Almost every conversation about battery storage is a conversation about cells. Which chemistry, whose cells, how many cycles, what the degradation curve looks like.
Meanwhile the piece of equipment that determines how fast your project can deliver power, how it behaves when the grid wobbles, and how much energy you lose on every single cycle gets a line on the datasheet. That piece is the power conversion system, and it deserves better.
What it actually does
Batteries store direct current (DC) — a steady one-way flow. The grid runs on alternating current (AC), which reverses direction fifty times a second in India. Something has to translate between them, in both directions.
That is the PCS: the core power electronics device sitting between the battery and the network. During discharge it converts DC from the cells into grid-compatible AC. During charging it rectifies AC from the grid back into DC. Same hardware, working both ways — which is why it is called bidirectional, and why it is not simply a solar inverter.
But conversion is only half the job. A modern PCS also provides precise active and reactive power control, frequency regulation, and energy management based on real-time grid conditions and the battery’s state of charge. When a tender asks a battery to respond to a frequency event in under a second, it is asking the PCS to do it. When a grid operator wants reactive power to hold voltage, that comes from the PCS too.
The cells store the energy. The PCS decides what happens to it.
MW and MWh: the PCS sets one of them
This is the most useful practical thing to understand.
The PCS determines the power rating. How many megawatts can flow in or out at any instant is a property of the converter, not the cells. The cells determine the energy — how many megawatt-hours are available before you run out.
A 100 MWh battery behind a 25 MW PCS is a four-hour system. The same 100 MWh behind a 50 MW PCS is a two-hour system. Nothing about the cells changed; the converter did. That trade-off is the subject of our guide to two-hour versus four-hour duration, and it is set at the PCS.
Grid-following versus grid-forming
PCS units come in two behavioural types, and the distinction matters more every year.
Grid-following units need an existing, stable grid to synchronise with. They read the grid’s voltage and frequency and inject current in step with it. If the grid disappears, so do they. This describes most inverters installed anywhere in the world today, and it is perfectly adequate on a strong network.
Grid-forming units establish their own voltage and frequency reference. They do not follow the grid; they help define it. That lets them support weak grids and run isolated microgrids where there is nothing to follow. Manufacturers have been pushing hard here — Huawei launched a grid-forming PCS for utility-scale storage in June 2026 combining conversion hardware, plant-level controls and AI-based energy management on one platform.
We go deeper on why this matters for grid stability in grid-forming inverters for BESS. The short version for a buyer: grid-forming capability is what makes microgrids and islanding through an outage possible, and it usually costs more. Specify it because you need it, not because it sounds better.
Efficiency, and why silicon carbide keeps coming up
Every conversion loses something as heat. Those losses land twice per cycle — once charging, once discharging — every day for twenty years.
India’s CEA technical standards now set a minimum PCS efficiency of 95 per cent with an isolation transformer, alongside a minimum AC-to-AC round-trip efficiency of 70 per cent for the whole system. Those are floors for excluding poor equipment; see CEA technical standards for connecting a battery for the full set.
The technology moving those numbers is silicon carbide (SiC). SiC MOSFETs are increasingly replacing traditional silicon IGBTs in PCS platforms, switching faster with lower losses. Practically, that means higher efficiency, less waste heat to remove — which eases the thermal management problem — and often a physically smaller unit.
What to ask a supplier
- Guaranteed efficiency across the load range, not just at the single optimal operating point. Batteries spend a lot of time part-loaded.
- Grid-following or grid-forming, and whether grid-forming can be enabled later or requires different hardware.
- Reactive power capability, and whether it is available at zero active power — increasingly asked for in Indian grid codes.
- Ambient temperature derating. A PCS rated at 25°C that throttles at 45°C is a different product in an Indian summer.
- Service and spares in India. Power electronics fail more often than cells do, and a converter waiting on an imported board is a project not earning.
What this means for you
- If you are a developer: specify the PCS as carefully as the cells, and make efficiency a guaranteed, tested number in the contract. It is one of the few specifications that affects revenue on literally every cycle.
- If you are a C&I buyer: the honest question is whether you need to keep running when the grid fails. If yes, you are in grid-forming territory and should say so early. If no, a well-specified grid-following unit is the sensible, cheaper answer.
- If you are comparing quotes: check that PCS ratings are quoted at a realistic ambient temperature, and that the MW and MWh figures are consistent with the duration you asked for. Our ADESS product range is specified with both numbers stated plainly.
- If you are sizing a system: the savings calculator starts from your load, and our engineering team can specify the converter side with you — get in touch.
Product specifications, semiconductor technology and grid code requirements move quickly, and the standards cited change by notification. Treat this as an August 2026 snapshot and confirm current requirements and guaranteed equipment performance with your suppliers.