Four-hour BESS projects are gaining project pipeline share over 2-hour systems in 2026 because the marginal capex buys a wider evening arbitrage window and a 1.5-percentage-point IRR uplift in the German market, where 4-hour units deliver 13.7% unlevered IRR at 2026 COD against 12.2% for 2-hour systems [S3]. The headline cost is 34% higher capex (€935k/MW vs €700k/MW), which Modo Energy's Q1 2026 outlook confirms is more than offset by the additional energy throughput [S3].
Duration is set by the energy-to-power ratio: usable kWh divided by rated discharge kW, and the same 1 MWh of cells can be sold as a 1 MW / 1 MWh (1-hour), 500 kW / 1 MWh (2-hour), or 250 kW / 1 MWh (4-hour) system [S1]. Picking the wrong duration is a sizing failure, not a chemistry failure, and it locks in the wrong C-rate before the first cell is ordered.
How duration is defined and why kW vs kWh trips up project teams
BESS duration is the approximate number of hours a system can discharge at its rated power before stored energy is depleted: usable energy capacity (kWh) divided by rated discharge power (kW), with 261 kWh at 125 kW working out to 2.09 hours [S1]. That ratio also sets the C-rate on the DC bus, so a 2 MWh / 1 MW system runs at 0.5C for two full hours of output [S2].
Common 1-hour, 2-hour, and 4-hour reference configurations include 500 kW / 500 kWh and 1 MW / 1 MWh (1-hour), 100 kW / 200 kWh, 500 kW / 1 MWh, and 1 MW / 2 MWh (2-hour), and 250 kW / 1 MWh (4-hour) [S1]. A 100 kW / 400 kWh battery stores enough energy for four hours at its rated output, but it cannot offset a 400 kW instantaneous demand spike because its power electronics cap the discharge at 100 kW; a 400 kW / 400 kWh unit can offset the spike but only for about one hour at full rated output [S1].
Decision matrix: 2-hour vs 4-hour BESS in 2026 economics
The two configurations solve different revenue problems, and the gap widens sharply between 2026 and 2030 as ancillary service markets saturate. In Germany, ancillary services account for 55% of BESS revenues in 2026, with FCR and aFRR combining to a 4.5 GW market; once connected battery capacity exceeds that, wholesale trading becomes the primary income stream and wholesale arbitrage reaches 95% of revenue by 2030, stabilising around €125k/MW [S3]. A 4-hour system can shift more energy per cycle and captures a wider day-ahead spread, which is the structural reason 2-hour near-term revenues of €240k/MW halve by 2030 [S3].
On capex, 4-hour German builds run €935k/MW against €700k/MW for 2-hour, a 34% premium that returns 13.7% unlevered IRR vs 12.2% under Modo Energy's central assumptions for 2026 COD [S3]. Round-trip efficiency for a modern LFP system sits at 88–92% at the AC terminals, and the losses concentrate in the PCS and thermal loop rather than the cells [S2].
The decision breaks down to four criteria, with the German 2026 numbers from [S3] anchoring each row:
1. Capex per MW: 2-hour €700k/MW vs 4-hour €935k/MW (34% premium for 4-hour) [S3].
2. Unlevered IRR at 2026 COD: 2-hour 12.2% vs 4-hour 13.7% [S3].
3. Revenue durability: 2-hour near-term €240k/MW halves by 2030 as ancillary markets saturate, with wholesale arbitrage stabilising near €125k/MW; 4-hour systems capture wider day-ahead spreads and hold revenue share better through the 2030 transition [S3].
4. Offtake fit: physical tolling dominates German offtake, with seven of nine 2025 deals fixing 70–100% of capacity for 5–10 year tenors and unlocking gearing up to 85% [S3].
A 50% fall in gas prices cuts day-ahead revenues by 37%, so 4-hour arbitrage-heavy stacks carry more commodity exposure than 2-hour ancillary-heavy stacks, and overbuild compresses 2030 revenues by 17% across both durations [S3].
Use case routing: which duration fits which load profile
Two-hour BESS remains the workhorse for commercial and industrial peak shaving because the daily 15-minute peak window, plus ramp and recovery, fits inside a two-hour discharge envelope, and demand-charge reductions of 30–50% are achievable [S2]. Time-of-use arbitrage on a typical 500 kW / 2 MWh rig spans the full evening rate window on a 2-hour configuration [S2].
Four-hour BESS is the right pick when the revenue line depends on shifting solar generation into the evening peak, providing microgrid resilience, or capturing deep day-ahead spreads in markets where renewables are growing fast, conditions Germany is forecast to hit with renewable generation rising 150% to 695 TWh by 2040 and demand climbing 70% to over 1,000 TWh through electrification [S3]. German BESS capacity scales from 5 GW at end-2026 to 40 GW by 2040, with 4-hour+ systems rising to 80% of the fleet [S3].
One-hour BESS still fits short-duration services: FCR-style frequency response, renewable smoothing, and power support where high discharge power matters more than long discharge duration, but it should not be selected for C&I peak shaving simply because the kWh number looks small [S1]. The first sizing question is always how long the site's actual peak lasts.
Component sizing and integration constraints that lock in duration
Every commercial BESS is built from the same five blocks, and three of them set a hard floor on duration [S2]. LFP cells in a typical C&I rack store 100–400 kWh and the BMS keeps a 10-year LFP system above 80% rated capacity at end of life, which is the chemistry choice that delivers long cycle life, thermal stability, and lower fire risk versus older NMC [S2].
The power conversion system is a four-quadrant inverter that converts DC battery power to grid-quality AC, and the kWh-to-kW ratio on the DC bus sets the C-rate directly [S2]. Liquid-cooled containers maintain cell temperatures at 25–35°C, which extends cycle life and makes the duration assumption hold across a 10–20 year operating horizon [S2]. The energy management system is the application layer that dispatches the battery against tariffs, weather, building loads, and grid signals minute by minute, and a weak EMS can leave 30–60% of project value on the table even with perfect hardware [S2]. For projects that need finer dispatch control, the same architecture pattern shows up in adjacent battery-powered systems like AGV opportunity charging vs battery swap, where the power electronics and charging profile set the usable duty cycle.
Limitations, failure modes, and project risks for each duration
No duration is risk-free. The 2-hour German BESS near-term revenue of €240k/MW is forecast to halve by 2030 as ancillary markets saturate, so projects banked on FCR and aFRR revenue carry concentration risk once connected battery capacity exceeds the 4.5 GW combined FCR plus aFRR market size [S3]. A 4-hour system captures more wholesale spread, but that revenue is commodity-exposed: a 50% fall in gas prices cuts day-ahead revenues by 37%, and overbuild compresses 2030 revenues by 17% across both durations [S3].
Over 700 GW of battery storage sits in the German grid queue against just 2.5 GW connected, so queue risk and connection timing are first-order project risks, and Flexible Connection Agreements can reduce IRR by up to 5 percentage points on either duration [S3]. Grey co-location delivers 13.7% BESS-only IRR, in line with standalone returns, but it improves queue position under Germany's proposed maturity-based connection system, which is a non-financial reason to consider co-location even when the standalone numbers tie [S3].
On the equipment side, real-world discharge time diverges from the kWh divided by kW label because state-of-charge limits, system efficiency, temperature, battery degradation, and auxiliary consumption all erode the headline number [S1]. Round-trip efficiency of 88–92% at the AC terminals is the operating envelope, not a guarantee, and the losses sit in the PCS and thermal system rather than the cells [S2].
Specification checklist and what to verify before locking duration
For a 2-hour build, verify: usable kWh at the AC terminals (not nameplate DC), continuous and peak kW ratings, PCS overload capability, round-trip efficiency at the operating point, and EMS dispatch logic against the actual tariff [S2]. For a 4-hour build, add: arbitrage revenue modelling under stressed gas and overbuild scenarios, with at least 37% and 17% downside cases applied to the day-ahead line [S3].
For any duration, the capex benchmark for German projects in 2026 is €700k/MW for 2-hour and €935k/MW for 4-hour, against an unlevered IRR target of 12.2% and 13.7% respectively at 2026 COD, and the financing case typically relies on physical tolling offtake fixing 70–100% of capacity for 5–10 years to support gearing up to 85% [S3]. The decision to extend duration from 2-hour to 4-hour should track the structural shift from ancillary services (55% of 2026 revenue) to wholesale arbitrage (95% of 2030 revenue), not just the capex delta [S3].
Trackable signals over the next reporting cycle: ancillary service price clearing in Germany against the 4.5 GW FCR plus aFRR threshold, the share of new grid-connection applications that are 4-hour or longer, and the average tolling tenor in disclosed offtake deals, all of which will indicate whether the 80% 4-hour-plus share in the 2040 German fleet is being pulled forward or pushed back [S3].
Component reference pages worth checking: ready mix concrete, measurement test 2, and electronic test measurement 2.