Texas subsidizes chargers at gas stations, not Tesla CCS, but we learn Tesla chargers cost 1/4 those of others

[bradtem]

Here's a story I released today with some tidbits learned from the grants given by Texas to pay 70% of the cost of charging stations.

Tesla’s Texas Charger Grant Applications Fail; It’s Bad For Texas But Reveals Tesla’s Super-Low Costs

A Texas grant program gave $21M to install charging, but all the money went to put chargers at gas stations that can charge only 30% of EVS on the road. Tesla offered to put in chargers that can charge all cars at a price that was less than 1/4 of what other stations asked for. Lots to unpack.

www.forbes.com

Notably:

• Texas took $21M of the Dieselgate money to allow companies to get up to $150K per charger to pay 70% of the cost
• Tesla applied to build 4 stations with 54 chargers which would include CCS or CHAdeMO connectors, a first in the USA
• They allocated the money to build 170 more standard DC Fast chargers at 41 locations, almost all of which were gas stations (lovely places to spend 40 minutes.)
• Tesla's applications were not granted, Texxas says it was first come first served and Tesla didn't apply on time
• Very interesting -- Tesla's grant applications asked for just $30K per charger (at a station of 17) while almost all other applications needed more than $100K per charger (at smaller stations) and most asked for the full $150K implying a cost per charger of $215K or more -- Tesla seems to put in chargers at 1/4 the typical price of others.
• Tesla probably would have just charged CCS cars a bit more than Teslas, though if they wanted to they could have used tricks to scare CCS cars away from their chargers a some Tesla drivers are worried that Tesla chargers will get busier when they accept non-Tesla cars
• and more....

 

[e of pi]

The article doesn't mention this, but there's another possible explanation for Tesla's $30k/stall bid. It sounds like companies could ask for up to 70% of the cost of a stall, not that they had to ask for 70% of the cost. Thus, Tesla's $30k/stall might be asking for much less than the 70% allowable cost per station, aiming to capture more of a monopoly on charging and then being able to make it up through things like charging CCS cars more to charge at their stations if/when Tesla stations with CCS become available. Is it worth taking a loss of possible subsidy of $10, $20, $40, $80k, $150k per stall for Tesla to ensure another network doesn't get funding to build that stall instead? That may have been Tesla's calculation, especially since as I understand it some companies like Electrify America and EVGo are using the exact same installation contractors Tesla uses for Supercharger installs. Thus, while there may be differences in hardware costs and how site costs and project management controls overhead, I'd shy away from assuming total station cost is really 5x lower for Tesla.

 

[Mockingbird]

One thing you have to understand is that around half of the cost of getting DCFCs up and running is something else than the cost of the equipment.

That includes permits, utility hookup, service plan, networking plan, etc.

Mathematically, it would be impossible to do the job at 1/5 of the cost.

 

[bradtem]

The article doesn't mention this, but there's another possible explanation for Tesla's $30k/stall bid. It sounds like companies could ask for up to 70% of the cost of a stall, not that they had to ask for 70% of the cost. Thus, Tesla's $30k/stall might be asking for much less than the 70% allowable cost per station, aiming to capture more of a monopoly on charging and then being able to make it up through things like charging CCS cars more to charge at their stations if/when Tesla stations with CCS become available. Is it worth taking a loss of possible subsidy of $10, $20, $40, $80k, $150k per stall for Tesla to ensure another network doesn't get funding to build that stall instead? That may have been Tesla's calculation, especially since as I understand it some companies like Electrify America and EVGo are using the exact same installation contractors Tesla uses for Supercharger installs. Thus, while there may be differences in hardware costs and how site costs and project management controls overhead, I'd shy away from assuming total station cost is really 5x lower for Tesla.

This seems like a strange approach. First of all, it did not work -- Tesla didn't get any grants. Grants were not given to the lowest price applicant (though they should have.) Had Texas said they would pick the lowest price applicants it would have been wise and you might have gotten people bidding lower.

Oddly, though, that makes it more likely people ask for grants on stations they already wanted to do anyway. The giant $150K/charger grants were most likely to trigger new construction. Some might have just taken grants to expand stations they were already building.

The grants required the stations to have CCS or CdM, no "if/when" about it. This would be a new unit for Tesla -- it makes CCS2 stations in Europe but makes no CCS1 stations.

Right now, Tesla just builds chargers for Teslas only, in order to sell more Teslas. They have no desire for a monopoly that I can see.

 

[bradtem]

One thing you have to understand is that around half of the cost of getting DCFCs up and running is something else than the cost of the equipment.

That includes permits, utility hookup, service plan, networking plan, etc.

Mathematically, it would be impossible to do the job at 1/5 of the cost.

The rules allowed you to spend 20% of the grant max on upgrading electrical service. You have to demonstrate you spent the money.

Even so, I can't see a motive for Tesla to want to take less grant money than the grants allowed. They asked for $500K of subsidy on a 17 stall charger, while most others asked for $600K on a 4-stall charger. You could not legally ask for more than 70% of your costs but I don't see a reason to ask for less.

In addition, Tesla seems to have a big economy of scale here. They requested $500K to do a 17-stall charger, and $383K to do a 9-stall charger. Nobody else put in for more than 6 stalls (asking the $600K max.)

 

[e of pi]

This seems like a strange approach. First of all, it did not work -- Tesla didn't get any grants. Grants were not given to the lowest price applicant (though they should have.) Had Texas said they would pick the lowest price applicants it would have been wise and you might have gotten people bidding lower.

Oddly, though, that makes it more likely people ask for grants on stations they already wanted to do anyway. The giant $150K/charger grants were most likely to trigger new construction. Some might have just taken grants to expand stations they were already building.

The grants required the stations to have CCS or CdM, no "if/when" about it. This would be a new unit for Tesla -- it makes CCS2 stations in Europe but makes no CCS1 stations.

Ah, I misread and didn't realize it was your article! It was well written and I appreciate the research in bringing it to light. Out of curiosity: you mentioned EVGo and Chargepoint bid but were also rejected. Did Electrify America get any?

Anyway, it's possible that Tesla's strategy was what I describe and it simply didn't work. Many things have been proposed that sound good in board rooms and yet turn out not to work in practice or in bid evaluation. I've been part of engineering teams which have rejected selecting purely the lowest bid because of some concern about if the bid was legitimate or met the actual intent. Part of the reason you have source selection teams is because it's not merely a matter of selecting the envelope with the paper with the smallest dollar number written on it. It's also possible (though you may have better access to the primary sources to confirm or deny) that Tesla's $30k/stall was for expanding stations that already exist or including CCS stalls in stations they already plan to build, thus meaning that those additional stalls don't necessarily need to bear the cost of electrical service upgrades, negotiating site access, and so on, such that while Tesla might not be taking a massive loss by bidding only $30k to add each CCS stall, this also doesn't represent what their true cost of a pure-CCS greenfield site would be compared to others. It's also possible that I'm wrong, that wasn't Tesla's intent, and Tesla actually can install a station for a fifth what anyone else can. I lack the data you have researched, and can't make any definitive statements. You may be able to tell me I'm wrong.

Right now, Tesla just builds chargers for Teslas only, in order to sell more Teslas. They have no desire for a monopoly that I can see.

Right now, but "having the best charging network" is part of how the Superchargers sell Teslas. If in 5 years most cars have networks available to them that are nearly as large as the options available to Tesla, that advantage in charging availability is eroded. Having 10,000 more stalls than the competition matters when there's only 20,000 stalls total and you have 75% of them. Having 10,000 more stalls than the competition when there's 100,000 stalls total, and it's a matter of being able to access 90% of a much larger number or 100% depending on your car selection matters less. Since no one but Tesla is installing stalls with >50 kW Tesla proprietary connections, it could be said to be valuable to see that as many of those exist as possible and thus that Tesla builds as many of the total number of new chargers being installed as it can.

And if Tesla was in a position where they had CCS open to other car companies but was charging more to use their stations with CCS plugs, that'd mean doing so was a revenue stream, so aside from the marketing of "best available charging network" there'd also be a financial incentive to make sure as many CCS cars used Tesla chargers instead of EA or Shell or Brand X chargers.

 

[bradtem]

This is the Dieselgate money. Some of it went into funding EA. Other portions were allocated directly to states to do things to make the world greener. This is a tiny portion of what Tesla got. It presumably could not go to EA.

Texas did not publish the addresses for failed grants so we won't know if it was expansion or new stations. Texas required stations be near major highways, but that's normally the case for Tesla stations.

Tesla *says* it just wants more EVs in the world, but for now they have not on their own accord made chargers that support non-Tesla.

In Europe, Tesla charges more for non-Tesla cars. Tesla has said it tries to keep the cost low (even break even) for Tesla owners, because the network is there to make a Tesla the best car to buy. They could keep doing that by using the various tricks I describe -- higher prices for non-Tesla, queue priority for Teslas etc.

But I still don't see a motive for applying for a grant of $383K for a 9-stall station when the rules would have allowed you to ask for $600K.

 

[bradtem]

The article doesn't mention this, but there's another possible explanation for Tesla's $30k/stall bid. It sounds like companies could ask for up to 70% of the cost of a stall, not that they had to ask for 70% of the cost.

Has anybody seen the budget of a supercharger station? Remember, $30K/stall is 70%, which is $42,000 per stall. That actually doesn't seem like too ridiculous a price. Tesla makes more fast charging stations than anybody else (except perhaps in China) and their stations have no screens, card readers or other UI -- just the cable and of course the charging electronics. They have made them in quantity 30,000 -- I could easily see them costing well under $42,000 to make, leaving the rest for construction and electrical service (which is only allowed to be 20%.)

What is more surprising is that the other stations like Chargepoints and EVgos cost $215K or more all-in. Just where is that going?

 

[Mockingbird]

The rules allowed you to spend 20% of the grant max on upgrading electrical service. You have to demonstrate you spent the money.

Even so, I can't see a motive for Tesla to want to take less grant money than the grants allowed. They asked for $500K of subsidy on a 17 stall charger, while most others asked for $600K on a 4-stall charger. You could not legally ask for more than 70% of your costs but I don't see a reason to ask for less.

In addition, Tesla seems to have a big economy of scale here. They requested $500K to do a 17-stall charger, and $383K to do a 9-stall charger. Nobody else put in for more than 6 stalls (asking the $600K max.)

Eligible project costs are those that are directly connected to the acquisition, installation, operation, and maintenance of new light-duty ZEV supply equipment.

Eligible project costs may include up to $20,000 per charging unit for the upgrade of off-site power converters (e.g., three-phase power converters) to accommodate the installation of DCFC light-duty ZEV supply equipment.

Eligible project cost categories include:

o Equipment includes tangible personal property having a unit acquisition cost of $5,000 or more (including sales tax and delivery) with an estimated useful life of over one year. Equipment purchased with grant funds should be budgeted as Equipment if the sum of the separate component parts (including sales tax and delivery) has an original value of $5,000 or more. Equipment costs that do not involve an arms-length transaction (e.g., use of inventory without a proof of purchase) are not reimbursable.

o Supplies and Materials includes non-construction related costs for goods and materials having a unit acquisition cost (including sales tax and delivery) of less than $5,000 per unit.

o Construction includes the costs for the enhancement or building of permanent facilities. Construction costs may include:
 planning, designing, and engineering;
 materials and labor;
 subcontracts for services in connection with the construction; and
 facility improvements, such as paving, foundations, and covers.

o Contractual includes non-construction related costs for subcontracted or hired-out professional services or tasks provided by a firm or individual who is not employed by the applicant. Each subcontractor/consultant should be listed separately.

o Other includes costs that do not fall under the equipment, materials and supplies, construction, or contractual categories.

 

[Mockingbird]

Let me put it this way:

Even if Tesla is providing the equipment for free, doing the project at 1/5 of the cost is impossible because other associated costs (other than the cost of the equipment itself) is already more than that.

 

[RTPEV]

Has anybody seen the budget of a supercharger station? Remember, $30K/stall is 70%, which is $42,000 per stall. That actually doesn't seem like too ridiculous a price. Tesla makes more fast charging stations than anybody else (except perhaps in China) and their stations have no screens, card readers or other UI -- just the cable and of course the charging electronics. They have made them in quantity 30,000 -- I could easily see them costing well under $42,000 to make, leaving the rest for construction and electrical service (which is only allowed to be 20%.)

What is more surprising is that the other stations like Chargepoints and EVgos cost $215K or more all-in. Just where is that going?

Well I think you're somewhat right, but if you consider that a 4-stall V3 Supercharger is really a single 1MW (kind of) charger with 4 simultaneous connectors, versus the ChargePoint and EVgo units which tend to be single 125kW, 150kW or 350kW chargers with 1 or 2 simultaneous connectors (many have 2 connectors -- CCS & CHAdeMO -- that may or may not be able to be simultaneously used). So a 4-stall ChargePoint site may in fact be more comparable to a 16-stall Tesla site (okay, maybe 12-stall to account for the higher power of the Supercahrger). Combined with the other factors you mention (and an additional one or two):

• Not having to buy equipment from a 3rd party (this alone is probably a huge factor)
• Volume manufacturing / economies of scale: while the number of Tesla vs. non-Tesla stations may be similar, the non-Tesla stations are made by a dozen or more suppliers, so Tesla really does have what is probably a 10X volume advantage over their competitors
• Manufacturing maturity advantage: Tesla has been at this longer than the other suppliers and no doubt has more streamlined designs and manufacturing capabilities
• Modular design reduces install costs
• Familiarity with permitting processes/requirements: Tesla's first mover advantage here is that they can more readily point to previous installations in the area to help communities that may never have seen a charging station before develop policies that lead to necessary permits. Tesla probably believes their spend on bureaucratic needs is going to be relatively low
• Optimized back-end with no need for card readers, screens
• Only need one connector per stall, with a relatively short cord (except for the token one that will support CCS) and simple cable management system

Some of those are bigger cost savers than others, but they all add up.

 

[bradtem]

Let me put it this way:

Even if Tesla is providing the equipment for free, doing the project at 1/5 of the cost is impossible because other associated costs (other than the cost of the equipment itself) is already more than that.

I agree that it's a huge margin. Now, to be fair, the 1/5th number comes from Tesla's 17-plex stations taking a grant of $500K and other people's 4-plex stations taking the max grant of $600K. But there is lots of variety. A Circle K asked for $297K for their 4-plex, and Tesla asked for $383K for their 9-plex. In that case, Tesla is only coming in at half the price of the Circle K. A small number of other applicants were in the range of $80-$90K per stall, such as GCRE solar, El Paso Electric and Texas Petroleum group. Raswan Inc submitted a 2-plex station for only $131K or $65K per station.

If you consider the Rashwan it's only 50% more than the Tesla 9-plex per stall. That's entirely credible unless you think they were also deliberately not asking for the full subsidy.

But if a Circle K can put in somebody's station for $74K per stall, then I think Tesla, which has built 30,000 stalls can pull it off at a 17 plex for $30K per stall.

The real question is, how did the others all bump up their costs to $150K per stall ($215K)

 

[RTPEV]

I also think you are correct that the grants are probably having the effect of keeping costs artificially high. Many of the traditional gas stations that won awards look like they probably just went out and requested the maximal amount of money they could for the smallest possible sites and likely did not shop around for competitive bids on charging stations.

Several of the Texas Petroleum Group projects showed up on AFDC today (added to fastcharger.info) as did one of the TA projects. They also have entries on Plugshare.

 

[bradtem]

Well, they do have to show actual costs. Since they have to pay 30% of the cost, they still have an incentive to get a cost effective unit, but not nearly as much of one. And given the notorious unreliability of many DC Fast stations, paying more has its attractions if it gets better quality. Eventually we'll see what brands of DC Fast they put in these units.

I will note that EVgo and Chargepoint also asked for $150K of subsidy on their applications, and they should have good economies of scale, though not as good as Tesla.

 

[bradtem]

This link gives costs for chargers that got rebates in California. The rebates are also fat and no doubt bring the price up.

CALeVIP DC Fast Chargers, Average Rebate, Unit Cost, and Total Project Cost per Charger

www.energy.ca.gov

Strangely, we see the total cost per charger lowest on single chargers, and slightly higher on 3 plexes than on 2 and 4-plexes which does not make a lot of sense, though larger installations will have higher costs to bring in new electricity, but it should be lower per stall. Possibly the solo chargers don't need to get new electrical service at many of the installations? My understanding is typically you get 480v service which most commercial locations don't have though industrial ones do.

Anyway, $25K for the box, and $75K in "other" which will be service, wiring, trenching/construction and the master-box. So if Tesla has that down to $42K in a 17-plex that doesn't surprise me, they have more skill and do it in higher volume.

 

[Mockingbird]

Look at what cost is eligible for the grant:

Eligible project costs are those that are directly connected to the acquisition, installation, operation, and maintenance of new light-duty ZEV supply equipment.

Eligible project costs may include up to $20,000 per charging unit for the upgrade of off-site power converters (e.g., three-phase power converters) to accommodate the installation of DCFC light-duty ZEV supply equipment.

Eligible project cost categories include:

o Equipment includes tangible personal property having a unit acquisition cost of $5,000 or more (including sales tax and delivery) with an estimated useful life of over one year. Equipment purchased with grant funds should be budgeted as Equipment if the sum of the separate component parts (including sales tax and delivery) has an original value of $5,000 or more. Equipment costs that do not involve an arms-length transaction (e.g., use of inventory without a proof of purchase) are not reimbursable.

o Supplies and Materials includes non-construction related costs for goods and materials having a unit acquisition cost (including sales tax and delivery) of less than $5,000 per unit.

o Construction includes the costs for the enhancement or building of permanent facilities. Construction costs may include:
 planning, designing, and engineering;
 materials and labor;
 subcontracts for services in connection with the construction; and
 facility improvements, such as paving, foundations, and covers.

o Contractual includes non-construction related costs for subcontracted or hired-out professional services or tasks provided by a firm or individual who is not employed by the applicant. Each subcontractor/consultant should be listed separately.

o Other includes costs that do not fall under the equipment, materials and supplies, construction, or contractual categories.

What is likely happening is that Tesla is not claiming every eligible costs.

 

[bradtem]

Look at what cost is eligible for the grant:



What is likely happening is that Tesla is not claiming every eligible costs.

So you say, but I can't think of a reason for them to do that. Quoting less had no bearing on winning the grant. Why would they not take the money? They are not altruists.

 

[Mockingbird]

So you say, but I can't think of a reason for them to do that. Quoting less had no bearing on winning the grant. Why would they not take the money? They are not altruists.

It wasn't known until after all the applications were submitted the state was giving grants on a first come first serve basis, rather than to the most qualified applications.

 

[bradtem]

It wasn't known until after all the applications were submitted the state was giving grants on a first come first serve basis, rather than to the most qualified applications.

I don't actually know what they said at the time, but I suspect they did not say that price would be a factor in who was granted either. Certainly nobody else felt they should submit a smaller request in the hope of getting it. FCFS was indeed a stupid way to allocate it -- I would have done it on price, station utility, station quality and station location. (They used station location, it wasn't pure FCFS.)

It is perhaps a debate about the choice between a Tesla+CCS+CdM station, which can charge all cars which can fast charge, and regular CCS/CdM which can charge only 30% of cars. From the standpoint of serving the public, you would want the stations that can charge anybody. From some viewpoints though, they might say that there are already lots of places for Teslas to charge and the combined stations would get lots of Teslas at them, so if I wanted to represent non-Tesla interest I would want to beef up the other network. That might make sense when Tesla was from California, but does not when it's HQed in Texas.

On station quality it's pretty clear Tesla is the leader, in part because they maintain better and also because they make larger stations where broken stalls are not so big an issue.

But they didn't use these criteria, only location (near highway) and FCFS. They didn't look at the location criteria I would advise looking at, namely:

1. Does this location serve an area that is not served or underserved by chargers
2. Does this location have reasonable activities for a 40 minute pause (good selection of restaurants, large general/grocery stores, tourist attractions, high speed internet.)

 

[bradtem]

FYI, I added a section to the article about how a state should prioritize these grants in the future. Frankly any charging network, including Tesla, should follow these principles.

If you are a state with grant money to use to improve charging in your state, you should allocate it according to the following factors:

1. Cost: Get more stations for your money. If a supplier will do it for less, weigh them favorably.
2. Station utility: Get stations that support the most cars. If it can’t charge Teslas it will only charge 30% of the cars sold in the USA. Tesla+CCS are good, but you can also support CCS+CHAdeMO chargers which stock the adapters that let most Teslas use the station. They add very little to the cost of a station.
3. Station reliability: Get promises of station uptime. Many subsidized stations end up broken with nobody in a rush to repair, as the subsidy paid for them but not for operating them. Larger stations with many stalls are more reliable because having one station broken is not a big deal.
4. Ease of use: Insist on Plug-to-charge or a very simple way to pay with a credit card. This is to be preferred over specialized apps.
5. Location: Look for locations and routes that don’t yet have chargers. Fast chargers should be rural (for road trippers) not urban, generally. Though urban locations off major highways can also work.
6. Amenities: All stations must have several restaurants, but ideally also other activities for 25 to 45 minutes. They should have bathrooms and high-speed free wifi. Also picnic tables or eating spaces for people to bring their own food or take-out. Also provide a list of restaurants that deliver to the charging station for eating at those tables. Also consider charging at the most popular tourist stops, such as state or national parks and major attractions. Places were a large fraction of road trippers going by want to stop already. Other useful amenities include large grocery stores and general stores, not just overpriced convenience stores as found at gas stations. Eating tables should e pleasant, not too loud. They might need to be covered or indoors.

Also consider not doing fast charging. Most EV charging should be slow. Consider using the funds or laws to encourage regular Level 2 charging at hotels or clusters of hotels. These may provide far more bang for the buck, though they do only benefit private entities. Hotel chargers should have reservation systems as drivers want to be sure they will get a charge overnight.