A stated volume would be more useful in describing the needed battery size.

While the article doesn't state requirements in terms of volumes, it does say that for NYC a 13 GWatt requirement "would fill nearly 60,000 square meters."

Doing a quick back-of-the-envelope calculation, this works out to a little more than a city block of area (660ft x 660ft).  How much of that could be stacked and built underground is not known, but depending on how much room you need for interfacing the system to the power grid, buildings and parking for workers, security spaces around the facility, etc., for a 4 square-city-block area this would be about 37% of the available space.

The real driver on renewable power based on wind and photovoltaic systems (after technical feasibility) are the costs of storing a kwh of power, the efficiency of the charge discharge cycle, and the life of critical components. Typically a storage system is only going to go through one or two charge discharge cycles per day. So if the system costs $100 per kwh of storage capability and is 99% efficient kwh in to kwh out(not likely), it costs $1200 to $2400 per daily kwh to run as part of a system. A simple cycle gas turbine system even with fairly expensive gas will be a cheaper way to go for peak needs. It is clear that storage and dispatch capability is the key to a functioning wind or photovoltaic system. We can only hope that this effort will bring a new set of possibilities.

The big problem with the electric power grid is producing power when you have little demand. A smaller problem is power produced intermittently.

Roughly two thirds of our power comes from coal and nuclear power plants. Because it takes days to heat them up or cool them down, they are never turned off, except for problems or some types of maintenance. These are called base-load plants.

This means they produce lots of energy in the middle of the night when demand is quite low. The incremental cost of power during peak periods is usually at least an order of magnitude greater than during the least used off peak periods. If your first sentence was correct, we would have none of these plants now.

How does the grid handle this problem now?

First, lower prices for using off peak power induce large industrial customers to shift their use to the middle of the night. We should be doing more of this, including many more commercial and residential customers, as time-of-day meters become less expensive. An important new use would be plug-in hybrid vehicles.

Second, natural gas power plants can be turned from off to full capacity, or the reverse, fairly quickly. They have relatively low fixed costs but relatively high incremental costs. They are called peaking plants.

Third, hydro is used extensively in reverse. Pump water above the dam so more power can be produced during the day. Because of long-distance transmission, this is used for large areas of the country. It's quite efficient.

The same three approaches would be used for wind or solar. Neither could be relied upon for providing peak power, though when available they would substitute for more expensive natural gas power.

The intermittent nature of their power introduces no big problem for the grid either. The power from a portfolio of wind or solar facilities would vary significantly, but it would almost always be slow enough for a natural gas plant to compensate for changes.

If the level of power changes very quickly now, power is quickly brought in from a longer distance or heavy industrial customers that have already agreed have some heavy power uses quickly cut off. The grid has a reserve capacity to adjust to such situations and assure reliability. This would continue.

All of the problems you discuss, we deal with now. Batteries, supercapacitors, peak/off-peak pricing, long-distance DC transmission, and thermal storage should all play a larger role in addressing this problem. They are each quite important, especially if there are any cost breakthroughs—but big changes in these technologies are not essential to absorb large amounts of wind and solar into the grid.

With liquids sloshing around in giant batteries, what is the likelihood that in an earthquake the metallic electrodes might be shorted out?  Could this lead to an explosion?  How does a liquid battery respond in such situations?  With the "huge" batteries envisaged, safety is also a huge issue.  Small individual cells might fare better, and the liquids might need to be stabilized in some kind of porous inert solid to prevent such accidents.

In addition, magnesium is highly reactive and accidental exposure of hot molten magnesium to air, as might happen if the battery is damaged by an earthquake, could result in an extremely dangerous fire.

What are the main differences between this liquid battery and a flow cell device? The Regenesys system sounds more like a large liquid filled tank of electrolytic materials which gets changed out periodcally by flushing the old solution and filling the container back up again. These liquid battery systems look like the fluid flows in continually through some type of filler passage. Correct?

What are the practical sizes that either can be made? If we're to look at electrcial vehicles as our primary mode of transportation in the future we'll all need power systems at home capable of storing inexpensive power at night so we can charge our cars up for travel.

Besides, distributing emergency power house by house removes and SPOF in the power grid from a consumer perspective. A 10 to 20KVA system for home emergency power may be pretty cheap to produce, and low enough to pay for with a HELOC.

Why not available in a year

So the next step is to adapt this storage technology into a Gel, then retrofit the gasoline storage tanks at petrol stations to hold this gel.  Then electric cars could easily get a recharge by pumping out the spent gel and pumping in the charged gel.  The stations could charge from the grid any time the price was right...

How can I help make this happen? 

_Vaughn Micciche

I did some reading and compared to the flow batteries I read up on (VBR).  This baterry does need a membrane buts uses a salt instead. A  pump is not needed with this battery, instead gravity is used.  So it is mechanically simple and could have lower maintence cost.

As for using it as a car battery VBR have already been tested in electric golf cart style along with refueling by putting changed battery liquid in . The problem seems to be weight. What is the power density/kg? VBR did not match solid batteries. Also the mechanical simplicity would probably not work in a moving vehical, flow battery might be better in case where the liquid will be moved around, and not sit still.

This would be great on trains, to regenerate braking energy and turn locomotives into large hybrids!
  Most locomotives are diesel-electric and could easily put electricity into batteries when breaking instead of burning off electricity via a heating coil.
  The liquid battery could be put into three or four liquid containers behind the locomotives, and when the train brake the electrical motor generate energy lowing the train and charging the batteries. Moving forward the train would suck current from the batteries.
  If they can get it to a couple million (10million) per unit, it would sell, because of the return on investment and aid in lowering fuel costs!
http://science.howstuffworks.com/diesel-locomotive4.htm
Brian Glassman
Commercialization
Innovation 

Aside from the obvious day/night storage problem, there's also the seconds-to-minutes variability of wind and (to a lesser degree) solar output, which causes dispatch and stability problems for the grid. 

Being electrochemical, batteries (these, or any others sufficiently cheap) should be capable of sub-second changes in power flow and direction, AND -- unlike supercapacitors -- store enough energy to address the day/night problem as well.

They would seemingly be co-located remotely, at the wind or solar farm, so peak power levels don't have to be sent far and transmission lines from farm to city can be steadily (and efficiently) loaded.  Not to mention the safety aspects of keeping gigajoules of stored energy away from populated areas.  (With apologies to any country folks here).