Voltage or State of Charge - which is more important?

[Irish Rover]

It's a dull cloudy overcast day here today. My solar is producing almost nothing. When I looked at the voltage on my house batteries it was 12.7v and the shunt was showing 92% SOC. I plugged in the kettle to boil and the voltage dropped to 11.6 and by the time it boiled SOC had dropped to 91%. When it finished boiling the voltage immediately went back to 12.5 and slowly climbed to 12.7 in the time it took me to write this post. Is 11.6v OK for a shortish period while there's a big draw and SOC is healthy?

 

[Stemar]

Short answer - yes, assuming we're talking about lead acid batteries.

A kettle is a pretty big draw, so I'd expect the voltage to drop while you're using it. 12.7 is the fully charged voltage for a healthy battery.

AIUI, it's actually quite hard to get LA batteries to 100% SoC, certainly, it takes a long time, as the charging current drops off as the battery gets full, so 91% doesn't sound too bad.

Paul Rainbow will be along in a minute to point out the bits I got wrong

 

[PaulRainbow]

It's a dull cloudy overcast day here today. My solar is producing almost nothing. When I looked at the voltage on my house batteries it was 12.7v and the shunt was showing 92% SOC. I plugged in the kettle to boil and the voltage dropped to 11.6 and by the time it boiled SOC had dropped to 91%. When it finished boiling the voltage immediately went back to 12.5 and slowly climbed to 12.7 in the time it took me to write this post. Is 11.6v OK for a shortish period while there's a big draw and SOC is healthy?

Voltage is only truly accurate if the battery is rested and has no loads. Apply a charge and the apparent voltage rises, but that's more about the charger voltage than the actual battery voltage. Apply a load and the voltage will apparently drop, but that depends on the load, the bigger the load the bigger the drop, but it will recover, at least in part, when the load is turned off.

Think of a sponge, if the sponge is dry (flat battery) and you spray it with water it looks wet (battery charged). But stop the hose and the water soaks into the sponge, the surface isn't so wet now. Take a wet sponge that won't absorb any more water (fully charged battery), lay some blotting paper on it and let it soak some water up and the sponge look dryer, leave it for a minute and sponge looks wet again.

SOC can be accurate, but only really so if the batteries are regularly fully charged, so as to re-sync the battery monitor. The longer the gap between full charges the more the accuracy of SOC will drift.

 

[Irish Rover]

Voltage is only truly accurate if the battery is rested and has no loads. Apply a charge and the apparent voltage rises, but that's more about the charger voltage than the actual battery voltage. Apply a load and the voltage will apparently drop, but that depends on the load, the bigger the load the bigger the drop, but it will recover, at least in part, when the load is turned off.

Think of a sponge, if the sponge is dry (flat battery) and you spray it with water it looks wet (battery charged). But stop the hose and the water soaks into the sponge, the surface isn't so wet now. Take a wet sponge that won't absorb any more water (fully charged battery), lay some blotting paper on it and let it soak some water up and the sponge look dryer, leave it for a minute and sponge looks wet again.

SOC can be accurate, but only really so if the batteries are regularly fully charged, so as to re-sync the battery monitor. The longer the gap between full charges the more the accuracy of SOC will drift.

Thanks. I only fitted the shunt a few weeks ago and up to yesterday the SOC trend has been 100max every day and min ranging 92 - 94%.

 

[Dockhead]

Voltage is a great indication of SOC, in fact, I would say with lead batteries, the best indication. But you need to know how to read it.

The theoretical holy grail is "resting voltage" with batts sat in the dark for 48 hours with no loads on it.

That's fine in the lab.

In fact so long as the loads on it are small -- so maybe less then 0.05C -- and have been small for a while, you can get a very valid measure of SOC from voltage.

A load will pull the voltage down, but N.B. -- only DOWN. So if you compare your system voltage with a small load on, to an open circuit voltage table, there will be a small error -- IN THE CONSERVATIVE DIRECTION. That is, your SOC will be slightly better than the table says. What this means, is that what you read off the voltage table is the worst case.

That's plenty close enough for good management of a lead acid battery bank. When you see 24v (or 12v in a 12v system), it's time to charge.

The exception to all of this is what Paul was talking about -- surface charge. This doesn't work when there has been charging and no loads. So you need to get the surface charge off before you can evaluate the SOC.

 

[Stemar]

I only have voltmeters for my battery banks, so I work on the basis that 12.7v = fully charged. 12.0 with the fridge running, it's time to turn the fridge off until we can get a few amp/hours back in, and when it gets down to 12.0v, with no load =~50%, it's time to put some juice back in PDQ,

 

[Dockhead]

I only have voltmeters for my battery banks, so I work on the basis that 12.7v = fully charged. 12.0 with the fridge running, it's time to turn the fridge off until we can get a few amp/hours back in, and when it gets down to 12.0v, with no load =~50%, it's time to put some juice back in PDQ,

I have been through almost every type of battery monitoring during my sailing life on several boats, including Victrons, Balmars, Xantrex, Merlin . . .

At the end of all of that, after many years, I realized that the method you just described cannot practically be improved upon. For lead batteries, anyway.

 

[MikeBz]

We have had a NASA BM2 for 13+ years. Its SOC is invariably inaccurate (or should I say variably inaccurate!) but it displays the voltage and charge/discharge current in nice big digits and I’ve calibrated myself to use those numbers to tell me where the batteries are at.

 

[rotrax]

I have been through almost every type of battery monitoring during my sailing life on several boats, including Victrons, Balmars, Xantrex, Merlin . . .

At the end of all of that, after many years, I realized that the method you just described cannot practically be improved upon. For lead batteries, anyway.

Exactly what we are experiencing with our new-to-us narrowboat.

The only instrument is a simple car dashboard type voltmeter.

440AH House bank, LA, 400W solar. As long as we have enough in the bank overnight we can run two fridges, an upright 12V LEC and our Dometic 12/240V Fridge/freezer 24/7.

With sunny days recently all is good. I do suspect a lack of capacity in the bank, but every day things get better. We are on our first 'shakedown' cruise, a three week trip from Oxford to Coventry and return. Engine has two alternators, a big one for the house bank. Four or five hours engine use when moving, perhaps an hour or an hour and a half when stationary. Last night at bedtime voltmeter was almost 13V. This morning just the needle width on the good side of 12V. My old school Wilkson meter confirms these readings.

We also have a Bluetti 180 power station. It is proving to be a good buy. We almost exhausted it yesterday, air fryer and Instant pot on for some time. First Mate was in a cooking mood! It was as low as we have ever seen it, 19% only. Three hours motoring and two hours in the sun, back to 100%. The laptop I am using now, plus the router are charging from the Bluetti.

I think I will check each battery for capacity once we are back on our mooring. I suspect one or perhaps two are not up to scratch.

I will not go too technical though, I will fit a really good ampmeter and voltmeter, analogue type.

I understand and trust them!

 

[Refueler]

Voltage vs SoC .... both are suspect ways to determine a batterys condition.

The only real way to determine a batterys condition is 'Load Test' ... that is a given draw over a measured time to calculate state ...

Monitors and so on are fine when regularly calibrated and set ... but as battery health fails - many cannot keep up and it thenj becomes a lottery as to batterys real state.

There's a lot to be said for the old Breakers Yard Two Handle Battery Tester .... I've tried digital / electronic versions - but that old version was never equalled.