Exploring the economics of TOU Battery Arbitrage

[ronaldraygun]

Doing some back of the envelope math to assess the viability of using TOU Battery Arbitrage--no solar--to power my home.

I'm here in NYC, where the spread between peak and off peak is as high as 23.7 cents.

Going off the complete system blueprint, comes out to $13451 for ~21 usable kwh (assume 80% depth of discharge max, and another 10% in losses)

That then means, the cost of the system would require 56,755kwh in order to break even.

1 Year of usage is 9,102 kwh.

So that then implies 6.23 years for the system to break even.

Am I doing this math correctly?

 

[Ampster]

Am I doing this math correctly?

I think the unknown is how many hours and kwhs to you need to cover usage at the high rate. That will give you an estimate of battery capacity you will need and the peak usage will be how you size the inverter. That will help you cost the system and the rate differential times the daily kWhs will tell you how much you save per day. Then do you have enough hours at the low rate to charge batteries from the grid. The savings per day divided into the cost of the equipment will give you the number of days for payback.

 

[Hedges]

Probably.
Can you do it with some 5kWh server rack batteries for $1500 each, and an inverter/charger?
If the price could be 50% to 66% of your figure, break-even is all that much shorter.

What is round-trip efficiency? One guy calculated low 90%'s, but measured low 80%'s.
I figured that was due to non-ideal load waveforms (but just a guess.)

That linked system says off-grid. Best setup would be grid-tied either exporting (if utility allows you to sell back their own power) or zero-export. Then it supplies your whole house, but doesn't need any particular peak wattage. If you have 5 hours peak rates, only need 4kW inverter to drain battery in that time.

 

[ronaldraygun]

I think the unknown is how many hours and kwhs to you need to cover usage at the high rate. That will give you an estimate of battery capacity you will need and the peak usage will be how you size the inverter. That will help you cost the system and the rate differential times the daily kWhs will tell you how much you save per day. Then do you have enough hours at the low rate to charge batteries from the grid. The savings per day divided into the cost of the equipment will give you the number of days for payback.

Noted on the peak vs off-peak usage. Definitely didn't consider that.

With a payoff that long, probably not going to make a material difference unless NYC tries to compete with San Diego for most expensive electricity.

 

[DIYrich]

30% federal tax credit for batteries, if you can benefit.

 

[ronaldraygun]

Probably.
Can you do it with some 5kWh server rack batteries for $1500 each, and an inverter/charger?
If the price could be 50% to 66% of your figure, break-even is all that much shorter.

What is round-trip efficiency? One guy calculated low 90%'s, but measured low 80%'s.
I figured that was due to non-ideal load waveforms (but just a guess.)

That linked system says off-grid. Best setup would be grid-tied either exporting (if utility allows you to sell back their own power) or zero-export. Then it supplies your whole house, but doesn't need any particular peak wattage. If you have 5 hours peak rates, only need 4kW inverter to drain battery in that time.

30kwh (really, 24kwh with 20% discharge buffer), is one server rack battery kit from signature solar. I unfortunately don't have hourly peak vs off peak usage so hard to know precisely daily usage patterns. Using the 9102kwh figure from above, that implies a daily consumption of 24.93kwh/day. Some subset of which will be during the off-peak time. So it seems that 30kwh EG4 server rack battery tower will be just sufficient to cover the average day--theoretically 50% of days. I may end up having to add one more server rack battery to increase the percentage of days that don't require paying for peak power at the cost of a longer payback period.

No idea what the round trip efficiency is, I figured 10% was appropriate, but if those measurements are true, then I'll almost definitely need another server rack battery to bring total capacity up to 35kwh.

My understanding of the linked system's inverter is that it acts as a load from the grid, never backfeeding power. So hypothetically, it's no different from some other high amperage power drawing appliance, like a water heater, or maybe electric dryer. I don't intend to backfeed into the grid, although it would be tantalizing to sell off any excess battery storage at the peak rate.

In terms of layout, I imagine some sort of transfer switch between the meter and the panel such that power is either flowing through the inverter and battery system before reaching the power, or directly from the meter to the panel. That should theoretically allow for any maintenance / upgrades on the inverter and/or batteries without disrupting the rest of the house.

I admit to be in the, "this-could-be-a-good-idea" stage. Still a long way to go before execution and installation.

 

[ronaldraygun]

30% federal tax credit for batteries, if you can benefit.

Doesn't that also apply to the inverter?

Assuming it does, that $13k figure goes down to $9,415. Which in turn means 39,725kwh to break even or about 4 years. Definitely helps the math a bit.

 

[zanydroid]

I’ve done the exercise myself for California, in two different price regions for electricity. Doesn’t look good for the maximally legal systems.

How are you getting to 20kWh per day spread between peak and off peak usage in an urban property? I guess you would need heat pump heat and AC to hit that.

Urban setting also means you may need to get creative on where you put the batteries. There are constraints in where they can go in habitable space. Going by 2019/2021 international residential code.

Along those lines either your inverters or batteries need to be more expensive to pass UL9540 listing, also required by more recent IRCs. With the currently available products it’s probably around $10K for a base 10kWh configuration. IOW that complete system blueprint off grid standard set up is only suitable for some states / some living situations.

 

[Surferdudemi]

How much can you save by just shifting usage patterns? I've been on net metering for years, and then they made ToU compatible with the net metering rate (that somewhat makes up for taking away $0.06 distribution charge from the export price). I've been able to shift about 90% of my usage to the off-peak hours just by doing stuff like laundry and most cooking outside peak hours. Once you've done that, the economics change because you're using fewer expensive kWH for zero investment.

Another way to look at it is maybe you are willing to pay for the battery and inverter so you don't have to care about when you use most electricity. Or maybe you need A/C desperately in the summer or electric heating in the winter (I have gas heat, so only need blower power).

You also cannot ignore the value of having electricity when there are outages due to storms. Just make sure you do the wiring to code with a transfer switch or proper inverter type and sizing.

 

[DIYrich]

Doesn't that also apply to the inverter?

Yes. Whatever you need for the batteries to power your house, and to recharge.