Storing the Energy

---

I have replaced the theories on this page with my personal experience. Two years ago I equipped my Ligier Optima with 14 45Ah ACM batteries. 14x12Vx45Ah means this battery pack could store 7.5kWh. But it can't. There is something called the Peukert effect. It basically says that when you discharge your battery slowly you get more energy from it then discharging it quickly. Namely, the batteries I use can deliver 45Ah when discharged over a period of 20 hours. The car discharges the battery within 30 minutes though. I quick glance at the datasheet gives a rated capacity of 30Ah for a 20 minute discharge. That only leaves us with 5kWh to go. Therefor, even when the batteries were new, my range was limited to about 40km.
Meanwhile, the batteries are 2 years old and I've driven 6000km on them. All thats left of the initial 40km of range is 15km. So if you consider the financial and environmental cost of replacing the pack every 6000km, running an electric car off lead acid batteries is hardly an improvement over petrol. Financially it is grossly expensive.
Luckily, new players have arrived. LiFePo4 batteries from China cost around 0.35€ per Wh as opposed to 0.12€ for lead acid. But then those a real kWh, hardly influenced by the Peukert effect.

	Lead Acid	LiFePo4
Price per Wh	0.12€	0.35€
Capacity Derating at 30min. discharge	70%	100%
Corrected price per Wh for EV use	0.17€	0.35€
Rated cycle life	500@50% DoD	5000@70% DoD
Lifetime energy throughput of 10kW pack	1750kWh@50% DoD, 30 min. discharge	35000kWh@70%DoD
Lifetime range @ 20kWh/100km	8750km	175000km
Battery cost per km	0.19€	0.02€
Battery cost for 175,000 km	34,000€	3500€

So, if you plan to drive your electric car 175,000km you will go through one set of LiFePo4 batteries but through 20 sets of lead acid batteries! Maybe there are lead batteries that are better suited for the job but still, it's an order of magnitude. For some initial experiments it maybe ok but then it is time to get rid of them.

Voltage considerations

There is many discussions on what voltage to use in an electric car. There is no definitive advantage for high voltages and none for low voltages.

	High Voltage (e.g. 500V)	Low Voltage (e.g. 96V)
Number of series cells	156	29
Current at 20kW	40A	208A
Minimal diameter DIN57100/523	4mm²	70mm²
Inverter switching losses	204W	144W
Inverter conduction losses	97W	636W
Inverter total losses	302W	780W

Some say that the inverter losses will remain unchanged no matter which voltage you use. According to the Semisel simulation that is not correct. The conduction losses increase linearly with current whereas the switching losses remain almost unchanged. It is like that because the switching losses are proportional to the output power which is 20kW in both cases.
The downside of a high voltage system is the large number of cells needed to provide the high tension. Each of them can break individually, so the error probability increases. Each of them (might) need individual voltage monitoring so cost increases.
Also, everything needs to be "high voltage". The cable insulation, spark gaps, circuit breakers.
The main reason for me to go high voltage is because I want to use an industrial motor "as is". Those are commodity items and thus really cheap.

Monitoring considerations

Two extremes exist when it comes to taking care of the battery pack: don't monitor anything at all (BMS-less) up to monitor and control everything (Individual cell voltage monitoring and charge balancing).
I don't like any of those extremes. The first one feels too unsecure to work in the real world, the latter seems too expensive relatively to the battery pack.
So theres the idea of block monitoring. The theory is, that if you split a healthy battery pack in half then both halfes should have the same voltage. If they differ that indicates a problem, charging or discharging must stop and the pack must be inspected manually. As pack voltage increases it is probably a good idea to further split the pack. This makes it easier to locate the bad cell.