Plan: Off grid solar with a Model S battery pack at the heart

[Pollux]

"Whole house UPS mode."

There's definitely a commercial market for this by itself. For example, my parents live in Washington, DC, and have told me many times of the notorious unreliability of the DC grid. Many people there would love to have power backup options. My parents investigated a gas-powered generator, but were foiled by gas company restrictions and pricing.

Of course, the issue of total-cost-installed for a "whole house UPS" solution would be of critical importance.

And absent some research, I certainly couldn't guess how *large* this particular market is.

Until my solar panels are installed my system is operating in whole house UPS mode. Works as expected. I cut off my main grid breakers and everything keeps ticking like nothing happened. I cut them back on and back to grid power with the inverters topping off the batteries. Pretty cool.

 

[cashie]

Depends on where you live. In my country we pay about $0.15 in tax for every kWh we draw from the grid. The power we sell back to the grid does not refund this tax. So with a kWh price of $0.30 when purchased and a $0.15 price when sold it becomes interesting to store the energy.

We also tend to live in smaller homes and could conceivably go off-grid with 10-15 kWh of storage.

Also here in Australia we get a very poor tariff for exported power. I currently pay around 25c/kWh and only get 8c/kWh for exported power (supposedly the "wholesale" price of power from the provider).
Basically forces us to try and use every kWh we generate during the day where the generated power save us 25c (as opposed to 8c exported). Pretty poor situation as it promotes energy use as opposed to savings.
Decently priced storage would be awesome here, I could store power all day (lots of sun here) and use power day or night.
Our power use appears a fair bit lower than you guys in the states, I'm currently using about 17kW/h per day to run a large home with a 3-phase electric reverse cycle ducted A/C, a pool, large shed etc.
Just looked at my bill, in the last 4 months (summer), I have exported more power than I used.... :cursing: I need some batteries.

 

[Robert.Boston]

Also here in Australia we get a very poor tariff for exported power. I currently pay around 25c/kWh and only get 8c/kWh for exported power (supposedly the "wholesale" price of power from the provider).

8c/kWh wholesale is very generous. Look at the actual data: AEMO Average Price Tables

The annual average wholesale price of power in, say, New South Wales was 3.5c/kWh. Where you live in Western Australia the average wholesale price in 2014 was 5.13c/kWh. The big gap between 25c and 8c is to pay for all the wires and services that keep your power available almost anytime you want it.

 

[cashie]

Yeah, fair enough, I do understand that. It's just disappointing that there is zero incentive for me to export any of my power, it actually promotes me to use all the generated power that I can during peak load time (when they should be promoting me to boost the grid).

 

[rjcbox]

In my area of U.S., we pay ~20c/kwh, utility company pays 4c/kwh for excess power put back into the grid

 

[none12345]

Love your setup, i only wish i could do something similar. But couldn't afford to.

One concern....

You're battery rack. I like the layout but im concerned about its mechanical load carrying capacity. While it looks fine for a static load. It doesn't look strong enough for an earthquake. Ive used that exact aluminum stock for projects, and i don't think id trust it for 2000 lbs worth of batteries during an earthquake. Not ones that can put out 3000-6000 amps or so per module in a dead short situation. I don't know how common earthquakes are where you live, but where i do, i wouldn't trust it at all. You may want to consider some reinforcing. I am not a structural engineer, so this is just coming from working with that material and a gut feeling.

 

[Lessmog]

Love your setup, i only wish i could do something similar. But couldn't afford to.

One concern....

You're battery rack. I like the layout but im concerned about its mechanical load carrying capacity. While it looks fine for a static load. It doesn't look strong enough for an earthquake. Ive used that exact aluminum stock for projects, and i don't think id trust it for 2000 lbs worth of batteries during an earthquake. Not ones that can put out 3000-6000 amps or so per module in a dead short situation. I don't know how common earthquakes are where you live, but where i do, i wouldn't trust it at all. You may want to consider some reinforcing. I am not a structural engineer, so this is just coming from working with that material and a gut feeling.

I can sympathise with that quaking horror. How about some cushioning? I hear spring is in the air, so maybe a bit of air suspension it the ticket? As long as it doesn't spring a leak ... :wink:

 

[wk057]

I had thought about the load issue with the rack. Earthquakes are super uncommon here, so, shouldn't be an issue. However, I've added all of the bracing the floor and walls for extra safety. The modules themselves, if say they were all of a sudden to collapse and the terminals somehow short out. The cell level fuses would certainly all fail in a dead short situation. With the ridiculously tough plastic between the cells and the aluminum bus bars I would expect a short wouldn't get too far, beyond destroying the module(s) itself that is.

It's pretty sturdy, even without the wall bracing. Before loading the modules I climbed it like a ladder and was pressing against walls and such trying to break it. No such luck.

In any case, I'm going to keep an eye on it for a while. If it seems anything is having problems with the load I'll just have a better rack professionally built for the task and install it.

 

[Tedkidd]

Looks like it may also be strapped to the ceiling? Really is a thing of beauty.

NY is going through "Reforming the Energy Vision" - Public Service Commission wants to bring us into the 21st century with distributed resources. They are very interested in creating a market where all parties get fair reward for the value they bring, be it energy, capacity, efficiency, etc...

What we don't see is when peak events occur utilities have to pay a very high price for energy (50¢ wholesale?), and in cases where they can' t get enough they institute rolling brown/blackouts. Distributed capacity increasing supply could benefit them greatly in terms of cutting costs and improving service.

Expect the day to come where you set a price on your backup energy, and when that price hits you start delivering it to the grid. I suspect with Elon's packs we could all be charging our home packs off peak for 3¢ and selling back at 15¢ on peak.

 

[oneday]

Since its in "whole house UPS mode," does that mean you are pulling current from the battery and the utility is keeping you at constant SoC? Does this put any negligible strain on the battery and what's the ideal SoC for this situation?

 

[electrodacus]

A more simple calculation related to the cost of energy storage using Tesla battery.
Tesla S has a 200000km warranty on the 60kWh battery (I do not think there is any warranty to the capacity loss just battery defects) based on wikipedia page.
Range seems to be around 350km so 200000km / 350km = 571 full charge cycles (of course there will be more smaller charge cycles in real life).
571 x 60kWh = 34260kWh over the life (warranty of the battery)

Say is about $300/kWh the 60kWh battery will cost $18000 then cost of energy stored during the life of the battery (life based on warranty probably at 80% original capacity best case)
$18000 / 34260kWh = $0.52 / kWh

Then there is this story not sure how real Tesla's residential battery system leaked, here's everything we know - TechSpot
Something about a 10 and 15kWh battery for a cost of maybe $13000 but there is some 50% ? rebate.
Then something about $1500 down and $15 / month for 10 years (something wrong here $15 x 120 months = 1800) no where near the $13000 or even half that.
To make things worse he makes monthly savings of $10 to $12

Based on first calculation battery should be at $300 / kWh so $3000 to $4500 for battery only there must be a relatively large inverter in there and some additional control logic so maybe another $2000 or more
Then $13000 with 50% rebate will look more likely as the real price of the unit $6500 not sure if for 10kWh or 15kWh version
Say is for the 15kWh unit and the same battery they use on the car then since 15kWh is a quarter of that 34260kWh / 4 = 8565kWh then

$6500 / 8565kWh = $0.76 / kWh

Are there any mistakes in my calculations ?

 

[Ampster]

You didn't calculate the cost savings of peak shaving or load shifting. That would depend on a TOU rate and the size of the delta between peak and off peak rates.
For example, if the delta was $0.25/kWh and I could shift 10kWhrs per day that would save me $900 per year. I could probably do that with a 15kWh battery costing $4500 plus inverter.
That would be a 7 year payback.

 

[wk057]

A more simple calculation related to the cost of energy storage using Tesla battery.
Tesla S has a 200000km warranty on the 60kWh battery (I do not think there is any warranty to the capacity loss just battery defects) based on wikipedia page.
Range seems to be around 350km so 200000km / 350km = 571 full charge cycles (of course there will be more smaller charge cycles in real life).
571 x 60kWh = 34260kWh over the life (warranty of the battery)

Say is about $300/kWh the 60kWh battery will cost $18000 then cost of energy stored during the life of the battery (life based on warranty probably at 80% original capacity best case)
$18000 / 34260kWh = $0.52 / kWh

Then there is this story not sure how real Tesla's residential battery system leaked, here's everything we know - TechSpot
Something about a 10 and 15kWh battery for a cost of maybe $13000 but there is some 50% ? rebate.
Then something about $1500 down and $15 / month for 10 years (something wrong here $15 x 120 months = 1800) no where near the $13000 or even half that.
To make things worse he makes monthly savings of $10 to $12

Based on first calculation battery should be at $300 / kWh so $3000 to $4500 for battery only there must be a relatively large inverter in there and some additional control logic so maybe another $2000 or more
Then $13000 with 50% rebate will look more likely as the real price of the unit $6500 not sure if for 10kWh or 15kWh version
Say is for the 15kWh unit and the same battery they use on the car then since 15kWh is a quarter of that 34260kWh / 4 = 8565kWh then

$6500 / 8565kWh = $0.76 / kWh

Are there any mistakes in my calculations ?

Where do you get these numbers sometimes? lol. First, I haven't seen a "street price" for a 60kWh pack yet. But I have actually bought 2.375x 85kWh packs. My cost was actually just north of $200/kWh in total including all of the hardware needed to put the battery bank itself together (wiring, fuses, rack, etc). Good luck finding a price per kWh much better than that for equally good tech, especially installed price of the battery bank.

The inverters are an independent cost and do not factor into battery life or lifetime cost per kWh pulled from the battery with regard to the battery itself. For the total system, sure, but your unfounded arguments have simply been against Tesla's cells anyway.

 

[FlasherZ]

Are there any mistakes in my calculations ?

Using the warranted life is never a good starting point.

I'm sure the original White House solar PV system was warranted only for a decade or so, yet here they are 38 years later, still producing energy for Unity College.

 

[electrodacus]

Is still hard for some of you to understand my calculations even if I made them reasonably easy.
Battery cost per capacity as wk mentioned $200/kWh for a used battery has no meaning.
Say that battery was good for only 10 cycles will you buy that ?
The important aspect is cost for each kWh stored during the life of the battery.
Batteries degrade and the Tesla battery was selected for energy density in the detriment of cycle life. There are better batteries for stationary energy storage and those will not be great for cars because of much lower energy density.
An EV car manufacturer wants to offer the best warranty that he can based on projected typical use of the car.
If you look at all commercially available EV cars, BMW, Kia, .... all offer around the same 7 to 8 years warranty on battery and all use about the same Lithium-ion batteries optimised for range so good energy density. Also most have a smaller pack than Tesla and about half the range so they offer 100000km warranty. Those others usually specify that 80% of original range should still be available in warranty. Tesla does not make any claims abut capacity but I assume they can do that if they want.

Anyway LiFePO4 will usually last 5 to 10x more cycles compared to Lithium Cobalt based batteries used in most electric cars so they are more suited for stationary energy storage. They are also greener and safer no thermal runaway.
There are already solution on market for energy storage even if is mostly sold in Europe for quite a few years.
Here is an example (sorry the page is in german) but you can see the numbers 6535 Euro for 4.8kWh storage capacity based on LiFePO4 (LFP) and the important part up to 20 years life and 8000 cycles (this are full 100% DOD with still 74% of original capacity after this)
Sony fORTELION Stromspeicher 4.8 kWh / Olivine LiFePO4 Technologie (LFP) (Eigenstromnutzung SONY fORTELION Technologie) - solarshop.net: Solarmodule, Solarstromanlagen, Photovoltaikanlagen günstig online kaufen

And I do not advertise Sony or any other LiFePO4 manufacturer I have no interest in this.
I designed a Solar BMS charger that just got full funding on Kickstarter but that can work the same with LiFePO4 or any other type of Lithium-ion including Tesla battery so I have no reason to say a certain type of Lithium is better.
When it comes to grid tie I think not even the less expensive LiFePO4 will make sense at this point in most circumstances.

You didn't calculate the cost savings of peak shaving or load shifting. That would depend on a TOU rate and the size of the delta between peak and off peak rates.
For example, if the delta was $0.25/kWh and I could shift 10kWhrs per day that would save me $900 per year. I could probably do that with a 15kWh battery costing $4500 plus inverter.
That would be a 7 year payback.

If you shift 10kWh/day from that 10kWh battery you will get to end of life within one to two years. I'm sure they will have a stipulation in the warranty not only about time but also kWh used same as for cars you get time + max km.

 

[Robert.Boston]

The $200/kWh matters because this thread is about @wk057's installation, and that's what he paid. Feel free to start a thread to discuss your ideas for a different system, but don't hijack his.

 

[Tedkidd]

Understanding what it might cost me for a pack is interesting and warrants its own thread. But its not relevant to this project.

Anyone else finding it distractingly off topic?

Would it be possible for a mod carve out this discussion of small residential packs and tiny off grid packs from this discussion wk started?

 

[ItsNotAboutTheMoney]

6535/4.8 ~= 1361! That is ridiculously expensive.

 

[oneday]

Understanding what it might cost me for a pack is interesting and warrants its own thread. But its not relevant to this project.

Anyone else finding it distractingly off topic?

Would it be possible for a mod carve out this discussion of small residential packs and tiny off grid packs from this discussion wk started?

+1
Any mention of lifepo4 should go to its own thread.

 

[wk057]

Have to admire this guy's persistence, I suppose. He seems to have joined the forum for the sole purpose of arguing against my setup. lol.

In any case, the argument using EV battery warranties is interesting. I love how you somehow know that the batteries will reach "end of life" based on this data.

<WARNING - COMPLETELY USELESS DATA FOLLOWS>
DO NOT REFERENCE THE FOLLOWING DATA AS IT IS COMPLETELY MEANINGLESS. DO NOT READ IF USELESS DATA MAKES YOU ANGRY. lol.

Ok, so, Let's use Tesla's 8 year/unlimited mile warranty on the 85 kWh pack as an example. 8 years @ average mileage per year of 12500 miles would be 100000 miles. That's at least ~450 discharge/charge cycles using realistic numbers. ~375 using rated miles, ~350 if you use 0.3kWh per mile and 85 kWh as the pack usage. We'll go with 350 since that's the best case in favor of your arguments against the tech here.

Now, in 8 years/100000 miles the battery will in no way be at "end of life." Best estimates in an EV usage scenario based on data from this very forum puts the worst case at about 20% degradation at this point.

An unrealistic over estimate for my system will be a 50% cycle per day (in reality will be much less since night time/no sun loads will dwarf daytime loads by a huge factor).

We won't even consider that the charging/discharge rates will be minuscule compared to demands placed on the pack by an EV.

So, 350 cycles would take ~2 years. Then, after two years using those numbers I'd still have 80% of the pack available for use.

Let's also say, unrealistically, that this figure is linear (even though it's common knowledge that the majority of degradation happens in the first portion of the cycles and slows afterward). 2 years for 20% at that rate in 10 years the batteries will hold 0kWh. (lol)

Based on this completely contrived and unrealistic super worst case scenario that is actually pretty impossible/improbable, 1750 cycles = no more storage. Using that, we can average 0.05714% capacity loss per cycle.

Using that we get about 74417kWh worth of lifetime capacity usage. Let's say this 85kWh pack costs $25,000. $25,000 / 74417 kWh = $0.3359/kWh.

So, even using some completely contrived and unrealistic super worst case scenario that is actually pretty impossible/improbable I still come up with a useless price per kWh that is a) way higher than real world and b) still cheaper than electricity in Hawaii or on-peak many places making things like ToU grid arbitrage practical even at this inflated cost.

lol

</COMPLETELY USELESS DATA>

Worth repeating: Do not use this data for anything. It is completely useless and based on contrived super worst case scenarios that have no basis in reality.

I just figured I would join in on the throwing around of pointless calculated figures yet still bolster my point even with the ridiculous figures.

I personally expect a *significantly* longer life out of my storage batteries as well as a *significantly* lower overall lifetime cost per kWh than described above, based on all of the real world data available currently (Tesla EV usage data and some published studies of LiCo chemistry and cycle life), all of which is much harder on the cell chemistry than my setup will ever be.


Edit: Mods: I somewhat agree that most of the posts by 'electrodacus' and subsequent replies by others and myself relating to those posts in this thread are mostly off topic here. I leave it up you you guys on whether or not it warrants carving to a 'Lithium Iron Phosphate (LiFePo4) vs Lithium Cobalt variants (LiCo*) for Stationary Storage' type spin-off thread. I have no objection either way.