Batteries, economical or not?

[koputai]

Naturally when I was thinking about having a PV system installed at home I also thought “I should probably get a battery as well”. A couple of d friends have Tesla batteries and rave about them, so I was keen to convince myself that I needed one too.

Now, I always spend heaps of time analysing my projects, thinking of the various ways it could be done, done better, done prettier, and occasionally in that process I reconsider whether it should actually be done at all.

Price definitely comes in to play, but I don’t like to do things cheap, that’s usually just the long way to spending more.

The whole purpose of adding a battery to a home power system is to save money by storing power from a cheaper source, and using it later when power is more expensive, so in that respect you have to know how much any bit of power you can use is going to cost. This is how I decided whether a battery was a worthwhile investment.

Firstly I took my electricity rates. I have a time of use meter (TOU) as follows:

55c/kWh Peak

25c/kWh Shoulder

15c/kWh Off-peak

And my contract with Origin here in NSW gives me a 21c/kWh feed-in tarrif, meaning every kWh I feed to the grid earns me 21c.

So now I know that my cheapest power is 15c/kWh, which is during the Off-peak period from 10pm to 7am every day.

OK, I can hear you screaming “but the power generated by your solar is the cheapest, it’s free!”. No it’s not, as I get paid 21c for every kWh I feed back to the grid, every one of those that I use myself instead of feeding back actually costs me 21c in lost payment.

So how does this all relate to a battery? The electricity rate used to charge your battery, and the rate at the time you use your battery’s stored power both influence the actual cost of that power.

Now to the battery itself. The purchase and install price of a Tesla Powerwall 2 battery as per my quotes was $14,000 and the battery is warranted for 37Mwh (37,000 kWh) of aggregate throughput. If we divide $14,000 by 37,000kWh we arrive at a purchase price of $0.378 per kWh. That means even if you charge your battery with free power, it still costs you 38c/kWh.

Next you have to work out how much it costs to charge your battery. The Powerwall 2 has a round trip efficiency of 90%, so it takes 11% more kWh in than it gives back. Another way of putting this is that the power going in costs 11% more.

So, now we can see that one kWh of power from the battery will cost us:

38c + 55c = 93c per kWh if charged using Peak power

38c + 25c = 63c per kWh if charged using Shoulder power

38c + 15c = 53c per kWh if charged using Off-peak power

and finally,

38c + 21c = 59c per kWh if charged using our own ‘free’ power we’ve generated from our own solar PV system.

As you can see, for me, power from a battery would be the most expensive power I could use, costing me more than my Peak rate.

So what if your rates are different than mine?

Say you’re on a fixed rate of 25c/kWh with a feed-in tarriff of 7c/kWh, then:

38c + 25c = 63c per kWh if charged from the grid

38c + 7c = 45c per kWh if charged from your solar PV

Therefore a battery still isn’t economical as it costs 20c/kWh more than power from the grid.

Using the method above, you should be able to work out the costs for your own situation.

Cheers,

Jason.

Sorry about the font size and colour, I don’t seem to be able to fix it using this  Marc, help!

Edited July 25, 2020 by Marc

[gibbo9000]

Interesting and thought provoking analysis Jason.  Need to think about for a while and apply it to my situation and see where it comes out.  That peak rate looks high relative to Vic.  We are in suburban Melbourne and more like peak 37c off-peak 22c and FIT 12c.

My two immediate observations:

• Wonder if there is any evidence on Warranted versus Actual life for a battery - and whether depth of cycles matters.  Agree it should be in the analysis though.
• I agree with the 'solar isn't free' argument - but for two reasons.  The first is the opportunity cost of the FIT as you say.  The other implication of free is the capital cost of the system is written off as a sunk cost (a decision you have made and can't reverse - so shouldn't influence future decisions). Not sure whether that is wrong or right in pure economics, but doesn't sit well with me.  I spent the money and am looking for return on Solar PV investment, and that is through a cost to its output.

Gibbo 

[koputai]

I agree totally Gibbo. The only reason I didn’t include the sunk cost of the solar, and even its opportunity cost, is that I was purely analysing the battery as a separate entity to the solar. Of course, I didn’t include the opportunity cost of the battery either, which at around $700 per year is not to be sneezed at.

 

Cheers,

Jason.

 

[koputai]

Sorry, I mentioned the round trip efficiency but forgot to use it in the calculations. With that corrected:

38c + (55c x 1.11) = 99c charging with Peak

38c + (25c x 1.11) = 66c charging with Shoulder 

38c + (15c x 1.11) = 55c charging with Off-peak

38c + (21c x 1.11) = 62c charging with solar generation.

 

If you want to take the opportunity cost in to account, the battery can do 5,000kWh per year, and $14,000 invested at 5% returns $700p.a.

$700 / 5000kWh = 14c per kWh to be added to the above figures.

Cheers,

 Jason.

 

[rmpfyf]

Did a ton of work on this in the last job, I cluding developing code to work out when it is cost effective.

Hit the mark a few years ago at peak prices with some batteries and since energy prices have cooked more than battery costs.

Still, storage doesn't need to be a battery.

[Marc]

2 hours ago, rmpfyf said:

Still, storage doesn't need to be a battery.

I think that's a really good point. What I am doing by diverting surplus into the car, the pool etc, is "storage", however it doesn't help offset my grid consumption post 6PM. Only batteries are going to help that (or switching the mains off, which I have contemplated but I'd probably have a mutiny on my hands!).

[rmpfyf]

49 minutes ago, Marc said:

I think that's a really good point. What I am doing by diverting surplus into the car, the pool etc, is "storage", however it doesn't help offset my grid consumption post 6PM. Only batteries are going to help that (or switching the mains off, which I have contemplated but I'd probably have a mutiny on my hands!).

Got some good gains with thermal - intelligent room heating particularly where very insulated. It's helping my post-6PM well enou

I did a build a few years ago on a trial home that went hard into water thermal storage - two tanks, one for DHW and the other for space heating - and a hydronic system accordingly. Was brilliant, worked very well and not expensive for a new build. The radiators had fan coils and the driving heat pump was reversible, so it did 'cold' very well also 

[Marc]

Another good point you raise if justifying the outlay and investment here. I've already over-capitalised on this place (major extension on 35+ year old house), and probably even in my renewable projects. But I never really got into any of this for the cost reduction as the primary driver. So much more would be easier to justify at the time of design and build (+ associated finance).

[rmpfyf]

30 minutes ago, Marc said:

Another good point you raise if justifying the outlay and investment here. I've already over-capitalised on this place (major extension on 35+ year old house), and probably even in my renewable projects. But I never really got into any of this for the cost reduction as the primary driver. So much more would be easier to justify at the time of design and build (+ associated finance).

I was totally floored at the difference in cost in build vs reno. I am a twinge regretful on the current reno (went 8x multi split on two machines) though it's a very efficient setup. Done again I'd have done more with heat transfer and hydronics.

 

[Marc]

Don't even get me started on that one. We doubled our home size. We had ducted gas plus simple splits in main areas previously, which allowed for tailored zoning.

Our builder (a family member) convinced us to remove the ducted gas and splits and replace it all with one big Lennox ducted heating and cooling system (with three zones). The zoning is a joke thanks to only relying on one centrally located thermostat, and terrible control over the whole system. Our running costs hover somewhere between double and triple what they were before, which granted, our house is nearly twice the size (including a 2.5 car garage), but the new system should have been far more efficient. The Heating/Cooling is now the biggest consumer of power and not something I can work out how to resolve easily.

[rmpfyf]

Oh noooooooooo. I deleted our central unit the second I started renovating; my family would come home and hit the 'make hot' button which has coarse spatial resolution and much energy input. 

Multisplits are friggn awesome by comparison save for the notion of their being so visible. You can of course get hidden heads, if the reno allowed more time, carpentry and finishing work this would have been good. I ended up with LG which is good (but I won't pretend is best) however all heads are WiFi and connect back through their ThinkQ interface, in turn Google Home... it works really well.

I guess you could install more dampers and multi-zoning, go to a control system with independent temp sensors but... yeah. Work and cash.

[Marc]

The problem is the system was flawed from delivery, to installation, and execution so much so we fell out with the installer.

Now trying to find someone who can take a look at the system and advise improvements (installation of genuine Lennox zoning system and multiple thermostats rather than the aftermarket 'dumb' zone controller), we keep getting referred back to the same installer, as this is "his area". 

How my blood pressure is normal, I have no idea 🙂

[rmpfyf]

Won't be the genuine zoning system that makes or breaks it IMHO. 

Could recommend a few, don't know what parts they work it. If you want Lennox you're constrained.

[Tezza]

On 24/07/2020 at 12:10 PM, koputai said:

The whole purpose of adding a battery to a home power system is to save money by storing power from a cheaper source, and using it later when power is more expensive, so in that respect you have to know how much any bit of power you can use is going to cost.

That may be true in the city but my main motivation for getting one now is that the numerous blackouts we get in storm season are a pain in the a*** and cost me money (as my office is at home).  Happy to pay a premium to avoid those.

On 24/07/2020 at 12:10 PM, koputai said:

Now to the battery itself. The purchase and install price of a Tesla Powerwall 2 battery as per my quotes was $14,000 and the battery is warranted for 37Mwh (37,000 kWh) of aggregate throughput. If we divide $14,000 by 37,000kWh we arrive at a purchase price of $0.378 per kWh. That means even if you charge your battery with free power, it still costs you 38c/kWh.

 

I think this logic is a bit flawed.  Like anything a battery is not going to get to the warranted life and then just fail.  If they guarantee 37Mwh then it is likely that the expected life is a fair bit beyond that (as they don't want to replace more than a small portion).  I know capacity will be lower at the end of that period (maybe 80% of new) but it will still work as a battery its just your daily throughput will be lower.  If you assume it can do say 70Mwh before it actually fails then your battery cost is $0.20 per kWh.

If you used that in your analysis then in your situation it could make sense to have a battery but you would charge it from off-peak power and then use it during the peak times.   Would be 20c + 15c*1.1 = 36.5c/kWh instead of paying 55c/kWh.  Of course if the difference between your cheapest and most expensive power rates is less than 20c (as it is in my case) then battery still doesn't make financial sense on that basis but that ignores the benefit of having backup power available in a blackout situation.

[koputai]

I agree with everything you say @Tezza, however if you want solid figures, you have to use guaranteed quantities, which is why I used the warranted throughput.

 

70Mwh is a bit of a stretch, pretty much double the warranted throughput. If it has degraded to 80% after ten years (and 37Mwh throughput), it would take something like thirty years to reach 70Mwh, by which time something else will have blown up long ago.

I suspect the break-even point is somewhere between, like 15-20 years, in a full charge-discharge cycle per day installation, which is probably about the same time the machine fails.

 

Cheers,

Jason.

 

[Prof Tournesol]

On 10/08/2020 at 10:09 AM, Marc said:

I think that's a really good point. What I am doing by diverting surplus into the car, the pool etc, is "storage", however it doesn't help offset my grid consumption post 6PM. Only batteries are going to help that (or switching the mains off, which I have contemplated but I'd probably have a mutiny on my hands!).

This week we are increasing our rooftop PV from 4.5kW to 10kW, mostly to reduce the load on the battery in winter and also to dump excess generation into the car during the day. We have a 17kWh lead-acid gel battery but the load on it was excessive with our current generation/use. Hopefully that'll change with increased generation capacity. We mainly have the battery to guarantee power for electric roof sprinkler pump, we haven't been able to draw more than 3kWh without the battery overheating

[koputai]

Well my first year is almost up, and Origin have decided that my 21c feed-in tarrif won't continue. They don't appear to offer any more than 12c to anyone any more.

I've run my spreadsheet over all the offers I can find, and staying with Origin seems to be my best option.

 

My new electricity rates will be as follows:

55c/kWh Peak

25c/kWh Shoulder

15c/kWh Off-peak

The new contract with Origin will give me a 16c/kWh feed-in tarrif.

 

So now I know that my cheapest power is 15c/kWh, which is during the Off-Peak period. Power from my own solar production now effectively costs me 16c/kWh if I use it myself.

 

So, now we can see that one kWh of power from the battery will cost us:

(38c + 55c) x 1.11 = $1.03 per kWh if charged using Peak power

(38c + 25c) x 1.11 = 70c per kWh if charged using Shoulder power

(38c + 15c) x 1.11 = 59c per kWh if charged using Off-peak power

and finally,

(38c + 16c) x 1.11 = 60c per kWh if charged using our own ‘free’ power we’ve generated from our own solar PV system.

 

Therefore a battery still isn’t economical as it costs 5c to 45c/kWh more than power from the grid.

 

Cheers,

Jason.

Edited December 16, 2020 by koputai
Updated pricing for actual contract

[Prof Tournesol]

We've recently changed from Powershop to Amber, who give us wholesale pricing. At least in summer, this is usually <20c kWh, often less than 12c kWh, (and once it was -2c kWh!!!) apart from 3-9pm when it can peak at 60c. Our PV covers us for most of that time in summer and the battery covers us for the rest. My powerbill has dropped from around $200 a month with Powershop (including EV charging) to under $70 a month. It may be different in winter, I'll have to wait and see, but so far the battery has been very useful

[rmpfyf]

8 minutes ago, Prof Tournesol said:

We've recently changed from Powershop to Amber, who give us wholesale pricing. At least in summer, this is usually <20c kWh, often less than 12c kWh, (and once it was -2c kWh!!!) apart from 3-9pm when it can peak at 60c. Our PV covers us for most of that time in summer and the battery covers us for the rest. My powerbill has dropped from around $200 a month with Powershop (including EV charging) to under $70 a month. It may be different in winter, I'll have to wait and see, but so far the battery has been very useful

Sure, though Amber isn't making money and if you get caught out when the market goes to vol ($14.10/kWh) it can get dicey. 

Still, whilst the ride lasts...

[Prof Tournesol]

58 minutes ago, rmpfyf said:

Sure, though Amber isn't making money and if you get caught out when the market goes to vol ($14.10/kWh) it can get dicey. 

Still, whilst the ride lasts...

If I'm seeing those sort of market prices I'll switch, but I should be able to cover those times by drawing on the battery

[rmpfyf]

51 minutes ago, Prof Tournesol said:

If I'm seeing those sort of market prices I'll switch, but I should be able to cover those times by drawing on the battery

Be kooky - write a script to command your discharge. AEMO's predispatch signal (the 'what the market should do') is freely available to anyone. 

A number of players (e.g., Reposit, Evergen, Members, etc) use that same signal as a basis for their 'what we do next' engines. They don't offer these with Amber's tariff, yet, probably because Amber's working on their own version of same. 

Don't get me wrong I think exposure to real market behaviours are a good things towards more efficient use of markets and networks. I just don't think mainstream customers - even those with batteries - are ready for exposure to market volatility.

[Prof Tournesol]

20 hours ago, rmpfyf said:

Don't get me wrong I think exposure to real market behaviours are a good thing towards more efficient use of markets and networks. I just don't think mainstream customers - even those with batteries - are ready for exposure to market volatility.

I suppose Amber's business model is predicting that you are wrong😊 It may cause problems in winter, there's no contract so we aren't locked in, but so far it's been far better for our circumstances that Powershop, even with the car charging.

[rmpfyf]

15 minutes ago, Prof Tournesol said:

I suppose Amber's business model is predicting that you are wrong😊 It may cause problems in winter, there's no contract so we aren't locked in, but so far it's been far better for our circumstances that Powershop, even with the car charging.

Not really - they're losing money at the moment (effectively paying their customers) though this is not uncommon as energy businesses start and aren't yet scaled up. There's hope for them yet. 

What they do is not for everyone just yet. But - with a bit of tech and more smarts - it should be.

[Prof Tournesol]

We've just bought our second home battery. The first one, a lead acid gel battery, has reached EOL. We've replaced it with 4x 3kWh GenZ LFP batteries. Each module has its own BMS, so if one module fails, the rest of the battery won't - vital when it's an hour and a half drive for our installer to fix the problem. It's connected to our 9kW of roof panels, DC coupled through a Selectronic SP PRO inverter. The system needs to be reliable as it powers our electric pump for our roof sprinklers, but as an added bonus, almost all our power bills are car charging power imports. We have a Tesla EVSE, but it's controlled by an app called ev.energy that only charges when Amber prices are low.