Hacker News Reader

Comment by misterbowfinger | Hacker News Reader

Isn't this supposed to be solved by better batteries? I'm confused why the supply of energy matters, unless there's a limited amount of time to use it.

I'm not an electrical engineer, so if anyone understands this better, please correct me :-)

View on HN

Batteries simply don't really exist to store grid level amounts of energy. You have to stack lots and lots of batteries which costs lots and lots of money and still only stores something like ~10 MW when we have 1500 MW coal plants. Electricity is used immediately, so we really can't economically store it in the levels we need. In the future (this is just starting to occur), large batteries will charge (as a load) when the price is low and discharge (as a generator)when it is high which allows the owner to make money and helps the grid as these resources can provide power nearly instantaneously. Right now they're small and few in number, but they will get bigger and more prevalent assuming battery technology continues to get better.

-Source electrical engineer in this industry

Isn't 10 MW a unit of power and storing energy would be in MWhs or Joules? I understand that the grid needs to balance the power it is using, but to provide that power over time you need to store energy. Surprised an "electrical engineer in this industry" would use these terms, but maybe I just don't quite grok the lingo. An explanation would be great.

You are correct, I used to work with batteries for 3 years, people count capacity in Ah or Wh, but never in W. More often in Ah.

Why Ah instead of Wh? Because V is constant, making Ah proportional to Wh. Wh = Ah * V; V is the rated voltage the battery is designed for (12V for instance). By definition the capacity is how much energy you can get out of your battery at the rated voltage under standard temperature at a standard discharge rate. Voltage will go down as the battery depletes. When the voltage output is too low and cannot power the circuit anymore we consider the battery empty. Therefore, as we stop using the battery when V is too different, we can consider V constant over the usage period.

So Ah = k x Wh, k = 1/ V being almost constant. Ah is a easier unit to manipulate in charge rates, lifecyle etc.

One important note: due to technology limitations, energy one can get out of a batter greatly depends on the discharge rate (+-50% energy) and environment temperature (+-50% energy). Capacity cannot be taken literally as how much energy you can store and restitute. Design a grid that can still operate with lower voltage, charge at a lower rate and operate in hot weather and you suddenly have way more capacity (x3).

You're correct, but in the industry we really just talk in terms of Power at the transmission grid level and usually in millions of watts or MW. So if a coal plant is currently generating 2000 MW and you have a battery that can only consume 5 MW or 0.25%, then that isn't very much storage. Put another way, let's say the load in your region is 60 GW and you have 200 MW of energy storage. That is only a tiny fraction of storage relative to what the demand is.

So the main constraint on the battery is throughput, not storage capacity?

Nope, the main issue is indeed the fact that it can't store a whole lot of energy (also important that they can't maintain a high output for too long), but we still talk in MW as that is what we're primarily concerned with (MW flow on a line going over the line's rating...not having enough generation (MW) to meet load (MW)...etc). Saying a battery can't store enough MW isn't scientifically accurate, but it is the slang of the industry if you will. If you worked in the water industry they're probably very concerned with water flow rates. In high voltage transmission we're concerned with the flow of power (MW (real power) & MVAR (reactive power))

You have the read in the implied math about the length of the demand cycle and how long it takes to fill capacity at the quoted throughput. for example, 12MWH battery can handle 12MW of peak throughput for 12 hours, and then discharge over 12 hours of trough throughput. "12MWH" and "12 hours" are ouputs to the battery formula, "12MW" is the output.

Currently, it seems like thermal storage is likely to achieve widespread grid-scale capacity before battery storage.

But there are battery storage systems that have capacities over 100MW, so even today it's not just in the 10-20 MW range.

How many 100 MW batteries exist and how economical are they? 100 MW is much better than 10 MW, but still small if you look at the current load in MISO, SPP, PJM, ISO-NE, NYISO, CAISO...etc. So we have a long time before we can get rid of all fossil based sources...and even then how long can that battery provide 100 MW?

The new Tesla installation in Southern Australia is 100MW. I don't know the specifics, but the combined battery capacity of the two big Southern California utilities is at least 120MW, including one 80MW installation (also by Tesla). The three big CA utilities are mandated by law to have at least 1,800MW of battery storage by 2024.

None of this is intended to provide a consistent source of electricity. They're intended to fill in the gaps from uneven renewable energy production or other disruptions. Apparently, it's been working very well in Southern Australia.

Thermal storage is a promising solution because it will likely end up being considerably cheaper than batteries. The most promising type is salt storage for concentrated solar plants. They can potentially store enough heat energy to maintain consistent output all night long. Wind doesn't get that same deal - currently unused Wind power can only be stored as electricity.

Out of curiosity, why isn't "physical" storage used more?

Like pumping a bunch of water up high, or spinning a big ass flywheel.

I understand a flywheel might not be realistic at these scales, but water pumping seems like it could scale stupidly high.

Pumped storage has the same problem as hydro, namely, the number of places suitable to placing the large reservoirs required by it is limited.

Pumped storage is actually quite common where you have a lot of dams. The Bonneville Power Administration in the upper Westside of the US can essentially press a button and have quite a bit of power due to all the dams.

The article touches on the battery that is being used:

> In one case, it takes advantage of negative prices by being paid to pump large volumes of water into a mountain lake in Austria. When prices are higher, the company releases the water, using turbines to generate electricity.

This is Pumped Storage hydroelectricity - https://en.wikipedia.org/wiki/Pumped-storage_hydroelectricit...

Pump the water up when the power is cheap, use the stored water in a hydro dam when the power is expensive. The storage capacity is quite significant and while not as fast to respond to the grid as batteries, its substantially faster than other power generation systems when wind or solar power generation dips.

I think the problem is that it's early days with larger grid-scale batteries, and deployments are limited so far, but I expect this will change quickly. One example project that's rolled out is in Australia: http://money.cnn.com/2017/12/01/technology/tesla-battery-sou...

Yes thats exactly correct. There is a balance between what is made and what is used, it must balance out. So it takes time to ramp up or ramp down, or even control ramping when the wind is blowing...it seems. So to keep that balance (i guess) at times it makes more sense to drop the cost to below zero.

At least that's what i picked up from a youtube video about the topic a few weeks ago.

Reply on news.ycombinator.com