Technological Complications of Energy Storage

Date: 21 May 2012 | posted in: Energy, Energy Self Reliant States | 7 Facebooktwitterredditmail

In the long run, there’s no avoiding energy storage for a 100% renewable energy society.  The two major sources of renewable power are wind and sun, and they are either fickle or reliably not available at night.

The problem is that the simplest energy storage option for electricity is batteries, and this image from Wikipedia (hat tip to Robert Rapier) illustrates a significant technical barrier: our simplest option is also among the least energy dense material we have.

There are two likely paths to a 100% renewable energy future in these circumstances: mass distribution of low-density, low-cost storage or higher density storage.

In some respects, we’re already moving along the first path.  Widespread availability of battery-powered iPads and laptops has led to great strides in greater energy density of batteries and lower cost. The following chart (used in our Democratizing the Electricity System report) illustrates the changes in the past 15 years.

Electrified transportation is the next iteration, using batteries that are orders of magnitude larger (e.g. a Nissan Leaf battery with 23 kW-hour capacity has 300 times the storage capacity of a Macbook Pro laptop battery).  These are 1st generation commercial batteries, with enormous improvements in capacity and cost likely.  Furthermore, with hundreds of millions of cars, the sheer storage capacity of the U.S. vehicle fleet will be tremendous (over 4 billion kilowatt-hours) as electric vehicles become the drive train standard.  And a recent study has shown that the storage capacity of 2.1 million vehicles can enable an additional 10 gigawatts of wind power on the grid (in the Northwest).

The Germans, ever the clean energy integration leaders with over 15% of their electricity sourced from wind and solar, have also looked at electricity to hydrogen storage (a look at the above chart suggests the energy density of hydrogen has some advantages).  While not as efficient as batteries (two-thirds of power is lost, compared to 10-20% round trip for batteries), the resulting hydrogen can be used in natural gas power plants to provide backup power or piped into the natural gas network for building heat.  It’s not only flexible but it also could be useful because wind power in particular can peak during periods of low electricity use (nighttime).

The good (or bad, depending on your perspective) news for the U.S. is that renewable energy is such a small fraction of total electricity generation that energy storage isn’t yet necessary in any significant quantity.  Existing power plants have sufficient spare capacity to fill the gaps left by variable renewables.  While this state of affairs doesn’t endear the U.S. power industry to environmentalists, it does mean there is time to see storage technology improve.

I’m optimistic.


7 Responses

  1. Peter Breithaupt
    | Reply

    impressive overview of electrochemical options, but unfortunately or better say positively it only tells you half the story. Energy storage will become a mediator linking communities at distributed power storage centres.
    2 technology options are emerging which already overcame the issues you mentioned.
    1. electrochemical conversion at high temperatures (MIT, Sadoway, see:
    2. electromechanical storage, i.e. flywheels (see
    Rather than by individuals, investments in energy storage will be done by local communities either through SPV’s or existing cooperatives.

  2. Maury Markowitz
    | Reply

    Just one important note about the graph, it does not included the addition of oxygen from the air. If you factor that in, all the dots move quite far down and to the right.

  3. Bob Arrington
    | Reply

    Electrical energy is abundant in nature.

    We have:

    Electrical energy from wind and solar.

    We do not have:

    Electrical energy direct from lightning (weather created discharge of electrical potential)
    The ability to effectively store electrical energy in huge quantities such as needed to put on grids or at continuous high voltage.

    We have:

    Electrical energy storage in less than efficient electrical power to pump water from low level to high level and water flow through turbines to regenerate electricity (Georgetown Xcel facility). Excess generation is slowly stored.
    Batteries (not up to grid capacity)
    Minor use of superconductivity for specific applications.(very expensive to maintain cryogenic temperatures and room temperature application still in research).

    Could have:

    Using graphane (adaptation of graphene) better ultracapacitors can be made. Capacitors store electrical power and can be charged rapidly and many more times than batteries. This material also appears to be free of electrolytic fluid and thus higher voltages. However, using multiple capacitors in series solves voltage problems. Now because capacitors charge fast, the electrical energy of storms, in gigawatts could be captured; and wind and solar could be stored. (graphane) (read comments also) (superconductivity comes into play also)

  4. John Farrell
    John Farrell
    | Reply

    @Peter – excellent point. I’ll have to add more on those options.

    @Maury – I’d love to know more about what you mean. That sounds out of my area of expertise, but very relevant.

  5. Bob Carver
    | Reply

    Batteries have a serious problem: they don’t work at low temperatures. In fact, capacitors do not suffer from this problem. Moreover, batteries can only go through a charge-discharge cycle a few thousand times, while capacitors likely can do so into the millions of cycles. And, capacitors equivalent to the storage capacity of Li-ion batteries already work in the lab and should be appearing on the market within the next year. The only disadvantage capacitors have at the present time appears to be that they weigh more per unit charge than Li-ion batteries. But, that isn’t a drawback for fixed storage.

  6. Maury Markowitz
    | Reply

    Hey John, missed that ping there.

    The issue is that li-ion looks bad because it is self-contained. Compared this with hydrogen gas, which requires oxygen. The graph is only listing the fuel, not the oxidizer, so only one part of the weight is being accounted for in the energy density.

    If you want a practical example, consider the difference in size between a 747 and a Saturn V. The Saturn had to carry both the fuel and the oxygen, which greatly increased the size and weight of the stack.

Leave a Reply

Your email address will not be published. Required fields are marked *

18 − 9 =