PCS Dispatch Efficiency

How one PCS efficiency curve turns into different realized values depending on dispatch.

By Sergey Syrvachev — Director of Product Engineering, grid-scale BESS · 10+ GWh delivered

0.4 MVA PCS / weighted eta
realized one-way eta98.49%
one-way loss1.51%
mean load38%
probe point98.59%

PCS efficiency is not a single number — it changes with load. The realized efficiency is the curve weighted by where the PCS spends its time, so the dispatch histogram below, not the peak, sets the headline value.

Typical cycling152 kW mean / peak 98.6% near 38% load
Conversion curve and dispatch histogram0.4 MVA PCS / time weighted
97.898.098.298.498.6one-way efficiency (%)10%40 kW20%80 kW30%120 kW40%160 kW50%200 kW60%240 kW70%280 kW80%320 kW90%360 kW100%400 kW0%20%40%60%time at load - drag barspeak 98.6%steep part-load skirtmean 38%realized 98.49%50% / 98.59%
A realistic mix of low-load and mid-load operating time.realized = curve weighted by time / mean load 38%
Drag bars to set time-at-load. Drag the dot on the curve to inspect a single operating point. The headline efficiency is the curve weighted by dispatch time.
Equipment visualization - select a dispatch profile, drag histogram bars, or move the curve probe to see realized PCS efficiency.

Download the PCS Dispatch Efficiency diagram

BESS PCS efficiency curve peaking at 98.6% one-way conversion, with dispatch histogram and 98.49% realized eta
A grid-scale BESS PCS efficiency curve overlaid with a dispatch histogram, showing how a 98.6% peak eta becomes a 98.49% realized value weighted by time at load.

Free to download and reuse — including commercially — under CC BY 4.0, with credit to BESS.engineer. Licence & attribution →
Browse all BESS diagrams →

Download diagram

What it shows

Realized PCS efficiency is a single 0.4 MVA power-conversion-system efficiency curve, weighted by how much time the PCS actually spends at each load. The curve runs roughly 98.0% to 98.6% one-way, rising off a part-load skirt to a mid-load peak (about 98.6% near 38% load) and easing slightly past it. A dispatch histogram below marks time-at-load, and the realized figure is the time-weighted average of the curve. Pick a profile, drag histogram bars, or move a curve probe.

Why it matters for BESS

Datasheets quote a peak or headline PCS efficiency, but a battery almost never sits at that load. The efficiency it realizes is set by its dispatch pattern: an idle-heavy profile parks most of its hours on the low-load skirt where conversion is worst, while a full-load or sweet-spot profile keeps time near the peak. This one-way conversion loss is separate from the battery's round-trip efficiency and compounds with it on both charge and discharge.

How to read it

The green line is the efficiency curve; the amber histogram is time-at-load. The realized-eta readout is the curve weighted by that histogram, with mean load shown alongside. Move the accent probe to read the curve at any single load point, and compare it to the realized value: they only match if all time sits in one bin. Switching from Sweet spot to Idle heavy drops realized efficiency as weight shifts left onto the part-load skirt.

Frequently asked

Why is realized PCS efficiency different from the peak on the datasheet?
Because efficiency changes with load, and a battery rarely holds the load that produces peak efficiency. The realized value is the efficiency curve weighted by how much time the PCS spends at each load, so an idle-heavy dispatch that dwells on the low-load skirt realizes a lower number than the nameplate peak. Only a profile concentrated at the peak load would come close to the headline figure.
Is this the same as round-trip efficiency?
No. This is a one-way PCS conversion efficiency — the loss the power-conversion system takes in a single direction, charge or discharge. Round-trip efficiency (RTE) is roughly charge efficiency times discharge efficiency across the whole system, including the battery. The PCS one-way figure shown here is one factor inside RTE, applied on each leg, not the round trip itself.
How does dispatch profile change the realized efficiency?
Dispatch decides where the histogram weight sits on the curve. A Full-load or Sweet-spot profile keeps most hours near the mid-to-high load peak, so realized efficiency stays high. An Idle-heavy profile puts most time at very low load on the part-load skirt, where a fixed offset loss is a large fraction of a small output, dragging the weighted average down.
What does the curve probe tell me?
The probe reads the efficiency curve at one specific load point, independent of dispatch. It answers 'what is efficiency at this load' rather than 'what does the whole profile realize'. The probe and the realized-eta readout only agree when all the dispatch time sits in a single load bin; any spread of time across loads makes the weighted realized value differ from any single probe point.

References

Standards and authoritative sources this visual is built on:

  1. EN 50530: Overall efficiency of grid connected photovoltaic inverters — CENELEC, 2010
  2. IEC 61683: Photovoltaic systems — Power conditioners — Procedure for measuring efficiency — IEC, 1999
  3. Performance Test Protocol for Evaluating Inverters Used in Grid-Connected Photovoltaic Systems (origin of the CEC weighted-efficiency scheme) — Sandia National Laboratories, 2004
  4. IEC 62933-2-1: Electrical energy storage (EES) systems — Part 2-1: Unit parameters and testing methods — General specification (round-trip efficiency) — IEC, 2017

← Explore all BESS visuals