It’s amazing how fast we went from rarely seeing an electric Amazon van near me to seemingly most of them being EVs.
Makes so much sense. I wonder how many (tens of?) thousands a single warehouse location might save per day in gas alone. Last mile in urban/suburban areas is nearly ideal for an EV.
The upfront is very expensive with all the ancillary costs, but it’s hot to pay off handsomely after a while. And if a real carbon tax were to ever appear it would be all the better.
It pays off fast. For my situation in Washington, I'd be paying ~$4.50/gal for an ICE car that gets ~28mi/gal, or ~$0.11/KWh (at the tier-2 rate) for an EV that gets 3.5mi/KWh.
For the ICE car, that's ~6.2 mi/dollar, for the EV it's ~31.8 mi/dollar. ~5x cheaper per mile to drive the ICE. I drive ~15k miles per year, so ~$2.4k in gas, or $470 in electricity - saving ~$2k/year.
This doesn't take into account the lower total cost of ownership of EVs in general. They're simple. They don't need oil changes. The transmission can't go out.
And that's just for me driving once or twice a day; I'm sure the Amazon drivers are putting factors more miles on their cars, so a faster break-even point.
It’s even better for delivery vans, too: EVs thrive in repeated stop-start conditions, which is what Amazon cans frequently do.
It seems like only a matter of time before other similar vehicles that are more industrial make this shift (eg, garbage/recycling trucks). Imagine no more super loud garbage trucks in the mornings…
I chatted with a friend who used to be a garbage man in a mid-sized west coast city. Most drivers in our area have a daily route, and when they are done, they are done, but paid for the whole's day work.
So they have an incentive to floor the throttle, and slam the brakes. The drivers don't pay for fuel or brake pads out of their salary, the company does, but if they drive like a maniac, they can finish a few hours earlier.
Are accelerometer logs widespread now? Even with 100 USD phone hardware, companies could check drivers are not causing too much wear or causing disomfort to passengers.
Are transmission failures really that common? In nearly fifty years of driving and owning eight different cars I've never had a transmission failure. One of the cars (Volvo Amazon estate car) did 350k miles including the previous owner rallying it for a couple of years. Apart from fixing a leaking radiator when I bought it I never had any problems with the motor or gearbox. We scrapped it just because my wife wanted a newer more comfortable car.
I’ve owned 3 cars in my life, one of them driven into the ground, one high mileage and given away when we moved to the city, and the third we still own. A transmission is the one thing I’ve never had to fix. Now that’s just me. Obviously transmissions do fail, but I’d bet in aggregate, it’s not nearly the most pressing issue. But it is the second most complicated set of moving parts in the car, so when it goes, it’s not cheap.
All that said, electric cars really shine in city driving. No transmission to go, and regenerative braking to save the pads. It’s valid to call it out if that’s your use case like these Amazon trucks.
A lot of the major cars makers have had years of bad transmissions. Just Google “ transmission failure class action” with manufacturer names and you’ll see how many settle. But not all years and not at the same time across manufacturers, so it’s easy to miss bad ones.
A bad transmission design can really leave a mark, particularly when the company using it tries to sweep it under the rug. Talk to owners of vehicles with Ford's "Powershift" dual-clutch automatic transmission and you'll hear nothing but horror stories.
There was (is?) a time where if you replaced the tranny fluid with anything BUT the exact Chrysler one, it would kill it dead pretty quick. Some people would buy up minivans with dead transmissions at 80k miles or whenever the quick lube place told them to flush the tranny, then clean it out and put the right fluid in and it'd work fine again.
My mom didn't even get one home from the dealership when it seized up on the highway. Something about the solenoid pack but don't take my word for it, I was all of 8.
Transmissions tend to have a lot of things that move around. Gears might be ok for a long time by themselves, but when there's gear selection, you get extra wear and tear when things don't meet up quite right. And everything gets hot, so you need lubricant/cooling and that can get gunked up by the wear and tear...
And then sometimes things aren't built as designed. The ford HF35 hybrid transmission should be pretty reliable as designed, but many of them were built with bearings (or something) that went bad, leading to an early salvage date for many C-Maxes.
You never owned a Chrysler product, did you? They were notorious for taking a transmission from a much smaller car or truck and putting it into a bigger one because on average they would last the life of the 3-yr/36k mile warranty, and not much longer. Same with things like shocks and struts and springs.
GM on the other hand went the exact opposite direction most of the time and put an overengineered heavy duty transmission with maybe slightly cheaper internals into their smaller vehicles.
> GM on the other hand went the exact opposite direction most of the time and put an overengineered heavy duty transmission with maybe slightly cheaper internals into their smaller vehicles.
Eh, the owners of 80's era Chevy Celebrities/Pontiac 6000s/Oldsmobile Cieras/Buick Centuries might like a word.
The one car that I sold because it died on me was due to transmission failure. I’m also at 8, and do not expect either of my current cars to ever have a transmission failure.
I come from a country where most people drive manual transmissions, and the rule of thumb is that you're replacing the clutch at 150-250k kilometers. Automatics have torque converters instead of clutches, so that's probably a non-issue. I do imagine that they're much more integrated, and without a sacrificial clutch plate to take load shocks, the entire system will give out after a certain distance. This is all conjecture though.
Automatics used to have torque converters. These days they've got multiple clutches are are mechanically more efficient and have more gears than a manual transmission. I still like my manual transmission car though.
My 1990 Nissan pickup made 380,000 km on its original clutch. 2.4 liters of 4 cylinders totally underpowered will do that. At least a starter every 100k and alternator every 150k though !
In my anecdotal experience I've found it to be not uncommon in work trucks used for commercial duty over 200k miles. Not sure about in the delivery van space.
It's a different level of wear and tear than a normal commuter passenger vehicle.
I think the bigger savings on a delivery van would be brakes. In constant stop-start driving using the motor for regenerative braking is both an efficiency saver and reduces wear and tear on the mechanical brakes.
"efficiency saver" kinda downplays the fact that the energy is being recovered which was previously just wasted. Not throwing all that energy away is material at Amazon scale.
I had a 2002 Saab 9-5 Sedan automatic transmission blow out on the day I bought it. ODO said about 300k km (190k miles give or take). The dealer replaced the car with a younger 9-5 Wagon with lower mileage so I'm kinda glad it did.
That 2002 model is supposed to be a lemon transmission-wise though (even so, 300k seems decent as far as lifespans go).
My Chevy cavalier died due to a transmission issue at around 125k miles. The Schuylkill expressway in rush hour didn't help. Very clonky shifts near the end and it would stall out when trying to find the right gear.
Energy costs in most of the US are favorable to this math. The cost of energy supply should be pushed down as renewable deployment continues indefinitely.
Unconstrained distribution costs because of lack of political will to aggressively regulate (PG&E, etc). Mostly a California problem, as long as PG&E is permitted to extract executive comp and profits while minimizing infrastructure investment. Supply costs are plunging during the day due to utility scale solar output.
> Spot power prices have been trending lower year on year on lower natural gas spot prices and more renewable generation supply, including hydropower following years of drought.
> CAISO has 19.011 GW of installed solar generation capacity, which accounts for the majority of renewable resources on the grid at 62%, according to the Key Statistics report. Solar capacity has increased nearly 15% compared to a year ago and 26% from two years ago.
> There are currently 28.6 GW of solar projects in the CAISO generation queue, including 13.8 GW that have executed interconnection agreements and nearly 4.5 GW of those have on-line dates expected listed for 2024. Only two solar projects slated to come online in 2024 do not have batteries connected.
It’s common to have a mix where some trucks make 1 route per day and others do several. So you can optimize by rotating loading times, spare vehicles, and having some onsite batteries.
Solar > battery means DC fast charging gets to skip AC>DC losses which are significant. But you don’t need to use 100% of solar’s output to save a lot of money.
I’ve already got them (and a battery), but it ain’t cheap! They’re also pressuring everyone to go electric/heat pump for everything, but then I’d have to start paying the extortionate electric rates here once again.
The lower TCO depends on the intended lifetime of the vehicle. For delivery use cases where the expected lifetime is probably on the order of 5-10 years, it’s believable that TCO is lower.
But if the expected lifetime is “as long as I can drive it,” then TCO is probably not lower, simply because the maximum lifetime of an EV is lower than that of an ICE vehicle. At some point the car will fail – it will need battery replacement (every X years), or it will depend on obsolete infrastructure, etc.
You can drive a 1960 Cadillac today. Will you be able to drive a 2020 Tesla in 2084?
A use case like this is probably one of the best for EV's. Presumably Amazon will be able to have a centralized maintenance facility where new batteries can be installed. They'll be able to price in their own maintenance and resale value will be of no concern.
I happen to live rather close to a large(?) Amazon fulfillment center they just built a few years ago. Our electricity isn’t very expensive either, though it’s not what hydro areas pay.
I don’t know if we’re a test site or it was just easy to get the power installed or what. Perhaps being a new facility meant they just needed new vehicles anyway so it made sense to not buy gas.
But once they started appearing they kept getting more common.
It would be interesting if Amazon would calculate how much CO2 they save because their customers don't drive to the supermarket but order from Amazon. With the data that Amazon certainly has access to, it should be possible to calculate this. Certainly considerable in the US.
1. The CO2 of the Amazon delivery needs to be taken into account (apportioned to the number of drops), vs delivery to store. Delivery to stores is highly efficient compared to last mile.
2. Most Amazon purchases probably wouldn't directly result in less trips to the supermarket (i.e. am I driving to the supermarket just to pick up one item, or am I going there anyway and picking up everything I need when I am there?)
Delivery to stores might be more efficient, but then the last mile cost doesn't go away. Most of the stuff that I would buy from Amazon isn't available at the stores that I go to so it would require an entirely separate trip to somewhere that I would not go otherwise. Often multiple stores.
It depends on where your stores are and how frugal the rest of your shopping is. My Trader Joes/Costco (five minutes away) is in an eight block shopping center with Best Buy, Home Depot, and a dozen other big box stores which I think is fairly common in America.
Between that and shopping for fresh meat and produce at ethnic markets that are all over the city, the vast majority of my shopping is amortized over staple groceries that I can’t or won’t buy online.
And in the EU at least, it's not as unfeasible as in most American cities to just do your grocery shopping with a backpack, since there are tons of walkable cities. I go about twice a week, usually when I pass one anyway on the way home, and haven't needed a car for the last fifteen years.
For use cases like this, the last mile is carbon neutral.
Edit:
Re-reading my comment, it comes across as if there were no walkable cities in the US. That was not my intention. Of course there are, but from what I've seen, not that many. Also, somehow there is a lot of political opposition to them, because of certain conspiracy theorists who seem to have taken over one of the political parties. And since there are only two worth mentioning, the situation is unlikely to improve.
Meanwhile, I would never dream of ordering groceries outside of specialty items through Amazon. They are prohibitively expensive on most food items that have any sort of weight compared to Walmart and Instacart. Granted, I don't have a Whole Foods nearby.
Me too. But coming back with 40 lb of groceries in hand is a chore. And if I can consolidate the list to a 24-hour schedule for one week period, that hammers it out majority of cases where emergency shopping is once in a blue moon.
Well, it cost me $15 more to deliver food from there. I save roughly an hour of time and do something more interesting instead. Also, not walking through the isles saves me from buying random junk.
Jokes aside, I save money by using grocery delivery.
Instacart already delivers within an hour to my home from a store of my choice. This means in addition to the regular grocery store items, I can get speciality items from the ethnic Indian, Chinese, Korean stores delivered too. With the subscription ($9.99/month) there are no delivery fees either. That’s going to be hard to beat.
Are you sure about that? Because if so I would reconsider my use of instacart. We buy a lot of stuff from Costco and it’s a 45 min drive away so we almost never go into the store.
EVs are cheaper per mile than ICE, with the notable drawbacks that the range is lower, the refill time is longer, and the batteries mean the vehicle weighs more. None of those 3 matter for Amazon deliveries - they’re doing local deliveries (in stop & go traffic, where the efficiency benefits of EVs are even more pronounced), they sit in one place overnight, and for cargo vehicles the weight difference isn’t as much of an issue. Amazon’s got a history of investing in long-term logistics projects (there was an article here a while back about their investment in shipping paying off during COVID), so I’d imagine they’ve penciled this out and are doing it for financial reasons. The company is notoriously focused on financial efficiency, but not necessarily short-term financial efficiency.
They may be cheaper per mile, but I doubt they are cheaper per delivered package when you take into account the purchase price of the vehicle at amazons 5.3% corporate bond rate (effectively the cost-of-money). The Rivian+EV charger probably costs ~$100k, whereas a regular mercedes would cost $60k.
Again, I’m sure that’s something that Amazon’s famously meticulous finance people missed in their calculations. What do you think the payoff horizon is for that kind of purchase for them?
Are they more expensive year one? Sure. New vehicles + chargers + infrastructure upgrades.
Year 5? Maybe it’s paid off. Energy costs, brake pads, no oil changes or other frequent maintenance.
When you refresh the EVs the infrastructure is still there, that’s one time. Gas prices go up? Your investment looks even better. Carbon tax appears? You’re golden.
Amazon must know it’s worth it. You don’t spend that kind of money for good PR alone.
And then, after putting solar on top of the warehouses and charging the batteries of the vans that aren't currently making deliveries, they get to have a closed system so they don't need to pay an external party.
Not sure about the US ones, but in Ireland they seem to use Mercedes eSprinters. These cost about 45k after grants (the new version, with a significantly bigger battery, is more like 55k, but I think they're on the old one). A normal Sprinter is more like 30k. It's not a huge difference, given the fuel cost diff.
Huh, that's a _huge_ gap. I'd wonder are those higher-spec/prosumer variants; I was using fleet prices.
EDIT: Hrm, no, a Ford Transit is also way more expensive in the US than in Europe. Wonder what's going on there. Is it just limited competition/more niche part of the market or something?
They're doing it, so obviously they think it makes financial sense. And considering how insanely successful and good Amazon is in logistics and online retail, I'm inclined to believe they know what they're doing over some random HNer who probably doesn't know the first thing about the topic. What's your field exactly? Moreover, where are you consuming your opinions about EV from?
Even if all that is correct, 300 miles a day or so could eat up a $40k difference really fast (gas alone could be $60 a day, or $15k a year - so even baking in the cost of electricity you could save the difference in only a few years).
Last-mile delivery vans are a pretty good use-case for EVs; they're relatively cheap from a capital perspective (they don't need huge batteries; they are not typically travelling hundreds of km/day) and they're way, way cheaper to fuel.
Even if it was more expensive, it could still make sense. It could be a hedge against a forecasted rise in gas prices. They could be planning to add power generation and get long-term price stability. They could believe stricter regulations over emissions are coming in a lot of their markets.
If they saved a bit of money on EV delivery vans and instead spent a little more on toilets[1], it would probably be far better use of money to buy good PR...
It would be interesting to know what the net gain is. Obviously there is the materials and labor cost of installation, but then you also have the battery cost. I assume most of the trucks are out during the day, so you need batteries to store the generated power until the truck can be charged at night.
And does Amazon own the warehouses outright, or do they lease them?
> I assume most of the trucks are out during the day, so you need batteries to store the generated power until the truck can be charged at night.
Why not just double your fleet size instead?
That's a bit extreme, but this is the kind of optimization that your bean counters are good at. You probably don't need to double your fleet but just keep enough that you can shift the consumption curve or take advantage of price changes.
Those numbers are very inflated, tho. A lot of Charge Point locations aren't open to the public. Some Ikea stores in the US use them to charge delivery vans, and those aren't open to the public.
I wouldn't be surprised if Tesla has more public chargers than anyone else on this list.
Yeah, so that just means that you can use your personal ChargePoint card to use the station. I once went through 5 different "public" charge point locations that weren't publically accessible at all.
I'm currently living in an apartment complex that has "public" charging stations, and most of them are on reserved parking spaces.
AWS was built from the start to be a sold product, whereas it sounds like this massive charging network is purely internal. Seems like a very big difference between wiring up your warehouse lots for charging, versus building power infrastructure across the country.
Can't find any specific history of EC2, but the concept of AWS as being a commercial offering seems like the idea from the outset: https://aws.amazon.com/blogs/aws/welcome/
Even if they dog-fooded EC2 internally first, all the public documentation I can find seems to suggest that they were always considering it as a product they would eventually sell.
Regardless of the specific timeline of AWS and its associated products, I think it's a bad (or at best shallow) comparison if you're trying to forecast Amazon's relationship to EV charging. Most tech companies build massive infra for their internal use and never open it to the public, seems way more likely that their EV charging infra would fall into this case than become anything similar to AWS. Especially when you consider the magnitude of difference between filling your parking lots with chargers, vs. brokering power between utilities and EV drivers. Nobody's arguing that Amazon is going to open a chain of gas stations, despite the fact that a few years ago they probably had a massive fuel distribution infrastructure.
Definitely the idea of selling EC2 was there from the start, relevant quote is:
> Pinkham thought about this issue and, in 2003, started trying to build an "infrastructure service for the world". His hope was that he could develop a service that would not only deal with Amazon's infrastructure, but also help developers.
It wouldn’t be the craziest thing to happen. People were surprised when a book store entered the cloud services business. Amazon has pattern of solving internal problems and turning the solutions into businesses.
Please don’t tell me the author believes “Amazon had to learn how to . . . call the power company.” Sure the folks at fleet services are different from AWS, warehouses, etc. in the big picture all of Amazon is devoted to transforming electricity into valuable services. Maybe the question is whether or not Amazon sees value in taking over the power companies jobs.
To get there, Amazon had to learn how to pick up the phone and call the power company. Electric utilities, which to that point primarily dealt with electric vehicles through powering the odd home car charging setup, encountered a new type of customer in Amazon.
I think the point the author is making here is that calling the electric utility about charging dozens or hundreds of vans is different than checking in to make sure your home setup doesn't violate utility policy.
Amazon is very often in a position to dictate terms of an agreement, but here they needed the utility's assistance. Amazon had to figure out how to negotiate the landscape.
The article even mentioned the difference involved. You buy a warehouse you might need 50 or 100 KWh. Everyone planned for that when the facility was built.
Want to charge 200 vans? Now you’re talking megawatts. You have to talk to the utility about upgrading their lines and stations and such.
“Hi. PG&E? We want to 100x our overnight power usage! That’s good right?”
That’s not a normal call they’d have had to do before.
That is good news all around. But we need to also realize when you have as many as 100,000 trucks, battery failures and battery replacements are nearly 1000 times a year. Those costs add up to the ghg. It might be definitely less than ICE
Cars but if your goal is to offset, then this is a cost to offset too.
Any failed or end of life batteries will be feedstock for the recycling stream. To drive down lifecycle green house gas emissions, there should of course be focus on using clean energy through the process whenever possible. We are building an enormous energy storage manufacturing and deployment flywheel, improvement should be continuous as the Machine is built.
I imagine they break-even in GHG emissions in most places in the US within a year or two of ownership. These trucks will likely be in service for over a decade. Last-mile delivery trucks get absolutely horrible mileage, constantly having to accelerate and brake all day long with a large amount of mass.
Then you also need to factor in the recycling opportunity for the production.
An EV may have fairly high CO2 emissions during production, but much of that is associated with mining of metals/minerals. Those are generally recyclable. Large scale commercial recycling of EV batteries is already happening. So those CO2 emissions are not for just one vehicle. It’s for many vehicles far into the future.
The CO2 emissions from burning fossil fuels will never benefit anything but the one vehicle it’s propelling forward.
You’ve also gotta keep in mind that the mining and industrial transportation sector is also being decarbonised, sometimes with the same EV technology that EV cars are made with. Electric mining dump trucks is now a reality. Keep the long term picture in mind: we actually have a very viable pathway to eliminate whatever minor CO2 emissions are associated with EVs by just pouring more investments and R&D into EV technology.
When making the comparison, don't forget to tally the effects and emmisions of drilling, pumping, refining, and delivering oil and gasoline to the point-of-sale.
Methane leaks are often by design when it's an oil well.
That's why you need to enforce flaring (burning the methane and releasing CO2 into the atmosphere) because that's better than simply venting the methane.
> Gas venting, more specifically known as natural-gas venting or methane venting, is the intentional and controlled release of gases containing alkane hydrocarbons - predominately methane - into Earth's atmosphere. It is a widely used method for disposal of unwanted gases which are produced during the extraction of coal and crude oil. Such gases may lack value when they are not recyclable into the production process, have no export route to consumer markets, or are surplus to near-term demand.
You are not the first person to think this thought. It is standard greenhouse gas inventory methodology. Quoting from Amazon's GHG report:
"""Transportation activities generate carbon emissions through the combustion of fossil fuels when vehicles are driven, the manufacturing of vehicles, and the supply chain for fuels. The emissions model for transportation covers emissions associated with “well-to-wheels” impacts (e.g.,extracting, refining, distributing, and consuming transportation fuels) and “vehicle” impacts (e.g., manufacturing, maintaining, and disposing of vehicles). These impacts are reported as emissions factors based on grams of CO₂e per kilometer basis."""
VW put out a report on the topic a few years back.
Tldr; for consumer vehicles, break even point for emissions was roughly 40k miles in the EU, and 50k in the US. After that it was a ~net positive vs equivalent gasoline model. For commercial use cases where 50k miles can be done many times over in a single year (and expected vehicular lifecycle is much much longer), it should be pure gravy.
Makes so much sense. I wonder how many (tens of?) thousands a single warehouse location might save per day in gas alone. Last mile in urban/suburban areas is nearly ideal for an EV.
The upfront is very expensive with all the ancillary costs, but it’s hot to pay off handsomely after a while. And if a real carbon tax were to ever appear it would be all the better.
For the ICE car, that's ~6.2 mi/dollar, for the EV it's ~31.8 mi/dollar. ~5x cheaper per mile to drive the ICE. I drive ~15k miles per year, so ~$2.4k in gas, or $470 in electricity - saving ~$2k/year.
This doesn't take into account the lower total cost of ownership of EVs in general. They're simple. They don't need oil changes. The transmission can't go out.
And that's just for me driving once or twice a day; I'm sure the Amazon drivers are putting factors more miles on their cars, so a faster break-even point.
It seems like only a matter of time before other similar vehicles that are more industrial make this shift (eg, garbage/recycling trucks). Imagine no more super loud garbage trucks in the mornings…
So they have an incentive to floor the throttle, and slam the brakes. The drivers don't pay for fuel or brake pads out of their salary, the company does, but if they drive like a maniac, they can finish a few hours earlier.
https://group.vattenfall.com/press-and-media/newsroom/2023/s...
Are transmission failures really that common? In nearly fifty years of driving and owning eight different cars I've never had a transmission failure. One of the cars (Volvo Amazon estate car) did 350k miles including the previous owner rallying it for a couple of years. Apart from fixing a leaking radiator when I bought it I never had any problems with the motor or gearbox. We scrapped it just because my wife wanted a newer more comfortable car.
All that said, electric cars really shine in city driving. No transmission to go, and regenerative braking to save the pads. It’s valid to call it out if that’s your use case like these Amazon trucks.
https://en.wikipedia.org/wiki/Ford_PowerShift_transmission
And then sometimes things aren't built as designed. The ford HF35 hybrid transmission should be pretty reliable as designed, but many of them were built with bearings (or something) that went bad, leading to an early salvage date for many C-Maxes.
GM on the other hand went the exact opposite direction most of the time and put an overengineered heavy duty transmission with maybe slightly cheaper internals into their smaller vehicles.
Eh, the owners of 80's era Chevy Celebrities/Pontiac 6000s/Oldsmobile Cieras/Buick Centuries might like a word.
Clutches wear out, and it is possible to wear out an automatic transmissions' clutches, but I've never had it happen.
Every car I've sent to the junkyard has had a working transmission.
It's a different level of wear and tear than a normal commuter passenger vehicle.
I think the bigger savings on a delivery van would be brakes. In constant stop-start driving using the motor for regenerative braking is both an efficiency saver and reduces wear and tear on the mechanical brakes.
That 2002 model is supposed to be a lemon transmission-wise though (even so, 300k seems decent as far as lifespans go).
Transmissions are tough because they endure a lot of extreme forces. Shift points have the risk of slamming components together.
But for Amazon their drivers are constantly doing stop/starts, which is horrible for fuel efficiency.
So I bet their numbers are way better. And I wasn’t including maintenance which as you mentioned should be tremendous.
https://www.globalenergyinstitute.org/average-electricity-re...
The reason PG&E rates are skyrocketing in California - https://news.ycombinator.com/item?id=40128003 - April 2024 (26 comments)
After 111 years, SF finally moving to oust PG&E and create a public-power system - https://news.ycombinator.com/item?id=39587901 - March 2024 (13 comments)
https://www.spglobal.com/commodityinsights/en/market-insight... | https://archive.today/PeDeH ("April 2024: CAISO sets back-to-back solar peak records in April; power spot prices plunge")
> Spot power prices have been trending lower year on year on lower natural gas spot prices and more renewable generation supply, including hydropower following years of drought.
> CAISO has 19.011 GW of installed solar generation capacity, which accounts for the majority of renewable resources on the grid at 62%, according to the Key Statistics report. Solar capacity has increased nearly 15% compared to a year ago and 26% from two years ago.
> There are currently 28.6 GW of solar projects in the CAISO generation queue, including 13.8 GW that have executed interconnection agreements and nearly 4.5 GW of those have on-line dates expected listed for 2024. Only two solar projects slated to come online in 2024 do not have batteries connected.
https://app.electricitymaps.com/zone/US-CAL-CISO ("ElectricityMaps.com: California CAISO Zone")
As long as you’re using less electricity from the grid overall per day, you’ll save money in the long run.
Solar > battery means DC fast charging gets to skip AC>DC losses which are significant. But you don’t need to use 100% of solar’s output to save a lot of money.
But if the expected lifetime is “as long as I can drive it,” then TCO is probably not lower, simply because the maximum lifetime of an EV is lower than that of an ICE vehicle. At some point the car will fail – it will need battery replacement (every X years), or it will depend on obsolete infrastructure, etc.
You can drive a 1960 Cadillac today. Will you be able to drive a 2020 Tesla in 2084?
Interesting to hear you say that. I've been looking for them but I've still never seen an Amazon EV in the DC metro area.
I don’t know if we’re a test site or it was just easy to get the power installed or what. Perhaps being a new facility meant they just needed new vehicles anyway so it made sense to not buy gas.
But once they started appearing they kept getting more common.
1. The CO2 of the Amazon delivery needs to be taken into account (apportioned to the number of drops), vs delivery to store. Delivery to stores is highly efficient compared to last mile.
2. Most Amazon purchases probably wouldn't directly result in less trips to the supermarket (i.e. am I driving to the supermarket just to pick up one item, or am I going there anyway and picking up everything I need when I am there?)
Between that and shopping for fresh meat and produce at ethnic markets that are all over the city, the vast majority of my shopping is amortized over staple groceries that I can’t or won’t buy online.
For use cases like this, the last mile is carbon neutral.
Edit: Re-reading my comment, it comes across as if there were no walkable cities in the US. That was not my intention. Of course there are, but from what I've seen, not that many. Also, somehow there is a lot of political opposition to them, because of certain conspiracy theorists who seem to have taken over one of the political parties. And since there are only two worth mentioning, the situation is unlikely to improve.
Different views of the future.
Jokes aside, I save money by using grocery delivery.
Careful with that.
The delivery fees may not be explicit, but they definitely charge more for items at some stores, with Costco being one of the more egregious ones.
https://www.reddit.com/r/Frugal/comments/1cal0ff/comparing_c...
Short answers:
- Yes, they mark up. 24%ish.
- Markup may be different when shopping on IC directly versus Costco same day delivery (via IC).
- Coupons like 50% off orders over $xx changes that math.
- Factor in tip into the equation.
Does Amazon hope that this investment in EV's and EV infrastructure will put them in a good position to sell EV's at some future point?
Maybe it'll give them a few years of de-facto monopoly in some cities/states that pass laws requiring all deliveries be carbon free?
Maybe being top of 'low carbon' company lists alone is worth it if that means they attract environmental conscious shareholders?
Maybe this is cheap positive PR which is cheaper than just spending an equivalent amount on adverts?
For a company so laser focussed on financial efficiency, the move to EV's doesn't seem to make a lot of sense.
Are they more expensive year one? Sure. New vehicles + chargers + infrastructure upgrades.
Year 5? Maybe it’s paid off. Energy costs, brake pads, no oil changes or other frequent maintenance.
When you refresh the EVs the infrastructure is still there, that’s one time. Gas prices go up? Your investment looks even better. Carbon tax appears? You’re golden.
Amazon must know it’s worth it. You don’t spend that kind of money for good PR alone.
EDIT: Hrm, no, a Ford Transit is also way more expensive in the US than in Europe. Wonder what's going on there. Is it just limited competition/more niche part of the market or something?
And long-long-term maybe some kind of integration with Zoox for an tsla/uber like delivery experience (with drones?).
[1] https://youtu.be/3CWCqJl0BEs
You can ignore any environmental impact, the main driver for this is that it's much cheaper to operate an EV van for last-mile deliveries.
[1]: https://www.bbc.co.uk/news/world-us-canada-56628745
And does Amazon own the warehouses outright, or do they lease them?
Why not just double your fleet size instead?
That's a bit extreme, but this is the kind of optimization that your bean counters are good at. You probably don't need to double your fleet but just keep enough that you can shift the consumption curve or take advantage of price changes.
(Not private thank you very much)
I wouldn't be surprised if Tesla has more public chargers than anyone else on this list.
I'm currently living in an apartment complex that has "public" charging stations, and most of them are on reserved parking spaces.
You don’t want it for road trips or in place of a home charger. It’s too slow. You want DC fast charging.
But for something like Amazon who has vans sitting overnight it’s much cheaper and more than adequate.
Even if they dog-fooded EC2 internally first, all the public documentation I can find seems to suggest that they were always considering it as a product they would eventually sell.
Regardless of the specific timeline of AWS and its associated products, I think it's a bad (or at best shallow) comparison if you're trying to forecast Amazon's relationship to EV charging. Most tech companies build massive infra for their internal use and never open it to the public, seems way more likely that their EV charging infra would fall into this case than become anything similar to AWS. Especially when you consider the magnitude of difference between filling your parking lots with chargers, vs. brokering power between utilities and EV drivers. Nobody's arguing that Amazon is going to open a chain of gas stations, despite the fact that a few years ago they probably had a massive fuel distribution infrastructure.
edit: found more detail here: https://www.zdnet.com/article/how-amazon-exposed-its-guts-th...
Definitely the idea of selling EC2 was there from the start, relevant quote is:
> Pinkham thought about this issue and, in 2003, started trying to build an "infrastructure service for the world". His hope was that he could develop a service that would not only deal with Amazon's infrastructure, but also help developers.
To get there, Amazon had to learn how to pick up the phone and call the power company. Electric utilities, which to that point primarily dealt with electric vehicles through powering the odd home car charging setup, encountered a new type of customer in Amazon.
Amazon is very often in a position to dictate terms of an agreement, but here they needed the utility's assistance. Amazon had to figure out how to negotiate the landscape.
Want to charge 200 vans? Now you’re talking megawatts. You have to talk to the utility about upgrading their lines and stations and such.
“Hi. PG&E? We want to 100x our overnight power usage! That’s good right?”
That’s not a normal call they’d have had to do before.
For a company at scale as amazon, they are a major consumer of equipment and the sourcing of that equipment is also a corporate operation.
Draw attention to https://www.epa.gov/greenvehicles/electric-vehicle-myths#Myt... specifically the lifecycle green house gas chart.
https://www.redwoodmaterials.com/recycle-with-us/
(breakeven is 15k miles with Redwood's recycling)
An EV may have fairly high CO2 emissions during production, but much of that is associated with mining of metals/minerals. Those are generally recyclable. Large scale commercial recycling of EV batteries is already happening. So those CO2 emissions are not for just one vehicle. It’s for many vehicles far into the future.
The CO2 emissions from burning fossil fuels will never benefit anything but the one vehicle it’s propelling forward.
You’ve also gotta keep in mind that the mining and industrial transportation sector is also being decarbonised, sometimes with the same EV technology that EV cars are made with. Electric mining dump trucks is now a reality. Keep the long term picture in mind: we actually have a very viable pathway to eliminate whatever minor CO2 emissions are associated with EVs by just pouring more investments and R&D into EV technology.
That's why you need to enforce flaring (burning the methane and releasing CO2 into the atmosphere) because that's better than simply venting the methane.
https://en.wikipedia.org/wiki/Gas_venting
> Gas venting, more specifically known as natural-gas venting or methane venting, is the intentional and controlled release of gases containing alkane hydrocarbons - predominately methane - into Earth's atmosphere. It is a widely used method for disposal of unwanted gases which are produced during the extraction of coal and crude oil. Such gases may lack value when they are not recyclable into the production process, have no export route to consumer markets, or are surplus to near-term demand.
"""Transportation activities generate carbon emissions through the combustion of fossil fuels when vehicles are driven, the manufacturing of vehicles, and the supply chain for fuels. The emissions model for transportation covers emissions associated with “well-to-wheels” impacts (e.g.,extracting, refining, distributing, and consuming transportation fuels) and “vehicle” impacts (e.g., manufacturing, maintaining, and disposing of vehicles). These impacts are reported as emissions factors based on grams of CO₂e per kilometer basis."""
Tldr; for consumer vehicles, break even point for emissions was roughly 40k miles in the EU, and 50k in the US. After that it was a ~net positive vs equivalent gasoline model. For commercial use cases where 50k miles can be done many times over in a single year (and expected vehicular lifecycle is much much longer), it should be pure gravy.