The previous post ended in a cheap Hollywood cliffhanger: 1 BMS was supposed to "rule all them cells", but we are going with... 2 ? Sorry, yeah... The reason is that pack modularity is a top objective for this build.
2 modular packs, so 2 BMSs. Let's just
call that an Executive Decision !
But won't that possibly be an issue for the Summer pack which, used year round, will likely age much faster ?
Absolutely. But that's okay. For 2 reasons:
- The Load Low Voltage Cutoff (LLVC) on the load side will be set higher than the BMSs' Discharge Low Voltage Cutoff
(DLVC) parameter
- Each BMS will be rated for 300+A. So if for any unexpected reason the Summer pack ever depletes down to DLVC before the other pack, it'll just go offline, waiting for a charge. Meanwhile the Winter pack will still happily supply 250A as needed (*)
Ensuring LLVC > DLVC is possible because all the 12V loads are behind a
Victron Battery Protect 220A device, and all the 110V loads are behind a
Victron Multiplus inverter. Both have a programmable LLVC setting (I think...). So
they'll stop drawing current before the BMSs reach their DLVC value.
 |
| Victron Battery Protect BP220 |
Note that this is a belt and suspenders
approach. If the BMS does a good job 100% of the time, one could
regularly deplete the battery down to DLVC, counting on the BMS to yank power, wait for charging, and happily resume operation.
But a BMS should be the last line of defense to protect the cells. Instead, I prefer to handle End Of Discharge (EOD) daily protection via graceful load shutdowns and restarts, via devices designed for that.
Brutally yanking power repeatedly, like a BMS is supposed to do, is ok for electric scooters. But the inductive spikes in a large electrical system
like a van's, could be a bit much for some everyday electronic devices (USB, laptop, camera, cell, wifi...).
Also, many of the affordable BMSs do not seem
designed to be 100% reliable over 100s of string shutdowns at full current.
Yes, it is overkill if EOD events are expected to be rare, or at low currents.
But I can't guarantee that the future owners of this van won't deplete
the system on a daily basis, with big loads. And some suppliers, like
Victron, recommend this very LLVC > DLVC strategy in some cases. So, I am going with it for additional safety &
peace of mind.
Cows on bicycles, these last 2 posts feel like a legislative body wrote them...
Yeah, sorry. It took me months to piece all this together. There is a ton of confusing, and sometimes contradictory, information on the BMS & String topics out there.
Especially
in the DIY RV & PV solar spheres, from people with very different
backgrounds, skills, contexts and objectives. Although I am an
Electrical Engineer, it still was a struggle to get a clear picture. I can only imagine how other folks might feel about this disconcerting cacophony...
So I spent time to painstakingly study the topic, and then boil it down
for you to key pieces of information, choices and safety
considerations. Still, not a fun read. But, hopefully, this will be
helpful for some ?
Anyhoo, are we there yet ? Yes ! All that to finally settle on this battery & BMS architecture:
(*)
SAFETY WARNING: there is a potentially dangerous corner case when using
BMSs in parallel and one goes off due to a depleted string. We need to
ensure it won't come back on in a situation where the other strings are
at a much higher voltage. That would result in massive currents between
the strings. That's also a major NO-NO with modular packs like here,
where a pack can be added / removed on the fly. They should never be
connected in parallel until their respective voltages are the same.
We'll cover this in a future post.
Comments
Post a Comment