5 MW Industrial Battery Storage: Pre-Design Survey Checklist

Measure the Site and Existing Infrastructure

Start with the physical boundaries. For a 5 MW system, you typically need 200–500 m² depending on container vs cabinet layout. Check access for delivery vehicles – a 40 ft container is 12 m long and weighs around 30 t when loaded with batteries and cooling.

Measure the distance to the medium-voltage (MV) substation. At 10 kV or 35 kV, a 5 MW unit draws roughly 290 A or 83 A per phase respectively. Cable losses and voltage drop over long runs are significant – at 0.4 kV this will not work, the current would be ~7200 A per phase, impossible for a practical cable.

Record the available short-circuit power at the point of common coupling (PCC). It affects the transformer impedance and protection settings. Also note the ambient temperature range – batteries derate above 35°C and below 0°C.

Pull Grid Connection Documents and Verify Constraints

Request the technical conditions (TU) from the local DSO or transmission operator. They specify the allowed power flow, reactive power capability, and protection requirements. For FCR/aFRR, the response time is typically under 1 second – the BMS and inverter must be capable of that.

Check the voltage class: 5 MW systems are almost always connected at 10 kV or 35 kV via a step-up transformer. Verify the grid code for harmonic limits, voltage flicker, and frequency response. If you plan to provide reactive power, the inverter size and transformer taps must be coordinated.

Obtain the load flow model of the feeder. This tells you if the line can handle the additional power without overloading. Also check the fault level – if it is low, the transformer inrush current might cause voltage dips that violate the grid code.

Define Power vs Capacity and Choose C-Rate

Power (MW) is the rate of energy transfer; capacity (MWh) is the stored energy. The C-rate is the ratio of power to capacity. For a 5 MW system with 10 MWh capacity, the C-rate is 0.5C – meaning it can discharge fully in 2 hours. For 20 MWh, it is 0.25C – 4 hours.

Your application dictates the C-rate. FCR/aFRR often requires high power for short bursts – 0.5C to 1C. Virtual power plants may need longer duration – 0.25C to 0.5C. Choosing a lower C-rate (larger capacity) increases CAPEX per kW but may be necessary for revenue stacking.

Battery cells degrade faster at higher C-rates. A 1C system might lose 20% capacity after 5000 cycles, while a 0.5C system might last 8000 cycles. The warranty typically specifies a throughput limit – e.g., 80% of rated energy throughput over the contract term.

Assess Footprint and Siting Options

For 5 MW, two main configurations exist: containerized (20 ft or 40 ft) and outdoor cabinets. A 40 ft container holds about 2.5–3 MWh of LiFePO4 with liquid cooling – so a 10 MWh system needs 3–4 containers plus a transformer and switchgear. Cabinets are smaller but require more civil works and shading.

Indoor rooms are possible but rare at this scale due to ventilation and fire suppression costs. Most utility-scale sites are outdoor. Consider the fire code – battery rooms need gas suppression and smoke detection, which adds cost.

Siting must account for blast zones (if required by local regulations), drainage, and soil bearing capacity. A 40 ft container filled with batteries weighs ~30 t – the ground must be level and compacted. Also check for overhead lines – you need clearance for cranes during installation.

Estimate CAPEX and What Moves It

For a 5 MW / 10–20 MWh system, the CAPEX range is honest: $350–600 per kWh, depending on configuration and market. That puts a 10 MWh system at $3.5–6 million. What moves it? The battery cell price (currently volatile), the C-rate (higher C-rate costs more per kWh), and the balance of plant (transformer, switchgear, EMS).

Civil works, grid connection fees, and engineering costs add 10–20%. If you need a high-voltage transformer (35 kV), that adds ~$100k–200k. The EMS and SCADA add ~5%.

Without a load profile and revenue model, that figure is rough. A 2-hour system for frequency regulation will cost more per MWh than a 4-hour system for peak shaving. Be wary of quotes that ignore site-specific costs – they are not realistic.