DIY Powerwall

[Batty]

I live in Perth, Australia where the sun is always shining! I have a rooftop solar system with 6.5kW of panels feeding the house via a 5kW 3phase Fronius inverter. I export surplus power to the grid and get paid 7c/kWh for this, versus 30c/kwh for buying grid power.

My mate has recently installed a Tesla powerwall which means he can store power instead of exporting it. However, the cost of the power wall has increased dramatically (about AUD$16,000 plus installation) and hence it cannot be justified on economics.

I was wondering how easy it is to build my own powerwall. I have already have some experience having recently updated an electric motorcycle to 48V LiFePO4 power - increasing the range to about 150km per charge. I incorporated a 150A BMS.

I am pretty familiar with the battery and BMS side of things. I am thinking of using 16 x 3.2V x 205Ah LiFePo4 cells giving a maximum theoretical capacity of 10.5kWh. The charging side is presumably straight forward, I could hopefully use my existing 48V 25A charger for this purpose, giving a charge time of about 8 hours during the sunny part of the day 8.00-16.00.

I am also familiar with BMS functionality and battery protectors, etc.

I will also need a very large inverter. However, I notice that a Tesla powerwall only has a 5 kW inverter, which means that it is not able to power any where near as much as the grid can (3 phases x 240V x 23A each = 16.5kW). However, I guess it is possible to make do with about 5kW.

The other issue is how it would be connected to the house power and whether it would be allowed under the Australian regulations. Presumably I would have to disconnect from the grid and connect to the DIY powerwall instead?

Currently I have an old 48V 16 cell LifePO4 battery pack with a capacity of about 3kWh powering a 3000W inverter. It has a 100A BMS fitted. In power outages, I simply connect the TV, coffee machine, computer, etc to the inverter via an extension lead and power board. Simple!

Is it worth trying to build my own powerwall/

Cheers

 

[chrisski]

I have been thinking of a similar project. My idea would 16 x 280 ah cells, which comes to sa,e capacity as a Tesla Powerwall. I’m considering a 6 kw grid assist inverter. The only caveat to that idea is if I could do that to avoid permitting issues.

This project with a homemade battery in a grid tied house is probably not going to be legal, and that would kill the build.

 

[Batty]

Yes - 280Ah even better! Thats about 13.5kWh.

 

[Tim Tim]

Hi Batty I think @wattmatters will be able to help you regarding the regulations.

I can’t understand why people install Tesla Powerwalls when you can just install the largest solar system possible and sell back to the grid to offset the cost of any electricity you use from the grid. It would take a very long time for me to recoup the outlay with my quarterly bill being $62 for winter and $50 for summer.

 

[Batty]

Yes, its hard to justify batteries on economics. However, our government owned power utility have just lowered the buy-back to 2.5c kWh for new installations claiming that they have too much solar generated power between 9am to 4pm and its overloading the grid!!! I suggested they might consider reducing the cost of grid power by 50% during this period to increase demand for the consumers that do not have solar, but I doubt common sense will prevail!

The government are also planning huge battery farms like the one in S Australia. However, surely it would be more economical to subsidise individual home batteries instead, since it requires no additional real estate and expensive wiring connections.

Most people I know who have powerwalls also have electric cars and tend to be electric evangelists.

However, I am still interested in building my own battery as I am a retired engineer looking for things to waste my time on!

 

[chrisski]

Since you’re getting no Austrailian response, I recommend looking at this guy‘s YouTube channel since he is in Australia

https://youtube.com/@OffGridGarageAustralia?feature=sharese

 

[wattmatters]

I will also need a very large inverter. However, I notice that a Tesla powerwall only has a 5 kW inverter, which means that it is not able to power any where near as much as the grid can (3 phases x 240V x 23A each = 16.5kW). However, I guess it is possible to make do with about 5kW.

The Tesla Powerwall, like your Fronius PV inverter, operates in parallel with the grid supply - the supplies are in sync. So if the Powerwall's inverter is maxed out, the grid supplies the balance of power demand.

This is why you can have smaller inverters with systems which operate in parallel with the grid, because by and large most power consumption in a typical home will be under 5 kW, so a Powerwall battery will be able to cover it. But when demand peaks (e.g. turn on the kettle to make a cuppa with the aircon going while the electric stove is cooking dinner), then it's no big deal, the grid supplies the balance beyond what the Powerwall can. In this way most of the total energy is supplied by the battery but the grid helps out when needed.

It's also why the Gateway wiring for when the Powerwall is operating during grid outages is restricted to a few limited essential circuits so as to reduce the chance of an overload when there is no grid to cover those peak loads. Often high power circuits such as ovens, aircon, etc will be excluded from a Powerwall's backup circuits

With an off-grid set up, the power demand you can supply is limited to the output capacity of the inverter. They can pass though grid power but the output is limited to the rating of the inverter. Hence if you expect to power your whole home with an off-grid set up, by necessity it may well need to be of larger output capacity inverter than your Fronius (or a Tesla Powerwall). You can of course choose to only run selected circuits via an off-grid set up. That's what I have done.

 

[Batty]

Many thanks for the learned insights.

I have done a few calculations (all in Australian Dollars (AUD) which is with about 67% of a USD.

My current power bill is about $860, made up of $975 for power (@30c/kWh) plus $365 standing charge, minus $480 credit for exported solar power (@7.14c/kWh).

The installed price of a Powerwall is currently $19,000. This would enable me to use stored power instead of grid power saving $741 a year (allowing for the reduced export rebate). So payback period would be 26 years! The break-even point would be if I could make a DIY Powerwall for $741 x 10 = $7410, assuming a 10 year life. Hardly worth the effort!

 

[wattmatters]

So payback period would be 26 years!

Yep, that's the sad reality of professionally installed home battery systems in Australia. They just cost too much to make much sense. And the reality of batteries like this is you will never really eliminate all grid imports (e.g. not every day will have the PV generation to fully charge a battery).

Which is why I never bothered with it and focussed on load shifting instead (e.g. hot water storage using solar PV to heat).

Meanwhile grid-tied PV is cheap.

I also build a small off-grid backup power system (we get regular grid outages) using sealed lead acid batteries designed for just this purpose and a small all-in-one (AIO) inverter, plus a small solar PV array to keep the batteries topped up and capable of powering the home via a transfer switch.

But I soon realised it was capable of doing much more than backup duties, so I added server rack LiFePO4 battery capacity so it could do daily energy cycling, which worked a treat so I upgraded the inverter to a higher capacity model and it's now managing our daily energy needs really well.

It even makes financial sense, just. Mainly because the backup power system was a sunk cost so it was mostly just the cost of the added battery (A$4k for 10 kWh). It doesn't eliminate our imports, as it doesn't cover our ducted aircon system nor our oven, but it powers most other things.

In the first four months it has reduced our bills by an average of ~$1.65/day compared with previous same periods last couple of years. But that's over Spring/early Summer, so will have to wait for a full year to get a better take on its impact. Even $1.50/day is passable. $4k / $1.50/day ~= 7 years.

 

[Batty]

Yes - I also use my solar to heat hot water (on a timer from 8 am to 4 pm). I bought a secondhand 125litre electric storage heater for $50 and changed the heating element to a lower wattage (1.8kW) to spread the load.

I also heat a hot tub with solar, this has a timer (8am to 4pm) plus a sensor fitted to the output from the solar inverter so as the spa is only heated when there is solar available. We don't heat it at all in the winter since we have only about 50% (20kWh) available versus a maximum of 40 kWh in the summer.

 

[sunny-spark]

Many thanks for the learned insights.

I have done a few calculations (all in Australian Dollars (AUD) which is with about 67% of a USD.

My current power bill is about $860, made up of $975 for power (@30c/kWh) plus $365 standing charge, minus $480 credit for exported solar power (@7.14c/kWh).

The installed price of a Powerwall is currently $19,000. This would enable me to use stored power instead of grid power saving $741 a year (allowing for the reduced export rebate). So payback period would be 26 years! The break-even point would be if I could make a DIY Powerwall for $741 x 10 = $7410, assuming a 10 year life. Hardly worth the effort!

Wow - our power is cheap here (10-12 USA-cents / kWh), so while your powerwall is more expensive, it's actually relatively more valuable given the difference in grid rates. What kills me here, though, is that our utility is going to eventually win on reducing the "surplus rate" as you call it (eg net metering) especially withOUT doing anything about time-of-use cost (which is level cost 24/265 today). In my opinion, since the stated problem is clearly claimed to be a mismatch between generation (solar peak at noon) and load (AC peak at dinner time etc), they should be addressing that with time-of-use first. But the regulatory agency apparently does not see it that way.

Anyway - I'm not surprised to see gist of the numbers you; but it would be really helpful to see what you got priced out for the DIY option - would you mind sharing the parts cost you found for the batteries and BMS (if you were to buy it)? Thanks!

 

[wattmatters]

Wow - our power is cheap here (10-12 USA-cents / kWh), so while your powerwall is more expensive, it's actually relatively more valuable given the difference in grid rates.

Cost of electricity seems highly variable across the US judging by some posts of prices in places like California. It is in Australia as well and though perhaps not as variable as in the US, it's sure not cheap.

We have just had some very nasty tariff increases across the board. Time of use tariffs (or demand tariffs) are normal here, most of those remaining on flat rate tariffs have older meters which will be upgraded if you install a grid-tied PV and/or battery system such as a Tesla battery.

Even so, Tesla's batteries here are very expensive (even after recent price drops) and so even with the new tariffs they just don't make financial sense.

My current tariffs are:

Tariff      $US/kWh (equivalent)
Peak         0.401
Shoulder     0.303
Off-Peak     0.218
CL           0.138
FIT          0.079
Service      1.359/day

CL = controlled load
FIT = Feed in tariff