top of page

Apr 27, 2026

Round-Trip Efficiency Is Not a Spec Sheet Number - It's a System Behavior Under Load

When we started looking at the data from ERCOT more closely, we couldn't help but notice that the AC RTE across the fleet is in the low-80s or high 70s, underscoring that RTE isn't a fixed property - it's an operating point. Here's what actually determines it:


Cell/Module/Pack:

  • I²R losses scale quadratically with current — high C-rate dispatch is inherently less efficient

  • Internal resistance rises with decreasing temperature and SOH degradation

  • Efficiency varies non-monotonically across SOC; mid-SOC operation generally minimizes losses


Power Conversion System (PCS):

  • Inverter efficiency is load-dependent — partial load (frequency regulation) can drop well below 90%; high load (energy arbitrage) approaches 97–98%

  • Switching losses scale linearly with power; conduction losses scale quadratically — distinct mechanisms, distinct mitigation strategies

  • Reactive power dispatch increases apparent power through the PCS without contributing to metered real energy output — a direct RTE penaltyFixed standby draw amortizes poorly over short or infrequent cycles


Thermal Management:

  • HVAC auxiliary load is a direct RTE deduction, highly climate- and architecture-dependent, and routinely underestimated in project models

  • Liquid cooling typically carries a lower parasitic load than air-cooled equivalents while providing tighter thermal control


Balance of Plant:

  • Transformer no-load (core) losses are present even at zero throughput — continuous and unavoidable

  • Conductor losses, site auxiliaries (BMS, EMS, SCADA, fire suppression) add a persistent baseline draw often excluded from headline RTE figures


Dispatch Profile:

  • RTE is path-dependent: same energy, different C-rate profiles → different losses

  • Low average utilization (peakers, ancillary services) amplifies the relative weight of standby and self-discharge losses

  • Cell-terminal, DC-meter, and AC-meter RTE can differ materially on identical hardware. This single variable explains most vendor datasheet discrepancies.


A quoted AC RTE without a defined C-rate, SOC window, ambient temperature, dispatch profile, and metering boundary is a marketing number.


What assumptions do you see most often buried in BESS efficiency specs? Email us at hello@camelotenergygroup.com for any questions!


Raafe Khan

bottom of page