Battery storage: calculate usable lifetime energy throughput
A battery can retain a specified share of its initial capacity while delivering much less lifetime energy than a presentation implies. Capacity describes an energy quantity at a defined test point. Throughput accumulates energy moved over time. A warranty may constrain either measure, both measures, or a separately defined operating regime. Research should preserve those three objects instead of translating a cycle-life headline directly into revenue-generating megawatt-hours.
The deliverable is a yearly energy ledger with a separate contract panel. It should show what the system could discharge under explicit operating assumptions, what it actually discharged where metered evidence exists, and what the supplier has promised under specified conditions. None of these automatically establishes an asset’s future profitability.
Locate the measurement boundary
The DOE 2022 Grid Energy Storage Technology Cost and Performance Assessment discusses cycle life, depth of discharge, calendar life, augmentation and efficiency. Locate the technology-specific performance sections and Section 6.1.3, printed pages 123–124, for the distinction between depth of discharge and cycles expressed on a 100%-depth basis, as well as Chapter 6’s levelized-cost assumptions, rather than treating a summary chart as a universal battery specification. It is an assessment using stated assumptions, not a guarantee for a particular installed project.
Start with the equipment schedule and commissioning test. Record rated DC cell energy, usable DC energy, AC energy delivered at the contractual meter, power rating, permitted state-of-charge range, temperature conditions and test duration. Ten megawatts is power; ten megawatt-hours is energy. A two-hour power rating does not tell you whether auxiliaries and conversion losses have already reduced the energy available to the customer.
Choose one output boundary for the worksheet. Here, usable discharge energy is measured at the AC delivery meter after discharge-side losses. Round-trip efficiency compares that discharged energy with compatible AC charging energy. Consequently, do not multiply AC discharge energy by round-trip efficiency again. That would deduct losses twice. If only cell-level specifications exist, keep the AC conversion unresolved until a supported loss bridge is available.
Reproduce a three-year operating case
All figures below describe a fictional system. Assume initial usable capacity before the permitted discharge window is 10 MWh. The worksheet applies an 80% usable window to that stated capacity, with 300 actual charge–discharge cycles each year, each completing the 80% usable window rather than a 100%-depth equivalent full cycle. Assume the compatible capacity basis falls to 9.5 MWh in Year 2 and 9 MWh in Year 3, no augmentation occurs, and the discharge-meter boundary is consistent throughout.
| Year | Capacity basis, MWh | Window | Actual cycles | Discharged MWh |
|---|---|---|---|---|
| 1 | 10.0 | 80% | 300 | 2,400 |
| 2 | 9.5 | 80% | 300 | 2,280 |
| 3 | 9.0 | 80% | 300 | 2,160 |
| Total | — | — | 900 | 6,840 |
The calculation is capacity × window × actual cycles for each year, followed by addition across years. On a fixed initial 10 MWh reference, 6,840 MWh corresponds to 684 equivalent full cycles; the 900 actual cycles are not 900 equivalent full cycles. Using initial capacity for all 900 actual cycles, each completing the 80% window, would produce 7,200 MWh, overstating this scenario’s output by 360 MWh, or 5.26% of the modeled total. The error arises from ignoring the changing energy available per cycle, not from a change in electricity prices.
Assume a constant 88% AC-to-AC round-trip efficiency. Compatible charging energy is 6,840 / 0.88 = 7,772.73 MWh. The simplified energy difference is 932.73 MWh. At an assumed charging price of 40 currency units per MWh, charging cost is 310,909.09. It is not 6,840 × 40: that calculation purchases only the energy later delivered and omits the assumed losses.
Next assume equipment, installation, maintenance and replacement costs allocated to this three-year study total 1,200,000 currency units before charging energy. The undiscounted cost per discharged MWh is (1,200,000 + 310,909.09) / 6,840 = 220.89. This deliberately simplified metric excludes financing, taxes, residual value and revenue timing. Name those omissions beside the result; do not label it an official levelized cost calculation.
The example uses one representative capacity per year. For actual evidence, monthly or test-to-test intervals may be necessary. A beginning-of-year capacity multiplied by a full year’s cycles can overstate output when degradation occurs throughout the year. Replace the approximation only when the finer inputs share a compatible boundary.
Keep capacity retention and throughput guarantees separate
Open the actual supply agreement or warranty and extract its defined terms. A capacity-retention floor might be measured after a prescribed test and subject to temperature, use and maintenance requirements. A throughput ceiling may define cumulative warranted operation. Passing a retention test does not establish unlimited further warranted throughput, and reaching a throughput ceiling does not prove the battery physically stops working.
For a separate hypothetical contract, suppose protection ends at the earlier of ten years or 20,000 MWh of warranted throughput. At 2,500 MWh annually, the throughput condition is reached after eight years under a constant-use assumption. A presentation showing ten years of retained capacity cannot silently extend the contractual protection to Year 10. Conversely, at 1,000 MWh annually, the calendar condition arrives first. These are invented contract terms, not industry-standard warranty limits.
Check whether throughput counts charged energy, discharged energy, both directions, or equivalent full cycles based on initial capacity. Translate only using the agreement’s definitions. If the public warranty summary omits that definition, mark contractual coverage unknown rather than select the most favorable interpretation.
Test a different mechanism: augmentation and idle aging
Suppose 2 MWh of new cells are added at the start of Year 3. They are a new age cohort with a new cost and potentially different warranty coverage. Do not rewrite the original cells’ capacity back to 10 MWh. Keep original and added energy in separate rows, apply compatible operating limits, and include commissioning downtime and integration constraints where evidenced.
An alternative system may perform fewer cycles yet age during long idle periods. In that case, lower cumulative throughput does not guarantee proportionally longer service. Calendar aging and operating exposure answer different questions. A reported lifetime-cycle estimate cannot determine both without the relevant technical and contractual evidence.
These cases also prevent a misleading comparison between two operators. One might preserve capacity by using a narrower operating window, sacrificing annual output. Another may support output through paid augmentation. The same reported retention percentage can accompany very different commercial energy and cash requirements.
Use an evidence-ready verification table
| Check | Obtain | Recalculate or resolve |
|---|---|---|
| Energy boundary | Commissioning report and meter diagram | DC, AC and auxiliary coverage |
| Capacity history | Dated test results and test conditions | Comparable retention series |
| Operating exposure | Interval discharge and charge meters | Output, input and equivalent cycles |
| Guarantee | Executed terms and amendments | Calendar, throughput and exclusions |
| Augmentation | Added-cell records and invoices | New cohorts, cost and downtime |
Treat an investor slide and a contract as different evidence roles. If the slide’s capacity exceeds the test result, first check whether it reports nameplate DC energy instead of usable AC energy. If two tests disagree, inspect temperature, discharge rate and measurement point before averaging them. Preserve missing months as gaps; a meter failure is not evidence of zero cycling.
Update the ledger after a comparable capacity test, a changed dispatch regime, new augmentation, a warranty amendment or persistent divergence between charging and discharge meters. State exactly which row changes and why. The useful conclusion is a reproducible operating range and a contractual coverage boundary, not an assumed future selling price or investment return.
Audit the cycle counter against the contract
A final miniature check concerns the cycle denominator. Suppose a system discharges 900 MWh during an interval. Dividing by an initial 10 MWh capacity gives ninety equivalent full cycles; dividing by a later 9 MWh test gives one hundred. The meter observed exactly the same output. The apparent difference comes from the reference energy, so neither counter can be applied to a warranty without its definition.
Ask for the accumulator’s start value, end value, reset events and configuration. A controller replacement can restart a displayed counter without erasing historical throughput. Retain the earlier and later device records and join them at the documented replacement event. Where only a screenshot exists, record its timestamp and cumulative scope before calling it annual operation. This check prevents a software counter change from masquerading as a new battery life or a sudden jump in physical usage.
If the separate augmentation scenario adds 2 MWh in Year 3 and the same window and 300-cycle assumptions apply without any power or integration constraint, the added cohort contributes 480 MWh. Year 3 becomes 2,640 MWh instead of 2,160. Keep that conditional increment and the added cost together; it is not evidence that old cells recovered capacity.
FAQ
Does 90% retained capacity mean 90% of original lifetime revenue?
No. Revenue also depends on dispatch, prices, availability and contractual access. Capacity is only one physical input.
Should efficiency be deducted from delivered AC energy?
Not again when that energy is already measured after discharge losses. Use round-trip efficiency to reconcile compatible charging input.