Tesla Energy and utility scale projects

[jhm]

Cool! I was just about to post on Adam Browning's tweet. These deals are incredible.

Notice the pricing scheme in many of these. There is a per MWh price solar energy plus a per MW-month price on battery capacity. So a typical set up is like 100 MW PV at $27/MWh plus 25 MW / 100 MWh battery at $6200/MW-month. The capacity payment comes close to covering interest plus maybe a little principal on battery capex, say about 7.5% on $1M for a 1MW battery. So the developer/operator seems to be able to take whatever battery revenue from selling services and power at pure profit. Basically any power price above $27/MWh they can get with sores solar is pure profit, but naturally they will discharge when prices are highest.

This is simply going to crush peak gas power. They will be completing with fully charged batteries that are willing to take the peak power price down to about $30/MWh. The cost of fuel alone does not even allow a gas peaker to go that low. This is going to be brutal on gas and coal.

 

[jhm]

With this sort of deal flow for solar+storage, demand for Powerpacks will be explosive.

The pricing issues here also lend some insight into how Tesla may have arrived at 7c/kWh price for charging Tesla Semi trucks. Solar peakers will be pecking off the high prices in wholesale markets everywhere. Creating demand at $70/MWh for stored solar actually helps to support the value of these solar+storage systems. Basically, Tesla can build out a solar+storage system to support a Megacharger for the trucks. This could be at a large enough scale that it also participates in the wholesale power market. Whatever is not sold to truckers at $70/MWh will be sold into the grid for prices mostly between $20 and $60 per MWh. So $70 is a really good price for both Tesla as power producer and trucker who pays about 11c/mile (7c/kWh) for fuel rather than 50c/mile ($3/gal) or more for diesel. The key here is that the Megachargers and supporting power resources need to be at a scale that Tesla is an effective independent power producer. This puts wholesale power prices within reach of charging.

 

[EinSV]

With this sort of deal flow for solar+storage, demand for Powerpacks will be explosive.

The pricing issues here also lend some insight into how Tesla may have arrived at 7c/kWh price for charging Tesla Semi trucks. Solar peakers will be pecking off the high prices in wholesale markets everywhere. Creating demand at $70/MWh for stored solar actually helps to support the value of these solar+storage systems. Basically, Tesla can build out a solar+storage system to support a Megacharger for the trucks. This could be at a large enough scale that it also participates in the wholesale power market. Whatever is not sold to truckers at $70/MWh will be sold into the grid for prices mostly between $20 and $60 per MWh. So $70 is a really good price for both Tesla as power producer and trucker who pays about 11c/mile (7c/kWh) for fuel rather than 50c/mile ($3/gal) or more for diesel. The key here is that the Megachargers and supporting power resources need to be at a scale that Tesla is an effective independent power producer. This puts wholesale power prices within reach of charging.

@jhm, really appreciate this analysis. One thing these recent solar plus storage packages drive home is that the economics of storage are much better when you factor in all of the services storage can provide rather than just the most obvious one -- load shifting. We already had a very striking example of this with the Australia big battery but these new packages with a very low added cost from storage suggest that these other functions of storage can provide significant and very tangible economic benefits that should drastically reduce the cost of storage and drive demand.

This has significant implications for Powerwall. Powerwalls will be much cheaper where they can be integrated as a virtual power plant, as they can provide these other services as well as taking additional pressure off the grid in many places because they are distributed. No wonder Tesla is pushing the virtual power plant concept so much (initially in Puerto Rico and South Australia, but also smaller projects), as this should drastically affect the economics.

Tesla is already installing lots of Powerwalls in California, which has had a very favorable incentive program for solar owners to install storage. I wonder whether at some point Tesla will find a path to networking Powerwalls in parts of California into virtual power plants -- it seems like an obvious thing to do and I would assume the PUC would approve it. I'm sure Tesla could create an attractive program for existing Powerwall owners to sign up and share the benefits.

More generally, I would think that as they install Powerwalls in an area at some point they could reach a high enough density that it would make sense to operate them as a virtual power plant and monetize some of these other services. Assuming regulatory issues could be overcome, I can't see any reason why they couldn't just flip a switch to create a nice high margin revenue stream with little added cost. They could create an opt-in program on the Tesla app that would allow customers to share in the benefit and make it easy to sign up.

 

[TheTalkingMule]

So what's the proportion of non-storage value for grid level(or aggregated distributed) battery packs now? These insane load balancing savings figures out of Australia and shockingly low solar+storage PPA bids lead me to believe utilities will get 50% or more of their value out of "non-storage" services.

Feels like this is especially true in the early stages when the amount of storage is small relative to total peak load. That is an unanticipated game changer, no? When the first 50MWh of storage provide twice as much value as the next 50MWh, you've got an infinitely more compelling angle to get your foot in the door.

There are so many self-reinforcing advantages to this new distributed dynamic, I don't think therthe a way to avoid underestimating the speed of this transition. Even for the fanboys here. Gas is going to be done much soon that we thought a year ago.

 

[SW2Fiddler]

So what's the proportion of non-storage value for grid level (or aggregated distributed) battery packs now?

Please enjoy The Brattle Group’s report on stacking benefits (calculated for the “deregulated” Texas grid) at surprisingly achievable price points.

In this context of declining battery costs, Oncor Electric Delivery Company, a Transmission and Distribution Service Provider (TDSP) in Texas, has engaged us to explore the economics of grid- integrated storage deployment in Texas. We evaluate this question first by estimating whether storage could be cost-effectively deployed on the distribution systems in the state from the perspectives of retail customers, wholesale electricity market participants, and the combined system or “society as a whole.”

I don't think therthe a way to avoid underestimating the speed of this transition. Even for the fanboys here. Gas is going to be done much soon that we thought a year ago.

.

Being a mule, I’ll forgive you the “F-word” LOL.

 

[SageBrush]

Notice how high the AC (inverter) capacity factor is. Clearly there is an effort to economize around inverter cost.

I read it as a desire to extend the useful output of PV from say 8 hours a day to closer to 11. Some of the additional hours are from extra panels and the remainder from battery. As Californians know so well, peak hours (and peak pricing) are from ~ 4pm to 9pm. It only makes sense for the developers to take a cut of that pie if they are able.

I suppose you can translate that market signal into "economizing around the inverter" but I think that suggests that the inverter is expensive when the actual impetus is high wholesale rates during evening hours.

 

[SageBrush]

So what's the proportion of non-storage value for grid level(or aggregated distributed) battery packs now? These insane load balancing savings figures out of Australia and shockingly low solar+storage PPA bids lead me to believe utilities will get 50% or more of their value out of "non-storage" services.

Australians pay up to $1/kWh for electricity at the margin of supply due to games played by the NG incumbents. Be careful what you wish for.

 

[SageBrush]

Please enjoy The Brattle Group’s report on stacking benefits (calculated for the “deregulated” Texas grid) at surprisingly achievable price points.

Required reading. Thanks for the link

 

[jhm]

I read it as a desire to extend the useful output of PV from say 8 hours a day to closer to 11. Some of the additional hours are from extra panels and the remainder from battery. As Californians know so well, peak hours (and peak pricing) are from ~ 4pm to 9pm. It only makes sense for the developers to take a cut of that pie if they are able to.

Yes, it optimizes both the value of solar power and the cost of inverter+interconnection capacity. Both sides are important.

One of the classic arguments against solar power is that as more is added to the grid, the value of that power is reduced. Of course, this is true of any power source. But what that naive argument missed was the role that inverters and batteries would play. Once the value of incremental peak solar is below the incremental cost of the inverter capacity, DC:AC ratios shoot up. And as the cost of batteries come down, they step in to soak up the excess and tap the value of peak power prices. So super cheap panels don't so much undermine the value of solar, rather they increase the value of batteries and inverters. Ultimately the cost of peak power comes down to the cost of the battery for storage because the surplus capacities of panels and inverters provide nearly zero marginal cost power for charging the batteries and inverting its power for distribution. Thus, value erosion for solar really leads to solar providing cheap power 24x7, including the peak power. The classic solar value erosion argument was that solar power really would no value when the sun is not shining and so would be limited to just fraction of the day.

Also I think the electrolyzers have an important role to play for balancing seasonal excess supply. I just posted this too. Shorting Oil, Hedging Tesla

 

[SageBrush]

Ultimately the cost of peak power comes down to the cost of the battery for storage because the surplus capacities of panels and inverters provide nearly zero marginal cost power for charging the batteries and inverting its power for distribution

Lets flip this: storage is competitive when ancillary services value and peak power revenue exceed cost.

Say cost is $750 per kWh and that allows 0.5 kW of ancillary services
Say 4000 cycles and 15 cents per kWh, then peak power revenue is $600
Then ancillary services have to be worth over $300 per kW over the lifetime of the battery.

 

[jhm]

Lets flip this: storage is competitive when ancillary services value and peak power revenue exceed cost.

Say cost is $750 per kWh and that allows 0.5 kW of ancillary services
Say 4000 cycles and 15 cents per kWh, then peak power revenue is $600
Then ancillary services have to be worth over $300 per kW over the lifetime of the battery.

Yep, and the cost of the battery is getting cheaper, and longevity is improving too. So the new solar+storage PPAs are effectively delivering peak power at 5c/kWh. This follows from $250/kWh for battery (after 30% ITC), 5000 cycle life, and effectively zero marginal cost for the power to charge or hardware to distribute. For comparison the LCOE on a new gas peaker plant is about 14 c/kWh. So stored solar is beating this hands down. Even at $750/kWh before the 30% ITC, this would work out to 10.5 c/kWh for peak power after ITC. So it is largely cheap solar that is unlocking the value of the battery. If solar were twice as expensive, this would not be happening. But the battery is also preserving the value of the direct feed of solar by making it firm and avoiding oversupply that undermines power prices. So there is a beautiful synergy starting to take shape.

 

[SageBrush]

This follows from $250/kWh for battery

I should have clarified that my $750 per kWh cost guess was turn-key.

It does not seem like a bad guess to say that today battery storage is nipping at the heels of NG peakers and in the 2020 - 2022 timeframe that these bids are meant for, battery storage looks very competitive.

Thanks for the discussion. Quite informative and uplifting

 

[GleanerC]

"Texas to get it largest battery, coupled with largest solar power plant"

"Vistra Energy plans to add a 10 MW / 42 MWh lithium-ion battery to the 180 MW-AC Upton County Solar 2 solar power plant. The energy storage project is scheduled to come online in Q4 2018."

I haven't found where this project lists it's storage provider and I haven't found any specific price/cost numbers at this point.

Texas to get its largest battery, coupled with its largest solar power plant

This link is to a power point Vistra Energy presented regarding storage and specifically this project.

https://s21.q4cdn.com/410616722/fil...erging-Technologies_June-2018-Analyst-Day.pdf

I'm amazed at how quickly these announcements have been coming on top of each other.

 

[jhm]

I should have clarified that my $750 per kWh cost guess was turn-key.

It does not seem like a bad guess to say that today battery storage is nipping at the heels of NG peakers and in the 2020 - 2022 timeframe that these bids are meant for, battery storage looks very competitive.

Thanks for the discussion. Quite informative and uplifting

Well, the PPAs in Colorado and neighboring states are so low it's got a lot of us scratching our heads wondering how in the world are they making this work. What's implied is about $360/kWh (including installation), but I'm not sure that even Tesla is quite there yet. But by 2020, perhaps this looks possible. So all these prices are quite fluid right now. I think the big Tesla battery in South Australia, the Hornsdale Power Reserve, was around $750/kWh. I think Tesla got about $100M for 100MW/129MWh fully installed. So it is certainly not a bad assumption!

 

[jhm]

"Texas to get it largest battery, coupled with largest solar power plant"

"Vistra Energy plans to add a 10 MW / 42 MWh lithium-ion battery to the 180 MW-AC Upton County Solar 2 solar power plant. The energy storage project is scheduled to come online in Q4 2018."

I haven't found where this project lists it's storage provider and I haven't found any specific price/cost numbers at this point.

Texas to get its largest battery, coupled with its largest solar power plant

This link is to a power point Vistra Energy presented regarding storage and specifically this project.

https://s21.q4cdn.com/410616722/fil...erging-Technologies_June-2018-Analyst-Day.pdf

I'm amazed at how quickly these announcements have been coming on top of each other.

Cool. So this is adding storage to an existing solar farm. This chart is a useful illustration of what is going on. That clipped peak is getting spread out over the shoulders.

 

[SageBrush]

Well, the PPAs in Colorado and neighboring states are so low it's got a lot of us scratching our heads wondering how in the world are they making this work. What's implied is about $360/kWh (including installation), but I'm not sure that even Tesla is quite there yet. But by 2020, perhaps this looks possible. So all these prices are quite fluid right now. I think the big Tesla battery in South Australia, the Hornsdale Power Reserve, was around $750/kWh. I think Tesla got about $100M for 100MW/129MWh fully installed. So it is certainly not a bad assumption!

I don't think Colorado has peak rates anywhere near 15 cents a kWh wholesale so that remains a head scratcher ... but then I really have no idea how much ancillary services are worth*. My comments were with California in mind.

*well, perhaps a hint: in a market where all ancillary costs of the grid eventually flow to the retail kWh cost for the consumer, the ancillary costs must be less than the retail cost per kWh less the wholesale cost per kWh. Substantially less, since the utility also has maintenance and profit folded in.

 

[jhm]

https://www.bloomberg.com/amp/news/...by-cheap-renewables?__twitter_impression=true

Solar PV will soon become cheaper than wind. Batteries are coming at just the right time. I think that, with solar at lower cost, batteries will pair more frequently with solar than with wind. When we get to a point where we need a regular supply of stored power, the lowest cost source is compelling. Additionally, solar is a more predictable daily supply for charging than wind. In some places wind can cease for more than a week and be hard to predict. But if a sufficient amount of solar is paired with a battery, you can pretty reliably recharge on a daily basis.

At any rate, I'm thinking that solar installations could surge in the coming years. It will be accelerated by being the low cost leader and by capital efficient pairing with batteries.

In the Shorting Oil thread, I recently posted simply projections of the energy mix 20 years out based on trends in log-share in the last 5 years. One of the surprises it that solar looks to provide about 2/3 of total energy consumed (not just electricity) in twenty years. This is such strong growth that it causes all other energy sources to peak in less time. Even wind peaks by 2035 according to this simple projection. Is this really possible? I'm not sure. But if solar become cheapest by far and the preferred source for charging batteries and powering electrilyzers, maybe 2/3 in 20 years is not that far off. We shall see.

 

[neroden]

Unless there's some weird supply chain bottleneck we don't know about, like, I don't know, an oncoming boron shortage, then I'm pretty sure the projections for solar are correct. We've been following that exponential curve since the *1970s*, and even at 2/3 of total energy... we're nowhere near maxing out the amount of energy we *want* to use, so we're not going to be approaching market saturation.

 

[adiggs]

Unless there's some weird supply chain bottleneck we don't know about, like, I don't know, an oncoming boron shortage, then I'm pretty sure the projections for solar are correct. We've been following that exponential curve since the *1970s*, and even at 2/3 of total energy... we're nowhere near maxing out the amount of energy we *want* to use, so we're not going to be approaching market saturation.

This is how I see the energy system developing long term (say 30-100 years+). We're so deeply embedded in the idea of energy as a precious, finite, and vanishing resource that as a species, we aren't yet incorporating the idea of "approximately zero marginal cost energy".

An analogy - imagine it's the 1970's; you may recall that telephones were ubiquitous. We can make long distance calls all over the country and they have an expense associated with them that is well within the budget of most Americans. But we don't just call long distance willy-nilly either. This is an amazing communications achievement in the context of history. And in the context of today, 2018, where the marginal cost to send a text or email is approximately zero for a really high % of the worldwide population, it's ancient history. Quaint. The bad old days.

The internet and it's associated technologies have brought "approximately zero marginal cost communication" to approximately everybody. And we're still in the early stages of discovering just what all you can do with approximately zero marginal cost communication that continues to plunge in price. It's amazing how much communicating needs to be done to create a market for companies to compete in where what's being provided is approximately zero marginal cost.


Anyway - this is the future I see coming in our energy system. I can see a faint echo of it from the 10kw solar system on the roof of my house. It produces about 12-14MWh of electricity each year. To get this year's 12-14MWh of electricity, I will need to do approximately nothing, because it's already been done. The system is already fully paid for, so the energy generated won't cost me anything to produce. I'll probably go crazy and climb up on the roof (it's flattish and easy to walk around on) to wash off the pollen and dust when it accumuulates enough. But I really don't HAVE to.

It'll keep on going generating electricity for decades with little or no effort.


We're still so early in the energy transition that we're in our energy scarcity mind set. In a fossil fuel energy system, burning extra fuel to produce excess electricity is just wasteful. It's important to approximately balance energy generation with consumption, and its important that the uses we put that energy towards be at least economically as valuable as the fuel we burn to produce the electricity.

But if the energy is produced whether we want it or not (solar panels), and available for use whether the use is economically worthwhile or not, what new systems WOULD we build to make use of that waste / excess energy? When there is approximately zero marginal cost energy available seasonally and reliably, in really big quantities, what new economic activity will grow up around that zero marginal cost energy resource that sounds insane in our context today?


So I see the next 10-40 years developing along the lines of what @jhm's been talking about, especially in the recent post talking about how solar even starts crowding out wind in a finite and not too far away timeline. But after that, as solar has another 20 years of getting dramatically cheaper year after year, entrepreneurs are going to start thinking of ways of running seasonal businesses that can start and stop reasonably easily, and that can make use of all that waste energy from when the sun is shining (because we have enough solar to power society when its cloudy). Any energy generation that isn't approximately zero marginal cost will be crowded out, and others like hydro and wind may be marginalized but still present because they'll have a different cycle (the wind blows when the sun doesn't shine type of dynamic).

I really think Elon sees this coming, and his "generous" offer to provide megacharging rates of $.07/kwh in a few years will prove to be .. excessively expensive.

 

[mspohr]

This is how I see the energy system developing long term (say 30-100 years+). We're so deeply embedded in the idea of energy as a precious, finite, and vanishing resource that as a species, we aren't yet incorporating the idea of "approximately zero marginal cost energy".

An analogy - imagine it's the 1970's; you may recall that telephones were ubiquitous. We can make long distance calls all over the country and they have an expense associated with them that is well within the budget of most Americans. But we don't just call long distance willy-nilly either. This is an amazing communications achievement in the context of history. And in the context of today, 2018, where the marginal cost to send a text or email is approximately zero for a really high % of the worldwide population, it's ancient history. Quaint. The bad old days.

The internet and it's associated technologies have brought "approximately zero marginal cost communication" to approximately everybody. And we're still in the early stages of discovering just what all you can do with approximately zero marginal cost communication that continues to plunge in price. It's amazing how much communicating needs to be done to create a market for companies to compete in where what's being provided is approximately zero marginal cost.


Anyway - this is the future I see coming in our energy system. I can see a faint echo of it from the 10kw solar system on the roof of my house. It produces about 12-14MWh of electricity each year. To get this year's 12-14MWh of electricity, I will need to do approximately nothing, because it's already been done. The system is already fully paid for, so the energy generated won't cost me anything to produce. I'll probably go crazy and climb up on the roof (it's flattish and easy to walk around on) to wash off the pollen and dust when it accumuulates enough. But I really don't HAVE to.

It'll keep on going generating electricity for decades with little or no effort.


We're still so early in the energy transition that we're in our energy scarcity mind set. In a fossil fuel energy system, burning extra fuel to produce excess electricity is just wasteful. It's important to approximately balance energy generation with consumption, and its important that the uses we put that energy towards be at least economically as valuable as the fuel we burn to produce the electricity.

But if the energy is produced whether we want it or not (solar panels), and available for use whether the use is economically worthwhile or not, what new systems WOULD we build to make use of that waste / excess energy? When there is approximately zero marginal cost energy available seasonally and reliably, in really big quantities, what new economic activity will grow up around that zero marginal cost energy resource that sounds insane in our context today?


So I see the next 10-40 years developing along the lines of what @jhm's been talking about, especially in the recent post talking about how solar even starts crowding out wind in a finite and not too far away timeline. But after that, as solar has another 20 years of getting dramatically cheaper year after year, entrepreneurs are going to start thinking of ways of running seasonal businesses that can start and stop reasonably easily, and that can make use of all that waste energy from when the sun is shining (because we have enough solar to power society when its cloudy). Any energy generation that isn't approximately zero marginal cost will be crowded out, and others like hydro and wind may be marginalized but still present because they'll have a different cycle (the wind blows when the sun doesn't shine type of dynamic).

I really think Elon sees this coming, and his "generous" offer to provide megacharging rates of $.07/kwh in a few years will prove to be .. excessively expensive.

Elon is known to be a fan of Iain M. Banks' Culture series which is scifi which posits a post-scarcity world.