SAE TECHNICAL
`PAPER SERIES
`
`2000-01-0895
`
` Studies on Carbon Canisters to Satisfy LEVII
`EVAP Regulations
`
`Hideaki Itakura, Naoya Kato and Tokio Kohama
`Nippon Soken.Inc.
`
`Yoshihiko Hyoudou, Toshimi Murai
`Toyota Motor Corp.
`
`Reprinted From: LEV–II Emission Solutions
`(SP–1510)
`
`SAE 2000 World Congress
`Detroit, Michigan
`March 6-9, 2000
`
`Tel: (724) 776-4841 Fax: (724) 776-5760
`400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A.
` MAHLE-1028
`U.S. Patent No. RE38,844
`
`

`

`The appearance of this ISSN code at the bottom of this page indicates SAE’s consent that copies of the
`paper may be made for personal or internal use of specific clients. This consent is given on the condition,
`however, that the copier pay a $7.00 per article copy fee through the Copyright Clearance Center, Inc.
`Operations Center, 222 Rosewood Drive, Danvers, MA 01923 for copying beyond that permitted by Sec-
`tions 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying such as
`copying for general distribution, for advertising or promotional purposes, for creating new collective works,
`or for resale.
`
`SAE routinely stocks printed papers for a period of three years following date of publication. Direct your
`orders to SAE Customer Sales and Satisfaction Department.
`
`Quantity reprint rates can be obtained from the Customer Sales and Satisfaction Department.
`
`To request permission to reprint a technical paper or permission to use copyrighted SAE publications in
`other works, contact the SAE Publications Group.
`
`All SAE papers, standards, and selected
`books are abstracted and indexed in the
`Global Mobility Database
`
`No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise, without the prior written
`permission of the publisher.
`
`ISSN 0148-7191
`Copyright © 2000 Society of Automotive Engineers, Inc.
`
`Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE. The author is solely
`responsible for the content of the paper. A process is available by which discussions will be printed with the paper if it is published in
`SAE Transactions. For permission to publish this paper in full or in part, contact the SAE Publications Group.
`
`Persons wishing to submit papers to be considered for presentation or publication through SAE should send the manuscript or a 300
`word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board, SAE.
`
`Printed in USA
`
`

`

` Studies on Carbon Canisters to Satisfy LEVII EVAP Regulations
`
` 2000-01-0895
`
`Hideaki Itakura, Naoya Kato and Tokio Kohama
`Nippon Soken.Inc.
`
`Yoshihiko Hyoudou, Toshimi Murai
`Toyota Motor Corp.
`
`Copyright © 2000 Society of Automotive Engineers, Inc.
`
`ABSTRACT
`
`INTRODUCTION
`
`Recently, the California Air Resources Board (CARB) has
`proposed a new set of evaporative emissions and “Useful
`Life” standards, called LEVII EVAP regulations, which are
`more stringent than those of the enhanced EVAP
`emissions regulations.
`If
`the new regulations are
`enforced, it will become increasingly important for the
`carbon canister to reduce Diurnal Breathing Loss (DBL)
`and to prevent deterioration of the canister. Therefore,
`careful studies have been made on the techniques to
`meet these regulations by clarifying the working capacity
`deterioration mechanism and the phenomenon of DBL in
`a carbon canister.
`
`It has been found that the deterioration of working
`capacity would occur if high boiling hydrocarbons, which
`are difficult to purge, fill up the micropores of the
`activated carbon, and Useful Life could be estimated
`more accurately according to the saturated adsorption
`mass of the activated carbon and the canister purge
`volume. As a result, it is presumed that a more
`adaptable, longer Useful Life can be realized by providing
`a sufficient purge.
`
`It has been also found that the butane diffusion in a
`carbon canister during vehicle parking which is loaded to
`the canister during the DBL test, is the main cause of
`evaporative emissions from the canister. To prevent such
`diffusion, it is effective to divide the carbon bed into
`separated segments and insert some “labyrinth” between
`such carbon beds. Compared with the conventional
`canister, the improved canister was able to reduce DBL
`by half._Furthermore it became clear that DBL is reduced
`to approximately 1/3 when the gasoline fuel vapor is
`loaded to the canister instead of butane, which is the
`main cause of DBL. It was also concluded that the
`evaluation method should be reconsidered to account for
`real world conditions.
`
`CARB has proposed for 2004, which is when both the
`enhanced EVAP emissions regulations and On-Board
`Refueling Vapor Recovery (ORVR) regulations will once
`be settled, a new set of LEVII regulations including a
`more stringent set of evaporative emissions and Useful
`Life standards as compared to those of the enhanced
`EVAP emissions regulations. In these regulations CARB
`also has required the evaporative emissions to be
`restricted to 1/4th of the parameter stated in the 1995
`regulations and the Useful Life to be extended from 10 to
`15 years.
`
`The sources of evaporative emissions generated in DBL
`test have already been investigated and it is clarified that
`a canister is one of the typical sources of the evaporative
`emissions. To meet the LEV II regulations, it has become
`more important for the canister to reduce DBL and to
`extend its useful life span.
`
`The useful life span of the canister depends on that of the
`activated carbon which adsorbs the hydrocarbons(HC).
`Therefore, several studies have been made on the
`deterioration of adsorption performance of the carbon
`canister withdrawn from the vehicles after long usage
`and the change in adsorption performance during the
`adsorption/desorption cycle
`tests
`in
`the
`laboratory.
`However, little research analyzing the relation between
`characteristics of activated carbon and its deterioration
`phenomenon has been done, and a general method for
`estimating canister life span has not been developed.
`
`Furthermore the effects on DBL by external factors such
`as temperature and the length of parking have been
`clarified, and this information has led to the formulation of
`the present Federal Test Procedure. Recent studies show
`the effects of the design factors of canister configuration
`and purge amount on DBL. However, detailed analysis
`on the adsorption states of activated carbon in the
`canister and on the mechanism of DBL have been few.
`
`1
`
`

`

`Therefore, the authors tried to clarify the mechanisms for
`canister life span and DBL. In addition, the authors
`demonstrated by a simple model, and tried to make clear
`a canister design that would comply with the LEVII
`regulations.
`
`EXPERIMENTAL METHODS
`
`The test equipment used in the experiment is shown in
`Figure 1. The equipment continuously measures the
`canister conditions, as recording the adsorption mass in
`the canister, the breakthrough concentration from the
`canister, fuel temperature, and fuel tank pressure, and
`then logs these data onto a computer file.
`
`Figure 1. Test Equipment
`
`DE TE R IORA TION AN AL YS I S – The adsorption/
`desorption cycles test procedure is shown in Figure 2. By
`using a 0.12 liter, small-sized canister enclosed with
`activated carbon, gasoline temperature was adjusted to
`35°C and gasoline fuel vapor was generated by bubbling
`dry air through gasoline in the fuel tank. 1.0-1.2g/min HC
`were loaded by breakthrough at a room temperature of
`25°C. Then the canister was purged with dry air at 10 liter
`per minute for 3.6 minutes(300 bed volume). This
`adsorption/desorption was carried out for 240 cycles.
`Reid Vapor Pressure_(RVP) of the gasoline used for this
`test was 62KPa, and it was exchanged with new gasoline
`every thirty cycles. The gasoline working capacity of the
`canister was prescribed by the amount of adsorption of
`each cycle. Residual HC components adsorbed in the
`activated
`carbon were
`extracted
`by
`using
`dichloromethane (CH2Cl2) as a solvent, and were then
`analyzed by gas chromatography.
`
`Figure 2. Test Pattern (Adsorption/Desorption Cycles)
`
`DBL ANALYSIS – The diurnal test procedure is shown in
`Figure 3, and the conditions of each process are as
`follows.
`
`fuel vapor
`1. Canister stabilization: The gasoline
`generated by bubbling dry air through gasoline was
`loaded by breakthrough to the 2.0 liter canister which
`enclosed activated carbon. Then the canister was
`purged with dry air at 20 liters per minute for 30
`minutes(300
`bed
`volume). This
`adsorption/
`desorption_was carried out for 6 cycles at 25°C.
`2. Canister
`loading and purging: Loading was
`conducted by 2 grams breakthrough using a 50/50
`percent by volume mixture of n-butane and nitrogen
`at a flow rate equal to 40g butane/hour. Then the
`purging was conducted with dry air at 20 liter per
`minute for 15 minutes(150 bed volume).
`3. Environmental conditions at parking: After soaking
`for 12 hours at 25°C, the canister went through two
`24-hour EPA diurnal temperature cycles.
`
`The canister was divided into 5 layers, as shown in
`Figure 4, and the activated carbon in each layer was
`removed. After measuring the adsorption mass of the
`activated carbon
`removed
`from each
`layer,
`the
`components adsorbed in the activated carbon were
`analyzed by gas chromatography.
`
`Figure 3. Diurnal Test Procedure
`
`Figure 4. Canister Configuration Used in Diurnal Test
`
`CANISTER PERFORMANCE DETERIORATION
`MECHANISM
`
`After long usage, canister performance may occasionally
`show reduced working capacity which could sometimes
`result
`in aggravated DBL as well. This
`is called
`“deterioration”. In this experiment, the canister was
`repeatedly
`loaded and purged. The
`results of
`compositional analysis of the fuel and its vapor applied in
`this adsorption test are shown in Figure 5.
`
`2
`
`

`

`Figure 6 shows the transitions of the adsorption mass. In
`the first cycle of adsorption/desorption, approximately 1/3
`of the adsorption mass that was unable to desorb
`remained after purging. The residual HC continued to
`climb after every adsorption/desorption cycle, but after
`the 10th cycle, this trend leveled out considerably. The
`adsorption mass at such time was also found to be
`stable. After such stable conditions continued for a while,
`the adsorption mass began to reduce. This is called a
`“deterioration phase”. This entire transition span of
`working capacity from the beginning of a stable state to a
`definite deteriorating state is counted as one span of the
`“Useful Life”.
`
`Figure 7 shows the transitions of mass and components
`of the residual HC after purging in the canister. The
`residual HC in the initial aging phase was mostly
`dominated by C4,C5,C6. Likewise, in the stabilized
`phase, all such components were replaced with high
`boiling hydrocarbons of C7 and over for a long time. In
`the deterioration phase, C4, C5 were not apparent in the
`residuals, but hydrocarbons of C7 and over continued to
`increase.
`
`Figure 8 shows the transitions of pore conditions of the
`activated carbon. In the initial aging phase, micropores
`over 13 angstroms decreased whereas
`those of
`approximately 9 angstroms to 11 angstroms increased.
`In the stabilized phase, the pores of 9 angstroms to 11
`angstroms continued to decrease. The decrease in
`micropores of larger size was small. In the deterioration
`phase, micropores of the size equivalent to hydrocarbon
`molecules in the gasoline fuel vapor began to decrease.
`
`Figure 6. Transition of Working Capacity during
`Adsorption/Desorption Cycles
`
`Figure 7. Transition of Mass and Components of
`Residual HC in Canister
`
`Figure 5. HC Components of Fuel and Vapor Applied in
`Adsorption Cycle
`
`Thus, based on the above-mentioned results, it can be
`concluded that deterioration is caused when high boiling
`hydrocarbons of C7 and over begin to accumulate in, and
`fill up those pores effective for adsorption, subsequently
`causing the working capacity to drop. (See Fig.9)
`
`Figure 8. Transition of Pore Conditions in Activated
`Carbon
`
`3
`
`

`

`Figure 9. Activated Carbon Deteriorating Mechanism
`
`the above-
`USEFUL LIFE ESTIMATION – Based on
`mentioned results, the canister adsorption performance
`is assumed to deteriorate when the residual mass of
`the_high_boiling_hydrocarbons
`reaches a
`specific
`amount. Such mass is proportional to the saturated
`adsorption mass_which is the adsorption mass in the
`case that all pores are filled by gasoline. According to
`Figure 7, on completion of Useful Life (140 cycles), the
`residual mass of high boiling_hydrocarbons was found to
`be 8.2g/100ml carbon. Depending on the pore volume
`per 100ml and the test gasoline density, the saturated
`adsorption mass was calculated to be 26.5g/100ml
`carbon. Therefore, the residual rate of 0.30(8.2/26.5) was
`assigned as the standard value of Useful Life. The rate of
`accumulated pore volume of micropores under 20
`angstroms in activated carbons used for automotive
`canisters is almost 30%.(Fig.10) Therefore, it is assumed
`that the deterioration occurs when the high boiling
`hydrocarbons fill up the micropores equal to those
`molecule size.
`
`Figure 10. Relationship between Pore Diameter and
`Accumulated rate of Pore Volume
`
`Figure 11. Purge Influence on the Residual HC
`
`Saturated adsorption mass varied according to types of
`activated carbon, and purge amount influenced the
`residual mass of high boiling hydrocarbons. The
`temperature change of the canister installed in close
`proximity to the fuel tank was not large enough to
`influence
`the desorption ability of high boiling
`hydrocarbons of C7 and over, but the purge amount was
`noteworthy.
`In Figure 11,
`the purge amount
`influence_to_the_residual_high boiling_hydrocarbons in
`the stabilized_phase is shown. It became clear_that 90%
`of the high boiling hydrocarbons of C7 and over could be
`desorbed by increasing number of purge bed volumes.
`
`A_formula for estimating the canister Useful Life was
`devised. The flow mass of C7 and over for each refueling
`is Vr(g), and the flow mass of C7 and over flowing into
`the canister in the state of driving and parking, until the
`next refueling is Vd(g). R(P) is the rate of residual mass
`of C7 and over according to the purge amount P as
`indicated in Figure 11, and is set at 0.1. If Q(L)
`represents carbon volume, then the residual mass Rm (g/
`100ml carbon) of C7 and over upon one refueling can be
`expressed as the following:
`
`(1)
`
`4
`
`

`

`Provided that a given vehicle can travel a distance of 300
`miles after
`refueling, by
`the point
`the vehicle
`accomplishes the designated DD (driving distance)
`miles, it will have been refueled DD/300 times. Even if the
`residual mass of components of C7 and over at such
`time-point has not reached the standard value of Useful
`Life, the canister Useful Life is guaranteed all the way up
`to such driving distance.
`
`(2)
`
`Following this estimation formula, an adsorption mass of
`the canister in every refueling cycle that would guarantee
`a Useful Life after 150,000 miles of driving was deduced.
`The factors for this calculation are presented in Table 1.
`Compared with
`the previously mentioned activated
`carbons, the pore volume on these activated carbons
`was increased by 30%.
`
`Calculations show that (Vr+Vd) should be under 5.1
`grams. As shown in Figure 5, since the proportion of
`components of C7 and over in the gasoline fuel vapor is
`6.6%, the adsorption mass of the canister at the refueling
`interval should be under 78(5.1/0.066) grams. The
`calculated value(5.1g) requires a purge amount in excess
`of 2000 bed volume. This is equivalent to a purge amount
`of 6.7 bed volume per 1 mile. Based on such results, it is
`clear that by guaranteeing a reasonable purge amount,
`canister life can be extended, or in other words,
`deterioration can be more restrained.
`
`Next, in order to verify the DBL reduction effects of the
`anti-deterioration efforts, DBL of a canister in the Useful
`Life phase and another canister in the deteriorating
`phase were compared. The amount of DBL from the
`canisters are shown in Figure 12. Compared with the
`canister in the Useful Life phase, the canister in the
`deterioration phase proves to have two days worth of
`DBL. Furthermore, DBL of the second day exceeded the
`amount of the first day. As a result, it was confirmed that
`restrained deterioration leads to greater assurance of
`DBL during Useful Life.
`
`5
`
`Figure 12. Comparison od DBL Performance of Canister
`in Stabilized Phase and in Deterioration
`Phase
`
`CANISTER DBL MECHANISM
`
`Figure 13 shows an example of DBL measurement and
`results that are apparent under a condition lower than the
`working capacity of the canister. The weight of the
`canister increased when the fuel temperature rose during
`the diurnal cycle. On the contrary, when the fuel
`temperature dropped, the weight of the canister reduced.
`This is because the fuel tank breathes according to the
`change in the fuel temperature. The total vapor flow mass
`upon two day diurnal was 52g. In spite of the fact that it
`was lower in capacity than the 80g working capacity of
`the canister, DBL already occurred in the first cycle. An
`interesting point is that evaporative emissions occurred
`only upon rising of the fuel temperature. Figure 14 shows
`the composition of the evaporative emissions. They were
`of low boiling points under gasoline fuel vapor such as
`C4and C5.
`
`In order to clarify the causes of such phenomenon, the
`changing trends of HC mass and composites inside the
`canister were examined. The results are shown in
`Figures 15 and 16. Low boiling hydrocarbons of C4 and
`C5 have been found to diffuse slightly on the outlet end of
`the canister under the state of soak and in the diurnal
`cycle performed inside the canister.
`
`Thus, based on the above results, the DBL mechanism
`can be explained as
`in Figure 17. Low boiling
`hydrocarbons, which adsorbed on the canister inlet side,
`diffuse to the outlet side of the canister during soak and
`during the diurnal cycle. The hydrocarbons in the
`gasoline fuel vapor, which flow into the canister upon
`rising of the fuel temperature in the diurnal cycle, adsorb
`on the canister inlet side. Therefore only the air flows into
`the outlet side of the canister. By this air, low boiling
`hydrocarbons are desorbed, and thus leakage occurs.
`
`

`

`Figure 13. Temperature, HC Mass, and Breakthrough
`HC Concentration Profiles during Two Day
`Diurnal
`
`Figure 16. Transition of HC Mass and Components of
`Each Layer during First Diurnal
`
`Figure 14. Breakthrough HC Components
`
`Figure 15. Transition of HC Mass and Components of
`Each Layer during Soak
`
`Figure 17. Mechanism of Diurnal Emissions from
`Canister
`
`6
`
`

`

`CANISTER DBL REDUCTION TECHNIQUES
`
`REDUCTION OF DBL BY IMPROVING CANISTER
`CONFIGURATION – It has been verified that the main
`factor causing DBL in the canister is caused by the in-
`canister diffusion of low boiling hydrocarbons such as C4
`and C5 that are present slightly in the residual HC after
`purge and in the HC adsorbed during the DBL test. For
`this reason, reducing the residual HC after purge and HC
`diffusion inside the canister have been found to be
`effective techniques for the DBL reduction. Because the
`amount of DBL increased along with the number of days,
`restraining
` HC diffusion
`inside
`the canister was
`particularly important. An equation explaining this
`diffusion is as follows:
`
`(3)
`(4)
`
`prepared. The length of the air bed was made to be
`variable. Only one carbon bed was adsorbed with HC,
`then aged afterwards inside a thermostatic chamber for a
`specific amount of time. Finally, the mass of the two
`activated carbon beds were measured. The results are
`shown in Figure 19. The HC migration to another carbon
`beds was found to be smaller when the air bed was
`longer and when the temperature was lower. At the same
`temperature, diffusion mass between carbons separated
`by an air bed of 20 cm was found to be only 1/4 the
`diffusion mass of carbons without an air bed.
`
`According to EQ 3, for preventing diffusion it is effective
`to minimize D, A, and ∂C/∂x. However, A should not be
`minimized because the pressure loss of the canister
`could be increased. Therefore, a low pressure loss and
`an anti-diffusion technique was developed. Figure 18
`shows the pattern of HC diffusion between activated
`carbons.
`
`Figure 19. Relationship between Length of Air Layer and
`Diffusion Mass
`
`Based on these results, the improved canister which has
`a labyrinth air bed was designed as taking into account
`the adsorption mass upon actual DBL test, and its DBL
`became 1/2 of a conventional canister. Such results are
`shown in Figure 20.
`
`Figure 18. HC Diffusion Pattern between Activated
`Carbons
`
`Therefore, even if a massive amount of HC is adsorbed in
`the micropores of activated carbon, provided that the HC
`concentration in the ambience of such activated carbon is
`low, it then becomes easy for the HC to desorb from the
`micropores. Furthermore, if the adsorption HC mass in
`other activated carbon micropores
`is small,
`then
`adsorption of HC in the ambient space by such pores
`becomes possible. In this way, the HC seemingly diffuses
`as if it migrates among the activated carbon micropores.
`
`In order to restrict such diffusion, activated carbon with
`large adsorption and that with small adsorption should be
`separated as much as possible. For confirmation, a
`canister with two activated carbon beds were separated
`by an air bed along the vapor flow direction was
`
`Figure 20. Comparison of DBL between Improved
`Canister and Conventional Canister
`
`REVISION OF DBL TESTING CONDITIONS – As seen
`in the diurnal test procedure (Fig. 3), there are several
`cycles of adsorption test prior to the DBL test. More
`officially speaking, this is done to adsorb the gasoline fuel
`vapor, but the butane loading is widely conducted as an
`optional method.
`
`It has been shown by this investigation that DBL occurs
`only because of the diffusion of low boiling hydrocarbons
`such as C4 and C5. For reference, in Table 2 the diffusion
`coefficients indicate how easily the diffusion of saturated
`hydrocarbons can take place. Based on the fact that the
`smaller the amount of carbon in components, the faster
`
`7
`
`

`

`the speed of their diffusion will be, it can be understood
`that the difference of HC adsorbed before the diurnal
`cycle can no longer be ignored. Only in the canister
`loading stage, the gasoline fuel vapor loading and the
`butane loading could be studied comparatively. All other
`adsorption in the canister stabilization stage had to be
`tried with the gasoline fuel vapor. As seen in Figure 21,
`the gasoline fuel vapor loaded canister showed 1/3 the
`DBL amount for 2 cycles compared to that of butane
`loaded canister. Also, with the gasoline fuel vapor loaded
`canister, the DBL fluctuation appeared to be more
`reduced in comparison with the butane loaded one.
`
`Table 2. Coefficient of Diffusion at Standard conditions
` (0°C, 1atm)
`
`Figure 21. Comparison of DBL between Butane Lording
`and Gasoline Fuel Vapor Lording
`
`Outside of the laboratory, DBL actually occurs after the
`gasoline fuel vapor loading, and such results are usually
`more emphasized. If a canister must be designed
`according to that result gained by employing a butane
`loading method, an extra performance margin ends up
`being created in order to compensate for that aggravated
`DBL and its fluctuations. To resolve this situation, one
`idea is to clarify the amount of such margin caused by the
`butane loading method, which consequently leads to the
`belief that gasoline fuel vapor loading will definitely
`guarantee higher precision than the butane method.
`
`CONCLUSION
`
`By a detailed investigation and analysis of carbon
`adsorption/desorption
`states,
`the
`deterioration
`mechanism and DBL mechanism were clarified.
`
`(cid:127) Deterioration of adsorption abilities is caused when
`the micropores effective for adsorption are closed
`due to accumulation of high boiling hydrocarbons of
`C7 and over which are present in low concentrations
`in the gasoline vapor.
`(cid:127) The main cause of evaporative emissions from the
`canister is butane-diffusion occurring inside the
`canister while the car is parked.
`
`Based on this mechanism a technology responding to
`extended Useful Life and enhanced evaporative
`emissions was clarified and the following points were
`acknowledged
`
`(cid:127) Deterioration of a current carbon canister can be
`restrained by increasing the purge amount and thus
`Useful Life can be extended
`(cid:127) To design a canister reasonably , canister loading in
`real world conditions must also be taken into account
`for evaluation.
`
`REFERENCES
`
`1. Harold M. Haskew and William R. Cadman,
`“Evaporative Emissions Under Real
`Time
`Conditions”, SAE Paper No. 891121,1989.
`2. Harold M. Haskew , William R. Cadman, and
`Thomas F. Liberty, “The Development of a Real-
`Time Evaporative Emission Test”, SAE Paper
`No.901110,1990.
`3. Harold M. Haskew , William R. Cadman, and
`Thomas F. Liberty,
`
`
`“Real-Time Non-Fuel
`Background
`Emissions”,
`SAE
`Paper
`No.912373,1991.
`4. J. E. Urbanic, E. S. Oswald, N. J. Wagner, and H. E.
`Moore,
`
`“Factors Affecting
`the Design and
`Breakthrough Performance of Evaporative Loss
`Control Systems for Current and Future Emissions
`Standards”, SAE Paper No.890621,1989.
`5. H. R. Johnson and R. S. Williams, “Performance of
`Activated Carbon
`in Evaporative Loss Control
`Systems”, SAE Paper No.902119,1990.
`6. Ames A. Grisanti, Ted R. Aulich, and Curtis L.
`Knudson,
` “Gasoline Evaporative Emissions –
`Ethanol Effects on Vapor Control Canister Solvent
`Performance”, SAE Paper No.952748,1995.
`7. Michael E. Payne, Jack S. Segal, Matthew S.
`Newkirk, and Lawrence R. Smith, “Use of Butane as
`an Alternative Fuel Emissions from a Conversion
`Vehicle Using Various Blends”, SAE Paper
`No.952496,1995.
`8. Philip J. Johnson, James R. Jamrog, and George A.
`Lavoie, “Activated Carbon Canister Performance
`During Diurnal Cycles: An Experimental and
`Modeling Evaluation”, SAE Paper No.971651,1997.
`
`8
`
`

`

`9. Philip J. Johnson, Roger J. Khami, Jeffrey E.
`Bauman, Thomas D. Goebel, Vernon L. Clark, David
`L. Hirt, and Paul J. Luft,
` “Carbon Canister
`Development for Enhanced Evaporative Emissions
`and
`On-Board
`Refueling”,
`SAE
`Paper
`No.970312,1997.
`10. Marek C. Lockhart, “Predicting Tank Vapor Mass for
`On-Board Refueling Vapor Recovery”, SAE Paper
`No.970308,1997.
`11. George A. Lavoie, Philip J. Johnson, and Jeffrey F.
`Hood, “Carbon Canister Modeling for Evaporative
`Emissions: Adsorption and Thermal Effects”, SAE
`Paper No.961210,1996.
`
`9
`
`