Tesla Energy and utility scale projects

[jhm]

California recently passed legislation that requires the PUC to develop rules that facilitate the development of clean microgrids.

Microgrid Legislation in California Could Become National Model
http://leginfo.legislature.ca.gov/faces/billTextClient.xhtml?bill_id=201720180SB1339

Depending on how the legislation is interpreted and rules are structured, it may permit Tesla and other companies to install solar and storage at residential and commercial customers and provide a wide range of grid services. This could make home solar/storage much more affordable and accelerate adoption.

Hopefully the PUC rules will encompass virtual powerplants. Maybe @jhm or one of our other technically savvy members can weigh in on whether a Tesla managed solar/storage VPP would qualify as a microgrid under the legislation:

"(d) “Microgrid” means an interconnected system of loads and energy resources, including, but not limited to, distributed energy resources, energy storage, demand response tools, or other management, forecasting, and analytical tools, appropriately sized to meet customer needs, within a clearly defined electrical boundary that can act as a single, controllable entity, and can connect to, disconnect from, or run in parallel with, larger portions of the electrical grid, or can be managed and isolated to withstand larger disturbances and maintain electrical supply to connected critical infrastructure."

Given California's strong policy in favor of developing robust distributed energy resources hopefully the PUC will find another way to develop rules and regulations that support VPPs even if they don't meet the definition of a microgrid. Seems insane not to.

It looks like batteries will be the brains of clean microgrids, as fossil-fuel based generators are generally excluded. I suspect that Tesla Energy will have no particular regulatory problem competing in this space. While the grid is viewed as backup, I think things will operate differently than thinking of the grid as some sort of NG backup generator. With batteries as the intelligent core for power management within the microgrid, the macro-grid serves more as a low cost way to maintain the SOC in local batteries. So the microgrids may draw grid power opportunistically, i.e., they mostly buy grid power when it is cheapest through the day. If this becomes the case, only a very low tariff may be needed. Essentially, the microgrids will make it easier for the macrogrid to manage it's power supply. For example, it will reduce the need for peaking generation capacity. It will also assure that demand on the transmission grid is more consistent throughout the day. This means greater utilization of existing infrastructure and less wear on components during times of stress.

I actually think virtual power plant is a bad metaphor for a distributed storage system. The basic objective is buffer power locally so that local consumption can be served at lowest cost. The idea that you want to store power locally so that it can be transmitted across the state at another time is silly. It's a bit like saying I'll save groceries in my refrigerators so I can take veggies back to the market and ship them off to another state when they need them. Even if I have a small garden, this is a silly idea. There is a necessary one-way flow of power from centralized generators to consumers. So localized storage is only really needed for local consumption. Amassing more storage locally than what is needed to buffer local consumption means that storage is not well distributed. While local consumption may be the intent of VPP, the metaphor of a power plant creates a false impression. One would never call a refrigerator a virtual farm or even a virtual grocery store. We don't need silly metaphors like that because we know exactly how useful a refrigerator is. I submit that batteries don't need a metaphor. It is not a virtual anything, it's just a power storage device. Once we get used to how local batteries can optimize local power consumption, we'll have a hard time remembering how we got by without it.

 

[jhm]

"A trip at KIUC’s Kapaia Power Station at 12:04 p.m. this afternoon led to a power outage affecting all of KIUC’s 25,800 members. Power continued to be supplied to the grid for approximately nine minutes following the trip, in large part due to KIUC’s battery storage facilities."

http://website.kiuc.coop/sites/kiuc/files/documents/pr2018-1105-Islandwide outage.pdf

Nice case in point. Imagine if the island also had microgrids. In such an event, the microgrids would cut off immediately from the larger grid and use their own storage and other resources to sustain local power for an hour or more. Meanwhile the batteries in the larger grid would also last longer supplying just the portion of the island not served by a microgrid. This demonstrates how microgrids can provide benefits to all grid users not just those directly served by the microgrid. If all users are in a microgrid, then they all have very high reliability.

 

[EinSV]

I actually think virtual power plant is a bad metaphor for a distributed storage system. The basic objective is buffer power locally so that local consumption can be served at lowest cost. The idea that you want to store power locally so that it can be transmitted across the state at another time is silly. <snip>

Curious if you feel the same way for projects like the proposed South Australia VPP that also provide power from distributed resources -- in the case of the SA project 250 MW/650 MWh from 50,000 homes with rooftop solar and storage, but controlled centrally. The project is expected to provide 20% of the power needs of South Australia.

To me the "virtual power plant" concept for a project like this that generates a large amount of power with distributed resources but can be controlled centrally seems helpful conceptually as the project has much in common with a large utility scale solar/storage project except the resources are distributed. The project description says:

1. What is a virtual power plant? How does it work?

A virtual power plant is created by a network of home solar photovoltaic (PV) and battery systems all working together to generate, store and feed energy back into the grid. This is similar to the ‘big battery’ at Jamestown, but instead of being in one location, the systems are spread over thousands of houses.

Energy from the home solar PV and battery systems installed as a part of this virtual power plant will provide electricity for the house on which they are installed. Any excess energy generated by the system will be automatically dispatched to the grid. This dispatched energy will be centrally controlled to meet the needs of the grid, providing additional energy to the rest of the state, when it is required.

South Australia's Virtual Power Plant | Virtual Power Plant
​

One thing I like about the "virtual power plant" terminology for large distributed rooftop solar/storage projects like this is that it conveys to the public the idea that this sort of project can replace a traditional power plant.

 

[jhm]

Curious if you feel the same way for projects like the proposed South Australia VPP that also provide power from distributed resources -- in the case of the SA project 250 MW/650 MWh from 50,000 homes with rooftop solar and storage, but controlled centrally. The project is expected to provide 20% of the power needs of South Australia.

To me the "virtual power plant" concept for a project like this that generates a large amount of power with distributed resources but can be controlled centrally seems helpful conceptually as the project has much in common with a large utility scale solar/storage project except the resources are distributed. The project description says:

1. What is a virtual power plant? How does it work?

A virtual power plant is created by a network of home solar photovoltaic (PV) and battery systems all working together to generate, store and feed energy back into the grid. This is similar to the ‘big battery’ at Jamestown, but instead of being in one location, the systems are spread over thousands of houses.

Energy from the home solar PV and battery systems installed as a part of this virtual power plant will provide electricity for the house on which they are installed. Any excess energy generated by the system will be automatically dispatched to the grid. This dispatched energy will be centrally controlled to meet the needs of the grid, providing additional energy to the rest of the state, when it is required.

South Australia's Virtual Power Plant | Virtual Power Plant
​

One thing I like about the "virtual power plant" terminology for large distributed rooftop solar/storage projects like this is that it conveys to the public the idea that this sort of project can replace a traditional power plant.

Being centrally controlled is a good feature, especially if you are running the grid. But this is inherently tied to the idea that a grid ought to be centrally controlled. The microgrid concept is much closer to the concept of a decentralized grid. The macro-grid ultimately becomes a clearing house managing excess supply and demand across of wide area. Most of the power control is actually in the microgrids. So rather than having some centralized dispatcher control the operation of your battery, your microgrid is more simply responding to price signals. Thus, demand for grid power becomes much more elastic, and this will drive down the cost of power enormously. The old centralized grid model makes demand almost completely inelastic and tries to balance supply and demand almost entirely on the production side. The reason why a VPP needs to be dispatchable is because price signals would not otherwise trigger distributed storage to buy and sell when the market most needs it to offer that correction. So I think that VPP along with the whole centralized grid model is doomed to be less than financially efficient.

 

[neroden]

OK, I'm impressed. The entire *island* of Kauai went out due to a switch issue and the battery kept the island going for nine minutes? They're not done installing batteries! And they are still putting in more wind and solar.

 

[Dave EV]

"A trip at KIUC’s Kapaia Power Station at 12:04 p.m. this afternoon led to a power outage affecting all of KIUC’s 25,800 members. Power continued to be supplied to the grid for approximately nine minutes following the trip, in large part due to KIUC’s battery storage facilities."

http://website.kiuc.coop/sites/kiuc/files/documents/pr2018-1105-Islandwide outage.pdf

It looks like if Kauai had a few more storage installations like the Tesla install already there, the batteries could supply 100% of the load for an extended period of time which could eliminate outages like this one.

I'm actually a bit surprised that the combination of batteries and solar wasn't sufficient to keep the grid running for longer considering that the outage happened in the middle of the day when solar should be producing. KIUC claims that when it's sunny 90% of the energy on the grid is from renewables, mostly solar with over 50 MW of solar installed. http://website.kiuc.coop/renewables The plant that tripped is a 39 MW oil burning power plant.

Would be very interesting to get a more detailed report.

 

[Dr. J]

To me the "virtual power plant" concept for a project like this that generates a large amount of power with distributed resources but can be controlled centrally seems helpful conceptually as the project has much in common with a large utility scale solar/storage project except the resources are distributed.

Perhaps "distributed power plant" would be better terminology.

 

[jhm]

Perhaps "distributed power plant" would be better terminology.

My objection is more to the "power plant" part than to the "virtual" part. In my view it is a "distributed energy storage system." An storage system does much more than just generate power. A DESS does more than just replace a few power plants. It can balance load, solve distribution and transmission problems and improve reliable of service to end users. We distribute storage, not because we lack for centralized locations to put batteries, but because placing them in the distribution system creates value that centralized resources cannot deliver.

How Dumb Distribution Spending Crowds Out a Smart Clean Energy Future
DESS is a species of Distributed Energy Resources (DERs). DERs are a key strategy for making the distribution system more efficient. This article points out that distributional costs have become a larger share of the cost of grip power as generation costs have declined.

As a result of this investment strategy shift, the cost of delivering energy through transmission and distribution systems has grown relative to the cost of generating electricity. Major investor-owned utility data shows what percentage of each delivered unit of energy (kilowatt-hour) is due to generation costs and delivery costs. Delivery costs have expanded from 22 percent of overall costs in 2006, to 36 percent in 2016, while generation costs have fallen from 69 percent to 54 percent.

This otherwise great article does miss the particular benefits that a DESS can bring. For example, it can cost a utility millions of dollars to upgrade transformer at a substation. The challenge is to have enough capacity to handle peak demand times. If peak demand grows, the existing transformers are aged more quickly operating to close to peak capacity, and that peak capacity may even become insufficient. Rather than spending millions on early transformer upgrades, the utility can install batteries within the substation. These batteries are then used to provide peak shaving for the transformers. So collectively even higher peak loads can be served and the transformers are operated at lower levels to enhance their longevity. So this can save substantial distribution costs, but it also has the side effect of avoiding need for centralized peakers because high demand within the substation will often coincide with use of centralized peakers in the transmission system. This also avoids congestion in the transmission system.

So a DESS can improve distribution, transmission and generation all at the same time. To sell it as a Virtual Power Plant is to narrow attention to just the generation benefits, but really the benefits to distribution are primary. Transmission and generation are secondary benefits in my view. Both DERs and microgrids deliver distribution benefits primarily, though they coordinate in different ways.

 

[jbih]

Tesla promotional video about Gigafactory 2 (Buffalo):

 

[electracity]

Tesla promotional video about Gigafactory 2 (Buffalo):

Apparently the robots were sleeping when the video was made. Or Tesla solar roof will be marketed as "Artisan".

 

[replicant]

Apparently the robots were sleeping when the video was made. Or Tesla solar roof will be marketed as "Artisan".

I guess the video does not target investors but parties most interested in local jobs.

 

[Dr. J]

Tesla promotional video about Gigafactory 2 (Buffalo):

Is the audio missing from this, or is that just me?

 

[electracity]

I guess the video does not target investors but parties most interested in local jobs.

Yes, it was from a press even targeted at New York. Obviously making the tiles is a processes that is well suited to automation.

Interestingly, one of the local papers wrote an article claiming that Musk has never visited the factory.

 

[electracity]

Is the audio missing from this, or is that just me?

It's just video. It's intended primarily to be cut into local broadcast news stories.

 

[jbih]

Interestingly, one of the local papers wrote an article claiming that Musk has never visited the factory.

Yes, and the local workers don't seem to be happy about it. On the other hand, Musk seems to be having weekly online meetings and to be always available when needed. And has personally given feedback on the progress of the tiles... Or rather sent them back to the drawing board when he wasn't happy. See Bloomberg - Are you a robot? (The start of the article sounds like FUD, but it contains a lot of info on Gigafactory 2/Tesla Energy, solar division.)

 

[jhm]

Yes, and the local workers don't seem to be happy about it.

They should be quite happy if Musk never camps out on the production line.

 

[electracity]

Yes, and the local workers don't seem to be happy about it. On the other hand, Musk seems to be having weekly online meetings and to be always available when needed. And has personally given feedback on the progress of the tiles... Or rather sent them back to the drawing board when he wasn't happy. See Bloomberg - Are you a robot? (The start of the article sounds like FUD, but it contains a lot of info on Gigafactory 2/Tesla Energy, solar division.)

I'm not sure what the Buffalo people are looking for . There are 800 jobs that they would not have without the factory. They will likely have a few thousand jobs in a couple of years.

 

[aubreymcfato]

I'm not sure what the Buffalo people are looking for . There are 800 jobs that they would not have without the factory. They will likely have a few thousand jobs in a couple of years.

They were promised more jobs sooner, as well as a growing "collateral" economy for local suppliers, vendors, etc. Between Solarcity and Tesla, these promises are late quite a few years. As an investor, I can "forgive" Tesla, especially given the circumstances that we know of, but I think it's perfectly fine fair for them to be criticized. It appears they are doing the legal minimum not to be sued.

 

[jbih]

They were promised more jobs sooner, as well as a growing "collateral" economy for local suppliers, vendors, etc. Between Solarcity and Tesla, these promises are late quite a few years. As an investor, I can "forgive" Tesla, especially given the circumstances that we know of, but I think it's perfectly fine fair for them to be criticized. It appears they are doing the legal minimum not to be sued.

We should remember that SolarCity was under attack from short sellers and probably wouldn't exist without Tesla now. So it's 800 jobs that wouldn't otherwise exist.

 

[EinSV]

They were promised more jobs sooner, as well as a growing "collateral" economy for local suppliers, vendors, etc. Between Solarcity and Tesla, these promises are late quite a few years. As an investor, I can "forgive" Tesla, especially given the circumstances that we know of, but I think it's perfectly fine fair for them to be criticized. It appears they are doing the legal minimum not to be sued.

I can see why you have this impression given some of the hit-pieces about GF2 in the press recently (including from Bloomberg) but it is not true, at least with respect to Tesla.

Under its agreement with the state of NY, Tesla promised 500 jobs at GF2 by April 2019. Tesla has two years to bring 1,460 jobs to Buffalo. Can they do it?

They already have met that target way ahead of schedule with 800 employees hired at GF2 between Tesla and Panasonic. Tesla offers glimpse inside Buffalo gigafactory

They also promised 1460 jobs at GF2 by April 2020 and 5000 jobs in NY w/n ten years. They are well ahead of schedule to hit the 1460 employee goal in 2020. The other one is a long-term goal so I think it is fair to give them time to reach it.

Tesla has also now trimmed a tremendous amount of fat from the solar sales organization to get rid of unsustainable door-to-door sales that cost an estimated $4500 per customer and an arrangement with Home Depot that cost an estimated $7,000 per customer. Tesla has two years to bring 1,460 jobs to Buffalo. Can they do it?

Selling through Tesla stores and the internet instead of expensive methods like door to door and Home Depot has already allowed it to cut solar panel prices by $3000-$5000. A lean sales structure should set the stage for the solar business to be very profitable once solar roof ramps up over the next year or two.

There obviously have been delays in ramping solar roof production, but Tesla has met and exceeded its job commitments to New York and has restructured its sales organization to reduce OpEx, cut prices and set the stage to become a profitable business. Seems to me they are on the right path.