PowerWall and "The Missing Piece..." Event

[jhm]

Is $175/kWh not somewhat optimistic at this time? Do we have any data points to suggest costs are really below 200 already? I find 33% GM very high also. Of course this would be wonderful but I prefer to be cautious.
My semi-conservative numbers would be $4500 20kWh cells @ $225/kWh with the other numbers being about right (I think $500 is probably an overestimate for case+controller, but let's stick with it). Where I disagree with you is that Tesla needs to offer the system at a lower price point than competition. It doesn't do so for its cars either, and we can be pretty certain that Tesla's product will have an edge on competitors in convenience features and clever technology (plus almost certainly some neat perks for Tesla car owners).

So let's say Tesla's sales price is also $400/kWh, so $8000 for a 20kWh system. With the numbers I suggested above, this gives $2000 gross profit for a 25% GM.

Gerasimental,
It's interesting how close we are per kWh, $400 you vs. $375 me. If you're trying to be conservative and think that this product does not need priced so competitively, why not assume a higher GM? $9000 on a 20 kWh device would net $3000 for a GM of 33%.

I certainly agree that the pricing strategy for stationary does not need to be the same a for autos. One big difference is that Tesla can begin with a high price like $9950 for a 20 kWh system and see how it sells. They can progressively lower price as production increases, manufacturing costs decline, and to fend off competition. This is in contrast to the situation with the Model S where Tesla does not reduce the price so as to preserve resale value for buyers. With stationary, there is not much concern about resale value, at least not in a way that would motivate Tesla to never reduce the price of a 20 kWh system. So it they start high, they can get a really nice GM and tap into markets that are most willing to pay that price. But longer run they will want to address bigger markets. So each time they lower the price or improve the product they expand the addressable market. Another basic problem with starting with too low of a price is that they will simply create a huge back order, but will not be able to plow a fat GM back into ramping up production faster. So if Tesla really wants to optimize the roll out of capacity, it seems better to price as high as the market is able to support and cut the price over time.

So if I had another guess, I would put a 20 kWh system at $8995, or $450/kWh. Let's see how this plays out on 4/30.

 

[stevej119]

The possibility of the announcement including hooking Tesla vehicles up to the grid allowing to sell power back this way haven't been talked about a lot, but couldn't that be pretty big? Now I am not sure how much wear and tear this could do to the battery but I could imagine it would be far worth it in states like California with a high peak to off peak energy price discrepancy. Say you traded 50 kwh of power with the grid every day at a price difference of 15c per kwh (not sure if this is realistic, please correct me if you know better), that would be $7,5 of profit a day, or $2700 per year, pretty serious cash, it could potentially pay for a new battery in 8 years time. For this to be viable it would probably need some nice software to come with the system allowing you to decide how low the battery is allowed to go in capacity depending on your expected needs to use the car, an option to have it at a certain charge at a certain time when you expect to use it, stuff like that which I am sure would be easy for Tesla to configure.

That would be great, but PG&E (the utility company in Northern California), will only buy the power back from us at the low wholesale price of approximately 4.5 cents per kilowatt, regardless of what time it's generated. You do get the full peak energy price for the power you generate in the daytime credited against your off-peak usage, but anything you generate in excess of your overall usage is purchased from you at a very, very low rate.

 

[Perfectlogic]

That would be great, but PG&E (the utility company in Northern California), will only buy the power back from us at the low wholesale price of approximately 4.5 cents per kilowatt, regardless of what time it's generated. You do get the full peak energy price for the power you generate in the daytime credited against your off-peak usage, but anything you generate in excess of your overall usage is purchased from you at a very, very low rate.

Not sure I understand you correctly but I wasn't talking about being a net seller to the utilities, but charging your battery with cheap off peak energy, and sell it back at peak hours, like this guy Tesla Home Battery Details Emerge - Gas 2

 

[jhm]

The possibility of the announcement including hooking Tesla vehicles up to the grid allowing to sell power back this way haven't been talked about a lot, but couldn't that be pretty big? Now I am not sure how much wear and tear this could do to the battery but I could imagine it would be far worth it in states like California with a high peak to off peak energy price discrepancy. Say you traded 50 kwh of power with the grid every day at a price difference of 15c per kwh (not sure if this is realistic, please correct me if you know better), that would be $7,5 of profit a day, or $2700 per year, pretty serious cash, it could potentially pay for a new battery in 8 years time. For this to be viable it would probably need some nice software to come with the system allowing you to decide how low the battery is allowed to go in capacity depending on your expected needs to use the car, an option to have it at a certain charge at a certain time when you expect to use it, stuff like that which I am sure would be easy for Tesla to configure.

I think Musk has ruled out the possibility of selling power back to the grid on the basis of it being too dangerous. However, I believe their are sound ways Tesla could make money from offering a dispatchable load to the grid. That is, controlling the rate of charge across a fleet of vehicles so that it helps stabilize the grid. The first step in this direction is simply to provide advanced charging features in vehicles through software upgrades that optimize the ratepayer's power bill. They've done similar programing with DemandLogic. Basically, you would plug your car in at night and indicate the time by which you want charging to finish. The software already knows the parameters of your rate plan and will optimize the charging session to minimize your cost. With a fleet of intelligent charging vehicles in place, Tesla could consult with utilities on how to offer special rate plans that would be a win for both the utility and the EV owner. This is the simplest set up I see.

To take this to the next level, suppose Tesla has some sort of device that knows what the load is from the rest of your house and output from solar if available. I would expect the stationary products to be unveiled to have this functionality as well, but it is not actually necessary that a stationary battery be involved. In any case, charging can be even more intellegent and manage peak demand for the whole house. So a rate plan that includes demand charges can use this technology to minimize those demand charges, i.e. peak shaving, even without a stationary battery. Again Tesla can consult with utilities to design mutually beneficial rate plans.

The most complex set up I see is where Tesla delivers real time load management services to utilities. Here's how it works, Tesla owners enroll their cars in the load the dynamic load management program. This allows Tesla to manage the rate of charge in real time, in exchange you get a rebate on your power bill based on how much dynamic charging you receive at the dynamic discount. Tesla coordinates with the utility monitoring spot prices and system load. When load and spot prices drop a discount is offered and Tesla sends signals to each vehicle to increase the charging load. So basically your car charges at a higher rate as the spot price rises. Meanwhile Tesla and your car are actively managing the charging process so that you get your full cage by the time you need it. So Tesla earns a fee from the utility for this service, the car owner gets a rebate on their power bill, and the utility is better able to manage its load and power production costs. This is a triple win. Notice also that this framework can apply to charging both auto batteries and stationary batteries. It is not necessarily for utilities to buy grid batteries to get many of the benefits of batteries that are networked and dispatchable on the charge cycle.

 

[Perfectlogic]

I think Musk has ruled out the possibility of selling power back to the grid on the basis of it being too dangerous.

Dangerous how? People are already selling back power like I described (see my link).

 

[bonnie]

That would be great, but PG&E (the utility company in Northern California), will only buy the power back from us at the low wholesale price of approximately 4.5 cents per kilowatt, regardless of what time it's generated. You do get the full peak energy price for the power you generate in the daytime credited against your off-peak usage, but anything you generate in excess of your overall usage is purchased from you at a very, very low rate.

True. But if I could load the battery with off-peak power at the lowest rate & use that to offset my daytime use during those times of year where I'm not producing excess during the day, it's still a win.

 

[bonaire]

A peak load shaving battery for CA residents can do well. The idea is to sell it on the premise that CA utilities are already on board and ready to offer TOU rates that make the units virtually free over their lifetime.

One thing to know, however, is there is a charging loss during night time charging. To output 10kWh in the daytime peak, it will take over 11, perhaps almost 12 kWh to charge it up nightly. This aspect will not be something people discuss much. Fox News might use that to play it down.

it really is not selling power back but lowering peak demand such that power plants need to produce less during daytime hours. The reason it "works" is the daytime consumer rate plans have high discouraging kWh pricing to try to get customers to use less power during the day. Power companies will want to do this on a larger scale.

 

[Perfectlogic]

it really is not selling power back

Is anyone going to explain why this is the case?

 

[bonaire]

Is anyone going to explain why this is the case?

Peak load shaving is not selling power back. Neither is storing Solar PV energy for use later in the day. It is expense avoidance for the rate-payer as well as a general benefit to the power company who will pay less for their electricity from power-generators. There will be limits to what a site can do - and power companies in conjunction with the Public Utility Commissions, can limit how much night-time energy is captured at low rate and output during the day at high rate. If the idea is every homeowner installs 10 units (ie. if they are 10 kWh each) and dims the lights at night to charge them up for 100 kWh and then dumps out the power during the day along with their Solar PV for net-metering to try to make a profit, both PUCs and power companies will react. Even in California.

In other states right now, a homeowner cannot install a Solar PV array that is larger in output than their typical usage pattern - or face fees or other wholesale-sellback limits. The reason is they get unjust benefits. PUCs and other law makers want to avoid unjust benefits garnered by those who can afford to game the system.

In other words, if a homeowner plans on buying 100 kWh of standby batteries and has a TOU rate of .05/kWh at night and daytime rate of .25/kWh retail pricing, they won't be allowed to net-meter off that 100 kWh and make $20 a day selling electricity ($2/day with one 10kWh unit). There will be metered limits. It's not me talking - it is the power companies who will fight this issue and generally, the one entity that benefits most is the one who installs the system due to the economic benefit of the situation.

Let's say the 100 kWh system garners $4000-5000 a year in "power sales" for the homeowner who usually paid $200-300/month in utility bills. What power company, state PUC or any logical lawmaker would allow for this? And 10 10kWh units will take up a good amount of space, so this type of install is really for someone with a barn in a warm area, large basement or big garage. Or, a store - that is the point of the commercial unit - they use a lot of power during the day and can get a big payback for installing such a unit as long as the price is right.

As you can see, it is up to how well the local laws are setup, how the power company billing structure is configured and what is in place now, or in place later. Don't get any "big ideas" just yet on how much power you can "sell" before understanding how your local power company will view your side of the meter. If I can think of this, power company rate planners and PUC members can also. Look into how badly the SREC market fell apart for those who planned on making a lot of money selling SRECs when they installed their over-sized home Solar PV system.

If thousands of people in California are thinking "give me one big-honkin battery - I will make a ton selling peak period power to the power company" - do some research on net-metering limitations now and what is yet to come. Talk to your Power Utility Commission's customer advocate and/or ombudsman.

Would you buy a home battery if you cannot make excess profit selling power to the power company? Would you buy it for what it is meant for - standby power and peak load shaving as a common good to do your small part in a much larger plan? The one big question is just how much is a single 10kWh unit going to cost (total install price and/or by-month lease cost). Until that is known, it really is a big unknown right now on the economics of all this.

Also, saying batteries can return a higher value to a homeowner than installing Solar PV itself (which it might if the economics are right) means that it could slow-up the adoption of Solar PV. Batteries must be allowed with or without Solar PV, of course. Solar PV must also be allowed without batteries. A very short-sighted state PUC would require that both be installed together by mandate. With wide TOU rates, California may be the only state where selling back during peak-load periods actually makes economic sense right now. Many states don't even offer a differential between day and night rates to homeowners but may do so as more smart-grid products are deployed - such as two-way smart meters to handle Solar PV and other sell-back situations.

Keep in mind that base-load power plants buy power at a basic wholesale price usually under $40/MWh from system generators. It is the peaker plants (also known as synchronous standby as well as other asynchronous standby) who are on contract at much higher rates that are the impetus of peak-load shaving. Power companies pay these generators at higher rates for short periods of time during the peak power hours of the day. If they can reduce their downstream demands in the later afternoons on summer days, they will be able to turn a higher profit at the power company. However, are they willing to pay Retail prices of up to $250/MWh to ratepayers (.25/kWh) for this feedback service even on days when they are paying only baseload prices of $40/MWh? I doubt that holds up as a smart business decision and the power companies would surely work with PUCs to "make it right". This is also why power companies like SCE and PP&E are working to try to limit Net-metering payments and go with gross metering for excess kWh when a rate-payer reaches net-zero. That means battery storage will also come into play in this area. A power company installing their own series of 16-32 MWh demand-response systems may cost them much less than agreeing to pay rate-payers retail prices to install the same 16,000 to 32,000 individual 10kWh units. As this technology plays out, the larger the system, the lower the cost per kWh. The tipping point is when can a power company make it work for them economically and a supplier build them a system where both parties can profit (a "win-win"). 32,000 individual rate-payers with 10kWh systems selling power back at $2/day opportunity-cost adds up fast. Tesla will offer the battery system(s) based on today's market situation. But as Gretzky says "skate to where the puck will be" - the power companies will be looking at what the market will look like in years to come and how their revenue stream is impacted by battery storage on the customer-side of the meter.

And so - that is why battery storage is not "selling power to the utilities". It is meant to smooth out the power curve overall by a variety of demand sites cutting back during peak load periods.

Readings and research:
FERC: Electric Power Markets - National Overview
http://ferc.gov/market-oversight/mkt-electric/california/2013/08-2013-elec-ca-archive.pdf
(below, look at the hourly pricing peaks - that is what the power companies want to reduce)
http://ferc.gov/market-oversight/mkt-electric/california/2011/08-2011-elec-CAISO-dly.pdf

SCE is already doing larger-scale system procurement. Some of this is being legislated into place by CA government and other parties.
Breaking: SCE Announces Winners of Energy Storage Contracts Worth 250MW : Greentech Media
North America’s Largest Battery Energy Storage System Now Operational
California Passes Huge Grid Energy Storage Mandate : Greentech Media

As the Greentech media article says, the future is still being planned and how you may be able to "profit" from it remains unknown:

Specific details pertaining to how the CPUC will regulate customer-owned storage assets, beyond existing programs like the state’s Self-Generation Incentive Program, will be “addressed in a future rulemaking,” Thursday’s decision notes. The ruling creates three separate classes of storage at the transmission-connected, distribution-connected, and customer-side levels. But just how utilities and third parties will manage the interplay of ownership and operation of those distinct assets remains a question to be answered in the future as well.

 

[dweeks]

This in today's (Sunday's) New York Times:

Solar Power Battle Puts Hawaii at Forefront of Worldwide Changes

Log In - The New York Times

 

[atang]

This in today's (Sunday's) New York Times:

Solar Power Battle Puts Hawaii at Forefront of Worldwide Changes

Log In - The New York Times

Thanks for that article dweeks! Makes one think about rushing out to buy more SCTY/TSLA stock. :wink: The rEvolution is on!

 

[Perfectlogic]

@bonaire

Thanks for the long thoughtout response, I do think there has been some misunderstandings though. I was suggesting the possibility of a car to grid channel, allowing you to sell back power stored in your Model S battery, I agree buying a 100 KwH storage system wouldn't be a wise choice as it would carry a high amount of risk due to the high upfront cost and not knowing hwether you can make the same money off it 5 years down the road as you can today, the car battery however is most of the time just sitting in your garage. You mention solar a lot which isn't really relevant to my suggestion, well true if you are already producing all the power you use from your own solar panels then apparently you can't use the scheme I suggested, but this is still only a small minority of home owners.

I don't think the utilities are particularly interested in shaving the peaks of demand, yes they pay a lot more for peak power but they are just sending that bill to the consumer, with a markup ofcourse, so as long as they don't take a serious margin hit on the higher peak power revenues, if any, they are making more money with the peaks. It would save the consumer money to shave the peaks though so if you could sell back power from your car battery it would be advantageous to everyone, but the utilites, who would earn a bit less money due to a revenue cut.

"However, are they willing to pay Retail prices of up to $250/MWh to ratepayers (.25/kWh) for this feedback service even on days when they are paying only baseload prices of $40/MWh?" I don't expect the utilities to pay more for power from the customer selling back than they could have done elsewhere, the scheme only works when they otherwise would have used those peak plants.

 

[bonaire]

PL, right. However the catch 22 is the car itself would make the perfect V2G solution, but puts a hurt on the future sales of storage batteries and also would cycle the car battery a lot more, impacting longevity. I think a lot of people want to sell energy to the grid and make some sort of pay back from it and that is their main reason both for investing in the stocks (tsla and scty) whilenot fully understanding why the grid providers are considering storage. They utilize contracted synchronous standby generators who run generators even if not outputting to the grid. To be greener, batteries can serve that standby service without burning fossil fuels. Some generators are not hot and need some time to come up. Batteries help lower the need for synchronous standby and can be placed closer to the customer, such as at substations or even in customer buildings. I suspect companies already with solar pv will add big batteries, such as walmart in California. They have 100 or more walmart stores with large solar arrays. They still can get charged a peak power price in later afternoons even if solar is there on hot cloudy days. If they can charge the batteries late at night, they can peak load shave in later afternoons in response to grid signals asking customers to cut demand. Battery companies have already been helping companies do this load shaving for years. That is the smart grid of tomorrow. It is one which asks customers to decrease load when demand is ramping up.

 

[mspohr]

Peak load shaving is not selling power back. Neither is storing Solar PV energy for use later in the day. It is expense avoidance for the rate-payer as well as a general benefit to the power company who will pay less for their electricity from power-generators. There will be limits to what a site can do - and power companies in conjunction with the Public Utility Commissions, can limit how much night-time energy is captured at low rate and output during the day at high rate. If the idea is every homeowner installs 10 units (ie. if they are 10 kWh each) and dims the lights at night to charge them up for 100 kWh and then dumps out the power during the day along with their Solar PV for net-metering to try to make a profit, both PUCs and power companies will react. Even in California.

In other states right now, a homeowner cannot install a Solar PV array that is larger in output than their typical usage pattern - or face fees or other wholesale-sellback limits. The reason is they get unjust benefits. PUCs and other law makers want to avoid unjust benefits garnered by those who can afford to game the system.

One man's "gaming the system for unjust benefits" is another's "free market". It seems that the free market should be able to deal with this without a lot of artificial regulations. If a utility is paying $250/MWh during the day from a peaking plant, they should also be happy to pay me the same amount from my home battery storage. If I can buy power from the grid at $40 at night and sell it back at $250 during the day, the power company should be happy. If they aren't, they need to change their prices (or set up their own batteries to go into competition with me).

 

[bonaire]

Msphor, i gave some links above indicating power companies (PoCo) are already going to at least setup initial test systems of large scale batteries. They have existing buildings, sites, engineers and so on. They can implement a sizable system for less costs than mandated subsidies to get behind the meter customers implementing them. This is easier than setting up solar pv generation where acres of land is needed versus using customer rooftops. It is easier for a PoCo to become a competitor for battery storage than one for solar pv generation. Especially when generators charge as low as $20-30 per MWh for base load power. Interesting market where the more solar pv is installed, the less peak power plants are needed...except when they are on those few but troublesome hot sultry summer afternoons when the clouds roll in and the wind turbines slow down. Those few days are critical for capacity of the standby systems or at least to have compliant customers who can go to an island and get off the grid. Hospitals and schools as well as some factories can go offline and get paid to do so as a demand-response compliant customer.

if distributed batteries do output their energy store in the later afternoons, profits at power companies can go up as less peak power is demanded and paid for. This is important for when Natural Gas generators pricing goes up should NG fuel prices rise again to the $10 or higher price as seen in 2007 during the heavier demand periods which also was pre recession. Since then many factories have moved overseas and thus less grid demand since then. Our corporations moving manufacturing off shore in the last 15 years has been quite nice for lowering our grid demands and helping keep electricity prices down. Including the drop in price of Natural Gas.

 

[jhm]

So long as Tesla is underwriting a warranty on auto batteries, I don't think that putting power back into the grid makes sense. I certainly do not like the idea of utilities having direct control over both charging and discharging. The utilities could work the batteries like rented mules and Tesla would be on the hook for replacing batteries under warranty.

On the other hand, allowing Tesla to coordinate charging so as to provide load management to utilities in exchange for rebates to the car owner and fees to Tesla makes sense to me. Tesla can be expected not to overwork the batteries since they are underwriting the warranties on them.

Curiously if utilities are unwilling to provide fair rebates, under this set up of coordinated charging on a sufficient volume of batteries, Tesla can do economic harm to the distribution utilities. They can ramp up load when the spot market is high and they can also pulse load to force utilities to pay more for standby power. Basically the utilities can work with Tesla and smooth out the load, or Tesla can make things very expensive for them. So the key thing is that Tesla needs to put alot of cars on the road and batteries in homes and business. For example, 100,000 vehicle plugged in at a potential charge rate of 10 kW each represents a dispatchable load of 1GW which is comparable to the output of an average nuclear power plan. Being able to control 1GW of load is huge buying power. Eventually the utilities will have to reckon with this buying power.

Right now Tesla does not need to do too much. Just play along with the utilities while building up their fleet, but once Tesla is pumping out 1 million cars, 100 GWh per year, Musk can retire to his volcano lair.

 

[PaulusdB]

Now that's an original thought! Thx for that, jhm.

 

[Perfectlogic]

PL, right. However the catch 22 is the car itself would make the perfect V2G solution, but puts a hurt on the future sales of storage batteries and also would cycle the car battery a lot more, impacting longevity.

I know, I mentioned the wear and tear of the battery in my first post, but unless you know how much it deteriorates the battery exactly, its might still be very worth it to use the battery for that purpose, if you can earn 15c or more per kwh you sell back its possible to make a lot of money doing it, possibly paying for a new battery in 8 years. I am pretty sure most understand how batteries can be used to smooth out the demand curve and lower the need for those expensive peak plants, its pretty straight forward. Money can be made/saved by saving energy from night time and using it during the peak demand period like you say I am not disputing that fact, but on top of that, which would accelerate the flattening of the demand curve even further, one could store a significant amount of cheap energy and pump it back into the system when there is a demand/supply imbalance.

@jhm

I'm don't think anyone has suggested that the utilities should be in control of the car batteries, the software controlling charge/discharge would ofcourse be in the hands of the car owner, he would be able to control how much he wants to sell if the car is at home during peak energy demand depending on his need to use the car, and know how much this sale would net him in profit. I doubt your Tesla bullying the utility scenario is very likely, and frankly I believe it would be illegal.

I guess whether it makes sense or not to use the car battery to sell back energy relies on how much it deterioration it causes.

 

[austinEV]

I think we are forgetting that if you could sell power back to the grid lots of people would tent their fingers together and plan to charge at superchargers and then go sell the power to the grid. Or, charge at work and do the same. Unless you have a solution for that it is an uncomfortable situation.

 

[bonaire]

AustinEV, I read somewhere else today that Tesla may be limiting the charge speed of owners at superchargers near their homes. I do not know if this is true or not but it is a lot of work for someone to shuttle their car back and forth to earn a few bucks a day and the car must be at home during peak times to do so. May look good economically but only helps a select few.

the issue of selling power to the grid actually only works economically with a wide spread between high daytime and low night time rate plans. In my state, no TOU plan exists for homeowners due to lack of smart meters. A few business rate plans offer about a 2-3 cent spread and that is not enough. California is currently the target market and Tesla is its own first customer because superchargers are the first good location for something like a 200kWh battery system to curb peak demands when multiple cars are charging during peak demand periods of later afternoons. California appears to have the wide spreads because of the limited amounts of power producers and the large number of electricity users.