BMV712 Battery Monitor: Capacity Testing & Peukert Setting

The default Peukert setting is 1.25, which is said to be appropriate for flooded lead acid batteries like our 150AH Trojan 1275, but it should be lower for other battery types or if specified by the manufacturer.  My basic understanding of this value is that it is a mathematical term used in an equation which defines how the available Amp-Hour capacity of a battery changes when it is discharged at various rates (currents). Since the standard rating is for a fairly high current that will discharge the battery in 20 hours, I wanted to know how battery capacity changes at the lower usage rates we might use when camping off-grid or even just sitting there idle (and how that affects the monitor's status report).  

By the way, one article I enjoyed reading on this topic, both for its nerd humor & practical information is found here:

https://www.batterystuff.com/kb/tools/peukert-s-law-a-nerds-attempt-to-explain-battery-capacity.html

The first issue I ran into was that even though our battery is made by a respected company like Trojan, whose specs provide more capacity ratings than most (see green box above with 5, 10, 20 & 100-hour ratings), they do not provide a Peukert value to use in the Victron monitor settings.  I suspect this is a common problem for BMV users who probably mostly just accept the default as I had done.

However, at the above referenced site, I found the calculator shown below which allows you to plug in two sets of manufacturer battery ratings to calculate not only a Peukert value, but how long your battery will last at whatever discharge current you choose..

I'm not sure exactly why two sets of ratings are needed, but guessing it's the same reason you need two points to determine a line. In any case, this was interesting, first because our monitor has never shown anything but 10 days for the “Time remaining” status, presumably because it was not designed to show anything greater than the default 10d and we had never dischaged the battery below around 90% before.  Our solar panel always got us recharged, so the battery never got low enough to drop below that.

As shown above,  I tried three pairs of ratings in the online calculator, covering the range of usage rates tested by Trojan, and to my surprise they yielded 3 different Peukert values, so I'm not quite sure what to make of that.  Moreover, the predicted time remaining was opposite from what I expected, giving a longer battery life at a faster usage rating. I can only guess that it is because my actual usage rate input was very low, while the ratings are given for much higher currents.  My current was only about a quarter of an amp, which is just the background electronics running.

So, at this point I was wondering what value to use in my monitor settings, or should it be changed to fit a particular situation?  There are six combinations of 4 ratings, if used two at a time, so I ran them all through the calculator..

I got a variety of Peukert values ranging from 1.07 to 1.19, mean=1.15, but none were close to the often quoted default of 1.25. I decided to do some further testing comparing the 1.12 value derived from using the highest & lowest ratings at 5 & 100 hours.  My plan was to let the trailer sit without shore power, fully charged battery connected, and only background electronics running at about 0.25A (250mA); then wait and see how long it takes to discharge the battery until I see less than ten days left on the monitor or the voltage gets fairly low.

Above is the monitor status after 11 days of discharging the battery.  I tried changing my Peukert setting to see the effects at right. You can see how going from 1.25 to 1.12 changed both the percent state of charge (80% -> 74% @ P=1.12) and the “Time remaining” changed from 10d to 8d.  This was the first time I ever saw it drop below 10d, but is it accurate? To try & answer this question, I thought I would try manually calculating the time remaining using the Peukert equation below (from wikipedia). I used a scientific calculator (RealCalc app) with a Y^x function to resolve the weird exponents like 30 to the 0.12 power equals 1.5.

I got a revised amp-hour capacity of 225 at 250mA rate of usage with the Peukert value of 1.12 and using the 20hr rating as a starting point.  This resulted in a calculated time of 19 days to reach the 50% level, and Lo & behold - this matched the monitor status report which showed 8 days remaining after 11 days!  So, the theoretical prediction seems to match my test results at this Peukert value. At 12 days (below), the reported 7 days remaining also totals 19 days at 50% discharge.

And “On the 12th day of discharge, my monitor gave to me..  7 days remaining, and a low voltage alert warning.” 🎵

I had forgotten that I set a low voltage alarm for 12.2V, so this verified that it worked. SOC was still showing 71 percent.  It would have been about 77% if I had left the Peukert setting at 1.25, but I plan to leave it at 1.12 for a more conservative lower percentage SOC indicator.

When I went to hook up the smart battery charger, I also found I had set the monitor for both audible & visual alerts on the display as shown in this video..  The audio alert is a very high pitched beeping - not to be confused with the louder birds singing in the background.

https://drive.google.com/file/d/1csMLG6wFzcXl11X-EtqdZl2XbbgdSoTd/view?usp=drivesdk

For what it's worth, I plotted this linear regression line for the manufacturer's capacity rating test values (AH vs Current).  It shows a strong 98% correlation, despite having only 4 points. I'm still curious whether there might be any loss of linearity at points outside this range, and I wish they would have rated the battery at, say, half an amp or 250mA.

After the battery was recharged, the monitor status display showed the curious behavior in this video..

https://drive.google.com/file/d/1uL2qBggXSZFAVQzF4-jeTqyJWGWyuTMa/view?usp=drivesdk

I assume the flashing values in the display is due to a pulsating charge from the NOCO Genius 7200 charger.  I found that with the T@B plugged back in to house power & the battery switch off, I could plug the charger into the 120 outlet above the sink & run the cord out through the locked front window & into the tub where I have a charger pigtail available for a full multistage charge cycle.

In any case, whether the Peukert value adjustment I've made is valid or not, I have at least learned a bit more about how it affects the monitor status report.  And now I know that the projected time remaining does, indeed, go below 10 days, and that if our 150AH battery is left to discharge from background electronics, it should reach the 50% mark in approximately 3 weeks. (Not the 10 days I usually see on the status report - consider that a minimum, but only at the current usage rate).

References:

https://www.batterystuff.com/kb/tools/calculator-for-load-specific-run-time.html

https://www.batterystuff.com/kb/tools/peukert-s-law-a-nerds-attempt-to-explain-battery-capacity.html