Levelising the cost of stored energy: Why $/kWh just doesn’t cut it

The expected battery cycle lifetime and corresponding LCOE for the lead acid battery option discussed above, given as a function of DoD. The blue curve represents the expected number of cycles and the red squares represent the corresponding LCOE.

Solar Choice

With the range of energy storage options available to Australians steadily growing, choosing the ideal battery storage solution for a home or business can be difficult. Even if you know roughly what size energy storage system that you are in the market for, consumers need a metric with which they can compare solutions regardless of a particular product’s specifications or chemical composition.

The most important thing is that the selected solution provides electricity reliably and when needed at the best value price. In order to meaningfully compare value, customers (and installers) have to understand three key performance characteristics: nominal cycle lifetime, recommended depth of discharge (DoD) and round-trip storage efficiency.

$/kWh is not useful in apples-to-apples comparisions

Comparing energy storage options can be complicated, so it’s not uncommon for media outlets to simplify the conversation by using the $/kWh metric when talking about the cost of energy storage. The calculation is pretty straightforward: Divide the cost of the system by its nominal storage capacity at full charge. While it may seem useful in comparing batteries across brands and models, $/kWh is at best inaccurate and at worst misleading because it does not account for the following three crucial factors:

  • Battery cycle life: Two batteries may have the same $/kWh cost, but this means very little if one lasts only one cycle and the other lasts 5,000 cycles! Cycle life is usually given in terms of the number of cycles it would take for the true battery storage capacity at full charge to reach 80% of its nominal capacity, given as a function of DoD (discussed below).
  • Depth of discharge (DoD): DoD is a measure of the portion of battery capacity that has been used. If, for example, a battery bank with 5kWh of nominal capacity has 2.5kWh stored within it, the current DoD is 50%. If it has 1kWh remaining, then DoD is 80%. In most cases it is not advisable to drain a battery of its full capacity, which can potentially cause permanent damage to its components. This is why batteries are ordinarily specified a recommended DoD for their nominal cycle lifetime (e.g. 3,000 cycles at 50% DoD) and the expected battery life tends to be shorter as DoD is increased (see the chart below).
rsz_expected-average-cycles-vs-dod
An example graph of battery cycle life as a function of DoD [Image Source: Electropaedia]
  • Round-trip storage efficiency: You can think of round-trip efficiency as the percentage of energy it takes a battery to store and release energy. For example, if you feed 1kWh of electricity into a battery but get only 0.8kWh back out, then the round-trip efficiency is 80%. The loss can be due to a number of factors, including heat or other system inefficiencies. Round-trip efficiency is actually highly dependent on a battery’s DoD, but on battery spec sheets you will ordinarily see only the average.

So as we can see, $/kWh is not a fantastic measure to use when comparing various types of battery types and models. To do this effectively, we need to use a better metric – levelised cost of energy (LCOE) – which gives us a much better idea of the true cost of a given energy storage solution.

LCOE for energy storage: Some sample calculations

In the table below I’ve listed three hypothetical batteries with different specifications for the purpose of showing how the factors discussed above can impact energy storage system value (and appearance of value) – and to give the buyer as much insight as possible before deciding on a system. This analysis offers a more accurate view as to which battery stores energy at the lowest cost compared to the others; in other words, which one has the lowest LCOE. Once they know the specifications of the products they’re looking at, anyone can use this or a similar approach to meaningfully compare the relative value of their choices.

Example battery specifications

Battery-storage-example-specifications

We can use the figures from the table above to calculate LCOE values for the three batteries by using the below equation. Please keep in mind that this is only intended to provide a rough estimate, and assumes that battery capacity does not degrade towards 80% as the number of cycles increases. If a more accurate calculation is required, the formula can be modified accordingly.

LCOE-for-energy-storage-calculation

Summary Table: LCOE for three sample batteries

Battery-storage-example-LCOE

As you can see, in this particular case the least expensive battery option does not appear to offer significantly more value than the slightly more ‘expensive’ lead acid option. With this, it should be fairly clear why LCOE should be the preferred metric, but unfortunately this is not yet something that manufacturers generally provide as a matter of course on their spec sheets, marketing materials or product labeling.

How DoD impacts LCOE

As the graphs above suggest, a battery’s cycle life is highly dependent on the DoD for all of those cycles. With the majority of batteries, higher average discharge depth means shorter usable lifetime. The table below illustrates the impact that DoD can have on LCOE, using the lead acid battery mentioned above as an example.

The expected battery cycle lifetime and corresponding LCOE for the lead acid battery option discussed above, given as a function of DoD. The blue curve represents the expected number of cycles and the red squares represent the corresponding LCOE.
The expected battery cycle lifetime and corresponding LCOE for the lead acid battery option discussed above, given as a function of DoD. The blue curve represents the expected number of cycles and the red squares represent the corresponding LCOE.

This chart highlights the idea that there is an optimal DoD operating point where the LCOE (and therefore the real cost of the battery) is minimised. Charts like these should be something that all consumers look at before deciding on a battery option – and should probably also be used as a guideline for best practice usage and design of an energy storage system. The DoD for this particular battery should be maintained at around 30% to ensure a long and productive life, or alternatively, the system should be designed such as to keep DoD around 30%.

Other factors that may influence battery choice

In addition to the three discussed above, there are many plenty of other factors to keep in mind in battery selection, including maintenance requirements, toxicity, safety and disposal. But as the examples above have highlighted, cycle life, depth of discharge and round-trip storage efficiency are the most important metrics to use to narrow down your battery storage options to those that offer the best value.

 

Source: Solar Choice. Reproduced with permission.

Comments

15 responses to “Levelising the cost of stored energy: Why $/kWh just doesn’t cut it”

  1. Khalid Amir Mahmoud Abdulla Avatar
    Khalid Amir Mahmoud Abdulla

    Great article. Totally agree with the importance of levelised costs. However, I would prefer if we called this levelised cost of storage. It is not really the cost of a kWh; but the cost of moving a kWh from a point in time when its available to one where its needed.

    1. Mike Dill Avatar
      Mike Dill

      The numbers above reflect the storage cost, for ‘free’ electrons. We need to add the base cost of the KWH to find a number closer to the ‘gross’ cost.

      1. Jacob Avatar
        Jacob

        No I think we are intelligent enough to add LCOS + LCOE ourselves.

        The LCOE from solar panels is between 15-22c/kwh for rooftops. Utility scale PV is a lot cheaper.

        The LCOS is still a mystery because nobody has tested the Powerwall to destruction yet. It is between 12-25c/kwh.

  2. Jo Avatar
    Jo

    I agree. The new metrix is good and important.
    But the name ‘ levelised cost of energy (LCOE)’ is misleading, because it sounds like levelised cost of energy (pun intended). The latter is a much more complicated formula because it regards also the time value of money ($100 today is worth more than the promise of $100 in 10 years).

  3. Jacob Avatar
    Jacob

    LCOS should be the name. Levelized cost of storage!

    Good info though, regarding battery life. There should be an option on our phones to only charge up to 80% and stop.

    Keeping the battery at 100% is not good for battery life.

    1. Diego Matter Avatar
      Diego Matter

      I agree.

      And an option to discharge only to 20% DoD

      1. Jacob Avatar
        Jacob

        That cannot be an option. Unless you want to turn off your phone when the battery gets to 20% charge remaining.

        1. Diego Matter Avatar
          Diego Matter

          Make the battery 20% bigger and let the operating system shut the phone down at 20% DOD telling the user that the battery is empty.
          Battteries would last more than 2 years.

  4. Jacob Avatar
    Jacob

    The figures above are pessimistic. The 7kwh Powerwall can be cycled 5000 times, not 2000 times as this article suggests.

    1. John Avatar
      John

      The pricing and storage capacity looks similar doesn’t it? =)
      All the batteries in the article are just generic examples – not reflective of Tesla’s or any other manufacturers products.

  5. Gerry Avatar
    Gerry

    Great article! There is however, one additional layer of complexity which is very important but difficult to model. That is: the average discharge rate. Batteries are rated for a specific amount of energy in Amp hours based on the rate at which they are discharged. So if a battery is discharged at its C3 rate (the amperage at which the battery would fully discharge in 3 hours) we would expect to have less energy available than if we discharged that same battery at its C10 rate (the amperage at which the battery would fully discharge in 10 hours). This means the real DOD is a constantly moving target as the battery experiences different discharge rates, which is why it is hard to pin down the actual amount of energy and cycles that a battery can achieve (at least without modelling).

    1. Gordon Avatar

      Is anyone still installing Lead-acid batteries? 😉 The reduction in capacity at higher discharge rates is due to Peukert’s Law- very important for Pb-acid, but pretty safe to ignore with Lithium batteries, which are generally rated at C1, a much higher discharge rate than domestic storage will encounter.
      92% round trip efficiency seems rather poor, I’m seeing (and others I know of as well) 99% with 2.5 year old LiFePO4 cells.

      1. Chris Baker Avatar
        Chris Baker

        I’m also seeing 99% efficiency of energy in and out of LiFePO4 cells, when I measure the DC energy exchange right at the battery. This is measuring the energy loss in converting electrical energy in the wires to chemical energy in the battery. I expect though in the examples in the article they are also including the losses in charge controllers and inverters so as to account for the full round trip cost, so I think the 92% quoted is probably on the money.

  6. John Avatar
    John

    Thanks all for the comments. You’re right it should be LCOS – I’ll see if we can get that corrected and reposted.

    1. Diego Matter Avatar
      Diego Matter

      And we also need to look at the kW a battery-SYSTEM can deliver and the kWh needed per day. The Tesla PowerWall is only delivering 2kW and max 7 kWh/day (at what DOD?), therefore you probably will need two of them, what in turn is rising the investment and therefore the LCOE.

Get up to 3 quotes from pre-vetted solar (and battery) installers.