How large of a battery storage facility (in MW) would be needed to supply 11,200 MWh of electricity for 4 hours before becoming completely discharged.

The facility would be comprised of lithium ion batteries.

The text below may supply some context: "The Applicants assume that a smaller-scale, alternative energy battery storage involves the installation of smaller-scale batteries and associated equipment to supplement the gas supply system at times when additional capacity is needed... assume that smaller-scale battery storage would supply four hours of electric supply, including approximately 11,200 MWh of energy storage capacity."

Please don't put on hold. there was some incredibly useful discussion coming from this horribly presented question.

  • 2
    \$\begingroup\$ You are mixed up. Define the maximum load power required (MW) the run time (h) and the product of the two gives you the required capacity (MWh). If the load is 11,200 MW then for a four hour run you need 44,800 MWh capacity. A question edit is required. \$\endgroup\$ – Transistor Sep 20 '17 at 21:54
  • \$\begingroup\$ I realize that its on a massive scale. It is a hypothetical at the moment. So, a 2,800 MW facility would have approximately 11,200 MWh of storage capacity? \$\endgroup\$ – gyoung1986 Sep 20 '17 at 22:06
  • 1
    \$\begingroup\$ @gyoung1986 Well, if it is only momentarily hypothetical, do keep me near the top of the list when it moves towards reality. I've always wanted to own a small, modest-sized continent somewhere. The way to word it is: "In order to be able to comply with a continuous demand for 2800 MW for a period of up to 4 hours, you would also need a facility able to safely and continuously maintain 11200 MWh of battery storage capacity." \$\endgroup\$ – jonk Sep 20 '17 at 22:11
  • 1
    \$\begingroup\$ @gyoung1986 Dividing MWh by h gives you MW. (Seem simple enough?). So 11200 MWh equals 2800MW for 4 hours, or 5600MW for 2 hours, or 1 MW for 11200 hours. I assume that's the theoretical ballpark calculation you're looking for. Now, if you are to actually build a storage plant of 11200 MWh you'll want to be a lot more sure about the details, like efficiency, and maximum discharge rate. \$\endgroup\$ – user253751 Sep 20 '17 at 22:38
  • 1
    \$\begingroup\$ @gyoung1986 That applicant statement is a bit vague on details. If this is legit work, you need to enter into a discussion to hash out exact details. \$\endgroup\$ – jonk Sep 20 '17 at 22:45

It's actually kind of a fun experimental thought, because I actually don't have any experience with anything of this magnitude (or several orders below it, to be honest.) This is way out of my experience. Which is why it's fun. (I must admit though that I've been involved in developing a product used by power companies in the US and in China for monitoring the oil temperatures at many internal points within large power transformers used for energy distribution. For those engineers out there... consider the difficulty of placing wires and/or anything electronic inside of a \$2500\:\textrm{kVA}\$ power transformer.)

Your specifications:

  • \$11.2\times 10^9\:\textrm{W-hr}\$ of energy storage capacity.
  • \$2.8\times 10^9\:\textrm{W}\$ power delivery capacity.
  • Able to sustain #2 for \$4\:\textrm{hr}\$,
  • Lithium ion battery storage technology (perhaps Tesla PowerPack)

The above is able to completely replace a 36" natural gas pipeline, where all of the full-up gas capacity of that pipeline is used to generate electricity, for a period of up to four hours.

Tesla's PowerPack specifications are:

  • Energy Capacity: \$210\times 10^3\:\textrm{W-hr}\$
  • Power: \$50\times 10^3\:\textrm{W}\$
  • System Efficiency: 89% round-trip (4 hour system)
  • Depth of Discharge: 100%
  • Dimensions: 51.5" x 32.4" x 86", or about \$2.4\:m^3\$
  • Weight: \$1622\:\textrm{kg}\$

Tesla's Inverter specifications are:

  • Power: \$50\times 10^3\:\textrm{W}\$
  • Dimensions: 39.9" x 49.4" x 86.3", or about \$2.8\:m^3\$
  • Weight: \$1200\:\textrm{kg}\$
  • (Scalable Inverter Maximum: \$625\times 10^3\:\textrm{VA}\$)

From the above, we can make some deductions.

  • Tesla's PowerPack modules provide a nifty full-load to full-discharge time of 4.2 hours. Quite convenient. I strongly suspect that Tesla's specifications have informed the situation here.
  • Your specifications use watts and fail to take into account the difference between watts and volt-amps. Practical systems delivering this kind of power will need to pay very close attention to these differences and size themselves appropriately. They may also require added components to support power-factor adjustments. These cost money, take space, and require support and maintenance.
  • Tesla's standard inverters can't scale to your needs. They fall short by a factor of almost 20000. You'll need to find out what kind of inverters can be scaled up in order to achieve what you need.
  • Assuming Tesla's inverters could be scaled up, you'd need 56,000 of them.
  • You'd also need 56,000 of their PowerPack storage modules.
  • If everything were stacked side-by-side, with no hallways or room for getting in there and repairing or replacing anything or for dissipation of heat, you'd need almost 300,000 cubic meters of volume to store them. If you avoid stacking them, that's 34 acres -- jammed tight. Clearly, you'd need a LOT more than 34 acres.
  • None of this deals with the requirements for distribution and support for that distribution from the inverters. Nor does it deal with building space for personnel, their support (food, airconditioning, workspaces, etc.), the storage needed for equipment and parts, shipping docks, roads, etc.

And like I said, you need to isolate these units from each other. You would NOT want the explosive failure of a unit or two to cascade into failures of nearby units.

There are lots of other details. I assumed 100% efficiency above. Tesla claims 89%. Assuming Tesla takes into account all of the final system's losses (which I'm sure their figures don't), you'd need 63000 pairs of modules and not 56000. (I'd say you should tentatively plan on 75000, until you can refine the figure down more precisely.) You also need to work out how to shift loads onto and off of the grid and how to deal with replacing failed distribution components while they are operating. Things fail under heavy load and when they fail, if you must maintain continuous operation, you need to be able to move that load onto another unit (without risking its destruction, too) and isolate the unit you need to replace and bring back on line. In the meantime, the temporarily "overloaded" unit may heat up beyond its specifications. Or if it doesn't, it will certainly age faster than before. So you need to keep track of the rate of aging of all of these components so you can anticipate and avoid, rather than respond to disasters.

And I really have to wonder about a system that only gets used once in a while. If it only gets used during extreme cases of need, odds are that is also exactly when you will find out there are problems. So I'm pretty sure the system needs regular testing and use to maintain it's ability to function in a pinch.

I can only begin to see some of the fun, here. I hope you have a good team on hand.

| improve this answer | |

I'm looking right now at an 18650 battery, and it says on the side it has a capacity of 8.14Wh.

So assuming your question is actually about supplying 11,200MWh of electricity over a period of 4 hours, you would need 11,200,000,000/8.14=1,375,921,376 batteries.

Four batteries fit into a holder measuring 80mm x 75mm x 20mm, giving a volume of 0.00012m3, so the total volume of your facility in terms of batteries alone would be 1,375,921,376/4*0.00012=41,277m3.

The question now is how much extra space you need for the charging and protection circuitry, physical housing and support, cabling, heating, cooling, ventilation and physical access gangways. Let's go for a factor of 5, which might work if we don't make the building too high; that gives a total volume of 206,385m3.

If we make the building 10m high that means a floor area of just under 144m square. Clearly a lot of guesswork involved but it gives you an idea. It's big but it seems possible assuming there's enough lithium available in the world for the batteries and you can wait long enough for them to be manufactured.

| improve this answer | |

AES is building the largest today in SD area.

  • 120 MWh/30 MW – for San Diego Gas & Electric in Escondido, California.

The AES Energy Storage systems utilizes 400,000 lithium-ion cells supplied by Samsung SDI in 20,000 modules and 24 containers.

Scale up vertically requires safety inspection of battery containers and weight management, not to mention it would be safer under-ground from environmental threats on earthquake proof platforms.

In 2012 in Ontario, Canada, costs for nuclear generation stood at 5.9¢/kWh while hydroelectricity, at 4.3¢/kWh, cost 1.6¢ less than nuclear. By September 2015, the cost of solar in the US dropped below nuclear generation costs, averaging 5¢/kWh.

India has Solar slightly cheaper than coal at 4¢/kWh but no storage.

Nuclear power in Ontario with several GW stations is still profitable according to my Nuclear Eng friend in the biz at Pickering, inspite of being only 50% efficient and heating up Lake Ontario with a 10’C rise on cooling water return to the Lake.

| improve this answer | |

Not the answer you're looking for? Browse other questions tagged or ask your own question.