Emergency start trips BMS

[Pizote]

Question:

I have two 170ah renogy LiFePo4 batteries on the house side. Max continuous discharge is 125amps each. They are connected in parallel - so continuous should be about 250amps. I don’t know the momentary discharge rate.

Two lead-acid start batteries (1150cca and 950cca) that I am replacing with two Odyssey ODP-AGM31. One of them is dead-dead, boils at 15amp charge after 30min. Voltage drops from 13.4 to 11 in about 45 minutes.

Between the two I have a Cyrix li-ct - it opens/closes corretly when the voltages are right And it closes the circuit when the emergency start is enabled.

Between the Cyrix and the house bus I have a 200amp breaker. Between the house bus and the batteries I have a 400amp breaker.

The issue I have is if I try to start the engine (with the house batteries) something on the house side trips and the batteries shut down. None of the breakers are tripped and to restart the batteries, I turn them off with a switch and back on. The batteries have an internal BMS so prevent over/under charge but I’m not sure how it really works with current draw.

I would think that if it went above the 200amps, the breaker would trip - but it doesn’t.

Any ideas?

PFA I’ve changed it up a bit but it’s mostly the same.

[Bert06840]

Well. A circuit breaker or a fuse has a curve. Say you draw 500A while starting, an overload factor of two or three means it takes easily 10-20s before it even considers tripping. If it ever does, at 2x it may take indefinitely.

It’s a terrible idea to start your engine on small house batteries. They aren’t made for that purpose and you’re blowing stuff up. The batteries themselves and all the expensive stuff attached to it.

Yes, emergency start is a thing, but you CHARGE your starter lead-acid batteries with your house batteries of any chemistry, so the starter batteries can take the hit of the engine. You don’t START your engine with house batteries.

A starter motor is also a highly inductive load, meaning it dumps a huge voltage spike back into the batteries when you release the key. If it’s not the current overload, it’s is the voltage surge that’s breaking your stuff.

So unless you are stranded in the desert with a scorpion gnawing on your leg and about to die, don’t use emergency start.

[Pizote]

Thanks for the input. As you say about emergency start, agreed. But in tech moloch we don’t just build disaster recovery and hope it works. We test it to make sure it works and we understand the limitations and behaviors.

The problem - through other forums - is likely the interval BMS shutting down the batteries due to over spike in current draw. Luckily, it’s a fairly smart system and restarts it not killin the batteries.

[Bert06840]

I hear you, I work on mission critical power systems, and you could not be more correct.

Your BMS definitely reacts to excessive draw here. I’d advise against repeatedly tripping your BMS overcurrent, but even more so against cranking on your LFP batteries even without an overcurrent situation. The MOSFETs in your BMS will have some current margin, but not much voltage margin.

This random picture off the internet shows what such a case looks like. This is about driving injectors, but the principle is the same. If your BMS is made with 40V-60V MOSFETs, a common value for 12V systems, an disconnect-during-cranking pulse could easily exceed that.



If you crank off a regular lead acid battery, there are no issues because the battery will always be able to sink such spikes. A disconnected LFP (because the BMS opened up) provides no such sink.

[john61ct]

There is no problem starting even a big diesel engine from a set of small LFP cells.

The problem here is that there is inaccessible circuitry internal to that "drop in" battery that has very low ampacity.

A DIY bank could handle whatever current you decide, build in a BMS bypass, not just higher quality cells, much longer lifespan, but cheaper and much more flexible.

With drop-ins you also need to verify if they can be paralleled and if so how many. Serialing even more so.

[Bert06840]

I need no convincing of the benefits of a component-based build, but he has what he has and he will have to work with that.

Or maybe he is going to burn out the BMSes in his house batteries and then he’s exactly at that point!

However. I radically disagree with you on the BMS bypass, especially *BECAUSE* you can easily start a big diesel with some tiny packs.

What do you think happens when you connect a 3S pack to a car, fire up the engine effortlessly, right after which the alternator puts out 14.4V, and the current flows into the pack?

[john61ct]

I would never use li-ion chemistries for a House bank

encapsulated in a dedicated purpose-built jumpstarter powerpak, sure

But 4S LFP is perfectly fine for cranking engines

Alt output should be run through a DCDC for charging purposes

[Rucker]

Originally Posted by Pizote

Thanks for the input. As you say about emergency start, agreed. But in tech moloch we don’t just build disaster recovery and hope it works. We test it to make sure it works and we understand the limitations and behaviors.

The problem - through other forums - is likely the interval BMS shutting down the batteries due to over spike in current draw. Luckily, it’s a fairly smart system and restarts it not killin the batteries.

What is Tech Moloch?

[Rucker]

Originally Posted by john61ct

I would never use li-ion chemistries for a House bank

encapsulated in a dedicated purpose-built jumpstarter powerpak, sure

But 4S LFP is perfectly fine for cranking engines

Alt output should be run through a DCDC for charging purposes

If not Li-Ion what would you use? I mean, everyone and their brother uses this-has used this-for years now and swears by it for house batteries. Unless I misunderstand this has to be one of the best contrarian positions I've seen on this forum.

[Pizote]

Who’s using li-ion?

LiFePo4 is lithium iron. Different technology. LiFePo4 is more stable and less expensive - less energy dense as well, than li-ion.

The OP (me) never mentioned li-ion.

[Bert06840]

ALL secondary Lithium batteries *ARE* types of Lithium-ion.

Lithium Cobalt Oxide cells, Lithium Iron Phosphate cells; “LFP’s”, Lithium Titanate cells; things commonly called “18650s”, “Pouch cells”, “Prismatic cells”, “LiPos”, “Battle Borns”, are all a type, a form factor, a category, a brand, or some other kind of lithium-ion battery.

ALL CAN BLOW UP WHEN EXPOSED TO VOLTAGES OVER 4.3V NO MATTER HOW BRIEFLY. This is because metallization between the positive and negative terminal occurs above that voltage. That creates a short made out of lithium metal.

LFPs don’t suffer from thermal runaway. But if you short them (or expose them to higher voltages, which eventually results into the same thing), you still can get a very violent fire.

Using secondary lithium batteries without a BMS is crazy.

[john61ct]

Yes technically all Lithium chemistries are li-ion.

However, that label now more commonly refers to

those higher voltage and higher density chemistries

other than LFP and LTO.

The latter are lower voltage much longer lasting and much safer than those at nominal 3.6 - 3.7V

commonly used for use cases where the higher risk is worth getting higher energy density, as for example propulsion.

Only LFP (or less commonly LTO) are suitable for House banks used in or close to human living spaces.

[john61ct]

Originally Posted by Bert06840

Using secondary lithium batteries without a BMS is crazy.

The term secondary is ambiguous, House bank as opposed to Starter is better.

That statement may be OK for careless noobs, but overly general.

A BMS is just a collection of functionalities to protect a battery

and differ widely between various instances and price ranges. And all these functions can be implemented without looking to get them all in a single all in one circuit board.

LVC is to stop discharging before voltage gets so low as to cause damage

HVC is to stop voltage going too high while charging

Either or both can be based off the lowest/highest **cell level** voltage respectively

usually both.

Good BMS let you both see the voltage values and adjust the setpoints.

Some just open/close external contactors (relays, solenoids), most have internal MOSFETs, and thus charge more to handle high currents.

Very few actually limit current, usually just shut down like a breaker, some burn like a fuse.

Some also have temperature sensor input.

Above are "protective BMS" functions.

Balancing can be done in many ways, few BMS do it well.

Start voltage, delta for stopping, and balance current are the key ratings, and again better ones make those adjustable.

Balance chargers, dedicated active balancers and manual balancing are other alternatives.


> ALL CAN BLOW UP WHEN EXPOSED TO VOLTAGES OVER 4.3V NO MATTER HOW BRIEFLY.

None blow up. Burn fast and hot, yes a semantic difference.

LFP and LTO have completely different, lower voltage specs, and are many orders of magnitude less susceptible to thermal runaway.

Of course any large capacity battery is dangerous if shorted or abused.

Lead banks can even explode, happens all the time, with owners pushing usage past EoL.

[Rucker]

Originally Posted by Pizote

Who’s using li-ion?

LiFePo4 is lithium iron. Different technology. LiFePo4 is more stable and less expensive - less energy dense as well, than li-ion.

The OP (me) never mentioned li-ion.

Understood-this was a question to John61ct, who said he'd never use it, which resulted in a flurry of emails clarifying the different chemistries.

And I'm still curious what you meant by 'But in tech moloch we don’t just build disaster recovery and hope it works. '

[john61ct]

The original mistaken conflation of li-ion with LFP has since been edited out, but hopefully the clarification was also useful to others

[Rucker]

Originally Posted by john61ct

The original mistaken conflation of li-ion with LFP has since been edited out, but hopefully the clarification was also useful to others

Thanks John. It's a great community here, and I think almost everyone is trying, like you, to provide the best information possible.

[Bert06840]

You know what. This is very difficult post for me to write. John is a good guy, he contributes a lot to this forum and I hope he continues to do so. Additionally, I'm the non-confrontational type, and there is so much stupid bickering ongoing on the internet already that I don't want to add to it.

But this is about people's lives. I cannot in good conscience ignore the fact that some of the advice given here really has the potential to put people in searious jeopardy.

I hate people that go around yelling loudly that they know what they are talking about. And they should be listened to. And they, and only they, should be trusted. I don't want to be one of them.

But here is my dilemma. In this respect, I really do know what I'm talking about. I have spent more than a decade designing Lithium-ion battery management systems, and more specifically, on the analytics that predicts their behavior. The customers that buy the products from the company I work for build things that float and fly. Safety and reliability is important to them. Many, many millions of dollars sort of important.

Tens of thousands of lithium-ion cells of various chemistries have passed my desk. Thousands of them have been treated and mistreated in every conceivable way, with the specific and only intent to understand their behavior better.

And here is the key problem, as it relates to this specific discussion. You can do something real mean to a li-ion battery, like expose it to 5 volts for a minute. Everything will seem fine. You might not even know you overvolted it. You might say, nothing did happen. But inside the cell, a microscopic structure, also known as a dendrite, might have started to form.

If you want to know what they look like, read this paper https://www.batterypoweronline.com/n...endrites-grow/

It's a lousy paper, but it has nice pictures. Pictures always win.

Nothing seems to happen for hours. Weeks. Months. Full capacity is retained. Everything works fine. Until - for whatever reason, temperature, shock, age, stress, or no apparent reason whatsoever - an actual chain reaction starts to occur. Your cell shorts out. Heat is generated, things melt, a bus catches fire. People die.

This is not a thermal excursion event, i.e. "thermal runaway". Dendrite formation happens to most any Li-ion battery flavor, including the iron phosphate and lithium titanate ones.

Originally Posted by john61ct

The term secondary is ambiguous, House bank as opposed to Starter is better.

For literally centuries, a "Secondary Cell" has meant that the cell chemistry is designed for being recharged. Primary means "use once". Like a regular alkaline AA or a zinc-manganese lantern battery.

https://en.wikipedia.org/wiki/Primary_cell

In this context, starter battery vs. house battery is irrelevant. Using a lithium cell once (like in a hearing aid, a smoke alarm, or niche uses like Lithium thionyl chloride cells in toll tags) without a BMS is fine.

Originally Posted by john61ct

That statement may be OK for careless noobs, but overly general.

It is a little bit like saying "my uncle smoked two packs a day and he lived to eighty-five. Smoking clearly does not cause lung cancer!"

Abusing a lithium cell doesn't immediately result into a catastrophic explosion 99% of the time. But while nothing happened, it did substantially increase the chance that at some point in the future, near or far, a fault will develop that could into harm or destruction of property.

If it is about a phone... you get the Samsung Galaxy Note 7: https://www.nbcnews.com/tech/tech-ne...y-mess-n710581
If it is about a car... you get the Chevy Bolt:
https://my.chevrolet.com/how-to-supp...y/boltevrecall
If it is about your family sleeping within feet from an abused multi-kilowatt-hour pack...

[john61ct]

Absolutely nothing there I disagree with, in a quick scan anyway.

Obviously that is the correct use of Secondary

but since single-use cells are 100% irrelevant to this use case I am flabbergasted you actually meant it that way.

Most non-USians in the forums use the term as a synonym for "leisure" aka House bank, using Primary for engine or Starter batt.

So best to use "rechargeable" or House to avoid ambiguity.

The extreme danger of using 3.6-3.7V nominal LI chemistries in a House bank cannot be emphasized enough, even with a knowledgeable and diligent user.

However mentioning LFP and LTO in the same breath is alarmist scaremongering.

Yes the need for proper knowledge and diligence with safety precautions remains, but 99.9999% of the time the consequence of failure with LFP and LTO

compared to a lead bank

is simply losing your rather large investment in those cells.

[Rucker]

Good conversation, and for all the right reasons. Is there a layman's version of this?

My question is really 'What should we be much less casual about when it comes to House batteries?' That should be an interesting and helpful thread.

Thanks in advance.

[john61ct]

That question needs greater specificity in order to solicit better quality answers.

less compared to who/what?

Lead banks have very very different care factors, many in fact opposed to LFP's.

And the overall context is critical, specific use case data points including usage patterns, resources available, owner preferences etc, lead to widely varying recommendations.