Glubux's Powerwall (2016)

>Glubux even began disassembling entire laptop batteries, removing individual cells and organizing them into custom racks. This task, which likely required a great deal of manual labor and technical knowledge, was key to making the system work effectively and sustainably.

This kind of thing is cool as a passion project, but it really just highlights how efficient the modern supply chain is. If you have the skills of a professional electrician, you too can spend hundreds of hours building a home battery system you could just buy for $20k, but is less reliable.

That is, in my opinion, the worst feature of this entire project. It is cool and nice and fun. But it takes a lot of time to research, acquire skills, get tools and build.

> you could just buy for $20k

I agree with a broader point but that particular price is extremely high and far from reality.

A reasonably good 18650 cell has a capacity of ~12 Wh (~3300 mAh * ~3.7 V = ~12.2 Wh). The battery mentioned in the article consists of "more than 1000" such cells. Let us assume 1200 cells. That would mean it has a capacity of ~14.4 kWh (1200 * 12).

It is possible to get a pre-assembled steel battery case on heavy-duty wheels for 16 LiFePo cells, with a modern BMS with Bluetooth and wired communication options, a touchscreen display, a circuit breaker and nice terminals for ~ $500. And it is also possible to get 16 high quality LiFePo cells with a capacity of ~300 Ah each, like EVE MB31, for significantly less than $100 each. This means that for less than ~$2000, it is possible to get all components required to assemble a fully working ~15 kWh LiFePo battery.

- That assembly would take a few hours rather than weeks.

- It will have new cells rather than used ones.

- It will be safer to use than a battery with Li-Ion cells.

- It will likely take much less space.

- It will be easy to expand.

I will point out that in 2016 when they started this project, the cost of new batteries would have been multiple times higher than it is today, so it would have been a moderately more "sensible" thing to do than it currently seems.