01.12.2012 Views

Effects of Low-Purge Vehicle Applications and ... - MeadWestvaco

Effects of Low-Purge Vehicle Applications and ... - MeadWestvaco

Effects of Low-Purge Vehicle Applications and ... - MeadWestvaco

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Copyright © 2007 SAE International<br />

ABSTRACT<br />

20077051<br />

2007-01-1929<br />

<strong>Effects</strong> <strong>of</strong> <strong>Low</strong>-<strong>Purge</strong> <strong>Vehicle</strong> <strong>Applications</strong> <strong>and</strong> Ethanol-<br />

Containing Fuels on Evaporative Emissions Canister<br />

Performance<br />

The California Air Resources Board (CARB) LEV II<br />

regulations require less than 500 mg <strong>of</strong> vehicle<br />

evaporative emissions which, typically requires canister<br />

emissions <strong>of</strong> less than 200 mg. PZEV regulations<br />

require less than 350 mg vehicle emissions <strong>and</strong> zero<br />

emissions (< 54 mg) from the fuel system.<br />

Activated carbon canister emissions <strong>of</strong> less than 20 mg<br />

are typically required in order to meet PZEV regulations.<br />

LEV II canister emissions levels can typically be<br />

achieved through a combination <strong>of</strong> canister design,<br />

sufficient purge volumes, use <strong>of</strong> the appropriate<br />

activated carbon or combination <strong>of</strong> carbons, <strong>and</strong> in some<br />

cases an activated carbon honeycomb.<br />

INTRODUCTION<br />

Activated carbon honeycombs are typically required to<br />

further reduce canister emissions to levels necessary to<br />

meet PZEV emissions regulations. Recent trends in<br />

automotive technology have led to a reduction in the<br />

amount <strong>of</strong> air available for regenerating or purging the<br />

carbon canister. The new technologies include hybridelectric<br />

vehicles <strong>and</strong> gasoline direct injection engines.<br />

Hybrid vehicles have reduced gasoline engine runtime<br />

<strong>and</strong> therefore use less intake air. The gasoline direct<br />

injection technology requires less throttle air to operate<br />

the engine. Under these reduced purge volume<br />

conditions, current activated carbon canister <strong>and</strong><br />

honeycomb configurations may not be adequate to meet<br />

PZEV or even st<strong>and</strong>ard LEV II requirements.<br />

A relationship between canister purge volume <strong>and</strong><br />

emissions has been developed. The data show that<br />

emissions for both the base canister <strong>and</strong> the canister<br />

with honeycomb attached increase significantly with<br />

decreasing purge volume. Relationships between<br />

activated carbon honeycomb properties <strong>and</strong> canister<br />

emissions have also been developed. This work<br />

Reid Clontz, Moe Elum, Peter McCrae, Roger Williams<br />

<strong>MeadWestvaco</strong> Corp.<br />

demonstrates that properties such as total part<br />

volume/capacity, part length, <strong>and</strong> diameter affect<br />

honeycomb performance. Manipulation <strong>of</strong> these<br />

properties led to improvements as compared to current<br />

part design. Honeycomb configuration was also<br />

investigated as a method <strong>of</strong> improving performance.<br />

Multiple honeycombs tested in series also showed<br />

improvements over current part design. These<br />

performance improvements enable conformance to LEV<br />

II <strong>and</strong> PZEV emissions st<strong>and</strong>ards under incrementally<br />

reduced purge volume levels.<br />

Activated carbon honeycombs were also resistively<br />

heated <strong>and</strong> evaluated under severely limited purge<br />

conditions. Test results indicate that resistively heating<br />

the honeycomb enables achievement <strong>of</strong> PZEV<br />

emissions levels from the canister under these restrictive<br />

conditions.<br />

Due to increased interest in <strong>and</strong> dem<strong>and</strong> for renewable<br />

fuels, the effects <strong>of</strong> ethanol content on the performance<br />

<strong>of</strong> activated carbon canisters <strong>and</strong> activated carbon<br />

honeycombs have been investigated <strong>and</strong> relationships<br />

developed. The effects <strong>of</strong> fuel vapor pressure on<br />

canister emissions were also investigated <strong>and</strong><br />

differentiated from the effects caused by ethanol<br />

content.<br />

As expected, the data showed that a strong relationship<br />

exists between diurnal vapor generation <strong>and</strong> fuel<br />

volatility regardless <strong>of</strong> ethanol content. Furthermore, the<br />

canister emissions obtained were similar when fuels <strong>of</strong><br />

equivalent volatility but varying ethanol content were<br />

utilized during a simulated diurnal emissions test. The<br />

results indicate that the use <strong>of</strong> a honeycomb can<br />

facilitate achievement <strong>of</strong> LEV II emissions levels from<br />

the carbon canister tested under CARB flexible fuel<br />

vehicle test conditions.<br />

In a previous paper, the impact <strong>and</strong> control <strong>of</strong> bleed<br />

emissions from evaporative emissions control devices<br />

were discussed (1). A technique was developed to


study the level <strong>of</strong> bleed emissions specifically from<br />

canister vent ports. Carbon type, canister design <strong>and</strong><br />

purge volume were shown to have a significant impact<br />

on bleed emissions. Further, the incorporation <strong>of</strong> a<br />

small auxiliary chamber in series with the primary<br />

canister was shown to decrease bleed emissions<br />

significantly (1).<br />

Activated carbon honeycombs, typically incorporated as<br />

separate auxiliary chambers, are currently being<br />

successfully used to meet LEV II <strong>and</strong> PZEV regulations<br />

in many vehicles. This technology has been shown to<br />

be an effective means to control bleed emissions,<br />

particularly in systems where large purge volumes are<br />

not available.<br />

Recent trends in vehicle technology have created<br />

questions concerning the performance <strong>of</strong> the<br />

evaporative emissions control canister system. <strong>Vehicle</strong><br />

technologies such as hybrid powertrains, direct fuel<br />

injection, variable engine displacement, <strong>and</strong><br />

homogeneous charge compression engine technology<br />

can have a significant impact on canister performance<br />

due to reduced volumes <strong>of</strong> air available for canister<br />

purge.<br />

In addition to new developments in power-train<br />

technology, increased interest in renewable fuels such<br />

as ethanol have also led to questions concerning the<br />

performance <strong>of</strong> activated carbon <strong>and</strong> activated carbon<br />

honeycombs.<br />

It should be noted that canister design can have a<br />

significant impact on evaporative emissions. The<br />

baseline canister used in this study was not a design<br />

optimized to control bleed emissions. Thus, the<br />

emissions results should primarily be taken on a relative<br />

basis. The purpose <strong>of</strong> this paper was not to investigate<br />

the effect <strong>of</strong> base canister design, but to optimize the<br />

use <strong>of</strong> the activated carbon honeycomb under reduced<br />

purge conditions <strong>and</strong> under conditions <strong>of</strong> exposure to<br />

fuels <strong>of</strong> different volatility <strong>and</strong> ethanol content.<br />

EXPERIMENTAL METHODS<br />

Bleed emissions were evaluated using a laboratory test<br />

method involving an initial canister gasoline preconditioning,<br />

butane load, purge followed by a timecontrolled<br />

soak, <strong>and</strong> emissions measurement during a<br />

final canister vapor load simulating a diurnal loading<br />

event.<br />

The canisters used for testing purposes were all the<br />

same design <strong>and</strong> contained 2.1 liters <strong>of</strong> BAX 1500<br />

carbon. The canisters were used multiple times for<br />

repeat testing. To ensure fuel blending was not a<br />

concern, each canister was cycled with the appropriate<br />

fuel unitl equilibrium was reached.<br />

GASOLINE VAPOR AND BUTANE CYCLING<br />

Gasoline vapor cycling was performed using automated<br />

cycle test equipment that precisely controlled <strong>and</strong><br />

monitored all testing conditions. Gasoline vapors were<br />

generated by bubbling air at a rate <strong>of</strong> 200 cc/min through<br />

2 liters <strong>of</strong> certified test fuel heated to 36°C. For<br />

example, a 62.1 RVP-certified test gasoline typically<br />

resulted in a vapor generation rate <strong>of</strong> approximately 40<br />

g/hr with a hydrocarbon concentration <strong>of</strong> approximately<br />

50% by volume. The generated vapors were sent to the<br />

canister until a breakthrough concentration <strong>of</strong> 5000 ppm<br />

was detected using a flame ionization detector (FID). If<br />

breakthrough was not detected after 90 minutes <strong>of</strong> vapor<br />

loading, the liquid gasoline was replaced with a fresh 2liter<br />

aliquot. After breakthrough was detected, the<br />

canister was purged for a specified volume, typically 400<br />

bed volumes, using dry air at a rate <strong>of</strong> 22.7 liters per<br />

minute. The dwell time between load <strong>and</strong> purge events<br />

was no more than 5 minutes. The ambient temperature<br />

<strong>of</strong> the equipment, including the canister storage<br />

compartment, was maintained at 20°C during all stages<br />

<strong>of</strong> the testing.<br />

DIURNAL TESTS<br />

The diurnal test procedure, referred to as the simulated<br />

real-time diurnal test, was used to subject a fuel tank<br />

<strong>and</strong> canister to the CARB diurnal temperature pr<strong>of</strong>ile<br />

within a temperature-controlled environmental chamber.<br />

This test procedure was published previously (1). The<br />

test was designed to simulate the diurnal portion <strong>of</strong> the<br />

CARB vehicle emissions test procedure in order to<br />

generate quantitative emissions data isolated to only the<br />

evaporative emissions canister.<br />

Pre-conditioning for the simulated real-time diurnal test<br />

included multiple gasoline vapor load <strong>and</strong> purge cycles<br />

as described above. Following the gasoline cycles, the<br />

canister was loaded with 50% butane vapor at a load<br />

rate <strong>of</strong> 40 g/hr to a 5000 ppm breakthrough as measured<br />

by an in-line flame ionization detector in preparation for<br />

a two-day diurnal test. After the butane load, the<br />

canister was allowed to soak for 60 minutes before<br />

purging with a specified volume <strong>of</strong> air. Following the<br />

purge, the canister was soaked at 20°C for 24 hours<br />

before starting the diurnal test.<br />

After the 24-hour soak, the canister was attached to a<br />

commercial 60-liter steel fuel tank containing 24 liters<br />

(40% fill) <strong>of</strong> the certified test fuel specified. The fuel<br />

temperature was equilibrated to 18.3 °C overnight before<br />

beginning the test. A Tedlar® bag was attached to the<br />

atmosphere port <strong>of</strong> the canister as shown in Figure 1 to<br />

collect the hydrocarbon emissions. During the 11-hour<br />

temperature ramp from 18.3° C to 40.6° C, the canister<br />

weight <strong>and</strong> emissions were measured several times.<br />

During the portion <strong>of</strong> the diurnal when the temperature<br />

decreased, the Tedlar® bag was removed in order to<br />

allow the system to back purge.<br />

During each emissions measurement, the Tedlar® bag<br />

was removed <strong>and</strong> filled to a known volume with nitrogen.


The hydrocarbon concentration was determined by<br />

evacuation <strong>of</strong> the bag contents into a flame ionization<br />

detector. Once the concentration <strong>and</strong> volume were<br />

determined, the mass <strong>of</strong> hydrocarbon was calculated<br />

<strong>and</strong> recorded. The total test procedure incorporating<br />

gasoline aging <strong>and</strong> diurnal testing takes approximately<br />

one <strong>and</strong> a half weeks per canister test. Due to time<br />

considerations, only a limited number <strong>of</strong> repeat tests<br />

were performed. Where possible repeats were<br />

performed <strong>and</strong> the nominal test value was reported.<br />

The entire two-day CARB diurnal temperature pr<strong>of</strong>ile<br />

was used for canister purge volumes <strong>of</strong> 75 bed volumes<br />

or greater. Figure 2 represents the CARB temperature<br />

pr<strong>of</strong>ile used for the environmental chamber containing<br />

the test canister. It should be noted that canister design<br />

can have a significant impact on evaporative emissions.<br />

The purpose <strong>of</strong> this paper was not to investigate the<br />

effect <strong>of</strong> base canister design, but to optimize the use <strong>of</strong><br />

the activated carbon honeycomb under reduced purge<br />

conditions. All <strong>of</strong> the testing was performed with a<br />

commercial 2.1-liter carbon canister filled with BAX 1500<br />

automotive grade carbon.<br />

RESISTIVELY HEATED ACTIVATED CARBON<br />

HONEYCOMB<br />

An activated carbon honeycomb was prepared for<br />

resistive heating by attaching electrodes to the<br />

honeycomb using conductive epoxy. Electrical current<br />

was passed through the part using a constant 12-volt<br />

power supply. Electrical current was applied only while<br />

the canister was being purged.<br />

ETHANOL FUEL BLENDS<br />

In order to determine the effect <strong>of</strong> ethanol content <strong>and</strong><br />

fuel vapor pressure on canister emissions, fuels with a<br />

range <strong>of</strong> ethanol content <strong>and</strong> volatility were evaluated<br />

<strong>and</strong> are listed in Table 1.<br />

Test<br />

Fuel - #<br />

RVP<br />

(kPa) @<br />

37.8 o C<br />

1 34.5<br />

2 48.3<br />

3 48.3<br />

4 55.2<br />

5 62.1<br />

6 62.1<br />

Fuel Product<br />

Code<br />

Haltermann -<br />

HF032<br />

Haltermann -<br />

HF087<br />

Haltermann -<br />

HF004<br />

Haltermann -<br />

HF087 <strong>and</strong> 40<br />

CFR 86.1313-04<br />

(FFV)<br />

40 CFR 86.1313-<br />

04<br />

Haltermann -<br />

HF115<br />

Ethanol<br />

Volumetric<br />

Concentratio<br />

n (%)<br />

85<br />

85<br />

0<br />

10<br />

0<br />

10<br />

Table 1. List <strong>of</strong> fuels used to perform diurnal testing.<br />

Tedlar Bag<br />

Environmental Chamber<br />

Canister 60-Liter Fuel Tank<br />

Figure 1. Diagram <strong>of</strong> simulated CARB diurnal<br />

evaporative emissions test.<br />

Temperature ( o C) .<br />

45<br />

40<br />

35<br />

30<br />

25<br />

20<br />

15<br />

10<br />

5<br />

0<br />

0 3 6 9 12 15 18 21 24<br />

Time (hr)<br />

Figure 2. Diagram <strong>of</strong> CARB Diurnal Temperature Pr<strong>of</strong>ile.<br />

RESULTS AND DISCUSSION<br />

Previously published results showed that canister<br />

evaporative emissions can be controlled by utilizing<br />

high-capacity/low-heel wood-based carbons, optimizing<br />

canister geometry, using an auxiliary chamber, <strong>and</strong><br />

increasing purge volume available to the canister (1).<br />

This paper concerns the further optimization <strong>of</strong> bleed<br />

control under conditions <strong>of</strong> reduced purge <strong>and</strong> also<br />

describes results obtained with ethanol containing fuels.<br />

EFFECT OF PURGE VOLUME ON CANISTER<br />

EMISSIONS<br />

The two-day simulated real time diurnal test was used to<br />

evaluate the effect <strong>of</strong> purge volume on canister<br />

hydrocarbon emissions. Each test involved<br />

preconditioning <strong>of</strong> a 2.1-liter canister as described above<br />

in preparation for the simulated diurnal test. In addition<br />

to evaluating base canister performance, canisters were<br />

evaluated with activated carbon honeycombs attached<br />

to the vent port <strong>of</strong> the canister. The activated carbon<br />

honeycombs are high-carbon content monoliths


specifically designed to reduce canister emissions. Two<br />

commercially available activated carbon honeycomb<br />

parts were evaluated. Dimensions for these two<br />

cylindrical parts are 29 mm diameter x 100 mm length<br />

<strong>and</strong> 41 mm diameter x 150 mm length. Both parts had a<br />

nominal cell density <strong>of</strong> 200 cpsi. In order to evaluate the<br />

effect <strong>of</strong> purge volume, the final purge volume before the<br />

24-hour soak <strong>and</strong> subsequent diurnal test was varied<br />

from 200 bed volumes to 75 bed volumes. The purge<br />

time was kept constant at 21 minutes, while the purge<br />

rate was changed in order to achieve the desired purge<br />

volume. The test results are shown in Figure 3.<br />

Day 2 Emissions (mg) .<br />

500<br />

450<br />

400<br />

350<br />

300<br />

250<br />

200<br />

150<br />

100<br />

50<br />

0<br />

Base Canister<br />

29x100-200<br />

41x150-200<br />

75 100 150 200<br />

Bed Volumes <strong>of</strong> <strong>Purge</strong><br />

Figure 3. Effect <strong>of</strong> <strong>Purge</strong> volume with base canister <strong>and</strong><br />

commercially available honeycombs.<br />

The data in Figure 3 demonstrates the strong<br />

relationship between purge volume <strong>and</strong> canister<br />

emissions. The activated carbon honeycombs<br />

significantly reduced canister emissions with purge<br />

volumes <strong>of</strong> 100 bed volumes <strong>and</strong> greater. In addition,<br />

the larger 41 mm x 150 mm honeycomb exhibited lower<br />

emissions levels than the 29 mm x 100 mm honeycomb<br />

at purge volumes <strong>of</strong> less than 200 bed volumes. Due to<br />

the strong relationship between purge volume <strong>and</strong><br />

emissions, as purge volume is decreased, emissions<br />

increase. As a result, there is a point for each <strong>of</strong> the<br />

commercially available carbon honeycombs at which<br />

purge volume will be insufficient to meet either PZEV or<br />

LEV II emissions targets.<br />

OPTIMIZATION OF CARBON HONEYCOMB<br />

PROPERTIES FOR LOW PURGE VOLUME<br />

APPLICATIONS<br />

Alternative Honeycomb Geometry<br />

The commercially available activated carbon<br />

honeycombs evaluated with low levels <strong>of</strong> purge (


Honeycomb Configuration B (three 41x50 parts in series<br />

<strong>and</strong> rotated 45°) showed a 32% improvement over the<br />

baseline Configuration A. This performance<br />

improvement is assumed to be the result <strong>of</strong> flow<br />

redistribution <strong>and</strong>/or increased tortuosity <strong>of</strong> the flow path.<br />

Honeycomb Configurations C (41x200) <strong>and</strong> D (50x150)<br />

showed approximately 45% <strong>and</strong> 60% improvement as<br />

compared to the baseline Configuration A. Both <strong>of</strong> these<br />

honeycomb configurations have more total working<br />

capacity due to their larger size as compared to the<br />

baseline honeycomb. In addition, Configuration C has a<br />

longer flow path for vapors diffusing out <strong>of</strong> the main<br />

canister.<br />

Results obtained with Configurations C <strong>and</strong> D confirm<br />

earlier data indicating that increased diffusional flow path<br />

length <strong>and</strong>/or total part capacity have an impact on bleed<br />

emissions (2). This earlier work was performed with 27<br />

mm diameter honeycombs <strong>of</strong> varying length at higher<br />

purge volumes <strong>of</strong> 150 bed volumes. The results from<br />

the earlier work are presented in Figure 5.<br />

Day 2 Emissions (mg) .<br />

200<br />

180<br />

160<br />

140<br />

120<br />

100<br />

80<br />

60<br />

40<br />

20<br />

0<br />

No HCA 50 mm 75 mm<br />

Honeycomb Length<br />

100 mm 150 mm<br />

Figure 5. Results obtained with 27 mm diameter<br />

honeycombs evaluated with 150 bed volumes <strong>of</strong> purge.<br />

The results obtained with the developmental honeycomb<br />

configurations show that the following changes to the<br />

baseline part improve emissions performance.<br />

1. Increased honeycomb working capacity<br />

2. Increased diffusional flow path length<br />

3. Flow redistribution<br />

4. Increased tortuosity<br />

Further work is planned in order to better underst<strong>and</strong> the<br />

reasons for these performance improvements <strong>and</strong> to<br />

develop a honeycomb with more robust performance at<br />

low purge volumes.<br />

Resistive Heating <strong>of</strong> the Honeycomb to Reduce Canister<br />

Emissions<br />

Resistive heating <strong>of</strong> activated carbon honeycombs has<br />

been shown to improve performance over baseline<br />

activated carbon honeycomb performance. This is<br />

particularly effective at low purge volume levels because<br />

an un-heated honeycomb may fail to achieve desired<br />

performance. In order to demonstrate this concept, the<br />

two-day simulated diurnal testing procedure described<br />

earlier was repeated by resistively heating an activated<br />

carbon honeycomb during the preconditioning purge.<br />

This is possible with <strong>MeadWestvaco</strong> activated carbon<br />

honeycombs due to their electrical properties. The<br />

electrical conditions <strong>and</strong> temperatures achieved during<br />

the purge cycles are summarized in Table 4 followed by<br />

a summary <strong>of</strong> the results in Figure 6.<br />

HCA Size<br />

Heated HCA Configuration<br />

41x150-200<br />

<strong>Purge</strong> Air Temp ( o C) 70 - 90<br />

Max HCA Temp ( o C) 200<br />

Current (amps) @ 12v 1.8 -2.2<br />

Table 4. Heated honeycomb conditions during purge<br />

cycle.<br />

Day 2 Emissions (mg) .<br />

180<br />

160<br />

140<br />

120<br />

100<br />

80<br />

60<br />

40<br />

20<br />

0<br />

50 75 100 150 200<br />

Bed Volumes <strong>of</strong> <strong>Purge</strong><br />

Heated 41x150-200<br />

Baseline 41x150-200<br />

Figure 6. Summary <strong>of</strong> emissions results obtained with<br />

heated honeycomb configuration.<br />

Figure 6 shows that emissions measured with the<br />

heated activated carbon honeycomb were lower as<br />

compared to the non-heated honeycomb. The<br />

improvement in performance was most dramatic at lower<br />

purge volumes. Resistively heating the honeycomb<br />

allowed emissions <strong>of</strong> less than 20 mg for all levels <strong>of</strong><br />

purge evaluated. These emissions levels are significant<br />

because they correlate well with performance required<br />

for PZEV compliance on vehicles. The results obtained<br />

with the resistively-heated activated carbon honeycomb<br />

are due to a more complete removal <strong>of</strong> hydrocarbon<br />

heel from both the activated carbon honeycomb <strong>and</strong> the


portion <strong>of</strong> the canister carbon bed closest to the<br />

atmosphere port.<br />

EFFECT OF CARBON HONEYCOMB ON CANISTER<br />

WORKING CAPACITY<br />

As a result <strong>of</strong> testing for canister bleed emissions<br />

performance, it was discovered that activated carbon<br />

honeycombs have a significant positive impact on<br />

diurnal canister working capacity. The effect <strong>of</strong> the<br />

activated carbon honeycomb on canister gasoline<br />

working capacity is summarized in Figure 7.<br />

Canister G W C (g) .<br />

180<br />

160<br />

140<br />

120<br />

100<br />

80<br />

60<br />

40<br />

20<br />

0<br />

+4%<br />

+11%<br />

+13%<br />

+19%<br />

No HCA 29x100-200 41x150-200 50x150-200 Heated<br />

41x150<br />

Figure 7. Effect <strong>of</strong> activated carbon honeycomb on<br />

canister working capacity.<br />

The baseline gasoline working capacity <strong>of</strong> 138 g<br />

presented in Figure 6 was obtained with a 2.1-liter<br />

canister filled with BAX 1500 without an activated carbon<br />

honeycomb attached. An increase in overall canister<br />

working capacity <strong>of</strong> 4% to 19% was demonstrated with<br />

the addition <strong>of</strong> an activated carbon honeycomb. The<br />

increased capacity was proportional to the honeycomb<br />

size <strong>and</strong> therefore the capacity <strong>of</strong> the honeycomb to<br />

adsorb hydrocarbon vapors.<br />

EFFECT OF ETHANOL FUEL BLENDS ON CARBON<br />

BED PERFORMANCE<br />

Adsorption <strong>of</strong> Ethanol<br />

The adsorption <strong>and</strong> desorption isotherms for Ethanol<br />

<strong>and</strong> Butane on BAX 1500 at 25°C, is shown in Figures 8<br />

<strong>and</strong> 9.<br />

Loading g/g Carbon<br />

1.40<br />

1.20<br />

1.00<br />

0.80<br />

0.60<br />

0.40<br />

0.20<br />

Adsorption<br />

Desorption<br />

0.00<br />

10 100 1000 10000 100000 1000000<br />

Concentration (ppm)<br />

Figure 8. BAX 1500 Butane Equilibrium Isotherm.<br />

Loading g/g Carbon<br />

1.40<br />

1.20<br />

1.00<br />

0.80<br />

0.60<br />

0.40<br />

0.20<br />

Adsorption<br />

Desorption<br />

0.00<br />

10 100 1000 10000 100000 1000000<br />

Concentration (ppm)<br />

Figure 9. BAX 1500 Ethanol Equilibrium Isotherm.<br />

Comparisons <strong>of</strong> these figures clearly show that, at equal<br />

concentrations, ethanol is more readily adsorbed than<br />

butane, the st<strong>and</strong>ard used to evaluate automotive<br />

carbons (3, 4). More importantly, due to the<br />

mesoporous nature <strong>of</strong> <strong>MeadWestvaco</strong> automotive<br />

carbons, the desorption curves show that ethanol <strong>and</strong><br />

butane are readily removed from the carbon pores.<br />

Ethanol Fuel Blends<br />

There are two potential ways for ethanol to be present in<br />

fuels:<br />

1. Splash Blending. This is the straight addition <strong>of</strong><br />

alcohol to a non-alcohol-based gasoline. This has been<br />

shown to significantly affect the vapor pressure <strong>of</strong> the<br />

resultant mix as shown in Figure 10 (5).


Reid Vapor Pressure @ 37.8oC (kPa)<br />

.<br />

80<br />

70<br />

60<br />

50<br />

40<br />

30<br />

20<br />

10<br />

0<br />

0 10 20 30 40 50 60 70 80 90 100<br />

Ethanol Concentration (%)<br />

Figure 10: Effect on Vapor Pressure by ethanol addition<br />

to 9.0 RVP Fuel<br />

Figure 10 shows that splash blending <strong>of</strong> ethanol up to<br />

approximately 40% will cause an increase in the vapor<br />

pressure <strong>of</strong> the resultant blend.<br />

2. Blending to a fixed Vapor Pressure. This is the<br />

addition <strong>of</strong> an ethanol to gasoline fuel that has been<br />

altered so that the resultant solution has the same vapor<br />

pressure as the unblended gasoline. For example, in<br />

the case <strong>of</strong> Fuel - 6, the volatility <strong>of</strong> the non-ethanol<br />

components <strong>of</strong> the fuel is reduced to compensate for this<br />

effect <strong>of</strong> adding ethanol.<br />

Regardless <strong>of</strong> the approach used, the most important<br />

property <strong>of</strong> the fuel blend with respect to activated<br />

carbon canister performance is the vapor pressure. The<br />

vapor pressure, rather than ethanol content, determines<br />

the resultant diurnal vapor generation from the fuel tank<br />

<strong>and</strong> therefore the required working capacity <strong>of</strong> the<br />

carbon canister.<br />

Diurnal Vapor Generation<br />

A range <strong>of</strong> fuels was tested for vapor generation rates by<br />

following the diurnal test procedure <strong>of</strong> subjecting a fuel<br />

tank <strong>and</strong> canister to the entire two-day CARB diurnal<br />

temperature pr<strong>of</strong>ile as described earlier. The RVP <strong>of</strong> the<br />

fuels tested ranged from 34.5 to 62.1 kPa, with an<br />

ethanol content <strong>of</strong> 0 to 85%. The diurnal vapor<br />

generation results are shown in Figure 11.<br />

Average Daily Vapor Load (g) .<br />

40<br />

35<br />

30<br />

25<br />

20<br />

15<br />

Fuel - 1<br />

Fuel - 3<br />

Fuel - 4<br />

Fuel - 2<br />

Fuel - 5<br />

10<br />

20 25 30 35 40 45<br />

RVP (kPa)<br />

50 55 60 65 70<br />

Figure 11. Diurnal vapor generation <strong>of</strong> fuels tested<br />

As would be expected, this figure shows a strong<br />

relationship between RVP <strong>and</strong> vapor generation. More<br />

importantly, it shows that fuels <strong>of</strong> equivalent RVP will<br />

generate similar total amounts <strong>of</strong> diurnal vapors<br />

regardless <strong>of</strong> the ethanol content. For example, on the<br />

average Fuel -2 generated 27.5 grams <strong>of</strong> vapor<br />

compared to an average <strong>of</strong> 29.7 grams for Fuel - 3 with<br />

an equivalent RVP. While similar total vapor generation<br />

amounts have been observed for fuels with similar RVP,<br />

it is understood that the fuels may have different<br />

volatilities at the various temperatures experienced<br />

during the CARB diurnal temperature pr<strong>of</strong>ile. Based on<br />

these results, if a canister is properly sized for use with a<br />

62.1 RVP fuel, it should be adequately sized for ethanol<br />

blends <strong>of</strong> an equal or lower RVP.<br />

EFFECT OF ETHANOL FUEL BLENDS ON<br />

HONEYCOMB PERFORMANCE<br />

Initially, simulated diurnal testing was conducted using<br />

st<strong>and</strong>ard Fuel - 5 <strong>and</strong> Fuel - 3 as specified by CARB<br />

procedures for non-Flex Fuel <strong>Vehicle</strong>s (FFV’s) to<br />

establish baseline performance. Results obtained using<br />

the 2.1-liter canister filled with BAX 1500 both with <strong>and</strong><br />

without a st<strong>and</strong>ard 41x150 honeycomb (Configuration A)<br />

attached under these conditions were compared to<br />

those obtained by cycling with the same Fuel - 5<br />

followed by a two day diurnal using Fuel - 2. The results<br />

are compared in Figure 12.<br />

Day 2 Emissions (mg) .<br />

180<br />

160<br />

140<br />

120<br />

100<br />

80<br />

60<br />

40<br />

20<br />

0<br />

BAX 1500 w / 41x150 HCA<br />

BAX 1500<br />

Cyc Fuel - 5 / DBL Fuel - 3 Cyc Fuel - 5 / DBL Fuel - 2


Figure 12. Diurnal emissions for equivalent RVP E0 <strong>and</strong><br />

E85 fuel blends with 150 bed volumes <strong>of</strong> purge.<br />

Figure 12 shows that within the accuracy <strong>of</strong> the testing<br />

(+/- 7.5 mg), use <strong>of</strong> an E85 fuel blend rather than the<br />

st<strong>and</strong>ard CARB Fuel - 3 during the simulated diurnal test<br />

does not have a significant impact on emissions. Based<br />

on this data, bleed emissions appear to be independent<br />

<strong>of</strong> ethanol content for both the main canister <strong>and</strong> the<br />

honeycomb. In addition, the results show that the use <strong>of</strong><br />

a st<strong>and</strong>ard, commercially-available activated carbon<br />

honeycomb (Configuration A) makes it possible to<br />

reduce canister emissions to PZEV levels.<br />

The simulated diurnal testing was repeated by<br />

performing preconditioning cycling with the CARB Fuel -<br />

2 mileage accumulation fuel followed by a diurnal with<br />

the fuel type specified for evaporative emissions testing<br />

<strong>of</strong> flexible fuel vehicles, Fuel - 4. These results were<br />

compared to those obtained when a st<strong>and</strong>ard Fuel - 5<br />

was used for the two-day diurnal test following<br />

preconditioning cycles with Fuel -2 mileage<br />

accumulation fuel. The results are shown in Figure 13.<br />

Day 2 Emissions (mg) .<br />

300<br />

250<br />

200<br />

150<br />

100<br />

50<br />

0<br />

BAX 1500 w / 41x150 HCA<br />

BAX 1500<br />

Cyc Fuel - 2 / DBL Fuel - 4 Cyc Fuel - 2 / DBL Fuel - 5<br />

Figure 13. Diurnal emissions for Fuel - 4 <strong>and</strong> Fuel – 5<br />

with 150 bed volumes <strong>of</strong> purge.<br />

The results provide further evidence that emissions are<br />

independent <strong>of</strong> ethanol content for either the main<br />

canister or the main canister in combination with the<br />

honeycomb. More importantly, the results show that the<br />

use <strong>of</strong> a st<strong>and</strong>ard, commercially available activated<br />

carbon honeycomb (Configuration A) makes it possible<br />

to reduce canister emissions to LEV II levels for FFV’s.<br />

CONCLUSION<br />

Recent trends in automotive technology have led to a<br />

reduction in the amount <strong>of</strong> air available for regenerating<br />

or purging the carbon canister. Under these reduced<br />

purge volume conditions, current activated carbon<br />

canister <strong>and</strong> honeycomb configurations may not be<br />

adequate to meet PZEV or even st<strong>and</strong>ard LEV II<br />

requirements. Improvements over the current<br />

honeycomb technology such as the following facilitate<br />

achievement <strong>of</strong> target emissions levels under the<br />

reduced purge conditions:<br />

• Higher capacity honeycombs<br />

• Longer diffusional path length<br />

• Flow redistribution<br />

• Resistively-heated honeycombs<br />

Further work is planned in order to gain a better<br />

underst<strong>and</strong>ing about the relative importance <strong>of</strong><br />

diffusional flow path length <strong>and</strong> honeycomb capacity.<br />

With regards to trends in renewable fuels <strong>and</strong> in<br />

particular ethanol blends, the following was shown<br />

• <strong>MeadWestvaco</strong> mesoporous automotive<br />

carbons readily adsorb <strong>and</strong> desorb ethanol<br />

• Based on the range <strong>of</strong> fuels tested, ethanol<br />

content does not adversely affect the carbon<br />

canister performance<br />

• Canister emissions are driven primarily by the<br />

fuel volatility rather than ethanol content<br />

• <strong>MeadWestvaco</strong> commercially available<br />

honeycombs can enable FFV’s to attain LEV II<br />

compliance.<br />

ACKNOWLEDGMENTS<br />

The authors wish to express their sincere appreciation to<br />

Joe W. Snead, who aided in the design, operation <strong>and</strong><br />

maintenance <strong>of</strong> the equipment used in the study. Dr.<br />

James R. Miller, who provided support <strong>and</strong><br />

encouragement.<br />

REFERENCES<br />

1. R. S. Williams <strong>and</strong> C. Reid Clontz, “Impact <strong>and</strong><br />

Control <strong>of</strong> Canister Bleed Emissions,” SAE<br />

Technical Paper 2001-01-0733, March 2001.<br />

2. R. S. Williams, <strong>MeadWestvaco</strong>, Evaporative<br />

Emissions Seminar, 2002.<br />

3. Annual Book <strong>of</strong> ASTM St<strong>and</strong>ards 2006, Volume<br />

15.01, ASTM International, ISBN 0-8031-4224-2,<br />

West Conshohocken, PA.<br />

4. H. R. Johnson <strong>and</strong> R. S. Williams, “Performance <strong>of</strong><br />

Activated Carbon Canisters in Evaporative Loss<br />

Control Systems,“ SAE Technical Paper 902119,<br />

October 1990.<br />

5. Robert L. Furey, “Volatility Characteristics <strong>of</strong><br />

Gasoline-Alcohol <strong>and</strong> Gasoline-Ether Fuel Blends,”<br />

SAE Technical Paper 852116.<br />

CONTACT


<strong>MeadWestvaco</strong> Specialty Chemicals Division<br />

P.O. Box 140, Washington Street<br />

Covington, VA 24426<br />

800-336-2211<br />

DEFINITIONS, ACRONYMS, ABBREVIATIONS<br />

BAX 1500: 2mm pelletized MWV automotive carbon<br />

with an ASTM BWC <strong>of</strong> 14.8 g/dl minimum<br />

Bed Volumes: Volume <strong>of</strong> air used to purge the canister<br />

divided by volume <strong>of</strong> carbon in main canister volume<br />

excluding volume <strong>of</strong> honeycomb<br />

CARB: California Air Resources Board<br />

CPSI: Nominal cells per square inch<br />

Cyc: Abbreviation for “cycled” for simulated driving<br />

cycles used to precondition evaporative emission<br />

canister<br />

DBL: Diurnal Bleed Loss<br />

E-10S: Splash Blended Fuel using 62.1 RVP (E0) <strong>and</strong><br />

pure ethanol to 10% Ethanol Concentration, uncontrolled<br />

vapor pressure<br />

E85: 85% ethanol blend<br />

FFV 10%: “Flex Fuel <strong>Vehicle</strong> 10%,” Fuel blended to<br />

10% ethanol concentration using 48.3 RVP E-85 <strong>and</strong><br />

62.1 RVP (E0)<br />

GWC: Gasoline Working Capacity<br />

HCA: Hydrocarbon Adsorber or Honeycomb<br />

LEV: <strong>Low</strong> Emission <strong>Vehicle</strong><br />

PZEV: Partial Zero Emission <strong>Vehicle</strong><br />

RVP: Reid Vapor Pressure at 37.8°C (kPa)<br />

SHED: Sealed Housing Evaporative Determination<br />

Tedlar® : DuPont’s registered Polyvinyl Fluoride<br />

products that are chemically inert

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!