FLAWED EPA METHOD OF CALCULATING MPG FOR E10 VEHICLES AND MPGeq FOR EVs

The EPA has a unique method of calculating the mileage of an electric vehicle. The EPA calls its MPGeq. The method is a grand deception of the US public. It is used nowhere else in the world.

 

Auto manufacturers did not object, because they were allowed to use the inflated EPA MPGeq values to boost their fleet averages. However, when rosy EV sales projections did not happen, the “54.5 mpg fleet average by 2025” proved to be an off-the-charts fantasy, and auto manufacturers began to object.

 

After Trump became president, a more realistic mpg target of about “37 mpg fleet average by 2021” became necessary, because of much fever EVs averaged into the new car sales mix.

 

The EPA MPGeq Method

 

The EPA arbitrarily sets the lower heating value, LHV, of gasoline at 115,000 Btu/gal.

Then EPA states 115000 Btu/(3412 Btu/kWh) = 33.7 kWh.

 

NOTE:The LHV of E10 (a blend of 90% gasoline/10% ethanol) is 112114 Btu/gal.

 

When testing an EV, the EPA measures the kWh (AC) from a wall meter.

The vehicle state-of-charge meter measures the kWh (DC) in the battery.

 

If the vehicle meter indicates a fuel consumption of 0.300 kWh (DC)/mile, then the wall meter would indicate 0.300 x 1.21 = 0.363 kWh (AC)/mile.

Energy per mile would be 3412 x 0.363 = 1239 Btu.

 

EPA would claim the mileage of the EV as 33.7/0.363 (wall meter) = 92.8 MPGeq, a gross exaggeration.

NOTE: 

- Whether EVs are used, or not used, some of the electricity is taken from the battery to operate various systems. Just go on vacation for 2 weeks, park a fully charged Tesla Model S at the airport, and you will be shocked at how much charge is lost.

- Some of the electricity would still be lost, even if no systems were operated.

- Some of the electricity is lost during charging.

- The total of these losses is about 21% for the Tesla Model 3 in California. See Appendix.

- These resting/charging losses likely would be greater in colder climates, such as in New England

- These resting/charging losses are separate from increased consumption per mile when drivingduring hot and cold weather 

 

NOTE: The Btus in gasoline are thermal Btus, whereas the Btus in a kWh are electrical Btus. If an engineering student were to equate them, he/she would be awarded a D. Mixing thermal and electrical Btus is OK in politics, but a basic no-no in engineering.

 

NOTE: In Europe and elsewhere, EV consumption has been stated as liters/100 km, and CO2 emission as g/km well before the EPA came up with its MPGeq “method”. Foreign manufacturers just smile, while complying with the EPA, to profitably sell cars in the US.

EPA Does Not Adjust MPGeq For Energy Upstream of Wall Meter

 

NE grid (gas turbine efficiency at 50%): If one starts with 0.938 kWh/mile of primary energy fed to gas turbine power plants, about 0.350 kWh/mile is stored in batteries.

 

US grid (conversion efficiency at 37.79%): If one starts with 1.239 kWh/mile of PE fed to US power plants, about 0.350 kWh/mile is stored in batteries. See table 1

 

Table 1	G/Ts at 50%	G/Ts at 50%	US grid	US grid
LNG requirement		kWh/mile		kWh/mile
Primary energy; gas turbines		0.938		1.239
Efficiency	50.0%	0.469	37.79%	0.771
Electricity generation		0.469		0.468
Self-use loss	3.00%	0.013	3.83%	0.017
Fed to grid		0.455		0.451
T&D loss	7.50%	0.032	6.50%	0.028
To meters		0.424		0.424
EV charging loss	21.00%	0.074	21.00%	0.074
In batteries		0.350		0.350

- The US electrical system has a mix of generators, which typically have efficiencies of less than 50%. The US system efficiency is about 37.79%. See table 1.

- That means, EPA should reduce its MPGeq to 92.8 MPGeq/2.943 = 31.534 mpg, a far more realistic number, based on primary energy. See table 2.

- That means, EPA should reduce the above MPGeq to 92.8 MPGeq/3.178 = 29.198 mpg, a far more realistic number, based on source energy

 

Table 2	%
Source energy	108.000
Extract/Process/Transport loss, 8% of PE	8.000
Primary energy fed to generators	100.000
Efficiency, 37.79%	62.210
Gross electricity generation	37.790
Self-use loss, 3.83%	1.447
Net electricity generation fed to grid	36.343
T&D loss, 6.5%, US grid, per EIA	2.362
User electricity consumption	33.980
Primary energy factor, 100/33.980	2.943
Source energy factor, 108/33.980	3.178
Real mpg, PE basis	31.534
Real mpg, SE basis	29.198

Source Energy and Source CO2 factors for Gasoline, Ethanol and E10

 

Biofuels, including ethanol, require far more energy from various fossil fuels and chemicals to produce them than gasoline.

The combustion CO2 of ethanol is not counted, because the next crop reabsorbs the CO2 a year later, per international agreement. See table 3.

 

Producing 1 million Btu of gasoline, HHV, requires about 230000 Btu of various energy inputs.

Producing 1 million Btu of ethanol, HHV, requires about 914414 Btu of various energy inputs. See arb.ca URL.

Producing 1 million Btu of E10 requires 0.9 x 230000 + 0.1 x 914414 = 298441 Btu of various energy inputs.

A driver needs 1.0331 gallon of E10 to go the same distance as on one gallon of pure gasoline. See table 3.

If the combustion CO2 of E10 is not counted E10 has 194.839 lb CO2/million Btu and pure gasoline has 197.442 lb CO2/million Btu; the ethanol program is a gigantic, expensive, political scam.

 

https://www.afdc.energy.gov/fuels/fuel_comparison_chart.pdf

https://www.arb.ca.gov/fuels/lcfs/042308lcfs_etoh.pdf

https://h2tools.org/hyarc/calculator-tools/lower-and-higher-heating...

Table 3
	Ethanol	Ethanol	Gasoline	E10 (90/10)
ENERGY	With credit	No credit	No credit	No credit
Fuel produced, HHV, Btu. See URL	1000000	1000000	1000000	1000000
Co-products, Btu. See URL	97301	0	0	0
Primary energy, Btu	1097301	1000000	1000000	1000000
.
Cropping, processing, transport, Btu	914414	914414
Extraction, processing, transport, Btu			230000	298441
Source energy, Btu	2011715	1914414	1230000	1298441
Factor = SE/PE	1.8333	1.9144	1.2300	1.2984
.
CO2 EMISSIONS
HHV, Btu/gal. See URL		84530	124340	120359	LHV ratio
LHV, Btu/gal, See URL		76330	116090	112114	1.0355
Combustion CO2, lb/gal. See URL		12.720	19.640	18.948
Crop, process, transport CO2, lb/gal		13.556			0.25 x 19.640 = 4.9100
Extract, process, transport CO2, lb/gal			4.9100	5.775	13.556 x 0.1 + 4.91 x 0.9
Source CO2, bio CO2 counted, lb/gal		26.276	24.550	24.723
Factor = Source/Combustion CO2		2.0657	1.2500	1.3048
Source CO2, bio not counted, lb/gal		13.556	24.550	23.451	13.556 x 0.1 + 24.55 x 0.9
Factor = Source/Combustion CO2		1.0657	1.2500	1.2376	1.25 per EPA
					Gallon ratio
Fuel, HHV, gal/million Btu		11.830	8.042	8.308	1.0331
Total CO2, not counted, lb/million Btu		160.369	197.442	194.839

Adjusting EPA Combined Ratings For Upstream Energy

 

The EPA determines the EPA Combined MPGeq rating, tank basis, which enables fuel consumption comparison of one vehicle versus another. It ignores upstream energy for extraction, processing and transport to produce the fuel.

 

A better method, for environmental reasons, would be from source to wheel, which would enable energy consumption comparison of one pathway versus another, on an A to Z, lifecycle, cradle to grave basis.

 

If an EV were rated at 0.363 kWh/mile, wall meter basis, it would be rated 2.943, from table 2 x 0.363 kWh/mile = 1.068 kWh/mile on a PE basis.

 

Energy per mile would be 3412 Btu/kWh x 1.068 = 3645 Btu, PE basis. See table 2. 

 

If a future efficient compact E10 vehicle were rated at 38-MPG EPA Combined, tank basis, it would be rated

1000000, fuel PE/1298441, fuel SE x 38 = 29.3 MPG Combined on a PE basis.

 

Energy per mile would be 112114 Btu/gal/29.3 = 3831 Btu, PE basis. See table 4.

 

These two Btu/mile values are far more realistic than the 1239 Btu/mile concocted, in a politics-inspired manner, by the EPA method.

 

In each case, the E10 vehicle and EV would have to overcome about the same rolling and wind resistance to travel from A to B. It stands to reason they would use about the same energy to do that. See table 4.

 

Table 4	EV	E10 vehicle	EPA
Electricity, kWh/mile, wall meter basis	0.3630		0.3630
Primary energy factor, US grid, table 2	2.943
Primary energy, kWh/mile	1.068
Btu/kWh	3412		3412
Mileage, mpg		38.0
E10 upstream factor, see table 3		1.2984
EPA Combined, adjusted for upstream, mpg		29.3
Btu/mile, PE basis, including upstream	3645	3831	1239

    

Annual Cost of Driving and CO2 Emissions of Two EVs and a Future Compact IC Vehicle

 

- Because vehicles are very often driven in various NE states, the CO2 was assumed to be for the NE grid, which was 374 g/kWh, at the wall meter, SE basis, in 2016, per ISO-NE.

- The CO2 values are based on PE fed to power plants.

- The Prius plug-in was assumed to be driven 50% in electric mode and 50% in hybrid mode.

- EPA has not yet published the EPA Combined MPGeq for the 2018 Prius plug-in.

- The E10 CO2/gal includes upstream. See table 3

- It was assumed compact IC vehicles @ 38 mpg would be available within about 5 years.

- See Appendix 2.

 

Table 5/Prius plug-in		$/y	$/mile
Cost of travel/y, 50% electric mode	0.2533 x 18 c/kWh x 0.50 x 12000	274
Cost of travel/y, 50% hybrid mode	12000/54 mpg x 0.50 x $2.80/gal	311
Total cost of travel/y		585	0.0492
		g/y	g/mile
CO2 emission, SE basis, electric mode	0.2533 x 12000 x 374 NE grid x 0.50	568405	95
CO2 emission, SE basis, hybrid mode	12000/54 x 23.451 lb x 454 x 0.50	1182973	197
Total CO2 emission, PE basis		1751378	146
.
Tesla, Model S, 4wd		$/y	$/mile
Cost of travel/y, 100% electric mode	0.381 x 18 c/kWh x 12000	823	0.0686
		g/y	g/mile
CO2 emissions, SE basis	0.381 x 12000 x 374 NE grid	1709928	142
.
Tesla Model 3, 4wd, Edmunds road test		$/y	$/mile
Cost of travel/y, 100% electric mode	0.302 x 18 c/kWh x 12000	652	0.0544
.		g/y	g/mile
CO2 emissions, SE basis	0.302 x 12000 x 374 NE grid	1355376	113
.
Future compact IC vehicle @ 38 mpg		$/y	$/mile
Cost of travel	12000/38 x $2.80	884	0.0700
		g/y	g/mile
CO2 emissions, SE basis	12000/38 x 23.451 x 454	3362133	280
.
Subaru Outback, 4wd @ 30 mpg		$/y	$/mile
Cost of travel	12000/30 x $2.80	1120	0.0933
		g/y	g/mile
CO2 emissions, SE basis	12000/30 x 23.451 x 454	4258702	355

Long-Term Road Test of Tesla Model 3

 

Edmunds, in California, has been performing a long-term road test of a Tesla Model 3 since January 2018. Here are the latest results from the Edmunds website.

https://www.edmunds.com/tesla/model-3/2017/long-term-road-test/2017...

 

A recent road test of the Tesla Model 3, performed by Edmunds, showed 1388 miles of driving in California, some of it on hills

 

- Wall meter consumption was 30.2 kWh/100 miles.

- Vehicle meter consumption was 25.17 kWh/100 miles.

- The charging/resting time loss was 16.7% to 21.29%, much greater than the 15% assumed in these articles.

http://www.windtaskforce.org/profiles/blogs/replacing-gasoline-cons...

http://www.windtaskforce.org/profiles/blogs/replacing-gasoline-cons...

- The charging/resting time loss was over 20% with different drivers and different road trips.

- Winter driving would require about 0.400/0.301 = 33% more electricity per mile than summer driving. See next section about Tesla Model S and URL

http://www.windtaskforce.org/profiles/blogs/electric-cars-lose-rang...

- EV drivers know little of the charging/resting time loss; they rely on the lower numbers of the EV meter.

- February, March and April were not shown, because of missing data. See table 6 and URL

https://insideevs.com/monthly-plug-in-sales-scorecard/

 

Table 6/ Model 3	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct
Odometer	1388	2922	3937	5237	6009	6659	7679	9329	10307	11174
Test travel, miles		1534	1015	1300	772	650	1020	1650	978	867
Wall meter, kWh/100m
Lifetime average	30.20	30.90	31.70	31.70	31.40	31.80	31.70	31.00	31.10	30.80
Veh. meter, kWh/100m
Lifetime average	25.17				24.83	25.03	25.09	24.76	24.70	24.49
Best fill, period	20.00				28.50	28.60	28.00	26.70	25.60	25.60
Best fill, lifetime	25.60	25.60	25.60	25.60	25.60	25.60	25.60	25.60	25.60	25.60
Charge/rest time loss	5.03				6.57	6.77	6.61	6.24	6.40	6.31
Charge/rest time loss, %	16.66				20.92	21.29	20.85	20.13	20.58	20.49

One-Year Experience With a Tesla Model S

 

An upstate New York owner of a Tesla Model S measured the house meter kWh, vehicle meter kWh, and miles for one year (bold numbers in table). There was significant kWh/mile variation throughout the year. His real world annual average was 0.392 kWh/mile, house-meter basis, and 0.333 kWh/mile, vehicle-meter basis.

 

- The Model S has regenerative braking as a standard feature.

- The owner did not take into account the source-to-house electrical losses.

- Owners may use more or less than 0.392 kWh/mile in other US regions. 

- New EVs would have less kWh/mile than older EVs, due to battery system degradation.

- Data as measured by owner in New York State covers only the driving energy. The embedded energy and its CO2 are ignored.

 

See URLs, especially the second, which has a wealth of data.

 

http://www.greencarreports.com/news/1090685_life-with-tesla-model-s...

http://www.uniteconomics.com/files/Tesla_Motors_Is_the_Model_S_Gree...

 

NOTE: In these article, I used 0.350 kWh/mile, vehicle-meter basis, for a mix of NE LDVs (cars, SUVs, minivans, ¼-ton pick-ups, short- and long wheel base). As the Tesla Model S, with a very low drag coefficient, shows an annual average of 0.333 kWh/mile (vehicle meter basis), my assumed 0.350 kWh/mile likely is significantly too low. Table 2.

 

http://www.windtaskforce.org/profiles/blogs/replacing-gasoline-cons...

http://www.windtaskforce.org/profiles/blogs/replacing-gasoline-cons...

 

Table 2/Tesla, Model S
Electricity cost, c/kWh	19.0
Travel, miles/y	15243
Vehicle meter, kWh/y	5074
kWh/mile, vehicle meter	0.333	5074/15243
kWh/mile, vehicle meter	0.301	Apr-Oct
kWh/mile, vehicle meter	0.290	July
kWh/mile, vehicle meter	0.371	Nov-Feb
kWh/mile, vehicle meter	0.400	Jan
House meter, kWh/y	5969
Charging, resting time factor	0.85
kWh/mile, house meter	0.392	5969/15243
Travel cost, c/mile	7.4	5969 x 19/15243

APPENDIX 1

CO2 Emissions of US Grid: US CO2 Emissions Decreased Due to Less Coal and More Natural Gas.The URL shows the unusually rapid decrease of CO2 emissions during 2015 and 2016. Such a rapid decrease likely will not occur during the next few years, as natural gas prices likely will increase due to exports, and as changes in EPA rules likely will cause fewer coal plants to close.

 

A “cleaner” US grid would mean EVs would compare more favorable with E10 vehicles regarding emissions.

https://www.eia.gov/totalenergy/data/monthly/pdf/mer.pdf

 

Table 6/Year	2013	2016	2017
CO2, per EIA, million metric ton	2053	1821	1744
To meters, per EIA, TWh	3845.38	3855.41	3797.54
kg CO2/kWh	0.5339	0.4723	0.4592
lb/kg	2.2046	2.2046	2.2046
lb CO2/kWh, PE basis	1.177	1.041	1.012
g/lb	454	454	454
g CO2/kWh, PE basis	534	473	460
Upstream factor	1.08	1.08	1.08
lb CO2/kWh, SE basis	1.271	1.125	1.093
g CO2/kWh, SE basis	577	511	496

APPENDIX 2

CO2 Emissions of NE Grid: The NE grid, managed by ISO-NE, covers all of New England. ISO-NE issued its 2016 emissions report in January 2019.

ISO-NE considers pumped storage, nuclear, wind and solar as non-emitting sources. In 2016 they were about 41% of all NE generation. See page 2 of URL

 

The emissions of each state are based on the physical locations of the generating units (connected to the NEgrid) in each state. ISO-NE operates the NE power system as one unified grid, dispatching a unit physically located in one state to serve the entire system, not only the unit’s own state. This does not include northern Maine and the Citizens Block Load (in Northern Vermont), which is typically served by New Brunswick and Quebec. These areas are not electrically connected to the ISO-NE Control Area.

 

The NE grid is significantly cleaner than the US grid due to generation by gas 48.0%, nuclear 30.8% and hydro 8.4% of all NE generation. See tables 7 and 8, and pages 2, 19 and 20 of ISO-NE URL

NOTE:

- About 8% of the combustion CO2 of the fuel fed to power plants (primary energy) needs to be added due to the CO2 of extraction, processing and transport of the fuel. The ISO-NE g/kWh values in its 2016 report do not include the 8%. Source energy CO2 = 1.08 x primary energy CO2. See table 7.

- About 25% of the combustion CO2 of a gallon of E10 fed to vehicles (primary energy) needs to be added due to the CO2 of extraction, processing and transport of the E10. Source energy CO2 = 1.25 x primary energy CO2. See table 7.

 

https://www.iso-ne.com/static-assets/documents/2018/01/2016_emissio...

http://www.windtaskforce.org/profiles/blogs/natural-gas-is-good-for...

 

Table 7/NE system	2014	2015	2016	2016	2016
	PE basis	PE basis	PE basis	PE basis	SE basis
	Fed to grid	Fed to grid	Fed to grid	Fed to meters	Fed to meters
lb CO2/MWh	726	747	710	763	824
g/lb	454	454	454	488	527
g CO2/kWh	330	339	322	347	374

In 2016, total NE generation was 105,572 GWh, of which 62,284 GWh (59%) had CO2 emissions and 43,288 GWh (41%) did not.

 

ISO-NE calculates LMU marginal CO2/MWh and allocates CO2 to each state, based on the generators located in that state.

 

In table 8 the calculated 105553 GWh is slightly different from the actual 105572 GWh, due to rounding. See URL.

 

Table 8/State	CO2 emission	CO2	Generation
	US ton	lb/MWh	GWh
CT	10,179,000	572	35591
ME	2,994,000	678	8832
MA	15,011,000	897	33469
NH	5,508,000	573	19225
RI	3,044,000	930	6546
VT	732,000	775	1889
Total in 2016	37,468,000	710	105553
Total NE generation in 2016, GWh	105572
Generation with CO2, GWh	62284
Generation without CO2, GWh	43288
37468000 x 2000/(105572 x 1000) = 710

APPENDIX 3

CO2 Emissions of VT Electricity Sector, PE basis

Table 9 includes data from the three URLs.

 

- VT-ANR estimated, with input from the VT-DPS, the 2015 CO2 emissions of the VT electricity sector.

- ISO-NE monitors and records the output of VT generating plants connected to the NE grid.

- ISO-NE calculates the CO2 emission of these plants, based on fuel consumption and kWh produced at various plant outputs.

- ISO-NE calculated 732000 short ton (664060 metric ton) of CO2 from 1889 GWh of in-Vermont generation; 775 lb/MWh, or 352 g/kWh, fed to the grid basis. See note and pages 19 and 20 of ISO-NE URL.

- ISO-NE calculated CO2 emission intensity would be about 352 g/kWh x 1.075, T&D factor = 378 g CO2/kWh, at user meters.

- H-Q supply, about 1348 GWh in 2016, about 98% hydro, was assumed to have 0 emissions.

- Emission from other generation and from supply to utilities is 335940 metric ton for 2793 GWh (by subtraction), or 119 g CO2/kWh, fed to grid basis. See table 8 and URLs.

 

NOTE: It is not clear to me how ISO-NE determined there was 1889 GWh of in-Vermont generation.

 

https://www.iso-ne.com/static-assets/documents/2018/01/2016_emissio...

http://www.windtaskforce.org/profiles/blogs/vermont-far-from-meetin...

http://dec.vermont.gov/sites/dec/files/aqc/climate-change/documents...

 

Table 9	CO2 emissions	Utility supply	Fed to grid	At user meters
	PE basis	PE Basis	PE basis	PE basis
	metric ton	GWh	g CO2/kWh	1.075, T&D factor
VT-ANR, 2015	996000	6030	165	178
ISO-NE, instate generation, 2016	664060	1889	352	378
H-Q supply, about 98% hydro	0	1348	0	0
Emissions, other generation	335940	2793	119	128

APPENDIX 4

Table 10 was prepared from Energy Action Network data. The 2016 data were updated by EAN in 2018

 

Table 10/Energy Action Network
Vermont 2016 Electricity Sources	Electricity	% of total	Efficiency	Power plant input	% of total  PE
2018 update	MWh		%	million Btu
Biomass (wood chips)*	465470	7.7	22.75	6982047	18.0
Distillate (oil)	7288	0.1	31.89	77987	0.2
Farm methane*	22674	0.4	100.00	77364	0.2
H-Q system mix*	1347714	22.4	96.00	4790000	12.3
Hydropower*	720389	11.9	100.00	2457967	6.3
Landfill methane (muni refuse)*	95934	1.6	100.00	327327	0.8
Natural gas	17766	0.3	43.19	140350	0.4
Nuclear	773705	12.8	32.81	8046529	20.7
Solar*	256834	4.3	100.00	876318	2.3
ISO-NE Non-RE	1593721	26.4	43.19	12590386	32.4
ISO-NE RE*	250863	4.2	100.00	855945	2.2
Wind*	477332	7.9	100.00	1628658	4.2
Total	6029690	100.0	52.95	38850878	100.0
Total, RE	3610256	59.9	68.82	17899826	46.1
Total, Non RE	2419434	40.1	39.40	20951052	53.9

 

APPENDIX 5

2018 US Monthly sales of top 6 EVs and hybrids and total sales YTD. Prius Prime and Volt are hybrids. Tesla will sell about 200,000 EVs in 2018, based on present production rates. Hybrids are marked *. The low-priced Prius should be selling much better, but the styling is an abomination.

 

Table 11/2018	Jan							Aug	Sep	YTD
Tesla Model S	187	2485	3820	3750	6000	5902	14250	17800	22250	78,132
Toyota Prius Prime*	1496	2050	2922	2626	292	2237	1984	2071	2213	20,523
Tesla Model S	800	1125	3375	1250	1520	2750	1200	2625	3750	18,395
Tesla Model X	700	975	2825	1025	1450	2550	1325	2750	3975	17,575
Chevrolet Volt*	713	983	1782	1325	1675	1336	1475	1825	2129	13,243
Chevrolet Bolt	1177	142	1774	1275	1125	1083	1175	1225	1549	11,807