Power & energy Essential term
Duration
Duration is how long a grid-scale battery can sustain its rated power output before its usable energy is exhausted, expressed in hours: usable energy (MWh) divided by rated power (MW). A 200 MWh / 50 MW system is a 4-hour system; so is 400 MWh behind 100 MW.
The 4-hour shape dominates today's utility-scale market, but real projects span roughly 0.5-hour frequency-response assets to 8-hour-plus energy-shifting plants. Duration — not raw MWh — is the single parameter that fixes a project's cost structure, its revenue model, and how much firm capacity the market will credit it with, so it is the first number an engineer or offtaker interrogates.
Reviewed July 2026 by Sergey Syrvachev
New to BESS? Start free with the 7-email fundamentals course — no cost, no account.
What duration actually is
Duration = usable energy (MWh) ÷ continuous rated power (MW). The two ratings are independent design choices: a project specifies how much Power it can push to the point of interconnection (POI) and, separately, how many hours of that power it can hold.
Both should be stated at the POI, net of losses, so a DC Nameplate duration and the delivered AC duration differ once you account for power-conversion-system (PCS) efficiency, transformer losses, auxiliary loads (HVAC, controls), and degradation over the warranty term. A container fleet quoted at 4.0 hours DC at beginning of life can deliver noticeably less at the POI at end of warranty.
Because lithium iron phosphate (LFP) cells are not cycled from 0 to 100% state of charge in practice, Usable energy is the Energy available within the operating SOC window, not the gross installed energy. Duration is therefore a delivered-performance number, and a careful engineer always states whether a figure is beginning-of-life or end-of-life, AC or DC, and at what reference temperature.
Note also that duration is defined at rated power: the same 4-hour system discharges for roughly 8 hours at half power. Duration is the inverse of the discharge C-rate, so a 4-hour system runs at about 0.25C and a 1-hour system at 1C.
Why it drives the whole project
Duration sets the size of the battery relative to the inverters, and the battery is the dominant cost. Lengthening duration mostly adds cells, racks, and containers — pure energy cost — while power capability is set largely by the PCS and transformers, which are sized in MVA (Apparent power) to cover both Real power and Reactive power, not MW alone.
Duration does not touch that AC-side sizing at all. This is why a 4-hour system costs far more per installed MW than a 1-hour system but much less per installed MWh: the fixed power-conversion and balance-of-plant costs spread over more energy. As a rough shape, doubling duration typically raises total capex well under 2x, because only the DC block scales.
Duration is also dictated by the revenue model and market rules. A capacity-market or resource-adequacy obligation commonly requires a 4-hour sustained discharge to count as firm capacity — CAISO's resource-adequacy framework made the 4-hour battery the default US shape. An arbitrage or solar-shifting asset may justify 6 to 8 hours; a fast-frequency-response or ancillary-services asset may need only 0.5 to 1 hour.
Pick the wrong duration and the asset either carries energy it has no high-value hours to discharge into, or runs empty before those hours arrive. Capacity accreditation (ELCC-type methods) also derates shorter durations, and the marginal credit of any duration falls as storage penetration on the system grows.
duration (h) = energy (MWh) ÷ power (MW)
A 100 MW / 400 MWh system is a 4-hour battery. Power and energy are sized independently, so duration is a design choice, not a fixed property.
- Definition
- Usable energy (MWh) ÷ rated power (MW), in hours — 400 MWh / 100 MW = 4-hour
- Most common grid-scale shape
- 4-hour (≈0.25C discharge), the default US resource-adequacy shape
- Typical range
- ~0.5 h (fast frequency response) to 8 h+ (long-duration energy shifting)
- C-rate mapping
- 1 h ≈ 1C, 2 h ≈ 0.5C, 4 h ≈ 0.25C, 8 h ≈ 0.125C (inverse of discharge C-rate)
- Cost scaling
- Longer duration → $/kW rises, $/kWh falls; only the DC block scales, PCS MVA is fixed
- Capacity-market convention
- 4-hour sustained discharge is the common firm-capacity threshold; shorter durations get derated credit
- Reference basis (always state)
- AC vs DC, BOL vs EOL, ambient temperature, SOC window, at rated power
- Partial-power behaviour
- A 4-hour system runs ~8 h at half power — duration is defined at rated power
- Degradation effect
- Hours at rated power shrink each year unless the DC block is overbuilt or augmented
- Datasheet check
- Ask AC or DC, BOL or EOL, and site ambient vs 25 °C before accepting an hours claim
- Verification
- Commissioning capacity test: full discharge at rated power under contractual conditions
- Relevant standards
- UL 9540 (system certification), UL 9540A (fire-propagation test method), NFPA 855 (installation), NFPA 68/69 (explosion protection)
Typical values and standards
Utility-scale durations cluster at 2 and 4 hours, with 4-hour now the dominant new-build shape in the US and 2-hour still common in ancillary-heavy markets such as ERCOT and Great Britain; 6-to-8-hour projects are appearing in high-solar systems, and procurement programs for 8-hour-plus long-duration storage exist in several jurisdictions.
The C-rate mapping is fixed by arithmetic: 1 hour ≈ 1C, 2 hours ≈ 0.5C, 4 hours ≈ 0.25C, 8 hours ≈ 0.125C. Shorter durations mean higher C-rates, which raise thermal and degradation stress on the cells and push harder on HVAC sizing and warranty cycling limits.
There is no standard that certifies "duration" as such — it is a derived ratio, verified at commissioning by a full-discharge capacity test at rated power under contractual conditions.
The standards enter through the consequences: higher C-rates feed the UL 9540A fire-propagation test data that NFPA 855 uses for spacing and hazard mitigation, while longer durations pack more energy per enclosure, raising the stakes for NFPA 68 deflagration venting or NFPA 69 prevention systems. UL 9540 remains the system-level safety certification regardless of duration, and IEC 62619 covers cell and battery safety for industrial applications.
Where you meet it in specs, studies and contracts
On a datasheet, duration is rarely printed as one honest number, so trace the chain before you accept it: cell energy times count gives DC nameplate; the SOC window gives usable DC; PCS efficiency, transformer losses, and auxiliary consumption give usable AC at the POI.
Three questions catch most padding — is this AC or DC, is it beginning-of-life or end-of-warranty, and is it at the site design ambient or the fresh 25 °C lab figure. In the interconnection study only the MW at the POI matters, because that sets fault duty and thermal limits; the offtaker and the capacity market care about the MWh, and therefore the hours, behind that MW.
In contracts, duration shows up as a guaranteed usable-energy schedule at each contract anniversary, backed by an augmentation plan: because cells fade, a project that must deliver 4 hours in year 15 either oversizes the DC block on day one (overbuild) or adds racks over time.
Read four clauses together — the warranty energy-retention curve, its cycling and temperature assumptions, the augmentation trigger, and the commissioning capacity-test protocol (power level, SOC limits, ambient, rest periods) — because those, not the headline, define what "4-hour" means legally. Confirm the C-rate assumed in the warranty matches the dispatch the revenue model expects: a 2-hour asset earning frequency markets cycles differently from a 4-hour shifter, and a mismatch starves revenue or voids coverage.
Common pitfalls
The classic trap is mixing bases: quoting DC beginning-of-life nameplate hours against an AC end-of-life obligation. A "4-hour" DC BOL system can be a roughly 3.5-hour AC system at the POI after conversion losses and a few percent of degradation, which is a contract breach if the obligation was written at the POI.
Related traps: "up to X hours" marketing figures measured at reduced power; short-term overload ratings presented as continuous capability; and forgetting that round-trip efficiency means the grid must supply more energy — and often more clock time — to recharge than the discharge delivered.
Finally, duration is not static. Degradation shrinks the hours available at rated power every year unless augmentation restores them, high ambient temperatures can derate both power and usable energy, and auxiliary loads consume energy even while the plant idles. A duration figure without its reference conditions — AC or DC, BOL or EOL, temperature, SOC window, power level — is not an engineering number; it is a slogan. State the basis, every time, and demand it of every vendor number you are handed.
A bigger battery (more MWh) automatically means longer duration — and 100 MW with 400 MWh is a "2-hour" system because 400/100/2 feels right.
In reality: Duration is the ratio of energy to power: 400 MWh ÷ 100 MW = 4 hours, full stop. Adding MWh while proportionally adding PCS MW keeps duration constant; duration only lengthens when energy grows relative to a fixed power rating. That ratio — not raw MWh — is what markets accredit, warranties guarantee, and the commissioning capacity test verifies, which is why it is the first line an offtaker or interconnection engineer reads.
- Sizing a BESS: Power, Energy, Degradation & Augmentation Article
- Interactive: The Duck Curve Interactive visual · bess.engineer
- Interactive: Cell Losses & Efficiency Interactive visual · bess.engineer
Duration, in context.
The Grid-Scale BESS course covers duration — and the rest of the system — from the ground up, the way it actually gets deployed.