Power & energy Essential term
Power MW
Power is the rate at which a grid-scale BESS charges or discharges, in megawatts (MW; 1 MW = 1,000 kW). It is capped by the power conversion system (PCS) rating and by the conductors, switchgear and transformers between the battery and the point of interconnection (POI) — not by stored energy.
Power (MW) and Energy (MWh) are sized independently, so every project is quoted as a pair: 100 MW / 400 MWh is a 4-hour system, and Duration equals energy divided by power. No power figure means anything until you name its reference point — DC terminals, inverter AC output, or POI.
Reviewed July 2026 by Sergey Syrvachev
New to BESS? Start free with the 7-email fundamentals course — no cost, no account.
What it is (precise)
Power is the instantaneous flow of energy into or out of the system — on the DC side, battery voltage times current. The PCS converts that flow to AC, and every stage downstream subtracts a loss, so the reference point governs the number: power at the inverter AC terminals exceeds deliverable power at the POI, because the medium-voltage transformer, AC collection and station auxiliaries each take a slice — typically 2 to 4 percent total. This is the AC / DC question in practice: a DC-side MW and an AC-side MW are never equal, so a bare "100 MW" with no measurement point is incomplete.
Power is bidirectional and time-limited by available Energy. A 100 MW plant can absorb or deliver 100 MW, but for how long depends on its reservoir — that ratio is its Duration, drawing on Usable energy, not the full installed cell count. Sustained power also falls below peak: temperature, state of charge and PCS thermal limits force derates, so a unit rated 100 MW at 25 C may deliver less at high SOC, at altitude, or in a hot enclosure. Charge power is often more restricted than discharge near the top of the SOC window.
Strictly, the MW figure is Real power — the part that does work and moves energy into the battery or grid. The converter itself is rated in Apparent power, measured in MVA, combining real power with Reactive power through S squared = P squared + Q squared. A PCS at full MW and unity Power factor has no reactive headroom left; when the grid code demands reactive support at full output, the MVA rating, not the MW rating, binds — which is why the converter fleet is specified a step larger than the contracted MW.
Why it matters in a real grid-scale project
Power drives the AC balance of plant and most of the interconnection cost. PCS count and rating, MV transformer kVA, collection-cable ampacity, switchgear ratings, protection settings and the operator's injection limit all scale with MW, largely independent of MWh. Two plants with identical energy but different power carry very different transformer, cabling and switchgear bills, and different short-circuit contributions. The MW the operator writes into the interconnection agreement — what you may inject and withdraw — is the hard ceiling every downstream design decision must respect.
Power also decides which markets the asset can serve. Frequency response, fast-frequency response and synthetic inertia pay for MW and ramp rate; energy arbitrage and capacity accreditation pay for MWh and Duration. The power-to-energy ratio is therefore a day-one commercial decision: a 1-hour system (100 MW / 100 MWh) chases ancillary services, while a 4-hour system (100 MW / 400 MWh) chases capacity payments and arbitrage. Grid codes also require reactive capability and fault ride-through, so the converter fleet is sized in MVA above the contracted MW.
- Unit
- Megawatts (MW); 1 MW = 1,000 kW = 10^6 W
- Sizing basis
- Independent of energy — PCS, transformer and POI limit set MW; installed cells set MWh
- Typical PCS block
- ~1-5 MVA per inverter, aggregated to tens-hundreds of MW
- DC bus voltage
- Up to 1500 VDC typical; 2000 V architectures emerging
- Common duration
- 2-4 hours today, trending longer; MW is often 1/4 to 1/2 of the MWh figure
- C-rate equivalent
- 4-hour system = 0.25C; 2-hour = 0.5C (C-rate = power / energy)
- POI vs nameplate
- POI power typically 2-4% below summed inverter rating after transformer and aux losses
- Round-trip efficiency
- Typically ~85-90% (RTE ≈ one-way²), so charge MW slightly exceeds discharge MW
- Converter sizing basis
- Apparent power (MVA), with S² = P² + Q²; MW alone does not size the PCS
- Typical PF requirement
- Commonly ~0.95 leading to 0.95 lagging at the POI (grid-code dependent)
- Interconnection standards
- IEEE 2800 (transmission-connected IBRs), IEEE 1547 (distribution)
- Safety codes
- NFPA 855 (installation), UL 9540 (system cert), UL 9540A (propagation test method), NFPA 68/69
Typical values and standards
Memorize the anchors. A utility-scale PCS block runs about 1 to 5 MVA per inverter, fed from a DC bus up to 1500 VDC (2000 V architectures are emerging), aggregated into plants of tens to hundreds of MW; the largest operating sites have passed several hundred MW. Duration today is usually 2 to 4 hours, which in C-rate terms is roughly 0.5C down to 0.25C — gentle rates chosen for cycle life, not a cell limit. Round-trip efficiency is near 85 to 90 percent, so charging MW slightly exceeds discharge MW.
Expect POI power to sit 2 to 4 percent below the summed inverter Nameplate once transformer and auxiliary losses are netted, and always confirm the reference conditions — PCS ratings are quoted at a specific ambient temperature, altitude and grid voltage and derate outside them.
In North America, transmission-connected inverter-based resources follow IEEE 2800, while IEEE 1547 covers distribution interconnection. Grid codes commonly require full active power across a power factor band around 0.95 leading to 0.95 lagging at the POI, and that requirement, not the MW headline, drives the MVA oversizing.
Power ratings interact with safety codes, because bigger blocks pack more energy behind each enclosure door. NFPA 855 governs installation and separation of stationary ESS; UL 9540 is the product safety certification for the complete system, while UL 9540A is the fire and thermal-runaway propagation test method whose data feeds NFPA 855 — the two are never interchangeable.
NFPA 68 and 69 cover deflagration venting and prevention. LFP, dominant in stationary BESS for its higher thermal-runaway onset and no oxygen-releasing cathode, is the default in high-power blocks; NMC is the higher-density contrast.
How it shows up in specs, studies and contracts
On a PCS or block datasheet, power is a matrix: rated kVA or MVA is tabulated against grid voltage, ambient temperature, altitude and power factor, with derating curves above the reference point. Ask the vendor three things: is the headline continuous or a short-time overload; is it AC- or DC-referenced; and at what temperature does it hold. A "3,600 kVA" block rated only to 40 C, or only on a higher-voltage variant, quietly delivers less in your project. The binding number is the continuous rating at your site's hottest condition.
In interconnection studies, power appears as the MW injection limit at the POI — and, increasingly, a separate charging-load limit, since a 100 MW discharge asset is also roughly a 100 MW load. Ramp-rate caps, reactive-capability requirements and short-circuit contribution all key off the MW and MVA figures. In market registration, accredited capacity can differ from Nameplate after the operator applies its own duration tests, so the revenue-bearing MW is not automatically the datasheet MW — check which figure your offtake or capacity contract pays on.
In supply and EPC contracts, guaranteed power lands in the capacity test: the plant must demonstrate the contracted MW (and MWh) at the POI. Ask at what ambient temperature, at what SOC window, at what power factor, and with auxiliary load netted or excluded. A guarantee at the inverter terminals at 25 C and unity power factor is far weaker than the same number at the POI on a hot day with reactive dispatch. The most common trap is a reference point left unstated; insist it and the conditions are explicit before signing.
Common pitfalls
Two traps catch people who know MW is not MWh. First, reactive obligations eat converter headroom: a fleet whose MVA rating equals the contracted MW cannot meet a 0.95 power-factor requirement at full output, because S squared = P squared + Q squared leaves no room for Q. Size the Apparent power above the MW, never to it. Second, the sustainable MW depends on temperature, altitude, grid voltage and SOC; a plant that meets Nameplate on a cool commissioning morning can miss it in August. Model the derated matrix, not the brochure number.
A subtler slip is comparing power figures pinned to different reference points. A DC-side rating, an inverter AC rating and a POI rating for the same plant differ by exactly the losses between them, so a vendor's AC / DC number lined up against a competitor's POI number compares nothing. Whenever you read or quote an MW, fix it to one of three places — DC terminals, inverter output, or POI — and hold every comparison there. Stating the reference every time is what separates a rigorous spec from an unsizable headline.
A 100 MW / 400 MWh plant is "a 400 MWh battery" — that one number tells you its size.
In reality: The single MWh figure leaves the system unsizable. Power (MW) and Energy (MWh) are set by different hardware: MW is limited by the PCS, transformer and interconnection; MWh by the cells installed. Duration in hours equals MWh divided by MW, so 400 MWh could be a 400 MW / 1-hour plant or a 50 MW / 8-hour plant — utterly different projects, markets and costs. Adding cells raises MWh and Duration but never the MW unless the converters and AC balance of plant are upsized too. Always quote the pair.
- Interactive: Energy Station Structure Interactive visual · bess.engineer
- Interactive: Power Factor Triangle Interactive visual · bess.engineer
- Interactive: PCS Dispatch Efficiency Interactive visual · bess.engineer
Power, in context.
The Grid-Scale BESS course covers power — and the rest of the system — from the ground up, the way it actually gets deployed.