Hacker News Reader

Comment by brotchie | Hacker News Reader

How feasible will it be to use 60+kWh of home battery storage for time-shifting arbitrage (not pure arbitrage, but close enough)? Is it yet possible for a consumer to get direct exposure to the spot electricity price for both feed-in and feed-out?

It would be interesting a) What the payback period is on a strategy where you charge when the spot price is low, and discharge when high, and b) How much installed battery capacity would be required to eliminate this arbitrage? i.e. what MWh of installed battery capacity would flatten the 24 hour across-grid demand cycle.

View on HN

If it works out, wouldn't the utilities do it themselves? Why transmit that power for miles and then transmit it back. And even if they didn't, arbitrage means the price difference between times would fall as more people do it. Peak prices would move closer to night prices as demand equalizes between the two.

Exactly. Coal, wind, and nuclear plant owners would just start charging batteries.

Then the price will rise at night and lower during the day.

Anyone who buys one of these just for time-arbitraging electricity is taking a pretty big risk.

This smells of the Newman/Krammer bottle deposit arbitrage scam.

I think most people would be happy to just charge the thing up via solar during the day and use it at night.

Even if it was powered from the grid, if these were installed everywhere you could have distributed charging mechanisms, which would lower the peak load on the grid, which would mean less peak power required across the board.

There should be savings there as well!

It's the opposite though - during the night, you want to use power from the utility, when it's plentiful, and cheap (particularly if you have a TOU contract).

The objective would be charge the battery up at night from the grid, and then use that power during the day + your solar power, when power is expensive.

Once solar becomes widespread enough, the dynamic changes. In Germany, electricity spot prices can be lower around noon because of solar energy. A quick Google search brings up this example: http://cleantechnica.com/2012/03/23/german-solar-bringing-do... (note that this already happened in 2012!)

Spot prices could even go negative, depending on the rules of the market. That's certainly a feature of Australia's National Electricity Market (NEM) which covers the eastern states. The price can drop to a floor of around -$1000/MWh, so at these times the large generators are paying the network for the privilege of keeping their running - which they are generally forced to do, since such a period of negative prices doesn't tend to last long... certainly not long enough to justify shutting a coal-fired turbine down and then the expense of spinning it back up again.

If you have solar panels, daytime power is as cheap as it gets. The cost of daytime power is negative for me, since my home solar array produces more power than I can use.

Net metering is an accounting trick which lets you treat the grid as a big storage battery. My house "charges up" this battery during the day by running the meter backward; then I can "draw down" the battery at night, when my panels are inactive.

If I had a literal storage battery at home, the economics would work exactly the same way: but instead of externalizing the cost of the battery onto the public utility, I'd be paying the cost myself by purchasing the battery.

Very few people have an excess of power from their solar grid - unless they aren't at home (which, come to think of it, is an excellent use case for a home battery). For most people, the home battery is going to be useful from day one by entering a TOU contract with their local utility, and then charging at night, and using the power during the day.

Solar power, on top of that, will be a bonus.

For the summer, California E-6 Residential rates are $.32/kWh during peak times, $0.13/kWh during off peak. If your house is using, say, 4 kW, for 5 hours during peak - then that 20 kWh will cost save you 20 * ($0.32-$0.13) = $3.80/day in power costs, approx $115/month. Do that for 4 months/year ,that's $460/year. With a 5 year life, that battery would need to be less than $2300 to come out even.

http://cleantechnica.com/2014/09/05/teslas-gigafactory-may-h... suggests that $100/kWh is in the ballpark - so this might make sense right out the door.

My point is that most people would be perfectly happy just to say that they're no longer reliant on burning fossil fuels to power their homes. Life is just too short for me to care about arbing the residential electricity market.

That is effectively a form of arbitrage, I suppose. You're charging the battery when electricity is either cheap (middle of the night), or free (from solar on your roof) and then using it in place of expensive day time electricity.

You could argue using battery electricity in stead of bringing in grid electricity is functionally equivalent to selling it to the grid, both in terms of pricing incentives, and the effect on the grid – in the same way that people are discussing being able to switch fridges on and off minute by minute as being equivalent to building grid generators which can be switched on and off on that timescale.

This was my initial thought too, but I had another!

How feasible is it that the majority of households could have a massive home-battery installed in terms of natural resources? It seems like a lot of rare earth materials to build so many big batteries.

It seems like a lot of rare earth materials to build so many big batteries.

Lithium-ion cells actually don't necessarily require any rare earth metals at all. For instance, the batteries Tesla currently uses for their cars are made up of lithium, nickel, cobalt, aluminum, and carbon, none of which qualify as rare earth elements.

Lithium isn't considered a rare earth element, although it is relatively rare compared to, say, lead.

From Wikipedia: However, according to a 2011 study conducted at Lawrence Berkeley National Laboratory and the University of California Berkeley, the currently estimated reserve base of lithium should not be a limiting factor for large-scale battery production for electric vehicles, as the study estimated that on the order of 1 billion 40 kWh Li-based batteries could be built with current reserves.[85] Another 2011 study by researchers from the University of Michigan and Ford Motor Company found that there are sufficient lithium resources to support global demand until 2100, including the lithium required for the potential widespread use of hybrid electric, plug-in hybrid electric and battery electric vehicles. The study estimated global lithium reserves at 39 million tons, and total demand for lithium during the 90-year period analyzed at 12–20 million tons, depending on the scenarios regarding economic growth and recycling rates.[1]

Of course, lithium isn't the only element we can use to make batteries.

Note, I'm not making a claim one way or the other on the feasibility, just stating that resources don't appear to be a limiting factor for the next ~life-time.

1. http://en.wikipedia.org/wiki/Lithium

Here is your feasibility for using batteries in just cars: http://large.stanford.edu/courses/2010/ph240/eason2/

There's a lot of lithium, but it's unlikely there is enough Lithium in the world to provide everyone with large home batteries.

Other commenters disagree with this... but to pull in a concept from other stories on HN, I wonder if any near-Earth asteroids are rich in lithium?

There is a LOT of lithium in the world and that's just near the surface - you can always dig down.

But at that scale you don't have to use lithium anyway. When it comes to grid or off-grid storage, you don't need the best energy density as cost per watt-hour is probably the most important factor and there are already other chemistries on the market that will work just fine with many more on the way. Sodium or Aluminium based batteries seem particularly promising.

Reply on news.ycombinator.com