Why BESS is so hard to price

1. Component Architecture: What You're Actually Building

The fundamental difference is that solar is a pure power asset; BESS is a power-and-energy system with three distinct layers that each have their own supply chain, pricing, and technology curve.

Dimension	Solar PV	BESS
Core energy component	Solar modules (no storage)	Battery cells (electrochemical, $/kWh)
Power conversion	Inverters ($/MW)	Inverters / PCS ($/MW), DC or AC block
Thermal management	None	Required per container (HVAC, fire suppression)
Controls/software	Plant controller (simple)	Energy management system + BMS (complex)
Physical structure	Racking, mounting	Shipping containers / AC blocks
Grid interface	Transformer, substation	Transformer, substation
Primary cost driver	Modules ($/Wp)	Cells ($/kWh) + inverter/PCS ($/kW)
Cost scales with	MW (power only)	Both MW and MWh (duration)

Table: Solar PV vs. BESS component comparison

Solar's cost structure is essentially one-dimensional: nearly everything (panels, inverters, racking, cabling) scales with the peak power rating (MW). There is no storage dimension to price. That simplicity is why solar capex can be quoted as a clean £/kW or €/kW number across projects with very different physical sizes.

BESS has two orthogonal cost axes: power (MW) determines the inverter, transformer, and grid connection; energy (MWh) determines the cells, container count, and thermal management. These don't move in sync. According to GB BESS Outlook Q4 2024, grid connection and inverter costs scale with rated power, so doubling duration from 1h to 2h increases total costs by only ~83%, not 100%. Cells get more expensive but shared infrastructure doesn't. This makes per-kWh and per-kW figures structurally incomparable across projects of different durations unless you fix one variable.

2. Who Builds It: The Supply Chain Structure

This is where the pricing complexity really compounds.

Solar: A Commodity EPC Model

Solar PV construction has converged on a standardised EPC (Engineering, Procurement, Construction) model. Panels are a traded commodity with transparent spot prices. Inverters are similarly off-the-shelf. An EPC contractor can quote a turnkey project price with reasonable confidence because they are assembling known, interchangeable components from multiple competing suppliers. The developer typically has a choice of dozens of panel manufacturers and several inverter vendors with published specifications.

Importantly, the panel is not the "system", it's just a component. Panels from JINKO, LONGi, or First Solar are interchangeable in the racking. This means competitive procurement is easy and pricing is verifiable against spot markets.

BESS: Three Fundamentally Different Procurement Paths

BESS suppliers in Great Britain: Q3 2025 update and How much does it cost to build a battery energy storage system in 2024? document three distinct roles that may or may not be performed by the same company:

a) Cell Manufacturer

Makes the electrochemical cells that store energy. Almost entirely concentrated in China (80% of GB installed capacity from Chinese manufacturers; CATL alone supplies 81% of cells installed since 2024). Cell prices are set in China, priced in RMB, and subject to currency risk, trade policy, and Chinese industrial policy. Developers cannot directly compare cell prices without knowing cell chemistry, cycle life, degradation guarantees, and warranty terms. A cheaper cell may degrade faster and destroy the revenue model.

b) BESS Supplier / Container Manufacturer

Assembles cells into containerised systems (the "rack + BMS + thermal management" layer). Companies like Tesla, Wartsila, Sungrow, CATL, BYD, Canadian Solar, and Trina all supply containerised BESS. Critically, some cell manufacturers also supply containers (CATL, BYD), while others don't (CATL cells frequently appear inside Tesla Megapacks). The container includes battery management systems, fire suppression, and thermal control, proprietary technology that varies materially across suppliers.

c) Integrator

Ties container systems together at site level: inverters (if DC-block architecture), transformers, energy management systems (EMS), grid protection, and SCADA. Integrators like Fluence, Tesla (for its own AC blocks), and independent operators like Belectric have taken this role. The integrator's EMS determines how the battery actually earns revenue (dispatch optimisation, market participation, frequency response), making this arguably the most commercially sensitive layer.

These three roles can be bundled (Tesla sells an AC block that combines cell, container, inverter, EMS in one package) or fully separated (CATL cells → third-party container → independent integrator → separate EPC civil contractor). The pricing of a BESS project changes fundamentally depending on which path is chosen.

3. Why BESS Is So Hard to Price: The Core Problems

Problem 1: You Can't Benchmark What You Can't Separate

When a developer buys a Tesla Megapack, the invoice is a single line: $/kWh for an AC block that includes cells, inverter, BMS, and EMS. When a developer procures directly from CATL and contracts an independent integrator, those are three separate contracts with three separate margin layers. The "containerised BESS cost" that Modo Energy surveys cannot always be split into cells + inverter + integration; it depends on the procurement path chosen.

Modo Energy's 2024 GB survey explicitly separated the containerised BESS cost (cells + racks + integrating electronics) from the Balance of Plant (BOP) contract, which covered private civil works but excluded inverters and transformers. This is already a simplification; some projects bundle transformers into BOP, others don't.

Problem 2: The EPC Layer Adds a Margin That Varies

For solar, EPC margin is a known, competitive overhead of roughly 5-10% on a commodity-priced job. For BESS, the EPC margin sits on top of a proprietary product whose underlying cost the EPC contractor may not fully know either. The GB BESS Outlook reports an EPC margin assumption of ~10-11% of total project costs, but in practice this varies because:

• Not all containerised BESS solutions are available to all developers (they may fail procurement processes)
• Some suppliers offer preferred pricing to large buyers or strategic partners, unavailable on the open market
• Lead times create pricing uncertainty; a container ordered today may arrive at a different price than when it was quoted

Problem 3: Software Is Not Priced Separately, But It Matters

In solar, the plant controller is a trivial cost. In BESS, the EMS and trading algorithm are central to revenue generation, yet in almost every procurement structure, they are either bundled into the integrator contract or provided under a long-term O&M/optimisation agreement with a revenue share. Two identically specified BESS systems can earn materially different revenues depending on who is running the EMS and how good their dispatch is. This doesn't appear in a capex comparison at all.

Problem 4: Warranty and Degradation Terms Are Not Standardised

Solar module degradation is slow, predictable, and publicly warranted (typically below 0.5%/year). BESS cell degradation is faster (often 2-3%/year), chemistry-dependent, cycle-dependent, and covered by warranties that vary enormously across suppliers in terms of what triggers a claim, what the remedy is, and whether the warranty is backed by local entities (important for enforcement). A £100k/MW difference in container cost may be fully offset by a superior degradation warranty, but you cannot see that in the headline capex figure.

4. The "Own Procurement" vs. "System Integrator" Split

The most commercially significant choice is whether a large developer:

Buys a fully integrated AC block (Tesla Megapack, Wartsila GridSolv, Sungrow PowerTitan, Canadian Solar PowerBlock): Factory-tested, plug-and-play, reduces on-site installation time by several weeks. Higher per-unit cost but lower execution risk, faster commissioning, and modular resilience (one inverter fails, others keep running).

Or procures a DC-block system with separate integrator: More configuration flexibility (duration can be adjusted more easily), potentially lower component cost if cells are sourced directly, but requires separate inverter contracts, independent integration engineering, and more on-site commissioning work. Execution risk is higher, and warranty responsibility may be fragmented across parties.

For very large projects (500 MW+), direct cell procurement from CATL or BYD is an option for developers with sufficient scale and balance sheet to run their own supply chain. This cuts out the container supplier margin but requires the developer to own the integration risk entirely. This approach is rare in Europe but growing in markets like Australia, where project scale justifies it.

The net result is that two projects with identical MW/MWh ratings, built simultaneously in the same country, can have capex figures that differ by 30-40%, and both figures are "correct" for their respective procurement structures. That's why BESS capex surveys (like Modo Energy's) show 68% of projects in a £400-700k/MW band, not a tight cluster around a single number.

The short version: Solar is like buying a commodity and assembling it; everyone pays roughly the same for panels. BESS is like buying a custom-engineered system where the hardware, software, integration, and warranty are all bundled differently by each supplier, making like-for-like comparison genuinely difficult, not just imprecise.