,in a reasonable bore/stroke
ratio, giving a stroke of 3. 0 inches.
The resulting expander dimensions and design point operating
conditions are given as follows:
Configuration Ffcur: cylinders, 90 °V
Bore 4.42 inches
Stroke 3.0 inches
Displacement 184 in-'
BHP (design) 103 at 2000 rpm
IMEP (design) 127 psi
BMEP (design) 117 psi
4. 3. 3 Expander Intake Valving
The discussion of Section 5 has indicated the importance of
variable intake valve closing (or cut-off). The size of the intake valve
is determined from Figure 4. 3.3. once the piston area and maximum
piston speed are known. A valve providing an average opening area
smaller than that indicated in Figure 4.3.3 would give lower expander
efficiencies than were assumed in the design point calculations.
a. Comparison with Internal Combustion Engine Valving
The resulting inlet valve must be about 1. 25 inches in diameter
and have a lift of 0.30 inches. This valve is, therefore, comparable
in size (and weight) to the valve of an internal combustion engine of
the same bore. The stress level in a cam-operated valve is propor-
tional to the lift divided by the valve event squared. The cam event
in typical four-stroke internal combustion engines is on the order of
120", whereas in the Ran kin e- cycle expander the design point intake
4-Z9
-------
THKRMO BIECTRON
ratio of 13.7% implies a valve event of no more than 60 or 70°, One
other problem associated with expander valving which does not occur
in the internal combustion engine is the high pressure against which
the valve must open (or close, depending on whether the valve opens
inward into the cylinder or outward into the inlet port). This has led
to the development of two types of pressure-balanced valves, shown
in Figure 4.3.5.
b. Mechanically-Operated Valving
Two types of mechanically-operated systems were considered
in detail. The first was a three-dimensional cam system in which
the timing would be varied by sliding the cam shaft along its axis.
Analysis showed that such a system would not be feasible even if a
three-dimensional cam were economically practical, since the cam
base circle diameter would have to approach the size of the bore of
the engine itself to give reasonable loading on the cam.
The second, more promising approach consists of operating
two concentric inlet valves in a series arrangement, as shown
schematically in Figure 4. 3. 6. The number 1 valve in Figure 4.3.6
is driven by camshaft 1, which has a fixed angular relationship with
the crankshaft; the number 2 valve is driven by camshaft 2, which
has a variable angular relationship with the crankshaft. The total
valve event is then controlled by the opening of valve 1 and the closing
of valve 2, and the cut-off point is determined by the overlap of the
two valves. With this system, long cam events are possible, resulting
in much lower stresses in cam and valve gear and larger flow areas
(at low cut-off) than could be accomplished with a single valve (see
Figure 4. 3. 3), although this advantage would probably be partially
negated due to a lower effective flow coefficient for the two valves in
series.
4-30
-------
Figure 4. 3. 5 Alternative Balanced Inlet Valves.
4-31
-------
1-315
roc
CRANK AN
-------
THKItMO BIBCTMOM
(ORFOIATIO*
A design study of the two valves in series approach was carried
out on an in-line four cylinder expander with the same bore and stroke
as the V expander. This design is shown in Figure 4. 3. 7. Because
of the very large springs required to overcome inertia and pressure
forces (these valves cannot easily be pressure-balanced), a two-cam
(or desmodromic) approach was adopted, with one cam opening the
valve and the second closing it. Both cams are on the same shaft and
actuate a single rocker. Using a cam rather than a spring to close the
valve has the added advantage of giving more rapid closing and sharper
cut-off.
c. Hydraulically Operated Valving
(1) Directly Actuated
The approacn illustrated in Figure 4. 3. 8 consists of a cam-
operated plungjer pump which operates on a hydraulic column to actuate
the inlet valve. The angular orientation of the plunger in its bore de-
termines its. effective stroke and hence the intake valve event. This
system is similar in many ways to diesel engine injection equipment,
and a manufacturer of such equipment feels that this system is feasible.
The peak pressure in the hydraulic column would be on the order of
5000 psi, and a spring force of about 400 pounds would be required
to accelerate the intake valve during closing.
(2) Pilot Operated
This system is shown schematically in Figure 4. 3. 9. A single
rotary valve (two are shown in the figure for clarity) delivers high
pressure oil from a gear pump and accumulator to alternate sides of
4-33
-------
-aoo
•1
*•>
u>
o
I
O
\
z
INTAKE RATIO
CONTROL ROO
Figure 4. 3.7 Mechanically Operated Intake Valving System.
-------
1-550
Figure 4. 3. 8 Directly Actuated Hydraulic Valve.
4-35
-------
LOW
PRt&SURE
"560
Figure 4. 3. 9 Pilot Operated Hydraulic Valve.
-------
THKRMO KLBCTHOM
a piston which is solidly connected to the poppet valve. Thus, hydraulic
pressure both opens and closes the valve, and variations in timing are
achieved by sliding the rotary valve along its axis of rotation. There
are a number of variants of this system, some of which are currently
under development by injection equipment manufacturers as dies el
engine fuel injection systems. This approach would also require a
pressure balanced inlet valve, otherwise the pressure required to
close the valve (or open it, if it opened outward) would become so high
that the work required to operate the valves would become excessive.
Preliminary calculations indicate that with an overall hydraulic efficiency
of 50%, this system will require 2 to 3 hp at full output and 2000 rpm.
d. Recommended Approach
Detailed design studies of the most promising hydraulic system
should be carried out, possibly with the aid of a manufacturer familiar
with similar equipment. A mechanical system should also be examined
in detail (probably the two inlet valves in series approach). The most
promising of these two alternatives should then be constructed in the
form of a bench test rig and developed to the fullest extent possible
before being installed in an engine.
4.3.4 Expander Exhaust Valving
The exhaust valving is completely automatic, requiring no cams
or other actuating means. The method of operation is shown schematically
in Figure 4.3. 10. The exhaust valve for the actual expander is shown
in Section C-C of Figure 4. 3. 12.
4-37
-------
OJ
oo
POWER STROKE
EXHAUST VALVE
OPENED
RETURN STROKE
EXHAUST VALVE SEQUENCE
DISPLACEMENT
INDICATOR DIAGRAM
EXHAUST VALVE
. CLOSED
n OO
' I
Figure 4. 3.10 Schematic of Exhaust Valve Function.
-------
THBRMO BLBCTRON
4. 3. 5 Engine Bearings
A detailed analysis of the engine journal bearing loading was
undertaken. The connecting rod big end bearing was selected for
analysis since its loading is more severe than the main bearings. The
piston pin bearing is also severely loaded, but an analysis of this bear-
ing was not undertaken, due to the lack of analytical technique for
studying bearings where operation is almost entirely dependent on
squeeze and partial film effects.
Preliminary analysis at both high and low speeds showed that the
300 rpm, maximum torque condition (80% intake ratio) would result in
the most severe bearing loading condition. This should be contrasted
with the internal combustion engine, where minimum oil film thicknesses
in journal bearings usually occur at high speed during the intake or ex-
haust stroke and are due to inertia forces alone in almost all modern
engines.
The oil film thickness was calculated as a function of crank angle
and is plotted in Figure 4. 3. 11. Shown along with the Thermo Electron
analysis is an analysis by Clevite Corporation, a major supplier of auto-
motive engine bearings. Thermo Electron used a 0.002-inch diameter
clearance in their analysis, Clevite a 0.003-inch diameter clearance;
the Clevite figure is probably more realistic. Both analyses indicate
minimum oil film thicknesses considerably less than the 100 microinches
generally considered to be a safe minimum. An analysis of the 302 in
Ford V-8 connecting rod bearing at full throttle and 3000 rpm predicts
a minimum oil film thickness of about 83 microinches, whereas the
Clevite analysis predicts a minimum of 30 microinches for the Rankine-
cycle engine.
4-39
-------
O.K>
I
*•
o
THICKNESS RATIO v»
ROC- Bit END
- SOOR.PM
IMTA.HC
7OUBNAL CMA - iO
JDUBHAL WIDTH -O-7S
r-so
n - lo-Ct CCNTIPOlftC
- ' J T DIAMETRAL CLCAAANC*
FILM THICKNESS
MINIMUM/ ran »ae KWO
CALCULATKO AT 3OOO
•£ CAlCULATION BT TMCBMO ELECTttOt
jam CALCULATION BY CLtV'Tt CORP
i ;
ui-
I3O' iOO'
CRANK ANGLE (ATDC)
Figure 4. 3. 11 Oil Film Thickness as a Function of Crank Angle.
-------
THERMO ELECTRON
cotroiATto*
Reduction of the maximum intake ratio from 80% to 29% should
improve the situation. If a fluid coupling is used rather than a clutch,
there should be very little bearing problem, since the engine must
speed up considerably before very much torque can be extracted. The
use of a two-speed transmission should also alleviate the bearing loading
considerably, particularly if the maximum intake ratio is limited to
0. 29. The analysis shows that the bearings are quite conservative for
higher speed operation.
It would probably not be practical to increase the diameter of
the bearings since the connecting rod would not go out through the
bore; this would cause severe assembly problems. The bearings could
be increased in length, but the engine would undoubtedly become too
long if the bearings were designed to give a minimum film thickness of
100 microinches at 300 rpm and 80% intake ratio .
Journal bearings are not feasible on an engine directly connected
to the driveshaft, since the engine must start and carry heavy torque at
practically zero rpm. In this case, anti-friction or roller bearings
must be used.
The above analysis was carried out assuming that the lubricant
is pure oil. Under start-up conditions, special precautions must be
taken to ensure that this is the case, since the oil and thiophene are
completely miscible. This is accomplished by heating the oil prior to
start-up with the hot exhaust from the burner. In this way, any thiophene
would be boiled out of the lubricating oil, thus ensuring a good supply of
high viscosity lubricant to the bearings. A more complete discussion
of this system is given in Section 4. 11.
4-41
-------
T H•R M O BLKCTMON
The first prototype expander should be constructed with caged
ne«fte'beacri'ng3-rather tharr'with-hydrodyriamic bearings so that prob-
lems" in ~the lubricating system may be solved without risking seizure
of the expander. Analysis shows that this type of bearing should function
quite well in the same geometry designed for the journal bearings.
4. 3. 6 Final Expander Design
The final expander design is shown in Figures 4. 3. 12 to 4. 3. 15.
It is shown equipped with a hydraulically actuated valve of the type shown
in Figure 4.3.8. The valve actuating pump and feedpump are driven off
the front of the expander (see Figure 4. 3. 14) and are in direct communi-
cation with the expander crankcase. The single shaft seal, which is
described more fully in Section 4. 9.2, is just aft of the rear main bear-
ing. The engine is shown coupled to the Dana single-speed transmission.
The accessory drive is located in the rear bell housing, as. shown in
Figure 4. 3-. 15. Casting thicknesses and crankcase design conform to
current automotive practice (the peak cylinder pressure of 500 psi is
roughly equivalent to that of 1C engines). Three-cornered sealing sur-
faces, which are used in most 1C engines, are avoided here because seal
joints, particularly in the crankcase, are subjected to substantially higher
pressure differentials than occur in 1C engines and must be vacuum tight.
All of the major expander components are cast iron; the crank-
shaft would be surface-hardened to provide good bearing surfaces and
would be counterweighted at both ends to eliminate the primary un-
balanced moment. Static seals would be iron with molded rubber
"o-ring" type inserts.
4-42
-------
TMKHXIO •LBCTHOI
C O R P O R ATION
Z611-D
Figure 4. 3.12 V-4 Expander with Hydraulically
Actuated Valves, Front View.
4-43
-------
K
Figure 4. 3.13 V-4 Expander with Hydraullcally Actuated Valves, Side View.
-------
2613-D
Figure 4. 3. 14 V-4 Expander with Hydraulically Actuated Valves, Showing Feedpump and
Valving Drive.
4-45
-------
T M • mm o « i m CTHON
CORP O~R » T i o N
2614-D
Figure 4. 3.15 V-4 Expander with Hydraulically Actuated Valves, Showing
Accessory Drive from Transmission End.
4-46
-------
THBRMO ELECTRON
Major problem areas in the expander are as follows:
1. To develop a reliable valving system which will provide
good flow area over the required range of intake ratios.
2. To ensure that the lubricant delivered to the bearings
is of adequate viscosity, i. e., that it has a minimum of
thiophene dissolved in it.
3. To provide a reliable shaft seal.
4. To establish some special quality control of castings
in mass production to ensure vacuum tight expander
assemblies.
4-47
-------
TMKHIMO «l»CT*OM
(oirottrio*
4.4 FEEDPUMP DESIGN
Thermo Electron Corporation has tested several different types
of positive displacement pumps with thiophene. Based on this testing,
piston pumps are the only type presently giving satisfactory performance
with thiophene for use in a Rankine-cycle system. The proposed feed-
pump is therefore a positive displacement piston pump with five cylin-
ders driven by a wobble plate. The design drawings for pumps of
three different displacements are illustrated in Figures 4. 4. 1 through
4.4. 3, and their characteristics are summarized in Table 4.4. 1. Since
the feedpump must be able to supply 15 gpm pumping rate over a range
of main engine speeds, the size of the feedpump required depends on
how the feedpump is driven and the relationship between main engine
rpm and feedpump rpm.
The organic flow rate to the boiler must be varied in response
to the vapor demand in order to maintain constant boiler outlet pressure.
Since the required pumping rate at any given engine speed can vary
from zero to the maximum rate, depending on the intake ratio setting,
it is necessary that the feedpump have variable displacement, permitting
proper adjustment of the organic flow rate with no loss in efficiency.
The method used to obtain variable displacement is similar to that
used in diesel fuel injection pumps in which rotation of a piston with
ramp undercut is used to vary the effective displacement of the pump.
Rotation of all five pistons simultaneously is obtained by means of a
central gear which meshes with gear teeth in the piston skirt; the
central gear is controlled by a rack and pinion drive passing external
to the pump through a rolling diaphragm hermetic seal
4-48
-------
TMMRMO ELECTRON
CORPORATION
A computer analysis of the pressure transients produced by the
flow ripple of piston pumps was carried out. This analysis indicated
that at least 5 cylinders were required to limit the suction side
transients so that cavitation on the pump suction did not occur. The
net positive suction head to the feedpump is provided by subcooling
of the liquid to the feedpump.
The pump could have been either crank-driven or wobble-plate
driven. A wobble-plate drive was selected for the following reasons:
compactness, easier integration with the engine, lower weight, less
vibration, quieter operation, and more convenient geometry for
incorporating variable displacement.
4-49
-------
i
Ui
o
o
o
£ -
CVJ
I— O
i a
Figure 4.4. 1 2000 rpm Feedpump.
-------
I
Ul
t\»
l\»
UJ
1
D
c
o
o
a
•o
0
X
>
O
z
Q
H
Z
•
1
1
0
•
r
0
4
>
0
I
SKT/ONA-A
Figure 4.4.2 800-2000 rpm Feedpump.
-------
©5;
//rvr \ \
® ® / ® © ©3 & &
iTo
®
IV »
I TJ
O 3,
Figure 4.4.3 300-ZOOO rpm Feedpump.
-------
1-564
TABLE 4. 4,-l
CHARACTERISTICS OF FEEDPUMPS FOR 15 gpm PUMPING RATE
Volumetric Efficiency = 90%, 5 cylinders
Pump
Illustrated
in Fig. No.'
4.4. 1
4.4.2
4.4.3
rpm Range
For Maximum
Pumping Rate
2000 rpm
800-2000 rpm
300-2000 rpm
Total
Displace-
ment
1.91
4,78
12.78
Bore
in
1. 175
1. 595
2.08
Stroke
in
0. 352
0.478
0.750
System for Which
Pump Would be
Used
Driven at constant
speed by auxiliary
engine.
Driven by main engine
with maximum intake
ratio of 0. 29, 1/1
speed ratio.
Driven by main engine
with maximum intake
.ratio of 0. 80, 1/1
speed ratio.
4-53
-------
4. 5 COMBUSTOR DESIGN ANI> CHARACTERISTICS
The combuitor requirements are:
a. Low emission level for NO, CO, and unburned hydrocarbons
for all operating conditions.
b Turndown ratio of 15 to 1-
c. Low pressure drop in compact burner
d... Hifh reliability, low maintenance, and low cost.
In this lection, a description of the combustor design and its
operating characteristics it presented. Emission levels measured
from a 140,000 Btu/hr burner at TECO which indicate the low-emission
potential Of the Rankine-cycle system are also presented, as well as a
brief description of the fuels available for use in Rankine cycle propulsion
systems
4. 5. 1 Combustor Design and Fuel/Air Supply
The combustor design of this study is based on parametric in-
formation provided by the Marquardt Corporation. Van Nuys, California,
under subcontract to Thermo Electron The information is derived
from measurements and calculations on the Marquardt SUE, or sudden
expansion burner, illustrated schematically in Figure 45.1 In Figure
4 b. 2, the predicted SUE burner characteristics are presented This
parametric plot indicates quantitatively the effect of c o.nbustlon chamber
Uameter on pressure drop and combustion chamber li-ngth for a burning
rate of 2. 0 x 10^ Btu/hr and equivalence ratios of 0 6, 0. 8 and 1 0 The
•rombustion chamber volume in this plot is based on a burning density of
/. 8 x 10 Btu/hr-ft at 1 atmosphere pressure Marquardt has obtained
limited experimental data which agree with the plot ami .He experimental
-------
in
in
FUEL MANIFOLD
COMBUSTION ZONE
TO HEAT E)CHANGER
>-
IGNITER
MULTIPLY INJECTOR NOZZLES
Figure 4. 5. 1. Marquardt Sudden Expansion Burner (SUE)
-------
1-945
T
LENGTH OExperimental Pat same mass
Flow/Unit Area,0 =0.8
0=0-6
A Experimental Length at
Equivalent Heat Release
8 10
Diameter, inches
10
8
12 14
Figure 4 5. 2 Predicted SUE Burner Characteristics.
4-56
-------
THERMO ELECTRON
points are indicated. Marquardt has also obtained burning densities
up to 2. 5 x 10" Btu/hr-ft in a 500,000 Btu/hr version of this burner.
Thus, this parametric plot should provide a reasonably accurate method
for sizing of the burner. Scaling to different burning rates can be made,
since, for a given burner diameter, the AP varies as the square of the
burning rate and the burner length varies directly with the burning rate.
Modification of this burner concept has been used in the burner
design developed in this study. The modifications involve (1) curving
of the combustion chamber to conform to the boiler tube bundle shape
for integration with the boiler, (2) use of a single fuel nozzle using
compressed air for atomization, and (3) use of swirl vanes on the inlet
nozzle to provide better fuel-air mixing and more stable combustion.
In Figure 4. 5. 3, cross-sectional views of the automotive-size
burner are presented, and in Table 4. 5. 1 the design point burner char-
acteristics are presented. The burner shape is designed to conform to
the boiler tube bundle shape and sits directly on top of the boiler tube
bundle. Two combustion chambers, each with a maximum burning rate
of 1.05 x 10 Btu/hr, are used in parallel to provide a pressure drop of
only 1.5 inches w. c. at full burning rate within the allowable combustion
chamber diameter of 7.0 inches. The maximum burner diameter is
restricted because of packaging considerations. It is essential to main-
tain a minimum combustion side pressure drop through the burner-boiler
combination, since the combustion system must be operated electrically
on startup and it is desirable to run the combustion system full-out on
startup to minimize startup time. Care will be required, however, to
insure that instabilities in the operation of the two parallel burners do
not exist.
4-57
-------
TMBUMO •LBCTHOI
CORPORATION
101,1 -
L
Figure 4.5.3 Cross Sections Through Auton; itiv»-Size Burner.
4-58
-------
VHBRMO ELECTRON
1-1075
TABLE 4. 5. 1
BURNER CHARACTERISTICS
Maximum burning rate for each of two burners
Combustion chamber diameter
Maximum volumetric energy release rate
Combustion chamber length required
Maximum pressure drop through burner
Combustion chamber mass (22 Ga SS)
Air distributor mass (22 Ga CS, 20% free area)
Plenum cover mass (22 Ga CS)
Fastening hardware, nozzles, electrodes,
strengthening struts, etc.
Total combustor mass
Heat transfer coefficient to burner wall
(Measured by Marquardt Corporation)
Combustor wall operating temperature
Combustion air temperature rise at inlet of
nozzle
Average residence time in burner at maximum
burning rate
1.05 x 10 Btu/hr
7. 0 in.
2.8 x 106 Btu/hr-ft3
17. 0 in.
1. 5 in. w. c.
6.3 Ibs
5.4 Ibs
8.2 Ibs
2. 1 Ibs
22.0 Ibs
2.2 Btu
hr-ft2 °F
1500-1800°F
35-45°F
12. 5 millisec
4-59
-------
THKRMO BLBCTRON
During operation, combustion air from the blower is directed
into the outer container which serves as a plenum. The air then flows
through a perforated plate surrounding the combustion chamber and
around the combustion chamber to the inlet end of the two burners. The
combustion air is thus used to cool the combustion chamber wall: the
distributor plate permits regulation of the air flow to eliminate potential
hot spots.
The air flows through the nozzle into the combustion chamber
through swirl vanes (Section A-A of Figure 4.5.3). Fuel is sprayed
into the chamber by an air-atomizing nozzle; this type of fuel nozzle
was selected for the following^ reasons:
a. It gives a finer fuel spray than pressure nozzles.
b. It can operate well over fuel flow rates from zero to
rated capacity, i. e., it has a large turndown ratio.
c. "It has a much larger fuel orifice than a pressure nozzle,
insuring maximum freedom from clogging.
Combustion occurs and is completed in the combustion caamber, and
the hot combustion products are then directed downward into the central
plenum of the boiler tube bundle.
The requirements and characteristics of th,.>. fu" / >ir supply system
are summarized in Tables 4. 5. 2 and 4. 5. 3. The i^a.-* irr.v;m combustion-
side pressure drop through both burner and boiler is •* • i- %v c. The
maximum blower horsepower is 0.60 hp and t'...: .-orruu•<:bsor oower
1/3 hp. For startup, the battery must supply 99 ar: : ^. i£ V d^. i.u th*
combustion system. For normal operation, \.\\ • :!'• .it-.. ahu.n.i supplv
75.0 amps on the average to the combustion sy.'c .-.».
4-60
-------
TMKUMO •LKCTROM
COIFOHiriOl
TABLE 4. 5. 2
FUEL-AIR REQUIREMENTS AND CHARACTERISTICS
Fuel Composition (Assumed)
Carbon 85%
Hydrogen 15%
Higher Heating Value (Calculated) 21,600 Btu/lb
Lower Heating Value (Calculated) 20, 180 Btu/lb
Excess Air 33%
Air/Fuel Mass Ratio 19.8
Equivalence Ratio 0. 752
Reference Cycle Burning Rate 1.91 x 10 Btu/hr
(82.7% HHV Boiler Efficiency)
Maximum Burning Rate 2. 10 x 10 Btu/hr
Fuel Flow Rate (Max. Burning Rate)
Ib/hr 97.2
gal/hr (p = 47.6 Ib/ft at 95°F) 15.3
Air Flow Rate (Max. Burning Rate)
Ib/hr 1921
CFM (95 °F, 1 atm) 447
4-61
-------
1-1077
It M O ELECTRON
C"?6 in
page 4-73) M 4. 75 in.
K 3. 79 in.
Weight ~ 2. 0 Ibs
Motor Power Requirement
440 CFM Lcj, rr\:
Shaft Power 446 watte 2v vatca
Electrical Power
(60% Motor
Efficiency) 745 watts j.v.tcs
Amperage at
12 V dc 62. 0 amps '.:,. 1 arnps
Total Weight Including Moto i 2(.. 0 iba
4-62
-------
i- . J I O
TNBMMO BLHCTNON
Fuel Atomizing Air
Pressure Required
Air Flow - SCFM
Compressor Cast Oil-Less Model No.
Speed
Motor Size Specified
Ideal Power Required
Motor Shaft Power Requirement
Motor Efficiency
Electrical Input
Amperage at 12 V dc
Weight including Motor
Fuel Supply (Driven by Compressor Motor at 1725 rpm)
Pressure Required
Pumping Rate
Ideal hp
Shaft Power (50% Pump Efficiency)
Electrical Power (60% Motor Efficiency)
Amps at 12 Vdc
Weight of Pump
Ignition and Flame Sensing
Ignition
Electric Power Requirement
Amperage at 12 V dc
Flame Sensing
Electric Power Requirement
Amperage at 12 V dc
Total Weight
9.5 psig
4.6
0740
1725 rpm
1/3 hp
0.25 hp
248 watts
60%
414 watts
34. 5 amps
22 Ibs
25 psig
15 gph
0.00364 hp
0.00728 hp
9. 1 watts
0. 76 amps
0.65 Ibs
Electrode Spark
12 watts
1. 0 amps
CdS Cell
8.4 watts
0. 7 amps
0.35 Ibs
4-63
-------
Electric Requirement for Combustion System
Vellmge
Total
12 Vdc
Arnps
Startup
99. 0 amps
Op crating -Peak 97. 3 amps
Operating-Min 64.2 amps
Operating- Estimated Average 75. 0 amps
Total Combustion System Weight 43 Ibs
Z-c-.: - _.
Ar-;p
fa)
5\3f-
era.-
tj* —
/- ~- '
Px<"
V \ s
K— 0— «•
i . /.
^X
-V
-^
c .
-
\
'
1
1
-L
-- -
-
N
E
_i
A
i
V,, J
— - -1 1 - - <— .
(rad.)
(/
t
' L-
I
Ji
FOR DOUBLE
^INLET HOUSING . ' • : - -
\ I: g-
1 - :
JL
M
i
KM
SINGLE INLET HOUSINGS
4-64
-------
T HBI MO gi-BCTROM
4, 5, 2 Emission Levels from Rankine-Cycle Burners
Of great importance is the level of emissions that potentially
can be expected from Rankine-cycle automotive propulsion systems.
While a full-scale burner for this system has not yet been constructed
and tested, the experimental evidence available indicates strikingly
low emission levels for NO, CO, and unburned hydrocarbons in an
external combustion system without air preheat. In Figures 4. 5.4
through 4. 5. 9 and Table 4. 5. 4, measured emission levels from a
burner developed at TECO for use in a 3 kwe engine-generator set
are presented, indicating extremely low emission levels under both
steady-state and transient operation. This burner is l/9th the size
of each burner required for the automotive propulsion system. Also
presented are burner-on and burner-off transients indicating that
startup or shutdown of the system should not be troublesome.
The Marquardt Corporation has made some measurements on
a. 500,000 Btu/hr SUE burner, operating with primary air only, at an
equivalence ration of 0. 75. Typical steady-state results from this test
(8)
are:
Unburned hydrocarbons 4 ppm
CO 60 ppm
NO 90 ppm
Results presented by General Motors on the burner used in the
GM SE-101 steam-powered car are illustrated in Figure 4. 5. 10. At the
lowest fuel/air ratio used, at 60% of design air flow, UHC concentration
is 8 ppm, NO is 75 ppm, and CO is 300 ppm. The GM burner has a
quoted heat release rate of 3.4 x 10 Btu/hr, a pressure drop of 13. 6
6 , 3
in w. c. , and a burning density of about 5 x 10 Btu/hr-ft . This burner
4-65
-------
1-916
i
0.
Q.
Z
o
£
Ul
200-
180-
160-
140 —
120-
100-
80-
60 -
40-
20-
STEADY S
Q=50,000
FUEL JP-4
*^ * "r^t^
™ * *
U ' I I I I I I
0 10 20 30 40 50 60
x -
NO
- CO
- CH
EXCESS AIR (%)
Figure 4. 5. 4 Effect of Excess Air on Emissions, 50,000 Btu/hr
4-66
-------
I
o»
-J
5
Q_
g
Cfl
200-
180-
160-
140-
120-
100-
80-
60-
40-
20-
0-
I
0
STEADY STATE DATA
Q=105,000 BTU/HR
FUEL JP-4
NO
i
10
I
20
i
30
i
40
I
50
i
60
EXCESS AIR (%)
Figure 4. 5. 5 Effect of Excess Air on Emissions, 105, 000 Btu/hr
-------
1-918
i
CL
W
Z
g
i
ai
200-
18O-
160-
140-
12O-
100-
80-
60-
40-
9fU-l
T=O BURNER OFF (NOT COLD)
T=0+ BURNER ON
Q=105,000 BTU/HR
25% EXCESS AIR
FUfcL JP-4
I-
0 1.0
\
2.0
I
TIME T (MIN)
Figure 4. 5. 6 Burner On Test, 105, 000 :itu/hr
4-68
-------
260-
240-
220-
200-
~ 180-
5
S- 160-
Q.
— *
c/> 140 -
Z
2 120-
0)
c/3 100-
uj 80 —
60-
40 -
20,
— •*
i
MIJ i — v* i^wi«i«h.ri v/i i \i«vy
f x T=0+ BURNER ON
Q=50,000 BTU/HR
25% EXCESS AIR
FUEL JP-4
NOTE: CO ANALYZER C
DRIFTED
CALIBRATE
ZERO 0
SPAN 200 PPM
'l
CO
1 A
;|
':\
I V
f' V-^-«-« ^CO
>CHX
1 / 1
3 1.0 2.0
1 VrV/l.
ALIBI
Fir
-5
190
3.0
TIME T (MIN)
Figure 4. 5. 7 Burner-On Test, 50,OOOBtu/hr
•>£>
-------
1-920
T=0 BURNER ON
T=0* BURNER OFF
Q=105,000 BTU/HR
25% EXCESS AIR
0.
V)
EMISSIOIS
200-
180-
160-
140-
120-
100-
80-
60-
40-
20-
C
dl CAUY
*I*D A ^1 ^ 1 E
TRANSli
FUEL JP-4
CHx
j
* t \ * —
/\ NO
' \
1 \
1 I I
I 1.0 2.0 3.0
CO
CH
TIME T (MIN)
Figure 4. 5. 8 Burner-Off Test, 105, 000 Btu/hr
4-70
-------
1-921
5
o.
o.
z
0
00
-------
1-946
8 v
&•»&•>
I I
i-i
1C. 101 BURNER
«0ft DESIGN AIRFIOW
' UD°f BIT
»"HjA BIP
a. Variation of measured exhaust emissions
concentrations with overall air-fuel ratio.
(9)
rr
11
.'lU Mil
^
S^Hl
-------
1-906
TABLE 4. 5.4
TRANSIENT EMISSION DATA
FIRING
RATE
(BTU/HR)
105,000
50,000
50,000
50,000
50,000
105,000
105,000
105,000
50,000
50,000
50,000
50,000
105,000
105,000
105,000
50,000
50,000
50,000
EXCESS
AIR
25%
25%
25%
?5%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
25%
CH
X
(PPM)
6
-
5
-
4.5
401
7
5
5
4
4. 5
151
-
6
-
4. 5
4
CO
(PPM)
60
-
30
25
70
1801
90
-
80
25
20
1000+1
-
75
75
30
15
NO
X
(PPM)
-
-
-
-
, m
-
-
-
-
-
•
-
-
-
-
-
422
392
ELAPSED
TIME
(MIN)
0
.5
I. 0
1.5
2.0
2.5
3.0
3. 5
4.0
4.5
5.0
5. 5
6.0
6.5
7. 0
7.5
8.0
9.0
NOTES:
1. Short duration peak-^ 10 sec.
2. Exhaust gas sample collected during
the 9 minute run. Then two samples
were drawn and NO analysis performed.
4-73
-------
THBKMO BJ.BCTRON
COIPOIATIOI
has low UHC and relatively low NO emission with a relatively high CO
emission. The Marquardt burner has very low UHC and CO and
relatively high NO emissions. The TECO burner is simultaneously
low in all three contaminants at 25% excess air ratio. The results
from all three independently developed burners confirm, in general,
that very low emission levels will be obtained from a Rankine-cycle
automotive propulsion system, provided a reasonable overall system
efficiency is attained.
4.5.3 Some Considerations in Fuel Selection
The main considerations in selection of the fuel are:
a. Low ash and sulfur content.
b. Easy ignition and clean burning (volatility or low end of
fuel affects ease of burning).
c. Safety (flash point of fuel).
d. Availability.
Short-term: Available at local distribution for large
Otto-Rankine ratio.
Long-term: Available at local distribution for large
Rankine-Otto ratio with minimum
processing, considering other demands
such as kerosene for aircraft and No. 2
fuel oil for heating.
In Tables 4. 5. 5 and 4. 5. 6, some of the fuel characteristics influencing
(10)
these considerations are presented.
4-74
-------
TABLE 4. 5. 5
BROAD CONSIDERATIONS OF VARIOUS PETROLEUM FUELS
Product
Propane
Gasoline
JP-4
Present
Volume Availability
Limited
Ample
Ample
Retail
Distribution
Present Future
Limited
Wide
Spread
Nil Poten-
tially
Wide
Spread
Performance
Advantages
Clean Burning, Easy Ignition &
Startup
Easy Ignition and Startup, Good
Handling, Low Sulfur
Easy Ignition and Startup,
Good Handling, Low Sulfur,
Reduced Vaporization Lois
Apprnx.
Bulk
Disadvantages Price f/gal
Needs Pressure 7.Sd
System, Safety,
Luw BTD Content
Safety, Burner 12.25*
Deposits and
Corrosion,
Vaporization Loss
Safety
CoooMntary
Ideal from Standpoint of
Emissions, Startup, Evapor-
ation Loss
Probably Unacceptable lit'cause
of TEL and Halidc Scavengers.
Possibly will be Available
Unleaded (?)
Maxi.num Availability, Only
Mi.inr Processing Required
in Manufacture
Jet A, No. I Burner Limited
(or Dl«s«l), Kerosene
No. 2 Burner (or Diesel) Aaple
No. 3 CT (Turbine) Limited
Limited
Broad
but Thin
Poten-
tlally
Wide
Spread
Easy Ignition and Startup, Good
Handling, Low Sulfur, Safety,
Little Vapor Loss
Higher BTU/gallon, Basically
Same Advantages as No. I
None
10.75°
Nil Unlikely High BTU/gallon/S
Cold-Flow at 10.0
Extreme Temp-
eratures. Smoke (?)
Higher Emissions,
Handling (?),
Ignitabllity, etc.
Good Performance and Handling,
Serious Problem of Future
Availability. Could be
solved by Processing If I_H
Market Demand Justifies. '
*G
4*
C.>,>d Compromise between _^
Performance, Safety, Availa-
bility, Cost, and Distribution
Factors
Interesting "Economy" Fuel--
Essentlally a low-ash resld
Nos. 4,5,6 (Residual)
Ample
Nil
Unlikely High BTU/gallon/$
High Emissions,
Deposits and
Corrosion
BLM/kjk
10/20/69
a Oil .ind Gas Journal of October 6, 1969. "--quotations are realistic spot prices for
refined products moving interstate on Wednesday each week. They will differ from
refiners' prices for branded products. They should not be considered as postings."
This is for 96 octane regular. Chicago (vs. 21.08 at the pump ex tax, 33.74 Including
federal, state and local taxes on national average).
b Same source, Chicago area
c Same source, New York Harbor, No. 6 in barges
'I PI.'I I ' s Oil nr.im. Producers Propane Prices in tank curs, transport truck, or pipeline input,
Ni-u York H.irbor, effective Oct. 1, I'Jd1*.
5-6'
Probably Unacceptable on
r,iformance Basis
-------
TABLE 4. 5. 6
NOMINAL PHYSICAL PROPERTIES ACROSS THE FUELS SPECTRUM
Product Initial
Propane
Gasoline 90
JP-4 Jat Pual (or Type 3} 140
Type A Jat >ue! )30
No. 1 Burnar Fuar (or Dlesal) 140
•*»
i No. 2 Burner Fuai (or Dlaaal) 373
-J
No. 3-CT (Turbine)
No. 4 Residual
No. 5 (light) Residual
No. 5 (heavy) Residua?
No. 6 Residual
Distillation. *F
107. 501 9OT End Point
.44
120 210 330 410
210 290 390 460
370 410 480 520
380 430 490 530
430 510 590 630
..
--
--
..
Gravity
•API
145
62
54
43
42
34.)
19
17
14
10
Specific
0.510
.731
.763
.811
,816
.853
.940
.953
.973
1. 000
Viscosity. v cs
-40'F
1.5
3.0
11
15
40 b
>215C
4.300b
22.000b
--
--
125*F
0.5
0.8
1.2
1.3
2.1
10
25
55
330
Sulfur. 7.
0.001
0.03
0.05
0.07
0.10
0.30
1.0
1.7
1.6
1.8
Ash. 7.
o.i-
--
--
--
<.03
.01
.02
.04
.06
7M
18.2
14
IS
14
14
13
12
12
11
11
BTU/Cal. 1
90.000
123,700
1 2^.000
134.100
13* ,.700
139,100
145.000
148,800
150.000
152.000
154.600
a At 2.5 go/gal Lead
b Minimum, bsx separation In most product occurs above this temperature and
viscosities could be much higher.
c Corresponding to 45 SUS*? 100°F minimum In proposed ASTM specification*.
-------
THERMO ELECTRON
4. 6 BOILER DESIGN
6
The reference cycle boiler heat transfer rate is 1. 58 x 10
Btu/hr. The boiler must have some excess capacity for control
purposes, however, and the maximum boiler heat transfer rate has
6 .
been selected as 1. 70 x 10 Btu/hr. With a boiler overall efficiency
of 82.5%, this rate corresponds to a burning rate requirement of
2.06 x 10 Btu/hr (HHV). The boiler design goals were:
a.. Positive elimination of hot spots on the organic side.
b. Boiler efficiency of at least 82. 5% without air pre-heat
and in compact boiler with low pressure drops.
c. Low material cost and easy construction.
To positively eliminate hot spots, a double-tube boiler construction is
used with water sealed in the annular space between the tubes under its
own vapor pressure. Heat transmission through the narrow (1/16 inch
ave-rage thickness) water jacket occurs by boiling on the outside tube
inner surface and condensing on the inside tube outer surface without
any net circulation of water. Since the water is under its own vapor
pressure, the temperature which the organic sees can never exceed
the saturation temperature corresponding.to the pressure in the water
jacket, thereby positively prohibiting hot spots.
In Figures 4. 6. 1 and 4. 6. 2, cross sectional views through the
combined burner-boiler are presented. In Figure 4. 6. 3, a top view
of the tube bundle is illustrated. With reference to these figures, the
combustion gases at about 3300 °F flow downward through a duct into
the central plenum formed by the boiler tube bundle; a screen distributor
(ceramic-coated stainless steel) is used to provide a uniform flow through
the tube bundle. The combustion gases then flow radially outward through
4-77
-------
THBRMO ELECTRON
three stages of the tube bundle and, are directly exhausted on leaving
the tube bundle.
The organic and combustion gas flow paths are illustrated in
Figure 4. 6.4. The incoming organic first flows through the outer
tube bundle (Stage 1) from which the combustion gases are exhausted;
this provides the lowest organic temperatures in the boiler at the com-
bustion gas outlet and a high boiler efficiency. It is important that an
extremely compact heat exchange surface with high heat transfer rate
per unit volume be used in this stage to maximize the boiler efficiency
with an acceptably low pressure drop on the combustion side. In
Figure 4.6.5, a comparison is given of several compact surfaces with
the ball matrix in terms of the heat transfer rate per unit volume versus
the power requirement per unit volume; the superiority of the ball
matrix on this basis is readily apparent. In addition, the matrix can
be easily fabricated between the round tubes required because of the
High water jacket pressure.
Leaving Stage I, the organic next flows into stage 2 through
which the combustion gases first flow. Because of the high gas tempera-
ture and finned surface on the combustion side, coupled with a high heat
transfer coefficient on the organic side, a very high heat transfer rate
can be obtained in the second stage. The organic then flows through
the superheater coil, or stage 3. This stage is a bare tube coil, since
the controlling thermal resistance is on the organic side and an ex-
tended heat transfer surface is not required on the outside of the tube.
A boiler computer program has been developed for the detailed
heat transfer and pressure drop analysis of the boiler. In Table
4.6. 1, the boiler reference design point characteristics are summarized
4-78
-------
THKUMO •LKCTHOI
CORPORATION
2606-D
Figure 4. 6. 1 Cross Section Through Burner-Boiler, Long Axis.
4-79
-------
TMBUMO «L«CTHOM
CORPORATION
2607-D
23 IN
\ \\\ \\\\\\\\\\\\vc\ \\ \\v \
Figure 4. 6.2 Cross Section Through burner-tioiler, Short Axis.
4-80
-------
TMKMMO •L«CTROI
CORPORATION
2605-D
Figure 4. 6. 3 Top View of Boiler Tube Bundle.
4-81
-------
1-908
ORGANIC FLOW TO ENGINE
COMBUSTION
GAS
FLOW
ORGANIC FLOW
TO BOILER
STAGE NO. I 3 2
Figure 4. 6. 4. Organic Flow Path Through Boiler Tube Bundle
4-82
-------
*>.
GO
—. 10 -
•o
tt
O 0.050 " D SPHERES. 880 FT2/ FT
A 0.125 " D SPHERES, 351 FT2/ FT3
I I I I I III
0.01
1.0 10 10'
(P/A)$td. HORSEPOWER PER CUBIC FOOT
KEY TYPE OF SURFACE CODE NUMBER
X RUFFLED FINS 17.8 - 3/8 R
A IN LINE PIN FINS AP-2
• LOUVERED PLATE FINS 3/8-11.1
o PLAIN PLATE FINS 19.86
° INSIDE CIRCULAR TUBES ST-1
• FINNED FLAT TUBE 9.68-0.87
FT2/ FT3
514
244
367
561
208
305
Figure 4. 6. 5 Comparison of Compact Exchanger Surfaces Illustrating
High Efficiency of Ball Matrix.
-------
TABLE 4. 6. 1
BOILER REFERENCE DESIGN POINT CHARACTERISTICS
Stage
Ball Matrix
Finned
Superheater
Total
Heat Transfer
Rate
Btu/hr
359.000
834,000
383, 000
1, 576,000
Combustion Ga»
Temperature
•F
Entering
1 190
3330
1896
-
Leaving
490
1896
1190
-
Tubing
Length
ft.
26.0
17.0
35.0
78.0
Prenure Do»lgn
Combuttion
Side
In w. c.
2. 06
0. 136
0.288
2.48
Organic
Side
pii
1.96
21.22
23.32
46.5
Mati, Pound*
Outer Tube
Carbon Steel
32.4
21.2
43.6
97.2
Inner Tube
Carbon Steel
9.4
9.6
12.7
31.7
Matrix
Carbon Steel
25.4
-
-
25.4
External Fin*
Copper
-
14. 8
-
14. 8
Water
2. 3
1. 5
3.2
7.0
Total
69.5
47. 1
59. 5
176. 1
*.
oo
Tube Specification*
Inner Tube ID = 0. 930 in.
OD = I. 000 In.
Outer Tube ID = I. 125 in.
CD = 1.315 in.
Material Carbon Steel
Extended Surface Specification*
(1) Matrix
Ball Sice 3/32 in.
Matrix Thickne** 0. 5 in.
Matrix Height 0.935 in.
(Between Tube*)
Ball Material Carbon Steel
(2) Stage 2 External Radial
Fins/inch 10.0
Fin Thicknea* 0.012 in.
Fin Height 0. 356 in.
Fin Material Copper
Stage 2 Internal Longitudinal
Number of Fin» Around
Tube
Fin Thicknea*
Fin Height
Fin Material
16
0. 0312 in.
0. 120 in.
Carbon Steel
vO
-------
THBRMO BLBCTRON
and the physical specifications of the boiler heat transfer surfaces are
given. The three boiler stages have a combined mass of 176 Ibs; to
this value must be added the following part weights to obtain the total
burner-boiler weight:
Tube Bundle (combined three stages) 176 Ibs
Expansion Tanks for Water Jackets 12 Ibs
Thermal Insulation, Top and Bottom,
and Top and Bottom Plates plus
Exhaust Enclosure 15 Ibs
Combustion Chamber 22 Ibs
Air/Fuel Supply plus Ignition and
Flame Detection System 43 Ibs
Fitting and Supporting Hardware
(estimated) 5 Ibs
Total 273 Ibs
In Figures 4 6. 6 and 46. 7, the organic, water and average wall
temperature variations through the boiler are presented for the
reference design point (7377 Ib/hr throughput) and for 4000 Ib/hr
throughput, respectively. In Figure 4.6.8, the organic pressure
input to the boiler for a constant boiler outlet pressure of 500 psia
is presented as a function of throughput; at the reference design
point, the organic pressure drop is 46. 5 psi.
At the reference cycle condition, the boiler overall efficiency,
based on the higher heating value of the fuel, is 82. 7%, equivalent
to a temperature of 490 °F for the exhaust combustion products. In
Figure 4. 6. 9, the variation of the boiler efficiency with organic
throughput is presented. At low flows, the boiler efficiency approaches
86.0%. The boiler efficiency as well as organic pressure drop vary
only slightly with the organic inlet temperature.
4-S5
-------
1-909
690
CPFtovreti'7377 Ibi/lw
Fuel flowroi»'8a.2 Ibt/hr
CP-MPr«uurfS46.40mia
300 -
20 30 40 50
lifting Length from Inlet. Feel
60 TO 80
Figure 4. 6. 6. Design Point Boiler Temperature Profile
4-86
-------
1-910
630
600 -
CPFIo.Rotf4000lbs/hr
Ftif Flow Ron-470 Ibs/hr
CP-
f, •843%(HHV)
300
20 30 40 50
Tubing Length from Intel, Feet
Figure 4. 6. 7. Medium Load Boiler Temperature Profil.
4-87
-------
OD
oo
580
.9 560
o: -
o 540
-------
*>
(T
90
88
X
c
o
T3
(U
S86
DQ
o
.1 84
o
•4—
6 82
00
80
I
Boiler Inlet Temperature
a 250°F
O 286°F
A 300°F
I
0
2468
Organic Flow Rate, 1000 IbsYhr.
10
vO
Figure 4.6.9 Boiler Efficiency (HHV) versus Throughput.
-------
THBHIHO ElKCTKOM
4. 7 CONDENSER DESIGN
The condenser design in this study is based on the Ford radiator
with louvered fins for the following reasons:
a. This fin type is amenable to mass production techniques at
low cost as demonstrated by its use in automotive radiators.
b. The louvered fin heat transfer surface has acceptable pressure
drop and fan power for a given frontal area and heat rejection
rate. It does not necessarily represent the surface with mini-
mum fan power, since manufacturing experience was given
a high weight in selection of the fin type.
c. Heat transfer data on the finned surface were available from
the Ford Motor Company.
In arriving at the design, the approach followed was to use the maximum
frontal area available in the 1969 Ford Fairlane, with some sheet metal
and frame modifications allowable at the front of the engine compartment.
Use of the maximum frontal area minimizes the fan power required and
results in a reasonable condenser configuration.
In Figure 4. 7. 1, the heat transfer and friction factor used in the
design analysis is presented. These curves were derived from data
supplied by the Ford Motor Company.
In Figure 4. 7. 2, the condenser design is illustrated. The con-
denser core measures 50 inches wide by 19.9 inches high by 3 inches
deep; the basic core consists of copper fins, identical to those now used
in the Ford radiator, and flattened carbon steel tubes extending through
the depth of the condenser. The flattened tube has a greater thickness
4-90
-------
.10
.05
.03
.01
.003
I I
I II
I I I I I IT
i
vO
(Jl
I I I I
I I I I l
.30
.50
1.0
5.0
Figure 4. 7. 1 Heat Transfer Coefficient and Friction Factor versus
Reynolds Number, Ford Louvered Radiator.
-------
r
L _
I
•&
ro
i
O
Figure 4. 7.2a Condenser Design.
ft~* I*
-------
2616-D
__
Ll 2.&-I
VIEW 8-B
~
SECTION*-A
Figure 4. 7. 2b Condenser Design.
- s£~f view o
4-93
-------
1-1080
THBRMO • I. • C T R O N
TABLE 4.7. 1
CONDENSER PHYSICAL CHARACTERISTICS
Flattened Tubes
Number of tubes 30
Total Length 129 feet
Total Mass (Carbon Steel) 85. 6 Ibs
Fins
Fins/inch 14
Fin thickness 0. 0025 inch
Fin mass (copper) 25. 5 Ibs
Vapor Header Mass (Carbon Steel) 2. 0 Ibs
Liquid Header Mass (Carbon Steel) 1. 0 Ibs
Mounting Hardware 1. 0 Ibs
Total Condenser Mass (Carbon Steel
plus Copper) 11 5 Ibs
4-94
-------
TABLE 4.7.2
CALCULATED CONDENSER PERFORMANCE
AS A FUNCTION OF AIR FLOW RATE
ORGANIC RATE • 7377 UB/HQ
6RGANIC TEMPERATURE • 210 F
AIR TEMPERATURE . 95
; p- IN
PS I A
27.0
27.0
27.0
27.0
27.0
«ST6P» 0
AjK FI.OW
l_B/*3
aoooo
5
2>K
• C
LENGTH D.rST
VAP
1.0
1*0
1.0
1 .C
1 «C
COND
55.6
61.6
69«7
H7.9
99. C
LIO
*3.»
37.4
29*3
11«1
.0
PAN PW
HP
20. ft*
15.17
10«71
7.17
4.46
o
-------
THBWIHO •i.BCTHON
coorottrioii
than the tubing used in the radiator and has a greater wall thickness
(0.030 inch versus 0. 005 inch). Flow diriders are positioned in the
flattened tube to provide three organic passes through the condenser
and to serve as stays for the flattened tube walls. The thirty flattened
tubes in the condenser are connected to common vapor and liquid cir-
cuits, comprising 30 parallel organic flow circuits.
Some condenser physical characteristics are summarized in
Table 4. 7. 1, which shows that the total condenser mass when con-
structed of carbon steel tubing and copper fins is 115 Ibs.
An alternative which is attractive from both a cost and a weight
point of view is use of an all-aluminum Condenser, using the same
basic design with the following changes:
1. Increase wall thickness of headers.
2. Increase number of dividers in flattened tube.
3. Increase fin thickness from 0. 0025 inch to 0. 005 inch.
With these changes, the heat transfer performance should be approxi-
mately equivalent to that of the reference design and both the mass and
manufacturing cost should be reduced significantly. The mass of the
all-aluminum condenser would be as follows:
Tubes 30. 8 Ibs
Fins 15.5 Ibs
Headers 3. 0 Ibs
Mounting Hardware 1.0 Ib
Total 50.3 Ibs
The Ford Motor Company has indicated the aluminum condenser
should be considerably less expensive than the carbon steel and
copper condenser.
4-96
-------
TNBRMO BLBCTRON
A computer program for calculating the detailed condenser
performance has been completed and incorporated into the generalized
computer program for calculating the overall system performance. The
condenser computer program calculates the heat transfer performance
increment by increment, using the best techniques available for cal-
culating the condensing coefficient and two-phase pressure drop through
the organic side of the condenser. In Table 4,7. 2, a summary computer
printout is illustrated for an inlet pressure of 27. 0 psia and inlet organic
temperature of 230°F. Calculations were performed at different air flow
rates; the table gives a summary of the condenser performance. The fan
power is the shaft fan power, based on use of two Torrington A-2029-5
fans (20-. 0 inches O. D. and 3. 50 inch pitch) . The "X out" is the quality
of the organic effluent from the condenser; a negative quality refers to
subcooled liquid It is planned to use the same heat exchanger as a.
combination desuperheater, condenser, and subcooler. The design
criterion used is that 15% of the organic flow path length be used as
the subcooler. By interpolation from Table 4. 7. 2, this occurs at an
air flow rate of 61, 000 Ibs/hr, equal to 14, 200 CFM of air flow at 95° F
and 1 atmosphere The total heat rejection for this condition is 1. 20 x
106 Btu/hr.
4-97
-------
THBRMO KLBCTMON
4.8 REGJENERATOR
The regenerator design is based on obtaining 90% effectiveness
at the design point, with effectiveness defined as
h - h
E
Reg h. - h '
i sat
where h and h are the actual organic enthalpies in and out of the re-
i o
generator, and h is the* saturation enthalpy of the organic at the
od t
regenerator outlet pressure. Under part load conditions, the regenerator
effectiveness will generally be greater than 90%.
As illustrated in Figure 4. 8. 1, the regenerator design is based on
use of a porous ball matrix extended surface on the vapor side; four
stages are used on the vapor side, since this provides performance
close to a pure counter flow exchanger. Four parallel liquid passes
are used to minimize the liquid side pressure drop. The regenerator
design point characteristics are summarized in Table 4. 8. 1.
A computer program has been written and incorporated into the
generalized system model. This program treats each stage separately
in the calculation for increased accuracy.
The regenerator, as described in Section 5, is packaged on top
of the engine. While not shown in detail, flanges on the exhaust ports
will be cast in the engine block. The regenerator will be mounted
directly on the engine by use of mating flanges on the regenerator
vapor inlet. The vapor then flows upward through the ball matrix.
This arrangement permits the regenerator to function as an effective
oil separator, removing a major fraction of the engine lubricant
blowby back to the crankcase through a drain line. This drain line
4-98
-------
I
sO
M
O^
00
1
O
*
o
0
I
T)
0
X
O
z
U
X
m
a
1
0
r
•
O
-1
a
0
i
Figure 4. 8. la Regenerator Design
-------
TMBHMC •LBCTHOI
T •.. CORPORATION
2617-D
Figure 4. 8. Ib Regenerator Design.
4-100
-------
1-1081
THERMO ELECTRON
C01PO««tlON
TABLE 4. 8. 1
REGENERATOR DESIGN POINT CHARACTERISTICS
Heat Transfer Rate 249, 000 Btu/hr
Effectiveness 90%
Vapor Temperatures
Inlet 348. IT
Outlet 230°F
Liquid Temperatures
Inlet 199. 0°F
Outlet 285. 3'F
Vapor Pressure
Inlet 25. 0 psia
Pressure Drop 2.2 psia
Liquid Pressure
Inlet 546 psia
Pressure Drop 3. 9 psia
Number of Stages 4
Number of Parallel Liquid Passes 4
Tubing (Carbon Steel)
Total Length 133-1/3 ft.
OD 0. 550 inch
ID 0.500 inch
Weight 12.51bs.
4-101
-------
1-1082
THBNMO BLBCTMON
coiroiirion
TABLE 4. 8. 1 (continued)
Matrix (1/4 Copper Balls, 3/4 Aluminum Balls)
Ball Diameter 1/16 inch
Matrix Height Between Tubes 0. 303 inch
Matrix Thickness 0.29 inch
Weight, Carbon Steel Balls 17.9 Ibs
Weight, Copper Balls 6. 0 Ibs
Total Matrix Weight 23. 9 Ibs
Shell
Thickness 1/16 inch
Weight 16.0 Ibs
Total Regenerator Weight (with 1. 6 Ib
allowance for fittings, supports, etc.) 54. 0 Ibs
4-102
-------
THBRMO BLECTRON
CORPORATION
is also used as the crankcase vent line. Use of the regenerator for this
purpose eliminates the requirement for a separate oil separator.
4-103
-------
T H • H M O • i.BCTMOM
CO«fOi»TIO«
is also
5.9 ROTARY SHAFT SEAL AND STATIC SEALS
4. 9. I Rotary Shaft Seal
Slight leakage of the working fluid past the piston rings into
the crank case (blowby) is inevitable. Thus, the crankcase must be
considered part of the system volume within which the working fluid
must be confined. A shaft seal must be provided where the crankshaft
passes through the crankcase wall, since loss of working fluid
and leakage of air into the system are both unacceptable. The vapor
space in the crankcase is vented to the regenerator, making the
crankcase pressure essentially equal to the pressure in the regenerator.
During normal operation with thiophene, this pressure is higher than
atmospheric pressure, and the crankshaft seal must prevent the loss
of working fluid. During shutdown, the condenser pressure corresponds
to the vapor pressure of the working fluid at the prevailing ambient
temperature; since this pressure will generally be less than atmospheric,
the crankshaft seal must prevent the leakage of air into the crankcase.
Since the power plant is shut down for most of its lifetime,
leakage of air into the system represents the most serious difficulty.
Air leakage into the system has two detrimental effects. First, the
presence of oxygen accelerates the thermal decomposition of the organic
material. Also, the noncondensible gases (both the air leaking in and
the gases produced by thermal decomposition) collect in the condenser,
degrading its heat rejection capability and reducing the overall system
performance. Air leakage into the system must therefore be limited
to extremely low levels.
4-104
-------
THKUMO E L. B C T R O N
COi'OliTIOH
Shaft seals which permit only very low fluid leakage in one
direction (i. e. , either into or out of the crankcase) are readily avail-
able. However, where leakage in both directions must be minimized
due to a reversal of the pressure force, the double seal geometry
shown in Figure 4. 9. 1 is used. The pressure of the buffer fluid is
sufficient to ensure that both of the rotating shaft seals function as
unidirectional seals. These unidirectional seals could be either lip
seals or mechanical face seals. The choice between lip seals and face
seals for the individual rotating seals shown in Figure 4. 9. 1 is dictated
by the requirements of the application. The inboard seal is in a thiophene
environment, wherein Viton, the only thiophene compatible elastomer,
swells as much as 30%. A lip aeal is not suitable here, since a long
life, low leakage lip seal is not possible when the elastomer swells
appreciably. Thus, a mechanical face seal is selected for the inboard
seal.
The pressure of the buffer fluid in the seal cavity must be kept
above the crankcase pressure of approximately 25 psia. A buffer
fluid pressure of 30-35 psia is suitable. Therefore, the pressure
.difference across the outboard seal is 15-20 psi. Although lip seals
are capable of pressure differentials of this magnitude, they do not
provide long life at these relatively high pressure differentials. Thus,
the outboard seal should also be a mechanical face seal for good
performance of the system.
Figure 4.9.2 shows the recommended shaft seal design. The
sketch shows the use of a single rotating seal ring which minimizes
the axial length of the seal (approximately 2. 5 inches, as shown in
4-105
-------
CRANKCASE
i
»M
O
M
ft O
y . •••:
(*
I —
HIGH PRESSpRE,
BUFFER FLUID ;
ATMOSPHERE
ROTATING SHAFT SEALS,
Figure 4. 9. 1 Double Shaft Seal Concept*
p
o
I*'
•1
III
• 1
I-
(T
d
cn
in
«V
I"
:i i •
it
> i
il
I '•
O
i
kO
in
-------
STATIONARY SEAL RING
SEAL RING MOUNT
SEAL CASING
ATMOSPHERE
CRANKCASExN
O-RING STATIC SEAL
ROTATING
SEAL
RING
ROTATING SLEEVE
vO
Figure 4. 9. 2 Shaft Seal Design.
-------
THBMMO KLBCTRON
Figure 4.9.2). Of course, there are several alternatives to this basic
design. The retaining sleeve ensures correct axial positioning of the
'"• ' ^-*^"""'"*---_
rotating seal ring by forcing the ring against the shoulder in the shaft.
The five static O-ring seals shown in Figure 4. 9. 2 would use Viton
^^ r
O-rings. The stationary seal ring could be carbon and the rotating
ring could be hardened stainless steel.
The buffer fluid pressure must be maintained at 30-35 psia during
operation ancFat 5" minffnym of 20-25 psia during'shutdown. The simplest
way to do thie" may be to keep the pressure constant at 30-35 psia using
the system shown in Figure 4. 9. 3. The combination of the spring and
bellofram keep the pressure of the buffer fluid constant by allowing ;
enough volume displacement to compensate for leakage past tire shaft
seals. ' - . ~
4.9.2 Static Seals' "T1~— : .
Static seals, which are required jat several locations in the system,
K. _ , • ~
must be vacuum tight. While O-rings could technically be used, ma-
chining of O-ring grooves is too expensive; use of O-rings is thus limited
to die-cast aluminum parts into which the O-ring grooves can be cast
directly. For most of the seals, including those of the engine, a metal
backed and molded Viton seal will be used, as illustrated in Figure
4.9.4. This allows all sealing surfaces to be machined as a flat surface
with significant reduction in manufacturing cost, and still insures a goqd
seal. The seal would be preshaped, exactly as a gasket, for rapid
assembly on the production, line.. In design of the engine, the sealed
joints have been designed so that only simple "gasket" geometries are
required for sealing.
*
Typical of "Gask-O-Seals," manufactured by the Parker Seal Company,
Cleveland, Ohio.
4-108
-------
SPRING
ROLLING DIAPHRAGM
SEAL CAVITY
CRANKSHAFT
Figure 4. 9. 3 Seal Buffer Fluid Accumulator.
-------
1-955
I
••for* Fattening
After Fattening
Figure 4. 9- 4 Static Seal Concept.
4-110
-------
THERMO ELECTRON
4. 10 AUTOMATIC TRANSMISSION
A transmission permitting the engine to idle at zero vehicle speed
is required if the accessories are to be directly driven by the engine.
In addition, a transmission can be used to improve acceleration per-
formance and gradability of a Rankine-cycle propulsion system of a
given size. Two transmission approaches have been evaluated in this
study. In one, a slipping clutch transmission (either single-speed or
two-speed) is used; an overdrive is not required for the approach,
since the clutch size is reasonable at the design engine speed and a
differential with low ratio can be utilized. In the other, a conventional
torque converter is used with approximately 1. 88 overdrive (required
to reduce the converter size to about 12 inches); a forward-reverse-
neutral gear is used after the torque converter. Both approaches are
completely automatic with the driver only required to select forward-
reverse-neutral-park.
4. 10. 1 Slipping Clutch
Dana Corporation of Toledo, Ohio, has prepared a design of both
a single-speed and two-speed automatic clutch transmission, and the
control system for the two-speed transmission has been conceptually
prepared. As discussed in Section 5, the two-speed transmission gives
much better acceleration performance and gradability than the single-
speed and is the preferred approach. The estimated weight of the two-
speed transmission is 135 Ibs. The design is based on existing tech-
nology and is considered state-of-the-art by Dana Corporation. A
clutch transmission is completely locked, except at low speeds, giving
the equivalent of direct coupling, and should, therefore, be more
efficient than a torque-converter transmission.
4-111
-------
T H•M M O B L • C TRON
4. 10. 2:. /Torque Converter
Ihe Ford Motor Company has recommended a 12-1/4 inch con-
ventional torque converter for the Rankine-cycle automotove propulsion
ryatem with approximately 1. 88 overdrive required between the rela-
tively low-speed engine and the converter. The converter must also
be integrated with reverse-forward-neutral-park gearing to form a
complete transmission. A preliminary assembly drawing of the com-
p*ete?transmission has been prepared at Thermo Electron Corporation
acing required parts from a manual transmission supplied by the Ford
Motor Company.
tD"C">; - ~ " - "
' The torque converter transmission has the important advantage
of eliminating high-torque, low speed operation of the engine, thus
alleviating potential bearing difficulties in the engine.
4-112
-------
THBMMO ELECTRON
-4.11 CONTROL AND STARTUP OF SYSTEM
The objective of the control and startup systems selection has
been to find the simplest, least expensive and most reliable approach
which provides adequate control and startup of the system under all
possible types of transients and conditions. The system selected uses
mechanical control elements, with diaphragm actuators operated by
fluid pressures for modulating control. A complete electrical control
»y_stem was also evaluated, but was more expensive, more complex and
probably less reliable than the mechanical system selected. For a
prototype system, an all-electric control system does have the advantage
of permitting rapid changes in the control system to improve system
control or eliminate control problems encountered in the prototype
testing.
4.11.1 Controls for System Operation
In Figure 4. 11.1, a complete flow schematic of the system is
illustrated, including controls. The schematic does not include the
electrical system, startup sequencing controls, ignition system flame
sensor or other safety controls.
Probably the most serious control problem in the system is
control of the boiler outlet organic pressure and temperature within
specified limits regardless of the type of transient encountered by the
system. The approach being following is to maintain the burning rate
and the feedpump rate to the boiler at values corresponding to the
organic vapor flow at any time, that is, to maintain quasi-steady
state operation of the boiler. Since practically instantaneous changes
from zero to full flow and vice versa can occur, the time delay being
the time to depress or release the accelerator pedal, large power
4-113
-------
THBRMO ELECTRON
changes in the system must be sensed at the earliest possible time and
the burning rate and feedpump rate changed to the new values in a time
period of approximately 50 to 100 milliseconds.
The flow rate to the boiler is varied at any engine speed by varying
the effective displacement of the feedpump. The control of the feedpump
is greatly simplified, however, by the existence of an approximate linear
relationship between the mass flow rate through the engine and the intake
ratio of the engine, as is evident from Figure 4. 11.2. The engine mass
flow rate can therefore be expressed as:
^_ = C, (N^ ) (IR)
Eng 2 Eng'
where C = constant
1* = engine RPM
Eng
IR = engine valve intake ratio
The flow rate of organic from the feedpump can be expressed as:
where C = constant
0 = feedpump variable displacement control position.
For equal flow rates and with the feedpump directly driven by the
engine with a 1:1 speed ratio,
C2 (N) (IR) = Cl
"
* = -£ (IR)
4-115
-------
THBRMO ELECTRON
_ Thus,..it may be concluded that the feedpump control position should
be directly proportional to the engine intake ratio irrespective of the
engine-feedpump speed. The engine intake valve and feedpump variable
displacement levers can therefore be connected directly together and
operated as a unit. In the schematic, the engine intake valve and feed-
pump displacement control levers are directly connected to the
accelerator pedal through a governor-controlled bar linkage which
limits the maximum intake ratio as a function of rpm to prevent ex-
ceeding the boiler capacity. The forces required for controlling the
hydraulically actuated engine intake and the feedpump displacement
should be low enough so that connection can be made directly to the
accelerator pedal without power amplification required. Directly
connecting both the feedpump and engine to the accelerator pedal in-
sures instantaneous response of the feedpump rate to changes in engine
power produced by varying the engine intake valving. In Figure 4. 11.3,
a schematic representation of the linkage of the accelerator pedal to the
t
system is illustrated.
A diaphragm actuator is also included on the feedpump control
to provide a vernier control on the feedpump rate in response to the
variable being controlled, the boiler outlet pressure. This actuator
uses a spring loaded diaphragm (or bellows) with the boiler outlet
pressure applied directly to one side of the diaphragm as illustrated
schematically in the flow diagram. Bellow seals will be used in this
vernier actuator to eliminate sliding seals, thereby providing a
"hermetic" actuator. This control will correct for the relatively
small non-linearities between the intake ratio and organic flow rate
through the engine, as well as for imbalances which occur between
the mass organic flew rates in and out of the boiler (due,, for example,
4-116
-------
9,000
^ 7,000
I 6.000
Uj"
K
| 4,000
o
2,000
O.O
0.2
0.4
INTAKE RATIO
0.6
11
>o
in
0.8
Figure 4. 11.2 Mass Flow Rate versus Intake Ratio
for Various Engine rpms.
-------
-------
o
CD
Figure 4. 11. 3 Pedal Actuator with Over-Ride and Governor
-------
THERMO ELECTRON
to momentary loss of pumping due to cavitation) to bring the boiler
outlet pressure back to the control point.
The boiler outlet pressure is adjusted by varying the spring
force on the vernier controller. On startup, the spring provides
full pump displacement until the boiler pressure has increased to
500 psia.
In Figure 4. 11. 1, the complete burner control is illustrated
schematically; a preliminary design of the fuel control valve and
actuator in this control system is illustrated in Figure 4. 11.4.
The burner control uses an orifice in the organic line to the
boiler to detect changes in the organic flow rate to the boiler instan-
taneously. The orifice AP is applied across a diaphragm directly
by the thiophene; the force exertedbythis AP is balanced by the fuel
pressure on the discharge side of the fuel valve (fuel inlet pressure
is maintained constant); the fuel pressure varies with flow rate by
use of an orifice in combination with the fuel nozzle downstream of
the valve. As an example of control valve operation, if the organic
flow rapidly increases from low flow (low power) to a high flow (full
power), a AP increase occurs across the orifice. This provides a
force unbalance on the valve stem, and the valve opens, increasing
the fuel flow rate until the net force on the stem is again balanced.
The speed of response of the valve is increased by the fact that the
. AP increase across the orifice is momentarily larger than the steady-
state AP corresponding to the steady-state flow value. While a detailed
analysis of the control dynamics has not been made, extrapolation from
similar controllers indicates a speed of response from full closed to
full open of the order of 50 - 100 milliseconds. Total travel of the valve
is 0.025 inch from closed to full open.
4-H9
-------
THKRMO ELECTRON
A vernier temperature control similar in concept to the vernier
pressure control- is also used. A separate spring-loaded diaphragm is
used with a pressure proportional to the boiler outlet temperature,
provided by a bulb partially filled with Dowtherm A which is immersed
in the outlet organic from the boiler. The pressure on the diaphragm
is then equal to the vapor pressure of Dowtherm A. In low flow con-
ditions, where speed of response is relatively unimportant, this tem-
perature controller serves as the primary burner control since the
orifice AP at low flow conditions is low.
As illustrated in Figure 4. 11.5, a separate spring-loaded diaphragm
actuated by the fuel1 outlet pressure from the fuel control valve is used to
regulate the air flow so that a constant fuel/air ratio is maintained at
any burning rate. This actuator regulates the position of a damper in
the blower discharge line to the burner.
Both the fuel pump and blower are operated at constant speed by
dc motors. Thus, transient response of the burner control is not
limited by the need for acceleration or deceleration of these components,
but is determined solely by the response of the controller -valve combina-
tions.
Adjustment of the boiler outlet temperature is provided simply by
varying the spring force by means of an adjustment screw. On startup,
the spring also maintains the fuel valve and air damper in the fully open
position for full burning rate until the boiler outlet temperature has
reached 550 "F.
4.11.2 System Startup
The startup sequencing will be designed for the worst possible
condition, thus insuring rapid and positive startup of the system at all
times. Difficulties in startup result from two factors:
4-120
-------
f- ORiriCC COMNCCTIOM&
/7 ORGANIC »CML£.I« FttOUNt
FOR TE.MPC.RA.TUR.E CONTROL.
i
vO
VJl
Figure 4. 11.4 Burner Modulating Fuel Control Valve and
Actuator.
-------
1-1084
Figure 4. 11. 5 Air Flow Actuator and Butterfly
4-122
-------
THBRMO ELECTRON
(1) In tasting at TECO of the piston feedpump for the 3 kwe engine-
generator set under development, it has been found that about 1 psi differ-
ential is required for opening of the suction valving, On startup in an
isothermal sealed system, generation of pressure differentials much
less than 1 psi will result in cavitation in the pump, preventing effective
pumping. The difficulty is magnified under cold-ambient startup, when
the vapor pressure of the thiophene is very low; thiophene has a vapor
pressure of 0. 14 psia at 0°F.
(2) In a sealed Rankine-cycle system under normal operating con-
ditions, the concentration of thiophene in the miscible lubricant is low
because of the high thiophene vapor pressure at the operating lubricant
temperature. When the system is shut down and allowed to cool to an
isothermal temperature, however, the thiophene from all points in the
system will tend to migrate and to dissolve in the lubricant. If the
system is shut down for sufficient time, the entire thiophene inventory
*
in the sealed system can migrate to the lubricant in the crankcaseso
that both lubricant and working fluid are located completely in the engine
crankcase. While this migration can theoretically be controlled by
shutdown valves which block off parts of the system, failure or leak
development in these valves would result in startup failure of the system.
The startup procedure proposed is based on the assumption that the
entire working fluid-lubricant inventory is in the engine crankcase.
With reference to the flow schematic of Figure 4. 11. 1, two
centrifugal pumps are included in the flow schematic, driven by the
same dc motor One is a feedpump boost pump in the organic line
from the condenser to the feedpump section; a standpipe is also included
from the condenser outlet to minimize the fluid volume required to
4-123
-------
TNBRMO KLBCTMON
provide a liquid head to this pump. This pump will provide a 5 psia
pressure rise at the feedpump suction, insuring proper operation of
the feedpump suction valving on startup. In normal operation of the
system, the pumps are not operating and the feedpump boost pump is
designed to pass the full organic flow with negligible pressure drop.
The second pump is provided to circulate the liquid in the crankcase
through a small finned tube heat exchanger, located in the boiler
P* c i £ - ." - . . - -. - -
e"xhaust, and back to the crankcase. The thiophene will be boiled
from the lubricant and vented to the regenerator-condenser side of
the flow system. Condensation of the thiophene will occur in the
0ondenser~; the condensed liquid will be pushed into the standpipe,
providing liquid head to the feedpump boost pump.
A detailed hydrodynamic design of these pumps has been carried
out to insure proper operation of the pumps under the very low head
conditions which will be encountered in the system startup. In Figure
«
4". 11.6, an assembly cross section of the pumps is illustrated. The
pumps are constructed on a common shaft. Since the pressure differ-
ential between the two pumps is low (2 psia or less), and since slight
leakage between the pumps can be tolerated, the filled fluorocarbon
radial and thrust bearing between the pumps is used as the only seal.
To eliminate the rotary shaft seal, magnetic coupling across a non-
magnetic bore seal is used; the permanent magnet sizes illustrated
are sufficient to transmit the relatively low torque required. This
type of coupling is currently used on low cost hermetic centrifugal
pumps for home hot water heating systems and has been tested at
TECO for use on the startup pump for the 3 kwe engine-generator set.
The motor is a constant speed (1335), 12 V dc motor generating 0.071
horsepower with 60% efficiency.
4-124
-------
IZ.25
THIOPMtNC
CONDENSER
LUBRICA-MT
PLUS TM1OPUENC
FROM CRANKCASC
I
1
- IZ V DC SHUWT MOTOR
1335 RPM
.07O8 HORSEPOWER
6OX EFFICIENCY
88 WATTS INPUT
FILLED FLUROCARBON
RADIAL. BEAR ING
CERAMIC MAGNETS
FOR RADIAL GAP
SYNCHRONOUS DRIVE
NON MAGNETIC
BORE SEAU
FEED BOOST
PUMP
FILLED FLUROCARUON
RADIAL AND THRUST BEARING
tRANKCASE CLEAN UP
PUMP
I
sO
ro
7.OO
ESTIMATED WEIGHT £3.3 POUNDS
Figure 4.11.6 Startup Pump Assembly.
-------
THBMMO •LBCTMON
In Figure 4. 11. 7, the design operating characteristics of the
pumps are illustrated. The feedpump boost pump is designed to provide
7. 5 gpm of thiophene flow with 5 psi differential and with a suction head
of 7.2 inches of thiophene without cavitation. The crankcase clean-up
pump is designed to handle 6 gpm of a fluid-composed of 1/3 lubricant
and 2/3 thiophene to 100% lubricant with a pressure differential of
2 paia and with a suction head of 5.4 inches without cavitation. Both
pumps operate at 1335 rpm with the relatively low speed required be-
cause of cavitation.
In Figure 4. 11. 8, a schematic is presented of a high energy
spark generator for ignition, operating off of 12 V dc with a current
draw of approximately 1 amp. A breadboard version of this igniter
was constructed and operation was satisfactory. The unit uses low-
cost standard electronic components and would have overall dimensions
of approximately 2 inches by 2 inches by 3 inches. A high energy
•
igniter is very important in minimizing pollutant emission on startup
since it is desirable to achieve ignition with the first spray of fuel
into the combustion chamber.
In Figure 4. 11. 9, a schematic is presented of a photoresistor
flame sensor to stop fuel flow in case of a flameout. This unit would
be integrated with the spark generator and control relays for automatic
reignition in case of a flameout. Low cost standard components are
used in this unit.
The control and electrical system will be designed for automatic
startup of the system initiated by turning of the ignition switch by the
operator. The startup sequence is summarized below:
4-126
-------
1-958
FEED BOOST PUMP
• 7.5 GPM Thiophene
• 5 PSI Pressure Rise Across Pump
• Pass 15GPM@ Approximately ItoZPSI Pressure
Drop Across Starter Pump
• 40% Efficiency
• 0.0548 Horsepower Input
DETERMINATION OF DESIGN POINT
N = 1.335 RPM
4=20"
D,= L09"
(V 5.2"
t = 0.545"
d= 1.54"
8.4"
1D>h
R=.I5D,
Suction Head Req.
Impeller O.D.(D2)
Impeller Speed —
' Design Point
Suction Head Avail.
_ I
D,,inches
J"iJUre 4. 11. 7a Characteristics of Feedpump Boost Pump and Crank-
case Cleanup Pump Used in Startup of System.
4-127
-------
1-959
CRANKCASE CLEAN UP PUMP
• 6 GPM (From mixture of-5- Lubricant
^to all Lubricant )
• 2 PSI Pressure Rise Across Pump
• 40% Efficiency
• 0.016 Horsepower Input
DETERMINATION OF DESIGN POINT
-5- Thiophene
Impeller O.D.(D2)-_
Suction Head Req.
R=.I5D
0.8 1.2 1.6
D, inches
Figure 4. 11. 7b Characteristic* of Feedpump Boost Pump and Crank-
case Cleanup Pump Used in Startup of System.
4-128
-------
CM
-O
12 Vdc
1 A Nom
X Spark L
5000T
o
00
Ul
o
o
'Electrodes
Figure 4. 11. 8 Igniter - Oscillating Frequency = — 30 kc
-------
+
o-
t
Fuse - A
12 V dc
OJ
o
PhotoceU
O-
180
II '
*s
2W9104J
'i
IN4736
, cnm
2N1613
• •• ,0
o
o
15K
r
50 O
6 V dc
Figure 4. 11.9 Flame Sensor.
-------
THBRMO ELECTRON
OIPOIATIOI
a. On turning the ignition switch, the igniter, fuel pump com-
bustion air blower, and startup pumps are turned on.
b. After a short delay (a few seconds) to allow the fuel pump
and combustion air blowers to reach operating speed, the fuel shutoff
solenoid valve will be opened. If ignition is not achieved in about 5
seconds, the flame sensor will close the solenoid fuel shutoff valve,
stopping fuel flow.
c. Approximately 20 seconds after turning the ignition switch
on, the engine startup motor will be switched on, rotating the engine,
feedpump, hydraulic valve actuating pump and oil lubricating pump.
An engine startup speed of about 200 rpm would be used.
d. As boiler pressure builds up, the engine will start, bringing
the rpm up to'idle speed of 300 rpm.
e. A pressure switch on the boiler will turn off the igniter and
startup pumps when boiler pressure reaches approximately 400 psia.
This switch will also turn on a green light on the instrument panel,
indicating the automobile is ready for operation. On shutdown, turning
the ignition switch off will close the solenoid shutoff valve and stop the
fuel pump and combustion air blower. The engine will continue to
rotate at idle speed until the boiler outlet pressure and temperature
decrease as the boiler cools to a level where the accessory and frictional
loads stall the engine.
4.11.3 Safety Controls
The following safety and malfunction controls are proposed in the
system:
4-131
-------
Engine Low Oil Pressure
Condenser High Pressure
Boiler Water Pressure
a. High side
b. Low side
Boiler High Organic Pressure
Boiler High Organic Outlet
Temperature
'Snap-switch operates indicator
light for low oil pressure.
Snap-switch breaks ignition
circuit, shutting system down.
Rupture disc prevents rupture
of condenser.
Snap-switch breaks ignition cir-
cuit, shutting system down.
Rupture disc prevents tube
rupture.
Rupture disc prevents tube
rupture.
Relief valve vents liquid from
feedpump outlet to condenser.
Snap-switch breaks ignition
circuit, shutting system down.
Snap-switch on Dowtherm A for
fuel control valve breaks ignition
circuit, shutting system down.
4-132
-------
THBRMO BLBCTROM
COIPOIAIIOII
REFERENCES
1. Sax, N. I. , Dangerous Properties of Industrial Materials, Reinhold
Publishing Corporation, New York, 1957.
2. Personal Communication, Dr. D. R. Miller, Monsanto Company,
St. Louis, Missouri, February 9, 1970.
3. Steere, N. V. , Handbook of Laboratory Safety, The Chemical
Rubber Company, Cleveland, Ohio.
4. Flury, F. .. and Zernick, F. , "Toxicity of Thiophene," Chem, Ftg.
56. 149 (1932).
5. Christomanos, A,, "Action of Organic Sulfur Compounds on the
Dog Organism: Action and Fate of Thiophene in the Metabolism
of the Dog," Biochemistry Z. 229, 248 (1930).
6. Christomanos, A., "Experimental Production of Cerebellar
Symptoms by Thiophene," Klin. Wo-chschr. 9_, 2354 (1930).
7. Myers, P. S., "Automobile Emissions - A Study in Environmental
Benefits versus Technological Costs," SAE paper No. 700182, 1969.
8. Personal Communication, April 1970, Mr. Curtis Burkland,
Marquardt Corporation, Van Nuys, California.
9. Vickers P. T., et al. , "The Design Features of the CM SE-101 -
A Vapor-Cycle Powerplant," SAE Paper 700163, January 12-16,
1970.
10. Personal Communication, January, 1970. Dr. B. L, Mi>ckel,
American Oil Co., Hammond, Indiana.
4-133
-------
THKRMO EL.BCTRON
5. SYSTEM DESIGN AND EVALUATION
5. l'' INTRODUCTION
A number of alternatives exist with respect to integration of the
basic components described in the previous section into a complete
system packaged in an automobile. Among the more important choices
are:
a. Method of Driving Accessories at Zero Vehicle Speed
In an automotive system, provision;must be included for driving
the Rankine-cycle accessories as well as the automotive-accessories,
such as power steering, at zero vehicle speed. In one approach, a
relatively simple transmission is used, permitting the main propulsion
engine to idle at zero vehicle speed so that all accessories can be driven
directly by the main propulsion engine. In the second approach, the
main propulsion engine is directly coupled to the drive shaft (through a
forward-neutral-reverse gear), with an auxiliary engine running at
constant speed used to drive all accessories.
b. Type of Engine Intake Valving
The engine intake valving can be constant intake ratio with a
throttle valve for power control or variable intake ratio with power
control obtained by adjustment of the engine intake ratio.
c. Packaging of System
Since the Rankine-cycle system consists of several relatively
independent components, greater flexibility exists in packaging the
system than with an internal combustion engine. For example, the
5-1
-------
TM
•ngine could be integrated wiQCthel rear axle, .withihelboiler and the
condenser in the normal engine compartment at the front of the car.
With respect to driving accessories, use of the main propulsion
•ngine with a simple transmission, permitting'the engine to idle at
zero vehicle speed,'has been selected as the" optimum cKbicel "If an
accessory engine is used, its size must be sufficient to handle any
summation of peak loads by the accessories. Considering all possible
acces«oritefrr.~Mri engDtB-power~of approximately 20'hp would be required,
even though the average load would be considerably less. The engine
would therefore be operating at low throttle pressure most of the time,
with a poor thermodynamic efficiency. Use of the main propulsion
«uci- z.f-i --•>- -. ------L— ^. A=T_ :c~JL^ i-^.odi i:v-*-i-r «-E:--r^-~- i-
engine permits handling of these peaks so that the additional power
Tt-X...**.. z... .;. .-c. z**. :±.. .r*V.:=._. - i r /. .~-^.\" c.".- -'.£..: trcr _..5;:
required to handle accessories can be based on the average load rather
than the peak load. In addition, the power required to drive the
.accessories will be generated more efficiently by the main propulsion
engine. The factors of importance, in addition to efficiency, are cost,
packaging, and simplicity. The cost increment represents a tradeoff
between the cost of the auxiliary engine-throttle valve control plus
smaller accessory components vs. the cost of the transmission plus
larger accessory components. This tradeoff is difficult to establish
without a detailed design and cost study of all of the components involved,
although it would appear qualitatively that the transmission approach
cost should be considerably less than tV che auxiliary engine approach
Addition of the 20 hp engine provides a difficult packaging problem if the
complete aytrm ia to b- installed in the engine compartment of a con-
ventional automobile. The required power can be obtained from the main
propulsion engine, Tvith negligible increase in size, and the transrrr is ion
5-2
-------
TMBRMO ELECTRON
«.an be packaged exactly as in current 1C powered automobiles. The
factor of simplicity again represents a tradeoff between the engine-
throttle valve control and the transmission, and it is difficult to
establish which represents the simpler or more reliable approach.
A computer program has been written for a detailed analysis
of the system, using models developed for the various components.
Calculations have been used to prepare performance maps for
systems with three different types of intake valve operation. In
Figures 5. 1. 1 and 5. 1.2, performance maps are given for a variable
intake ratio engine with (IR) = 0. 8, selected as the maximum
max
practical intake ratio, and for (IR) = 0.29. The maximum power
r max r
curve (wide open throttle) on these plots is established by increasing
the engine intake ratio at each rpm until the boiler design capacity
ia reached. In Figure 5. 1.3, the maximum intake ratio and maximum
organic flow rate at which this boiler capacity is reached are illus-
trated for the two cases. The maximum power curve of Figure 5. 1. 1
with (IR) = 0. 8 represents the highest performance for a given
max
engine displacement and boiler capacity. In Figure 5. 1.4, the
maximum torque curves and maximum power curves are presented.
As illustrated in Figures 5. 1. 1 and 5. 1.2, the maximum
engine intake ratio can be decreased from 0. 8 to 0. 29 with a
relatively small decrease in performance, since intake ratios
higher than 0.29 can be used t>nly in the range of 300 - 800 rpm,
assuming an idle speed of 300 rpm. Use of an intake ratio of
0.29 results in a significant reduction in the feedpump displacement
and condenser load at low engine speeds. From Figure 5. 1. 5, which
presents the performance map for a system with throttle valve control
5-3
-------
THERMO ELECTRON
.a ^constant, engine intake ratio of 0. 137, .it is apparent that use of
-""•'•. *- r — - - • . 11 - 3 = 7 Y. c ; :' ;
-. .-.-a- constant intake ratio with throttle valve control results in a sig-
._ nificant decrea'se in both the efficiency and the performance of the
system.
From these performance maps, it is evident that a strong
incentive exists for development of an engine with variable intake
c.. .. valving. The choice of (IR) = 0. 8 or (IR) = 0. 29 is difficult to
£«-- = --- ? --- • max •_.-•: max _____
j-.c.~as~sess quantitatively since it represents both a'cost and a performance
tradeoff. In Figure 5. 1.6, a comparison is presented of the road load
plus grade load for a 1969 Ford Fair lane with the system maximum
_,,. power output .for the cases with (IR) = 0. 8 and (IR) = 0. 29,
-cu- . r ..__-.._£. __._-....- •_. max ••?. - -<• - • max
^.^ grespectiyely. . .Top speed of the vehicle is seen to be about 100 mph on
a level grade. The gradability of the system with (IR) = 0. 29 is
max
, __ ..^limited to about a 20% grade assuming a direct drive system. Use of a
__ ,. two- speed transmission can give gradability and performance equivalent
, .. to. or exceeding that of the system with (IR) =0. 8 when coupled directly.
It is recommended that the development be concentrated on the
(IR) = 0.29 valvine with two- speed transmission for the following
max
reasons:
(a) Performance and gradability equivalent to or better than that
of the (IR) = 0. 8 system with one-speed transmission can
be obtained by use of a two-speed transmission with
(IR) = 0. 29.
max
(b) The condensing load in full-throttle acceleration is easier
to meet with a fixed condenser size due to the higher con-
denser fan rpm at lower vehicle speeds.
5-4
-------
(c) Hydrodynamic journal bearing performance in the engine
should be more acceptable due to the shorter time for
application of full cylinder pressure coupled with the
higher engine speed at low vehicle speeds.
(d) The feedpump size is smaller, the required displacement
being less by a factor of 2. 66.
It should also be noted that engine valving developed for (IR) = 0. 29
max
can be easily extended to (IR) = 0. 8 since the primary valving
7 max
difficulties result from the shorter intake ratios.
5-5
-------
m
i
110
100
80
60
O
C 40
o
CO
20
I
in
in
•o
0 2OO 4OO 6OO 800 1000 I20O I4OO (600 I80O 2OOO 22OO
Engine RPM
Figure 5. 1. 1 Performance Map with 184 CID Engine Maximum
Intake Ratio of 0. *.
-------
no
too
80
g. 60
4>
O
I
C 40
o
20
0 ZOO 400
GOO 80O 1000 I2OO I4OO I60O I8OO 2OOO 22OO
Engine RPM
Figure 5. 1.2 Performance Map with 184 CID Engine and Maximum
Intake Ratio of 0. ?.Q.
-------
ut
00
€>
*
o
*
in
i.
O
X
.c
to
no
too
80
60
40
20
T
Full Power, Variable IR ^**
of O.BMax. ^ *•*
200 400 600 80O
1000 1200 I4OO
Engine RPM
1600 1800 2OOO 220O
Figure 5. 1.3. Performance Map with 104 CID Engine and Constant
Intake Ratio of 0. 137 (Throttle Valve Control)
-------
Ul
vO
10,000
- 8000
— 6000 £
0.4
0.2
- 4OOO
— 2000
2OO 400 6OO 800
1000 1200 1400 I6OO I8OO 2000 2200
Engine RPM
Figure 5. 1.4 Maximum Intake Ratio and Maximum Organic Flow Rate
as Functions of Engine Speed.
-------
Ul
i
900
800 -
700 —
600 —
o- 500
o
JC
400
300
200
100
Maximum HP fouri
-------
ui
i
tio
100 -
I ' I ' I '
C10% I* 13%
GnMlt I Grodt
Vfhiclt Load
System Power Output
10 -,
0
0
10 20 30 40 50 60 70
Vehicle Speed,MPH
80
90
100 110
Figure 5. 1. 6 Comparison of Road Load Plus Grade Load for
Ford Fairlane with System Maximum Power Output.
-------
THHMMO KLBCTRON
5. 2 PERFORMANCE IN REFERENCE AUTOMOBILE
A 1969 4-door Ford Fairlane has been selected as the reference
automobile for this study. Performance characteristics of the Rankine-
cycle system are presented in this section for a torque converter trans-
mission and single-speed clutch transmission with IR =0.8 and for
e r max
single and two-speed clutch transmissions with IR = 0.29.
max
Figures 5. 2. 1 through 5. 2.4 and Tables 5. 2. 1 through 5. 2. 5
present the performance and fuel economy calculations prepared by
the Ford Motor Company and the Dana Corporation, using existing
computer programs, for the Rankine-cycle system with different types
of transmissions and for the Ford production 302 - 2V engine with three
speed transmission. The basic input to these calculations was the per-
formance maps presented in Figures 5. 1. 1 and 5. 1.2. Fuel economy
calculations have been completed only for the Rankine-cycle system
with IR = 0. 8.
max
The most important conclusions from these calculations are:
1. The Rankine-cycle system with 184 CID engine should be
capable of providing 0-60 mph acceleration times of less
than 15.0 seconds, taken as the criterion for acceptable
performance. The acceleration performance is fairly
dependent on the type of transmission used. A maximum
level grade vehicle speed of 95 - 100 mph should be attainable,
irrespective of type of transmission used.
2. The Rankine-cycle system with two speed clutch transmission
and IR = 0.29 provides a close approximation to the
max
tractive effort delivered by the 302- 2V internal combustion
engine with three speed automatic transmission.
5-12
-------
THKRMO ELECTRON
OOIFORATIOI
3. The Rankine-cycle system with two-speed clutch transmission
and IR = 0. 29 provides a gradability of 49% with 54 ft-lbs
max
of torque subtracted for driving accessories.
4. To obtain performance with the Rankine-cycle system, with
IR = 0. 8 and single-speed clutch transmission, equiva-
max or i-i
lent to that of the 302 - 2V internal combustion powered system
would require an increase in the size of the Rankine-cycle
system of about 20% (engine displacement = 220 CID, with
appropriate increases in boiler, condenser, feedpump, and
other accessory sizes).
5. The customer average fuel economy for the Rankine-cycle
system with (IR) = 0. 8 and single-speed clutch trans-
max " r
mission is 24% less than the 302 - 2V internal combustion
system (15. 7 mpg vs 12. 7 mpg).
6. The suburban fuel economy for the Rankine-cycle system with
(IR) = 0. 8 and single-speed clutch transmission is 14%
max
less than the 302 - 2V internal combustion system (18. 0 mpg
vs 15. 8 mpg).
7. The city fuel economy for the Rankine-cycle system with
(IR) = 0. 8 and single-speed clutch transmission is 39%
max
less than the 302 - 2V internal combustion engine. It is
expected that this comparison will be improved by use of
the Rankine-cycle system with IR = 0.29 and two-speed
-- max
clutch transmission, since this eliminates the region of lowest
system efficiency in the low speed range and moves the overall
system operation into a more efficient region for the Rankine-
cycle system at low vehicle velocities (see performance maps,
Figures 5.1.1 and 5. 1. 2).
5-13
-------
THBRMO BLBCTRON
(8) The steady speed fuel economy for the Rankine-cycle system
with IR = 0. 8 and single-speed clutch transmission is
max
higher than that of the 302 - 2V internal combustion system
up to 50 mph and is lower above this speed.
5-14
-------
1-967
1800
1600
1400
1200
1000
LJ
0)
£
E 800
600
400
200
TWO SPEED CLUTCH
2.5/1 Ratio to 1800Engine RPM Downshift
I/ I After Downshift
2.0/1 Ratio to 1800 Engine RPM Downshift
I/I After Downshift
1
Single Speed Clutch
l
I
0
20 40 60 80 100
Vehicle Speed, MPH
120
140
Figure 5. 2. 1 Comparison of Tractive Effort for Single and Two
Speed Clutch Transmission.
5-15
-------
1-968
is:
KC:
2200
2000
1800
1600
1400
1200
bJ
5
o 1000
600
600
400
200
-.Ford 1969 Fairlane, 4 dr., 302-2V
Engine, Three Speed Transmission
TWO SPEED CLUTCH
(2. 5/1 Ratio to 1800 Engine rpm
'( downshift, 1/1 after downshift.
I ' I "
Two Speed
Clutch, 2. 5/1
Ratio to 1800 Engine
rpm downshift, 1/1
after downshift.
3420 Eng RPM
I
1
20 40 60 80 100 120 I4O
Vehiclt Speed, MPH
Figure 5. 2. 2 Comparisons of Tractive Effort for 1969 Ford Fairlane
Powered by 302-2V Engine and by Rankine-Cycle System
with 184 CID Engine, IRmax = 0. 29, with Dana Two Speed
Clutch Transmission.
5-16
-------
1-1087
2200
2000
1800
1600
1400
1200
*<
a
o 1000
u
rt
H 800
600
400
200
1969 Fairlane, 4 Dr., 302-2V Engine
Speed Transmission
r2.0/l Ratio to 1800 Engine Rpm
1/1 After Downshift
I
I
20 40 60 80 100
Vehicle Speed, Mph
120
140
Figure 5. 2. 3 Comparison of Tractive Effort for 1969 Ford
Fairlane powered by 302-2V Engine and by
Rankine-cycle system with 184 CID Engine,
IRmax = 0. 29, with Dana two-speed clutch
transmission.
5-17
-------
1-969
h ',
20??OQ
800
600
400
200
• FORD l969Foirlone,4Dr,302-2V Engine
Three Speed Transmission
Maximum Torque Capacityof System
' Rankine Cycle, 184 CID, I R^O.8
Single Speed Clutch Transmission
RanKine Cycle, l84CID,IRmfl^0.8
Torque Converter Transmission
(12 feD Converter)
Low Speed Torque
Used in Performance
Calculations
3420 Eng.RPM
2000 Eng. RPM
I
I
20 40 60 80 100
Vehicle Speed.MPH
120
140
Figure 5.2.4 Comparison of Tractive Effort for 1969 Ford Fairlane
Powered by 302-2V Engine and by Rankine-Cycle System
with 184 CID Engine, IR.max = °- 8- with Single Speed
Clutch Transmission or with Torque Converter Transmission.
5-18
-------
Ol
H*
vD
TABLE 5.2. 1
SUMMARY OF RANKINE-CYCLE ENGINE
PERFORMANCE AND ECONOMY PROJECTIONS
FROM COMPUTER PROGRAMS PB1111 AND PB1213
(IR) =0. 8
max
VEHICLE:
1969 Fairlane
4 -Door Sedan
ENGINE AND DRIVE
302- 2V Production Engine
Automatic Transmission
250- IV Production Engine
Automatic Transmission
182-CID TECO Engine,
Single Speed
Clutch Drive
220- CID TECO Engine
Single Speed
Clutch Drive
182-CID TECO Engine
Speed-up 1,88, Converter
PERFORMANCE PB1 1 1 1
'0-4 Sec.
Ft.
90
78
70
86
74
0-10 Sec.
Ft.
469
406
408
480
412
0-60 mph
Sec.
11.9
15.6
14.2
11.4
14.4
Passing
at 50 mph
Sec.
9.7
11.6
10.3
9.2
11.0
ECONOMY PB1213
City
mpg
13.3
12.9
9.6
-
10. 1
Suburban
mpg
18.0
19.6
15. .8
-
15.0
Customer
Average
15:7
16.3
12. 7
-
12.5
70 mph
mpg
16.2
17. 59
1
13 4
-
13.0
8
Z
o
r
N
0
H
JO
0
z
r\>
-------
n.A H:M i rv Lfr I > ilMft, ;
• \ i ii < < <
i : • A\ (IR) "•• «b
max ;
rn ,,,
.
Performance - Computer Program PB1111
0-4 Sec.
ft.
I i I.
89.9
78.2
70.0
77.2
86.0
94. 4'
74.2
79.8
0 - 10 Set.
i **• > •
1 .' •'
:469.2
405.7
407.9
442.0
'
480.4
516.7
412.2
414.6
0-60 mph
, sec.
1
1
11.9
15.6
14.2
12.7
11.4
10.3
14.4
14.3
1
0 - 1/4 Mile
: SBC.
1 .
18.8
20. 6
i •
20.3
19.5
18.7
18.0
20. 1
20. 1
!
. !
i
•
Passing a!t
50 mph;
1 1, i"- '
,,
9.'
11.!
1 1
[••
2
1
t
9 •
10.30
9.74
9.22
8.79
10.97
10.85
IM
o
-------
TABLE 5. 2. 3
VEHICLE ECONOMY PROJECTIONS
THERMO ELECTRON CORPORATION
RANKINE CYCLE ENGINE
VEHICLE: 1969 Fairlane 4-Door Sedan. Wheelbaee - 116 in.
TIRES: 7.75 x 14, Rolling Radius - 1.08ft.. Rev/Mile - 778
(IR) = 0. 8
max
Ford Production Engines
302-2 V 3-Speed Automatic
250- IV 3 -Speed Automatic
Rankine Cycle Engines
182-CID Clutch Drive, Single Speed
182-CID1. 88 Ratio Speed -Up Gear
182-CID 2. 77 Ratio Speed-Up Gear
Idling
Speed
rpm
500
600
300
300
300
Fuel Flow
Ibs/hr
3.75
2.68
2.00
2.00
2.00
Fuel Economy Computer Program PB1213
City
mpg
13.3
12.9
9.6
10.1
10.3
Suburban
mpg
18.0
19.6
15.8
15.0
15.2
Customer
Average
mpg
15.7
16.3
12.7
12.5
12.8
Steady Speed mph
30
mpg
27.3
27.0
33. 1
32.4
32.3
40
mpg
23.3
25.4
25. 1
24.6
24.8
50
mpg
20.4
22.9
20.4
19.7
19.5
60
mpg
18.0
20. 1
16.7
16.0
15.7
70
mpg
16.2
17.6
13.4
13.0
12.9
30-70 mph
Average
mpg
21.0
22.6
21.7
21.2
21.0
-------
TH
1-1086
TABLE 5.2,4
VEHICLE PERFORMANCE PROJECTIONS
IR =0.29
max
Full Torque Curve Used (No Allowance for Accessories)
Transmission
Single Speed
Two Speed
Two Speed
Two Speed
Two Speed
Two Speed
Gear
Ratios
1/1
2.5/1
1/1
2.25/1
1/1
2. 0/1
1/1
1.75/1
1/1
1.50/1
1/1
Engine Shift
Speed
-
1800
1800
1800
1800
1800
0 - 60 mph
Acceleration Time
Seconds
17.1
12. 5
12.. 4
li. 5
12. 8
13.2
Gradability
19.1%
56: 3%
49. 1%
42.4%
36 1%
30. 2%
5-22
-------
THBRIMO ELECTRON
1-966
TABLE 5.2.5
VEHICLE PERFORMANCE PROJECTIONS
IR =0.29
max
54 ft # Subtracted from Full Torque Curve at All Speeds for Accessories
•Transmission
Single Speed
Two Speed
Two Speed
Two Speed
Two Speed
Two Speed
Gear
Ratio
1/1
2.5/1
1/1
2.5/1
1/1
2. 00/1
1/1
1.75/1
1/1
1.50/1
1/1
Engine_ Shift
Speed
-
1800
1800
1800
1800
1800
0-60 mph
Acceleration Time
Seconds
19.6
14.4
14.5
14.6
14.9
15.5
Gradability
17.0%
48. 9%
42.5%
.37.2%
31.9%
26.7%
5-23
-------
THKWMO BLBCTRON
coiroitTioi
5. 3 PACKAGING OF SYSTEM AND SYSTEM WEIGHT
For the present study, the criterion used in packaging the
system has been that the entire system should be packaged in the
engine compartment of the"1969 Ford Fairlane with only minor
sheet metal and frame changes required for mounting of the Rankine-
cycle components and with rear wheel drive. This approach was
followed for the following reasons:
a. It permits use of the current production automobile
chassis, with only minor changes, for constructing
the first prototype.
b. The optimum system appears to be one in which the
components are packaged in close proximity to each
other, eliminating long fluid or vapor lines between
separated components and permitting the engine to
directly drive high power accessories such as the
condenser fans and feedpump. Use of a regenerator
increases the number of fluid and vapor runs required
when components are separated.
While it appears desirable to package the complete Rankine-cycle
system in one location in the vehicle, alternatives do exist for
incorporating the system into an automobile which have not been
investigated in this study. As an example, the fact that the
Rankine-cycle system is composed of several components, each
of which is relatively small compared to the equivalent internal
combustion engine-transmission system, would make it particularly
attractive to incorporate and package a front-wheel drive system for
an automobile.
5-24
-------
THRRMO ELECTRON
In Figure 5. 3. 1 the packaging of the major components for the
100 hpRankine-cycle power plant in a 1969 Ford Fairlane engine
compartment is illustrated. The engine-transmission assembly is
installed in exactly the Same relative position as the current 1C engines
in production in rear-wheel-drive automobiles. The much shorter
engine provides room for the boiler and condenser to be placed in front
of the engine as shown. The condenser is placed in the front of the car
in the same relative position as the radiator of an 1C car, facilitating
movement of air through the condenser. The condenser is wider than
the radiator for an 1C engine, however, and some modification of the
fender skirts will be required for incorporation of the condenser. The
regenerator is positioned directly above the engine, permitting a direct
run of the engine exhaust vapor into the regenerator. The regenerator
is mounted directly to the engine by means of flanges cast into the
engine block on the exhaust vapor lines and also serves as the oil
separator for the system.
In the photographs of Figure 5. 3. 2, a complete mockup of the
100 horsepower system installed in tne engine compartment of a 1969
Ford Fairlane chassis is illustrated. The major components, as
described in the component designs of Section 4, fit without difficulty,
with room remaining for convenient placement of the power system and
vehicle accessory components (including air conditioning). All power-
driven accessories are included at the front with the condenser fans so
that they may be driven by a single-power driveshaft. The accessory
power drive must come from the rear of the engine, since only a
single rotary shaft seal is desired. In the mockup, a flexible shaft
is used for this purpose; an alternative is to use a solid shaft drive
with universal joints.
5-25
-------
THBMMO BLBCTRON
The 1969 Ford Fairlane used a coil spring front suspension which
Ir. J,-_r
restricts apace in the engine compartment. Additional room, if required
IG~ c--. :.i .
at a later date, could be obtained by use of a torsion bar suspension.
«.•_'•.:-. . .
However, for the 100 hp system, this modification is not necessary.
i_ns ;z- - - ..
In Table 5.3.1, a tabulation of the total system weight is presented
and compared with that of the 302-ZV internal combustion system with
3-speed transmission.
. cycle system is estimated to weigh about 150 Ibs
more than the internal combustion engine system. This additional
weight may require use of a heavy duty suspension system in the front
fST'dptirnum vehicle riding quality.
In Table 5.3.2, the system thiophene inventory is given. The
total system inventory is 31 Ibs. The lubricant inventory is estimated
to be 4 quarts.
5-26
-------
1-970
•to 30 no 10
-50
Figure 5.3. 1 Position of Major Components for 100 hp Rankine-Cycle
Power Plant in a 1969 Ford Fairlane Engine Compartment.
5-27
-------
TABLE 5.3. 1
TABULATION OF TOTAL SYSTEM WEIGHT
AND COMPARISON WITH 3Q2-2V INTERNAL COMBUSTION
SYSTEM WITH 3-SPEED TRANSMISSION
v*
i
Engine Expander Assembly
Feedpump
Engine Subsystem
Transmission
Burner- Boiler
Regenerator
Condenser
Radiator with fan, connectors, and water
Controls, Exhaust, Electrical System,
Accessory Drives, and other
Miscellaneous Components
Working Fluid and Lubricant
Total
Rankine
Reference
Design
220
45
265 Ibs
135 Ibs
273 Ibs
54 Ibs
115 Ibs
75 Ibs
40 Ibs
957 Ibs
302 Cu. In. V-8
with 3 -Speed
Automatic
479 Ibs
159 Ibs
54 Ibs
114 Ibs
806 Ibs
H
Z
n
a
2
o
n
r
n
o
H
a
o
z
-------
THBRMO BLBCTItOM
r-~lT)88
j
£ £ TABLE 5.3.2
THIOPHENE INVENTORY IN SYSTEM
I — fc - —
Boiler
Regenerator
Condenser
Lines
Engine and Feedpump
Total
11.2 Ibs
10.6
1.2.
1.7
0. 1
30. 8 Ibs
5-30
-------
THBMMO BLKCTRON
5.4 EMISSION LEVEL FROM THE SYSTEM
Of great importance are the emission levels in gms/mile which
can be expected from a Rankine-cycle propulsion system. The
emission levels obtained with the burner developed at TECO for a
5 hp Rankine-cycle system have been used in making these projections.
The control system for the burner will be designed to maintain optimum
fuel/air ratio at any burning rate to minimize pollutant emissions. It
is believed that tha emission levels obtained from the TECO burner
can be attained, with development, in the full-size burner for the
automotive propulsion system.
In Table 5.4. 1, projections are made of the emission levels in
grams/mile from the Rankine-cycle propulsion system, using the
measured emission data reported in Section 4. 5. The projections are
based on a fuel economy of 10.0 mpg. The projected customer aver-
age fuel economy for the Rankine-cycle system is 12. 7 mpg, leaving
a 2. 7 mpg difference to account for driving accessories such as air
conditioning, power steering, and power brakes, as well as a safety
factor to account for any uncertainties in projected system fuel
economy. The average emission levels used in the emission projections
are:
Excess Air 33%
Unburned Hydrocarbons 15 ppm
Carbon Monoxide 60 ppm
Nitrogen Oxide 40 ppm
It is apparent that the Rankine-cycle system has the potential for
exceeding the projected 1980 Federal Standards in gms/mile, as given
in Table 5.4. 1, being a factor of 5 lower in UHC, a factor of 13.4
lower in CO, and a factor of 1.6 lower in NO.
5-31
-------
TABLE 5.4. 1
PROJECTION OF MEASURED EMISSION DATA TO
TECO RANKINE CYCLE PROPULSION SYSTEM
Excess Air = 33%
mpg - 10
i
UJ
Pollutant
UHC
CO
NO
Projected Emission Levels
for TECO Rankine Cycle
Propulsion System
ppm. Exhaust Gas
15
60
40
gms/mile
0.050
0.35
0.25
Projected Federal Standards
gms/mile
1975
0.5
11.0
0.9
1980
0. 25
4.70
0.40
Uncontrolled
1C Engine
(1967)
11. 5
85
4
IIEC
Goals for
1C Engine
0.82
7. 1
0.68
I
vO
-------
THKMIMO B&.BCTRON
toiroiATio
5. 5 RELATIVE COST COMPARISON WITH 302-2V FORD ENGINE
.Work on this task has been initiated with the Ford Motor
Company, but is not available for this report. The cost information
will be available in a later supplement to this report.
5-33
-------
-THKRMO KI.KCTRON
=5,6 ^GENERALIZED COMPUTER MODEL
VA generalized computer model for steady-state performance
-^calculations has been completed, providing refinement of the per-
formance calculations presented earlier. Any performance changes
from use of this more detailed model will be second-order effects
and should not affect the conclusions with respect to system per-
formance. Typical printouts are illustrated in Tables 5. 6. 1, 5. 6. 2,
and 5. 6. 3. This program uses detailed models describing the per-
formance of all major components including engine, boiler, feedpump,
regenerator, condenser, and condenser fans; pressure drops in con-
necting lines are also calculated. In the calculation, condenser
pressure is not permitted to fall below approximately 10 psia, and
the position in the condenser where 100% condensation is reached is
controlled at a constant point in the condenser to insure constant fluid
inventory in the system.
In Tables 5.6.1, 5.6.2, and 5.6.3, three runs from this program
are presented. Comparison of the efficiencies from these tables with
those given in Figure 5. 1. 2 for the same horsepower and engine rpm,
as-calculated with the preliminary performance program, indicates
a slight increase in overall system efficiency for the low power (part
load) operation and a decrease in high power (wide-open throttle)
operation. For low power operation, the condenser is oversized,
and thus operates at the minimum pressure (10 psia) allowed. This
factor, coupled with an increased boiler efficiency, accounts for the
improved efficiency under part-load conditions relative to the earlier
calculations. At the higher power levels, Che condenser pressure
must be considerably higher for the required heat rejection so that
5-34
-------
THKRMO ELBCTNOM
the overall system efficiency is decreased slightly relative to the
earlier calculations.
The results of Table 5.6.2 represent one of the poorer situations
for condenser operation: full power output (IR = 0. 29) at low vehicle
speed (5. 70 mph) with low engine speed (300 rpm). The condenser
size used in this report is adequate for this condition with a condenser
pressure of 62. 8 psia, well below the design pressure of 100 psia.
Thus, it can be concluded that the condenser is adequate for total
condensation in the completely sealed system for any condition which
can be encountered in operation of an automobile.
5-35
-------
1-1089
the overs. 11 systeir. eiLic-ieii
- r i. = _ ;_c..-. r^.i:.
PREDICTION FROM GENERALIZED MODEL
in* ret.;.
tor
1» 331*9* 313.13 •65.34 89IL 194.o33 13.125 183.757
*UT 330*00 300.00 123>*0 LIME il*2 1.002 .000
1*
t
t
i
* »EiB\r?t
3 8UT 330*22 62.73
397*** 42.96
152.»13 *36.013 3^.699
•013 .0*5 .000
• CSV
*
9
11
11
CrCCE
9UT
I*.
9UT
62.77
f'frr
• 006
• •1*1'
• 000
.124'
.020
• 351817
•026308
•089914
28.13
-c' cori'?-""-^-" - J ececv-s.:-
1 «.0*r 13J.815 t*SI*39
42.20 «9i».08 PUHP 2*8*9
91*.42 >90>33 LINE »001
.973
.177
28**90 91*.33 -90*33 «OL 49.0*1 .812 29.971 •089916
333*9* 911.32 -60-34 LINE -«000 .397 .000
3000
»96.8S
.037
.7190
-fl« 32.736
I».T«KC
000
• 000
• ;57
• 000
971
000
t<07
5.70
139.67
.000
22«8»5
8FH. 300
V2« .il88
01* «»836
.6979
21.088
Err. 9.68
V3* 1*64*
02* ».JBO
83.7(
10193
600*0
2.0
-tf»T
F*S
.18
8UT.263.16'
52.3
.8*97
TAK QP« .16
COME OP. .17IN
LI6LIO. *3.5
.7792
VEU
.01
.08MP
PSUB* .96 »S1
1.737
5-36
-------
1-1090
TABLE 5.6.2
PERFORMANCE PREDICTION FROM GENERALIZED MODEL
1*
1
»
3
%
5
4
11
12
13
MIL 1*
BUT
ENO IK'
BUT
P.EQV IK
BUT
CBND I>
BUT
PUMP IN
BUT
REGL I*
BUT
CVCLF
ENOINE
T * k
25»*75 911.59-102*99
550*00 900.00 123*»0
9.9*89 »99.21 123*»0
306*03 1».29 69*17
3QS*99 13.91 69*17
194*49 13.40 39*29
194*»» 13.59 39*29
173*69 t». 36-13*. 25
173.65 13.92-11«.25
176*53 912.7»-132**8
176*53 512. 53-13I' »8
25»« 75 511.89-10J.59
n.ewPAT£. 2690 LB/MH
PI* »*1.33
D.IV. 1.005
ErFTM* .9C26
JC»« 59. 2M
OT 0
§811.295.253 11
LINE .112
ENO 2*3.853 »8»
LINE .0*9
•EOV 109.939
LINE .Oil
CBND 22.786
LINE .000
HJHP Z.J79 ,73
LINE -.001
•EOL 78.217
LINE -.001
fPM. 60.00
M» 3**18
OUEV. .000
ErFME. .9135
hSHAFT* »8»8
f
.595
.789
.958
.272
.171
.051
.806
.4*0
.338
.213
.637
.298
»1
DM
229.99.
.000
98.232
• 000
29.8i2~
.000
169.935
.000
1.77*
.000
29.882
.000
0 HP
.591956
•012635 99.68
.079196
• M9323
1*19
.079196
HP. »9.02
V2. .188
01* »*767
EFFALL* .82*9
WN£T. .7.067
• 9.61
Err. 17.73
V3* 6.788
02* 2. 280
r«N
C1NOENSE1*
DATIB..0750
87.16
PL*.* »oi3»
»f» *00.0
lN.it6.oor
• 3.0
• 9700
[•»• .703 -BTu/MP.
.00
A|R 6uT.158.t9r
81.8
. .7997
.8515
r*s OP* .00
CBRE DP* 1.60IN
LIQUID* 15.2
EFFALL* .7797
VEL DP* 1.60
CB*e PftwER. 2.56MP
PSUB* 2.09 PS!
PUMP
t.77»
EL* >2661
RES • .2003E-0*
QES2».2193E>0*
ITL* 9
-------
1-1091
I-1'."-
TABLE 5.6.3
PERFORMANCE PREDICTION FROM GENERALIZED MODEL
OT
Of
14 MIL I"
1 9UT
:; I>G i»
5 " 8UJ
6 fC^O I'
9 ' 8UT
13 AJ*^ I*!
U """ £UJ
1» BEOL I*
13 " Jixr
CVCLE
t,OT*E
34S'C2 530.18 -60.37
550*00 500.00 123. »0
549-27 494.87 133.10
233; 30 .61 .06' . J9.47
393;?» .60.61 59«»7
293*71 59.96 58.96
271*11 61.0* -^6*58
27t»ll 58.33 -96.58
273«8» 535.72 -«1».89
3»5«02 ^3?. iO" •*»'«37
r'LBfBATE* 6897 LB/*"
P^.>79.50-
5.IV. .270
Errtk.. .7887
WF«. 3*. 197
I ..TAKE «ATin..2400
S8IL 204.976 30.178
LINE ..728 5.128
E^O 155.969 433.811
LJN£ • -.0*7 .455
BECV 99.503 .543
LINE : -.021 ' .J32
C9S3 22.605 -1.077
LINE .000 2.710
PUM" .2«726 404.517
LINE 1-.Q04 1.431
REGL 7l.l«6 3.621
LINE -.003 • 1.926
l rPM. 15.21
M« 13*>77
CWEV. .300
fffni* .9504
WSHAFT. 25*633
BPM« 800
187.772
.000
33.927
.000
30.518
• 009
155.533
.000
1.688
• 000
30.518
.000
1.295233
•047988 73.10
•210502
1.072826
4.57
•210532
MP. 61.60
V2. .190
01> 6.957
FTFALL* .7496
*"CT> 23.945
Err. 10.04
V3« 1.705
02- 3.380
FFF- 88.30
AlH FL9-. 30568
TAK P-PH- 1600.0
AI« IN- 95.OOF
2.0
3.30
At" 9UT-232
C4NQ- 04.t
P9ILE"
• 98QO
upAT IN. J.561 -HTu/MO
.7751
FAK OP- .97
C9RE DP. 1.07IN
LIQUID* 13.1
.7596
VEU DP- .11
PSUB* 7.52 PS1
PUMP 46RX. 1.688
5-38
-------
THBRMO •LBCTRON
aolPOIATIOII
6. CONCLUSIONS
The major emphasis in developing the conceptual design in this
study has been on the development of a system which has cost, per-
formance, and convenience competitive with current automotive in-
ternal combustion engines and which uses the current state-of-technology
to the fullest possible extent. The study has indicated that a Rankine-
cycle automotive propulsion system with reciprocating expander and
thiophene (or similar) working fluid has a very strong potential of
meeting these objectives, with emission levels for all three of the
major pollutants significantly less than the projected 1980 federal
standards.
Specific conclusions arrived at in the conceptual design study
are:
a. Packaging
A Rankine-cycle propulsion system competitive in performance
with a 302 cubic inch displacement internal combustion engine can be
completely packaged in the engine compartment of a 1969 Ford Fair-
lane with only minor internal sheet metal and frame modifications
required.
b. Weight
The Rankine-cycle propulsion system has a total weight, as
designed, approximately 20% greater than the equivalent internal
combustion system. Design refinement of the Rankine-cycle system,
coupled with changes required in the internal combustion engine system
to meet future pollution requirements, will decrease this difference.
6-1
-------
THBRMO ELECTRON
c. Ehgine Valving
A strong incentive exists for use of variable intake engine valving.
• Tr.e r- 3.; '.:•-: -\..-.i -..•-• • ;t--.-= ~ :• :
A maximum intake ratio of 0. 29 is preferable to 0. 8 because of a
FtM- ..-.-••-- ..... - ' -.
reduction in the feedpump size and a reduction in the condenser cooling
icrrr s.r. -. .- i." ; . : r . : : ::--..- :-r = -: z. -~: ~: '.'..'..
air requirement at low vehicle speeds with a relatively small effect on
system performance.
cvcl? 4»- • Transmission
f 2f^tran-Brnissi6n whi'ch permits the rriain* propulsion engine
.''.-. '
to idle at zero vehicle speed, allowing accessories to be driven directly
the engine, is preferable to use of a direct coupled engine requiring
y engine to drive the accessories. A two- speed transmission
provides- a significant decrease in the 0-60 mph acceleration time
and an improved gradability of the vehicle relative to a single- speed
transmission.
e. Condenser
A condenser -fan combination with sufficient capacity to handle the
peak-load condensing requirements in the completely-sealed system is
feasible for the 100 shp system using state-of-the-art technology.
f. Performance
A Rankine-cycle system of 100 shp net output (184 cubic inch
engine displacement) with two- speed transmission is almost equivalent
in acceleration performance and top speed to a 302 cubic inch displace-
ment internal combustion system with three-speed transmission (adver-
tised power of 220 hp).
6-2
-------
THBRMO ELECTRON
coifoitriod
g. Fuel Economy
With a maximum cycle temperature of 550 CF, the customer-
average fuel economy is 12. 7 mph with the Rankine-cycle system
and 15.7 mph with the 302 cubic inch displacement internal combustion
'engine system. A modest increase in maximum cycle temperature
to 600 °F for the Rankine-cycle system, coupled with reduced fuel
economy anticipated for the internal combustion engine as tighter
pollution emission restrictions are imposed, will decrease this
difference.
. h. Qniasion Level
Projections of current emissions data from burners suitable
for use in compact Rankine-cycle systems indicate a significant
reduction in pollutant emission levels below the current projected
1980 federal standards. Assuming, to be conservative, 10 mpg
fuel economy, the UHC gms/mile is a factor of 5 below the projected
1980 federal standards, the CO gms/mile a factor of 13.4 below,
and the NO gms/mile a factor of 1. 6 below.
i. Cost
A detailed, large-volume, manufacturing cost estimate for the
system is being prepared by the Ford Motor Company for the conceptual
design and will be available at a later date. The cost estimate will be
reported as a ratio relative to the Ford 302 cubic inch internal combustion
engine system with three speed automatic transmission.
j. Reliability and Maintenance
Since only limited test experience is available at this date, no
quantitative information on these factors is available. However, the
6-3
-------
THBMMO BLHCTNON
approach followed in development of the system is similar to that of
hermetically sealed air-conditioning systems. It is expected that,, as
in the air conditioning systems, the system failures and maintenance
«'»£.!.
requirements will occur primarily from control, motor, and accessory
a.r.: ..
failures rather than from failure of the major mechanical components
•uch as the engine or feedpump.
ec; -k. Working Fluid
•P°—-" -The primary consideration in the conceptual design and in the
-first experimental prototype is demonstration of a low cost Rankine-
cycle system competitive with the internal combustion system in
performance and drivability, using the current state-of-the-art to
the fullest extent. Thiophene, in our judgement, best fulfills these
— . . - i
goals, In view of the flammability and toxicity of thiophene, a question
does exist regarding its suitability for large scale use in automobiles
for the general public. It is, therefore, strongly recommended that
a comprehensive search and development effort be initiated for a fire
resistant and low toxic working fluid which retains the otherwise de-
sirable characteristics of thiophene.
6-4
-------
THBRMO KLBCTHON
APPENDIX A
PARAMETERS FOR CHARACTERIZING FLAMMABILITY
CHARACTERISTICS OF MATERIALS
-------
THKRMO KLKCTRON
COIPOII»TIUII
Flash Point
The lowest temperature at which a liquid will give off flammable
vapor at or near its surface. This vapor forms an intimate mixture
with air, and it is this mixture which ignites. The flash point of liquids
is usually determined by the Standard Method of Test for Flash Point
with the Tag Closed Cup Tester (ASTM D56-52). The Interstate
Commerce Commission uses the Tag Open Cup Tester, giving results
5-10'F higher (less flammable).
Fire Point
The lowest temperature at which a mixture of air and vapor
continue s to burn in an open container when ignited.
Autoignition Temperature
The lowest temperature at which a material will self-ignite
and sustain combustion in the absence of a spark or flame.
Explosive Range or Flammability Limits
Range of concentration of material vapor in air, expressed as
per cent by volume, over which the vapor-air mixture will burn when
ignited. The values are generally given for normal conditions of
temperature and pressure.
Hazard Rating of Flammable Fluids
(1) Interstate Commerce Commission
Flash Point < 80°F High Fire Hazard
Flash Point 80-350°F Moderate Fire Hazard
Flash Point >350°F Slight Fire Hazard
A-l
-------
TMBMMO •i.KCTMOM
conpoiirion
(2) National Fire Protection Association
Flash Point < 20°F High Fire Hazard
Flash Point 20-70°F Moderate Fire Hazard
Flash Point 70-200°F Slight Fire Hazard
Flash Point >200°F Not generally called flammable
A-2
-------
THBRMO ELECTRON
APPENDIX B
API TOXICOLOGICAL REVIEWS
THIOPHENE AND DERIVATIVES
AND
GASOLINE
-------
1-1105
API TOXICOLOGICAL REVIEW
THIOPHENE AND DERIVATIVES
SEPTEMBER 1948
W.G.K.
A?P. 1 1 1958
Note: This review summarizes the best available informa-
tion on the properties, characteristics, and toxicology of
tkiopfitne and derivatives. It offers suggestions and tentative
recommendations pertaining to medical treatments, medical
examinations, and precautionary measures for workers who
are exposed to thiophene and derivatives. It was prepared at
the Harvard School of Public Health, Boston, Mass., under
the direction of Professor Philip Drinker. The review has
been accepted for publication by the Medical Advisory Com-
mittee of the American Petroleum Institute. Anyone desir-
ing to submit additional information or proposed changes for
consideration prior to re-issuance of this review is requested
to send them to the American Petroleum Institute.
This review was prepared by Marshall Clinton, M. D.
AMERICAN PETROLEUM INSTITUTE
DEPARTMENT OF SAFETY
'50 WEST 50™ STREET
NEW YORK 20, N. Y.
B-l
Prie« 25 Ccnu
-------
1-1106
API TOXJCOLOG1CAL REVIEWS
THIOPHENE AND DERIVATIVES
TOXICOLOGICAL REVIEW OF THIOPHENE AND DERIVATIVES*
L Substanc*
Thiophene.
Formula: C4H4S.
Structural formula:
HC CH
•Ic
\/
S
Molecular weight=84.13.
Synonym: thiofuran.
II. Properties and Characteristic* >/ :* *•
Melting point =minu$ 38 deg C ( — 36.4 deg F).
Boiling point =84 deg C (183.2 deg F).
Refractive index=1.5285.
1 mg per liter =291 ppm; 100 ppm=0.344 mg
per liter.
Thiophene is a clear colorless heterocyclic com-
pound encountered as an important contaminant of
benzene. It is insoluble in water, but is readily soluble
in alcohol, ether, benzene, and roost hydrocarbons.
Thiophene is difficult to separate from benzene by
physio:! means because of their similar boiling points,
but can be separated fairly readily after reaction of
the more reactive thiophene with other substances,
such as mercury. Thiophene can be obtained from
crude benzene, and now can be synthesized without
excessive difficulty. Thiophene is highly reactive, and
is readily nitrated, sulfonated, haiogenated, or mer-
curated. It can be made to undergo ketone forma-
tion or aminomethylarion without difficulty. It is
usually removed from benzene by sulfuric-acid
treatment.
III. Probable Sources of Contact
Contact with thiophene may occur due to leaks
occurring in the course of its handling or manu-
facture. Contact may result from handling crude
coal-tar benzene, as this contains up to 0.5 per cent
rhiophene. It is not possible to state the most prob-
• Prepared under the auspices of the Subcommittee for Pu-
ouuiSte Concentration! of Toxic Substances in the Petroleum
Industry.
* Figures refer to bibliography on p. J.
able sources of exposure, as this substance is, com-
mercially, relatively new, and its uses are ill-defined
but growing. .
IV. Toxicology
•. C«o«r«l Coniid*r«t!*ai
A considerable amount of study has been devoted
to investigations of the toxicology and pharmacology
of thiophene and its derivatives. Most of the latter
studies have been comparisons of the thio homo-
logues of organic substances of known and, usually,
fairly marked pharmacological activity. Thiophene
in fairly high concentrations has, according to most
authors,4* * an acute narcotic effect greater than equal
concentrations of benzene. Flury and Zernik report
that the inhalation by mice of 2,900 ppm of thiophene
results in loss of consciousness and, in some instances,
death; whereas similar concentrations of benzene
can be tolerated without difficulty. Concentrations
of 8,700 ppm of thiophene caused death of mice in
20 min to 80 min; whereas benzene produced no
such effect.
The acute toxic action of thiophene appears to be
exerted primarily on the central nervous system. It
has a selective action on the equilibrium centers of
the cerebrum and cerebellum, producing severe ataxia
following repeated injections.8- • Thiophene pro-
duces fairly diffuse changes in the cerebellum, par-
ticularly the vermis, characterized by degeneration of
nerve cells in these areas. There may be superim-
posed vascular changes.1"* * The metabolism of thio-
ohene is poorly understood, although it is stated that
5 to 12 per cent is recovered in the urine in conjugated
form. Hov.-ever, there is no increase in the con-
jugated sulfates, and the total sulfate excretion de-
creases after administration of thiophene.%^
b. Aeui« Effects
As already noted, acute exposure to high concen-
trations of thiophene results in nervous-system de-
pression. Repeated daily injection of 2 g of thiophene
in dogs results in locomotor ataxia and paralysis,5
rather similar to those noted in severe CS2 poisoning.
The effects of thiophene on humans has not been
described.
B-2
-------
1-1107
API TOXICOLOGICAL REVIEWS
THIOPHENB AND DERIVATIVES
c. Chronic Effiett
No reports on the chronic effects of repeated or
prolonged exposure to low concentrations of thio-
phene are available.
4. $•!• Limit*
Extreme concentrations of thiophene are obviously
intolerable, as they produce acute poisoning. No
information is available, however, on the effects of
repeated exposure to lesser concentrations, such as
100 to 1,000 ppm. Therefore, no safe limits have
been or can be promulgated at present.
V. Treatment
No information on the possible therapy of thio-
phene poisoning is available,
VI. Examinations
The present state of knowledge concerning thio-
phene does not permit the establishment of any
special pre-employment or periodic examinations.
It appears sensible to employ only men in good health
to work with thiophene and to re-examine them fre-
quently for possible evidences of blood dyscrasia or
neurologic disturbances, but these measures may
prove unnecessary.
VII. Precautionary Measures
Thiophene should be handled with extreme care,
in closed systems or with adequate ventilation, until
its chronic toxicity is established or shown to be
absent.
VIII. Bibliography
1. C D. Hodgman and H. N. Holmes. Handbook of
Chemistry and Physics, 25th edn., Chemical Rubber
Publishing Co., Cleveland (1941).
2. P. Karrer, Organic Chemistry, Nordemann Publishing
Corp., New York, 330, 700 (1938).
3. Anon., Thiophene Chemicals, Socony-Vacuurn Oil Co.,
Inc. Research and Development Laboratories, New
York (1946).
4. F. Flury and F. Zernik, 'Toxicity of Thiophene," Chem.
Ztg. 56, 149 (1932).
3. A. Quistomanos, "ExperimentaJ Production of Cere-
bellar Symptoms by Thiophene," Klin. \\'o-cbtcbr. 9,
2334 (1930).
6. A. Christomanos, "Action of Organic -Sulfur Com-
pounds on the Dog Organism: Action and Fate of Thio-
phene in the Metabolism of the Dog," Biochem. Z.
229,248 (1930).
7. T. Upners, "Experimental Studies Concerning the Local
Action of Thiophene on the Central Nervous System,"
Z. ges. Neural. Psychial. 166, 623 (1939).
8. A. Christomanos and W. Scholz, "Electricity of Toxic
Substances for the Central Nervous System: Clinical
and Pathological Studies of Thiophene," Z. ges. Neural.
Psjcbiat. 144, 1 (1933).
B-3
-------
1-1108
API TOXICOLOGICAL REVIEW
GASOLINE
FIRST EDITION, 1967
The information and recommendations contained in this publication have been
compiled from sources believed to be reliable and to represent the best current
opinion on the subject. No warranty, guarantee, .or representation is made by the
American Petroleum Institute as to the absolute correctness or sufficiency of any
representation contained in this and other Toxicolosic.il Reviews, and the Institute
assumes no responsibility in connection therewith; nor can it be assumed that all
acceptable safety measures are contained in this and other Toxicological Reviews,
or that other or additional measures may not be required under particular or
exceptional conditions or circumstances. The American Petroleum Institute, as
sponsor of this review, takes no position as to whether or not any method contained
herein is covered by an existing patent, nor as to the validity of any patent alleged
to cover any such method. Furthermore, nothing contained in this review grants
any right, by implication or otherwise, for the manufacture, sale, or use in con-
nection with any method, apparatus, or product covered by letters patent.
This review was prepared by the Committee on Toxicology and accepted by the
Central Committee on Medicine and Health. Anyone desiring to submit additional
information or proposed changes for consideration prior to reisr.uance of this review
is requested to send them to the American Petroleum Institute.
AMERICAN PETROLEUM INSTITUTE
1271 AVENUE OK THE AMERICAS
NEW YORK, N. Y. 10020
B-4
-------
1-1109
API TOXICOLOGICAL REVIEW OF GASOLINE
i;.»siilinc is a refined petroleum product suitable for
t:.. operation of an internal-combustion engine. It is a
complex mixture of hydrocarbons to which arc usually
aJJc»! antiknock agents, inhibitors, and dyes. The hy-
drocrubons present are primarily paraffins, naphthcnes,
aromatics, and olefins. Widely varying amounts of the
individual hydrocarbons arc contained in typical gaso-
Imc blends, depending on sucli factors as the origin
of the blending streams, seasonal requirements, and
intended use.
II. Properties and Giaracteristics
Distillation range
(byASTMDSG)
Specific gravity
Flash point
(Tag closed cup)
= 32 C to 225 C (90 F to 437 F)
= approximately 0.71 to 0.77
May be as low as — 45 C
(-50 F)
Explosive limits = 1.3 to 6.0 percent by volume in air
Because the composition of gasoline is variable, it is
possible to make only general statements on the prop-
erties and characteristics.
It is an extremely flammable liquid and is water-white
to straw-tint in color before the addition of dyes.
HI. Uses and Probable Sources of Contact
Gasoline is intended for use as a motor fuel. Al-
though sometimes used as a cleaning agent and as a
substitute for other solvents, it should not be used other
than as a motor fuel unless adequate precautions are
taken to control both the potential health hazard and
the fire hazard.
In refineries, exposure to gasoline vapors may occur
at process units, in the repairing of refinery equipment,
in the cleaning of storage tanks, during the gaging of
tanks, and in the laboratory. Potential exposures may
also occur in the filling of tank cars, tank trucks, drums,
storage tanks, and in the fueling of automobiles and
aircraft
Gasoline exposures also occur from the use of gaso-
line as a cleaning agent or as a solvent substitute or
from careless handling and storage which may result in
accidental ingcstion.
IV. Toxicology
a. Gencr.nl Considerations
Gcrardc " points out tint although the hydrocarbon
composition of gasoline has changed over the years, its
basic pharmacology' and toxicology have not altered
significantly. The signs and symptoms of intoxication
from acute exposure to gasoline arc similar to those for
an exposure to heptane, namely, marked vertigo, inabil-
ity to walk a straight line, hilarity, and incoordination.
Depending on the mctliodr. of manufacture and on the
blending components, gasoline may contain benzene. In
cases of repeated exposure to significant amounts of
gasoline vapor, the potential hazard of benzene expo-
sure may have to be considered. Benzene is unique
among the hydrocarbons in its ability to depress the
hemopoietic system.*
Numerous additives may be present in the many
branded gasolines. In general, these materials are added
in very low concentrations and do not contribute sig-
nificantly to the toxicity of gasoline by inhalation and
skin contact. In the case of tetraethyllead (TEL).
Kehoe ••4 states that the low concentration of TEL in
gasoline effectively prevents the absorption of significant
quantities of TEL through the skin. Similarly, the va-
porization of TEL from gasoline is so low at ordinary
temperatures as to preclude its presence in more than
minute quantities in gasoline vapor. Hence, persons
dispensing gasoline as a motor fuel have no significant
exposure to TEL. However, if gasoline containing
TEL is spilled, sprayed, or otherwise vaporized in un-
ventilated or enclosed spaces, the concentration of TEL
may exceed safe levels. Similarly, a significant exposure
to TEL vapors occurs in the removal of sludge from
storage tanks which have contained leaded gasoline.
Because tetramethylkad (TML) is more volatile than
TEL, Kehoe, et al.,* recently investigated the handling
of gasoline containing this material. In a study of both
refinery workers and service station attendants, it was
concluded that exposure of the various groups to TML,
under the prevalent environmental conditions of the oc-
cupations, is negligible. The comments made in the
preceding paragraph concerning the hazard of TEL
gasolines in unventilatcd or confined spaces also apply
to gasolines containing TML.
fc. Aente Toxicilr
inhalation
Browning • states that many severe or fatal cases from
inhalation of gasoline vapors are reported but these
• Figures refer to BIBLIOGRAPHY on p. 5.
B-5
-------
I-1110
GASOLINE
have almost always involved men who entered tanks
containing high concentrations of gasoline vapors. Ac-
cording to von Oettingen,' inhalation of very high con-
centrations of vapors may cause sudden loss of con-
sciousness, coma, and sudden death. Browning* and
von Oettingen T report the following signs and symp-
toms: In severe cases, delirium with cyanosis, coma,
tonic and clonic convulsions, shallow and stertorous res-
piration, and a thready pulse. Vomiting, inward strabis-
mus, contracted pupils, and loss of reflexes have also
been observed. In less severe cases, headache, flushing
of the face, nausea, mental confusion and depression,
anorexia, blurring speech, and difficulty in swallowing
have been observed.
Oerarde * states that there are a number of reports
in the clinical literature which indicate that very acute
hydrocarbon intoxication may cause central nervous
lystem sequelae, such as convulsions or seizures, several
months after the initial acute exposure.
Wang and Irons • report a fatal case of gasoline in-
toxication in a man who entered an unpurgcd aircraft
wing tank in which the concentration of vapor was esti-
mated to be 0.5 to 1.6 percent Approximately 5 min
after entry, he was found unconscious within the tank
and was quickly removed. Artificial respiration was ap-
plied immediately but the patient died en route to
hospital. Autopsy revealed acute pulmonary edema,
acute exudative trachcobronchitis, passive congestion of
the liver and spleen, and early hemorrhagic pancreatitis.
The authors believe that the clinical history and patho-
logical findings were entirely compatible with a diagno-
sis of death due to hydrocarbon poisoning.
Necropsy, on a youth who died while using a ladle to
fill a 2-gal can with gasoline from a supply barrel, re-
vealed nothing abnormal other than raw areas of skin
on the wrists and upper arms. Concentration of vapor
was estimated to be 500 ppm to 30,000 ppm.»
MacLean " reports three cases of fatal aplastic ane-
mia which arc assumed to have occurred after siphon-
ing gasoline containing benzene or inhaling its vapors.
As the benzene content of the gasoline was approxi-
mately 10 percent, benzene rather than the gasoline
itself is thought to be the cause of the anemia.
Ainsworth n reports the following postmortem find-
ings in a young boy found unconscious in a pool of
gasoline: Hypcremia was present in all organs exam-
ined; lungs showed considerable edema, some intraal-
veolar hemorrhage, and necrosis of alveolar walls; su-
perficial epidermis was loose and could be stripped off
with ease.
imgution
The acute oral toxicity of gasoline for rats ranges
from 10 to 35 g per kg of body weight." Although
serious poisoning of humans may result from the ingcs-
tion of 20 g to 30 g of gasoline, the usual fatal dose for
adults is approximately 350 g. With children 10 g to
15 g may be fatal." The variation in susceptibility is
caused by a number of factors, including the presence
of food in the stomach and, most importantly, whether
respiratory aspiration occurs.
According to von Octtingcn,T the ingestion of gaso-
line causes a burning sensation in the mouth, pharynx
and chest, and intense irritation of the gastrointestinal
tract, with vomiting, colic, and diarrhea. Dizziness,
unconsciousness, and coma may also result. In ncn-
fatal cases, bronchitis, pneumonia, and nephritis may
develop.
In cases of ingestion, it is generally believed that the
resulting pncumonitis, if present, is caused by aspiration
of gasoline into the lungs. Gerarde " has shown that
the aspiration of as little as 0.2 ml by rats causes instan-
taneous death.
c. Chronic Toxicltr
Gerarde1 states that in service station attendants and
garage workers who arc exposed repeatedly to low con-
centrations of gasoline vapors with brief exposures to
higher concentrations, there is no conclusive evidence
of harmful health effects due to exposure to gasoline
vapors.
Browning* expresses the opinion that reports of
chronic poisoning are few and vague; most of the au-
thorities that mention them quote from earlier works
rather than from personal experience. Machle " reports
chronic poisoning to be rare. In his study, 2,300 refin-
ery workers showed no symptoms, nor did service sta-
tion attendants, tank wagon drivers, etc. However, he
states that "barrel fillers," who were exposed to concen-
trations which might well be intolerable to many people,
showed sighs of malnutrition, pallor, anorexia, nausea,
nervousness, and low hemoglobin in a high proportion
of the individuals involved.
MacLean l° reports that an oil company employee
developed hcmolytic anemia and myclofibrosis after
12 months' exposure to gasoline vapor resulting from
spills. He also reports a case of thrombocytopcnic pur-
pura in a man who had cleaned metal parts in gasoline
over a 2-year period. In the Orst case, the benzene con-
tent of the gasoline was less than 1 percent; in the sec-
ond case, benzene content may have been as high as
10 percent
Sterner " reports on a group of painters who were
exposed to vapors released during spray painting with
gasoline-diluted paints. The concentration of aromatic
hydrocarbons in the vapor was from 300 ppm to
800 ppm, and it was assumed that the total hydrocar-
bon concentration would be five to ten times more than
B-6
-------
I-1111
API TOXICOI.OGlCAt. KUVlliW
this range. The chief symptoms were headache, na
weakness, mental depression, anorexia, and inability for
sustained attention and activity. One case showed a
tremor and weakness of the arms and legs with multiple
librillary iwilchings on fatigue. A significant decrease
was noted in hemoglobin, crythrocytcs, and blood cell
volume values, with an increase in mean corpuscular
hemoglobin, mean corpuscular volume, and reliculocytc
count.
Oldham 1T reports on a 17-year-old girl who indulged
in the habit of repeated self-intoxication with gasoline
vapor, three or fuur times a week, o\\c a period of two
years. The patient described the effects as consisting of
"dreams" followed by giddiness, nausea, and vomiting.
There were no apparent chronic effects.
Drinker, et al.,li state that exposure to 1,000 ppm for
1 hr caused slight dizziness, nausea, and headache in
human subjects. When the concentration reached
2,600 ppm, all subjects were drunk and somewhat anes-
thetized. In a group exposed to 160 ppm and 270 ppm,
for 8 hr with a midday break of 1 hr, the most distinc-
tive symptoms were irritation of the eyes and throat.
Davis, ct al.,1* exposed human subjects to nonlcadcd
gasoline of the following approximate compositions:
Sample A
Sample B
Sample C
Paraffins
(Percent)
25
40
30
Naphthenes
(Percent)
30
35
5
Aromotics
(Percent)
40
20
65
Eye irritation was the only effect observed after Vi-hr
exposure to vapor. Irritation increased as the con-
centration increased from 200 ppm to 500 ppm to
1,000 ppm. There was no significant difference in the
irritating potential of the three gasolines.
. Skin Efftelt
Gasoline Is a primary skin irritant and prolonged con-
tact may dry and dcfat the skin with resultant derma-
titis. Scnsitizat^ion " to gasoline has been reported but
is not common.
Eyt Effect*
Gasoline causes smarting and pain on splash contact
with the eyes, but only slight transient corncal epithelial
disturbances."
Exposure of rabbit eyes to both leaded and unleaded
gasoline caused the conjunctiva to become moderately
edcmatous and hypcrcmic. However, the injury \vp.s
superficial and transient. There was no difference in
reactions between the leaded and unleaded gasoline.
<1. I'crnti.-siltlo Limit* of Exposure
. The threshold limit value adopted by the American
Conference of Governmental Industrial Hygicnists for
an S-hr day is 500 ppm."
V. Treat incut
InJia'nt'tOH
The first aid care of a victim of acute vapor exposure
requires his immediate removal from the contaminated
atmosphere. Rescuers should take suitable precautions
to prevent their bcin^ overcome by hir^i conccrtntio-is
of vapors. If breath:;!!; is interrupted, artifid:1' r.:^::iu-
tion should be applied immediately. A physician should
be called. Medical treatment is symptomatic. No specific
measures arc advocated.
Ingcstion of gasoline is rarely encountered in indus-
try. In the home it may be accidentally swallowed and
usually by very young children.
A person who has ingested gasoline should be given
olive oil or some other vegetable oil orally to retard
absorption of the gasoline. Gastric lavagc qnd the in-
duction of vomiting arc not advisable because of the
possibility of aspiration of gasoline and the subscc4i:cnt
development of chemical pneumonia. The use of oxy-
gen and antibiotics prophylactically for the prevention
of secondary bacterial pneumonia may be indicated.
VI. Examinntlono
Special health examinations for the determination of
toxic exposure arc unnecessary for most employees
engaged in the manufacture, distribution, and sale of
gasoline as a motor fuel.
However, persons who may be repeatedly exposed to
significant amounts of gasoline vapors or mists shou'.d
be given a prcernploymcnt examination, and those with
evidence of blood dyscrasias or serious systemic disor-
ders should be excluded from such work. Periodic
rccxaininations should be carried out at intervals to be
determined by the physician in charge. Particular atten-
tion should be paid to evidence of eye irritation., derma-
titis, or symptoms related to the nervous system.
VII. Precautionary Measures
In gasoline manufacture, there is very little actual ex-
posure to gasoline vnpor because the process is carried
out in a closed system.
Before entering r~..,\s, vessels, or other confined
space, adequate ventilation should be provided to main-
tain the gasoline vapor concentration at a safe level. In
the cleaning and repairing of storr.gc tar.'cs in kv.c'-.-d
gasoline service, special precautions r.rc also ncccssr.r.y
B-7
-------
1-1113
GASOLINE
hazard-front the lead-«ntiknocte
that may be present. . • : ... .-. '
^Adequate local exhaust, ventilation should be pro-
vi^jcd at daim-filling operations conducted indoors.
• Leaded gasoline should be used only as a motor fuel.
..Gasoline should never be siphoned by mouth.
-Gasoline should be stored in a cool, well-ventilated
place. To avoid accidental ingcstion, it should be stored
in clearly marked containers, well out of the reach of
children.
' If unleaded gasoline is used as a solvent substitute,
adequate ventilation should be provided, as well as
the .other-necessary precautions; to prevent?fire and.
ejipjosion, -: -• : :-: v.--'
•Repealed or prolonged skin contact should be
avoided. If such contact is necessary, protective cloth-
ing and gloves should be worn. Goggles may be worn as
protection against accidental splashes.
VDL Bibliograph)
\'~ ." '• •' '• '
1 H. W. Cerardc, "Aliphatic Hydrocarbons," Industrial
Hygiene and Toxicology 2nd cdn., ed. Frank A. Patty, II
Intencience Publishers, New York (1962).
* API' lexicological Review: Benzene. "2nd edn., Am.
Petrol. lost.. New York (1960).
•R. A. Kehoe, "Industrial Lead Poisoning," Industrial
Hygiene and Toxicology 2nd cdn., cd. Frank A. Patty. II
Intencience Publishers, New York (1962).
* R. A. Kehoe, J. Cholak, J. A. Spence, and W. Hancock,
Totcntial Hazard of Exposure to Lead I. Handling and
Use of Gasoline Containing Tetramcthyllcad," Arch. En-
viron. Health 6 [2] 239-54 (1963).
• R. A. Kehoe. J. Cholak, J. G. Mcllhinncy, G. A. Lof-
quist, and T. D. Sterling, "Potential Hazard of Exposure to
Lead II. Further Investigations in the Preparation, Han-
dling and Use of Gasoline Containing Tetramethyllead,"
Arch. Environ. Health 6 [2] 255-72 (1963).
**E. 'Browning; Toxicity of Industrial Organic Solvents.
Medical Research Council Report SO. Her Majesty's Sta-
tionery Office, London (1953).
T W. F. von Octlingcn, Poisoning. 2nd cdn., \V. D. Saun-
ders Co., Philadelphia (1958).
•C. C. Wang and G. V. Irons, Jr., "Acute Gasoline In-
toxication," Arch. Environ. Health 2 [6] 714-16 (1961).
• R. Aidin, "Petrol-Vapour Poisoning," Brit. Mcil. J. 2
369-70(1958).
'•J. A. MacLcan, "Blood Dyscrasia After Contact With
Petrol Containing Benzol," Med. J. Australia 47 [2] 845-49
(1960).
11 R. \V. Ainsworth, "Petrol-Vapour Poisoning." Hrlt.
Med. J. 1 1547-48 (1960).
11 W. B. Dicchmann and H. W. Gerardc, Symptomatol-
ogy _&. .Therapy of Toxicological Emergencies. Academic
PreJs^Ncw York (1964).
}*T. Sollman, A Manual of Pharmacology. 8th edn.,
Wf'B: Saundcrs Co., Philadelphia (1957).
14 H. W. Gerardc, "Toxicological Studies on Hydrocar-
bons, IX. The Aspiration Hazard and the Toxicity of
Hydrocarbons and Hydrocarbon Mixtures," Arch. En-
viron. Health 6 [3] 329-41 (1963).
"W.Machlc, "Gasoline Intoxication," /. Am. Med.
Aaoc. 117 [23] 1965-71 (1941).
"JlH. Sterner, "Study of Hazards in Spray Painting
with.Gasoline as Diluent," /. Ind. Hyg. Toxicol. 23 [9]
437-47(1941).
" W. Oldham, "Deliberate Self-Intoxication with Petrol
Vapour." Brit. Med. J. 21687(1961).
»• P. Drinker, C. P. Yaglow, and M. F. Warren, "The
Threshold Toxicity of Gasoline Vapor," J. Ind. Hyg. Toxi-
col. 25 225-32 (1943).
» A. Davis, L. J. Schafcr, and Z. G. Bell, "The Effects
on Human Volunteers of Exposure to Air Containing Gaso-
line Vapor," Arch. Environ. Health 1 [6] 548-64 (1960).
•• J. B. Bicdcrman, "A Case of Contact Dermatitis Pro-
duced at a Distance by the Sensitizing Agent," /. Am. Med.
Assoc. 106 2236-37 (1936).
n W. Morton Grant, Toxicology of the Eye, 414, C. C.
Thomas, Springfield, III. (1962).
*» "Threshold Limit Values for 1964" Arch. Environ.
Health 9 W 545-54 (1964).
B-8
------- |