What Is Depth of Discharge?

Depth of discharge is the percentage of capacity you've pulled out of a battery. If you start at full and run it down to 20%, that's 80% DoD. Simple enough.

The reason anybody cares about this number is cycle life. Every battery datasheet gives you a cycle rating, but that rating means nothing without knowing the DoD it was tested at. A cell rated for 4000 cycles at 50% DoD might only give you 1200 cycles if you're running it to empty every time. I've seen projects get burned by this. Someone specs a system based on the headline cycle number without reading the fine print, and two years later the pack is toast.

Here's the thing. Shallow cycling extends life, but it also means you're paying for capacity you never use. A 100 kWh system limited to 60% DoD gives you 60 kWh per cycle. You could have bought a smaller pack and run it harder. Whether that makes sense depends on cell costs, replacement costs, site constraints, and about fifteen other variables. There's no universal right answer.

LFP cells have changed the math on this. Five years ago, running NMC to 90% DoD was asking for trouble. The cathode side falls apart. Oxygen loss, structural degradation, all the usual suspects. LFP doesn't have those failure modes at high DoD. You can push it harder without the same penalty. That's why a lot of stationary storage projects have moved to LFP even though the energy density is worse.

The anode side gets stressed at both ends of the state of charge window. Down near zero, you're pulling lithium out of graphite that really wants to hold onto it. The SEI layer cracks and reforms. Each time it reforms, it eats more lithium and the capacity drops. Up near 100%, especially on fast charge, you risk plating. Neither extreme is good for longevity.

Cathode degradation depends heavily on the chemistry. NMC111 is more stable than NMC811 at high voltage, but NMC811 gives you the energy density. The high-nickel cathodes want to release oxygen when you delithiate them past a certain point. That oxygen reacts with electrolyte and generates heat. Keep doing that and the cell ages fast.

Mechanical stress matters more than most people realize. Graphite swells when you lithiate it. Silicon anodes swell even more. Deep cycling means bigger volume changes, more stress on the electrode coatings, more chance of delamination or particle cracking. Some of the newer silicon-blend anodes have real problems with this.

Depth of Discharge

Every battery management system has min and max SoC setpoints. The BMS won't let you go outside that window. What the customer calls "0%" on their display is probably 5% or 10% actual SoC. What they call "100%" might be 95%. Manufacturers do this to build in margin.

The complication is that capacity fades over time. A cell that started at 100 Ah might be at 85 Ah three years later. If your BMS is still using the original capacity value, your DoD calculations are wrong. Good systems recalibrate periodically. Cheap systems don't.

Temperature throws another wrench in. A cell at 0°C might only deliver 60% of its rated capacity even if you try to pull 100% DoD. The lithium can't move fast enough. Some systems derate the DoD limit automatically when it's cold. Others just let you try and fail.

Depth of Discharge

For grid storage, 80% DoD is pretty standard. You get decent utilization without killing the cells. Revenue stacking applications that need multiple cycles per day might go to 70% to preserve life. Backup power that sits idle most of the time can afford 90% or higher on the rare occasions it's needed.

Residential solar storage is usually 80-90%. The Tesla Powerwall runs close to 100% usable, but they're using their own cells and their own warranty math. Third-party integrators tend to be more conservative.

EVs are all over the map. The displayed 0-100% on your dashboard doesn't correspond to actual cell limits. Manufacturers hide capacity at both ends. How much depends on the pack design and how aggressive their warranty assumptions are.

The bottom line is that DoD is a design choice, not a fixed number. You pick it based on economics, risk tolerance, and application requirements. Anyone who tells you there's one right answer for every situation hasn't done enough projects.