Team-BHP > Electric Cars
Register New Topics New Posts Top Thanked Team-BHP FAQ


Reply
  Search this Thread
12,854 views
Old 7th December 2023, 11:09   #16
Newbie
 
Join Date: Nov 2023
Location: Chennai
Posts: 19
Thanked: 109 Times
Re: Electric Vehicle Architecture | Why does it matter?

As someone who has some experience working on battery packs, I would like to add a few technical clarifications here.

High Voltage (HV): Technically speaking, any system operating above 60 V DC is classified as HV. This is mainly because coming in contact with DC voltage above 60 V could harm human beings. So, 12 V system in ICE vehicles and even the 48 V systems fall under the LV category. All Battery Electric Vehicles in production are HV (typically in the range of 300 V - 800 V). The high voltage architecture being discussed here is in this context, even though technically 300 V is also HV. All HV connections used (including the harnesses/cables/busbars) must be rated for the operating voltage and current. There are certain IEC standards which regulate the insulation requirements, clearance and creepage distances, etc. under different environmental conditions.

Clearance: In simple terms, this is the minimum air gap between two electrical conductors to avoid arcing between them.
Creepage distance: Minimum length, along the surface, of insulating material required to avoid "tracking" of current between two electrical conductors.

Attachment 2540503

Whether the EV uses 300 V system or 800 V system, the design target is to keep the cells at the optimum temperature at all times. This calls for two things:
1. Optimum thermal management: Proper cooling/heating of the cells
2. Prevent overheating

Preventing overheating is addressed at multiple levels:
1. First is the basic design of all HV conductors with appropriate cross-sections to limit the maximum temperature they operate when the vehicle draws the maximum current. This applies to both 300 V as well as 800 V systems. Larger the conductor cross-section, less the electrical resistance, hence lower Joule heating. This is where 800 V systems hold an advantage, as their peak currents are lower for the same power (V*I). Hence 800 V systems have conductors of smaller cross-sections --> lower weight.
In my opinion, a properly designed battery pack of 300 V and 800 V must be operating at similar maximum temperatures of the conductors (HV wiring harness/busbar) to optimize the weight. The maximum temperature is also limited by the insulating material used on them.
2. The second method is through the battery management system which controls the coolant temperature/flow and also limits the current drawn if it senses a potential overheating concern.

Quote:
I don't know why Apple didn't engineer it differently but this is true even in my latest iPhone 15, maybe they benefit from batteries dying faster as firstly they sell their batteries at a ridiculous premium and their batteries cost as much as what complete entry level phones do, or even better for them the people will be compelled to sell their perfectly fine working phones instead of replacing the batteries, either way they win.
(My iPhone 11's battery lasted me 4 full years as opposed to 2.5 years in my sister's iPhone 12, the difference being me using slow charging and her using fast charging)

Won't batteries heating more be an issue here as well or is it something different?
I would like to add a few things here. The life of a cell or a battery is dependent on multiple things. Life = Capacity (in terms of the maximum Ah it can provide). At the beginning of life (BOL), i.e., a brand new cell, the capacity is 100%. Capacity fades over time and based on usage. Based on the application (mobile phone, EV, UPS, industrial, etc.), the end of life (EOL) capacity could differ. E.g.: An EV with 50% capacity is almost equivalent to saying it has 50% range (km) compared to when it was brand new. Cell life is usually mentioned in terms of charge-discharge cycles.

One full Charge-discharge cycle is defined as draining a cell from 100% charge to 0% charge. This is not achieved practically - the BMS itself stops such a deep discharge, and usually a full charge as well. So, if a cell is always charged to around 90% and discharged to around 40%, it is not completing a full charge-discharge cycle. BMS calculates the fraction of cycle and estimates the number of cycles it has completed.

As mentioned above, cell capacity fading depends on the usage.
1. Temperature
As mentioned earlier, cells operate the best at an optimum temperature range. If it goes below or above the limit frequently, its life reduces.
2. Depth of discharge
The more a cell is subjected to higher discharge (more depth of discharge - like 100% to 20%), the less its life
3. C-rate
C-rate is the amount of charging/discharging current. It is actually the ratio of current to the cell capacity in Ah. Higher C-rates correspond to drawing more current or in the scenario of fast charging. Every cell manufacturer defines the maximum C-rates allowed in charging and discharging, and this is controlled by the BMS. Higher C-rates frequently will reduce the cell life. In simple terms, frequent fast charging is not good for the cell. But this again depends on the C-rate limitations of a BMS-cell combination.

The discussion on cell life is based on many research articles on this topic, and it is an evolving field.

In mobile phones, there is something we ignore sometimes. Every mobile phone is recommended to be used with its own charger (including the cable and the power adaptor). This is because that combination only can control the current/power supplied to the phone battery. Frequently using random chargers could be detrimental to the phone battery. In EVs, these are standardized and hence controlled.
Prado is offline   (5) Thanks
Old 9th December 2023, 11:56   #17
BHPian
 
Join Date: Jan 2008
Location: Bangalore
Posts: 954
Thanked: 1,544 Times
Re: Electric Vehicle Architecture | Why does it matter?

Electric Vehicle Architecture | Why does it matter?-screenshot_20231209115320_whatsapp.jpg

Any merit in this argument?

Is there a valid counter argument that tilts the balance in favor of BEVs?
vharihar is offline  
Old 9th December 2023, 12:01   #18
Distinguished - BHPian
 
Join Date: Oct 2012
Location: Delhi
Posts: 8,463
Thanked: 55,046 Times
Re: Electric Vehicle Architecture | Why does it matter?

Quote:
Originally Posted by vharihar View Post
Attachment 2541196

Any merit in this argument?

Is there a valid counter argument that tilts the balance in favor of BEVs?
Yes, even taking all of that into account, EVs still work out more environmentally friendly.

Jeroen
Jeroen is offline   (2) Thanks
Old 9th December 2023, 16:41   #19
BHPian
 
Shresth_EV's Avatar
 
Join Date: Jan 2021
Location: Lucknow
Posts: 312
Thanked: 1,112 Times
Re: Electric Vehicle Architecture | Why does it matter?

Quote:
Originally Posted by vharihar View Post
Is there a valid counter argument that tilts the balance in favor of BEVs?
You don’t have to look any further than the ICE equivalent of the same ie oil extraction, refining and transport. TLDR for those who don’t wish to read a wall of text :

Crude oil and refined derivatives of oil make up 40% of global shipping by weight.

Global shipping produced 646M tons CO2 in 2020 (IEA) 40% of that is 258.4M tCO2 or 258.4trillion gCO2 https://www.iea.org/fuels-and-techno...ional-shipping

In same period, oil production was 93.9M barrels/day or 14390 Million L per day, or 5.45trillion L per year https://www.iea.org/reports/oil-2021

Adjusted emission in gCO2/litre = 47.4gCO2/litre

Applying the same to refining, at 614gCO2/kWh like the article assumes for the grid, is 1535gCO2/litre (assuming electric furnaces used for fractional distillation) in 100% coal power and for 60% coal this is 921gCO2/litre of oil.

Now adding 47g from shipping ie 968g.

Out of this, roughly 40% becomes petrol.

Net “emission backlog” = 387g/L (40% of 968)

If a car like Nexon petrol gives 15kmpl, the effective per km CO2 from oil supply chain is 26g/km, in addition to the 120-130g/km from tailpipe itself. Reference : Pdf ph11, indexed pg6 : https://theicct.org/wp-content/uploa...s-wp-FINAL.pdf

Even if you assume warranty = life of battery pack, at 160K km, the Nexon battery did 3tons of CO2 emissions, for same distance, petrol will do 4.16tons for oil production alone let alone tailpipe. And this battery is recyclable to 92% which means 100 old batteries can make 92 new batteries. Oil isn’t, it’s extracted throughout life of vehicle

This is BEFORE adding the pumps and oil rigs used to extract oil, and the last mile supply chain of tankers or rains which transport from refinery to pumps. Oil rigs are self sustaining work and accommodation for workers in middle of sea, how much metal used there? How much emissions by the ship which carried the parts of the rig there?

On topic of cobalt, The good news for ev’s is that cobalt is not a necessity and non cobalt chemistries exist — such as the widely use LFP (entire Chinese market, Tata, MG, BYD and lower trim teslas) and the bad news for ICE is that Cobalt is used in desulphurisation of crude oil as catalyst.

But how much, we don’t know exactly. There are very few sources regarding this, the limited ones that are present are paywalled research articles.

The once which aren’t are basically some version of “it’s a trade secret so we can’t tell you how much cobalt we use.” How convenient innit. OTOH, battery composition is in public domain and completely transparent. Who is having things to hide?

https://www.reddit.com/r/oil/comment...ion/?rdt=40490

Quote:
I work for a company that designs HDS units and sells HDS catalyst.

Basically what you are asking are trade secrets, so I can't give you a straight answer. You can however, do some research online of refining training courses to get a ballpark answer. You need to find:

The average Co content in a HDS catalyst; you can find some typical formulations online.
The fuel volume (diesel + kero + naphtha) sold annually.

The average catalyst life expectancy, which varies between services and licensors. As a hint, I can tell you that your average CoMo catlayst can be regenerated a few times before it's dead.
The average LHSV of a HDS reactor. This is the most technical part.
That should do it.
Other estimates refer to entire cobalt supply in catalyst sector so if we take that value ([190,000 metric tons](https://www.statista.com/statistics/...ne-production/)) and apply that to [catalyst sector (7%)](https://www.researchgate.net/figure/...fig1_326060301), that’s around 13,300 tons per year.

In same period, oil production was 93.9M barrels/day or 14390 Million L per day, or 5.45trillion L per year https://www.iea.org/reports/oil-2021

This comes to 2.44mg/litre of fuel refined

Although unlike a battery, this is a recurring expense (for each litre you need that much cobalt)

For battery, it’s a one time bulk usage. Visualizing the Key Minerals in an EV Battery

An NMC 60kWh battery like the one in model 3 will need around 8kg cobalt in it.

For driving an equivalent ICE like BMW 330i would, for argument sake, driven for the warranty period of the EV ie 1,60,000km, would need 16000L at 10kmpl. That comes to 39kg of cobalt catalyst usage for usable life of vehicle.
Shresth_EV is offline   (3) Thanks
Old 9th December 2023, 21:41   #20
Senior - BHPian
 
SKC-auto's Avatar
 
Join Date: Jan 2020
Location: KA01/AP31
Posts: 1,254
Thanked: 3,721 Times
Re: Electric Vehicle Architecture | Why does it matter?

Quote:
Originally Posted by vharihar View Post
Attachment 2541196

Any merit in this argument?

Is there a valid counter argument that tilts the balance in favor of BEVs?
There is tons of research available on the same, a simple search on Google shows actual research not memes.

Those minerals from the meme can be recycled in perpetuity(95% materials can be extracted today), ICE cars will only be cleaner if the tail pipe emissions are captured and converted to synthetic fuels using renewable energy.
Quote:
In 2020, building the world’s wind turbines, solar panels, EVs and other clean-energy infrastructure demanded 7 million tons of minerals, estimates the International Energy Agency. Roughly half of this was destined for batteries and EVs.

The oil, gas and coal industry, by contrast, extracted the equivalent of 15 billion metric tons in 2019. And the industry will need to extract it year after year to keep supplying energy. Clean-energy technology can use these materials for decades or, if recycled, in perpetuity.

Extracting 1 ton of copper, for example, requires digging up around 100 tons of ore. But even accounting for this, estimates Sam Calisch, a scientist at the nonprofit Rewiring America, mining minerals for the clean-energy economy amounts to extracting about five times less matter than what’s extracted by the fossil fuel industry. “This is still massive,” Calisch says.

Noah Horesh, a researcher at Colorado State University who studies life cycle emissions in the transportation sector, has analyzed vehicles’ emissions over their life spans. Horesh estimates fossil fuel vehicles generate roughly twice the emissions of an EV, even accounting for emissions from extracting added minerals and metals.

This difference will only grow as the electricity sector decarbonizes, and battery manufacturing becomes far more efficient. People recharging with clean electricity, or driving smaller vehicles, may already see a bigger difference today.

https://www.washingtonpost.com/clima...-fossil-fuels/

Last edited by SKC-auto : 9th December 2023 at 21:45.
SKC-auto is online now   (2) Thanks
Reply

Most Viewed


Copyright ©2000 - 2024, Team-BHP.com
Proudly powered by E2E Networks