Testing Battery State of Charge / Volts Only

[Maine Sail]

I have seen the discussion many times here where folks say they monitor their batteries state of charge via a volt meter. While this can be fairly accurate when the battery has been allowed to come to a resting voltage, the reality is that when you are on the boat and cruising, attaining a 12+ hour resting voltage can be very difficult.

I wanted to do a photo/time/voltage experiment to illustrate why using a simple voltage test on a boat may not be the most accurate way of checking the resting voltage of a battery especially with short resting times.

In order to check the state of charge (SOC) on a lead acid battery the battery should have rested, meaning no input voltage/amps or output/loads, not even the stereo memory, for a considerable amount of time. If the battery is not allowed to rest, and you have recently run the engine, or had a load turned on, the SOC will not be accurate for many hours.

To conduct this illustration I used a bench charger, stop watch and a DVM. I set my camera up on the tripod and photographed the battery over 24 hours (actually I still have 12 hours to go as of this posting).

The battery, a Trojan group 24 deep cycle, had been on my bench and drained to about a 60% SOC. It then sat for a week before I took my baseline SOC reading. The resting voltage was 12.24 volts or about 60% of charge according to Trojan Battery. I then plugged in the 20 amp charger for about 130 seconds and then unplugged it.

This was only 130 seconds of charging and it affected the resting SOC for at least 11 hours and was off by about 10% even after sitting for 11+ hours.

Battery SOC at Resting Voltage and 80f (Source Trojan Battery)

100% = 12.73
90% = 12.62
80% = 12.50
70% = 12.37
60% = 12.24
50% = 12.10
40% = 11.96
30% = 11.81
20% = 11.66
10% = 11.51

Resting voltage Fri 10:04 P.M. / 12.24 Volts:

Charge Applied Fri 10:05:42 P.M. Voltage rises quickly to 13.97 volts:

By 10:05:56, or within 14 seconds, the voltage was already 14.3 volts:

I had unplugged the charge at roughly 130 seconds and the voltage was already dropping back to 13.45 volts at 10:07:15 which was 1:57 seconds from where I turned on the charger:

I took the next photo on Saturday at 2:13 A.M. a little over four hours after turning off the 130 second charge. The voltage still read 12.44 volts or roughly 76% SOC:

Unfortunately I fell asleep and did not take the next photo until Saturday morning at 8:42 A.M. when the voltage still read 12.38 volts or roughly 70% SOC. Based on my calculations the battery should still be at about 12.24-12.25 resting based on 130 seconds of charging and what went into it:

Here's the photo from the 12+ hour mark, 10:26 A.M. Saturday, still reading approx 70% SOC:

Here's the photo after 24 hours rest. The battery is back to just about where I had calculated it would be. As of this morning at roughly 34 hours it was still at 12.26 volts and seems to have finally attained a resting voltage:

If you were to run your engine before bed at say 10:00 P.M. then wake up and take an SOC reading at 8:45 A.M. you could be off by as much as 10% in the wrong direction thinking you are at a higher SOC when in reality you are as much as 10% lower. If that lower range is 60% you only have 10% bank capacity not 20% before you hit the 50% re-charge threshold.

 

[Don S/V ILLusion]

Interesting analysis. I think most folks would be content knowing nothing more than an approx state of charge and the easiest way of knowing it for wet cells is to simply use a hydrometer which can be a fairly accurate means regardless of timing issues.

 

[Maine Sail]

Interesting analysis. I think most folks would be content knowing nothing more than an approx state of charge and the easiest way of knowing it for wet cells is to simply use a hydrometer which can be a fairly accurate means regardless of timing issues.

I agree but the vast majority of boats I've been on have volt meters and the owners don't even know what a hydrometer is or how to use one.

I am merely pointing out that a quick glance at your volt meter with loads, input or insufficient resting won't tell you much about the state of charge. This is something many boaters do not understand..

I'd be curious as to how many forumites other than you, myself, Bill, Stu. Rich H. and a handful of others actually own a hydrometer or know what to do with one vs. how many own a volt meter or have one installed in their vessel??

 

[Buck420]

Well...

Used to own a hydrometer, but can't really use it now w/ AGM's.... and w/ a Link 2000, (I think) I've got all the (accurate) battery info I need ...
but why do I have a feeling you're going to tell me ... "Nope, sorry, you don't" ???

 

[Maine Sail]

Not at all

Used to own a hydrometer, but can't really use it now w/ AGM's.... and w/ a Link 2000, (I think) I've got all the (axccurate) battery info I need ...
but why do I have a feeling you're going to tell me ... "Nope, sorry, you don't" ???

Not at all. I think the battery monitors are the closest to accurate SOC measurements we have. they are tremendously useful devices if programmed correctly.

I am merely trying to point out the pit falls of assuming your volt meter (not a battery monitor) is accurate for reading an SOC.

With a volt meter only if you shut off your engine after a 20 minute run and the voltage reads 12.7 after 15 minutes you could still have a battery at 50% SOC, because it takes hours for the battery to reach resting voltage, but one may incorrectly assume, by glancing at the volts, they have a full 100% charge.

 

[Warren Milberg]

A number of owners may have so-called maintenance free batteries (as I do) which do not allow the use of a hydrometer. So, while knowing the "state of charge" as Maine Sail describes is a useful thing to know using a DVM, that may tell you little about the ability of your batteries to put out cranking amps. And amps are what start, or do not start, your engine. Further complicating the issue is the fact that a battery rested and not under load may show a state of charge of 12.6-12.8 vdc, suggesting it is "fully charged." Yet, it may not be able to muster the necessary cranking amps to start your engine. My fairly good (Fluke) DVM does a lot of tasks aboard my boat, but it is not capable of measuring battery amps. While I could add an ammeter to my panel, I've found that just starting the engine with one or the other battery, but not both, will give me a pretty good sense of how easily, or not, a battery can put out cranking amps.

 

[Ross]

Monitoring battery voltage while cranking the engine will tell you much about your state of charge. Cranking without starting and watching the battery voltage recover is also informative.

 

[Maine Sail]

Can you

Monitoring battery voltage while cranking the engine will tell you much about your state of charge. Cranking without starting and watching the battery voltage recover is also informative.

Can you please explain how one can determine SOC using this method?

 

[Ross]

Can you please explain how one can determine SOC using this method?

You won't determine SOC but you will get a good indication of the internal resistance and the reserve recovery. If the battery voltage drops substantially when cranking there is reason to believe that they are nearing the end of their life. Further if they recover slowly it is just an additional indicator of their impending demise.

 

[Maine Sail]

You won't determine SOC but you will get a good indication of the internal resistance and the reserve recovery. If the battery voltage drops substantially when cranking there is reason to believe that they are nearing the end of their life. Further if they recover slowly it is just an additional indicator of their impending demise.

Battery voltage does drop substantially when you start a motor, even with a brand new battery at 100% SOC, and is in direct correlation to the starter motors load or Kw rating and the bank you have behind the stater. This voltage comes back up quickly, from its low, but it still takes that battery a long while to reach resting voltage to determine a more accurate SOC..

If we are trying to get a sufficient life out of our batteries by not regularly dropping our banks below 50% SOC using that method only tells you very little about actual SOC for longevity.

 

[ghost]

There have been so many threads about batteries & their state of charge etc that it makes me think battery technology is too complex that joe public will not be ready to handle hybrid & electric vehicles.

Boaters tend to be more "hands on" as far as batteries & general upkeep of things on their boats but I can't see the general public getting into the science of battery technology in a big way. Most just want to get in the car, turn the key & go. I predict the failure of hybrid & electric vehicles for the public at large because they are too complex.

 

[henkmeuzelaar]

Ohm's Law may be too darn hard for most of humanity....!

gghost,

I'm afraid you've got an excellent point there!!

Having directed instrument development and test labs, populated with graduate students and postdocs trained in various engineering and chemistry disciplines, for nearly 3 decades I had to learn the hard way that quite a few of them had less trouble understanding quantummechanics than Ohm's law!

Since the inevitable result would be seeing yet another valuable prototype go up in a hail of sparks, smoke and flames, our technical staff spent an inordinate amount of time trying to teach everyone "electricitiy 101".

As we often worked with batteries to power field-portable test equipment, I even remember buying Ample Power's little book entitled "Living with Ample Power" -- or something like that -- and putting it on the shelf in the lab.

Flying Dutchman

 

[Don12364]

An excellent experiment and report Main but did the temperature coefficient of the battery influence the results?
I am aware that a 12V lead acid battery has a negative temperature coefficient of about 2.6mV/°C and if the day/night variation of the environment was 10°C this would lead to the open circuit voltage being higher in the wee small hours by .026volts and it would drop again as the following day warmed up. Did you see any of this? If the variation was only 10°C it would have only affected the last digit on your meter by three places though.
Temp coeff also matters with high rate charging and a temperature compensated charger is seen as essential.
See the Battery Reference Book for electrochemistry too difficult for me to understand:-
http://books.google.co.uk/books?id=...=X&oi=book_result&ct=result&resnum=5#PPT42,M1

 

[Maine Sail]

An excellent experiment and report Main but did the temperature coefficient of the battery influence the results?
I am aware that a 12V lead acid battery has a negative temperature coefficient of about 2.6mV/°C and if the day/night variation of the environment was 10°C this would lead to the open circuit voltage being higher in the wee small hours by .026volts and it would drop again as the following day warmed up. Did you see any of this? If the variation was only 10°C it would have only affected the last digit on your meter by three places though.
Temp coeff also matters with high rate charging and a temperature compensated charger is seen as essential.
See the Battery Reference Book for electrochemistry too difficult for me to understand:-
http://books.google.co.uk/books?id=...=X&oi=book_result&ct=result&resnum=5#PPT42,M1

Heated garage. T-stat set at 72 so there was no temperature fluctuation during the 24 hours. The 15A charger was only on for 130 seconds at a 60% SOC which was clearly not enough to change any battery temps but enough to mess with the resting voltage.

I was quite surprised by the results. It took a full 24 hours for the battery to get back to 12.26 volts. I will upload the photos tomorrow. I'll be curious to see if it drops down to 12.25 overnight as that is what I figured it would return to based on what went in during that 130 seconds.

I know batteries self discharge faster in warmer temps & slower in cold but I wonder if colder temps would slow the return to a resting voltage as well? I was really surprised it took nearly 24 hours to fully return to a resting state for an accurate resting voltage.

 

[henkmeuzelaar]

Relative SOC vs voltage profiles are nice to have.....

What a cruiser often wants to know, however, is not so much the percentage of full charge remaining as well as the total Watt-hours available under the anticipated battery load & temperature conditions.
Under normal conditions relative SOC can be readily converted into available Watt-hours (or Amp-hours, etc) using empirical graphs and charts.

Over the years we have repeatedly found ourselves in situations where battery bank performance suddenly appeared to have gone out of the window -- preferably under the most taxing conditions -- and a quick estimate of remaining absolute battery reserves was necessary in order to differentiate between the most likely culprits; i.e. insufficient input from charging systems, failed battery storage or excessive loads & shorts.

Although I don't often happen to agree with Ross; the dynamic battery condition test approach he describes would seem to be the standard test approach in all situations where one can no longer simply assume that the nice SOC/voltage graphs drawn up several months earlier are still relevant, since the battery bank itself might have become damaged.......

Even if one were able to directly determine the volume and concentration (density) of the electrolyte, thereby measuring a battery bank's chemical reserves, this does not guarantee that such chemical potential can be adequately converted into electrical energy since the plates might have suffered mechanical or physiochemical damage and/or the electrolyte might have become contaminated.

One of the easiest dynamic battery condition tests is simply to simulate a 1-or 2-second engine start attempt (kill-knob out, transmission in neutral) while reading both batttery terminal voltage and starter motor current (e.g. using a clamp-on meter) . On most vessels this start attempt can be readily switched between starter battery and house bank.

In order to properly translate/interpret the readings of such dynamic load tests one needs to have prepared a table or chart with prior dynamic test results conducted under similar conditions or under conditions where battery charge was known with reasonable confidence.

Have fun

Flying Dutchman

 

[Benny]

Battery Monitoring.

Mainesail, years ago before we had sophisticated battery monitors we used an analog voltmeter. You are very right when you indicate that a reading taken when the battery is not at rest will yield a value not corresponding to the actual state of charge. Boats back then did not have sophisticated electronics and the needs to monitor battery charge was not that critical. We obtained more information from the voltmeter by comparing readings from one day to next than by the absolute value of the reading we were getting at a particula time. There are still boats with simple electrical systems and requirements that adequately monitor their batteries with a voltmeter. On the other hand boats with a lot of electronics and high end batteries do require an appropiate monitoring tool to protect the investment. It is not useful to generalize as not all boats are equal.

 

[John Nantz]

Details, details.

There's nothing like Donalex reading these posts to look out for the little details! The devil is in the details, as they say.

Henk & Ghost - you mean, there's no law against stupidity??? Could it be that AAA towing insurance might go up as we lead into the hybrid/electric vehicle age?

Main: I assume the voltage of the VOM batteries had no effect on the readings because it was turned off between readings. Otherwise, if the VOM was sensitive to it's own battery voltage, as the VOM batteries declined in voltage it would be possible that the readings could decline too.

Gezzz, who comes up with all these things?

 

[Maine Sail]

One of the easiest dynamic battery condition tests is simply to simulate a 1-or 2-second engine start attempt (kill-knob out, transmission in neutral) while reading both batttery terminal voltage and starter motor current (e.g. using a clamp-on meter) . On most vessels this start attempt can be readily switched between starter battery and house bank.

In order to properly translate/interpret the readings of such dynamic load tests one needs to have prepared a table or chart with prior dynamic test results conducted under similar conditions or under conditions where battery charge was known with reasonable confidence.

In order to properly measure this test one would need a DC clamp meter that could read inrush current for DC. These meters are not cheap, there are not many of them commercially available.

I would assume anyone going to these measures, charting performance when the bank was known good, may already understands resting voltage.

This post is for the owner who does not or did not understand that a battery needs to rest after a charging or discharging event to get a accurate SOC volt reading. The individual who did not already understand resting voltage may not own a DVM let alone an expensive DC clamp meter that can do measure DC inrush current.

There are many owners out there who will glance at a volt meter five minutes after shutting down the engine and see 12.7 and assume they have a 100% full bank. I am merely trying to show why this would be an incorrect assumption without allowing the battery to come to a resting voltage.

This is also not a post of battery monitor vs. volt meter. The individual who has not already sprung for a battery monitor is probably not going to do so. It is not my intent to change anyone's on board electronics but rather to help them understand how to get the most out of them. Learning how more accurately tell your SOC is a good thing to know, especially for bank longevity, and it can be done without a battery monitor provided you have patience.





John,

The DVM's displayed volt readings are independent of the meters internal battery state of charge until you get a low battery warning then all bets are off. Beyond that I had installed a new battery prior to this test and the batteries in this unit last about 300+ hours of use. It was turned off betwen readings though..

 

[Bill Roosa]

I'm with Benny

I don't own a hydrometer and probably never will. Don't see the usefulness of one.
I do, however, monitor my battery voltage ALL the time and note when I use things that can MATERIALLY effect the state of charge. By materially I mean stuff that is going to start sucking more than 10% of my bank max AH out. My battery bank is 410 AH. So running the stereo (<0.5 amps) is not something that I track since it can't suck out more than 12 AH if run the eniter 24 hours. Anchor lights are close to breaking the treshold (2.5 amps) at 30 AH each night. The reefer certianly does at 84 AH minimum a day in a moderate climate. The macerator does not even come close even at 25 amp cousumption since it is only on for 4/60ths of an hour for 0.41 AH

Since I know my boat systems (been playing with them for a while now) and I just pay attention to what I'm doing I have a pretty much "gut feeling" for how my SOC is.

Kinda like asking yourself can I make it to the marina in my car with 1/2 a tank of gas. You just know how much gas it takes to do that in terms of tanks not gallons and know if you are getting low and need to take some action.

 

[Witzend]

This is an interesting lab demonstration of the delayed effect on SOC as measured by a DVM from charging a wet cell battery. Why does the battery not instantly reflect the resting voltage after the input current source is removed? My guess, and I really have never read anything on this, but have been aware of it, is:
1) The time needed for the chemical reaction to take place with ions migrating from the plates to the electrolite.
2) The potential E is low between the source and the destination so the migration is slow.

Anyone care to take a stab at explaining this phenomenon?