Tesla Powerwall: Engineering Analysis

[russ_watters]

Anorlunda, we are talking about two different things. Yes, I agree that the price to sell the power at generally has to be above the highest price to generate the power (unless there is an unfixable surplus) and that changes throughout the day. But the price to generate the power does not change throughout the day and that's what I was asking about above. The shortage during the peak increases the profits for all plants because the average selling price goes way up (4x per Tesla) while the price to make it doesn't go up as much. The price to make it is what I'm trying to determine.

I haven't gone through the whole analysis though, so I don't know the actual economic case: that's why I'm asking for clarification of the input data.

 

[anorlunda]

Now, that last one implies a case that might favor your argument: let's say the utility is considering installing a new peaking plant that runs at .25 kW for 4 hours, 100 days a year. Or they can buy a 1 kWh battery from Tesla and use their already existing peaking plants to charge it at night.

You're still working under the assumption that it is the utility that owns the generation facilities. In many parts of the country, that is no longer true. Each generating facility is a business unto itself. The econmic decision is whether to invest at all, or invest in asset A, or asset B. If your business strategy is to buy energy when it is cheap, and sell it when it is expensive, you must deal with the fact that you can sell power to the grid at wholesale prices, (say $0.03 kWh) but unless you are a public utility, you must buy it at retail prices (say $.15 kWh). The net metering rules that allow solar panel owners to sell back at retail, don't apply to batteries.

Anorlunda, we are talking about two different things. Yes, I agree that the price to sell the power at generally has to be above the highest price to generate the power (unless there is an unfixable surplus) and that changes throughout the day. But the price to generate the power does not change throughout the day and that's what I was asking about above. The shortage during the peak increases the profits for all plants because the average selling price goes way up (4x per Tesla) while the price to make it doesn't go up as much. The price to make it is what I'm trying to determine.

If your only asset is a battery bank, then your cost is the capital cost divided by expected lifetime, add to that the fixed cost of meeting the grid's interconnection requirements to sell wholesale (for example, a transformer to boost the voltage to 115 KV, plus the costs for computers and interfaces to allow your batteries to be controlled by the grid operator, plus the engineering studies required to certify your design as safe and reliable), your variable costs are the cost to purchase the energy to charge the battery. How that energy was generated is irrelevant to you, only the price matters. As I said above, you'll have to pay retail price to buy that energy.

Even if you own another asset, such as a wind farm, your cost to charge your batteries is the "opportunity cost" lost by not selling the energy to the grid. Let's say you own a wind farm that can generate for zero incremental cost. It can charge your batteries, or it could sell the energy to the grid at night for $10/MWh. The "cost" to charge your battery is then $10/MWh, not zero. Or maybe you own a peaking gas turbine, that generates incrementally at $40/MWh. It would be foolish to run that at night for $40 when you could buy the energy from the grid for $10. So no matter what you own or don't own, your costs to charge the battery will be the grid's marginal price at that time of day.

I understand the logic you're trying to apply Russ. I'm trying to tell you that you have an unrealisticly overly simplified view of how it really works. The coplexities completely overwhelm the simple logic.

 

[anorlunda]

Or maybe you own a peaking gas turbine, that generates incrementally at $40/MWh. It would be foolish to run that at night for $40 when you could buy the energy from the grid for $10.

Whoops, I misspoke there. The complexities overwhelm me too at times. In the secenario above, you would be forced to buy retail at $150 rather than wholesale at $10.

 

[rollingstein]

@anorlunda Is it really true that there are today parts of Southern California where the retail day versus night price differs by 20 to 25 cents? I remember reading that somewhere.

If they really have such a large differential then battery options start making sense I guess for such niche markets. Though I'm not sure if Li makes sense or just lead acid.

I just find it mind boggling why such a large peak-non-peak differential can persist. At those margins won't more peaking stations make sense?

The battery was installed to allow the utility up to 20 minutes to do maintenance on the line without blacking out the city.

Intuitively I'm surprised that a battery was cheaper for this large application instead of a diesel / gas generator.

You're still working under the assumption that it is the utility that owns the generation facilities.

I was reading up on PG&E and it seems like in addition to electric distribution they also own several nuclear, coal & hydroelectric generation plants.

So can their distribution business not use the generation business' power directly? Must the PG&E Generation sell to the grid & the PG&E distribution buy from the grid?

How do things work when a utility also owns its own generation capacity?

 

[anorlunda]

How do things work when a utility also owns its own generation capacity?

My first hand knowledge is from New York State. There, the utilities are forbidden to own generation. In California, the rules of CALISO apply. I can't speak to that.

Remember also that the electrons flow though every possible path in the entire interconnected grid, which is owned by many parties. CA belongs to the interconnection that goes all the way to Colorado. You can't use that grid without following the rules of the grid operator, even if you own the source, the destination, and the most direct path between them. That's logical because what you do affects everyone.

But in New York, there was a case where power plants draw from the grid about 1% of the power they inject to operate their "house power". I.e. internal pumps, AC and lights (and let's inject charge their batteries) When the plant is shut down, it still draws house power from the grid. The ruling was that they had to buy that "house" power from the grid at retail prices. The "net metering" rules in CA that allow homeowners to buy and sell back at the same price don't apply except for homeowner solar/wind.

 

[dschlink]

Russ - my cabin is on time-of-use payment with PGE; this is particularly valuable for me as it has a heatpump and in the winter when I'm not there, I have the thermostat programmed to raise the temperature at night when I'm paying about 1/3 of the regular rate.

I could see charging the battery at night and using it to reduce daytime costs; but for power backup, I'd need 4KW or two units.

 

[nitsuj]

The problem for now is that there is no residential market that I'm aware of for this concept. In order to make it useful, your electric rate would have to vary from day to night (as it already does commercially). So this is a product without a market (except for the secondary purpose of being a whole house backup).

Ontario, Canada has "smart meters" for time of day costing, there is a price differential though I'd be surprised if the unit could pay for itself before the battery capacity becomes nill.

Also I haven't seen the very important value for it's wattage/amperage capabilities. Powering an oven/dryer maybe out of the question for a single unit, even a microwave takes a lot of wattage.

 

[rollingstein]

Also I haven't seen the very important value for it's wattage/amperage capabilities. Powering an oven/dryer maybe out of the question for a single unit, even a microwave takes a lot of wattage.

5.8 amp nominal, 8.6 amp peak output

Remember also that the electrons flow though every possible path in the entire interconnected grid, which is owned by many parties. CA belongs to the interconnection that goes all the way to Colorado. You can't use that grid without following the rules of the grid operator, even if you own the source, the destination, and the most direct path between them. That's logical because what you do affects everyone.

Have you seen situations where someone had to make an unscheduled shutdown of a baseload power station e.g. Coal just because demand dropped to a level so low that even after shedding all the peaking plants, generation exceeded the ability of the grid to absorb power?

 

[OmCheeto]

Ontario, Canada has "smart meters" for time of day costing, there is a price differential though I'd be surprised if the unit could pay for itself before the battery capacity becomes nill.

Also I haven't seen the very important value for it's wattage/amperage capabilities. Powering an oven/dryer maybe out of the question for a single unit, even a microwave takes a lot of wattage.

According to the U.S. Energy Information Administration:

In 2013, U.S. electric utilities had 51,924,502 advanced (smart) metering infrastructure (AMI) installations. About 89% were residential customer installations.

Residential  Commercial  Industrial   Transportation             Total
46,083,727   5,597,147     242,783              845        51,924,502

ref

Though this doesn't guarantee that "Time of Use" is available to all customers, but only that it could be.

It is available through my utility company:

https://www.portlandgeneral.com/residential/your_account/billing_payment/time_of_use/pricing.aspx

Peak   $/kWh
On     $0.12581
Mid    $0.07222
Off    $0.04195

Just eyeballing it, it looks as though it would be cheaper to be a mushroom.
hmmm...

My usage:

1000 kwh/month average = $126/month on-peak & $42/month off-peak = $84/month savings
$84/mo * 12mo/yr = $1000
Cost of Tesla battery = $3500
ROI = 3.5 years.​

hmmmm...

 

[rollingstein]

My usage:
1000 kwh/month average = $126/month on-peak & $42/month off-peak = $84/month savings
$84/mo * 12mo/yr = $1000
Cost of Tesla battery = $3500
ROI = 3.5 years.

1000 kWh/mo is approx. 40 kWh / day assuming a pretty invariant usage pattern.

That sounds like you'd need four Powerwalls? $14,000?

 

[nitsuj]

1000 kWh/mo is approx. 40 kWh / day assuming a pretty invariant usage pattern.

That sounds like you'd need four Powerwalls? $14,000?

What's more Om's ROI takes 3.5yrs to reach cost...what's the battery life expectancy after 3.5 years of daily use?...will it catch fire?

 

[OmCheeto]

1000 kWh/mo is approx. 40 kWh / day assuming a pretty invariant usage pattern.

That sounds like you'd need four Powerwalls? $14,000?

Ah! You are correct. So that would be 14 years for ROI, which is longer than the batteries last, so this would not be a good investment for me.
Unless of course, I switch to another form of heating.
My summer usage is only 14 kwh/day, so I could probably get by with 1 unit.

Wait a second. No, I don't need 4 units. I only need enough units to cover "On-peak" rates.
Good grief. This is worse than figuring out taxes.

This may take me a while to figure out...

 

[CWatters]

Here in the UK... Electricity consumption per head is lower than in the USA. Many people with PV generate more than they use at certain times of the day (or even when averaged over the year). The subsidy/financial incentive they give you to install PV assumes you will export a fixed percentage of the electricity you generate even if you use it all yourself. So it's attractive to store and use all that you generate rather than export any.

It doesn't normally make sense to use electricity to heat water (gas and oil are cheaper) but some people install an extra heating element in their hot water tanks and use this to store the excess electricity from their PV system. There are devices you can get that detect you are exporting and turn on and modulate the heating element.

I think some people will also use Tesla batteries to store the excess electricity they generate in the day to charge their electric car at night.

One question I have... Tesla are offering a 10 year warranty which seems generous given that many cells are only rated for around 1000-2000 cycles?

 

[nitsuj]

5.8 amp nominal, 8.6 amp peak output

less than half of a standard outlet /cry presuming normal voltage, this thing can can't even produce just one horsepower (ideal use)

I think they maybe really sacrificing available wattage for battery life / safety. The battery for my remote control car toy (1.5kg) has an 11.1 lithium polymer that can do 1400 watts all day and stay only "warm". though would be lucky to get 500 uses and still have >80% capacity.

One question I have... Tesla are offering a 10 year warranty which seems generous given that many cells are only rated for around 1000-2000 cycles?

It would be interesting to have more info on the batteries. I don't get why this is suddenly a viable tech, which makes me think marketing rather reduced cost / improved tech. Has Tesla come up with new battery formulations, or reach new "economies of scale" with their batteries?

When will Tesla sell batteries for my remote control car toys? lol I want 10 year warranty with 3000 cycles."It's[/PLAIN] like crazy off-the-hook," Musk said during an earnings presentation on Wednesday. "The sheer volume of demand here is just staggering."

Anyone can go online and place a reservation, years in advance, with no money down and no commitment to buy."

Oh Musk, you're the king of optimizing investors with your optimism to get them investing or to stop asking about some sort return...again & again. lol That said GO MUSK!

If I had "venture dollars" hearing "It's like crazy off-the-hook," would be a bit too much of a ...warning sign..."cause like trendy isn't for things that cost thousands, so like I so couldn't"

 

[rollingstein]

It would be interesting to have more info on the batteries. I don't get why this is suddenly a viable tech, which makes me think marketing rather reduced cost / improved tech. Has Tesla come up with new battery formulations, or reach new "economies of scale" with their batteries?

If you look at their Powerwall website, other than whatever details are on that one page there's no other Literature. No spec. sheets, no test results, no whitepapers, no pdf's, no best practices, installation instructions. Nothing. Zilch.

To me that's a big red flag. It has the classic signs of a "magic" technological leap.

 

[nitsuj]

If you look at their Powerwall website, other than whatever details are on that one page there's no other Literature. No spec. sheets, no test results, no whitepapers, no pdf's, no best practices, installation instructions. Nothing. Zilch.

To me that's a big red flag. It has the classic signs of a "magic" technological leap.

Ah..so I'll use it to charge my Segway.

The magical technological leap is upon purchase it makes you feel good being a solution to an energy "problem".

What's more Musk backers see shimmers of hope when Musk adds a Musk multiplier to that sale preventing a balance sheet disaster. So there is some "magic" there too. take a sale of $3,000...multiply it by the number from an internet poll...report to media "The sheer demand here is just staggering!" & "...off the hook!".

What's the carbon foot print of one of these units?

 

[OmCheeto]

If you look at their Powerwall website, other than whatever details are on that one page there's no other Literature. No spec. sheets, no test results, no whitepapers, no pdf's, no best practices, installation instructions. Nothing. Zilch.

To me that's a big red flag. It has the classic signs of a "magic" technological leap.

They listed specs near the bottom of the page. I'll list the clues as to why I believe a 10 year lifespan is doable:

Specs

Technology

Wall mounted, rechargeable lithium ion battery with liquid thermal control.​

Models

10 kWh​

Power

2.0 kW continuous
​

I've never done testing on lithium ion batteries, so I can only take other people's words for how they function:

Understanding the life of lithium ion batteries ...

Note to journalists: Please report that this research was presented at a meeting of the American Chemical Society. (Yay! I think I'll add "journalist" to my resume.)

NEW ORLEANS, April 10, 2013 — Scientists today answered a question that worries millions of owners and potential owners of electric and hybrid vehicles using lithium-ion batteries: How long before the battery pack dies, leaving a sticker-shock bill for a fresh pack or a car ready for the junk heap? Their answer, presented here at the 245th National Meeting & Exposition of the American Chemical Society (ACS), being held here this week, may surprise skeptics.

“The battery pack could be used during a quite reasonable period of time ranging from 5 to 20 years depending on many factors,” said Mikael G. Cugnet, Ph.D., who spoke on the topic. “That’s good news when you consider that some estimates put the average life expectancy of a new car at about eight years.”

Cugnet explained that the lifespan depends mainly on the battery’s temperature, state of charge and charge protocol. Battery performance begins to suffer as soon as the temperature climbs above 86 degrees Fahrenheit. “The higher the temperature, the lower the battery service life,” he said. “A temperature above 86 degrees F affects the battery pack performance instantly and even permanently if it lasts many months like in Middle East countries.”
​

This may take me a while to figure out...

This is worse than taxes and orbital mechanics, combined.

Winter and Summer consumption are best (June-Oct 14 kwh/day) and worst (Jan & Feb 50 kwh/day) cases.

Graphically, over my last 12 recorded billing cycles:

x = date y = kwh/day​

About the only things I've determined so far, is that:

1. a single Tesla Powerwall

a. will keep my refrigerator running during "On" and "Mid" peak hours during the summer (9.3 kwh)
b. will not cover my winter needs (33.3 kwh)​

2. I should do my laundry and dishwashing on Sundays (my water heater is set to 100°F, so I use the "Water Heat" mode on my dishwasher)​

So based on 1.b., I will have to determine a method of storing 23.3 kwh of energy via an alternative method, to make this "Wall" a viable option, for me.

Interesting that Tesla got nearly a billion dollars worth of orders for these type of things in the first week.

Tesla's Battery Grabbed $800 Million in Its First Week (BloombergBusiness)

 

[rollingstein]

a. will keep my refrigerator running during "On" and "Mid" peak hours during the summer (9.3 kwh)

How many Amps does your refrigerator pull? How many Amps will your Laundry / Dishwasher pull?

 

[OmCheeto]

How many Amps does your refrigerator pull? How many Amps will your Laundry / Dishwasher pull?

simple maths: Amps = Watts/Volts

You figure it out!

clues: I've never heard of a 220 volt home fridge, dishwasher, nor clothes washing machine. I seem to recall that there are 120 volt laundry dryers for apartment dwellers.
Ha! I just realized, that I've never been an "apartment" dweller.

hmmm...

 

[rollingstein]

simple maths: Amps = Watts/Volts

You figure it out!

Did you post the Watts somewhere? Maybe I missed that info.

I just wanted to point out that the Powerwall is rated for 6 Amps. I was wondering if you refrigerator + dishwasher / dryer combined load would be able to stay within those Amps. I didn't think it would work.

 

[OmCheeto]

Did you post the Watts somewhere? Maybe I missed that info.

I just wanted to point out that the Powerwall is rated for 6 Amps. I was wondering if you refrigerator + dishwasher / dryer combined load would be able to stay within those Amps. I didn't think it would work.

I don't know how much you eat, or roll in the mud, but I only have only one appliance that runs 24/7 in the summer.
The others can wait for Sunday.

 

[rollingstein]

I don't know how much you eat, or roll in the mud, but I only have only one appliance that runs 24/7 in the summer.
The others can wait for Sunday.

At least on Sunday you need something rated for the combined Amps?

 

[OmCheeto]

At least on Sunday you need something rated for the combined Amps?

I don't understand what you are asking/saying.

Or perhaps you don't understand my pictograph?

Summer rates start on May 1 ____ Winter rates start on Nov 1

What it means is that I can get off-peak rates all day Sunday. So I can run everything in the house, which is about 20,000 watts.

One funny thing is, that the way I cook, is already optimal for "Time of Use" metering. Ha!
I cook about 5 lbs of meat overnight at fairly low temperatures, and then take it out in the morning.
Then I store this in the fridge for the week, and heat it up in the microwave.

5 minutes of microwave cooking/day @ 500 watts = 42 watt hours
16 hours of refrigeration/day @ 500 watts = 8,000 watt hours
total = 8,042 watt hours

I'm thinking I will only need one of these things.
But I'm still working on the maths on how to heat my house in the winter.
It's been 25 years since I've done some of these calculations.

This is the problem I'm working on:

Outside temperature is 43°F
Specific heat capacity of my house is 5000 Btu/°F
R-value is 13
Conductive surface area is 1900 ft²
If the house is heated to 75°F at 6 am, what will be the temperature in the house, 16 hours later?
And if it's less than 60°F after 16 hours, how much water, heated to 150°F would be required to maintain the minimum temperature at 60*F?​

I should have the solution by noon tomorrow.
But right now, my brain hurts, and I feel like watching some Hulu TV before bed.
Ciao!

 

[nitsuj]

They listed specs near the bottom of the page. I'll list the clues as to why I believe a 10 year lifespan is doable:

Specs
Technology
Wall mounted, rechargeable lithium ion battery with liquid thermal control.Models
10 kWhPower
2.0 kW continuous

How many Amps does your refrigerator pull? How many Amps will your Laundry / Dishwasher pull?

apparently 2,000 watts (household = 120 * 15 = 1800)

10,000 / 2,000 = 5 hours at 2000 watts

I'm more confused now...

 

[CWatters]

If you need more than 2KW...

The extra can be drawn from the grid.
or
If you are off grid multiple Powerwalls can be connected together.

 

[nitsuj]

If you need more than 2KW...

The extra can be drawn from the grid.
or
If you are off grid multiple Powerwalls can be connected together.

If it's seamless ( powerwall <-> grid) that's perfect, if not I think that's more math than most want to do in determining what to power with it to ensure all energy is used before night charging. I think this is a step towards higher production volume of batteries as opposed to a step towards convenience / energy cost savings / meeting market demands.

 

[OmCheeto]

If it's seamless ( powerwall <-> grid) that's perfect, if not I think that's more math than most want to do in determining what to power with it to ensure all energy is used before night charging. I think this is a step towards higher production volume of batteries as opposed to a step towards convenience / energy cost savings / meeting market demands.

I think you are correct, in that this is "more math than most want to do".
It's taken me an hour just to convert from imperial to SI units.
It may be 2 pm, before I have my answer.

...the problem I'm still working on:

Outside temperature is 43°F ( 6.11 °C )
Specific heat capacity of my house is 5000 Btu/°F ( 2.6 kwh / °C )
R-value is 13 ( (2.3 m² °C) / watt )
Conductive surface area is 1900 ft² ( 176 m² )
If the house is heated to 75°F ( 23.9°C ) at 6 am, what will be the temperature in the house, 16 hours later?
And if it's less than 60°F ( 15.6 °C ) after 16 hours, how much water, heated to 150°F ( 65.6°C )would be required to maintain the minimum temperature at 60*F?​

Thank god those silly Europeans haven't yet devised their own SI time unit...

 

[nitsuj]

apparently 2,000 watts (household = 120 * 15 = 1800)

10,000 / 2,000 = 5 hours at 2000 watts

I'm more confused now...

Ah I see now, it's 350-400 volts...presuming for efficiency & reliability reasons (not bad if a few cells are poor).

This unit is getting a bit more impressive...wish there was more info on the batts though.

 

[James Alton]

Ah I see now, it's 350-400 volts...presuming for efficiency & reliability reasons (not bad if a few cells are poor).

This unit is getting a bit more impressive...wish there was more info on the batts though.

7KW of energy storage could be very interesting to people that live in Northern Climates. That is enough power to run an oil furnace or perhaps a wood pellet stove for an extended period of time in case a power loss and make the difference between freezing and not! In the event of an extreme winter with an extended power loss could even save some lives. Generators are another option but the beauty of a charged battery is that you don't have to get it started for it to start producing power and gasoline units especially can be pretty unreliable. I could imagine powering my home for about 3 days using one of these batteries after removing loads that were not critical. Personally if I had to choose between relying on a generator or a charged lithium battery for backup, I would take the battery. And no, a bank of wet cell lead acids of the same capacity would not be a suitable replacement. Besides the corrosive/explosive gases, those batteries self discharge at an unacceptable rate IMO. I think that there will be a market for Elon's batteries, I might even help a little! James

 

[OmCheeto]

Ah I see now, it's 350-400 volts...presuming for efficiency & reliability reasons (not bad if a few cells are poor).

This unit is getting a bit more impressive...wish there was more info on the batts though.

I was assuming Tesla was going to use the same technology as their vehicles. I don't know about the battery, but the cells used are currently made by Panasonic. And they have 19 different versions!

http://na.industrial.panasonic.com/search/18650


ps. interrupting me will only delay the solution to my problem.
pps. and then there's lunch... I'm shooting for 4 pm at the moment.