GREET_2021 LCA of Various Vehicle Technologies

wxman

Veteran Member
Joined
Oct 26, 1999
Location
East TN, USA
TDI
Other Diesel
The latest version of Argonne National Laboratory's GREET (Greenhouse gases, Regulated Emissions, and Energy use in Technologies) model was recently released (https://www.greencarcongress.com/2021/10/20211012-greet.html).

Based on this latest GREET model (GREET_2021), the following are graphical representations of health and environmental damages from various vehicle technologies and fuel pathways...








Damage factors used to calculate the health and environmental damages were obtained from a European Environmental Agency report (https://www.eea.europa.eu//publications/the-first-and-last-mile - Table 2.1 on page 24), which represent the latest data available.

The social cost of carbon (SCC) used for the second graphic was obtained from the current U.S. administration's value of $51/ton of CO2 (https://www.scientificamerican.com/article/cost-of-carbon-pollution-pegged-at-51-a-ton/).
 

nwdiver

Veteran Member
Joined
Sep 27, 2015
Location
Texas
TDI
2003 Jetta TDI (sold); 2012 Tesla Model S
Seems like a critical factor is going to be vehicle lifespan. From what I read GREET assumes a life of 150,000 miles? My EV is already at 195,000 on the original battery so vehicles exceeding this assumption and pushing the average up would have a significant impact on lifecycle emissions.

The marginal energy cost of charging is also difficult to really integrate. If they're charged using curtailed energy how would that be assessed? Is it really accurate to assume the lifecycle cost of the source if they're using energy that would be wasted if the vehicle didn't exist? To some extent the overall energy impact is actually negative.
 

wxman

Veteran Member
Joined
Oct 26, 1999
Location
East TN, USA
TDI
Other Diesel
The default vehicle useful life in GREET is 173,151 miles, but the bulk emissions of the vehicle manufacturing process are given for each vehicle technology...






You can divide each of those emission values by whatever lifetime in miles you want. Are there any good data on how EV UL differs from ICEV UL?

Using renewable electricity for charging an EV should correspond to the "BEV LR (solar)" bar in the graph even though e.g., wind, has slightly lower overall emissions than solar among the RE sources. Now sure how to handle the curtailed generation. Any suggestions?
 

nwdiver

Veteran Member
Joined
Sep 27, 2015
Location
Texas
TDI
2003 Jetta TDI (sold); 2012 Tesla Model S
Now sure how to handle the curtailed generation. Any suggestions?
Seems almost like it should count as 'negative emissions' or zero. If an EV is being used to reduce curtailment it basically becomes infrastructure. If not for EVs the utility would need to invest in grid storage or that energy simply would not exist.

There was an article published in 'Nature' pointing out that by oversizing renewables by 50% the grid can be 100% decarbonized with only ~12 hours of storage. So that would leave A LOT of free energy available for flexible loads like EV charging. Even now SPP has tossed ~20GWh of wind energy two nights in a row due to lack of demand. Enough energy for ~2M EVs.

If generating capacities are instead increased so that annual generation exceeds annual demand in each country by 50% (i.e., 1.5x generation), but without energy storage, the most reliable mixes meet 83–99% (average 94%) of electricity demand. The 1.5x generation most reliable mixes are substantially more reliable than in the 1x generation systems but include more wind power: 70–90% wind power (78% on average; Fig. 2d). These “overbuilt” systems are more reliable in all of these 18 countries than the systems with 12 h of energy storage but no excess generation (Fig. 2c). Adding energy storage to systems whose generation is 1.5x annual demand again increases both the system reliability (89–100%, average 98%) and the share of solar generation (most reliable mixes have 10–60% solar power, 36% on average; Fig. 2e, f).
 

wxman

Veteran Member
Joined
Oct 26, 1999
Location
East TN, USA
TDI
Other Diesel
Still don't know how to quantify that. The BEV charged by solar is effectively zero emissions in the fuel cycle; the only damages from emissions in the fuel cycle are from background emissions from producing the solar panels.

I meant to mention that GREET_2021 has 10 battery chemistry options, including LFP solid state. I used the default which is NCM111 if I recall correctly. Is that an appropriate battery chemistry to use?

There are also four range options for the BEV; 200 miles, 300 miles, 400 miles, 500 miles. I used 300 mile range option. Would that be the best choice?
 

nwdiver

Veteran Member
Joined
Sep 27, 2015
Location
Texas
TDI
2003 Jetta TDI (sold); 2012 Tesla Model S
Still don't know how to quantify that.
Really just need to view any life cycle analysis in the context of both the assumptions and technology path. Imagine a similar study conducted in 1980 when solar was 0.001% of the grid and ~50x more energy intensive to manufacture than today. But even then it was on a path to lower costs. Every year there is more wind and solar curtailment. By 2030 it's very likely that >50% of the energy used to charge most EVs will be energy that would be wasted without those EVs on the grid. Basically free energy. Meanwhile there's less oil and the remaining oil requires more energy to get to and refine
 
Last edited:

turbobrick240

Top Post Dawg
Joined
Nov 18, 2014
Location
maine
TDI
2011 vw golf tdi(gone to greener pastures), 2001 ford f250 powerstroke
It seems to me that LFP is likely to be the dominant chemistry this decade. I don't think any of the big players in the EV space are still using the old NCM 111 chemistry.
 

TDIMeister

Phd of TDIClub Enthusiast, Moderator at Large
Joined
May 1, 1999
Location
Canada
TDI
TDI

Takeways:
1. Demand-response based emission rates are not non-baseload emission rates. Non-baseload emission rates are calculated from underutilized power generation resources which are lower efficiency plants that typically have higher emissions than more modern plants. The demand-response based emission rates come from the plants that are regularly used.

2. Average mix emission rates are a lower accuracy estimate of the demand-response based emission rates and should only be used for present day analysis when hourly power plant generation and emissions data, and grid demand data is not available. In the U.S. this data is publicly available, so we have no excuse for using average mix emission rates in our vehicle LCAs.

3. The problem with average mix emission rates can be demonstrated with the removal of a 0.1 MWh load from a 1 TWh grid that has one non-emitting source of electricity (solar, wind, nuclear, hydro) and one emitting source (coal, gas, oil) which we'll assume produces 500 kg CO2/MWh. When we remove the small load, we turn off 0.1 MWh of emitting electricity production because we want to fully utilize the low carbon electricity at all times. The table below demonstrates the problem for two grid configurations: 100,000 MWh and 500,000 MWh non-emitting generation. In both cases the average mix emission rate (450 kg CO2/MWh and 250 kg CO2/MWh) under-predicts the CO2 for the 0.1 MWh load and the missing CO2 (5 kg and 25 kg) mathematically remains on the grid. Except that it doesn't, because we turned off that 0.1 MWh of generation. That's the mass conservation error in the calculation, which leads to an under-prediction of the appropriate emissions for the 0.1 MWh load of interest. We can fix the problem by correcting the final emission rates, and the last column shows how small the corrections are to the average mix emission rate, but it's important to recognize that the small correction applies to everything left on the grid (999,999.9 MWh) so it's a non-negligible amount of CO2. For the same scenario, the demand-response emission rate is 500 kg CO2/MWh.

4. The average mix error is worse at higher renewable generation levels so we need something much better for predictions of future emissions.
 

TDIMeister

Phd of TDIClub Enthusiast, Moderator at Large
Joined
May 1, 1999
Location
Canada
TDI
TDI
Further to wxman's graphs in the first post, a minority of people - of which I include myself - do not advocate singling out technologies depending on individuals' biases (EVs, ICEVs, FCVs, etc.) but rather promote an all-inclusive approach of diverse energy sources if we're all working in good faith toward the same shared objectives. #thefutureiseclectic
 

nwdiver

Veteran Member
Joined
Sep 27, 2015
Location
Texas
TDI
2003 Jetta TDI (sold); 2012 Tesla Model S
Further to wxman's graphs in the first post, a minority of people - of which I include myself - do not advocate singling out technologies depending on individuals' biases (EVs, ICEVs, FCVs, etc.) but rather promote an all-inclusive approach of diverse energy sources if we're all working in good faith toward the same shared objectives. #thefutureiseclectic
What's wrong with a 'first principles' approach. Every technology has a fundamental limit based on the physics it relies on to function. ICEVs are fundamentally limited to <40% efficiency. FCVs are fundamentally limited to a round-trip efficiency of ~50%. Best case for using H2 as a fuel is you need >2x more solar and wind vs EVs. Then there's the infrastructure bottle neck if you're comparing H2 to electrons. To be able to take advantage of 10kW of surplus wind or solar with H2 costs >$15k in electrolysis. While a 10kW EV charger costs ~$2k.

Not sure I would frame that as a bias. Just working from the ground up to determine which technology path holds the most promise.

And in your link.... why are they assuming demand-response would increase output from FF generators instead of from curtailed RE resources? That makes no sense. If you have 30GW of wind cutback to 20MW and you get another MW of demand you increase wind output by a MW... the marginal emissions of that MW are 0.
 
Last edited:

TDIMeister

Phd of TDIClub Enthusiast, Moderator at Large
Joined
May 1, 1999
Location
Canada
TDI
TDI
Well, your numbers above already demonstrate inherent bias rather than any first-principles approach. Where do you come up with ICEs with "fundamentally limited to <40% efficiency"? Where do you get FCVs "fundamentally limited to a round-trip efficiency of ~50%"? - the bias being on the optimistic side if round-trip means also the H2 production. PEMFCs achieve about 50-60% - SOFCs a bit more - just at the stack level, and the best efficiencies are achieved at low rather than high loads.

Nobody of any credibility discounts the top-line efficiencies of electrons directly generated and used to turn electric motors on a socket-to-wheels basis (yet a full accounting reveals that real efficiencies are still far lower than most of the BEV fanboys purport, that is, north of 75%). However, this is completely unrealistic and shows a lack of understanding of how electricity grids operate. If all the electricity generation were turned over to renewables in a short span, the result will be massive and severe outages. For every kWh of renewable electricity generation, about the same or more capacity of energy storage is required. If these were all provided by batteries, the environmental toll to extract that much battery raw materials would be a global disaster as big as the one it is purported to try solving (cf. the "battery manufacturing" component of the Health and Other Damages LCA graphs in the first post).

Conversion efficiency is not, should not and can not be the sole and facile merit that determines your choices if the energy resource is variable and mismatched to demand, and there are wide ranges of costs and hidden environmental impacts behind each option.
 
Last edited:

TDIMeister

Phd of TDIClub Enthusiast, Moderator at Large
Joined
May 1, 1999
Location
Canada
TDI
TDI
This is not at all to say that we should not go fully to renewables ASAP. I wholeheartedly agree that we should. But the path toward that should not be defined by one or just a limited few options but rather encourage a diversity of solutions, while not letting the perfect or best be the enemy of the good.

* Diverse circumstances require diverse solutions,
* What is best for the average is not best for all, and
* The solution to uncertainty is diversity.

 

nwdiver

Veteran Member
Joined
Sep 27, 2015
Location
Texas
TDI
2003 Jetta TDI (sold); 2012 Tesla Model S
Well, your numbers above already demonstrate inherent bias rather than any first-principles approach. Where do you come up with ICEs with "fundamentally limited to <40% efficiency"? Where do you get FCVs "fundamentally limited to a round-trip efficiency of ~50%"? - the bias being on the optimistic side if round-trip means also the H2 production. PEMFCs achieve about 50-60% - SOFCs a bit more - just at the stack level, and the best efficiencies are achieved at low rather than high loads.
It's not a bias. It's just physics. Carnot efficiency is a fundamental limit of all heat engines. Same with H2. There is a fundamental limit to how many kWh of H2 you can get from a kWh of electricity and an fundamental limit for how many kWh of electricity you can get from a kWh of H2. Realistically the limit is ~50%, kWh => H2 => kWh. If I chose Aluminum over lead for an Aircraft frame is that a 'bias' or just following the physics?

The fact we need storage to buffer renewables is exactly WHY EVs make the most sense. An EV is a battery. Instead of investing in 1MWh of grid storage why not invest in ~150 EVs on a demand response scheme?
 

TDIMeister

Phd of TDIClub Enthusiast, Moderator at Large
Joined
May 1, 1999
Location
Canada
TDI
TDI
It's not a bias. It's just physics. Carnot efficiency is a fundamental limit of all heat engines. Same with H2. There is a fundamental limit to how many kWh of H2 you can get from a kWh of electricity and an fundamental limit for how many kWh of electricity you can get from a kWh of H2. Realistically the limit is ~50%, kWh => H2 => kWh. If I chose Aluminum over lead for an Aircraft frame is that a 'bias' or just following the physics?
There are many corrections that I can make to your understanding of thermodynamics, but I get what I believe is the underlying point of your argument - that ICEs and FCs are less efficient than simply shuttling electrons from a generator somewhere to the motor driving the wheels of your EV with an on-board battery in between. I get that, and there's no argument of the in-use efficiency figures or the lack of tailpipe emissions per se, but you let these facts be license to gloss over the very significant INefficiencies and environmental impacts under the surface in the manufacture of those battery packs, in particular, which is the point of the OP. This does not require an understanding of physics - it requires seeing the bigger picture rather than a narrow focus.

The fact we need storage to buffer renewables is exactly WHY EVs make the most sense. An EV is a battery. Instead of investing in 1MWh of grid storage why not invest in ~150 EVs on a demand response scheme?
OK, suppose we have such a scheme in place. Will you be pleased to give the power and decision over to the electrical utility as to when or how much to charge your EV? Suppose you are looking forward to a road trip and you plug your EV to your charger at home, but because the utility determines that there is an acute shortage of electricity at the time, it delays or throttles the charging - or worse - temporarily "borrows" electrons from your batteries promising to "return" it but failing to do so in time for when you unplug and set off on your road trip, so that instead of an expected full charge, you are at a much lower SOC. Later, you stop somewhere along your trip and make an obligatory stop at a public fast charger, which you also double as a meal break. You return to your car to discover a lower SOC than expected because, again, the utility's algorithm has determined a more pressing need for the electrons elsewhere and throttled your station's charge rate or even leave it temporarily off-line (or you pay through the nose for at a variable rate for the privilege of the full charge). A small number of people might be willing to put up with this inconvenience for the conviction that they're saving the world - I assume you, @nwdiver, can be counted in this number; God bless them and you - but I suspect the vast majority will not.

The funny thing is, I'm completely on the same page as you on most things.
  • We need to transition to renewables even faster than ASAP
  • I'm actually quite bullish on EVs; I believe they play an important role in transport. Where we differ is that I believe there is an equally important role for all other propulsion and energy storage technologies for the right respective use cases and where they make the most economic sense.
What I'm not OK with is the self-righteous smugness that some - not all - EVers have over other technologies. I am equally if not more disgusted by the despicable actions of ICErs and people who defeat and delete emission control systems in their ICEVs.

The bottom line is, I believe we should see everyone who's working hard to solve the big problems of our time put our heads together in a spirit of cooperation, and regard each other healthy competitors rather than bitter rivals.

I invite you to watch this Youtube video in its entirety before forming a preconceived conclusion:
https://bit.ly/gzDfEhVcF6o
In it, I come straight out to say that I think Li-ion batteries are better used in consumer electronic products and electric cars than for grid electricity storage.
 

nwdiver

Veteran Member
Joined
Sep 27, 2015
Location
Texas
TDI
2003 Jetta TDI (sold); 2012 Tesla Model S
OK, suppose we have such a scheme in place. Will you be pleased to give the power and decision over to the electrical utility as to when or how much to charge your EV? Suppose you are looking forward to a road trip and you plug your EV to your charger at home, but because the utility determines that there is an acute shortage of electricity at the time, it delays or throttles the charging - or worse - temporarily "borrows" electrons from your batteries promising to "return" it but failing to do so in time for when you unplug and set off on your road trip, so that instead of an expected full charge, you are at a much lower SOC.
Why does it have to be all or nothing? ~90% of the time I just want ~20kWh over the next 4 days and ~99% of the time there's going to be surplus available on the grid for me to get it. Xcel has a plan in the works to pay $50/yr for an EV program. Not great but $50 is $50 and there really aren't any strings attached. If I need energy that 10% of the time I can take it. They still get to use my car as a buffer the other 90% of the time. And it doesn't have to be bi-directional. The biggest issue now is surplus. It's gonna be a while before feeding energy back into the grid really make that much sense.

I don't think it's common knowledge how much clean energy is thrown out. Keep an eye on the SPP portal. Over the course of this week they're easily going to toss >100GWh worth of wind energy. That's enough for nearly 2M EVs. At the end of the day money talks and utilities save a lot of money getting people to take this otherwise wasted energy. Increasingly this will be passed on as an incentive.

This study found that EVs pay off their embedded energy vs ICE after ~13,000 miles.
 
Last edited:

turbobrick240

Top Post Dawg
Joined
Nov 18, 2014
Location
maine
TDI
2011 vw golf tdi(gone to greener pastures), 2001 ford f250 powerstroke
My intuition tells me that the GREET models dramatically understate the total health and climate costs associated with anthropogenic global warming. Unfortunately, I believe time will bear that out. I don't think it's a linear relationship either, ie as we approach a tipping point, Carbon emissions become more and more costly. Snow cover melts, increasing solar absorbtion. Tundra thaws, releasing massive methane emissions, and various other climatic factors cascade in a domino effect that is understandably difficult to model.
 

wxman

Veteran Member
Joined
Oct 26, 1999
Location
East TN, USA
TDI
Other Diesel
I don't disagree that SCC may be underestimated, but it's the Biden administration that has set that $51/ton CO2 cost, not GREET...

...the Biden administration used the Obama-era formula for a central social cost of carbon for 2020 of $51 a ton, with methane and nitrous oxide, which both pack a stronger climate punch than CO2, at $1,500 a ton and $18,000 a ton in 2020, respectively. These would rise to $85 a ton for CO2, $3,100 a ton for methane and $33,000 for nitrous oxide by 2050 as damage from warming is expected to progress....

The GREET model calculates only criteria air pollutants and GHG emissions over the entire life-cycle of the vehicle.

The damage cost factors I used for the criteria air pollutants came from an European Environmental Agency report...



https://www.eea.europa.eu//publications/the-first-and-last-mile (Table 2.1 on page 24)

Conversion - €/kg X 1 kg/1000 g X 100 cents/dollar X 1.15 $/€ X 1.143 (inflation - https://www.usinflationcalculator.com/) = ¢/gram

VOC - €1.2/kg = 0.158 ¢/g

NOx (urban) - €21.3/kg = 2.800 ¢/g
NOx (rural) - €12.6/kg = 1.656 ¢/g

PM2.5 (metro) - €381/kg = 50.080 ¢/g
PM2.5 (city) - €123/kg = 16.168 ¢/g
PM2.5 (average urban) = 33.124 ¢/g
PM2.5 (rural) - €70/kg = 9.201 ¢/g

PM10 (PM10-2.5) - €22.3/kg = 2.931 ¢/g

SOx - €10.9/kg = 1.433 ¢/g

I used "City" damage factors for all vehicle operation emissions (a simple average of "Metropolitan" and "City" for PM2.5), and "Rural" damage factors for all upstream emissions included well-to-pump and vehicle manufacturing. Ammonia (NH3 ) is not calculated by GREET and is thus not included.
 

turbobrick240

Top Post Dawg
Joined
Nov 18, 2014
Location
maine
TDI
2011 vw golf tdi(gone to greener pastures), 2001 ford f250 powerstroke
Whomever came up with that $51/ton figure seriously missed the mark, imo. It's probably triple that number or more.
 

nwdiver

Veteran Member
Joined
Sep 27, 2015
Location
Texas
TDI
2003 Jetta TDI (sold); 2012 Tesla Model S
Whomever came up with that $51/ton figure seriously missed the mark, imo. It's probably triple that number or more.
Might have been one of them 'Texas Sharpshooter' things. You pick the target you think you can meet.
 

wxman

Veteran Member
Joined
Oct 26, 1999
Location
East TN, USA
TDI
Other Diesel
Here is an updated damage assessment of various vehicle technologies and fuel pathways using the latest version of GREET (GREET_2023 - https://greet.anl.gov/) depicted graphically. Used the same methodology as described in post #17.


 

TDIMeister

Phd of TDIClub Enthusiast, Moderator at Large
Joined
May 1, 1999
Location
Canada
TDI
TDI
RD looks very good! I fueled up with it exclusively whenever I could find it during a road trip to California last summer. The 76 app was very useful for the purpose and I also got significant discounts (as much as 70 cents per gallon at the time) for paying for fuel with it.
 
Top