United States
Environmental Protection
Agency
Environmental Research
Laboratory
Corvallis OR 97330
EPA-600/3-78-055
May 1978
Research and Development
vvEPA
Research on Control
Technology for
Ice Fog From Mobile
Sources
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3 Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6 Scientific and Technical Assessment Reports (STAR)
7 Interagency Energy-Environment Research and Development
8. "Special" Reports
9 Miscellaneous Reports
This report has been assigned to the ECOLOGICAL RESEARCH series This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials. Problems are assessed for their long- and short-term influ-
ences Investigations include formation, transport, and pathway studies to deter-
mine the fate of pollutants and their effects. This work provides the technical basis
for setting standards to minimize undesirable changes in living organisms in the
aquatic, terrestrial, and atmospheric environments
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161
-------
EPA-600/3-78-055
May 1978
RESEARCH ON CONTROL TECHNOLOGY
FOR ICE FOG FROM MOBILE SOURCES
by
Harold J. Coutts
Ronald K. Turner
Arctic Environmental Research Station
Corvallis Environmental Research Laboratory
College, Alaska 99701
CORVALLIS ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U. S. ENVIRONMENTAL PROTECTION AGENCY
CORVALLIS, OREGON 97330
-------
DISCLAIMER
This report has been reviewed by the Corvallis Environmental Research
Laboratory, U.S. Environmental Protection Agency, and approved for publica-
tion. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
-------
FOREWORD
Effective regulatory and enforcement actions by the Environmental Pro-
tection Agency would be virtually impossible without sound scientific data on
pollutants and their impact on environmental stability and human health.
Responsibility for building this data base has been assigned to EPA's Office
of Research and Development and its 15 major field installations, one of which
is the Corvallis Environmental Research Laboratory (CERL).
The primary mission of the Corvallis Laboratory is research on the ef-
fects of environmental pollutants on terrestrial, freshwater,,, and marine
ecosystems; the behavior, effects and control of pollutants in lake systems;
and the development of predictive models on the movement of pollutants in the
biosphere. CERL's Arctic Environmental Research Station conducts research on
the effects of pollutants on Arctic and sub-Arctic freshwater, marine water
and terrestrial system; and develops and demonstrates pollution control tech-
nology for cold-climate regions.
This report describes a two winter investigation of technology for con-
trolling ice fog emmissions from mobile sources.
A. F. Bartsch
Director, CERL
-------
ABSTRACT
Automotive generated ice fog i.s a form of air pollution that results when
exhaust water vapor freezes into minute particles which form a dense fog.
This study on control techniques was conducted by the U.S. Environmental
Protection Agency at its Arctic Environmental Research Station near Fairbanks,
Alaska.
The major control technique evaluated was the cooling of exhaust gases to
well below the dew point, thus condensing water vapor into a liquid stream
before final discharge.
During the winter of 1974-75 nine exhaust gas cooler-condensers were
installed on local vehicles and their water vapor removal performances were
evaluated. Based on these data three cooler-condensers were fabricated,
installed, and more intensely evaluated during the winter of 1975-76. The
sizing criteria developed the first winter were found inadequate because ice
film formation decreased heat transfer efficiency. Cooler-condensers must be
designed to avoid or to accommodate condensate freezing.
An ice fog mass emission reduction up to 80 percent was attained with
cooler-condensers on motor vehicles. However, the increase in visibility over
roads was not quite proportional because of the many other ice fog sources.
The overall impact of automotive ice fog control would be a visibility in-
crease of at least 70 percent in areas where motor vehicles create 50 percent
or more of the ice fog.
Control of automobile-generated ice fog would also mean cleaner air, but
perhaps more ice on the road. Cleaner air would result because sulfur oxides
and lead compounds would be absorbed in the condensate. This condensate, if
allowed to drip from the cooler-condensers, would freeze onto the road and
require a more intense snow removal effort.
This study has shown that cooler-consensers are effective ice fog control
devices for mobile sources. The next step is to further evaluate and demon-
strate the devices on fleet vehicles used in the dense ice fog areas.
This report covers a period of work from July 1974 to May 1976.
-------
CONTENTS
Foreword ii
Abstract iii
Figures vi
Tables vii
1. Summary and Conclusions 1
2. Recommendations 4
3. Introduction , 5
Background 5
Scope 8
4. Description of Ice Fog Control Device Installations 11
Test Methods and Instruments 11
Heat Exchanger Definition 13
Devices Evaluated by the Arctic Environmental Research Station . 14
Cooler-Condensers Evaluated by Private Contractors 17
5. Device Performance and Comparison 19
First Winter Results 19
Mist Coalescers 26
Second Winter Results 28
Constant Speed Performance 28
Drive Through Town 35
Backpressure and Tube Icing 35
Backpressure and Fuel Economy 41
Passenger Compartment Carbon Monoxide Measurements 41
Overall Ambient Ice Fog Reduction 42
-------
6. Future Applications 45
Selection of Heat Transfer Medium 45
Mounting Locations 46
Corrosion 46
Costs 47
Recommendations 47
7. Environmental Considerations 48
Additional Ice on the Road 48
Reduction of Other Air Pollutants 49
Condensate Quality 51
References 52
APPENDIX
A. Detailed description of ice fog control device installations. . 54
B. Calculation of condensation curve and heat exchanger duty.. . . 79
C. Heat exchanger design techniques 82
D. Estimation of road icing from ice fog control on automobiles. . 86
E. Low temperature psychrometric chart, example illustrating use.. 88
VI
-------
FIGURES
Number Page
1. Condensation curves for three automotive fuels. 9
2. Exhaust gas heat content, mole weight and mass. Percent water 20
vapor condensed.
3. First winter cooler-condenser exhaust temperature ranges. 24
4. First winter cooler-condenser exhaust temperatures. 25
5. First mist coalescer. 27
6. 1968 Chevy Carryall (4x2) second winter cooler-condenser performance. 30
7. 1974 Chevy Nova second winter cooler-condenser performance. 31
8. 1967 Mercedes Benz Diesel second winter cooler-condenser performance. 33
9. 1971 CMC Jimmy with di1utor-ambient air heater. 34
10. 1971 CMC Jimmy without di1utor-ambient air heater. -^
11. Cooler-condenser performance during a drive through town. 37
12. First winter cooler-condenser back pressure. 39
13. 1974 Chevy Nova second winter cooler-condenser back pressures. 40
VI 1
-------
TABLES
Number Page
1. Effect of fuel economy on automotive ice fog (H20) emission. 7
2. First winter performance of prototype automotive ice fog control 21
devices.
3. Second winter comparison of overall heat transfer coefficients at 36
64 km/h.
VI
-------
SECTION 1
SUMMARY AND CONCLUSIONS
Ice fog is a form of air pollution caused by water vapor released into
air too cold to retain the water in the vapor phase. Thus the vapor condenses
into ice particles. In cold climates where this condition is common during
the dark winter months, the reduced visibility is a major citizen complaint.
There are many strategies available to reduce water vapor emissions from
all sources. The two main strategies for reducing automotive-generated ice
fog are: (1) reducing combustion engine vehicle use or (2) reducing com-
bustion engine water vapor output.
Successful implementation of the first strategy would require an effec-
tive mass transit network and/or use of electric powered vehicles. Such
traffic modifications were not included in this study. The second strategy
would require some means of dehydrating vehicle exhaust. This study has shown
that cooler-condensers can eliminate up to 80 percent of the ice fog caused by
vehicular emissions.
A research effort encompassing two winters was directed toward finding
and evaluating methods to reduce automobile-generated ice fog. Four methods
are discussed: (1) allow ice fog particles to form, then capture them with
particulate vtraps; (2) capture the water vapor with a dessicant; (3) remove
the water vapor as a liquid condensate by use of cooler-condensers; and (4)
warm the ambient air with an exhaust dilutor-air heater thus allowing it to
accept the dispersed exhaust water vapor without immediately forming ice
fog.The first two methods would require such cumbersome equipment that they
were considered impractical. Only methods 3 and 4 were evaluated.
Cooler-condensers, devices for removing water vapor from gases, were
evaluated for their ice fog control possibilities. These cooler-condensers
are heat exchangers in which the exhaust gas is cooled by cold ambient air or
coolant from the automobile's radiator.
The first winter's effort was spent adapting, modifying and attaching
existing heat exchangers to vehicle exhaust systems. With the resulting field
performance data, the overall heat transfer coefficients were calculated.
Also, independently, overall heat transfer coefficients were estimated from
engineering data books. These latter coefficients resulted in required heat
transfer surface areas much larger than those indicated by the first winter's
-------
field data. The second winter's experience indicated that the actual exchan-
ger surface area required is closer to that calculated from coefficients
estimated from the engineering data book. Therefore, it was found that for
effective ice fog removal from a standard size gasoline-fueled light duty
vehicle with a 250 cubic inch displacement engine a cooler-condenser would
need 1.9 square meters (21 square feet) of heat transfer surface area.
It was found that ice film formation limited the performance of unpro-
tected bare tube, air cooled, cooler-condensers with 1 cm (1/2 in.) tubes.
Because of ice plugging problems, smaller tube sizes are not recommended.
The cooler-condenser should be designed for efficient condensate drainage.
Weep holes (at low points in the cooler-condenser system) were required to
drain condensed water pockets which would otherwise freeze and block the
outlet manifold.
Using a baffled automotive radiator as a cooler-condenser with antifreeze
coolant might reduce the icing problem and make the lowest cost cooler-
condenser. Because of increased engine vulnerability and unknown (first
winter) radiator performance, they were not as thoroughly researched as the
air cooled cooler-condensers.
Location of weep holes and the final exhaust outlet is critical, because
air currents around the vehicle may cause exhaust gases to enter the passenger
compartment and raise carbon monoxide levels.
With the cooler-condensers only, much of the resultant condensed water
formed a fine mist; therefore, use of coalescers was necessary to transform
the mist into liquid water.
One cooler-condenser was installed on a diesel sedan, but it did not
perform as well as those on gasoline engines. Also, its coalescer became
plugged with soot. Another type of ice fog control device evaluated was a
perforated spiral wound flexible metal exhaust hose coiled under a vehicle
and attached to its exhaust pipe. Its function was to serve as an exhaust
dilutor-ambient air heater. It significantly reduced visible ice fog emission
but raised the passenger compartment carbon monoxide levels. However, it was
not as effective in reducing ice fog as were the cooler-condensers.
To quantify the actual on-the-road performance of a cooler-condenser, an
equipped vehicle was driven through urban Fairbanks under normal ^inter driv-
ing conditions. The overall water vapor condensed (ice fog removed) was 81
percent at an ambient temperature of -16°C (4°F).
The increased exhaust system back pressure due to cooler-condensers was
found to have an insignificant effect on fuel economy.
The authors estimate that mobile sources are responsible for about fifty
percent of the Fairbanks ground level ice fog at temperatures below about
-40°C, if cooling pond emissions are excluded. Ice fog control to yield 80
percent reduction on all vehicles in the Fairbanks area would therefore
reduce ground level ice fog emissions by about 40 percent. Because of in-
creased stationary source emission at lower ambient temperatures the overall
-------
ice fog reduction would be less, but the ice fog would not be as dense over
roadways. The 40 percent reduction in ice fog emissions would theoretically
yield a 67 percent increase in visibility.
The condensation process would also tend to remove toxic exhaust products
such as sulfur oxides and lead compounds. The sulfur oxide removals varied
from 1 to 20 percent. The lead removals varied from 6 to 49 percent of that
in the gasoline. At 85 percent water vapor removal the condensate to be
disposed of amounts to 0.8 gallon per gallon of gasoline burned.
If not captured by the automobile, the condensate ends up as ice on the
road. This additional ice amounts to about 20 percent of the normal snowfall
in Fairbanks, Alaska for the four coldest winter months. It would probably
require more effort by the road maintenance crews to keep it from accumulating
at some intersections.
This limited research study has shown that cooler-condensers can be
effective in limiting ice fog from automobiles and trucks. fjowever, the
problem of controlling condensate freezing in the cooler-condensers was not
thoroughly investigated. More experience is needed before possible regulatory
action can be considered. A further evaluation and demonstration of the
devices on fleet vehicles used in the dense ice fog areas should be carried
out.
-------
SECTION 2
RECOMMENDATIONS
Ice fog is a form of air pollution that becomes a problem when tempera-
tures drop below -18°C (0°F). It is estimated that motor vehicle exhaust
contributes nearly half the ground level ice fog.
Exhaust gas dehydration by use of water-condensers can effectively
eliminate 75 percent of the automotive generated ice fog. If possible regula-
tory action is to be considered, then the next step would be to evaluate long
term cooler-condenser operation and maintenance problems. This could best be
accomplished by using the information supplied in this report to design and
attach prototypes to 20 or more fleet vehicles that are routinely operated in
dense ice fog areas. That evaluation could be used to determine which coolant
works best and the degree of temperature controls needed to prevent cooler-
condenser freeze ups. The result could also be used to derive a cost-benefit
ratio for automotive ice fog control.
Because of the possibility of carbon monoxide poisoning, undercarriage
exhaust discharge is not recommended for passenger vehicles.
Coalescers were necessary to eliminate mist emissions and are recommended
for each vehicle equipped with a cooler-condenser.
-------
SECTION 3
INTRODUCTION
BACKGROUND
Alaska is the largest, most sparsely populated, least industrialized
state in the nation. Yet its major cities, Fairbanks and Anchorage, have
winter time air pollution levels which rival those of New York and Los An-
geles. The air quality of these Alaskan cities is degraded mainly by three
types of pollutants: ice fog, other particulates, and carbon monoxide (1).
The toxic health effects, if any, of ice fog have not been documented. Ice
fog is most severe in Fairbanks but is increasing in Anchorage. This study
deals with controls for ice fog from mobile sources. Ice fog air pollution is
unique to regions with extremely cold climates. The nature of ice fog has
been well defined (2). The main objection to this cold weather phenomenon is
that it severely restricts visibility during abnormally difficult driving
conditions. It limits commerce by closing airports and increasing automobile
traffic accident rates. During ice fog conditions there are often thermal
inversions which trap the fog near the ground.
Ice fog is a winter phenomenon typical of inhabited Arctic regions. It
is composed of minute ice crystals that are produced when water vapor is
released in ambient air that is too cold to hold it in solution. The water
vapor separates into a liquid or solid phase. If cold enough, it will soli-
dify into very small ice crystals which seem to hang in the air.
As the urban population in Alaska has increased, this fog has created
serious problems for the people who attempt to live comfortably in this cli-
mate. Ice fog, capped by atmospheric thermal inversions, is known to increase
the ambient levels of other pollutants such as lead compounds and toxic
gases, including nitrogen and sulfur oxides, aldehydes, and halogenic acids.
Carbon monoxide is a major air pollutant in Fairbanks, but it is not
directly related to ice fog. The higher levels of carbon monoxide are caused
by thermal inversions that start at ground level. However, ice fog is caused
by low temperatures not thermal inversions. When dense ice fog is present
the thermal inversions are usually the strongest at the top of the ice fog
layer. Therefore, because of the larger dilution volume under the inversion,
the higher levels of carbon monoxide are not necessarily present during ice
fog. Carbon monoxide is a known health hazard; the others are potential
health hazards. These pollutants in the Fairbanks area have been measured by
other investigators (1, 3, 4).
-------
There are three major sources of ice fog. Ranked in decreasing order of
their vision obscuring effect to the subarctic city resident, they are:
1. automobile and truck exhausts,
2. open water surfaces ( such as cooling ponds), and
3. exhaust gases from heating and electrical power plants.
The relative impact of each source depends upon the individual point of
view. To persons downwind from a cooling pond, the pond appears as the most
important source. But, when motoring in heavy traffic, automobiles appear to
be the most important source.
In the combustion sources 1 and 3 above, the water vapor is created by
the oxidation of the hydrogen in the hydrocarbon fuels (gasoline, fuel oil,
and coal). In the case of soft coal, much of the water vapor comes from
hydrocarbon oxygenates and trapped moisture.
During winter, waste heat from power plants prevents total freeze over
of the Chena River and the Fort Wainwright cooling pond. These open waters
yield considerable ice fog due to high evaporation rates.
The ice fog created by home furnaces is usually injected into the atmos-
phere at heights of 3 to 5 meters above ground. This contribution to reduced
visibility is omnipresent rather than concentrated at any one locale. Ice
fog particles in the larger coal-fired power plant plumes tend to increase
their overall density, thus dragging some of the toxic combustion products
into the lower air layers as they settle. Because thermal inversions inhibit
vertical mixing they, at times, limit some of this plume fallback into the
immediate area. The extremely stable air in thermal inversions also severely
limits plume dilution by dispersion. However, if the ice fog layer is deeper
than the plume height, the plume is usually trapped in the ice fog thus
raising the ground level sulfur dioxide concentration. To the work-a-day
commuter the most significant source of ice fog is automobiles and trucks.
Vehicles emit ice fog along the road network and the result is greatly reduced
driver visibility. The ice fog is usually much denser at intersections.
This reduced visibility in turn forces the operators to drive so slowly that
increased fuel consumption and more ice fog result. The only compensating
effect is that most people try to limit their driving during the extreme cold
periods when ice fog is very dense.
During the four coldest months consumption of petroleum products in the
Fairbanks area is about half fuel oil and half gasoline. Therefore, the ice
fog contributions from heating and mobile sources are roughly equal. Together
with cooling ponds they comprise most of the ground level ice fog problem.
Ice fog is not as much of a problem in the other 49 states. However,
this research performed in the Fairbanks, Alaska area is applicable to any
cold region where water vapor emissions are a problem.
In recognition of the severity of the ice fog problem, the U.S. Environ-
mental Protection Agency's (EPA) Arctic Environmental Research Station (AERS)
decided to apply the resources ot its Technology Research Branch. Others
-------
have done an admirable job in defining the problem (2). But what is needed
now is the development and demonstration of some effective, and low-cost,
control hardware.
Before discussing control methods, it is interesting to look at the
relative ice fog emissions for the various fuel types. These emissions are
listed in Table 1, which is extrapolated from reference (5).
Table 1. EFFECT OF FUEL ECONOMY ON AUTOMOTIVE ICE FOG (H20) EMISSIONS
FUEL
Diesel (Fuel Oil)
Gasoline (Small Vehicle
Gasoline (Standard vehi
Propane
ASSUMED
MILEAGE
)
cle)
km
1 iter
11
11
6.8
4.7
mi
gal
(26)
(26)
(16)
(11)
RESULTANT q H?0
EMISSIONS km
90
79
130
180
,oz H20,
^ mi J
(5.1)
(4.5)
(7.4)
(10)
The three most common types of automotive fuels used in the Fairbanks
area are propane and gasoline for the spark ignition engine and fuel oil for
the diesel engine. The ice fog emission is the water vapor emission in grams
per kilometer (ounces per mile). Note that the emissions are directly related
to the fuel economy. In generating this table a gasoline fuel economy of 6.8
kilometers per liter (16 mi/gal) for a standard size automobile was assumed.
For propane the same motive energy requirement (Joules/mi) was used resulting
in 4.7 km/1 (11 mi/gal). In the above case propane would emit the most ice
fog, and diesel the least. Initially, the diesel powered vehicle was assumed
to be 48 percent more efficient than gasoline. Recently some of the newer,
smaller gasoline vehicles have mileages comparable to the larger diesels--
11 km/1 (26 mi/gal). In those cases, the gasoline vehicle would emit less
ice fog than the diesel. This is because fuel oil contains more hydrogen per
gallon than does gasoline. Fuel economy as it relates to gross vehicle
weight or passenger comfort is not considered here.
In comparing emissions for any one fuel type, the water vapor (ice fog
emission) is directly related to fuel economy. For example, a vehicle yield-
ing 17 km/1 (40 mi/gal) will emit only half as much ice fog as a vehicle
yielding 8.5 km/1 (20 mi/gal) and one-quarter as much ice fog as a vehicle
attaining only 4,3 km/1 (10 mi/gal).
-------
SCOPE
There are several alternate ways to reduce automotive created ice fog
without applying controls on individual vehicles. These include pooling, buses
and/or electric vehicles. If any mixed application of the three above methods
would result in fewer hydrocarbon powered vehicles on the streets, then the
automotive-generated, on-the-road ice fog would be reduced proportionally.
However, more electric power plant ice fog would result if electric powered
cars come into general use. Electric automobiles use gas heaters to keep the
occupants warm. Although these heaters emit ice fog the amount is much less
than for gasoline powered automobiles, and it can be controlled by methods
similar to those used for other automobile ice fog problems.
The AERS research effort on automotive ice fog control is the subject of
this paper.
One method of controlling ice fog would be to allow the ice fog to form at
a certain distance from the tail pipe, then trap it as particles. These partic-
les could be removed with large filter assemblies such as those used in heating
ventilating ducts to remove dust from air. Research has shown that in some
cases electrostatic precipitation will work (2). Electrostatic precipitators
are common fly ash control -devices on large coal-burning power plants. The
major problem with these methods would be the requirement for equipment to mix
the exhaust gas with the cold air to first form the ice fog and then to capture
it on filters or electrostatic precipitators. The equipment would have to be
sized to handle more than 28 cubic meters per minute (1000 CFM) at a pressure
drop of less than 0.25 cm (0.1 in) water column. In this case, the filtration
plenum would probably be larger than the vehicle which created the exhaust.
Therefore, this method would not seem to be too practical.
It is easier to limit water vapor emission than to attempt to clean up the
resultant ice fog after it has formed. Two major methods of removing vapor
from a gas stream are: (1) using a hydroscopic media (desiccant) such as
glycol to absorb the water, or (2) cooling the gas stream below the dew point
to condense out part of the water vapor. The desiccant, which may also be in e
solid form (such as silica gel), is usually more effective in gas drying (dehy-
dration) since dew points down to -40°C (-40°F) are attainable. The lower the
dew point, the dryer the gas. However, the desiccant requires a contactor to
absorb the water and a stripper to remove water (regenerate) before reuse.
Also, exhaust contaminations such as lead compounds, soot, and mineral acids
may contaminate and/or coat the desiccant.
To avoid the above complexity it was decided to concentrate on a cooler-
condenser type of heat exchanger and to control the exhaust gas outlet tempera-
ture to a range of 1.7 to 4°C (35 to 40°F) to prevent freezing. Even at this
high an outlet temperature, over 94 percent of the water vapor is condensed
out. Figure 1 depicts three curves showing the percent of water vapor con-
densed for the various automotive fuel types. The calculations for these curves
are in Appendix B.
-------
100
80
O
LU
oo
UJ
Q
o
o
o
O-
o:
60
40
UJ
a.
20
Dew point
DIESEL
230 Percent
excess air
GASOLINE
0 Percent
excess air
PROPANE
0 Percent
excess air
I
-7(20)
I
I
4(40)
16(60) 27(80)
TEMPERATURE °C (°F)
I
38(100;
I
49(120)
Figure 1. Condensation curves for three automotive fuels.
9
-------
Because these heat exchangers cool' the exhaust and condense the water
vapor they are called cooler-condensers. This technique has proven successful
for ice fog control from oil-fired bofler stacks (6). In the past there have
been private innovators who have assembled cooler-condensers and used them
successfully. The U. S. Army Cold Regions Research Engineering Laboratory is
the only organization which has reported on its device (7).
Because of high local interest in the automotive ice fog problem, it was
decided to support several local research contracts and to have a limited
inhouse (AERS) effort during 1975. These efforts resulted in the design,
construction, installation, and demonstration of nine different ice fog
cooler-condensers on nine different automobiles and light duty trucks. Four
were constructed by AERS and five were constructed by local engineering
contractors.
10
-------
SECTION 4
DESCRIPTION OF ICE FOG CONTROL DEVICE INSTALLATIONS
TEST METHODS AND INSTRUMENTS
Methods
In this research effort it was decided to build and install the cooler-
condensers and evaluate them by road testing. Ice fog removal efficiency is
assumed to be the same as water vapor removal efficiency which is the percent
of water vapor condensed in Figure 1. Therefore, the cooler-condenser effi-
ciency could be determined by simply measuring outlet temperature.
The road testing was performed at constant speeds of 32 and 64 km/h (20
and 40 mi/h). It was intended that the cooler-condensers be sized mainly to
control ice fog in the urban-suburban areas since that is where the problem
is concentrated. In these areas the speed limits are generally limited to
less than 64 km/h (40 mi/h). To control ice fog emissions at higher speeds
would require much larger cooler-condensers because the exhaust heat to be
removed increases exponentially with vehicle speed.
Since no one drives in urban areas at a constant speed, cooler-condenser
performance during a drive through town was also evaluated.
Dial Thermometers and Thermocouples
Temperature data were used to evaluate the heat exchanger (cooler-
condenser) performance. Reliable temperature data were easily obtained by
installing thermocouples in the exhaust gas stream, both upstream and down-
stream from the heat exchangers. For the higher inlet temperatures a hole
was drilled at the exhanger inlet and a chromel-alumel thermocouple was
inserted. To measure the low outlet temperatures an iron-constantan, ther-
mocouple was inserted into the coalescer. Lead wires were run into the cab
compartment and connected to a pair of Leeds and Northrup dial potentiometers.
To measure back pressure a hole was drilled in the exhaust pipe between
the engine and the heat exchanger. A 1/8 inch N.P.S. pipe nipple was welded
in with a plastic tube running into the cab to connect to a Bourdon tube
pressure gauge calibrated in inches of water.
11
-------
Ambient air temperatures were measured with a dial thermometer secured
through an open window. The stem extended at least 15 cm (6 in.) from the
body of the vehicle. In some cases the National Weather Service temperatures
were used since their thermometers were within a few miles of the test road.
Each interval lasted between one and two minutes to allow the changing temper-
atures to stabilize with respect to the thermal-kinetics of the heat exchange
system.
Readings were first recorded in a log book. Later it became more feasi-
ble to use a portable tape recorder in the vehicle and transcribe the data in
the office.
All measurements were in English units which in this report are in
parenthesis following their equivalent metric units. Therefore, it should be
realized that both 0°F and 1°F are -18°C and nominal units such as 1/2 inch
EMT for example are rounded off to 1 cm EMT.
Fuel Flow Rate Meters
Fuel consumption tests were performed to determine the effect of in-
creased back pressure upon fuel economy. These tests were conducted with the
heat exchanger connected and the normal exhaust plugged. The test was then
repeated with the heat exchanger plugged and normal exhaust unplugged. A
constant speed of 80 km/h (50 mi/h) was maintained over a 33 km (20 mi) level
section of the Richardson Highway. A stop watch was used to determine exact
speed from mileage as indicated by an odometer. Exhaust back pressure was
monitored and found to be constant on the level stretches.
Two fuel meters were installed in the gas line in series to check com-
parative precision of each. A Columbia system meter was first, followed by a
Kent-Moore meter. The Columbia systems meter recorded only half the fuel
actually consumed and its figures were disregarded. It was later found to
have excessive internal wear.
12
-------
HOT
EXHAUST
GAS
1
n
i
SHELL SIDE-,
1
HEAT EXHANGER DEFINITION
A heat exchanger is a device which allows heat to be passed from one
fluid to another without permitting the fluids to mix. A metal wall tube
usually contains one of the fluids and allows only heat to pass through.
Examples of heat exchangers are automobile radiators, refrigerator condenser
coils, and combustion chambers of some furnaces. A tube and shell heat ex-
changer can be diagramed as such:
AIR
COOLED
EXHAUST
GAS
—»—
INCLUDING I /- , J INCLUDING
WATER VAPOUR / IfJ LIQUID HATER
TUBE SIDE | (CONDEMSATE)
COLD
AIR
Heat is transferred from the hot fluid, tube side in this case, through
the tube wall to the shell side fluid. If there is considerable film resis-
tance to heat transfer, the tube surface may be extended by the use of fins.
The fins are on the side with the highest film resistance. If the fluids are
pumped through the exchanger then it is called forced convection; if not, it
is free convection. The amount of heat (BTU or Kcal) that is transferred is
called the duty of the heat exchanger.
The automobile radiator is a forced convection shell-less heat exchanger
in which hot antifreeze solution is pumped through the externally finned
tubes. For a discussion of heat exchanger sizing techniques, see Appendix C.
Twelve of the thirteen ice fog control devices investigated here were
cooler-condensers, which are heat exchangers that could cool a vehicle's ex-
haust enough to condense out the combustion created water vapor. This vapor,
when mixed with air at sub-zero temperatures, becomes ice fog. The other
control technique was an exhaust di1utor-ambient air heater. This discussion
describes thirteen such devices which were installed on nine vehicles. Eight
of the control devices were constructed and/or modified by the authors. Two
different types were evaluated on each of the following four vehicles equipped
with the following engine cubic inch displacement (CID):
(1) Metal coil flex hose cooler-condenser, first, and combination tube
cooler-condenser second, on a 1968 Chevrolet Carryall (4X2)
(utility vehicle), 250 CID 6 cylinder.
(2) Brazed radiator cooler-condenser first, and exhaust dilutor-
ambient air heater second, on a 1971 CMC Jimmy (4X4) (utility
vehicle), 250 CID 6 cylinder.
13
-------
(3) Brazed radiator cooler-condenser first, and single pass stainless
steel cooler-condenser second, on a 1974 Chevrolet Nova (Sedan),
250 CID 6 cylinder.
(4) Modified finned oil cooler cooler-condenser first, and a four
pass stainless steel cooler-condenser second, on a 1967 Mercedes
Benz (D-200), 135 CID 4 cylinder.
The other five installations were performed under contract by the follow-
ing local engineering outfits. They were:
(5) Fan tube cooler-condenser on an Arctic Studies Group 1970 Volvo
144S 4 cylinder.
(6) Finned copper tubing cooler-condenser on an AE Research, Inc.
1974 Datsun, B-210, 1300 cc 4 cylinder.
(7) Liquid cooled, cooler-condenser on a H and S Research 1968 Jeep
Wagoneer, 6 cylinder.
(8) Finned pipe cooler-condenser on a Scarborough & Associates 1968
Chevrolet Carryall (4X4), 307 CID V8 cylinder.
(9) Louvered shell cooler-condenser on a Simplex-Standard 1968 IHC
Scout, 266 CID V8 cylinder.
All installations were designed to demonstrate the effectiveness and
practicability of an ice fog removal device on an in-service automobile.
All contractors were issued guidelines by the USEPA Arctic Environmental
Research Station which specified certain criteria for the scope of their
project. Each contractor indicated what type of vehicle they had access to,
its engine displacement, etc. In November, 1974, the contracts were awarded.
By March 31, 1975, the demonstration of an ice fog removal device was required
on an in-service vehicle. By May 1, 1975, a written report covering design,
material cost, installation time, data obtained, and the practicality and
anticipated service life of the unit was required. Photographic evidence of
the system at work under sub-zero weather conditions was also required.
A short description of each device and its apparent performance follows.
A more detailed description of each device is in Appendix A.
DEVICES EVALUATED BY THE ARCTIC ENVIRONMENTAL RESEARCH STATION
Metal Coil Flexhose Cooler-Condenser
The first attempts to find a solution to the Fairbanks ice fog problem
were performed on a 1968 Chevrolet 1/2 ton Carryall (4X2). Its muffler was
removed and 46 m (150 ft.) of flexible exhaust hose were coiled beneath the
vehicle. Flexhose was chosen because it was easy to install and would not
present any back pressure problems. The system proved that the exhaust tem-
perature could be reduced to a low enough point so little or no visible ice
fog was being emitted from the vehicle. Even though the hose appeared to be
filled with frost there was no noticeable power reduction from back pressure.
The system was only temporary and corrosion weakened the coils after two
winters' use.
14
-------
Combination Tube Cooler-Condenser
In the winter of 1975-76 the metal coil flexhose was removed and two
conventional steel mufflers were mounted in series in the exhaust pipe, A
cooler-condenser fabricated from 1 cm (1/2 in.) electrical metallic tubing
was mounted between the radiator and fan. In conjunction, a section of
flexhose was mounted in front of the radiator in a switch back configuration.
The exhaust was directed through the mufflers, through the flexhose then into
the cooler-condenser. The outlet was extended to the center of the vehicle
and condensate distributed against the vehicle's drive shaft. There was no
visible ice fog at any temperature at different speeds. Although this setup
was over-sized intentionally, it demonstrated that all visible traces of ice
fog could be eliminated. There were drawbacks such as excessive build up of
condensate during long periods of idling which, if not drained out (while
still in the liquid state), would freeze the entire system closed, preventing
the engine from operating.
Brazed Radiator Cooler-Condenser
Late in 1974 the 1971 GMC Jimmy (4X4) was the first vehicle on which a
compact cooler-condenser installation was attempted. Its first cooler-con-
denser was a small radiator in which solder joints were replaced with brazing
alloy It was connected in lieu of the muffler It was mounted between the
drive shaft and the frame. It was quite successful at idle; but exhaust tem-
peratures above 1000°F (approx. 540°C) melted joints at the inlet, header,
creating excessive noise and high temperatures on the floorboards.
Free Convection-Finned Oil Cooler Cooler-Condenser
Next, early in 1975, a cooler-condenser was assembled from ten 100 cm
(40 in.) lengths of 1 cm (1/2 in.) electrical metallic tubing (EMT). It
lacked sufficient surface, so one half of a mobile oil cooler (Young Radiator
Company MOC #6) was mounted in series. Both heat exchangers were mounted
under the Jimmy between the drive shaft and frame channels; the EMT on the
passenger side, and the one half MOC #6 on the driver's side. At an ambient
temperature of -18°C (-1°F) this assembly would condense out 55 percent of
the water vapor at idle and none at 64 km/h (40 mi/h).
Exhaust Dilutor Ambient Air Heater
Because of the relative ineffectiveness of the large surface heat ex-
changer mounted under the Jimmy, it was decided to try a different approach
during the winter of 1975-76. This approach is based upon the principle that
heating cold ambient air will increase its ability to accept water vapor
without forming ice fog. The hardware involved consisted of 8 m (26 ft.) of
5 cm (2 in.) perforated spiral wound flexible metal hose connected at: the
tail pipe. This hose was wired to the frame channels, behind the transfer
case and under the rear bumper. It functioned as a heat exchanger and ex-
haust gas distributor first by heating the ambient, air so it could take more
water vapor into solution, thereby dispersing the moist exhaust gas into the
heated air. The cooler section of the metal hose also condensed some of the
exhaust water vapor. With this setup the trail of visible exhaust was about
half as long as without.
15
-------
Because most of the exhaust was released under the floorboards, some
carbon monoxide leaked through. Therefore, this system is not recommended
when occupants must ride in the contaminated airspace. However, it would be
quite satisfactory under the open cargo bed of a truck.
Brazed Radiator Cooler-Condenser
Another small radiator was prepared for high temperatures by brazing its
seams. It was then mounted between the radiator and the grill on the 1974
Chevy Nova Sedan where the ambient air would flow over the cooling fins. In
conjunction, a coalescer was fabricated out of a 61 cm (2 ft) length o-f 10 cm
(4 in) stove pipe. An 8 cm (3 in) thick plug of furnace air filter was
placed inside the pipe and the outlet exhaust directed to flow through it.
Minute drops of water mist impinged on the fiberglass which caused them to
coalesce and run off in a liquid stream. Condensate freezing in the radiator
tubes and resultant excessive back pressure was the main problem with this
application. However, at -29°C (-20°F) visible ice fog was negligible.
Single Pass Stainless Steel Cooler-Condenser
For the winter of 1975-76 a heat exchanger was built from 1 cm (1/2 in.)
stainless steel tubing and mounted on the front of the Nova between the
radiator and grill. Metal flexhose was used to pipe the exhaust into and out
of the exchanger. A coalescer was mounted on the rear bumper and thermo-
couples inserted at the inlet and the outlet. Approximately 85 percent of
the exhaust water vapor was condensed at temperatures below -25°C (-13°F).
The addition of chains inside each tube increased internal surface area and
acted as gas turbulators but did not significantly increase back pressure on
the engine.
Modified Finned Oil Cooler, Cooler-Condenser
During the first winter (1974-75), a mobile oil cooler was mounted in
front of the radiator on a 1967 Mercedes Benz Sedan for evaluation with a
diesel engine. The exhaust was piped directly from the engine's manifold.
Excessive back pressure required the removal of internal tubulators from the
oil cooler. Visible ice fog diminished within a foot of the vehicle at
temperatures below -18°C (0°F). One problem was the smell of exhaust fumes
in the cab compartment. This could be eliminated by ensuring a leak proof
mounting and extending the cooler-condenser outlet past the passenger com-
partment.
Four Pass Stainless Steel Cooler-Condenser
For the winter of 1975-76, a cooler-condenser with 1 cm (1/2 in.) stain-
less steel tubes was built and installed in place of the mobile oil cooler on
the Mercedes diesel. It was designed from calculations using test data from
the previous season. The problem of excessive back pressure on a diesel en-
gine was taken into account but internal chains in the last tube pass presen-
ted too much of an obstruction for the large volume of excess air required by
a diesel engine. This is not a problem for cooler-condensers used on gaso-
16
-------
line engines. This stainless steel cooler-condenser effectively removed 70
percent of any visible ice fog. Under heavy acceleration a visible mist
plume was forced out through the coalescer. However, the mist was too heavy
to remain in the atmosphere and was not considered to be troublesome ice fog.
COOLER-CONDENSERS EVALUATED BY PRIVATE CONTRACTORS
Fan Tube Cooler-Condenser
Students in the University of Alaska mechanical engineering program de-
signed and fabricated an air cooled cooler-condenser for a 1970 Volvo Sedan
using 16 pieces of 1 1/2 in. X 15 in. EMT tubes enclosed in a rectangular
sheet metal box (8). The exhaust gases circulated around the outside of the
tubes. The device was mounted on the rear bumper. Exhaust heat was removed
by cold ambient air drawn through the tubes to further reduce its relative
humidity. This device prevented visible ice fog from being released into the
atmosphere. Material costs were quoted at approximately $50.00 with an an-
ticipated life of no more than three seasons.
Finned Copper Tubing Cooler-Condenser
Since the engine displacement of a 1974 Datsun Sedan was small, a mani-
fold style cooler-condenser was fabricated from the standard copper tubing,
aluminum-finned baseboard heater pipe (9). Three sections, each 91 cm (3 ft.)
long, were connected in parallel and mounted under the rear of this vehicle
at a slight angle to drain condensate. At idle some aerosol fog was visible.
At higher speeds the greater fuel consumption increased the heat exchanger
load and visible ice fog increased accordingly. To function more effectively,
increased surface would be required.
Liquid Cooled Cooler-Condenser
A liquid cooled condenser rather than an air cooled type was designed by
H & S Researchers. It consisted of an enclosed radiator mounted on the front
of a 1968 Jeep Wagoneer (10). The coolant antifreeze solution was connected
in series with the vehicle's engine cooling system. Exhaust gases were
baffled back and forth through a radiator encased in a sheet metal box. The
outgoing exhaust was directed at the vehicle's radiator to reevaporate mist
droplets not removed in the condenser. Although visible ice fog was reduced
to a minimum, this last step was undesirable since carbon monixide gas was
drawn into the vehicle's cab. Safety being prerequisite, the unit was relo-
cated to the rear of the vehicle. This made it inconvenient to use the
engine's cooling system. A second radiator was incorporated using a 12 volt
DC pump to circulate the antifreeze. Warm weather necessitated testing in a
cold cell laboratory. No fog was visible from the mock-up.
Finned Pipe Cooler-Condenser
Scarborough & Associates built an air cooled cooler-condenser using a
1.3 m (4 ft.) section of 5 cm (2 in.) aluminuim pipe with large 15 cm X 15 cm
(6X6 in.) fins attached to the outside (11). A spiral strip of steel
17
-------
inserted inside acted as a gas turbulator. It was mounted in series with the
muffler under a 1968 Chevy Carryall (4X4). Effectiveness was limited because
of inadequate heat transfer surface area.
Louvered Shell Cooler-condenser
Using no welding or machining of parts, a cooler-condenser was fabrica-
ted by Simplex-Standard Company using 1/2 in EMT encased in a sheet metal
housing (12). This simplified method of construction demonstrated that a
cooler-condenser could be built with limited tools. The unit was designed so
that header covers could be removed for inspection and cleaning. It was
mounted on the front bumper of a 1968 IHC Scout and fed by a section of flex
exhaust hose extending from the tail pipe. Air flowing across the tubes
could be externally controlled by a cable attached to adjustable louvers in
front of the tubes. The unit reduced a substantial amount of visible water
vapor at idle and was quite effective at higher speed.
18
-------
0
SECTION 5
DEVICE PERFORMANCE AND COMPARISON
When considering the cooler-condenser type of ice fog control device the
only performance criterion is the fraction of exhaust gas water vapor conden-
sed (removed). The percent of the water vapor condensed is directly related
to the outlet temperature (assuming constant pressure) as shown in Figures 1
and 2. The calculations are in Appendix B. Therefore, the cooler-condenser
(.heat exchanger) performance can be measured in terms of the one most impor-
tant parameter, outlet temperature. Efforts were directed toward reducing
the temperature enough to condense out 95 percent of the water vapor. From
Figure 1 that temperature must be 4°C (40°F) or less for gasoline. Because
of excessive ice formation it is not practical to condense more than 85
percent of the moisture in diesel engine exhaust.
FIRST WINTER RESULTS
The results from the first winter's (1974-75) installations will be dis-
cussed first. The coiled flexhose under the 1968 Chevy Carryall (4x2) will
not be compared with the other cooler-condensers because it was too large to
tie considered practical and it clogged easily with frost.
The comparative performances of the remaining eight, devices are tabula-
ted in Table 2. Listed with each cooler-condenser is the vehicle on which it
was installed.
In all the cooler-condensers except the one on the Jeep the exhaust heat
was transferred directly to ambient air Free air convection means that the
cooler-condenser was mounted parallel to the air stream or such that one tube
shaded the rest from the air movement. Forced convection means that air was
forced across the tubes either by vehicle movement or a fan. The forced
convection condensers had to be mounted in front of the vehicle of have a fan
blowing air across the tubes.
The 1968 Chevv Carryall (4x4) belonged to a contractor and had only one
finned tube hanging below the vehicle, so it was called a forced convection
cooler-condenser
The percent water condensed (Columns 1 and 2) was obtained by measuring
the outlet temperature then reading the percent condensed off of Figure V.
Sixty-four km h (40 mi hi was chosen as the upper reasonable speed limit for
siring the cooler-condenser duty since urban traffic seldom exceeds that
speed during ice fog conditions.
-------
oo
h-
CJ
rs
Q
o
o:
•=t
X
-o
OJ
03
-M
C
o;
o
OJ
Q.
en o
i—. j
LlJ
3 CO
OQ
O
D-
O LLJ
CL. ^
"^^ i—t
Q _l
z: o
rD oo
o <
Q- CD
100
80
60
40
20
0
30
29
28
15.5,
15.0.
14.5.
1000-
500.
DEW POINT
•3000
.2000
00
LU
00
CO
h-
00
ro
cn
CQ
<
o
1000
I I I I I
-18(0) 38(100) 93(200) 149(300) 204(400) 260(500)
Temperature °C(°F)
Figure 2. Exhaust gas heat content, Kcal/Kgram of gasoline (C7H13) consumed.
Exhaust gas vapor mole weight and mass per mass gasoline.
Percent water vapor condensed vs. temperature.
20
-------
TABLE 2. FIRST WINTER PERFORMANCE OF PROTOTYPE AUTOMOTIVE ICE FOG CONTROL DEVICES
r-o
Percent exhaust H20
Type of
Cooler-
Condenser
and
(Vehicle)
Finned Tube,
free convection
(1971 GMC Jimmy 4x4)
Brazed Radiator
forced convection
(1967 Chevy Nova Sedan)
Modified finned oil
cooler, forced convection
(1967 Mercedes Benz D200)
Fan Tube, forced
convection
(1970 Volvo Sedan 144S)
Finned Copper Tubing,
mixed convection
(1974 Datsun B-210 Sedan)
Louvered She! 1 ,
forced Convection
(1969 IH Scout 4x4)
Finned Pipe,
forced convection
(1968 Chevy Carryall 4x4)
Liquid Cooled,
forced convection
(1 968 Jeep Wagoneer)
Vapor Condensed
and Ambient
at Idle
55%
-18°C
(- 1°F)
95%
-19°C
(- 3°F)
25%
-24°C
(-12°F)
98%
-18°C
(- 1°F)
97%
-17°C
( 1°F)
90%
-17°C
( 0°F)
47%
-20°C
(- 5°F)
47%
-12°C
( 10°F)
Temperature
at 64 km/h
(40 mi/h)
0%
-18°C
(- 1°F)
83%
-18°C
(- 1°F)
0%
-24°C
(-12°F)
No Data
0%
-14°C
( 6°F)
88%
-17°C
( 0°F)
0%
-20°C
(- 5°F)
No Data
Heat
Transfer
Surface-
Hot-
Gas Side
1 .1 m2
(11.8 ft2)
0.72 m2
(7.8 ft2)
0.21 m2
(2.3 ft2)
1.0 m2
(11 ft2)
0.33 m2
(3.6 ft2)
0.93 m2
(10 ft2)
0.25 m2)
(2.7 ft2)
1.75 m2
(18.9 ft2)
Heat
Transfer
Coefficient
kcal/m2-h-°C
(BTU/h-ft2-°F)
43
( 8.9)
100
( 21)
127
( 26)
No Road Test
54
(11)
73
(15)
42
( 8.6)
Depends
Required
Heat
Transfer
Surface
at -29°C
(-20°F)
2.1 m2
(23 ft2)
0.78 m2
(8.4 ft2)
0.50 m2
(5.4 ft2)
Data
0.79 m2
(8.6 ft2)
0.92 m2
(9.9 ft2)
0.48 m2
(5.2 ft2)
on
Exhaust
System
Ratio of Back
Finned to Pressure
Tube Area at 64 km/h
(40 mi/h)
5:1 56 cm
of water
5:1 106 cm
of water
8:1 105 cm
of water
1:1 N11
13:1 Nil
1:1 35 psig
(see text)
15:1 Nil
5:1 Nil
Antifreeze temperature
-------
Mathematics detailing the method for calculating the overall heat trans-
fer coefficient (Column 4) are presented in Appendix C. This coefficient is
the transferred heat rate (duty) from Figure 2 divided by the temperature
difference (log mean) and the surface area (inside hot gas side, Column 3).
The coefficient was calculated from the ambient temperatures listed in Col-
umns 1 and 2 and the inlet and outlet temperatures. The surface required to
condense out 95 percent of the water vapor at -29°C (-20°F) (Column 5) am-
bient was then calculated using the above heat transfer coefficient (Column
4). All calculations were based upon the inside surface area since data from
the Engineering Handbook (13) show the inside (exhaust gas) film as having
the most resistance to heat transfer. This is true only when the outside air
velocity across the outside tub surface is high enough to significantly
reduce the outside film thickness. For a definition of film resistance, see
Appendix C.
The ratio of the extended surface to the internal surface is listed in
Column 6. The condensers that were modified radiators had ratios of 5:1 or
greater. This is because they were designed for liquid in the tubes and air
outside. The resistance of the liquid film to heat transfer is small when
compared to gas film resistance. Therefore, to compensate, the gas film sur-
face area is made much larger.
The only cooler-condenser that did not use ambient air to cool the
exhaust was on the 1968 Jeep. This condenser was a modified automobile
radiator in which cold antifreeze from the Jeep's normal radiator was used as
the cooling medium. Exhaust gas was baffled through the finned area of the
radiator. Because of increased duty this system would be at a disadvantage
when the normal radiator isn't able to cool the antifreeze to below 5°C
(40°F). Under that condition 95 percent of the exhaust water vapor could not
be condensed out (Figure 1).
To find out whether or not this is a significant hindrance, a thermo-
couple was wired to the radiator return antifreeze hose on a 1968 Chevrolet
Carryall with a 250 CID engine. Insulation was placed over the probe so that
its temperature would be close to that of the antifreeze. At road speeds
from idle to 64 km/h (40 mi/h) the antifreeze temperature varied from 5 to
8°C (9 to 14°F) above ambient at ambient temperatures less than -18°C (0°F).
But when the vehicle was under heavy load, such as accelerating up a 10
percent grade, the return antifreeze temperature rose to 30°C (54°F) above
ambient. For exhaust condensation it is only necessary to consider normal
road load since ice fog is not as much of a problem in hilly terrain. There-
fore, at normal road load, the antifreeze temperature is low enough for
efficient condensation when the ambient temperature is below -18°C (0°F).
Automobile created ice fog is generally not a problem at temperatures above
-18°C (0°F).
A lower tendency for ice accumulation is a major advantage of passing
exhaust gas through the shell side of a cooler-condenser. The flow area on
the shell side is much larger than that through the tubes and will accomodate
much more ice before blocking. Exhaust gases on the 1970 Volvo flowed through
the shell side of its cooler-condenser.
22
-------
A first year quantitative performance comparison among the various
cooler-condensers is difficult because of lack of consistent test procedures,
different ambient conditions, and lack of accurate temperature data. However,
some statements can be made. All the cooler-condensers were successful in
removing some of the water vapor, particularly the cooler-condenser on the
Nova, Datsun, and Scout, which condensed over 90 percent of the water vapor
at engine idle. Only the condensers on the Scout and Nova were effective at
64 km/h (40 mi/h). Direct cooler-condenser comparisons are made difficult
because the exhaust gas into the Nova's cooler-condenser was above 260°C
(550°F) while the temperature into the Scout's was less than 93°C (200°F).
The Nova's exhaust was tapped before the muffler; the Scout's after. The
short connecting hose from the exhaust system to the cooler-condenser on the
Nova did not perform much cooling because of its small surface; compared to
that of a muffler. The 0.2 to 0.4 square meters (2 to 4 square ft) muffler
surface on the Scout removed considerable superheat from the exhaust.
For the heat transfer calculations, the estimated exhaust temperatures
into the condenser at 64 km/h (40 mi/h) were 150°C (300°F) for the Scout and
430°C (800°F) for the Nova. Because of low inlet temperature, -the cooler-
condenser on the Scout was the only one adequately sized to do the job at
64 km/h (40 mi/h) at ambients of -29°C (~20°F) or less.
The effect of the type of coolant circulation on the overall heat trans-
fer coefficient is shown in Column 4. All cooler-condensers with forced
convection had heat transfer coefficients of 73 kcal/h-m2-°C or more. The
1968 Chevy Carryall 4X4 transfer coefficient was evaluated at idle-free con-
vection. For the other cooler-condensers with free convection, the transfer
coefficients were 54 kcal/h-m2-°C or less. The cooler-condenser on the
Datsun is a mixture of forced and free convection because one tube partially
shelters the rest from the wind flow.
From the above discussion, it is obvious that to be effective, the
cooler-condensers must be exposed to the wind or have some other means, such
as a fan, to create air turbulence. For example, without forced convection,
the surface requirement on the Jimmy cooler-condenser is 2.1 square meters
(23 square ft).
The superiority of the Nova cooler-condenser is further exemplified by
Figure 3 which is a plot of cooler-condenser outlet temperatures vs. vehicle
speed. This figure also shows that the outlet temperature may vary by as
much as 19°C (35°F) at any one speed. That is because of varying cooler-
condenser duty caused by slight road slope and/or imperceptible acceleration-
deceleration.
The effect of engine load is further demonstrated in Figure 4 which
shows that for a fixed vehicle speed, higher engine rpm places more duty on
the cooler-condenser resulting in a higher outlet temperature. For a given
load and vehicle speed, an engine running at a higher rpm (lower gear) will
put out more waste heat because it has to overcome increased internal fric-
tion.
23
-------
x - 1967 Mercedes Benz 134CID @ -24°C jf-12°F)
•* - 1971 GMC Jimmy 250CID @ -23°C (-10°F)
o - 1974 Chevy Nova 250CID @ -19°C (-3°F)
96(60) -
80(50) -
64(40) -
48(30) -
32(20) -
16(10) -
X X
^ 0 00 00 X * X *
i
OOO X * X
-C
E
^
Q" 00 OO-M-X X*
UJ
LU
Q_
0 -H- X
-7 4 15 27 38 49 60 71 83 93 105 116 127 138 149 160
(20) (40) (60) (80) (100) (120) (140) (160) (180) (200) (220) (240) (260) (280) (300) (320)
EXHAUST TEMPERATURE AT OUTLET OF COOLER-CONDENSER, °C (°F)
Figure 3. First winter cooler-condenser exhaust temperature ranges.
-------
C_3
O
OO
ID
<=t
31
X
149(300)
93(200)
66(150)
38(100)
10(50)
A
B
C
1967 Mercedes Benz 134CID @ -20°C(-4°F)
1971 GMC Jimmy 250CID @ 18°C( 1°F)
1974 Chevy Nova 250CID @ -23°C(-106F)
I I I I I I
16(10) 32(20) 48(30) 64(40) 80(50) 96(60)
SPEED, Km/h (mi/h)
Figure 4. First winter cooler-condenser exhaust temperature
25
-------
Measuring the outlet temperature and calculating the percent water vapor
condensed does not always show the whole picture because it was noticed that
some cooler-condensers had exhaust that was less visible than others even
though both had condensed over 90 percent of the water vapor. This difference
is thought to be caused by relative humidity of the ambient air. If the air
is saturated (relative humidity = 100%), then any water vapor emission, no
matter how small, will at low temperatures appear as ice fog. The relative
humidity appears to be less during late winter then early winter. Therefore,
ice fog control generally appears more effective in Februrary then in Decem-
ber, even though the same percent water vapor is condensed in both cases.
Mist Coalescers
Cooling the exhaust gas to well below the dew point does not ensure that
all the condensed water will collect as one easily dischargeable aliquot.
Some of the water vapor condenses into discrete minute droplets (mist) which
when emitted appear as a dense ice fog. This problem was first noticed with
the Nova and Jimmy cooler-condensers. A first attempt in removal was to
direct the outlet stream against a large cold metal surface hoping the drop-
lets would freeze to it. This was done by directing the Jimmy's cooled ex-
haust against a large metal skid plate. Success was limited, probably because
the exhaust was not cold enough to begin with.
Next, a coalescer (Figure 5), was fabricated for the Nova cooler-con-
denser. This coalescer was a 61 cm (24 in.) length of 10 cm (4 in.) diameter
stove pipe. A 8 cm (3 in.) thick expanded fiberglass air filter plug was
placed near one end. The condenser exhaust was injected by a 4.4 cm
(1-3/4 in.) hose into the other end of the stove pipe and directed toward the
plug.
The injected exhaust plus some inspired ambient air passed through the
coalescing fiberglass plug which captured (by impingement) the minute droplets
causing them to grow until they were heavy enough to drip out of the coales-
cer. If for some reason the fiberglass plug should freeze solid, then the
exhaust would simple exit the stove pipe at the inlet .end since the annular
space between the inlet hose and the stove pipe was open. The coalescer
performed well; there was no longer any readily visible persistent mist
during vehicle operation.
The investigators with the Volvo eliminated the mist problem by reevapor-
ating it. They mixed the exhaust with ambient air being drawn through the
tubes. On the Jeep, the mist laden exhaust was directed at the vehicle's
radiator where it was reevaporated and mixed with the warm dry air. One
major problem with this is the possibility of carbon monixide poisoning of
occupants in the vehicles, especially if the fire wall between the engine and
passenger compartments is not airtight.
26
-------
1
/O cm (4m-)
\
OUTLET-
cm
r
Oft PLAST/C
PIPE SECT'ON
61 cm (2.4 <».
POROUS
Figure 5. First mist coalescer.
-------
In summary, some method of removing the mist that escapes the cooler
condenser must be devised. Reevaporation or coalescence will work, but re-
evaporation will only add to the atmospheric water vapor, part of which will
result as ice fog at some distance from the vehicle. Therefore, coalescence
is the best method.
The coalescer mounting direction is critical. Mounted paralled to ve-
hicle movement, it tends to freeze shut; mounted perpendicular there have
been no problems. It is also thought that 8 or 10 cm (3 or 4 in.) plastic
pipe makes a better coalescer shell since it accumulates less ice than does
metal. During the second winter, coalescer lengths of 35 cm (14 in.) were
found to be adequate.
SECOND WINTER RESULTS
For the winter of 1975-76, the AERS extrapolated the results from the
first winter (Table 2 column 5) to size and fabricate three cooler-conden-
sers. Later it was necessary to add more surface because the extrapolated
values from the first winter proved inadequate. Also, one exhaust dilutor-
heat exhanger was attached to the 1971 Jimmy.
The second winter results were analyzed in more detail because by then
the field evaluation techniques had become more standardized.
Constant Speed Performance
All the constant load cooler-condenser output temperatures were measured
by driving at a constant speed on the near level section of the Chena Pump
Station Road, a few kilometers south .of the AERS. Temperature measurements
were taken going out as well as coming back. The constant speed results were
usually within a few °C of each other.
After changing speeds, it would take several minutes for the cooler-
condenser outlet temperatures to reach equilibrium. In most cases after the
rate of temperature change dropped to less than 1°C per minute, the tempera-
tures were recorded. In later tests it was discovered that the temperatures
would slowly continue to drop by as much as 7°C after they had been recorded.
For example, with the gasoline engine this error would cause the actual
percent condensed to be 6 percent low for a hastily read outlet temperature
of 21°C (69°F). At low temperatures the temperature error has little effect
on performance (percent condensed) because the condensation curves flatten
out when more than 85 percent of the water vapor is condensed.
In actual vehicle use there is no such thing as constant speed level
road driving. The engine load seldom remains constant long enough for the
cooler-condenser to attain equilibrium. Therefore, it is considered better
to be on the conservative side when describing cooler-condenser performance.
To verify actual situation performance a drive-though-town test was made.
28
-------
Combination Coil/Tube-Cooler-Condenser--
The cooler-condenser on the 1968 Carryall consisted of three separate
heat exchangers: two mufflers in series, then a looped length of spiral
wound flexhose, followed by an EMT cooler-condenser then more flexhose. The
system was evaluated with and without the two lengths of spiral wound flex-
hose. Each respective total heat transfer surface was 1.9 and 1.1 square
meters (21 & 12 square feet).
Performance of the two systems is shown in Figure 6. At temperatures
below -37°C (-35°F) and speeds below 20 km/h (12 mi/h), both systems conden-
sed out over 90 percent of the water vapor. At -21°C (-6°F) the smaller
system was not capable of condensing out any water vapor. It had inadequate
surface. The total system of 1.9 square meters (21 square feet) was more
than adequate since it condensed over 80 percent of the water vapor at 64 km/h
(40 mi/h) at an ambient of -11°C (+12°F). Its condensation performance at
lower temperatures was slightly better but never exceeded 95 percent; probably
because internal ice films built up and decreased the heat transfer efficien-
cy.
In sizing a cooler-condenser for this vehicle it appears that 1.1 square
meters is not enough surface. But, 1.9 square meters will provide for greater
than 80 percent condensation at speeds up to 64 km/h (40 mi/h) and ambients
temperatures -11°C (+12°F) or less. The muffler, but not the connecting
tailpipe surface, is included in the 1.9 square meters.
When making extrapolations to other vehicles it must be recognized that
a cooler-condenser's surface is directly related to its duty. For example, a
heavier vehicle with a larger engine would require a proportionally larger
surface.
Single Pass Stainless Steel Cooler-Condense)—
The second winter cooler-condenser performance on the 1974 Chevy Nova is
shown in Figure 7. Curves 1 through 5 were results with the front outlet
coalescer mounted below the bumper on the driver's side. The total surface
area including the inlet flexhose amounted to 1.3 square meters (14 square
feet). It was capable of about 60 percent condensation at 64 km/h (40 mi/h)
for temperatures below -19°C (-13°F). The poorest high speed performance,
steepest curve, is at the lowest temperature, -42°C (-43°F). This was proba-
bly due to ice film formation. The ice film creates more resistance to the
flow of heat through the tube walls. For more discussion on icing see the
section on back pressures. Also note at this ambient that the idle perfor-
mance exceeded 90 percent condensed. This value may be high because the ice
formed at the higher speed is probably melting and increasing efficiency at
idle.
29
-------
o
LU
UJ
Q
O
O
O
Q_
oo
o
0
o:
00-^
_
90 —
80—
70—
-
60-
50 —
40 —
30 —
20 —
10 —
" ""~2H_ ^ • ^-L^_
V^ T
Two muffl ers ,
Ambient
1. -H°C (
2. -21°C (
3. -29°C (
4. -42°C (
Two mufflers &
Ambient
5. -37°C (
6. -21°C (
6.
r i i
0 16(10) 32(20) 48(30)
— — »__
^^^-^
" ^^
flexhose & EMT:
temperature
12°F)
-6°F)
-20°F)
-43° F)
EMT:
temperature
-35°F)
-6°F) 0% condensed *
i T~
64(40) 80(50)
SPEED Km/h (mi/h)
Figure 6. 1968 Chevy Carryall (4 x 2) second winter cooler-condenser performance.
-------
100 —i
Q
LU
GO
O
O
o;
o
Q-
-------
When temperatures are near freezing (-1°C, 30°F) the idle performance is
poor because the temperature difference across the cooler-condenser is low.
But at speeds greater than 35 km/h (22 mi/h) the performance exceeds that for
all the lower temperatures, probably because there is no ice film to restrict
the flow of heat at the -1°C (30°F) ambient temperature.
Adding enough spiral wound flexhose to extend the outlet to the rear of
the vehicle increased the cooler-condenser surface to 1.8 square meters
(19 square feet). This increased surface' resulted in 80 percent or more
condensation at speeds less than 64 km/h (40 mi/h) for ambients -30°C (-22°F)
and warmer- At -40°C (-40°F), 68 percent was condensed at 64 km/h (40 mi/h).
Because it would not hold water, the spiral wound flexhose never filled
with frozen condensate.
Four Pass Stainless Steel Cooler-Condenser--
The performance of the cooler-condenser on the 1967 Mercedes Benz diesel
is shown in Figure 8. Curves 1 through 6 were results when a 3.8 cm (1.5 in.)
diameter spiral wound metal flexhose was used to connect the cooler-condenser
outlet to the muffler inlet. Curve 7 illustrates data when the connection
was made with a plastic suction hose. The spiral wound flexhose added 0.14
square meters (1.5 square feet) of heat transfer surface and, since it leaked,
reduced the condensate freezing problems in the muffler and downstream. The
cooler-condenser performed well only at idle. Its lower performance under
load at the lower temperatures probably indicates ice formation on the inside
tube wal1s.
The diesel engine is the only automotive type engine that operates with
excess air. Its exhaust carries a larger fraction of noncondensables than a
gasoline engine exhaust. The diesels exhaust must therefore be cooled to a
lower temperature to achieve the same percentage condensed as a gasoline en-
gine's exhaust. In other words, for the same percentage condensed, the
diesel exhaust condensate is nearer freezing; see Figure 1.
Exhaust Oil utor-Ambient Air Heatei—
The eight meters (26 ft) of perforated spiral wound flexhose connected
to the tailpipe and mounted under the 1971 Jimmy functioned as an exhaust
dilutor-ambient air heater. It heated ambient air so exhausT water vapor
would be accepted without forming ice fog. There is also some associated
exhaust condensation, but that is not the device's principal function. The
exhaust is being dispersed along the hose length and diluted as its tempera-
ture drops. There is much more visible ice fog reduction than a condensation
curve would show. The device performance is best displayed by the with and
without photos, Figures 9 and 10. Both photos were taken within five minutes
of each other at an ambient of -29°C (-20°F). In both cases the vehicle was
idling about 1100 rpm. In figure 9 the dilutor-ambient air heater hose can
be seen slipped over the tailpipe below the left tail light. Without the
device the person standing at the left of the vehicle is almost completely
obscured. The visual effect during driving is similar, but not as spectacu-
lar.
32
-------
C/
O
O
o;
o
O-
ct
o:
CO
CO
Ambient temperature
-15°C
-21°C
-23°C
-28°C
-29°C
-32°C
(5°F)
(-20°F)
(-25°F)
with plastic hose
-42°C (-43°F)
64(40)
80(50)
16(10) 32(20) 48(30)
SPEED, Km/h (mi/h)
Figure 8. 1967 Mercedes Benz diesel second winter cooler-condenser performance.
-------
Figure 9. 1971 GMC Jimmy with dilator -
ambient air heater
Figure 10. 1971 GMC Jimmy without
dilutor - ambient air
heater
34
-------
Cooler-Condenser Si zing--
When comparing sizing criteria for the exhaust gas to air forced convec-
tion cooler-condenser, two different estimates were obtained. One was extra-
polated from the first winter's results. The other from the Engineering
Manual (13). The first winter's results indicated that about 0.93 square
meters (10 square feet.) of transfer surface was needed to condense out 95
percent of the water vapor at 64 km/h (40 mi/h) and -29°C (-20°F) ambient.
On the other hand, calculations based upon data in the Engineering Manual
yield about 1.9 square meters (21 square feet) as the required surface
(Appendix C). In field tests about 1.9 square meters (21 square feet) of
transfer surface was found to be required. The probable reason for this
discrepancy is that the first winter's data were obtained at warmer tempera-
tures when internal surface icing would not have reduced the overall heat
transfer coefficient.
Diesels have a high percentage of non-condensables (non water) in their
exhaust and they usually emit comparatively less ice fog than gasoline or
propane engines (Table 1). Because of the condensate freezing problem and
low water vapor exhaust concentration, it is expected that the air against
exhaust gas cooler-condensers will require more surface on diesel powered
vehicles. This fact is evident from the tabulated coefficients in Table 3,
which is a comparison of the overall heat transfer coefficients for the
second winter's cooler-condensers. Because the fuel economies were assumed,
the heat transfer coefficients have an accuracy of only one significant
figure.
Drive Through Town
To demonstrate the actual performance of the ice fog cooler-condensers a
special test was performed by driving the 1974 Chevy Nova through town. Data
were gathered under normal city driving conditions. Temperatures, speed,
location and time were recorded every 15 seconds. Odometer readings were
taken at different intervals so that an average town speed could be calcu-
lated.
Speed varied between 0 and 64 km/h (40 mi/h) with an average intown
speed of 13 km/h (22 mi/h). Figure 11 is a graph showing percent water vapor
condensed and vehicle speed vs. time. Outlet temperatures ranged
between 9°C (4S°F) and 36°C (97°F) with an overall average of 21 °C (60°F).
Ambient air temperature was 16°C (4°F) Amounts of water vapor condensed
ranged from a low of 57 percent under heavy acceleration to a high of 92
percent at idle. The overall amount condensed for the total run was 81
percent. The day was clear with open roads and occasional ice patches.
Traffic was moderate to heavy at the busier intersections.
Back Pressure and Tube Icing
Another important consideration is the added back pressure on the engine
caused by the cooler-condensers. The Jimmy (first, winter). Nova and Mercedes
had their exhaust manifolds connected directly to a cooler-condenser in place
35
-------
TABLE 3. SECOND WINTER COMPARISON OF OVERALL HEAT TRANSFER COEFFICIENTS AT 64 km/h (40 mi/h)
TYPE OF
COOLER-CONDENSER
Vehicle and Exchanger
Surface m2 (ft2)
Gasol ine Engines
Combination Coil/tube on
1968 Chevy Carryall
0.79 (8.5)
0.79 (8.5)
0.79 (8.5)
Single Pass
Stainless Steel
on 1974 Nova
0.97 (10.4)
1.42 (15.3)
0.97 (10.4)
1.42 (15.3)
Diesel Engine
Four Pass Stainless Steel
on 1967 Mercedes Benz
2.28 (24.5)
2.28 (24.5)
1.91 (20.5)
TEMPERATURES
AMBIENT
AIR
°C (°F)
-9( 15)
-29(-20)
-41 (-43)
-19( -3)
-29(-20)
-40(-40)
-40(-40)
-20( -5)
-31 (-25)
-41 (-42)
COOLER-CONDENSERS
OUT IN
°C (°F)
16(60)
12(53)
7(45)
34(94)
27(80)
51(123)
32(90)
24(75)
22(71)
27(88)
°C (°F)
74(165)
171(340)
107(225)
238(460)
260(500)
341(645)
299(570)
249(480)
282(540)
271(520)
ASSUMED
FUEL
ECONOMY
km/1
(mi/gal)
7.7(18)
7.7(18)
7.7(18)
8.5(20)
8.5(20)
8.5(20)
8.5(20)
13(30)
13(30)
13(30)
EXHAUST
HEAT
REMOVED
kcal/h
(BTU/h)
in 1000's
5.2(21)
8.4(33)
7.0(28)
7.1(29)
8.2(32)
6.5(26)
8.7(35)
4.1(16)
5.0(20)
3.4(14)
OVERALL
COEFFICIENT
U=O/AAT
kcal fBTU \
h-m^-°C lh-ft^-°FJ
» /
90(18)
120(25)
120(24)
60(12)
40(8)
30(7)
30(7)
15(3)
15(3)
10(2)
-------
o
LU
OO
L±J
O
21
O
(_>
o;
o
D_
<:
n>
o;
UJ
h-
^
h-
-^r
*^_.
UJ
CJ
or
100
90
80
70
60
—
—
—
—
™
64(40)
UJ
Q_
32(20}
16(10)
20
TIME, minutes
Figure 11. Cooler-condenser performance during a drive through town.
-------
of the muffler. Extremely high back pressure readings on the 1968 Scout
(Table 2) were probably caused by some tubes freezing shut. All other contrac-
tors reported insignificant back pressures with the addition of a cooler-
condenser.
The back pressures attributable to the first winter's cooler-condensers
are shown in Figure 12. Addition of the cooler-condenser to the Jimmy's
exhaust system increased the back pressure by 4-1/2 times at 80 km/h
(50 mi/h). However, except for high speeds, its back pressure was similar to
that on the Nova which had the same displacement engine.
The first cooler-condenser on the Nova was the only one which did not
substantially increase back pressure much above the normal exhaust system.
However, twice during highway runs at temperatures below -29°C (-20°F) the
cooler-condenser plugged with ice and caused the pressure relief cap on the
end of the normal exhaust pipe to blow off. Apparently the small, thin tubes
on this first condenser (modified radiator) froze shut.
The back pressure on the diesel was higher than for the gasoline powered
vehicles because, for the same fuel consumption, the diesel engine has about
three times the exhaust volume (200 percent excess air).
The effect of engine rpm for a given road load (constant at any one
speed) is shown by the third and fourth gear curves (Figure 12). Use of the
third gear at 64 km/h (40 mi/h) creates more exhaust gas than fourth gear,
which results in more back pressure. A discussion of the diesel back pres-
sure as related to chain turbulators in its cooler-condenser during the
second winter is in Appendix A.
Because of greater temperature differences, the overall condensation
performance as indicated in Figures 6, 7, and 8 should have increased as the
temperatures decreased. But they did not. It was speculated that there was
some internal ice film formation increasing the thermal resistance (recipro-
cal of the heat transfer coefficient). If so, this ice film would have
reduced the flow passage (cross-section area) resulting in increasing back
pressure with decreasing ambient temperature (more ice). The effects of this
icing are demonstrated in Figure 13, which shows that at the lower ambient
temperatures there is a significant increase in back pressure.
The icing effects did not appear to be accumulative over short periods
because the stainless steel cooler-condenser on the Nova never plugged with
ice. However, the vehicle was parked in a heated garage about once every 5
to 15 days. But most of the time the garage temperature at the cooler-
condenser level never exceeded 0°C (32°F). Therefore, any ice should not
have melted in the garage.
On the morning of February 12, 1976 the 1968 Carryall (4x2) was left
idling at about -30°C (-22°F) for 30 minutes while a local television camera
crew made a filmed newscast for the evening news. The vehicle was shut off
without blowing out the liquid condensate which had accumulated at the bottom
38
-------
oo
1_U
cc:
OO
LU
CfL
o
-------
60 -,
gauge limit
00
LU
DC
Qi
Q_
O
=C
CO
50 —
40
30 —
102
-------
of the cooler-condenser. Later attempts to use the vehicle were futile be-
cause the cooler-condenser system was plugged with ice. The back pressure
was so great that the vehicle would only idle. It was therefore driven, at
two miles an hour, to a heated garage where it thawed open after sitting
inside over a weekend. Even with the garage thermostat set at about 20°C
(68°F) it took three days to melt the ice. This situation was probably
precipitated as a result of intentional plugging of the weep hole on the
bottom header. The hole had been plugged because it was too large, allowing
excessive exhaust gas leakage. There were no other ice plugging problems
with the EMT cooler-condenser on this Carryall.
Back Pressure and Fuel Economy
Exhaust systems are generally designed for low back pressure because it
is known that high back pressures rob power and decrease fuel economy. The
cooler-condensers increase back pressures, as shown in Figures 12 and 13.
During the second winter it was decided to quantify the back pressure effect
on fuel economy for two vehicles.
Test runs were made in the Nova and Jimmy over a 32 kilometer (20 mile)
level test section on the Richardson Highway, south of Fairbanks. Fuel economy
was measured both with and without cooler-condensers in the exhaust system.
The back pressure with the cooler-condenser was from 20 to 150 percent
more than without it. For example, back pressure on the Jimmy at 80 km/h
(50 mi/h) was steady at 51 cm of water column (20 in. H20) with the cooler-
condenser connected. Without the cooler-condenser it was 36 cm of water
column (14 in. H20). For the Nova, at 80 km/h with the cooler-condenser the
back pressure was 76 cm of water column (30 in. H20), and 31 cm water column
(12 in. H20) without.
Calculations with data from the Kent-Moore fuel meter (readable to 0.001
gallon) showed there was only a 0.5 percent increase in fuel consumption with
the cooler-condenser on the Nova, and 3,2 percent increase with the exhaust
dilutor-ambient air heater on the Jimmy. The precision and accuracy of the
instruments were each less than 4 percent. Therefore, this suggests that
additional back pressure due to installation of cooler-condensers will proba-
bly not reduce fuel economy by a significant amount.
Passenger Compartment Carbon Monoxide Measurements
When installing a cooler-condenser, the final exhaust outlet was re-
located to the front of the vehicle in some cases. This relocation might
increase the risk of introducing more exhaust fumes into the passenger com-
partment, especially if the exhaust is directed out the drivers side, against
an adjacent wall such as at a bank drive up window.
To quantify this risk, interior carbon monoxide (CO) levels were moni-
tored periodically on test runs. Measurements were taken with an Ecolyzer,
ambient carbon monoxide analyzer, model #2900. Values did not follow any
pattern except at speeds above 32 km/h (20 mi/h). With the windows closed
41
-------
and heater running the CO level remained at or near zero parts per million
(ppm) on the 1974 Chevy Nova, which was equipped with a front mount cooler-
condenser and front exhaust outlet. At idle the reading ranged between 3 ppm
and 30 ppm with an average of 22 ppm. For speeds up to 32 km/h (20 mi/h) the
average was about 7.5 ppm.
The 1968 Chevy Carryall cab readings were higher because the two cooler-
condensers mounted at the front of the vehicle were not leakproof and there
were a number of holes in the floor boards through which exhaust gases entered.
The exhaust outlet was directed at the drive shaft behind the front seat. At
idle the CO readings ranged between 10 ppm and 40 ppm with an average of 18
ppm. At speeds up to 32 km/h (20 mi/h) the levels ranged between 5 ppm and
20 ppm with an average of 11 ppm.
The 1971 GMC Jimmy also had high passenger compartment CO levels with
the exhaust dilutor-air heater mounted under the vehicle. One reason for
this was that some of the distribution holes were directed towards the floor
boards, giving exhaust gases more opportunity to enter the cab. For periods
of idle lasting over two minutes, CO levels reached a maximum of 100 ppm.
Other idle data gave values between 50 ppm and 10 ppm with an average of 33
ppm. Speeds of between 16 and 32 km/h (10 and 20 mi/h) had CO readings
between 50 ppm and 15 ppm with an average of 26 ppm. Above 32 km/h (20
mi/h), the CO level dropped below 5 ppm.
In all of the above measurements, none of the vehicle's occupants were
smoking. For comparison, one pipe smoker in a pickup truck raised the cab CO
levels at different rates, depending on certain variables. For example, at
idle, with the heater on the CO was 25 ppm. Turning the heater off increased
the level to 45 ppm. Cruising at 56 km/h (35 mi/h) with the heater on, the
CO was 5 ppm; however, shutting off the heater increased it to 15 ppm. A
puff of smoke blown in the direction of the CO analyzer shot the CO level to
over 60 ppm.
Overall Ambient Ice Fog Reduction
In discussing ice fog control techniques, the first question one asks
is, "How effective is it or how much will the fog be reduced?"
Ice fog is generated by heating and power plants, coolijig ponds, and
motor vehicles. Ice fog from the tall stacks of power plants does not readily
add to the ground level problem until the temperature drops to -40°C (~40°F)
or less. Again, except for very low temperatures, cooling pond ice fog is
mainly concentrated adjacent to and down wind of the ponds. There are only
two cooling ponds of significance in the Fairbanks air shed. They are the
Ft. Wainwright pond which serves the South power plant, and the Chena River
which acts as a receiver for the cooling water from the Fairbanks Municipal
Utilities System.
When considering on-the-road ice fog, there are two significant sources;
low level home heating stacks and motor vehicles. At this point, we will
consider only these latter two sources in calculating ice fog reduction.
42
-------
From discussions with the fuel suppliers, the authors estimate that the
amounts of heating oil and motor fuel consumed in the Fairbanks air shed are
about equal; 50 percent each. The amount of water vapor created is roughly
the same for each of these fuels. Therefore, if controls were applied to
motor vehicles to reduce their water vapor emission by 80 percent, then the
overall ground level water vapor emission would be reduced by one half of 80
percent, or 40 percent.
Let us consider two conditions, one with incipient ice fog and the other
with ice fog already present.
As a first example, without ice fog control, assume conditions are such
that incipient ice fog occurs at -32°C (-25°F). Incipient means 100 percent
relative humidity @ -32°C. And also assume 90 percent of all water vapor in
the affected area is a result of man's activities. Now, if ice fog control
is applied to all motor vehicles, when would ice fog begin to form? From the
psychrometric charts (Appendix E @ -32°C [~25°F]) the saturation (incipient
ice fog) humidity ratio is 0.0002 gm of water per gm of dry air. Now if ice
fog controls were to decrease the water vapor input by 40 percent, the humi-
dity ratio would be reduced to 90 percent of (1 - 0.4) 0.002 = 0.001 which
would not appear as ice fog (100 percent relative humidity) until tempera-
tures reached -37°C (-35°F). Because the air is very stable there will still
be isolated spots, such as at intersections, where ice fog (relative humidity
>100%) will still appear. The above calculations assume complete phase
equilibrium; which means that there is no supercooled water vapor.
When ice fog is present the air is saturated with water vapor. There-
fore, any water vapor emission will immediately be converted into more ice
fog. The Lambert-Beer Law can be used to better estimate the visibility
effect of reducing the ground level ice fog input by 40 percent. This law of
light absorption states that: _ ,
Light intensity after absorption = initial intensity x 10 , where a
is the extinction coefficient which depends upon the ice fog particle morphol-
ogy, b is the sight distance, and c is the concentration of ice particles.
When it is desired to calculate the effect of a new concentration (c1) on
sight distance (b;) for the same light intensity the relationship reduces to
the follow ing form:
a'b'c' = abc
When consider! ng ice fog from only one source a' = a; then: c'/c = b/b' .
If control reduces ambient ice fog by 40 percent then c1 = 0.6c. The
effect upon the sight distance is b1 - b/0.6 = 1.67b. Therefore, for this
case the visibiltiy is increased by 67 percent. Based on all assumptions
discussed above, this means that if visibility were 50 meters (150 ft.)
before control it would then be 83 meters (250 ft.) after control. The
visibility increase would be slightly greater than 67 percent because the
extinction coefficient for motor vehicle created ice fog is higher than for
ice fog from other sources. Exact values of this coefficient for the various
types of ice fog are unknown.
43
-------
The overall impact of automotive ice fog control would be a visibility
increase of 70 percent or more in areas where motor vehicles create 50 per-
cent or more of the ice fog. The picture will not always be as favorable as
shown because of the many other sources of ice fog. However, the motor
vehicle ice fog control devices would greatly reduce the roadway ice fog and
prevent dense ice fog from obscuring busy intersections.
If the water vapor is no longer put into the air, where does it end up?
The ice fog control device condenses it to liquid water which readily freezes
and if not retained by the vehicle undercarriage, it falls onto the road.
The road icing effect will be discussed in Section 7.
44
-------
SECTION 6
FUTURE APPLICATIONS
SELECTION OF HEAT TRANSFER MEDIUM
Both the exhaust dilutor-air heater and the cooler-condensers are effec-
tive mobile source ice fog control devices. But because of its intentional
exhaust dispersion beneath the vehicle's floorboards, the exhaust dilutor-air
heater should not be used under any passenger compartment.
The cooler-condenser is the only remaining safe, effective device. To
keep the physical size within reason requires forced convection. However,
ice films may develop which cause poor performance and increased back pres-
sure. Weep holes in the manifolds are not effective in reducing ice films in
condenser tubes. There are three possible solutions. One is to add control
louvers, as with the 1968 Scout, to block the cold air. Or the exhaust flow
passages can be made so large that they will hold a winter's ice film build-
up. It would then be necessary to oversize the surface to compensate for
increased heat transfer resistance. The last solution would be to use cold
antifreeze for the coolant as was done on the 1968 Jeep, where a baffled
radiator was the cooler-condenser. Freezing problems would be less since the
flow area for the exhaust gas would be 5 to 10 times that of a nominal 5 cm
(2 in.) tail pipe. It would take much longer for ice to bridge across the
fins. Also, additional freeze protection could be accomplished by incorpora-
ting a temperature controller which would block off air flow to the normal
radiator causing the antifreeze to warm up enough to melt ice in the cooler-
condenser.
The other advantages of the baffled radiator cooler-condenser are: low
space requirement because of the extended surface, low pressure drop because
of large flow area, and low cost because of mass production of radiators. If
automotive radiator solders, which soften at about 200°C (400°F) can withstand
the high exhaust temperatures on one side and cold antifreeze on the other,
then economical cooler-condensers could be fabricated by adding baffles to an
auxiliary radiator as was done on the 1968 Jeep. The automobile's regular
radiator would cool the antifreeze before being pumped through the cooler-
condenser (baffled radiator). The disadvantage of this system is increased
vulnerability to the engine cooling system. The additional hoses and connec-
tions to the antifreeze system increase the chance of leaks or total loss of
coolant.
45
-------
MOUNTING LOCATIONS
When considering air to exhaust gas cooler-condensers the first question
that comes to mind is where to put them. To obtain the best advantages of
forced convection they should be mounted in front or behind the vehicle's
radiator. Many American-made vehicles have a removable spacer on the radiator
fan hub. Moving the fan back allows mounting room for a cooler-condenser.
The EMT cooler-condenser on the 1968 Chevy Carryall (4x2) was mounted in
front of the radiator. To check for possible hot weather problems a long
2 km (1.2 mi) run up a 12 percent grade during a warm 24°C (75°F) day did not
cause the engine to overheat even with hot exhaust flowing through the EMT
cooler-condenser. Cooler-condensers could also be designed to fit into the
space reserved for the automotive air conditioner condenser, since air condi-
tioners are unnecessary in arctic regions.
CORROSION
The cooler-condenser condensate is corrosive because it contains car-
bonic halogenic, sulfurous(ic), and nitrous(ic) acids. One contractor
reported the condensate pH to be consistently between 3 and 5.5 which is
highly acidic. He recommended that anodized aluminum be considered for these
cooler-condensers (1). Because of the difficulty in maintaining the integrity
of the protective aluminum oxide film during fabrication, the authors feel
that anodized aluminum would quickly corrode. Regular or mild carbon steels
would also not be expected to last long when subjected to this acidic conden-
sate.
A recent material engineering symposium addressed the corrosion problems
in flue gas scrubbers (14). The chemical quality of the scrubbing liquid
would be similar to the cooler-condenser condensate. The following recommen-
dations were made: use stainless steels such as Carpenter 20 or Uddeholm
9041 when the pH is below 4 and no halides are present. This would mean the
use of unleaded gasoline, because leaded gasolines contain bromine compounds
to act as lead scavengers. For condensate from leaded fuel, Hastelloy "C" or
Inconel 625 should be used. These alloys are very expensive, costing more
then $1 per pound when fabricated into tubing.
An exhaust system manufacturer recommended the use of a special "muffler
steel" known as type 409 Stainless. Since its cost is much less than the
other stainless steels it was decided to use it in the second winter's cooler-
condensers. Type 409 is a 12 percent chromium ferritic steel used in the
exhaust systems of some of the 1975-76 automobiles. It costs less than $1.50
per kg (70
-------
The tubes for the second cooler-condensers on the Chevy Nova and Mercedes
diesel were fabricated from type 409 Stainless. So far there has been no
evidence of corrosion. The EMT and galvanized flexhose have shown consider-
able corrosive attack in the same time period.
Old automobile radiators have been used as the heat exchanger for the
antifreeze coolant cooler-condensers. There has been no corrosion evaluation
but it is expected that the copper alloys would stand up as well as the
stainless steel.
COSTS
The installed costs of the ice fog control devices varied from about
$100 for the flexhose exhaust dilutor-air heater to about $850 for the stain-
less steel cooler-condenser. It should be remembered that these prototype
devices were one-of-a-kind fabrication. If they were mass produced the
installed cost would probably be 1/3 to 1/2 the above. As an example of the
cost reduction by mass production, consider the $90 automobile radiator (heat
exchanger). If it were to be fabricated as a one-of-a kind, the shop cost
would exceed $500.
The lowest cost, most readily available, solution to automotive ice fog
may be to use automobile radiators (as with the 1968 Jeep) as the cooler-
condensers. This is because they are mass produced and contain more surface
per dollar than any other heat exchanger. But they should be evaluated with
higher temperature exhausts to see if solder melting and corrosion are pro-
blems.
As mentioned in Section 6 the flexhose exhaust dilutor-air heater should
not be considered for use under passenger vehicles.
RECOMMENDATIONS
All the control devices worked to some extent on all the vehicles.
There were no insurmountable problems, although condensate freezing could be
a real problem in improperly designed cooler-condensers. What is now needed
is a demonstration-evaluation on fleet vehicles that are routinely operated
in the dense ice fog areas. Some vehicles should try antifreeze as the
coolant, while others could use the forced convection air cooled cooler-
condensers. The fleet testing would further demonstrate the practicability
of automotive ice fog control to the general public and would provide enough
information to select the best coolant and to decide if automobile ice fog
control regulations are desirable.
47
-------
SECTION 7
ENVIRONMENTAL CONSIDERATIONS
ADDITIONAL ICE ON THE ROAD
The accumulation of ice on public roads and at intersections due to
deposits made by vehicles equipped with cooler-condensers has given rise to
this question, "What increased dangers will be encountered by the additional
ice in relation to the hazards of poor visibility created by ice fog?" The
overall percentage of ice added to the highway has been calculated to be
about 20 percent more than what Mother Nature deposits over the four month
winter period (Appendix D). Assuming the condensate leaving the cooler-
condenser hits the road surface in the form of water, then freezes, the total
accumulated ice is approximately 1.4 cm (0.54 in.) from November through
February. This estimated condensate accumulation amounts to about 1/3 cm
(1/7 in.) per month. Mixed with accumulated hoar frost and snowfall it would
be difficult to identify and measure. Some of the smaller vehicles with
large exhaust systems, which act as cooler-condensers, presently deposit some
condensate on the road during extreme (-40°C) weather. This deposit is
presently concentrated.at left turn lanes.
The Alaska Department of Highways indicates that intersections are
sanded or plowed according to need. No other criteria are considered. Since
sanding is done only at intersections, bus stops, and curves, very little
economic impact should be felt by the state Department of Highways with the
incorporation of cooler-condensers on automobiles. Danger from the additional
ice is minimal in comparison to the hazards of low visibility driving in ice
fog. But the effects may be conflicting. Better visibility means many
drivers will increase speed, which may be more dangerous on the increased
ice.
To further qualify the geometry of this fallen condensate, a piece of
sheet metal plate 1 x 1.3 m (3 ft x 4 ft) was mounted under the 1968 Chevy
Carryall (4x2) to simulate a roadway. Its purpose was to catch condensation
leaving the cooler-condenser to determine the shape and pattern of the conden-
sate hitting the road, and whether or not the condensate would freeze before
reaching the road.
First tests were done using a 8 cm x 45 cm (3x18 in.) piece of plastic
pipe coalescer mounted against the vehicle frame. Condensate leaving the
coalescer when dropped a distance of 5 cm (2 in.) ran onto the metal plate
48
-------
forming a sheet of ice in an alluvial fan pattern approximately 3.8 cm
(1.5 in.) thick and 25 cm (10 in.) wide. The coalescer was removed to provide
a long drip ( distance) resident time. It was thought that the latent heat
of freezing would be dissipated at ambient temperatures below -23°C (-10°F)
before striking the metal sheet. The condensate fell in a liquid state
forming an ice sheet as before. The vehicle was driven at approximately
32 km/h (20 mi/h) with many stops and turns to simulate city driving.
1.9 liters (0.5 gallons) of gasoline were consumed. This should have yielded
approximately 1.8 liters (0.4 gallons) of condensate, 2/3 of which was
collected on the metal plate.
Next, a chain followed by a wire screen was placed between the outlet
and the metal plate. A long icicle formed betewen these intermediate devices
and led to a pool which froze as a sheet of ice on the metal plate.
The outlet was then placed at a 30° angle with respect to the vehicle
frame and 0.6 cm (1/4 in) above the drive shaft. Ice formed around the spin-
ning shaft much like the quills on a porcupine. Ice formed a rough surface
on the metal sheet in a scattered pattern radial with respect to the outlet
above the drive shaft. Some ice formation accumulated on the frame and the
undercarriage as the condensate spun away from the drive shaft.
Experience with the above techniques indicates that most of the conden-
sate would fall upon the road surface as droplets and freeze into rime ice.
During extreme cold the condensate would probably freeze before reaching the
road surface. When the ambient temperatures are only slighty below freezing,
the condensate would freeze to the road as clear ice. However, where there
is heavy traffic, tires would erode the forming ice through attrition and
cause some of it to be swept to the roadside. Also, some of the condensate
would freeze before reaching the road surface. It would not then easily
adhere, but would be partly swept aside by traffic.
REDUCTION OF OTHER AIR POLLUTANTS
Passing gas through a water spray, known as gas scrubbing, is one of the
oldest techniques for cleaning gases. The condensing section of a cooler-
condenser is an excellent scrubber because some of the small submicron parti-
cles that might not be caught in a water spray will act as condensation
nuclei and thereby be removed. The condensing section should theoretically
remove the majority of the exhaust particulates such as lead compounds and
soot. Since carbon monoxide is not appreciably soluble in water, it is not
expected to be removed in the condensation process. There should be partial
removal of the water soluble toxic gases such as halogenic acids, sulfur
oxides and nitrogen oxides.
Carbon dioxide is also absorbed in the condensate to form carbonic acid.
The fact that these acids are present in the condensate is indicated by its
low pH which ranges from 3.5 to 5.5.
A test was run on the 1974 Chevy Nova cooler-condenser to determine what
fraction of the exhaust lead compounds and sulfur oxides were actually removed
49
-------
with the condensate. The vehicle was allowed to idle at an ambient tempera-
ture of -36°C (-32°F) while its condensate was collected. The gasoline
consumption was measured with a 0.1 gallon (0.4 liter) burette and the resul-
tant condensate with a graduated cylinder. The test was divided into three
consecutive runs so that the amounts of collected condensate could be used to
verify steady state conditions. The calculated percent water vapor condensed
varied from 85 to 88.
The sulfur oxides come from the combustion of sulfur compounds in the
gasoline. They are measured as sulfates in the condensate. One contractor
performed such analysis and reported sulfate levels from 8 mg/1 (milligram
per liter) to 134 mg/1 (1).
The sulfur removal efficiency for the run on the Nova was calculated by
a sulfur balance which says that all the sulfur in the fuel must show up as
sulfur oxides in the exhaust. The test was run with the vehicle idling at
-36°C (-32°F). Exactly 1.1 liter (0.3 gallons) of gasoline was burned. The
fuel supplier said it contained 0.2 percent sulfur. Assuming complete
combustion the sulfur oxide emission would be 170 mg as sulfur. The collected
condensate amounted to 0.695 liter which had a sulfur compound concentration
of 32 mg/1 as sulfate; 7.4 mg as sulfur. The analysis was the Barium Chloride
Turbidimeter method. Therefore, four percent of the exhaust sulfur compounds
were captured in the condensate. Extrapolating for the range of reported
sulfate concentrations the cooler-condenser can be expected to capture from
one to 20 percent of the sulfur oxides.
With the introduction of catalytic converters on the newer vehicles some
of the sulfur dioxide is further oxidized to sulfur trioxide which reacts
with exhaust moisture to form a sulfuric acid mist. The trioxide and sulfuric
acid mist have a much higher affinity for water than does sulfur dioxide.
Therefore, on converter equipped vehicles, cooler-condensers would be expected
to remove a larger fraction of the sulfur oxides.
Atomic absorption analysis for lead in the Nova condensate yielded
24 mg/1. But for condensate from the 1968 Chevy Carryall values of 54 and
190 mg/1 were obtained. The fuel supplier indicated that the gasoline con-
tained 1.4 ± 0.28g tetraethyl lead (TEL) per gallon. Since TEL is 64.1 per-
cent lead (Pb), the Pb in the consumed gasoline (0.3 gallon) was
1.4 (0.3) 0.641 (1000) = 269 mg. At 24 mg/1 the Pb in the 695 ml of Nova
condensate amounted to 24 (0.695) = 17 mg. Therefore, the Pb removed in the
condensate was 6 percent of that in the gasoline. That is much lower than
expected. Corresponding removals for the cooler-condenser on the Carryall
were 14 and 49 percent for the 54 and 190 mg Pb/1 condensates, respectively.
Record checks indicated that the Nova had not been fueled with lead free
gasoline. Since its exhaust system and cooler-condenser had not been in
service as long as those on the Carryall, they may have still been accumula-
ting an internal coating of leaded combustion products. This could explain
the discrepancy between the lead concentrations in the different condensates.
50
-------
A simple nitrogen balance to determine the efficiency of nitrogen oxide
removal by the cooler-condenser is not possible because an undetermined frac-
tion of the atmospheric nitrogen is oxidized to nitric oxide. The solubility
of nitric oxide (NO) is about 1/2000 that of sulfur dioxide. Therefore, the
cooler-condenser removal efficiency for nitrogen oxides is expected to be
insignificant.
Hydrogen halide (halogenic acids -HX) appears in the exhaust because
halides are added to gasolines as lead scavengers. Their removal with con-
densate is expected to be relatively high since HX are from two to ten times
more soluble than sulfur dioxide.
Other non-water soluble gases such as the lighter hydrocarbons (methane
through propane) are not removed by cooler-condensers.
CONDENSATE QUALITY
As mentioned in the section on Corrosion, the condensate,is very corro-
sive because it contains combustion products such as carbonic, hydrohalic,
sulfuric, and nitric acids. Its pH varies from 3.5 to 5.5. It also contains
lead (leaded gasoline) compounds and soot. When released into the environment
its pH will slowly approach neutral as the carbonic acid decomposes into
carbon dioxide and water.
The condensate total solids content varies from 800 to 3900 mg/1. The
larger number represents some of the cooler-condenser corrosion products
(rust). About 60 percent of the solids are volatile (burn off at 600°C),
which indicates that bicarbonates and soot may be the major solids in the
condensate.
Without any ice fog control most of these combustion products are cap-
tured in the ice fog particles which eventually settle on the Chena River
flood plain. During breakup these compounds are released and captured by the
soil or its vegetative mantle or they flow into the local streams. This
acidic ice fog melt is probably well buffered before it reaches any water
course.
With ice fog control the frozen acidic condensate will be removed along
with snow cleared from local roads. This snow is hauled to snow dumps --
some on the banks of the Chena River. Upon melting it will drain directly
into the river. This low pH melt will probably not cause any pH drop in the
river because there is enough alkalinity available to act as a buffer. Also
the sulfur, nitrogen, and halide load to the river would show a short term
increase during snow dump melt. The increase of lead and soot would probably
be less since considerable lead and soot already end up in the snow dump.
51
-------
REFERENCES
1. Holty, J.G. Air Quality in a Subarctic Community: Fairbanks, Alaska.
Arctic Journal of the Arctic Institute of North America. 26,
4:292-302, Dec. 1973.
2. Ohtake, T. , Studies on Ice Fog. APTD-0626 U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, July, 1971.
3. Jenkins, T. F. , R. P. Murrmann, and B. E. Brockett. Accumulation of
Atmospheric Pollutants Near Fairbanks, Alaska During Winter SR
225, U.S. Army Cold Regions Research and Engineering Laboratory,
Hanover, New Hampshire. April 1975. 27. pp.
4. Winchester, J.W. , W.H. Zoller, R.A. Duce, and C.S. Benson. Lead and
Halogens in Pollution Aerosols and Snow from Fairbanks, Alaska.
Atmospheric Environment (1):105-119. 1967.
5. Coutts, H. J. , L. E. Leonard, and K. W. MacKenzie Jr. Cold Regions
Automotive Emissions. Working Paper #19, Arctic Environmental
Research Laboratory, U.S. Environmental Protection Agency, College,
Alaska. August, 1973.
6. Coutts, H. J. , and C. D. Christiansen. A Flue Gas Heat Exchanger for
Ice Fog Control. Working Paper #25, Arctic Environmental Research
Laboratory, U.S. Environmental Protection Agency, College, Alaska.
February, 1974.
7. Tedrow, J.V., Exhaust Moisture Reduction by Prototype Heat Exchanger.
U.S. Army Cold Regions Research and Engineering Laboratory, Alaska
Field Station, Fairbanks, Alaska. April, 1969. (Unpublished
report).
8. Schmidt, R. Evaluation of Automobile Ice Fog Removal Device. Arctic
Studies Group, University of Alaska, College, Alaska. May 1975.
(Unpublished report).
9. Holty, J. A Report on Design and Testing of an Automobile Exhaust Gas
Moisture Condenser. A. E. Research, Inc., College, Alaska. May,
1975. (Unpublished report).
10. Schmidt, G. Evaluation of an Ice Fog Suppressor for Automobiles at
Fairbanks, Alaska. H & S Research, College, Alaska. May, 1975.
(Unpublished report).
53
-------
11. Scarborough, T. N. Demonstration of Ice Fog Removal Device on Automobile.
Scarborough and Associates, College, Alaska. May, 1975. (Unpub-
lished report).
12. Borghorst, J. T. Demonstration of Automobile Ice Fog Removal Device.
Simplex-Standard, Fairbanks, Alaska. May, 1975. (Unpublished
report).
13. Perry, R. H. , editor. Engineering Manual, 2nd Edition, McGraw-Hill Book
Company, New York, N.Y. 1967.
14. Troubleshooters Swap Data. Chemical Engineering. 81 (9):52. April 15,
1974.
54
-------
APPENDIX A
DETAILED DESCRIPTION OF ICE FOG
CONTROL DEVICE INSTALLATIONS
COILED METAL FLEXHOSE COOLER-CONDENSER
Research on solutions to the automotive ice fog problem was started at
the AERS in 1972. During that winter, a cooler-condenser heat exchanger was
mounted underneath the carriage of a 1968 Chevy 1/2 ton Carryall with a
250 CID 6 cylinder engine. This cooler-condenser consisted of 150 feet of 2
in. spiral-wound, galvanized, flexible exhaust hose. The hose hung from the
floorboards in four loops. The loops were hung over the propshaft under the
frame channels and behind the transmission. The condenser was supported by
1-1/2 in. x 1-1/2 in. angle iron rests which hung from the floorboards by 3/8
in. threaded rods. The muffler was removed and the flexhose connected direct-
ly to the exhaust manifold pipe. The layout is shown in Figure A-l.
In this first attempt, the flexhose was chosen because it was easy to
install. It was thought that since it would not hold water, it would not
freeze shut causing excessive engine back pressure.
The installation time was three man-days.
The flexhose appeared to be condensing out all the water vapor at ambi-
ents of -12°C (10°F) and colder. Operation of the vehicle at below freezing
temperatures caused massive icicle formations which clung to the flexhose,
breaking off during travel over rough spots, such as crossing railroad tracks.
The outlet temperatures were so low that the condenser became plugged with
frost. There was no noticeable increase in back pressure since the connecting
joints, every 50 feet, and the flexhose itself were not gas tight. The
following data were obtained at -12°C (10°F) ambient:
SPEED 40 MPH 50 MPH
Inlet temperature °F 1100 1200
Outlet temperature °F 10 10
Back pressure inches of water 12 27
The exhaust noise at idle measured 70 db at 4 feet on the A scale of a
Scott instrument, Lab Model 451 ANSI type S3a sound intensity meter. A 1968
pickup with the same 250 CID engine, but conventional muffler gave approxi-
mately the same noise level.
55
-------
46 /7? x 5 cm
t x 2 in.}
OUTLET
Figure A-l. Coiled metal flexhose cooler-condenser on 1968
Chevrolet Carryall (4x2).
56
-------
The exhaust back pressure was about the same as that with a conventional
muffler. For comparison, a 1974 Chevy Nova with the same displacement engine
has a back pressure of 28 in. water at 50 mph.
COMBINATION COIL/TUBE COOLER-CONDENSER
In the spring of 1975, corrosion had weakened the coiled flexhose
causing brittle failure at stress points. With the idea of trying out
another heat exchanger, one was built from 1 cm (1/2 in.) electrical metal
tubing (EMT) welded to two 5 cm (2 in.) header pipes (Figure A-2). Chains
were inserted inside each tube to provide gas tubulation. It was mounted
between the radiator and cooling fan. The fan hub spacer was removed to
provide clearance. The outlet was located over the right side steering
knuckle.
The EMT condenser did not provide enough surface area to adequately cool
the exhaust gases. In November of 1975, two conventional steel mufflers were
mounted in series in the exhaust pipe between the engine and the EMT conden-
ser. Work was done by a local muffler shop. The mufflers increased the
surface area to 1.1 square meters (12 square feet).
Although more surface area was provided, mounting of the mufflers under
the vehicle prevented maximum air flow around them. The EMT outlet tempera-
ture was too high to allow condensation at ambient temperatures of -21°C
(-6°F).
Another cooler-condenser was fabricated using 5 cm (2 in.) diameter
flexhose mounted in a switch-back configuration in front of the radiator;
adding one square meter ( 12 square feet) between the mufflers and the EMT.
Also, the outlet was extended 2 meters (7 ft.) to the center of the vehicle.
The total amount of surface area increased to 2.2 square meters (24 square
ft.). An average outlet temperature at 64 km/h (40 mi/h) was 12°C (53°F)
condensing out 85 percent of the water vapor at an ambient air temperature
of -21°C (-6°F).
BRAZED RADIATOR COOLER-CONDENSER
The AERS automotive ice fog research effort was formally initiated
during the winter of 1974-75. In that program, the first ice fog cooler-
condenser was installed on a 1/2 ton CMC Jimmy with a 250 CID 6 cylinder
engine. The first cooler-condenser was a radiator from a small foreign
vehicle. Its overall dimensions were 13-1/2 in. x 12-1/2 in. x 4 in. Its
solder joints were brazed to prevent joint melting. The solder melts at
about 232°C (450°F) while the brass brazing rod would withstand exhaust
temperatures to 454°C (850°F). The radiator was connected in front of the
muffler where exhaust temperatures of 426°C (800°F) were expected. For
protection, it was mounted between the drive shaft and the frame because the
vehicle was used off the road and for snow plowing. The cooler-condenser
performed satisfactorily, but had many attendant problems.
57
-------
MQUNT/NG
BRACKETS
^Scm
(2/r,.)
DIA.
(-CHAIN
TUR8ULATOQS'
FRONT VIEW
OUTLET
CUA/N
RETAINERS
20 EACH
- in. EMT
WELDED
SIDE VfEW
Figure A-2. EMT cooler-condenser on 1968 Chevrolet Carryall (4 x 2)
58
-------
At an ambient air temperature of 2°C (35°F), the cooler-condenser
dropped the exhaust from 321°C (610°F) to 32°C (90°F) at idle and from 482°C
(900°F) to 54°C (130°F) at 64 km/h (40 mi/h). Extrapolating to an ambient
temperature of -32°C (-25°F), it was estimated that the condenser exhaust
temperature would be 2°C (35°F) at idle and 21°C (70°F) at 64 km/h (40 mi/h).
The corresponding vapor (ice fog) removal would therefore be 95 percent and
82 percent.
It appeared that the condenser was going to be a success except that the
high temperatures reached, 648°C (1200°F) at 88 km/h (55mi/h), caused the
brazed header joints to melt open. These leaks resulted in excessive exhaust
noise and high temperatures on the floorboards, so the cooler-condenser was
removed.
FREE CONVECTION/FINNED TUBE COOLER-CONDENSER
Next, a heat exchanger was fabricated out of 1 cm (1/2 in.) electrical
metallic tubing (EMT). The condenser had 10 tubes 40 in. long with 9 pieces
of 1-1/2 in. x 3 in. slotted angle iron welded across the tubes to extend the
surface area. The inlet and outlet manifolds were 5 cm (2 in.) exhaust
tubing. This condenser was assembled and installed in about 1-1/2 man-days.
It is shown in Figure A-3.
It did not provide enough surface area so one-half of a Young Radiator
Company mobile oil cooler (MOC #6) was mounted in series with the EMT conden-
ser. To fit, it was cut in half, parallel with the finned tubes; the internal
tube turbulators were removed. Prior experience with the turbulators indi-
cated they caused too high a back pressure. Overall dimensions of the 1/2 MOC
#6 were 14-1/2 in. wide x 29 in. long x 1-1/2 in. thick, as shown in Figure
A-4. The EMT condenser was hung in place of the normal muff lei—between the
frame and the drive shaft on the driver's side. The 1/2 MOC was mounted on
the passenger side between the frame and drive shaft. The 1/2 MOC plus EMT
cooler-condenser system condensed out 55 percent of the water vapor at idle.
The total material cost is estimated at $200. The excessively large cooler-
condenser surface area under this vehicle was not very effective because it
was mounted high up between the frame channels. This shielded location
prevented cold ambient air from effectively removing the exhaust heat.
EXHAUST DILUTOR-AMBIENT AIR HEATER COOLER-CONDENSER
During the winter of 1975-76 a decision was made to try a different
method to solve the problem of ice fog control on the 1971 CMC Jimmy (4x4).
It must be realized that ice fog, by definition, occurs only when the air is
over-saturated with water vapor. For example, from the psychrometric chart,
Fig. E-l assumes an air temperature of -32°C (-25°F). At this temperature
air can hold up to 0.0002 g of water per g of dry air. The 0.0002 g/g repre-
sents 100 percent relative humidity. Any additional water vapor input will
form ice fog. But if the air is warmed to say -18°C (0°F) with the same
water vapor content, the relative humidity will be reduced to 23 percent.
59
-------
1 12 cm
(44 in.)
102 cm ^
^ (40 in.)
/-IO EACH i in- EMT
r 2
\. x x x \ \ \ \\x\\\\x\x\xx
xxxxxxxx x \ x \ xxxxxxxx
\X\\\ X X X X \\XX\\\X\ X. X
\. XXXXXXXXX X X X X \X\XX X
S. X XXXXXXXXX X X X X XX XXX
\\\ xxxxx xxxxxxxx xx\x
XXX X X X X X X X X X X X X X X XXX
XXXXX XXXXXXXXX X X X X XX
XXXXX \\X\XX\\\X X X X X X
XXXXXX X XXXXX X X X X XXXX
p
.~
\
ou
1
(
30 cm
:«2 in.)
INLET
61 cm
(24 in.)
ii ii
, 41 cm
(16 in.)
3 cm x 8 cm x 30 cm
(l in. x 3 in. x 12 in.)
SLOTTED ANGLE
5 cm (2 in.)
EXHAUST
PIPE
Figure A-3. First winter EMT cooler-condenser on 1971 CMC Jimmy.
60
-------
BRACKET
5° cfrj (2 in.)
INLET
HEADER
00
~74cm (29in.)
69cm (27'In.)
/V/°T PIPE PLUG
iiiiiiiniiiiiii
uiwwu'
w
CORE
HEADER
CORE
(2 in.)
OUTLET
Figure A-4. Free convection/finned tube (MOC-6) cooler-condenser on 1971 GMC JIMMY.
-------
At -18°C (0°F) the air is capable of holding up to 0.0008 g of water per
g of dry air before ice fog will form. Therefore, by taking 1.0002 g of
saturated air at -32°C (-25°F) and warming it to -18°C (0°F), it could
accept 0.0006 g of water vapor before forming any fog. However some ice fog
will form as the air cools from -18°C. This principle was tried on the 1971
Jimmy during the winter of 1975-76.
Eight meters (26 ft.) of perforated spiral wound flexhose were connected
to the tail pipe. The perforated holes, four per foot, were 0.6 cm (1/4 in.)
in diameter, approximately 15 cm (5 in.) apart. See Figure A-5. The hose
was wired to the frame channels, behind the transfer case and under the rear
bumper. The setup was quite effective in reducing visible ice fog by almost
one-half.
The hose was wired below the rear bumper so that it could easily be
slipped on or off the tail pipe for comparison purposes. Forcing the exhaust
to flow out the 0.6 cm (1/4 in.) holes was accomplished by partially plugging
the hose end with a paper towel; then condensate froze shut the last
one meter (3 ft.) of hose. During operation, exhaust heat would be trans-
ferred to the ambient air through the metal hose walls. Simultaneously moist
exhaust would be dispersed into this warm air because it could now accept
more water vapor before becoming saturated and showing visible fog behind the
vehicle. Also during extreme cold -34°C (-30°F) or less, some of the exhaust
water vapor condensed out and dripped from the metal hose. Several ice
stalagmites 10 to 15 cm (4 to 6 in.) high formed on the ground under the
perforated hose when the vehicle was left idling for 15 minutes at -40°C
(-40°F). In one winter's operation the hose rusted through at the tail pipe
connection. Accurate costs for the perforated hose could not be determined
since this hose was salvaged from the initial coils under the 1968 Carryall.
The estimated new replacement cost with stainless steel spiral wound flexhose
is about $75.
The advantage of this system is the low cost. One disadvantage is the
danger of CO poisoning. Because most of the exhaust is released under the
floorboards, some of it leaks through. Therefore, this system is not recom-
mended when occupants could ride in any contaminated airspace. However, this
system would be quite satisfactory under the open cargo bed of a truck. The
other disadvantage is that this system is not as effective as cooler-conden-
sers in limiting exhaust water vapor.
BRAZED RADIATOR
Early success with using the radiator as an exhaust gas cooler-condenser
on the Jimmy indicated that the initial solution to the problem of automotive
ice fog may be close at hand. It was decided to install a small radiator
(cooler-condenser) with brazed joints on a 1974 Chevrolet Nova sedan with a
250 CID 6 cylinder engine. The cooler-condenser was mounted in front of the
normal radiator behind the grill so the ambient air would flow over the cool-
ing fins. For convenience it was mounted off center on the driver's side.
The exhaust pipe was taped just in front of the muffler and 4.4 cm (1-3/4 in.)
flexhose was used to convey the exhaust to the cooler-condenser. Its
62
-------
FRAME
CHANNELS
JUt ( It
..,,
( l( III IK
INLLT
7.9m x 5 cm
(26 ft x 2 in.)
FLEXHOSE
OUTLET
Figure A-5. Exhaust dilutor ambient air heater on 1971 GHC Jimmy.
63
-------
overall dimensions were 39.4 cm (15-1/2 in.) wide x 43 cm (17 in.) long x
5 cm (2 in.) thick. The exhaust flow was through the inside of the tubes.
Past experience had shown that even though over 90 percent of the water vapor
may have been condensed out, there were some minute liquid water droplets
(aerosols) suspended in the condenser exhaust which appeared as ice fog. One
solution was to impinge the droplets upon a surface causing them to coalesce
and run off in a liquid stream. A coalescer was fabricated out of a 61 cm
(24 in.) section of 10 cm (4 in.) stove pipe into which a 7.6 cm (3 in.) plug
of expanded fiberglass furnace air filter was placed. By means of a 4.4 cm
(1-3/4 in.) flexhose, the condenser exhaust was directed at the fiberglass
coalescer in the 10 cm (4 in.) pipe. The setup is show in Figure A-6. The
coalescer effectively removed the droplets without freezing shut. Visible
ice fog at -28°C (-20°F) was negligible. The cooler-condenser was more than
adequate. There were only two problems with this application, one was exces-
sive heat and the other was plugging. The cooler-condenser was mounted next
to a thermoplastic parking light which partly melted. The plastic grill was
unaffected. During long, high speed trips, the condensate would freeze in
the tubes, restricting them, causing enough back pressure to reduce engine
performance. Apparently the tubes which were 1.9 cm (3/4 in.) x 0.16 cm
(1/16 in.) in cross section were too small to drain before they froze.
SINGLE PASS STAINLESS STEEL COOLER-CONDENSER
During the winter of 1975-76 a new cooler-condenser was built for the
1974 Chevy Nova sedan using 1 cm (1/2 in.) diameter type 409 stainless steel
(Figure A-7). Engineering design and tubing were furnished by AERS and a
fabrication contract awarded to the University of Alaska, Geophysical Machine
Shop. Price for fabrication was $600. The tubing ends were beaded with a
parker beading tool and swaged into the headers. The unit was 89 cm (35 in.)
long with 25 tubes. Its surface area was 0.85 square meters (9.1 square
feet).
It was mounted between the radiator and grill on the front of the ve-
hicle. 2.9 meters (9.5 ft.) of flexible exhaust hose were used to pipe the
hot exhaust from the muffler inlet to the cooler-condenser inlet header. A
chromel alumel thermocouple was inserted in the inlet header. The outlet was
piped to a coalescer mounted under the bumper in front of the driver's side.
Later a 4.4 meter (14.5 ft.) length of flexhose was extended to the rear of
the vehicle adding 0.85 square meters (6.2 square ft.) surface and a coalescer
with an iron constantan thermocouple was inserted in the outlet exh*aust flow.
At idle the output temperature averaged 10°C (50°F) with inlet tempera-
tures of 180°C (350°F). At speeds of 64 km/h (40 mi/h) the output temperature
averaged 21°C (70°F). Back pressure at idle was 3-10 cm (1-4 in.) of water.
At 64 km/h (40 mi/h)it was 76 cm (30 in.) of water.
Each tube had chains inserted inside to increase internal surface area
and to turbulate the flow of gases. The cooler-condenser was over-sized in
design (flow area) to get the required surface area. Therefore, addition of
chains did not affect engine performance or increase back pressure.
64
-------
BRAZED
JOINTS
FIBERGLASS
(24 in.)
4.4cm
FLEXIBLE
EXHAUST
•OUTLET
FRONT V/EW
RADIATOR
6 IDE
Figure A-6. Brazed radiator cooler-condenser and coalescer on 1974 Chevrolet Nova
65
-------
l4in.NPS THREADED
COUPLING INLET
3.8cm(ll/zih.DIA.
5cm (2in. LONG
OUTLET
Figure A-7. Single pass stainless steel cooler-condenser on 1974 Chevrolet Nova
66
-------
MODIFIED FINNED OIL COOLER COOLER-CONDENSER
Demonstrations were progressing satisfactorily with the gasoline com-
bustion engines so it was decided to try a vehicle equipped with a diesel
engine. A 1967 Mercedes Benz D200 was volunteered. The exhaust cooler-con-
denser for this vehicle was center mounted between the grill and the radiator.
In this application, the exhaust was piped directly from the manifold to the
cooler-condenser. The exchanger used was a Young Radiator Company mobile oil
cooler #2 (MOC-2) as shown in Figure A-8. At first, the MOC-2 was not modi-
fied, but after measuring excessive pressure drop, the internal tube tabula-
tors were removed. The overall dimensions of the MOC-2 were 41 cm (16 in.)
wide x 43 cm (17 in. ) long x 3.8 cm (1-1/2 in. ) thick.
This method worked satisfactorily in that the visible ice fog diminished
within a foot or so of the automobile at temperatures below -18°C (0°F).
This allowed total visibility of the vehicle during hazardous winter driving
conditions.
One problem encountered was the smell of exhaust fumes inside the cab
compartment. Therefore, it is important for the comfort and safety of the
passengers that the front mounted cooler-condensers be leak proof and the
exhaust outlet extend so it will bypass the seating area. Exhaust leaking out
of weep holes should be directed such that it will flow under the passenger
compartment since the heater system usually takes in fresh air between the
hood and windshield.
FOUR PASS STAINLESS STEEL COOLER-CONDENSER
A new cooler-condenser was designed for the Mercedes during the winter
of 1975-76 using the calculated amount of square feet as a guide for maximum
efficiency. Stainless steel type 409 was selected for the 1 cm (1/2 in.)
tubing and 18 gauge stainless steel sheets were formed into two multi-chamber
headers. The direction of flow through each row of tubes switched after
entering each chamber. There were four flow passes; see Figure A-9. The
last row of tubes had chains inside each tube to increase turbulance and
internal surface area.
The material and design were furnished by AERS, and Midway Welding of
North Pole did the welding. Cost of fabrication was $700. Installations of
connecting piping were performed by the Fairbanks Muffler Shop at a cost of
$150.
The condenser was mounted on the front of the vehicle between the radia-
tor and grill. A short exhaust pipe was connected directly to the engine's
manifold and secured tightly to prevent any gases from being drawn into the
passenger compartment. The drawback of using a short rigid pipe was that it
transmitted engine vibration directly to the auto body, bypassing motor
mounts. Therefore, a one foot section was replaced with flexhose wrapped
with high temperature asbestos tape to prevent gas leaks and to reduce trans-
mission of motor vibration. The tape failed to eliminate fume leakage, but
it was satisfactory for testing purposes.
67
-------
•4
43tm (ilia.}
3/4;^NPT OUTLETS fl)
5c.mf2.io-)
— l.^em
Figure A-8. fbdified finned oil cooler (MOC-2) cooler-condenser
on 1967 Mercedes Benz.
68
-------
42. TUBES
I cm
41.9 cm
06i in.)
/O cm
(4 in.)
S~cm.
(2 /».)
9O* ELBOW
OUTLET
4.1 cm D/A. x 5~ cm
(/•§- /'f?. DlA. -x 2 in.)
/NLET
Figure A-9. Four pass stainless steel cooler-condenser on
1967 Mercedes Benz.
69
-------
A six foot section of plastic suction hose connected the cooler-conden-
ser outlet to the muffler. A plastic hose was used to reduce the amount of
condensate freezing to the inside surface and increasing back pressure.
Water freezes to a cold metal surface more readily than to plastic. Measure-
ments of outlet temperatures indicated that internal freezing was not to be
considered a problem so the plastic hose was removed and replaced with metal
flexhose, adding 0.3 square meters to the system, bringing the total surface
area to 2.3 square meters (24-1/2 square feet). A coalescer was mounted to
the back bumper with an iron constantan thermocouple inserted into the
exhaust stream. This coalescer was a 36 cm (14 in.) length of 7.5 cm (3 in.)
plastic pipe with a 7.5 cm (3 in.) thick fiberglass coalescing medium. The
fiberglass became plugged with soot and ice.
Inlet temperatures ranged between 93°C (200°F) at idle to over 260°C
(500°F) at 64 Km/h (40 mi/h). Back pressure measured between 200 and
500 kdy/cm2 (3 and 7 psig). Ambient air temperatures on different testing
days were from -15°C (5°F) to -40°C (40°F).
The cooler-condenser effectively removed approximately 80 percent of the
visible ice fog. Under heavy acceleration, a plume of water particles was
visible at the tail pipe.
Some power loss was observed while climbing hills, indicating excessive
back pressure. Removal of the unit and gathering amounts of condensation
which drained out after allowing it to thaw overnight only produced about
200 ml of water. This was not considered enough to account for the power
loss. Therefore, excessive back pressure had been built into the cooler-
condenser.
Since the device was designed to calculated specifications, further
testing was required to determine the cause for the increased back pressure.
The problem was narrowed down to the chains which had been added to the last
row of tubes. They had not been taken into account during the original
calculations and were introduced as a last minute suggestion.
A mock-up was built using a positive displacement rotary blower to pro-
vide sufficient air flow to simulate the engine's exhaust. This was attached
to a flow meter in series with a piece of 1 cm (1/2 in.) 409 stainless steel
tubing cut to the length of those used on the cooler-condense^. A pressure
gauge was attached immediately preceding the tube. Tests were run with and
without internal chains. At a constant air volume of 10 cmh (6 cfm) the back
pressure without a chain was 5 kdy/cm2 (2 in. water). At the same volume of
air with a chain, the back pressure was 120 kdy/cm2 (48 in. water), an in-
crease of 2300 percent. Because of warm weather, the unit was not replaced
on the vehicle for further testing. However, it was assumed that it would
have worked just as effectively without internal turbulators and there would
not have been any noticeable power loss due to back pressure during heavy
load requirements.
70
-------
FAN TUBE COOLER-CONDENSER
The Arctic Studies Group designed and fabricated a fan cooled shell and
tube cooler-condenser for a 1970 Volvo Sedan. It was fabricated from 16
electrical metal tubes (EMT), 1-1/2 in. x 15 in. enclosed in a 12 in. x
15 in. x 8 in. sheet metal box. A 12v DC fan was mounted at one end to draw
ambient air through each tube so a continuous flow could be maintained. See
Figure A-10. The cooler-condenser was mounted on the rear bumper of a 1970
Volvo 144S. Exhaust gas at temperatures between 93°C (200°F) and 148°C
(300°F) flowed into the shell and was baffled around the tubes. The exhaust
gas outlet was recirculated through the tubes, mixing with the ambient air
being drawn by the fan. This outlet temperature was in the range of 5°C
(40°F).
This device prevented any visible ice fog from being released into the
atmosphere. The condensed water remained in a liquid state and leaked out of
the condenser through the seams, freezing on contact with the road.
Simulated tests were performed at the University's cold cell laboratory
where room temperatures of -26°C (-15°F) to -31°C (-25°F) were maintained. A
steam generator was used to simulate automotive exhaust. This ingoing imita-
tion exhaust temperature measured between 93°C (200°F) and 148°C (300°F).
Test results produced no visible exhaust vapor in the form of ice fog.
However, the simulated test data could not duplicate road conditions.
The time involved for assembling required approximately 4 hours in con-
struction time, 1/2 hour for installation, 2 hours for simulated testing, and
2-1/2 days of road testing while mounted on the vehicle.
Material cost can be itemized as:
Sheet metal $15.50
Conduit 12.00
Fan & Motor 20.00
Miscellaneous 4.00
$51.50
Taking into consideration the corrosion factor, this system had an
anticipated life of no more than three seasons.
FINNED COPPER TUBING COOLER-CONDENSER
Early project plans by AE Research, Inc. involved the fabrication of an
air cooled cooler-condenser made from 3.2 cm (1 1/4 in.) aluminum tubing sur-
rounded by 10 cm (4 in.) square fins spaced 2.5 cm (1 in.) apart over a 1.3 m
(4 ft.) length (9). A twisted aluminum strap was inserted to create turbu-
lance. It was mounted on the rear of a 1972 Toyota with a 1400 cc engine.
This was discarded when access to the Toyota ended.
A 1974 Datsun with a 1300 cc engine was used as a replacement in the
project. Since the engine displacement was smaller, a second cooler-condenser
was designed. A manifold style of construction was incorporated using three
71
-------
^ O ^ Crn (
— ^ o ^» f«<« / / if* r* M ^
'24 in.)
"• jo cm (is m.) .
>
/
f. —
/
L
! /
i /
i /
i / /
i — : ~^~ ~^~.
/3
(S
I
/
o
o
o
cm
in.)
rz^;
o1
o
o
'2 VD
-AM
tfi
1
/S cm
(6 in.)
\
/NLET-/
BAFFLES-^
3 EACH
IB cm K 2O cm
(7in. x & //?.)
FRONT V/EW
S" cm (2/n) Of A.
OUTLET
/€ EACH
// in. EMT
o
20cm
(8 in.) u-
oooo
^DOO
OOOO
OOO
00
1
3C
C/2
i
i
) cm
in.)
r
SIDE
Figure A-10. Fan tube cooler-condenser on 1970 Volvo sedan.
72
-------
parallel standard aluminum tinned-copper tubing baseboard heating pipes 1 m
(3 ft.) in length. No internal turbulators were used as the manifold was
expected to produce internal turbulance (Figure A-ll). The condenser was
mounted on the rear of the vehicle at a slight angle to drain condensate from
the condenser. Samples of condensate were collected and frozen for analyses.
Tests for pH, conductivity, and sulfates were run. pH remained consistent
between 3 and 4 indicicating a potentially high corrosive liquid. Conduc-
tivity ranged between 500 and 800 umhos. Sulfates were between 14 and
30 mg/1.
Material cost for the fin tubing and miscellaneous items for construction
of the second unit came to about $55. No cost figures were given for the
first unit constructed. Installation time for mounting and testing were not
included in the report.
Heat loss calculations were done on theoretical assumptions of fuel eco-
nomy in conjunction with measured temperature differences. Total gas flow
rate was measured using an Alnor velometer and exhaust gas density relation-
ships were provided by supplements from the AERS.
At engine idle the percent water vapor condensed approached 100 percent,
but not all visible vapor had been dispersed. The remaining aerosol fog was
of such a minimal amount that the condenser could be considered a success.
However, at higher speeds requiring more fuel consumption, the amount of
visible ice fog increased. To function more effectively, a more sophisticated
design is required which would pass more air over the fins to take away the
heat.
LIQUID COOLED COOLER-CONDENSER
Working with the University of Alaska Mechanical Engineering Department,
H & S Research designed a liquid cooled cooler-condenser rather than an air
cooled type (10). Their version consisted of a water-antifreeze system
mounted on the front of a 1968 Jeep Wagoneer. The coolant was connected in
series with the vehicle's cooling system. The exhaust gases were routed into
a shell encasing a radiator; Figure A-12. The exhaust temperatures going
into the condenser ranged between 93°C (200°F) and 148°C (300°F) while the
incoming coolant was about 26°C (80°F). The outgoing coolant and exhast
gases measured approximately 37°C (100°F). The outgoing gases were then
directed at the vehicle's radiator which reevaporated all water particulates
which were not removed in the condenser. This last step was undesirable
because of the hazards of drawing poisonous exhaust gases into the vehicle's
cab. Also the reevaporated water would form ice fog once it cooled down.
Large icicle formations accumulated at the lower seams where the condensed
water weeped out into the cold atmosphere. These icicles broke off and fell
onto the roadside.
Since safety was a prerequisite along with eliminating visible exhaust,
the unit was relocated to the rear of the vehicle. From this location it was
no longer practical to use the Jeep's radiator coolant.
73
-------
INLET
9km
all joints sweat soldered 50/50
2.5cm
(l'"Oelbow
2.. 5cm
0">} tee
OUTLET
137cm
temperature sampling port
lin..) fin tubing
Figure A-ll. Finned copper tubing cooler-condenser on 1974 Datsun
74
sedan.
-------
ANTI FREE2E
OUTLET
EXHAUST
OUTLET
-ANTIFREEZE
OUTLET
GALVANIZED
$H£ET METAL
LOUVERS
SIDE VIEW
Figure A-12. Liquid cooled cooler-condenser on 1968 Jeep Wagoneer,
75
-------
Another configuration was developed using a 15 gpm 12v DC pump and a
second radiator. Due to warm weather this system was tested at the Univer-
sity's cold cell laboratory with room temperatures of -31°C (-24°F) to -35 C
(-31°F) A steam generator was used to simulate automobile exhaust. The
ingoing pseudo-exhaust temperature was 121°C (250°F). The outgoing exhaust
measured 2°C (35°F) and coolant temperature was 5°C (40°F). There was no
visible exhaust vapor with this mock setup.
Material costs for the first device, including antifreeze, totaled $169.
Additonal cost for the second device was $135. Total overall material cost
for this project was approximately $300.
The flexible exhaust hose was considered the weakest part of the system,
due to its susceptibility to corrosion and attack by sulfurous acids. How-
ever, the sheet metal shell is equally attacked by acidic condensate. Its
anticipated life was no more than two seasons.
FINNED PIPE COOLER-CONDENSER
Scarborough and Associates designed and fabricated a finned pipe cooler-
condenser using a 1.3 m (4 ft.) length of 5 cm (2 in.) aluminum pipe with
forty 15 cm (6 in.) square aluminum fins around it, spaced 2.5 cm (1 in.)
apart (11). A spiral strip of steel was inserted to act as a gas turbulator
(Figure A-13). It was mounted after the muffler on a 1968 Chevy Carryall
(4x4).
Although simplicity of construction was a major advantage, this condenser
appeared to be undersized since it failed to remove any substantial amounts
of visible exhaust vapor either at idle or higher speeds.
The ingoing exhaust temperature was measured at 96°C (205°F). The
outgoing exhaust temperature was 39°C (102°F) when the ambient air was -20°C
(-5°F). This value of exhaust gas temperature was too high to condense most
of the water vapor.
Construction and installation cost amounted to approximately $250. Since
the unit was custom made and parts were acquired locally, it is speculated
that commercially built cooler-condensers of this type could be produced for
approximately $80.
The condenser was used in service for a period of 30 days; approximately
1300 miles were accumulated. It was removed and inspected for corrosion and
any buildup of deposits. No pitting or other acidic damage was evident. The
interior wall and turbular strip had a uniform soot deposit. The aluminum
pipe had an anticipated service life of about three years. Recommendations
were made for construction using light-weight stainless steel.
LOUVERED SHELL COOLER-CONDENSER
A cooler-condenser fabricated out of 1.2 cm (1/2 in.) EMT (12) was de-
signed by Simplex-Standard to meet the requirement for a 1968 IHC Scout
76
-------
INLET
TWISTED
TAPE INSERT
40 ALUMINUM FINS
DETAIL
Figure A-13.
Finned pipe cooler-condenser on
1968 Chevrolet Carryall (4x4).
77
-------
equipped with a V8 266 CID engine. For simple construction there was no
welding or machining of parts (Figure A-14). The condenser was designed so
that the outlet header could easily be removed to allow for cleaning and
deposit removal inside the tubes. Mounting was easily accomplished by
bolting it onto the front bumper and running a 5 m (15 ft.) piece of flex
exhaust hose from the tail pipe to the inlet header.
Air flow across the tubes was controlled from inside the cab by a
mechanical cable attached to six adjustable baffles in front of the tubes.
Freezing of the two bottom tubes became a problem which was solved by blocking
the air flow across them. Materials for the condenser cost approximately
$85. Roughly 40 hours were required to fabricate and install the condenser.
The unit was tested for five weeks in daily routine traveling. Ambient
temperatures ranged from -7°C (20°F) to -37°C (-35°F) and at speeds from zero
to 72 km/h (45 mi/h). The average temperature drop across the condenser was
approximately 10°C (50°F). Inlet temperatures ranged from 54°C (130°F) under
heavy load to 36°C (98°F) at idle. Outlet temperatures were from 28°C (84°F)
to 6°C (42°F), respectively. Low inlet temperatures were explained in part
by the heat loss through the 5 m (15 ft. ) of exhaust flexhose.
An inspection of the tubes after five week's use revealed a small accu-
mulation of scale and powder residue inside the tubes and headers. From
this, an anticipated life of three to four seasons could be assumed by using
the most inexpensive materials. A considerably longer seasonal life could be
expected by using corrosion resistant alloys.
78
-------
5cm(2in.
INLET
(38 in}
fOcm
f4-.n0
23 each l/2in. EM'
30cm
OUTLET
6 each 16 gage
LOUVERS
control cable
Figure A-14.
Louvered shell cooler-condenser on 1968
International Scout
79
-------
APPENDIX B
CALCULATION OF CONDENSATION CURVE.AND HEAT EXCHANGER DUTY
Gasoline is a mixture of many hydrocarbons: it contains butene C4H8 at
the lighter end, and decane C10H22 at the higher boiling, heavier end. The
mass ratio of carbon to hydrogen C:H in commercial gasolines varies from 6:1
to 6.8:1.
For this example a gasoline with a C:H of 6.5:1 will be used, its em-
pirical chemical formula is C7H13.
A complete combustion chemically balanced equation with air is:
Fuel Air Exhaust
Form:C7H13 + 10.25 02 + 38.56 N2 •* 7.00 C02 + 6.50 H20 + 38.56 N2 Ibs.
Ibs: 97 + 328 +1080 = 308 +117 + 1080
Since there are no carbon monoxide or hydrocarbons in the exhaust, the
equation assumes complete combustion. Well tuned engines approach complete
combustion; therefore, their exhaust water concentration approximates the
above equation. The exhaust water vapor [H20 , ,.] is the ice fog.
The above equation is based on an air to fuel weight ratio (A/F) of
(328 + 1080)/ 97= 14.5:1, which is called STOICHIOMETRIC.
Most gasoline fueled automobiles operate at A/F between 12:1 and 16:1.
For purposes of sizing an exhaust gas cooler-condenser the heat contents and
water vapor condensation will be based upon the combustion of one pound of
gasoline. At the stoichiometric ratio it will yield 15.5 pounds of exhaust
gas of the following composition: 3.18 Ib. C02, 1.21 Ib. H20 and 11.14 Ib. N2.
The reference temperature for heat content of the noncondensables - C02
and N2 is 60°F. For H20, 32°F is used.
Table B-l Heat Content at 3QO°F, 149°C
Component BTU/1b°F Ib AT°F BTU
BTU/lb°
0.
0.
4.
215
249
46
F
X
X
X
3
11
1
15
Ib
.18
.14
.21
.5
X
X
X
AT°F
240
240
268
=
=
C02 0.215 x 3.18 x 240 = 164
N2 0.249 x 11.14 x 240 = 667
H20( ^ 4.46 x 1.21 x 268 = 1447
80
-------
As the exhaust gas is cooled, the water vapor starts to condense at the
dew point temperature. The dew point is reached when the vapor pressure (VP)
of water equals its partial pressure (mole fraction x total pressure).
Assume the pressure drop in the condenser to be 0.4 psi. Since at 500 feet
above sea level the atmospheric pressure is 14.5 psia then total presure (P)
half way through the condenser is 14.7 psia.
Dew point temperature (Td ) is temperature when:
VP = P x H9°
VP = (14.7) x
C02 + H20 + N2
6.50
7.00 + 6.50 + 38.56
= (14.7) (0.125)
= 1.84 psia
From steam table: the temperature at which VP = 1.84 psia is 123°F;
Table B-2 Mole weight and heat content @ 123°F, 51°C
therefore, T, = 123°F.
dp
Component Ib./mole wt.
C02 3
N2 11
H20(v) 1
Totals @ 123°
Mole weight: 15. 5
.18 =
44
.14 =
28
.21
F
- 9ft !•
moles
0.
0.
0.
0.
) 11-
.072
.398
.067
.537
i /Th mn 1 o
BTU/lb°F
0.2047 x 3
0.2488 x 11
12.25 x 1
1
Ib
.18 x
.14 x
.21 x
5.5
AT°F BTU
63 = 41
63 = 174
91 = 1349
1564
U.
Moles H20, , _ (C02 + N2) VP
At 60°F Ib H20(v) = 18 x moles H20(y) = (°" (^"^256)''"' = °"149
H20,, is liquid water (condensate)
H20(L) = H20 (total) H20(v)
therefore H20(L) =1.21-0.149-1.06
Amount crrrinnsed is 1.06 ,nn _
T-TF x IUO -
81
-------
Component
C02
Table B-3 Heat content @ 60°F, 16°C
BTU/lb°F Ib AT°F
H20
H20
(v)
(L)
0.20
0.24
38.9
1.01
x 3.18
x 11.14
x 0.149
x 1.06
x
x
X
0
0
28
28
BTU
0
162
30
192
For brevity many calculation steps have been omitted. The results of
the above computations have been plotted on Figures 1 and 2 (pages 9 and 20).
Similar calculations were performed for exhaust from diesel and propane
fueled engines in order to draw their respective condensation curves.
The diesel is the only piston engine that normally operates with excess
air. This excess air results in a lower exhaust moisture concentration;
hence a larger fraction will remain in the vapor phase at any given tempera-
ture when compared to exhausts with no excess air.
82
-------
APPENDIX C
HEAT EXCHANGER DESIGN TECHNIQUES
An automotive ice fog cooler-condenser is simply a heat exchanger with
exhaust gas on the hot side and ambient air or other coolant on the cold
s i de.
Heat exchangers are specified (sized) by the amount of surface area
(square feet) required to adequately transfer the heat. For forced convection
heat transfer, the surface, A, is related to: the amount of heat to be
transferred, Q; the temperature difference, AT; and the overall heat transfer
coefficient, U. This realtionship can be expressed thusly:
Transfer surface - ft.2, A = j. ,
Where: the units of Q (the exchanger duty) are BTU/hr.,
the units of AT, , (the log mean temperature difference-LMTD)
are °F L
the units of U are BTU/hr.ft.2°F.
Since U is the overall transfer coeficient, its reciprocal 1/U is the overall
thermal resistance.
To size the exchanger (A) one needs to know the duty, the heat transfer
coefficient, and the temperatures. First the exchanger duty Q will be calcu-
lated. It depends upon the exhaust gas flow rate and temperature. A ve-
hicle's exhaust flow is directly related to its fuel economy. For example,
say at 40 miles per hour a vehicle gets 20 miles per gallon and its exhaust
temperature is 500°F. At -25°F the density of gasoline is 6.4 pounds per
gallon; the gasoline consumption rate is therefore:
6.4 Jb x Iflal. ml = 1b gasoline
gal 20 mi hr hr
RTl)
From Figure 1 the heat content of the exhaust is 3280 yr p—
at 500°F and 80 1b gasoline at 40°F' Therefore the dutV. Q is 12-8 lb/nr- x
3200 BTU/lb = 41,000 BTU/hr. Note that at 40°F approximately 94 percent of
the exhaust water vapor has been condesnsed out. The calculations for making
Figure 1 are in Appendix B.
83
-------
Next, consider heat transfer through a metal tube wall. The overall
heat transfer coefficient U needs to be calculated. Its reciprocal, the
overall thermal resistance 1/U, is estimated from empirical data. 1/U is the
sum of several series resistances, expressed in the following formula:
1 AQ xAQ 1 1
Where: A. = inside surface area for heat transfer.
A1 = outside surface area for heat transfer.
A° - (A + A.)/2
ave ^ o i
h-j = inside film coefficient for heat transfer.
h = outside film coefficient for heat transfer.
h° = fouling film coefficient for heat transfer.
k = thermal conductivity of the tube wall.
x = thickness of the tube wall.
The reciprocal 1/h etc. are the respective thermal film resistances. The
term xA /k A , which is the thermal resistance of the metal wall is negli-
o ave
gible compared to the other resistances and therefore can be ignored.
For the thin wall tubes A /A. will be about 1.0. The film conductance
depends upon film composition, surface roughness, temperature, and fluid
velocity. The individual film coefficients are estimated by procedures
detailed in the Engineering Manual, Reference 13 (pages 2-65,70., tables
2-14, 15,. case 11 and 20). Assuming an average exhaust temperature of
200°F, the inside film base factor is 4.1. A gas velocity of 58 ft/sec
yields a correction factor of 2.3 for 1/2 in. diameter tubes. Therefore,
h. = 4.1 x 2.3 = 9.4.
Assuming ambient air temperature of 15°F, the outside film base factor
is 7.7. An air velocity of 40 mph yields a correction factor of 3 for 1/2
in. diameter tubes. Therefore, h = 7.7 x 3 = 23.
Lead and soot deposits will foul the inside surface, therefore a fouling
film coefficient (hf) of 500 is used. Substituting the coefficients into the
above formula:
1/U = 1/9.4 + 1/500 + 1/23
= 0. 1064 + 0.002 + 0.0435
= 0.1519
Therefore U = 6.6 BTU/hr-ft2-°F
Heat transfer by radiation is neglected.
Last, it is necessary to calculate the log mean temperature difference
(LMTD). Assume the condenser is to work at ambients of -15°F or less at
40 mph. So much air will be flowing across the condenser that the air tem-
perature will increase less than 5°F.
84
-------
= Greatest temp, diff. (GTD) least temp, diff. (LTD)
log GTD/LTD
e
Therefore LMTD - [500 - (-15)] [40 - (-15)] _ 460 _ 206°F
mererore, LMIU - -
Now the surface area required for heat transfer can be calculated:
A = Q - 41,000 _ 2
H U (LMTD) 6:7 (206) ~ JU Tt
In evaluating the overall coefficient it can be seen that the largest
resistance is the inside film, 1/h.. If the inside surface and velocity were
increased by adding a metal twisted tape insert in each tube or if other tur-
bulators were added, then the effective inside heat transfer coefficient (h.)
based on the outside surface could be increased to about 18 BTU/hr-ft2-°F.
The resulting overall coefficient and required transfer area would then
be:
U = 1/18 + 1/500 + l/23= 9'9
A = 41.°00 - 90 ft2
M 9.9 (206) ~ ZU Tt
The back pressure drop also needs to be calculated. The condenser
cannot practically be removed for higher speed highway driving. The design
data for back pressure at 55 mph are 12 mi/gal and 500°F exhaust:
Exhaust flow = 6.4 Ib gasoline/gal x 1 gal/12 mi x 55 mi/hr x
15.5 Ib exhaust/lb gasoline = 455 Ib/hr = 0.13 Ib/sec.
Using the ideal gas law for density (reciprocal specific volume, 1/v ):
Pressure (atm) x mole wt. (yr - ?— )
1/V = ID mo i e
s Gas constant (0.73) x temp. (°F + 460)
1.0 (28.9)
s 0.73 (960)
The condenser flow area will be the same as the most common tail pipe,
generally around two inches in diameter. Its flow area is 0.022 ft2. The
exhaust gas velocity is:
ft3 1
0.13 Ib/sec x 1/0.041 j^- x Q Q22 ft'2 = 144 ft/sec-
Assume the cooler-condenser tubes will be designed so that the 144 ft/sec
velocity will not be exceeded. It will have two sharp (90°) orifice bends,
giving a pressure drop of 1.9 velocity heads each. Assume those plus other
losses to total 5 velocity heads (K = 5).
85
-------
g = gravitational constant, 32 ft/sec2.
2
v .
Pressure drop = K |- = 5 = 1600 ft. of gas
= 1600 ft x 0.041 lb/ft3 = 67 lb/ft2 = 0.46 psi
= 13 inches of H20
This addition to the normal exhaust system, whould be tolerated since
normal back pressures may run 30 to 50 inches H20 at 50 to 60 mph. Actual
total back pressure will be less when the cooler-condenser is used in lieu of
the muffler.
86
-------
APPENDIX D
ESTIMATION OF ROAD ICING FROM ICE FOG CONTROLS ON AUTOMOBILES
The condensate from the cooler-condensers will most probably drop to the
road surface and form ice. In some cases the condensate droplets may freeze
before striking the road and roll off to the side out of the traffic lanes.
The amount of ice (condensate) formed depends upon four variables:
fuel economy - miles per gallon
cooler-condenser efficiency - percent water condensed
traffic density - vehicles per day
traveled road width - feet
The following ice accumulation calculations are for two cases, one for a
straight section of road where the average fuel economy is 16 miles per
gallon and the other a heavily traveled intersection where the fuel economy
is 8 miles per gallon.
CASE 1
A two lane road with a traveled width of 40 feet, a traffic density of
12,000 vehicles per day, all equipped with cooler-condensers that condense
out 85 percent of the exhaust water vapor. The average condensate yield is
then:
X
1 gal gas 1 mile 6.4 1b gas 1.21 1b H20 1 ft H20
16 veh mile X 5280 ft gas gas Ib gas 62.4 lbH20 40 ft
^
12,000 x 12 ines x 30 x 0.85 efficiency x 4 months = 0.54 in.
CASE 2
A heavily traveled intersection such as the College Road, University
Avenue intersection which is loaded at 25,000 vehicles per day. It had three
four-lane roads, each 60 feet wide, and one two-lane road 40 feet wide.
Total width is therefore 220 feet. Because of the lower average speed at an
intersection, the cooler-condenser efficiency will be 90 percent.
87
-------
3
1 gal gas 1 mile 6.4 1b gas , ?1 1bH20 1 ft H20 1_
8 veh mile x 5280 ft x gal gas X LZI Ibgas x 62.41t>H20 x 220 ft
25 000 ^ x 12 x 30 -- x 0.9 efficiency x 4 month =.0.42 inch
' day ft montn
For comparison, consider what Mother Nature puts on the ground ^during
the four winter months which have average temperatures of less than 5°F. The
precipitation is, in inches of water equivalent:
November: 0.69
December: 0.59
January: 0.90
February: 0.49
2.67
For the two cases the condensate on the road in relation to precipi-
tation wi11 be:
CASE 1 CASE 2
Condensate _ 0.54 x 100 _ 2Q% 0.42 x 100 _
(Nov. - Feb.) Precipitation 2.67 2.67
Therefore, if the cooler-condensers were in common use the additional on-the-
road precipitation over the natural accumulation would be approximately 16 to
20 percent.
It is a conservative assumption that the condensate would spread out
evenly over the total width of the road. If, on the other hand, the accumu-
lated ice formed ridges between the tire tracks it could easily be removed by
snow plow, particularly in the left-hand turn lane where the greatest amount
of ice accumulates.
88
-------
APPENDIX E
CO
EXAMPLE ILLUSTRATING USE:
Air that is just forming ice fog,saturated at -32°C is heated to -18 C.
How much moisture can it accept before reforming ice fog (resaturated)?
SOLUTION:
At -32°C the saturated humidity ratio is 0.0002 gmHjO/gm dry air.
At -18°C the saturated humidity ratio is 0.0008 gmHjO/grn dry air.
Therefore 1.0002KG of saturated air warmed from -320C to -100C can
accept 0.0006KG of addition moisture before becoming super-
saturated (forming ice fog).
Chart from Figure 3, McFadden, T., Ice Fog Suppression -
A Review of Techniques, The Northern Engineer 7(4) :29
University of Alaska, Fairbanks, Alaska.
RELATIVE
HUMIDITY
100%
-20 -15 -10 -5
Dry Bulb Temperature (°C)
0.0020
0.0005
£
a
a>
o.
o
Q.
o
a <
£
o
0.0010 -2
o
cr
"E
X
Figure E-l. Low Temperature Psychrometric Chart (Metric Units)
-------
TECHNICAL REPORT DATA
(Please read Ii-islntctions on the reverse before completing)
1 REPORT NO. J2.
EPA-6QQ/3-78-055 J
4 TITLE AND SUBTITLE
Research on Control Technology for Ice Fog from Mobile
Sources
7. AUTHOR(S)
Harold J. Coutts and Ronald K. Turner
9. PERFORMING ORGANIZATION NAME AND ADDRESS
U. S. Environmental Protection Agency
Arctic Environmental Research Station
College, Alaska 99701
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Con/all is Environmental Research Laboratory
200 S. W. 35th Street
Corvallis, Oregon 97330
3. RECIPIENT'S ACCESSI ON- NO.
5. REPORT DATE
Mav 1978
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
1AA602
11. CONTRACT/GRANT NO.
13. TYPE OF REPORT AND PERIOD COVERED
inhouse
14. SPONSORING AGENCY CODE
EPA/600/02
15. SUPPLEMENTARY NOTES
(16. ABSTRACT
Automotive generated ice fog is a form of air pollution that results when exhaust
water vapor freezes into minute particles which form a dense fog.
The major control technique evaluated was cooling the exhaust gases to well
below the dew point, thus condensing water vapor into a liquid stream before final
discharge.
During the winters of 1974-75 and 1975-76 the Arctic Environmental Research Sta-
tion evaluated 12 cooler-condensers on nine inservice vehicles. It was found that ice
film formation decreased heat transfer efficiency. An ice fog mass emission reduction
up to 80 percent was attained with cooler-condensers on motor vehicles. However, the
increase in visibility over roads was not proportional because of the many other ice
fog sources. The overall impact of automotive ice fog control would be a visibility
increase of at least 70 percent in areas where motor vehicles create 50 percent or
more of the ice fog.
Control of automobile-generated ice fog would also mean cleaner air, but perhaps
more ice on the road. Cleaner air would result because sulfur oxides and lead
compounds would be absorbed in the condensate. This condensate, if allowed to drip
from the cooler-condensers, would freeze onto the road and require a more intense
snow removal effort.
17. KEY WORDS AND DOCUMENT ANALYSIS
]a DESCRIPTORS
1 Automotive Emission Control
| Low Temperature Air Pollution
! Ice Fog Control
)
,1
•1i DISTRIBUTION STATEMENT
' Release Unlimited
b. IDENTIFIERS/OPEN ENDEDTERMS
19. SECURITY CLASS (This Report)
unclassified
20. SECURITY CLASS (This page)
unclassified
c. COSATI Field/Group
21. NO. OF PAGES
98
22. PRICE
90
U. S GOVERNMENT PRINTING OFFICE: I978—797-308/I97 REGION 10
------- |