The possibly impossible economics of DCFC

[nwdiver]

I thought it would be worth having a thread to discuss how difficult and perhaps impossible it's going to be for L3 charging to operate as a viable business. One fact about electric billing very few people have experience with is demand billing, being billed per kW in addition to per kWh. In SE NM Xcel charges $15.40/kW in the winter and $18.49/kW in the summer. So if you want to have 250kW fast charging a single car charging at that rate sets your electric bill at ~$4600/mo. Even if rates are set as high as $0.40/kWh and the average car takes 50kWh you would need to charge 230 cars per month or 7 per day just to break even on demand billing and that's just a single stall. Obviously more stalls = higher demand fees and that's not counting the cost of electricity per kWh or the cost of the equipment which is also substantial.

For DCFC to be economically sustainable it needs to collect revenue from all parties that share in it's benefit. The manufacturer, the host, the driver, other local businesses.

There's a great video on YouTube with the CEO of charge point where he points out that gas stations don't make money selling gas... they make their profits from the attached convenience store. Economically selling electrons is even worse than selling gasoline.

 

[MXLRplus]

If the government truly wants to support EV adoption and expansion of the charging grid, they need to limit demand-based tariffs at DCFC sites.

I don't see that happening though. Politicians aren't the sharpest knives in the drawer.

At this stage of the game, EVs aren't really boosting demand enough to justify needing more infrastructure, and that's what demand tariffs are for.

 

[miimura]

Battery storage makes a lot of sense for DCFC sites that have a limited number of charging sessions per day. The batteries can be put between the chargers and the meter or can be integrated into the charging station.

FreeWire Pro Series

Boost Power Pro is a fully-integrated ultrafast EV charging technology that can be deployed anywhere and everywhere.

freewiretech.com

 

[MXLRplus]

Boy don't I feel stupid. California changed their demand requirement for 2019-2024. They did the wise thing and removed it for now. When I researched this and wrote a letter to the CPUC in 2018(?), they had demand tariffs for EV sites. Now they don't:

https://library.sce.com/content/dam/sce-doclib/public/regulatory/tariff/electric/schedules/general-service-&-industrial-rates/ELECTRIC_SCHEDULES_TOU-EV-8.pdf

It's still a bit draconian for >500kW https://library.sce.com/content/dam...ustrial-rates/ELECTRIC_SCHEDULES_TOU-EV-9.pdf

The meter fee is $571.13/m on the TOU-EV-9. So if your 4 car, 150kW station costs $100,000 to build after incentives, and the cost of money is 6%, you need to make a $1,071.13/m in gross profit just to cover the meter fee and cost of capital, or $37/day over operational costs.

With the earlier demand tariffs, it was brutal, IIRC, $2400/m just in demand fees.

 

[nwdiver]

Battery storage makes a lot of sense for DCFC sites that have a limited number of charging sessions per day. The batteries can be put between the chargers and the meter or can be integrated into the charging station.

FreeWire Pro Series

Boost Power Pro is a fully-integrated ultrafast EV charging technology that can be deployed anywhere and everywhere.

freewiretech.com

The thing that sucks about batteries is to some degree it's all or nothing. You get a busy weekend that pulls your batteries down and there's one 30 minute window in a month where demand hits 500kW or whatever... and you're back to square one... it's idiotic. They really need to have TOU instead.

 

[Chisale]

At what electrical power level does the utility typically charge demand pricing? Is it regional or fairly uniform throughout the country. From what I understand it's based on 15 minute intervals, correct? So maybe something in the 10s of KWs consistent through that 15 minute period?

 

[SSonnentag]

Solar combined with battery storage solves the peak demand issue. It just requires proper scaling of the system.

 

[nwdiver]

At what electrical power level does the utility typically charge demand pricing? Is it regional or fairly uniform throughout the country. From what I understand it's based on 15 minute intervals, correct? So maybe something in the 10s of KWs consistent through that 15 minute period?

It varies. For Xcel in NM I think it's a 30 minute average >25kW. I think pretty much anywhere is going to have demand fees >100kW.

 

[gnuarm]

One of the interesting points is that for an application like this where the load needs to be shifted over the course of a day or even week, the type of battery in your car is not the right type of battery for use in the charging facility. There is a lot of work being done on vanadium flow batteries for utility storage because the energy is stored in the ions in solution rather than in the plates. So the kWh capacity is not directly tied to the kW rating. They can actually store electrolyte in tanks holding a full day's need for kWh without having to use huge amounts of the battery itself. These batteries are not practical for cars because of their weight and size, but these issues don't matter for stationary uses.

I think the demand charges will become moot. But I do not think they should be arbitrarily removed. They are present to recoup the fixed costs of operating a utility. Utilities are regulated and while the utility knows how to work the numbers, the bottom line is they don't make excessive profits. The only issue is where the money comes from, not how much money they are making. Reduce the demand charges and they will have to make it up with higher per kWh rates.

My concerns with charging costs is more along the lines of what happens to Supercharging if Tesla encounters financial trouble or simply decides to divest the charging network. I have free supercharging. What happens if the network is opened up to other automakers and they don't like providing me with free charging. Likely the network becomes a third party spin off with multiple owners and free supercharging dies. :-(

 

[miimura]

The thing that sucks about batteries is to some degree it's all or nothing. You get a busy weekend that pulls your batteries down and there's one 30 minute window in a month where demand hits 500kW or whatever... and you're back to square one... it's idiotic. They really need to have TOU instead.

The thing about that Freewire charger is that it can keep you off a demand plan completely. It only requires 208V 3-phase or 240V split phase. The input power is <=27kW. If you want to complain about something, it would be the recovery time to charge the battery up enough to charge a vehicle at full speed for the full duration that they want to charge.

 

[nwdiver]

If you want to complain about something, it would be the recovery time to charge the battery up enough to charge a vehicle at full speed for the full duration that they want to charge.

It's not the recovery time... it's the size of the battery. If you need to keep your peak demand <30kW to avoid high demand fees you need to keep your demand <30kW for the entire month. Not 90% of the month. Not 95% of the month. But 99.99% of the month. If demand is 500kW for 30 minutes your demand fee is 500kW even if the other 99.99% of the time it was <30kW. So let's say you invest in a 1MWh battery. That keeps demand <30kW for the first 25 days but then you get a stream of EVs because there's a concert that Saturday... after 5 hours the battery is drained and your demand goes to 500kW because the battery is gone. You get ZERO credit for the fact the 25 days before the 40 EVs drove through your demand was <30kW. That 1MWh battery accomplished nothing.

 

[miimura]

It's not the recovery time... it's the size of the battery. If you need to keep your peak demand <30kW to avoid high demand fees you need to keep your demand <30kW for the entire month. Not 90% of the month. Not 95% of the month. But 99.99% of the month. If demand is 500kW for 30 minutes your demand fee is 500kW even if the other 99.99% of the time it was <30kW. So let's say you invest in a 1MWh battery. That keeps demand <30kW for the first 25 days but then you get a stream of EVs because there's a concert that Saturday... after 5 hours the battery is drained and your demand goes to 500kW because the battery is gone. You get ZERO credit for the fact the 25 days before the 40 EVs drove through your demand was <30kW. That 1MWh battery accomplished nothing.

Certainly, the scenario you describe is a problem. However, you can just force the system to remain within the assigned demand envelope and just slow down the charging so that it is never exceeded. Of course, that could suck for your customers that get their charging throttled way down when your battery is depleted. For the Freewire charger I linked to, the demand is not the problem because its grid demand is fixed below 30kW per charger.

 

[nwdiver]

Of course, that could suck for your customers that get their charging throttled way down when your battery is depleted. For the Freewire charger I linked to, the demand is not the problem because its grid demand is fixed below 30kW per charger.

That's why demand fees need to either take into account the time of day or simply use TOU instead. 200kW demand on the grid at noon is generally going to be a good thing and pulling from an onsite battery when there's surplus energy on the grid would be a bad thing. During the 'neck' of the duck curve is a different story. That's why demand fees are generally a bad idea... they don't take into account WHEN that demand occurs. Pulling 30kW from the grid at 7pm should cost MORE than pulling 300kW from the grid at 11am.

If I'm fast charging in Oklahoma at 11pm it would be kinda silly for the charger to throttle to 30kW because the battery is drained while the wind turbine 500m away is getting cutback because there's too much wind....

 

[andrewFW]

If you have a megapack or similar utility scale battery storage, you should (as in, "it should be so", not "in all likelihood") be able to get reduced or eliminated demand charges by working out a demand sharing plan with the power company. That battery could just as easily soften grid demand changes as it could charge cars.

That said, I think the way this works out is by just building c-stores with the chargers. Large c-stores are all the rage in the gas station world right now. People love wawa, sheetz, buck-ees, etc. You could design a very nice, very intentional experience designed around the fact that your customers are there for 20-30 minutes instead of 2-5. The biggest challenge gas stations face is convincing the driver to walk inside. A nice coffee shop, clean bathroom, etc. could do a lot of that convincing. Sell the candy, soda, and chips. That'll always be where the margins are and what makes the numbers make sense.

 

[nwdiver]

If you have a megapack or similar utility scale battery storage, you should (as in, "it should be so", not "in all likelihood") be able to get reduced or eliminated demand charges by working out a demand sharing plan with the power company. That battery could just as easily soften grid demand changes as it could charge cars.

That said, I think the way this works out is by just building c-stores with the chargers. Large c-stores are all the rage in the gas station world right now. People love wawa, sheetz, buck-ees, etc. You could design a very nice, very intentional experience designed around the fact that your customers are there for 20-30 minutes instead of 2-5. The biggest challenge gas stations face is convincing the driver to walk inside. A nice coffee shop, clean bathroom, etc. could do a lot of that convincing. Sell the candy, soda, and chips. That'll always be where the margins are and what makes the numbers make sense.

I agree that DCFC are going to need to be sited near locations where people would commonly spend ~30 minutes or so. The trouble is the delegation of cost. To my knowledge none of these locations have invested ~$250k+ in the DCFC stations themselves... they've simply allowed Tesla or EA the ability to build their stations nearby. I would love to see what kind of arrangement they've made. I know that Tesla has actually paid some locations rent since one Supercharger in Utah was 'shutdown' when Tesla didn't pay their rent on time... it kinda needs to go the other direction... the host should be paying Tesla or EA for the additional foot traffic they bring in. I don't think that's happening yet.

 

[miimura]

That's why demand fees need to either take into account the time of day or simply use TOU instead. 200kW demand on the grid at noon is generally going to be a good thing and pulling from an onsite battery when there's surplus energy on the grid would be a bad thing. During the 'neck' of the duck curve is a different story. That's why demand fees are generally a bad idea... they don't take into account WHEN that demand occurs. Pulling 30kW from the grid at 7pm should cost MORE than pulling 300kW from the grid at 11am.

If I'm fast charging in Oklahoma at 11pm it would be kinda silly for the charger to throttle to 30kW because the battery is drained while the wind turbine 500m away is getting cutback because there's too much wind....

California utilities have separate demand charges for Total Demand, Peak Demand, and Off-Peak Demand. They are all additive. However, they also have commercial "Business Electric Vehicle" charging plan.

For comparison: B-20 1,000kW+ and BEV

So, a Supercharger site that routinely draws 1,500kW would have these charges:
B-20
Fixed Monthly: $50.05*30 days = $1,501.50/mo
Summer Demand: $26.38/kW * 1500kW + $5.57/kW * 1500kW + $22.81/kW * 1500kW = $39,570 + $8,355 + $34,215 = $82,140
Winter Demand: $1.80/kW * 1500kW + $22.81/kW * 1500kW = $2,700 + $34,215 = $36,915

BEV-2 (over 100kW)
Subscription Charge: $95.56/block * 30 blocks = $2,866.80

So, you can see that the fixed costs of the BEV rate are WAY lower than the otherwise applicable commercial rate. However, the per kWh prices for BEV are much higher - roughly double.

 

[Rocky_H]

That keeps demand <30kW for the first 25 days but then you get a stream of EVs because there's a concert that Saturday... after 5 hours the battery is drained and your demand goes to 500kW because the battery is gone.

You seem to not understand how that charger works. It won't spike up to 500kW, because it CAN'T. It is limited by the input power connection, which in this case is the 200-ish volt AC supply, so if the battery is depleted, all it can do is act as a pass-through, doing AC to DC conversion of about 20-30ish kW. But the point of these kinds of stations is MUCH cheaper installation costs because of the lower power connection requirements and not needing really large utility transformers, and then having the capability to most of the time provide faster charging than the input connection if the battery has buffered some.

 

[nwdiver]

You seem to not understand how that charger works. It won't spike up to 500kW, because it CAN'T. It is limited by the input power connection, which in this case is the 200-ish volt AC supply, so if the battery is depleted, all it can do is act as a pass-through, doing AC to DC conversion of about 20-30ish kW. But the point of these kinds of stations is MUCH cheaper installation costs because of the lower power connection requirements and not needing really large utility transformers, and then having the capability to most of the time provide faster charging than the input connection if the battery has buffered some.

Wasn't referring to that charger specifically or individually but a station of chargers. I don't think that not delivering power to vehicles that need to charge is a good solution to DCFC...

 

[Rocky_H]

Wasn't referring to that charger specifically or individually but a station of chargers. I don't think that not delivering power to vehicles that need to charge is a good solution to DCFC...

What is your deal? It IS delivering power to vehicles. It doesn't just sit there showing "Out of Order" until the battery refills, as you suggest. It has a minimum power level of about 20-30kW, if there isn't extra battery capacity, or up to 100kW+ if there is.

One needs to choose the right tool for the job, though. If one is going to talk about the paradigm of a "station of chargers", like with a dozen stalls, that's a high traffic site that isn't suited to this infrequent use, buffering type of device. Low use sites, with only one or two stalls, on lesser used routes, is where this is the right tool for the job.

 

[andrewFW]

I agree that DCFC are going to need to be sited near locations where people would commonly spend ~30 minutes or so. The trouble is the delegation of cost. To my knowledge none of these locations have invested ~$250k+ in the DCFC stations themselves... they've simply allowed Tesla or EA the ability to build their stations nearby. I would love to see what kind of arrangement they've made. I know that Tesla has actually paid some locations rent since one Supercharger in Utah was 'shutdown' when Tesla didn't pay their rent on time... it kinda needs to go the other direction... the host should be paying Tesla or EA for the additional foot traffic they bring in. I don't think that's happening yet.

I'm advocating for a chain of convenience stores to be built with the specific intention of hosting chargers. Whether these chargers are owned by the convenience store chain or by an established charging network seems inconsequential, could even be a mixture of both (privately owned CCS chargers, but also tesla superchargers on site as well). I believe there are enough gaps/not enough redundancy in the existing charging networks to where there is space for such an entity.