NOTE

DATE:		January 8, 2010

SUBJECT:	Analysis of Ethanol Evaporative Permeation Effects Update

TO:		RFS2 Docket EPA-HQ-OAR-2005-0161

FROM:	David Hawkins, Assessment and Standards Division, OTAQ

This memorandum serves as an update to the data analysis of the ethanol
effect on evaporative permeation.

Ethanol affects evaporative emissions from gasoline vehicles due to the
increased volatility of Ethanol blends, the increased permeation of fuel
vapors through tanks and hoses, and the increased vapor emissions due to
the lower molecular weight of Ethanol.  Each of these effects were
modeled using the draft MOVES model, which separates permeation
emissions from vapor venting emissions to allow better accounting for
these different processes.

DATA USED

Permeation effects were developed from data available as of June 2009
from CRC E-65, E-65-3, E-77, E-77-2, and E-77-2b programs, which
measured evaporative emissions.  The E-65 programs used ten fuel systems
that were removed from the vehicles and tested on E0, E5.7 and E10
fuels.  The E-77 programs used full vehicles, tested on E0, E10, and
E20.  An INNOVA sampler in E-77 was used to isolate and subtract
refrigerant-134a and Methanol should they contaminate the fuel
permeation measurements.  The canister was routed outside of the SHED,
to ensure that measured emissions within the SHED were solely from
permeation, rather than canister breakthrough.  All of the E-77 vehicles
underwent preliminary tests to ensure they did not contain any vapor
leaks that would confound the results in the SHED.  

Table 1. CRC E-65, E-65-3, E-77, E-77-2, and E-77-2b vehicles

Veh. No.	Yr	Make	Model	Evap Standard	Program

10	1978	Oldsmobile	Cutlass	Pre-enhanced	E-65

9	1985	Nissan	Sentra	Pre-enhanced	E-65

8	1989	Ford	Taurus GL	Pre-enhanced	E-65

7	1991	Honda	Accord LX	Pre-enhanced	E-65

3	1992	Honda	Accord	Tier 0	E-77

10	1992	Toyota	Camry	Tier 0	E-77

6	1993	Chevrolet	Caprice Classic	Pre-enhanced	E-65

5	1995	Ford	Ranger	Pre-enhanced	E-65

9	1995	Plymouth	Neon	Tier 1	E-77

1	1996	Chevrolet	S-10	Tier 1	E-77

6	1996	Chevrolet	Cavalier	Tier 1	E-77

7	1996	Chevrolet	Cavalier	Tier 1	E-77

8	1996	Ford	Explorer	Tier 1	E-77

202	1996	Ford	Taurus	Tier 1	E-77-2

216	1996	Ford	Taurus	Tier 1	E-77-2

4	1997	Chrysler	Town & Country	Pre-enhanced	E-65

3	1999	Toyota	Corolla	ORVR	E-65

4	1999	Dodge	Grand Caravan SE	Tier 0	E-77

204	1999	Honda	Accord	Tier 1	E-77-2

2	2000	Toyota	Tacoma	Tier 1	E-77

2	2000	Honda	Odyssey	Enhanced	E-65

220b	2000	Chevrolet	Malibu	Tier 1	E-77-2b

221b	2000	Mitsubishi	Galant ES	Tier 1	E-77-2b

1	2001	Toyota	Tacoma	Enhanced	E-65

205	2001	Toyota	Corolla	Tier 1	E-77-2

207	2001	Dodge	Caravan	Tier 1	E-77-2

206b	2002	Nissan	Altima	Tier 2	E-77-2b

208b	2002	Chevrolet	Trailblazer	Tier 2	E-77-2b

11	2004	Ford	Taurus

E-65-3

12	2004	Chrysler	Sebring

E-65-3

211	2004	Toyota	Camry LE	Near Zero	E-77-2

214	2004	Ford	Escape	Tier 2	E-77-2

215	2004	Toyota	High Lander	Tier 2	E-77-2

209b	2004	Chrysler	Stratus	Tier 2	E-77-2b

210b	2004	Chevrolet	Impala	Near Zero	E-77-2b

213b	2004	Dodge	Ram 1500	Near Zero	E-77-2b

222b	2004	Ford	Focus ZX3	Zero Evap	E-77-2b

14	2005	Chevrolet	Tahoe

E-65-3

212	2006	Ford	Taurus	Near Zero	E-77-2

5	2007	Ford	Taurus	Tier 2	E-77



ANALYSIS

There are several variables considered in determining the appropriate
aggregation level for permeation rate, including four evaporative model
year groups defined by MOVES (pre-1977, 1978-1995, 1996-2003, 2004 and
later), two vehicle classes (car/truck), and three Ethanol volumes (E0,
E5.7, E10).  E15 and E20 were not analyzed for the RFS2 rule.  Attempts
to find an effect for each combination of these variable failed to be
significant due to the relatively small amount of data; P-values of 0.6
or greater were calculated (Significance is generally accepted at
p-values < 0.05).   Data were then aggregated to determine significant
effects; cars and trucks were aggregated, and E5.7 and E10 fuels are
combined into a single “with ethanol” category to compare to E0
fuels.  Likewise, model year 1996 and later (nominally representing
evaporative enhanced vehicles) were combined to compare to pre-1996
vehicles.  These steps decrease the resolution of the results but
provide greater significance.

To determine the effect of the ethanol blend, the 65-105-65 F diurnal
cycle test results from either 24-hour or 72-hour tests are used.  The
dataset was analyzed using the Mixed procedure in SAS 9 (Statistical
Analysis System) to ensure significant (p < 0.05) differences between E0
and E5.7/E10 for both pre-enhanced and evaporative enhanced vehicles. 
Calculated p-values are <0.0001.  The adjustment in MOVES2010 is the
percent increase due to Ethanol in the fuel.

RESULTS

Table 2. Increase in emissions due to E5.7/E10 compared to E0 (gasoline)

Tech Group	Percent increase due to ethanol

Pre-enhanced	66

Enhanced	113.8



Due to the evaporative enhanced phase in from 1996 to 1999
(20/40/90/100%), the two fuel adjustments are appropriately weighted for
those model years.  

Table 3. Increase in emissions due to E5.7/E10 compared to E0 w/
phase-in (gasoline)

Model years (via fuelmygroupid in MOVES)	Percent increase due to ethanol

1995 and earlier	66

1996	75.5

19972000	107.1

2001 and later	113.8



   HYPERLINK "http://www.crcao.org/publications/emissions/index.html" 
http://www.crcao.org/publications/emissions/index.html  (E-77-2 and
E-77-2b reports to be available Mid-2010)

